Triple Negative Advanced Breast CancerSystemic Therapy
Giuseppe Curigliano MD PhDUniversity of Milano and European Institute of Oncology
Outline
bull Targeting TNBC by subtypesbull Immunotherapy ldquoHype or Hoperdquobull New antibody drug coniugates
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Metastatic triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
bull Triple negative breast cancer and BRCA-mutationsndash Clinical behaviorndash Genomic instability
Stephens et al Nature 2009 vol 462 (7276) pp 1005
Basal like 1 TNBC
Specific tumor cell killing
HR re
pair
Few normal tissue effectsNormal tissue cells
DNA repair
DNA repair
Base excision DNA repair
Homologous recombination(HR) repair
BRCA1BRCA2 deficientTumor cells
HR re
pair
PARP inhibitor
Base excision DNA repair
PARP inhibitor
Base excision DNA repair
HR re
pair
Tutt and Ashworth Cold Spring Harb Symp Quant Biol 200570139ndash148 McCabe N et al Cancer Res 2006668109ndash8115
The principle of synthetic lethal tumour targeting
Tutt et al The Lancet 2010 376(9737)235-44
To assess the efficacy and tolerability of oral olaparib in BRCA1 BRCA2mutation carriers with breast cancerProof-of-concept phase II study single-arm sequential cohort design
Confirmed BRCA1 or BRCA2 mutationAdvanced refractory breast cancer
(stage IIIBIIICIV) after failure of ge1 prior chemotherapy for advanced disease
Olaparib 400 mg po bid (MTD)28-day cycles n = 27
Olaparib 100 mg po bid28-day cycles n = 27
Cohort 2Cohort 1 (enrolled first)
400 mg BD 100 mg BD
Efficacy in BRCA12 breast cancer
ORR 41 ORR 22
Tutt et al The Lancet 2010 376(9737)235-44
Single agent olaparib in ovarian cancer
Moore K et al October 21 2018 DOI 101056NEJMoa1810858
OlympiAD
bull Tablet formulation (2 tablets twice daily)bull M=metastatic breast cancer HER2=human epidermal growth factor 2 TNBC=triple negative breast cancer TPC=treatment of physicianrsquos choice OS=overall survival PFS=progression-free survival PFS2=progression-free survival 2 ORR=objective response rate HRQoL=health-related quality of life FSI=first subject in RECiST= Response Evaluation Criteria in Solid Tumors ER=oestrogen receptor PR=progresterone receptor ECOG PS= Eastern Cooperative Oncology Group Performance Status gBRCAm=germline BRCA mutation po=oralbull 1 httpsclinicaltrialsgovct2showNCT02000622 2 Robson et al Poster OT1-1-04 San Antonio Breast Cancer Symposium 2014 3 AZ data on file (2017) 4 Robson et al N Engl J Med 2017 377523-533
Olaparib300mgpo bid
Treatment of Physicianrsquos
Choice (TPC)Capecitabine eribuline and vinorellbine
bull gBRCAm mBC
bull TNBC or HER2-negative ERPR positive
bull le2 prior chemotherapy lines for mBC
bull Previous treatment must include anthracycline and taxane
bull Hormone receptor positive (HR+) disease progressed on ge1 endocrine therapy or not suitable
bull If patients have received platinum therapy there should be
bull No evidence of progression during treatment in the advanced setting
bull At least 12 months since (neo)adjuvant treatment and randomisation
bull ECOG PS 0-1
bull At least one lesion that can be assessed by RECIST v11
Randomise 21
N=3024
Stratification by2
bull Prior chemotherapy regimens for metastatic breast cancer
bull Hormonal receptor (HR) status
bull Prior platinum therapy
Primary endpointbull PFS (RECIST 11
Independent Review)
Secondary endpointsbull OSbull PFS2bull ORRbull PFS PFS2 and OS
based on Myriad gBRCAm status
bull HRQoL (EORTC-QLQ-C30)
bull Safety and tolerability
FSI May 20143
Global Study in 19 countries and approximately 141 sites1
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Olaparib n=205n ()
TPCn=97n ()
Totaln=302n ()
Median age (min max) 44 (22 76) 45 (24 68) 44 (22 76)
Male 5 (24) 2 (21) 7 (23)
ECOG PS0 148 (722) 62 (639) 210 (695)
1 57 (278) 35 (361) 92 (304)
Race
White 134 (654) 63 (649) 202 (669)
Asian 66 (322) 28 (289) 94 (311)
Other 5 (24) 6 (62) 11 (36)
Olaparib n=205n ()
TPCn=97n()
Totaln=302n ()
Received previous chemotherapy formBC
Yes 146 (712) 69 (711) 215 (712)
No 59 (288) 28 (288) 87 (288)
Hormonal receptorstatus
HR+ 103 (502) 49 (505) 152 (503)
TNBC 102 (498) 48 (495) 150 (497)
Received prior platinum therapy for breast cancer
Yes 60 (293) 26 (268) 86 (284)
No 145 (71) 71 (73) 216 (715)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Number of metastatic sites
1 46 (224) 25 (258) 71 (235)
2 or more 159 (776) 72 (742) 231 (765)
Sites of metastatic lesions
Bone amp locomotor only 16 (78) 6 (62) 22 (73)
CNS 17 (83) 8 (82) 25 (83)
BRCA mutation status
BRCA1 117 (571) 51 (526) 168 (556)
BRCA2 84 (410) 46 (474) 130 (430)
Both 4 (19) 0 4 (13)
De novo metastatic BC 26 (127) 12 (124) 38 (126)
Progressive disease at randomisation 159 (776) 73 (753) 232 (768)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Received prior endocrine therapydagger
Adjuvantneoadjuvant 71 (689) 36 (735) 107 (704)
Metastatic 66 (641) 28 (571) 94 (618)
Total 97 (942) 45 (918) 142 (934)
Lines of previous cytotoxic chemotherapy for metastatic breast cancer
0 68 (332) 31 (320) 88 21-7(
1 80 (390) 42 (433) 011 3 -3(
2 57 (278) 24 (247) 81 (268)
Received prior platinum therapy for BCdagger
Metastatic 43 (210) 14 (144) 46 07-8(
Adjuvantneoadjuvant 15 (73) 7 (72) 11 6-2(
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205 n ()
TPCn=97 n ()
Totaln=302 n ()
Full analysis set (randomised ITT) 205 (100) 97 (100) 302 (100)
Patients who received study treatment 205 (100) 91 (938) 296 (980)
Number who received olaparib 205 (100) - 205 (679)
Number who received capecitabine - 41 (423) 41 (136)
Number who received eribulin - 34 (351) 34 (113)
Number who received vinorelbine - 16 (165) 16 ( 53)
Safety
Robson et al N Engl J Med 2017 377523-533
21
17
18
18
12
21
7
15
22
50
23
15
26
35
1
9
11
14
16
16
17
20
21
27
29
30
40
58
0 25 755075 50 25Adverse events ()
NauseaAnemiaVomitingFatigueNeutropeniaDiarrheaHeadacheCough
Decreased appetitePyrexiaIncreased ALTIncreased ASTHand-foot syndrome
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Safety
Robson et al N Engl J Med 2017 377523-533
2
2
3
0
1
3
1
10
26
4
0
1
1
2
2
2
3
3
9
16Anemia
Fatigue
Neutropenia
Increased AST
Hand-foot syndrome
Dyspnea
Decreased platelet count
Leukopenia
0 25 755075 50 25
Headache
Adverse events ()
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Progression free survival
Robson et al N Engl J Med 2017 377523-533
17783
15946
10
09
08
07
06
05
04
03
02
01
00
Prob
abili
ty o
f pro
gres
sion
-free
surv
ival
Time from randomisation (months)
14121086420 16 18 20 22 24 26 28
234
408
6911
9421
10725
15444
20597
214
111
41
31
21
10
00
214
367
6111
7313
10024
12929
20188
111
111
31
21
10
10
OlaparibTPC
Number of patientrsquos at risk Median PFS was improved by 69 with olaparib treatment compared to standard of care
chemotherapy
Olaparib TPC
n 205 97
Events () 163 (795)
71(732)
Median (m) 70 42
HR = 058 95 CI (043 080)
p=00009
PFS free at 6m()
541 329
PFS free at 12m()
259 150
Olaparib 300 mg bd (N=205)TPC (N=97)
Response rate
Robson et al N Engl J Med 2017 377523-533
Olaparib TPC
Response Evaluable Population n 167 66
ORR n () 100 (599) 19 (288)
Complete Response n () 15 (90) 1 (15)
Partial Response n () 85 (510) 18 (273)
Median Duration of Response months (95CI) 64 (29-97) 71 (32-122)
Median Time to Onset of Responsedays 47 45
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Outline
bull Targeting TNBC by subtypesbull Immunotherapy ldquoHype or Hoperdquobull New antibody drug coniugates
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Metastatic triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
bull Triple negative breast cancer and BRCA-mutationsndash Clinical behaviorndash Genomic instability
Stephens et al Nature 2009 vol 462 (7276) pp 1005
Basal like 1 TNBC
Specific tumor cell killing
HR re
pair
Few normal tissue effectsNormal tissue cells
DNA repair
DNA repair
Base excision DNA repair
Homologous recombination(HR) repair
BRCA1BRCA2 deficientTumor cells
HR re
pair
PARP inhibitor
Base excision DNA repair
PARP inhibitor
Base excision DNA repair
HR re
pair
Tutt and Ashworth Cold Spring Harb Symp Quant Biol 200570139ndash148 McCabe N et al Cancer Res 2006668109ndash8115
The principle of synthetic lethal tumour targeting
Tutt et al The Lancet 2010 376(9737)235-44
To assess the efficacy and tolerability of oral olaparib in BRCA1 BRCA2mutation carriers with breast cancerProof-of-concept phase II study single-arm sequential cohort design
Confirmed BRCA1 or BRCA2 mutationAdvanced refractory breast cancer
(stage IIIBIIICIV) after failure of ge1 prior chemotherapy for advanced disease
Olaparib 400 mg po bid (MTD)28-day cycles n = 27
Olaparib 100 mg po bid28-day cycles n = 27
Cohort 2Cohort 1 (enrolled first)
400 mg BD 100 mg BD
Efficacy in BRCA12 breast cancer
ORR 41 ORR 22
Tutt et al The Lancet 2010 376(9737)235-44
Single agent olaparib in ovarian cancer
Moore K et al October 21 2018 DOI 101056NEJMoa1810858
OlympiAD
bull Tablet formulation (2 tablets twice daily)bull M=metastatic breast cancer HER2=human epidermal growth factor 2 TNBC=triple negative breast cancer TPC=treatment of physicianrsquos choice OS=overall survival PFS=progression-free survival PFS2=progression-free survival 2 ORR=objective response rate HRQoL=health-related quality of life FSI=first subject in RECiST= Response Evaluation Criteria in Solid Tumors ER=oestrogen receptor PR=progresterone receptor ECOG PS= Eastern Cooperative Oncology Group Performance Status gBRCAm=germline BRCA mutation po=oralbull 1 httpsclinicaltrialsgovct2showNCT02000622 2 Robson et al Poster OT1-1-04 San Antonio Breast Cancer Symposium 2014 3 AZ data on file (2017) 4 Robson et al N Engl J Med 2017 377523-533
Olaparib300mgpo bid
Treatment of Physicianrsquos
Choice (TPC)Capecitabine eribuline and vinorellbine
bull gBRCAm mBC
bull TNBC or HER2-negative ERPR positive
bull le2 prior chemotherapy lines for mBC
bull Previous treatment must include anthracycline and taxane
bull Hormone receptor positive (HR+) disease progressed on ge1 endocrine therapy or not suitable
bull If patients have received platinum therapy there should be
bull No evidence of progression during treatment in the advanced setting
bull At least 12 months since (neo)adjuvant treatment and randomisation
bull ECOG PS 0-1
bull At least one lesion that can be assessed by RECIST v11
Randomise 21
N=3024
Stratification by2
bull Prior chemotherapy regimens for metastatic breast cancer
bull Hormonal receptor (HR) status
bull Prior platinum therapy
Primary endpointbull PFS (RECIST 11
Independent Review)
Secondary endpointsbull OSbull PFS2bull ORRbull PFS PFS2 and OS
based on Myriad gBRCAm status
bull HRQoL (EORTC-QLQ-C30)
bull Safety and tolerability
FSI May 20143
Global Study in 19 countries and approximately 141 sites1
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Olaparib n=205n ()
TPCn=97n ()
Totaln=302n ()
Median age (min max) 44 (22 76) 45 (24 68) 44 (22 76)
Male 5 (24) 2 (21) 7 (23)
ECOG PS0 148 (722) 62 (639) 210 (695)
1 57 (278) 35 (361) 92 (304)
Race
White 134 (654) 63 (649) 202 (669)
Asian 66 (322) 28 (289) 94 (311)
Other 5 (24) 6 (62) 11 (36)
Olaparib n=205n ()
TPCn=97n()
Totaln=302n ()
Received previous chemotherapy formBC
Yes 146 (712) 69 (711) 215 (712)
No 59 (288) 28 (288) 87 (288)
Hormonal receptorstatus
HR+ 103 (502) 49 (505) 152 (503)
TNBC 102 (498) 48 (495) 150 (497)
Received prior platinum therapy for breast cancer
Yes 60 (293) 26 (268) 86 (284)
No 145 (71) 71 (73) 216 (715)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Number of metastatic sites
1 46 (224) 25 (258) 71 (235)
2 or more 159 (776) 72 (742) 231 (765)
Sites of metastatic lesions
Bone amp locomotor only 16 (78) 6 (62) 22 (73)
CNS 17 (83) 8 (82) 25 (83)
BRCA mutation status
BRCA1 117 (571) 51 (526) 168 (556)
BRCA2 84 (410) 46 (474) 130 (430)
Both 4 (19) 0 4 (13)
De novo metastatic BC 26 (127) 12 (124) 38 (126)
Progressive disease at randomisation 159 (776) 73 (753) 232 (768)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Received prior endocrine therapydagger
Adjuvantneoadjuvant 71 (689) 36 (735) 107 (704)
Metastatic 66 (641) 28 (571) 94 (618)
Total 97 (942) 45 (918) 142 (934)
Lines of previous cytotoxic chemotherapy for metastatic breast cancer
0 68 (332) 31 (320) 88 21-7(
1 80 (390) 42 (433) 011 3 -3(
2 57 (278) 24 (247) 81 (268)
Received prior platinum therapy for BCdagger
Metastatic 43 (210) 14 (144) 46 07-8(
Adjuvantneoadjuvant 15 (73) 7 (72) 11 6-2(
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205 n ()
TPCn=97 n ()
Totaln=302 n ()
Full analysis set (randomised ITT) 205 (100) 97 (100) 302 (100)
Patients who received study treatment 205 (100) 91 (938) 296 (980)
Number who received olaparib 205 (100) - 205 (679)
Number who received capecitabine - 41 (423) 41 (136)
Number who received eribulin - 34 (351) 34 (113)
Number who received vinorelbine - 16 (165) 16 ( 53)
Safety
Robson et al N Engl J Med 2017 377523-533
21
17
18
18
12
21
7
15
22
50
23
15
26
35
1
9
11
14
16
16
17
20
21
27
29
30
40
58
0 25 755075 50 25Adverse events ()
NauseaAnemiaVomitingFatigueNeutropeniaDiarrheaHeadacheCough
Decreased appetitePyrexiaIncreased ALTIncreased ASTHand-foot syndrome
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Safety
Robson et al N Engl J Med 2017 377523-533
2
2
3
0
1
3
1
10
26
4
0
1
1
2
2
2
3
3
9
16Anemia
Fatigue
Neutropenia
Increased AST
Hand-foot syndrome
Dyspnea
Decreased platelet count
Leukopenia
0 25 755075 50 25
Headache
Adverse events ()
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Progression free survival
Robson et al N Engl J Med 2017 377523-533
17783
15946
10
09
08
07
06
05
04
03
02
01
00
Prob
abili
ty o
f pro
gres
sion
-free
surv
ival
Time from randomisation (months)
14121086420 16 18 20 22 24 26 28
234
408
6911
9421
10725
15444
20597
214
111
41
31
21
10
00
214
367
6111
7313
10024
12929
20188
111
111
31
21
10
10
OlaparibTPC
Number of patientrsquos at risk Median PFS was improved by 69 with olaparib treatment compared to standard of care
chemotherapy
Olaparib TPC
n 205 97
Events () 163 (795)
71(732)
Median (m) 70 42
HR = 058 95 CI (043 080)
p=00009
PFS free at 6m()
541 329
PFS free at 12m()
259 150
Olaparib 300 mg bd (N=205)TPC (N=97)
Response rate
Robson et al N Engl J Med 2017 377523-533
Olaparib TPC
Response Evaluable Population n 167 66
ORR n () 100 (599) 19 (288)
Complete Response n () 15 (90) 1 (15)
Partial Response n () 85 (510) 18 (273)
Median Duration of Response months (95CI) 64 (29-97) 71 (32-122)
Median Time to Onset of Responsedays 47 45
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Metastatic triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
bull Triple negative breast cancer and BRCA-mutationsndash Clinical behaviorndash Genomic instability
Stephens et al Nature 2009 vol 462 (7276) pp 1005
Basal like 1 TNBC
Specific tumor cell killing
HR re
pair
Few normal tissue effectsNormal tissue cells
DNA repair
DNA repair
Base excision DNA repair
Homologous recombination(HR) repair
BRCA1BRCA2 deficientTumor cells
HR re
pair
PARP inhibitor
Base excision DNA repair
PARP inhibitor
Base excision DNA repair
HR re
pair
Tutt and Ashworth Cold Spring Harb Symp Quant Biol 200570139ndash148 McCabe N et al Cancer Res 2006668109ndash8115
The principle of synthetic lethal tumour targeting
Tutt et al The Lancet 2010 376(9737)235-44
To assess the efficacy and tolerability of oral olaparib in BRCA1 BRCA2mutation carriers with breast cancerProof-of-concept phase II study single-arm sequential cohort design
Confirmed BRCA1 or BRCA2 mutationAdvanced refractory breast cancer
(stage IIIBIIICIV) after failure of ge1 prior chemotherapy for advanced disease
Olaparib 400 mg po bid (MTD)28-day cycles n = 27
Olaparib 100 mg po bid28-day cycles n = 27
Cohort 2Cohort 1 (enrolled first)
400 mg BD 100 mg BD
Efficacy in BRCA12 breast cancer
ORR 41 ORR 22
Tutt et al The Lancet 2010 376(9737)235-44
Single agent olaparib in ovarian cancer
Moore K et al October 21 2018 DOI 101056NEJMoa1810858
OlympiAD
bull Tablet formulation (2 tablets twice daily)bull M=metastatic breast cancer HER2=human epidermal growth factor 2 TNBC=triple negative breast cancer TPC=treatment of physicianrsquos choice OS=overall survival PFS=progression-free survival PFS2=progression-free survival 2 ORR=objective response rate HRQoL=health-related quality of life FSI=first subject in RECiST= Response Evaluation Criteria in Solid Tumors ER=oestrogen receptor PR=progresterone receptor ECOG PS= Eastern Cooperative Oncology Group Performance Status gBRCAm=germline BRCA mutation po=oralbull 1 httpsclinicaltrialsgovct2showNCT02000622 2 Robson et al Poster OT1-1-04 San Antonio Breast Cancer Symposium 2014 3 AZ data on file (2017) 4 Robson et al N Engl J Med 2017 377523-533
Olaparib300mgpo bid
Treatment of Physicianrsquos
Choice (TPC)Capecitabine eribuline and vinorellbine
bull gBRCAm mBC
bull TNBC or HER2-negative ERPR positive
bull le2 prior chemotherapy lines for mBC
bull Previous treatment must include anthracycline and taxane
bull Hormone receptor positive (HR+) disease progressed on ge1 endocrine therapy or not suitable
bull If patients have received platinum therapy there should be
bull No evidence of progression during treatment in the advanced setting
bull At least 12 months since (neo)adjuvant treatment and randomisation
bull ECOG PS 0-1
bull At least one lesion that can be assessed by RECIST v11
Randomise 21
N=3024
Stratification by2
bull Prior chemotherapy regimens for metastatic breast cancer
bull Hormonal receptor (HR) status
bull Prior platinum therapy
Primary endpointbull PFS (RECIST 11
Independent Review)
Secondary endpointsbull OSbull PFS2bull ORRbull PFS PFS2 and OS
based on Myriad gBRCAm status
bull HRQoL (EORTC-QLQ-C30)
bull Safety and tolerability
FSI May 20143
Global Study in 19 countries and approximately 141 sites1
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Olaparib n=205n ()
TPCn=97n ()
Totaln=302n ()
Median age (min max) 44 (22 76) 45 (24 68) 44 (22 76)
Male 5 (24) 2 (21) 7 (23)
ECOG PS0 148 (722) 62 (639) 210 (695)
1 57 (278) 35 (361) 92 (304)
Race
White 134 (654) 63 (649) 202 (669)
Asian 66 (322) 28 (289) 94 (311)
Other 5 (24) 6 (62) 11 (36)
Olaparib n=205n ()
TPCn=97n()
Totaln=302n ()
Received previous chemotherapy formBC
Yes 146 (712) 69 (711) 215 (712)
No 59 (288) 28 (288) 87 (288)
Hormonal receptorstatus
HR+ 103 (502) 49 (505) 152 (503)
TNBC 102 (498) 48 (495) 150 (497)
Received prior platinum therapy for breast cancer
Yes 60 (293) 26 (268) 86 (284)
No 145 (71) 71 (73) 216 (715)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Number of metastatic sites
1 46 (224) 25 (258) 71 (235)
2 or more 159 (776) 72 (742) 231 (765)
Sites of metastatic lesions
Bone amp locomotor only 16 (78) 6 (62) 22 (73)
CNS 17 (83) 8 (82) 25 (83)
BRCA mutation status
BRCA1 117 (571) 51 (526) 168 (556)
BRCA2 84 (410) 46 (474) 130 (430)
Both 4 (19) 0 4 (13)
De novo metastatic BC 26 (127) 12 (124) 38 (126)
Progressive disease at randomisation 159 (776) 73 (753) 232 (768)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Received prior endocrine therapydagger
Adjuvantneoadjuvant 71 (689) 36 (735) 107 (704)
Metastatic 66 (641) 28 (571) 94 (618)
Total 97 (942) 45 (918) 142 (934)
Lines of previous cytotoxic chemotherapy for metastatic breast cancer
0 68 (332) 31 (320) 88 21-7(
1 80 (390) 42 (433) 011 3 -3(
2 57 (278) 24 (247) 81 (268)
Received prior platinum therapy for BCdagger
Metastatic 43 (210) 14 (144) 46 07-8(
Adjuvantneoadjuvant 15 (73) 7 (72) 11 6-2(
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205 n ()
TPCn=97 n ()
Totaln=302 n ()
Full analysis set (randomised ITT) 205 (100) 97 (100) 302 (100)
Patients who received study treatment 205 (100) 91 (938) 296 (980)
Number who received olaparib 205 (100) - 205 (679)
Number who received capecitabine - 41 (423) 41 (136)
Number who received eribulin - 34 (351) 34 (113)
Number who received vinorelbine - 16 (165) 16 ( 53)
Safety
Robson et al N Engl J Med 2017 377523-533
21
17
18
18
12
21
7
15
22
50
23
15
26
35
1
9
11
14
16
16
17
20
21
27
29
30
40
58
0 25 755075 50 25Adverse events ()
NauseaAnemiaVomitingFatigueNeutropeniaDiarrheaHeadacheCough
Decreased appetitePyrexiaIncreased ALTIncreased ASTHand-foot syndrome
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Safety
Robson et al N Engl J Med 2017 377523-533
2
2
3
0
1
3
1
10
26
4
0
1
1
2
2
2
3
3
9
16Anemia
Fatigue
Neutropenia
Increased AST
Hand-foot syndrome
Dyspnea
Decreased platelet count
Leukopenia
0 25 755075 50 25
Headache
Adverse events ()
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Progression free survival
Robson et al N Engl J Med 2017 377523-533
17783
15946
10
09
08
07
06
05
04
03
02
01
00
Prob
abili
ty o
f pro
gres
sion
-free
surv
ival
Time from randomisation (months)
14121086420 16 18 20 22 24 26 28
234
408
6911
9421
10725
15444
20597
214
111
41
31
21
10
00
214
367
6111
7313
10024
12929
20188
111
111
31
21
10
10
OlaparibTPC
Number of patientrsquos at risk Median PFS was improved by 69 with olaparib treatment compared to standard of care
chemotherapy
Olaparib TPC
n 205 97
Events () 163 (795)
71(732)
Median (m) 70 42
HR = 058 95 CI (043 080)
p=00009
PFS free at 6m()
541 329
PFS free at 12m()
259 150
Olaparib 300 mg bd (N=205)TPC (N=97)
Response rate
Robson et al N Engl J Med 2017 377523-533
Olaparib TPC
Response Evaluable Population n 167 66
ORR n () 100 (599) 19 (288)
Complete Response n () 15 (90) 1 (15)
Partial Response n () 85 (510) 18 (273)
Median Duration of Response months (95CI) 64 (29-97) 71 (32-122)
Median Time to Onset of Responsedays 47 45
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Metastatic triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
bull Triple negative breast cancer and BRCA-mutationsndash Clinical behaviorndash Genomic instability
Stephens et al Nature 2009 vol 462 (7276) pp 1005
Basal like 1 TNBC
Specific tumor cell killing
HR re
pair
Few normal tissue effectsNormal tissue cells
DNA repair
DNA repair
Base excision DNA repair
Homologous recombination(HR) repair
BRCA1BRCA2 deficientTumor cells
HR re
pair
PARP inhibitor
Base excision DNA repair
PARP inhibitor
Base excision DNA repair
HR re
pair
Tutt and Ashworth Cold Spring Harb Symp Quant Biol 200570139ndash148 McCabe N et al Cancer Res 2006668109ndash8115
The principle of synthetic lethal tumour targeting
Tutt et al The Lancet 2010 376(9737)235-44
To assess the efficacy and tolerability of oral olaparib in BRCA1 BRCA2mutation carriers with breast cancerProof-of-concept phase II study single-arm sequential cohort design
Confirmed BRCA1 or BRCA2 mutationAdvanced refractory breast cancer
(stage IIIBIIICIV) after failure of ge1 prior chemotherapy for advanced disease
Olaparib 400 mg po bid (MTD)28-day cycles n = 27
Olaparib 100 mg po bid28-day cycles n = 27
Cohort 2Cohort 1 (enrolled first)
400 mg BD 100 mg BD
Efficacy in BRCA12 breast cancer
ORR 41 ORR 22
Tutt et al The Lancet 2010 376(9737)235-44
Single agent olaparib in ovarian cancer
Moore K et al October 21 2018 DOI 101056NEJMoa1810858
OlympiAD
bull Tablet formulation (2 tablets twice daily)bull M=metastatic breast cancer HER2=human epidermal growth factor 2 TNBC=triple negative breast cancer TPC=treatment of physicianrsquos choice OS=overall survival PFS=progression-free survival PFS2=progression-free survival 2 ORR=objective response rate HRQoL=health-related quality of life FSI=first subject in RECiST= Response Evaluation Criteria in Solid Tumors ER=oestrogen receptor PR=progresterone receptor ECOG PS= Eastern Cooperative Oncology Group Performance Status gBRCAm=germline BRCA mutation po=oralbull 1 httpsclinicaltrialsgovct2showNCT02000622 2 Robson et al Poster OT1-1-04 San Antonio Breast Cancer Symposium 2014 3 AZ data on file (2017) 4 Robson et al N Engl J Med 2017 377523-533
Olaparib300mgpo bid
Treatment of Physicianrsquos
Choice (TPC)Capecitabine eribuline and vinorellbine
bull gBRCAm mBC
bull TNBC or HER2-negative ERPR positive
bull le2 prior chemotherapy lines for mBC
bull Previous treatment must include anthracycline and taxane
bull Hormone receptor positive (HR+) disease progressed on ge1 endocrine therapy or not suitable
bull If patients have received platinum therapy there should be
bull No evidence of progression during treatment in the advanced setting
bull At least 12 months since (neo)adjuvant treatment and randomisation
bull ECOG PS 0-1
bull At least one lesion that can be assessed by RECIST v11
Randomise 21
N=3024
Stratification by2
bull Prior chemotherapy regimens for metastatic breast cancer
bull Hormonal receptor (HR) status
bull Prior platinum therapy
Primary endpointbull PFS (RECIST 11
Independent Review)
Secondary endpointsbull OSbull PFS2bull ORRbull PFS PFS2 and OS
based on Myriad gBRCAm status
bull HRQoL (EORTC-QLQ-C30)
bull Safety and tolerability
FSI May 20143
Global Study in 19 countries and approximately 141 sites1
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Olaparib n=205n ()
TPCn=97n ()
Totaln=302n ()
Median age (min max) 44 (22 76) 45 (24 68) 44 (22 76)
Male 5 (24) 2 (21) 7 (23)
ECOG PS0 148 (722) 62 (639) 210 (695)
1 57 (278) 35 (361) 92 (304)
Race
White 134 (654) 63 (649) 202 (669)
Asian 66 (322) 28 (289) 94 (311)
Other 5 (24) 6 (62) 11 (36)
Olaparib n=205n ()
TPCn=97n()
Totaln=302n ()
Received previous chemotherapy formBC
Yes 146 (712) 69 (711) 215 (712)
No 59 (288) 28 (288) 87 (288)
Hormonal receptorstatus
HR+ 103 (502) 49 (505) 152 (503)
TNBC 102 (498) 48 (495) 150 (497)
Received prior platinum therapy for breast cancer
Yes 60 (293) 26 (268) 86 (284)
No 145 (71) 71 (73) 216 (715)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Number of metastatic sites
1 46 (224) 25 (258) 71 (235)
2 or more 159 (776) 72 (742) 231 (765)
Sites of metastatic lesions
Bone amp locomotor only 16 (78) 6 (62) 22 (73)
CNS 17 (83) 8 (82) 25 (83)
BRCA mutation status
BRCA1 117 (571) 51 (526) 168 (556)
BRCA2 84 (410) 46 (474) 130 (430)
Both 4 (19) 0 4 (13)
De novo metastatic BC 26 (127) 12 (124) 38 (126)
Progressive disease at randomisation 159 (776) 73 (753) 232 (768)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Received prior endocrine therapydagger
Adjuvantneoadjuvant 71 (689) 36 (735) 107 (704)
Metastatic 66 (641) 28 (571) 94 (618)
Total 97 (942) 45 (918) 142 (934)
Lines of previous cytotoxic chemotherapy for metastatic breast cancer
0 68 (332) 31 (320) 88 21-7(
1 80 (390) 42 (433) 011 3 -3(
2 57 (278) 24 (247) 81 (268)
Received prior platinum therapy for BCdagger
Metastatic 43 (210) 14 (144) 46 07-8(
Adjuvantneoadjuvant 15 (73) 7 (72) 11 6-2(
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205 n ()
TPCn=97 n ()
Totaln=302 n ()
Full analysis set (randomised ITT) 205 (100) 97 (100) 302 (100)
Patients who received study treatment 205 (100) 91 (938) 296 (980)
Number who received olaparib 205 (100) - 205 (679)
Number who received capecitabine - 41 (423) 41 (136)
Number who received eribulin - 34 (351) 34 (113)
Number who received vinorelbine - 16 (165) 16 ( 53)
Safety
Robson et al N Engl J Med 2017 377523-533
21
17
18
18
12
21
7
15
22
50
23
15
26
35
1
9
11
14
16
16
17
20
21
27
29
30
40
58
0 25 755075 50 25Adverse events ()
NauseaAnemiaVomitingFatigueNeutropeniaDiarrheaHeadacheCough
Decreased appetitePyrexiaIncreased ALTIncreased ASTHand-foot syndrome
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Safety
Robson et al N Engl J Med 2017 377523-533
2
2
3
0
1
3
1
10
26
4
0
1
1
2
2
2
3
3
9
16Anemia
Fatigue
Neutropenia
Increased AST
Hand-foot syndrome
Dyspnea
Decreased platelet count
Leukopenia
0 25 755075 50 25
Headache
Adverse events ()
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Progression free survival
Robson et al N Engl J Med 2017 377523-533
17783
15946
10
09
08
07
06
05
04
03
02
01
00
Prob
abili
ty o
f pro
gres
sion
-free
surv
ival
Time from randomisation (months)
14121086420 16 18 20 22 24 26 28
234
408
6911
9421
10725
15444
20597
214
111
41
31
21
10
00
214
367
6111
7313
10024
12929
20188
111
111
31
21
10
10
OlaparibTPC
Number of patientrsquos at risk Median PFS was improved by 69 with olaparib treatment compared to standard of care
chemotherapy
Olaparib TPC
n 205 97
Events () 163 (795)
71(732)
Median (m) 70 42
HR = 058 95 CI (043 080)
p=00009
PFS free at 6m()
541 329
PFS free at 12m()
259 150
Olaparib 300 mg bd (N=205)TPC (N=97)
Response rate
Robson et al N Engl J Med 2017 377523-533
Olaparib TPC
Response Evaluable Population n 167 66
ORR n () 100 (599) 19 (288)
Complete Response n () 15 (90) 1 (15)
Partial Response n () 85 (510) 18 (273)
Median Duration of Response months (95CI) 64 (29-97) 71 (32-122)
Median Time to Onset of Responsedays 47 45
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
bull Triple negative breast cancer and BRCA-mutationsndash Clinical behaviorndash Genomic instability
Stephens et al Nature 2009 vol 462 (7276) pp 1005
Basal like 1 TNBC
Specific tumor cell killing
HR re
pair
Few normal tissue effectsNormal tissue cells
DNA repair
DNA repair
Base excision DNA repair
Homologous recombination(HR) repair
BRCA1BRCA2 deficientTumor cells
HR re
pair
PARP inhibitor
Base excision DNA repair
PARP inhibitor
Base excision DNA repair
HR re
pair
Tutt and Ashworth Cold Spring Harb Symp Quant Biol 200570139ndash148 McCabe N et al Cancer Res 2006668109ndash8115
The principle of synthetic lethal tumour targeting
Tutt et al The Lancet 2010 376(9737)235-44
To assess the efficacy and tolerability of oral olaparib in BRCA1 BRCA2mutation carriers with breast cancerProof-of-concept phase II study single-arm sequential cohort design
Confirmed BRCA1 or BRCA2 mutationAdvanced refractory breast cancer
(stage IIIBIIICIV) after failure of ge1 prior chemotherapy for advanced disease
Olaparib 400 mg po bid (MTD)28-day cycles n = 27
Olaparib 100 mg po bid28-day cycles n = 27
Cohort 2Cohort 1 (enrolled first)
400 mg BD 100 mg BD
Efficacy in BRCA12 breast cancer
ORR 41 ORR 22
Tutt et al The Lancet 2010 376(9737)235-44
Single agent olaparib in ovarian cancer
Moore K et al October 21 2018 DOI 101056NEJMoa1810858
OlympiAD
bull Tablet formulation (2 tablets twice daily)bull M=metastatic breast cancer HER2=human epidermal growth factor 2 TNBC=triple negative breast cancer TPC=treatment of physicianrsquos choice OS=overall survival PFS=progression-free survival PFS2=progression-free survival 2 ORR=objective response rate HRQoL=health-related quality of life FSI=first subject in RECiST= Response Evaluation Criteria in Solid Tumors ER=oestrogen receptor PR=progresterone receptor ECOG PS= Eastern Cooperative Oncology Group Performance Status gBRCAm=germline BRCA mutation po=oralbull 1 httpsclinicaltrialsgovct2showNCT02000622 2 Robson et al Poster OT1-1-04 San Antonio Breast Cancer Symposium 2014 3 AZ data on file (2017) 4 Robson et al N Engl J Med 2017 377523-533
Olaparib300mgpo bid
Treatment of Physicianrsquos
Choice (TPC)Capecitabine eribuline and vinorellbine
bull gBRCAm mBC
bull TNBC or HER2-negative ERPR positive
bull le2 prior chemotherapy lines for mBC
bull Previous treatment must include anthracycline and taxane
bull Hormone receptor positive (HR+) disease progressed on ge1 endocrine therapy or not suitable
bull If patients have received platinum therapy there should be
bull No evidence of progression during treatment in the advanced setting
bull At least 12 months since (neo)adjuvant treatment and randomisation
bull ECOG PS 0-1
bull At least one lesion that can be assessed by RECIST v11
Randomise 21
N=3024
Stratification by2
bull Prior chemotherapy regimens for metastatic breast cancer
bull Hormonal receptor (HR) status
bull Prior platinum therapy
Primary endpointbull PFS (RECIST 11
Independent Review)
Secondary endpointsbull OSbull PFS2bull ORRbull PFS PFS2 and OS
based on Myriad gBRCAm status
bull HRQoL (EORTC-QLQ-C30)
bull Safety and tolerability
FSI May 20143
Global Study in 19 countries and approximately 141 sites1
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Olaparib n=205n ()
TPCn=97n ()
Totaln=302n ()
Median age (min max) 44 (22 76) 45 (24 68) 44 (22 76)
Male 5 (24) 2 (21) 7 (23)
ECOG PS0 148 (722) 62 (639) 210 (695)
1 57 (278) 35 (361) 92 (304)
Race
White 134 (654) 63 (649) 202 (669)
Asian 66 (322) 28 (289) 94 (311)
Other 5 (24) 6 (62) 11 (36)
Olaparib n=205n ()
TPCn=97n()
Totaln=302n ()
Received previous chemotherapy formBC
Yes 146 (712) 69 (711) 215 (712)
No 59 (288) 28 (288) 87 (288)
Hormonal receptorstatus
HR+ 103 (502) 49 (505) 152 (503)
TNBC 102 (498) 48 (495) 150 (497)
Received prior platinum therapy for breast cancer
Yes 60 (293) 26 (268) 86 (284)
No 145 (71) 71 (73) 216 (715)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Number of metastatic sites
1 46 (224) 25 (258) 71 (235)
2 or more 159 (776) 72 (742) 231 (765)
Sites of metastatic lesions
Bone amp locomotor only 16 (78) 6 (62) 22 (73)
CNS 17 (83) 8 (82) 25 (83)
BRCA mutation status
BRCA1 117 (571) 51 (526) 168 (556)
BRCA2 84 (410) 46 (474) 130 (430)
Both 4 (19) 0 4 (13)
De novo metastatic BC 26 (127) 12 (124) 38 (126)
Progressive disease at randomisation 159 (776) 73 (753) 232 (768)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Received prior endocrine therapydagger
Adjuvantneoadjuvant 71 (689) 36 (735) 107 (704)
Metastatic 66 (641) 28 (571) 94 (618)
Total 97 (942) 45 (918) 142 (934)
Lines of previous cytotoxic chemotherapy for metastatic breast cancer
0 68 (332) 31 (320) 88 21-7(
1 80 (390) 42 (433) 011 3 -3(
2 57 (278) 24 (247) 81 (268)
Received prior platinum therapy for BCdagger
Metastatic 43 (210) 14 (144) 46 07-8(
Adjuvantneoadjuvant 15 (73) 7 (72) 11 6-2(
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205 n ()
TPCn=97 n ()
Totaln=302 n ()
Full analysis set (randomised ITT) 205 (100) 97 (100) 302 (100)
Patients who received study treatment 205 (100) 91 (938) 296 (980)
Number who received olaparib 205 (100) - 205 (679)
Number who received capecitabine - 41 (423) 41 (136)
Number who received eribulin - 34 (351) 34 (113)
Number who received vinorelbine - 16 (165) 16 ( 53)
Safety
Robson et al N Engl J Med 2017 377523-533
21
17
18
18
12
21
7
15
22
50
23
15
26
35
1
9
11
14
16
16
17
20
21
27
29
30
40
58
0 25 755075 50 25Adverse events ()
NauseaAnemiaVomitingFatigueNeutropeniaDiarrheaHeadacheCough
Decreased appetitePyrexiaIncreased ALTIncreased ASTHand-foot syndrome
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Safety
Robson et al N Engl J Med 2017 377523-533
2
2
3
0
1
3
1
10
26
4
0
1
1
2
2
2
3
3
9
16Anemia
Fatigue
Neutropenia
Increased AST
Hand-foot syndrome
Dyspnea
Decreased platelet count
Leukopenia
0 25 755075 50 25
Headache
Adverse events ()
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Progression free survival
Robson et al N Engl J Med 2017 377523-533
17783
15946
10
09
08
07
06
05
04
03
02
01
00
Prob
abili
ty o
f pro
gres
sion
-free
surv
ival
Time from randomisation (months)
14121086420 16 18 20 22 24 26 28
234
408
6911
9421
10725
15444
20597
214
111
41
31
21
10
00
214
367
6111
7313
10024
12929
20188
111
111
31
21
10
10
OlaparibTPC
Number of patientrsquos at risk Median PFS was improved by 69 with olaparib treatment compared to standard of care
chemotherapy
Olaparib TPC
n 205 97
Events () 163 (795)
71(732)
Median (m) 70 42
HR = 058 95 CI (043 080)
p=00009
PFS free at 6m()
541 329
PFS free at 12m()
259 150
Olaparib 300 mg bd (N=205)TPC (N=97)
Response rate
Robson et al N Engl J Med 2017 377523-533
Olaparib TPC
Response Evaluable Population n 167 66
ORR n () 100 (599) 19 (288)
Complete Response n () 15 (90) 1 (15)
Partial Response n () 85 (510) 18 (273)
Median Duration of Response months (95CI) 64 (29-97) 71 (32-122)
Median Time to Onset of Responsedays 47 45
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Specific tumor cell killing
HR re
pair
Few normal tissue effectsNormal tissue cells
DNA repair
DNA repair
Base excision DNA repair
Homologous recombination(HR) repair
BRCA1BRCA2 deficientTumor cells
HR re
pair
PARP inhibitor
Base excision DNA repair
PARP inhibitor
Base excision DNA repair
HR re
pair
Tutt and Ashworth Cold Spring Harb Symp Quant Biol 200570139ndash148 McCabe N et al Cancer Res 2006668109ndash8115
The principle of synthetic lethal tumour targeting
Tutt et al The Lancet 2010 376(9737)235-44
To assess the efficacy and tolerability of oral olaparib in BRCA1 BRCA2mutation carriers with breast cancerProof-of-concept phase II study single-arm sequential cohort design
Confirmed BRCA1 or BRCA2 mutationAdvanced refractory breast cancer
(stage IIIBIIICIV) after failure of ge1 prior chemotherapy for advanced disease
Olaparib 400 mg po bid (MTD)28-day cycles n = 27
Olaparib 100 mg po bid28-day cycles n = 27
Cohort 2Cohort 1 (enrolled first)
400 mg BD 100 mg BD
Efficacy in BRCA12 breast cancer
ORR 41 ORR 22
Tutt et al The Lancet 2010 376(9737)235-44
Single agent olaparib in ovarian cancer
Moore K et al October 21 2018 DOI 101056NEJMoa1810858
OlympiAD
bull Tablet formulation (2 tablets twice daily)bull M=metastatic breast cancer HER2=human epidermal growth factor 2 TNBC=triple negative breast cancer TPC=treatment of physicianrsquos choice OS=overall survival PFS=progression-free survival PFS2=progression-free survival 2 ORR=objective response rate HRQoL=health-related quality of life FSI=first subject in RECiST= Response Evaluation Criteria in Solid Tumors ER=oestrogen receptor PR=progresterone receptor ECOG PS= Eastern Cooperative Oncology Group Performance Status gBRCAm=germline BRCA mutation po=oralbull 1 httpsclinicaltrialsgovct2showNCT02000622 2 Robson et al Poster OT1-1-04 San Antonio Breast Cancer Symposium 2014 3 AZ data on file (2017) 4 Robson et al N Engl J Med 2017 377523-533
Olaparib300mgpo bid
Treatment of Physicianrsquos
Choice (TPC)Capecitabine eribuline and vinorellbine
bull gBRCAm mBC
bull TNBC or HER2-negative ERPR positive
bull le2 prior chemotherapy lines for mBC
bull Previous treatment must include anthracycline and taxane
bull Hormone receptor positive (HR+) disease progressed on ge1 endocrine therapy or not suitable
bull If patients have received platinum therapy there should be
bull No evidence of progression during treatment in the advanced setting
bull At least 12 months since (neo)adjuvant treatment and randomisation
bull ECOG PS 0-1
bull At least one lesion that can be assessed by RECIST v11
Randomise 21
N=3024
Stratification by2
bull Prior chemotherapy regimens for metastatic breast cancer
bull Hormonal receptor (HR) status
bull Prior platinum therapy
Primary endpointbull PFS (RECIST 11
Independent Review)
Secondary endpointsbull OSbull PFS2bull ORRbull PFS PFS2 and OS
based on Myriad gBRCAm status
bull HRQoL (EORTC-QLQ-C30)
bull Safety and tolerability
FSI May 20143
Global Study in 19 countries and approximately 141 sites1
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Olaparib n=205n ()
TPCn=97n ()
Totaln=302n ()
Median age (min max) 44 (22 76) 45 (24 68) 44 (22 76)
Male 5 (24) 2 (21) 7 (23)
ECOG PS0 148 (722) 62 (639) 210 (695)
1 57 (278) 35 (361) 92 (304)
Race
White 134 (654) 63 (649) 202 (669)
Asian 66 (322) 28 (289) 94 (311)
Other 5 (24) 6 (62) 11 (36)
Olaparib n=205n ()
TPCn=97n()
Totaln=302n ()
Received previous chemotherapy formBC
Yes 146 (712) 69 (711) 215 (712)
No 59 (288) 28 (288) 87 (288)
Hormonal receptorstatus
HR+ 103 (502) 49 (505) 152 (503)
TNBC 102 (498) 48 (495) 150 (497)
Received prior platinum therapy for breast cancer
Yes 60 (293) 26 (268) 86 (284)
No 145 (71) 71 (73) 216 (715)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Number of metastatic sites
1 46 (224) 25 (258) 71 (235)
2 or more 159 (776) 72 (742) 231 (765)
Sites of metastatic lesions
Bone amp locomotor only 16 (78) 6 (62) 22 (73)
CNS 17 (83) 8 (82) 25 (83)
BRCA mutation status
BRCA1 117 (571) 51 (526) 168 (556)
BRCA2 84 (410) 46 (474) 130 (430)
Both 4 (19) 0 4 (13)
De novo metastatic BC 26 (127) 12 (124) 38 (126)
Progressive disease at randomisation 159 (776) 73 (753) 232 (768)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Received prior endocrine therapydagger
Adjuvantneoadjuvant 71 (689) 36 (735) 107 (704)
Metastatic 66 (641) 28 (571) 94 (618)
Total 97 (942) 45 (918) 142 (934)
Lines of previous cytotoxic chemotherapy for metastatic breast cancer
0 68 (332) 31 (320) 88 21-7(
1 80 (390) 42 (433) 011 3 -3(
2 57 (278) 24 (247) 81 (268)
Received prior platinum therapy for BCdagger
Metastatic 43 (210) 14 (144) 46 07-8(
Adjuvantneoadjuvant 15 (73) 7 (72) 11 6-2(
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205 n ()
TPCn=97 n ()
Totaln=302 n ()
Full analysis set (randomised ITT) 205 (100) 97 (100) 302 (100)
Patients who received study treatment 205 (100) 91 (938) 296 (980)
Number who received olaparib 205 (100) - 205 (679)
Number who received capecitabine - 41 (423) 41 (136)
Number who received eribulin - 34 (351) 34 (113)
Number who received vinorelbine - 16 (165) 16 ( 53)
Safety
Robson et al N Engl J Med 2017 377523-533
21
17
18
18
12
21
7
15
22
50
23
15
26
35
1
9
11
14
16
16
17
20
21
27
29
30
40
58
0 25 755075 50 25Adverse events ()
NauseaAnemiaVomitingFatigueNeutropeniaDiarrheaHeadacheCough
Decreased appetitePyrexiaIncreased ALTIncreased ASTHand-foot syndrome
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Safety
Robson et al N Engl J Med 2017 377523-533
2
2
3
0
1
3
1
10
26
4
0
1
1
2
2
2
3
3
9
16Anemia
Fatigue
Neutropenia
Increased AST
Hand-foot syndrome
Dyspnea
Decreased platelet count
Leukopenia
0 25 755075 50 25
Headache
Adverse events ()
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Progression free survival
Robson et al N Engl J Med 2017 377523-533
17783
15946
10
09
08
07
06
05
04
03
02
01
00
Prob
abili
ty o
f pro
gres
sion
-free
surv
ival
Time from randomisation (months)
14121086420 16 18 20 22 24 26 28
234
408
6911
9421
10725
15444
20597
214
111
41
31
21
10
00
214
367
6111
7313
10024
12929
20188
111
111
31
21
10
10
OlaparibTPC
Number of patientrsquos at risk Median PFS was improved by 69 with olaparib treatment compared to standard of care
chemotherapy
Olaparib TPC
n 205 97
Events () 163 (795)
71(732)
Median (m) 70 42
HR = 058 95 CI (043 080)
p=00009
PFS free at 6m()
541 329
PFS free at 12m()
259 150
Olaparib 300 mg bd (N=205)TPC (N=97)
Response rate
Robson et al N Engl J Med 2017 377523-533
Olaparib TPC
Response Evaluable Population n 167 66
ORR n () 100 (599) 19 (288)
Complete Response n () 15 (90) 1 (15)
Partial Response n () 85 (510) 18 (273)
Median Duration of Response months (95CI) 64 (29-97) 71 (32-122)
Median Time to Onset of Responsedays 47 45
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Tutt et al The Lancet 2010 376(9737)235-44
To assess the efficacy and tolerability of oral olaparib in BRCA1 BRCA2mutation carriers with breast cancerProof-of-concept phase II study single-arm sequential cohort design
Confirmed BRCA1 or BRCA2 mutationAdvanced refractory breast cancer
(stage IIIBIIICIV) after failure of ge1 prior chemotherapy for advanced disease
Olaparib 400 mg po bid (MTD)28-day cycles n = 27
Olaparib 100 mg po bid28-day cycles n = 27
Cohort 2Cohort 1 (enrolled first)
400 mg BD 100 mg BD
Efficacy in BRCA12 breast cancer
ORR 41 ORR 22
Tutt et al The Lancet 2010 376(9737)235-44
Single agent olaparib in ovarian cancer
Moore K et al October 21 2018 DOI 101056NEJMoa1810858
OlympiAD
bull Tablet formulation (2 tablets twice daily)bull M=metastatic breast cancer HER2=human epidermal growth factor 2 TNBC=triple negative breast cancer TPC=treatment of physicianrsquos choice OS=overall survival PFS=progression-free survival PFS2=progression-free survival 2 ORR=objective response rate HRQoL=health-related quality of life FSI=first subject in RECiST= Response Evaluation Criteria in Solid Tumors ER=oestrogen receptor PR=progresterone receptor ECOG PS= Eastern Cooperative Oncology Group Performance Status gBRCAm=germline BRCA mutation po=oralbull 1 httpsclinicaltrialsgovct2showNCT02000622 2 Robson et al Poster OT1-1-04 San Antonio Breast Cancer Symposium 2014 3 AZ data on file (2017) 4 Robson et al N Engl J Med 2017 377523-533
Olaparib300mgpo bid
Treatment of Physicianrsquos
Choice (TPC)Capecitabine eribuline and vinorellbine
bull gBRCAm mBC
bull TNBC or HER2-negative ERPR positive
bull le2 prior chemotherapy lines for mBC
bull Previous treatment must include anthracycline and taxane
bull Hormone receptor positive (HR+) disease progressed on ge1 endocrine therapy or not suitable
bull If patients have received platinum therapy there should be
bull No evidence of progression during treatment in the advanced setting
bull At least 12 months since (neo)adjuvant treatment and randomisation
bull ECOG PS 0-1
bull At least one lesion that can be assessed by RECIST v11
Randomise 21
N=3024
Stratification by2
bull Prior chemotherapy regimens for metastatic breast cancer
bull Hormonal receptor (HR) status
bull Prior platinum therapy
Primary endpointbull PFS (RECIST 11
Independent Review)
Secondary endpointsbull OSbull PFS2bull ORRbull PFS PFS2 and OS
based on Myriad gBRCAm status
bull HRQoL (EORTC-QLQ-C30)
bull Safety and tolerability
FSI May 20143
Global Study in 19 countries and approximately 141 sites1
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Olaparib n=205n ()
TPCn=97n ()
Totaln=302n ()
Median age (min max) 44 (22 76) 45 (24 68) 44 (22 76)
Male 5 (24) 2 (21) 7 (23)
ECOG PS0 148 (722) 62 (639) 210 (695)
1 57 (278) 35 (361) 92 (304)
Race
White 134 (654) 63 (649) 202 (669)
Asian 66 (322) 28 (289) 94 (311)
Other 5 (24) 6 (62) 11 (36)
Olaparib n=205n ()
TPCn=97n()
Totaln=302n ()
Received previous chemotherapy formBC
Yes 146 (712) 69 (711) 215 (712)
No 59 (288) 28 (288) 87 (288)
Hormonal receptorstatus
HR+ 103 (502) 49 (505) 152 (503)
TNBC 102 (498) 48 (495) 150 (497)
Received prior platinum therapy for breast cancer
Yes 60 (293) 26 (268) 86 (284)
No 145 (71) 71 (73) 216 (715)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Number of metastatic sites
1 46 (224) 25 (258) 71 (235)
2 or more 159 (776) 72 (742) 231 (765)
Sites of metastatic lesions
Bone amp locomotor only 16 (78) 6 (62) 22 (73)
CNS 17 (83) 8 (82) 25 (83)
BRCA mutation status
BRCA1 117 (571) 51 (526) 168 (556)
BRCA2 84 (410) 46 (474) 130 (430)
Both 4 (19) 0 4 (13)
De novo metastatic BC 26 (127) 12 (124) 38 (126)
Progressive disease at randomisation 159 (776) 73 (753) 232 (768)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Received prior endocrine therapydagger
Adjuvantneoadjuvant 71 (689) 36 (735) 107 (704)
Metastatic 66 (641) 28 (571) 94 (618)
Total 97 (942) 45 (918) 142 (934)
Lines of previous cytotoxic chemotherapy for metastatic breast cancer
0 68 (332) 31 (320) 88 21-7(
1 80 (390) 42 (433) 011 3 -3(
2 57 (278) 24 (247) 81 (268)
Received prior platinum therapy for BCdagger
Metastatic 43 (210) 14 (144) 46 07-8(
Adjuvantneoadjuvant 15 (73) 7 (72) 11 6-2(
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205 n ()
TPCn=97 n ()
Totaln=302 n ()
Full analysis set (randomised ITT) 205 (100) 97 (100) 302 (100)
Patients who received study treatment 205 (100) 91 (938) 296 (980)
Number who received olaparib 205 (100) - 205 (679)
Number who received capecitabine - 41 (423) 41 (136)
Number who received eribulin - 34 (351) 34 (113)
Number who received vinorelbine - 16 (165) 16 ( 53)
Safety
Robson et al N Engl J Med 2017 377523-533
21
17
18
18
12
21
7
15
22
50
23
15
26
35
1
9
11
14
16
16
17
20
21
27
29
30
40
58
0 25 755075 50 25Adverse events ()
NauseaAnemiaVomitingFatigueNeutropeniaDiarrheaHeadacheCough
Decreased appetitePyrexiaIncreased ALTIncreased ASTHand-foot syndrome
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Safety
Robson et al N Engl J Med 2017 377523-533
2
2
3
0
1
3
1
10
26
4
0
1
1
2
2
2
3
3
9
16Anemia
Fatigue
Neutropenia
Increased AST
Hand-foot syndrome
Dyspnea
Decreased platelet count
Leukopenia
0 25 755075 50 25
Headache
Adverse events ()
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Progression free survival
Robson et al N Engl J Med 2017 377523-533
17783
15946
10
09
08
07
06
05
04
03
02
01
00
Prob
abili
ty o
f pro
gres
sion
-free
surv
ival
Time from randomisation (months)
14121086420 16 18 20 22 24 26 28
234
408
6911
9421
10725
15444
20597
214
111
41
31
21
10
00
214
367
6111
7313
10024
12929
20188
111
111
31
21
10
10
OlaparibTPC
Number of patientrsquos at risk Median PFS was improved by 69 with olaparib treatment compared to standard of care
chemotherapy
Olaparib TPC
n 205 97
Events () 163 (795)
71(732)
Median (m) 70 42
HR = 058 95 CI (043 080)
p=00009
PFS free at 6m()
541 329
PFS free at 12m()
259 150
Olaparib 300 mg bd (N=205)TPC (N=97)
Response rate
Robson et al N Engl J Med 2017 377523-533
Olaparib TPC
Response Evaluable Population n 167 66
ORR n () 100 (599) 19 (288)
Complete Response n () 15 (90) 1 (15)
Partial Response n () 85 (510) 18 (273)
Median Duration of Response months (95CI) 64 (29-97) 71 (32-122)
Median Time to Onset of Responsedays 47 45
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
400 mg BD 100 mg BD
Efficacy in BRCA12 breast cancer
ORR 41 ORR 22
Tutt et al The Lancet 2010 376(9737)235-44
Single agent olaparib in ovarian cancer
Moore K et al October 21 2018 DOI 101056NEJMoa1810858
OlympiAD
bull Tablet formulation (2 tablets twice daily)bull M=metastatic breast cancer HER2=human epidermal growth factor 2 TNBC=triple negative breast cancer TPC=treatment of physicianrsquos choice OS=overall survival PFS=progression-free survival PFS2=progression-free survival 2 ORR=objective response rate HRQoL=health-related quality of life FSI=first subject in RECiST= Response Evaluation Criteria in Solid Tumors ER=oestrogen receptor PR=progresterone receptor ECOG PS= Eastern Cooperative Oncology Group Performance Status gBRCAm=germline BRCA mutation po=oralbull 1 httpsclinicaltrialsgovct2showNCT02000622 2 Robson et al Poster OT1-1-04 San Antonio Breast Cancer Symposium 2014 3 AZ data on file (2017) 4 Robson et al N Engl J Med 2017 377523-533
Olaparib300mgpo bid
Treatment of Physicianrsquos
Choice (TPC)Capecitabine eribuline and vinorellbine
bull gBRCAm mBC
bull TNBC or HER2-negative ERPR positive
bull le2 prior chemotherapy lines for mBC
bull Previous treatment must include anthracycline and taxane
bull Hormone receptor positive (HR+) disease progressed on ge1 endocrine therapy or not suitable
bull If patients have received platinum therapy there should be
bull No evidence of progression during treatment in the advanced setting
bull At least 12 months since (neo)adjuvant treatment and randomisation
bull ECOG PS 0-1
bull At least one lesion that can be assessed by RECIST v11
Randomise 21
N=3024
Stratification by2
bull Prior chemotherapy regimens for metastatic breast cancer
bull Hormonal receptor (HR) status
bull Prior platinum therapy
Primary endpointbull PFS (RECIST 11
Independent Review)
Secondary endpointsbull OSbull PFS2bull ORRbull PFS PFS2 and OS
based on Myriad gBRCAm status
bull HRQoL (EORTC-QLQ-C30)
bull Safety and tolerability
FSI May 20143
Global Study in 19 countries and approximately 141 sites1
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Olaparib n=205n ()
TPCn=97n ()
Totaln=302n ()
Median age (min max) 44 (22 76) 45 (24 68) 44 (22 76)
Male 5 (24) 2 (21) 7 (23)
ECOG PS0 148 (722) 62 (639) 210 (695)
1 57 (278) 35 (361) 92 (304)
Race
White 134 (654) 63 (649) 202 (669)
Asian 66 (322) 28 (289) 94 (311)
Other 5 (24) 6 (62) 11 (36)
Olaparib n=205n ()
TPCn=97n()
Totaln=302n ()
Received previous chemotherapy formBC
Yes 146 (712) 69 (711) 215 (712)
No 59 (288) 28 (288) 87 (288)
Hormonal receptorstatus
HR+ 103 (502) 49 (505) 152 (503)
TNBC 102 (498) 48 (495) 150 (497)
Received prior platinum therapy for breast cancer
Yes 60 (293) 26 (268) 86 (284)
No 145 (71) 71 (73) 216 (715)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Number of metastatic sites
1 46 (224) 25 (258) 71 (235)
2 or more 159 (776) 72 (742) 231 (765)
Sites of metastatic lesions
Bone amp locomotor only 16 (78) 6 (62) 22 (73)
CNS 17 (83) 8 (82) 25 (83)
BRCA mutation status
BRCA1 117 (571) 51 (526) 168 (556)
BRCA2 84 (410) 46 (474) 130 (430)
Both 4 (19) 0 4 (13)
De novo metastatic BC 26 (127) 12 (124) 38 (126)
Progressive disease at randomisation 159 (776) 73 (753) 232 (768)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Received prior endocrine therapydagger
Adjuvantneoadjuvant 71 (689) 36 (735) 107 (704)
Metastatic 66 (641) 28 (571) 94 (618)
Total 97 (942) 45 (918) 142 (934)
Lines of previous cytotoxic chemotherapy for metastatic breast cancer
0 68 (332) 31 (320) 88 21-7(
1 80 (390) 42 (433) 011 3 -3(
2 57 (278) 24 (247) 81 (268)
Received prior platinum therapy for BCdagger
Metastatic 43 (210) 14 (144) 46 07-8(
Adjuvantneoadjuvant 15 (73) 7 (72) 11 6-2(
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205 n ()
TPCn=97 n ()
Totaln=302 n ()
Full analysis set (randomised ITT) 205 (100) 97 (100) 302 (100)
Patients who received study treatment 205 (100) 91 (938) 296 (980)
Number who received olaparib 205 (100) - 205 (679)
Number who received capecitabine - 41 (423) 41 (136)
Number who received eribulin - 34 (351) 34 (113)
Number who received vinorelbine - 16 (165) 16 ( 53)
Safety
Robson et al N Engl J Med 2017 377523-533
21
17
18
18
12
21
7
15
22
50
23
15
26
35
1
9
11
14
16
16
17
20
21
27
29
30
40
58
0 25 755075 50 25Adverse events ()
NauseaAnemiaVomitingFatigueNeutropeniaDiarrheaHeadacheCough
Decreased appetitePyrexiaIncreased ALTIncreased ASTHand-foot syndrome
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Safety
Robson et al N Engl J Med 2017 377523-533
2
2
3
0
1
3
1
10
26
4
0
1
1
2
2
2
3
3
9
16Anemia
Fatigue
Neutropenia
Increased AST
Hand-foot syndrome
Dyspnea
Decreased platelet count
Leukopenia
0 25 755075 50 25
Headache
Adverse events ()
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Progression free survival
Robson et al N Engl J Med 2017 377523-533
17783
15946
10
09
08
07
06
05
04
03
02
01
00
Prob
abili
ty o
f pro
gres
sion
-free
surv
ival
Time from randomisation (months)
14121086420 16 18 20 22 24 26 28
234
408
6911
9421
10725
15444
20597
214
111
41
31
21
10
00
214
367
6111
7313
10024
12929
20188
111
111
31
21
10
10
OlaparibTPC
Number of patientrsquos at risk Median PFS was improved by 69 with olaparib treatment compared to standard of care
chemotherapy
Olaparib TPC
n 205 97
Events () 163 (795)
71(732)
Median (m) 70 42
HR = 058 95 CI (043 080)
p=00009
PFS free at 6m()
541 329
PFS free at 12m()
259 150
Olaparib 300 mg bd (N=205)TPC (N=97)
Response rate
Robson et al N Engl J Med 2017 377523-533
Olaparib TPC
Response Evaluable Population n 167 66
ORR n () 100 (599) 19 (288)
Complete Response n () 15 (90) 1 (15)
Partial Response n () 85 (510) 18 (273)
Median Duration of Response months (95CI) 64 (29-97) 71 (32-122)
Median Time to Onset of Responsedays 47 45
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Single agent olaparib in ovarian cancer
Moore K et al October 21 2018 DOI 101056NEJMoa1810858
OlympiAD
bull Tablet formulation (2 tablets twice daily)bull M=metastatic breast cancer HER2=human epidermal growth factor 2 TNBC=triple negative breast cancer TPC=treatment of physicianrsquos choice OS=overall survival PFS=progression-free survival PFS2=progression-free survival 2 ORR=objective response rate HRQoL=health-related quality of life FSI=first subject in RECiST= Response Evaluation Criteria in Solid Tumors ER=oestrogen receptor PR=progresterone receptor ECOG PS= Eastern Cooperative Oncology Group Performance Status gBRCAm=germline BRCA mutation po=oralbull 1 httpsclinicaltrialsgovct2showNCT02000622 2 Robson et al Poster OT1-1-04 San Antonio Breast Cancer Symposium 2014 3 AZ data on file (2017) 4 Robson et al N Engl J Med 2017 377523-533
Olaparib300mgpo bid
Treatment of Physicianrsquos
Choice (TPC)Capecitabine eribuline and vinorellbine
bull gBRCAm mBC
bull TNBC or HER2-negative ERPR positive
bull le2 prior chemotherapy lines for mBC
bull Previous treatment must include anthracycline and taxane
bull Hormone receptor positive (HR+) disease progressed on ge1 endocrine therapy or not suitable
bull If patients have received platinum therapy there should be
bull No evidence of progression during treatment in the advanced setting
bull At least 12 months since (neo)adjuvant treatment and randomisation
bull ECOG PS 0-1
bull At least one lesion that can be assessed by RECIST v11
Randomise 21
N=3024
Stratification by2
bull Prior chemotherapy regimens for metastatic breast cancer
bull Hormonal receptor (HR) status
bull Prior platinum therapy
Primary endpointbull PFS (RECIST 11
Independent Review)
Secondary endpointsbull OSbull PFS2bull ORRbull PFS PFS2 and OS
based on Myriad gBRCAm status
bull HRQoL (EORTC-QLQ-C30)
bull Safety and tolerability
FSI May 20143
Global Study in 19 countries and approximately 141 sites1
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Olaparib n=205n ()
TPCn=97n ()
Totaln=302n ()
Median age (min max) 44 (22 76) 45 (24 68) 44 (22 76)
Male 5 (24) 2 (21) 7 (23)
ECOG PS0 148 (722) 62 (639) 210 (695)
1 57 (278) 35 (361) 92 (304)
Race
White 134 (654) 63 (649) 202 (669)
Asian 66 (322) 28 (289) 94 (311)
Other 5 (24) 6 (62) 11 (36)
Olaparib n=205n ()
TPCn=97n()
Totaln=302n ()
Received previous chemotherapy formBC
Yes 146 (712) 69 (711) 215 (712)
No 59 (288) 28 (288) 87 (288)
Hormonal receptorstatus
HR+ 103 (502) 49 (505) 152 (503)
TNBC 102 (498) 48 (495) 150 (497)
Received prior platinum therapy for breast cancer
Yes 60 (293) 26 (268) 86 (284)
No 145 (71) 71 (73) 216 (715)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Number of metastatic sites
1 46 (224) 25 (258) 71 (235)
2 or more 159 (776) 72 (742) 231 (765)
Sites of metastatic lesions
Bone amp locomotor only 16 (78) 6 (62) 22 (73)
CNS 17 (83) 8 (82) 25 (83)
BRCA mutation status
BRCA1 117 (571) 51 (526) 168 (556)
BRCA2 84 (410) 46 (474) 130 (430)
Both 4 (19) 0 4 (13)
De novo metastatic BC 26 (127) 12 (124) 38 (126)
Progressive disease at randomisation 159 (776) 73 (753) 232 (768)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Received prior endocrine therapydagger
Adjuvantneoadjuvant 71 (689) 36 (735) 107 (704)
Metastatic 66 (641) 28 (571) 94 (618)
Total 97 (942) 45 (918) 142 (934)
Lines of previous cytotoxic chemotherapy for metastatic breast cancer
0 68 (332) 31 (320) 88 21-7(
1 80 (390) 42 (433) 011 3 -3(
2 57 (278) 24 (247) 81 (268)
Received prior platinum therapy for BCdagger
Metastatic 43 (210) 14 (144) 46 07-8(
Adjuvantneoadjuvant 15 (73) 7 (72) 11 6-2(
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205 n ()
TPCn=97 n ()
Totaln=302 n ()
Full analysis set (randomised ITT) 205 (100) 97 (100) 302 (100)
Patients who received study treatment 205 (100) 91 (938) 296 (980)
Number who received olaparib 205 (100) - 205 (679)
Number who received capecitabine - 41 (423) 41 (136)
Number who received eribulin - 34 (351) 34 (113)
Number who received vinorelbine - 16 (165) 16 ( 53)
Safety
Robson et al N Engl J Med 2017 377523-533
21
17
18
18
12
21
7
15
22
50
23
15
26
35
1
9
11
14
16
16
17
20
21
27
29
30
40
58
0 25 755075 50 25Adverse events ()
NauseaAnemiaVomitingFatigueNeutropeniaDiarrheaHeadacheCough
Decreased appetitePyrexiaIncreased ALTIncreased ASTHand-foot syndrome
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Safety
Robson et al N Engl J Med 2017 377523-533
2
2
3
0
1
3
1
10
26
4
0
1
1
2
2
2
3
3
9
16Anemia
Fatigue
Neutropenia
Increased AST
Hand-foot syndrome
Dyspnea
Decreased platelet count
Leukopenia
0 25 755075 50 25
Headache
Adverse events ()
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Progression free survival
Robson et al N Engl J Med 2017 377523-533
17783
15946
10
09
08
07
06
05
04
03
02
01
00
Prob
abili
ty o
f pro
gres
sion
-free
surv
ival
Time from randomisation (months)
14121086420 16 18 20 22 24 26 28
234
408
6911
9421
10725
15444
20597
214
111
41
31
21
10
00
214
367
6111
7313
10024
12929
20188
111
111
31
21
10
10
OlaparibTPC
Number of patientrsquos at risk Median PFS was improved by 69 with olaparib treatment compared to standard of care
chemotherapy
Olaparib TPC
n 205 97
Events () 163 (795)
71(732)
Median (m) 70 42
HR = 058 95 CI (043 080)
p=00009
PFS free at 6m()
541 329
PFS free at 12m()
259 150
Olaparib 300 mg bd (N=205)TPC (N=97)
Response rate
Robson et al N Engl J Med 2017 377523-533
Olaparib TPC
Response Evaluable Population n 167 66
ORR n () 100 (599) 19 (288)
Complete Response n () 15 (90) 1 (15)
Partial Response n () 85 (510) 18 (273)
Median Duration of Response months (95CI) 64 (29-97) 71 (32-122)
Median Time to Onset of Responsedays 47 45
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
OlympiAD
bull Tablet formulation (2 tablets twice daily)bull M=metastatic breast cancer HER2=human epidermal growth factor 2 TNBC=triple negative breast cancer TPC=treatment of physicianrsquos choice OS=overall survival PFS=progression-free survival PFS2=progression-free survival 2 ORR=objective response rate HRQoL=health-related quality of life FSI=first subject in RECiST= Response Evaluation Criteria in Solid Tumors ER=oestrogen receptor PR=progresterone receptor ECOG PS= Eastern Cooperative Oncology Group Performance Status gBRCAm=germline BRCA mutation po=oralbull 1 httpsclinicaltrialsgovct2showNCT02000622 2 Robson et al Poster OT1-1-04 San Antonio Breast Cancer Symposium 2014 3 AZ data on file (2017) 4 Robson et al N Engl J Med 2017 377523-533
Olaparib300mgpo bid
Treatment of Physicianrsquos
Choice (TPC)Capecitabine eribuline and vinorellbine
bull gBRCAm mBC
bull TNBC or HER2-negative ERPR positive
bull le2 prior chemotherapy lines for mBC
bull Previous treatment must include anthracycline and taxane
bull Hormone receptor positive (HR+) disease progressed on ge1 endocrine therapy or not suitable
bull If patients have received platinum therapy there should be
bull No evidence of progression during treatment in the advanced setting
bull At least 12 months since (neo)adjuvant treatment and randomisation
bull ECOG PS 0-1
bull At least one lesion that can be assessed by RECIST v11
Randomise 21
N=3024
Stratification by2
bull Prior chemotherapy regimens for metastatic breast cancer
bull Hormonal receptor (HR) status
bull Prior platinum therapy
Primary endpointbull PFS (RECIST 11
Independent Review)
Secondary endpointsbull OSbull PFS2bull ORRbull PFS PFS2 and OS
based on Myriad gBRCAm status
bull HRQoL (EORTC-QLQ-C30)
bull Safety and tolerability
FSI May 20143
Global Study in 19 countries and approximately 141 sites1
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Olaparib n=205n ()
TPCn=97n ()
Totaln=302n ()
Median age (min max) 44 (22 76) 45 (24 68) 44 (22 76)
Male 5 (24) 2 (21) 7 (23)
ECOG PS0 148 (722) 62 (639) 210 (695)
1 57 (278) 35 (361) 92 (304)
Race
White 134 (654) 63 (649) 202 (669)
Asian 66 (322) 28 (289) 94 (311)
Other 5 (24) 6 (62) 11 (36)
Olaparib n=205n ()
TPCn=97n()
Totaln=302n ()
Received previous chemotherapy formBC
Yes 146 (712) 69 (711) 215 (712)
No 59 (288) 28 (288) 87 (288)
Hormonal receptorstatus
HR+ 103 (502) 49 (505) 152 (503)
TNBC 102 (498) 48 (495) 150 (497)
Received prior platinum therapy for breast cancer
Yes 60 (293) 26 (268) 86 (284)
No 145 (71) 71 (73) 216 (715)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Number of metastatic sites
1 46 (224) 25 (258) 71 (235)
2 or more 159 (776) 72 (742) 231 (765)
Sites of metastatic lesions
Bone amp locomotor only 16 (78) 6 (62) 22 (73)
CNS 17 (83) 8 (82) 25 (83)
BRCA mutation status
BRCA1 117 (571) 51 (526) 168 (556)
BRCA2 84 (410) 46 (474) 130 (430)
Both 4 (19) 0 4 (13)
De novo metastatic BC 26 (127) 12 (124) 38 (126)
Progressive disease at randomisation 159 (776) 73 (753) 232 (768)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Received prior endocrine therapydagger
Adjuvantneoadjuvant 71 (689) 36 (735) 107 (704)
Metastatic 66 (641) 28 (571) 94 (618)
Total 97 (942) 45 (918) 142 (934)
Lines of previous cytotoxic chemotherapy for metastatic breast cancer
0 68 (332) 31 (320) 88 21-7(
1 80 (390) 42 (433) 011 3 -3(
2 57 (278) 24 (247) 81 (268)
Received prior platinum therapy for BCdagger
Metastatic 43 (210) 14 (144) 46 07-8(
Adjuvantneoadjuvant 15 (73) 7 (72) 11 6-2(
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205 n ()
TPCn=97 n ()
Totaln=302 n ()
Full analysis set (randomised ITT) 205 (100) 97 (100) 302 (100)
Patients who received study treatment 205 (100) 91 (938) 296 (980)
Number who received olaparib 205 (100) - 205 (679)
Number who received capecitabine - 41 (423) 41 (136)
Number who received eribulin - 34 (351) 34 (113)
Number who received vinorelbine - 16 (165) 16 ( 53)
Safety
Robson et al N Engl J Med 2017 377523-533
21
17
18
18
12
21
7
15
22
50
23
15
26
35
1
9
11
14
16
16
17
20
21
27
29
30
40
58
0 25 755075 50 25Adverse events ()
NauseaAnemiaVomitingFatigueNeutropeniaDiarrheaHeadacheCough
Decreased appetitePyrexiaIncreased ALTIncreased ASTHand-foot syndrome
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Safety
Robson et al N Engl J Med 2017 377523-533
2
2
3
0
1
3
1
10
26
4
0
1
1
2
2
2
3
3
9
16Anemia
Fatigue
Neutropenia
Increased AST
Hand-foot syndrome
Dyspnea
Decreased platelet count
Leukopenia
0 25 755075 50 25
Headache
Adverse events ()
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Progression free survival
Robson et al N Engl J Med 2017 377523-533
17783
15946
10
09
08
07
06
05
04
03
02
01
00
Prob
abili
ty o
f pro
gres
sion
-free
surv
ival
Time from randomisation (months)
14121086420 16 18 20 22 24 26 28
234
408
6911
9421
10725
15444
20597
214
111
41
31
21
10
00
214
367
6111
7313
10024
12929
20188
111
111
31
21
10
10
OlaparibTPC
Number of patientrsquos at risk Median PFS was improved by 69 with olaparib treatment compared to standard of care
chemotherapy
Olaparib TPC
n 205 97
Events () 163 (795)
71(732)
Median (m) 70 42
HR = 058 95 CI (043 080)
p=00009
PFS free at 6m()
541 329
PFS free at 12m()
259 150
Olaparib 300 mg bd (N=205)TPC (N=97)
Response rate
Robson et al N Engl J Med 2017 377523-533
Olaparib TPC
Response Evaluable Population n 167 66
ORR n () 100 (599) 19 (288)
Complete Response n () 15 (90) 1 (15)
Partial Response n () 85 (510) 18 (273)
Median Duration of Response months (95CI) 64 (29-97) 71 (32-122)
Median Time to Onset of Responsedays 47 45
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Olaparib n=205n ()
TPCn=97n ()
Totaln=302n ()
Median age (min max) 44 (22 76) 45 (24 68) 44 (22 76)
Male 5 (24) 2 (21) 7 (23)
ECOG PS0 148 (722) 62 (639) 210 (695)
1 57 (278) 35 (361) 92 (304)
Race
White 134 (654) 63 (649) 202 (669)
Asian 66 (322) 28 (289) 94 (311)
Other 5 (24) 6 (62) 11 (36)
Olaparib n=205n ()
TPCn=97n()
Totaln=302n ()
Received previous chemotherapy formBC
Yes 146 (712) 69 (711) 215 (712)
No 59 (288) 28 (288) 87 (288)
Hormonal receptorstatus
HR+ 103 (502) 49 (505) 152 (503)
TNBC 102 (498) 48 (495) 150 (497)
Received prior platinum therapy for breast cancer
Yes 60 (293) 26 (268) 86 (284)
No 145 (71) 71 (73) 216 (715)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Number of metastatic sites
1 46 (224) 25 (258) 71 (235)
2 or more 159 (776) 72 (742) 231 (765)
Sites of metastatic lesions
Bone amp locomotor only 16 (78) 6 (62) 22 (73)
CNS 17 (83) 8 (82) 25 (83)
BRCA mutation status
BRCA1 117 (571) 51 (526) 168 (556)
BRCA2 84 (410) 46 (474) 130 (430)
Both 4 (19) 0 4 (13)
De novo metastatic BC 26 (127) 12 (124) 38 (126)
Progressive disease at randomisation 159 (776) 73 (753) 232 (768)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Received prior endocrine therapydagger
Adjuvantneoadjuvant 71 (689) 36 (735) 107 (704)
Metastatic 66 (641) 28 (571) 94 (618)
Total 97 (942) 45 (918) 142 (934)
Lines of previous cytotoxic chemotherapy for metastatic breast cancer
0 68 (332) 31 (320) 88 21-7(
1 80 (390) 42 (433) 011 3 -3(
2 57 (278) 24 (247) 81 (268)
Received prior platinum therapy for BCdagger
Metastatic 43 (210) 14 (144) 46 07-8(
Adjuvantneoadjuvant 15 (73) 7 (72) 11 6-2(
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205 n ()
TPCn=97 n ()
Totaln=302 n ()
Full analysis set (randomised ITT) 205 (100) 97 (100) 302 (100)
Patients who received study treatment 205 (100) 91 (938) 296 (980)
Number who received olaparib 205 (100) - 205 (679)
Number who received capecitabine - 41 (423) 41 (136)
Number who received eribulin - 34 (351) 34 (113)
Number who received vinorelbine - 16 (165) 16 ( 53)
Safety
Robson et al N Engl J Med 2017 377523-533
21
17
18
18
12
21
7
15
22
50
23
15
26
35
1
9
11
14
16
16
17
20
21
27
29
30
40
58
0 25 755075 50 25Adverse events ()
NauseaAnemiaVomitingFatigueNeutropeniaDiarrheaHeadacheCough
Decreased appetitePyrexiaIncreased ALTIncreased ASTHand-foot syndrome
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Safety
Robson et al N Engl J Med 2017 377523-533
2
2
3
0
1
3
1
10
26
4
0
1
1
2
2
2
3
3
9
16Anemia
Fatigue
Neutropenia
Increased AST
Hand-foot syndrome
Dyspnea
Decreased platelet count
Leukopenia
0 25 755075 50 25
Headache
Adverse events ()
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Progression free survival
Robson et al N Engl J Med 2017 377523-533
17783
15946
10
09
08
07
06
05
04
03
02
01
00
Prob
abili
ty o
f pro
gres
sion
-free
surv
ival
Time from randomisation (months)
14121086420 16 18 20 22 24 26 28
234
408
6911
9421
10725
15444
20597
214
111
41
31
21
10
00
214
367
6111
7313
10024
12929
20188
111
111
31
21
10
10
OlaparibTPC
Number of patientrsquos at risk Median PFS was improved by 69 with olaparib treatment compared to standard of care
chemotherapy
Olaparib TPC
n 205 97
Events () 163 (795)
71(732)
Median (m) 70 42
HR = 058 95 CI (043 080)
p=00009
PFS free at 6m()
541 329
PFS free at 12m()
259 150
Olaparib 300 mg bd (N=205)TPC (N=97)
Response rate
Robson et al N Engl J Med 2017 377523-533
Olaparib TPC
Response Evaluable Population n 167 66
ORR n () 100 (599) 19 (288)
Complete Response n () 15 (90) 1 (15)
Partial Response n () 85 (510) 18 (273)
Median Duration of Response months (95CI) 64 (29-97) 71 (32-122)
Median Time to Onset of Responsedays 47 45
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Olaparib n=205n ()
TPCn=97n()
Totaln=302n ()
Received previous chemotherapy formBC
Yes 146 (712) 69 (711) 215 (712)
No 59 (288) 28 (288) 87 (288)
Hormonal receptorstatus
HR+ 103 (502) 49 (505) 152 (503)
TNBC 102 (498) 48 (495) 150 (497)
Received prior platinum therapy for breast cancer
Yes 60 (293) 26 (268) 86 (284)
No 145 (71) 71 (73) 216 (715)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Number of metastatic sites
1 46 (224) 25 (258) 71 (235)
2 or more 159 (776) 72 (742) 231 (765)
Sites of metastatic lesions
Bone amp locomotor only 16 (78) 6 (62) 22 (73)
CNS 17 (83) 8 (82) 25 (83)
BRCA mutation status
BRCA1 117 (571) 51 (526) 168 (556)
BRCA2 84 (410) 46 (474) 130 (430)
Both 4 (19) 0 4 (13)
De novo metastatic BC 26 (127) 12 (124) 38 (126)
Progressive disease at randomisation 159 (776) 73 (753) 232 (768)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Received prior endocrine therapydagger
Adjuvantneoadjuvant 71 (689) 36 (735) 107 (704)
Metastatic 66 (641) 28 (571) 94 (618)
Total 97 (942) 45 (918) 142 (934)
Lines of previous cytotoxic chemotherapy for metastatic breast cancer
0 68 (332) 31 (320) 88 21-7(
1 80 (390) 42 (433) 011 3 -3(
2 57 (278) 24 (247) 81 (268)
Received prior platinum therapy for BCdagger
Metastatic 43 (210) 14 (144) 46 07-8(
Adjuvantneoadjuvant 15 (73) 7 (72) 11 6-2(
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205 n ()
TPCn=97 n ()
Totaln=302 n ()
Full analysis set (randomised ITT) 205 (100) 97 (100) 302 (100)
Patients who received study treatment 205 (100) 91 (938) 296 (980)
Number who received olaparib 205 (100) - 205 (679)
Number who received capecitabine - 41 (423) 41 (136)
Number who received eribulin - 34 (351) 34 (113)
Number who received vinorelbine - 16 (165) 16 ( 53)
Safety
Robson et al N Engl J Med 2017 377523-533
21
17
18
18
12
21
7
15
22
50
23
15
26
35
1
9
11
14
16
16
17
20
21
27
29
30
40
58
0 25 755075 50 25Adverse events ()
NauseaAnemiaVomitingFatigueNeutropeniaDiarrheaHeadacheCough
Decreased appetitePyrexiaIncreased ALTIncreased ASTHand-foot syndrome
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Safety
Robson et al N Engl J Med 2017 377523-533
2
2
3
0
1
3
1
10
26
4
0
1
1
2
2
2
3
3
9
16Anemia
Fatigue
Neutropenia
Increased AST
Hand-foot syndrome
Dyspnea
Decreased platelet count
Leukopenia
0 25 755075 50 25
Headache
Adverse events ()
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Progression free survival
Robson et al N Engl J Med 2017 377523-533
17783
15946
10
09
08
07
06
05
04
03
02
01
00
Prob
abili
ty o
f pro
gres
sion
-free
surv
ival
Time from randomisation (months)
14121086420 16 18 20 22 24 26 28
234
408
6911
9421
10725
15444
20597
214
111
41
31
21
10
00
214
367
6111
7313
10024
12929
20188
111
111
31
21
10
10
OlaparibTPC
Number of patientrsquos at risk Median PFS was improved by 69 with olaparib treatment compared to standard of care
chemotherapy
Olaparib TPC
n 205 97
Events () 163 (795)
71(732)
Median (m) 70 42
HR = 058 95 CI (043 080)
p=00009
PFS free at 6m()
541 329
PFS free at 12m()
259 150
Olaparib 300 mg bd (N=205)TPC (N=97)
Response rate
Robson et al N Engl J Med 2017 377523-533
Olaparib TPC
Response Evaluable Population n 167 66
ORR n () 100 (599) 19 (288)
Complete Response n () 15 (90) 1 (15)
Partial Response n () 85 (510) 18 (273)
Median Duration of Response months (95CI) 64 (29-97) 71 (32-122)
Median Time to Onset of Responsedays 47 45
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Number of metastatic sites
1 46 (224) 25 (258) 71 (235)
2 or more 159 (776) 72 (742) 231 (765)
Sites of metastatic lesions
Bone amp locomotor only 16 (78) 6 (62) 22 (73)
CNS 17 (83) 8 (82) 25 (83)
BRCA mutation status
BRCA1 117 (571) 51 (526) 168 (556)
BRCA2 84 (410) 46 (474) 130 (430)
Both 4 (19) 0 4 (13)
De novo metastatic BC 26 (127) 12 (124) 38 (126)
Progressive disease at randomisation 159 (776) 73 (753) 232 (768)
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Received prior endocrine therapydagger
Adjuvantneoadjuvant 71 (689) 36 (735) 107 (704)
Metastatic 66 (641) 28 (571) 94 (618)
Total 97 (942) 45 (918) 142 (934)
Lines of previous cytotoxic chemotherapy for metastatic breast cancer
0 68 (332) 31 (320) 88 21-7(
1 80 (390) 42 (433) 011 3 -3(
2 57 (278) 24 (247) 81 (268)
Received prior platinum therapy for BCdagger
Metastatic 43 (210) 14 (144) 46 07-8(
Adjuvantneoadjuvant 15 (73) 7 (72) 11 6-2(
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205 n ()
TPCn=97 n ()
Totaln=302 n ()
Full analysis set (randomised ITT) 205 (100) 97 (100) 302 (100)
Patients who received study treatment 205 (100) 91 (938) 296 (980)
Number who received olaparib 205 (100) - 205 (679)
Number who received capecitabine - 41 (423) 41 (136)
Number who received eribulin - 34 (351) 34 (113)
Number who received vinorelbine - 16 (165) 16 ( 53)
Safety
Robson et al N Engl J Med 2017 377523-533
21
17
18
18
12
21
7
15
22
50
23
15
26
35
1
9
11
14
16
16
17
20
21
27
29
30
40
58
0 25 755075 50 25Adverse events ()
NauseaAnemiaVomitingFatigueNeutropeniaDiarrheaHeadacheCough
Decreased appetitePyrexiaIncreased ALTIncreased ASTHand-foot syndrome
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Safety
Robson et al N Engl J Med 2017 377523-533
2
2
3
0
1
3
1
10
26
4
0
1
1
2
2
2
3
3
9
16Anemia
Fatigue
Neutropenia
Increased AST
Hand-foot syndrome
Dyspnea
Decreased platelet count
Leukopenia
0 25 755075 50 25
Headache
Adverse events ()
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Progression free survival
Robson et al N Engl J Med 2017 377523-533
17783
15946
10
09
08
07
06
05
04
03
02
01
00
Prob
abili
ty o
f pro
gres
sion
-free
surv
ival
Time from randomisation (months)
14121086420 16 18 20 22 24 26 28
234
408
6911
9421
10725
15444
20597
214
111
41
31
21
10
00
214
367
6111
7313
10024
12929
20188
111
111
31
21
10
10
OlaparibTPC
Number of patientrsquos at risk Median PFS was improved by 69 with olaparib treatment compared to standard of care
chemotherapy
Olaparib TPC
n 205 97
Events () 163 (795)
71(732)
Median (m) 70 42
HR = 058 95 CI (043 080)
p=00009
PFS free at 6m()
541 329
PFS free at 12m()
259 150
Olaparib 300 mg bd (N=205)TPC (N=97)
Response rate
Robson et al N Engl J Med 2017 377523-533
Olaparib TPC
Response Evaluable Population n 167 66
ORR n () 100 (599) 19 (288)
Complete Response n () 15 (90) 1 (15)
Partial Response n () 85 (510) 18 (273)
Median Duration of Response months (95CI) 64 (29-97) 71 (32-122)
Median Time to Onset of Responsedays 47 45
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205n ()
TPCn=97n()
Totaln=302n ()
Received prior endocrine therapydagger
Adjuvantneoadjuvant 71 (689) 36 (735) 107 (704)
Metastatic 66 (641) 28 (571) 94 (618)
Total 97 (942) 45 (918) 142 (934)
Lines of previous cytotoxic chemotherapy for metastatic breast cancer
0 68 (332) 31 (320) 88 21-7(
1 80 (390) 42 (433) 011 3 -3(
2 57 (278) 24 (247) 81 (268)
Received prior platinum therapy for BCdagger
Metastatic 43 (210) 14 (144) 46 07-8(
Adjuvantneoadjuvant 15 (73) 7 (72) 11 6-2(
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205 n ()
TPCn=97 n ()
Totaln=302 n ()
Full analysis set (randomised ITT) 205 (100) 97 (100) 302 (100)
Patients who received study treatment 205 (100) 91 (938) 296 (980)
Number who received olaparib 205 (100) - 205 (679)
Number who received capecitabine - 41 (423) 41 (136)
Number who received eribulin - 34 (351) 34 (113)
Number who received vinorelbine - 16 (165) 16 ( 53)
Safety
Robson et al N Engl J Med 2017 377523-533
21
17
18
18
12
21
7
15
22
50
23
15
26
35
1
9
11
14
16
16
17
20
21
27
29
30
40
58
0 25 755075 50 25Adverse events ()
NauseaAnemiaVomitingFatigueNeutropeniaDiarrheaHeadacheCough
Decreased appetitePyrexiaIncreased ALTIncreased ASTHand-foot syndrome
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Safety
Robson et al N Engl J Med 2017 377523-533
2
2
3
0
1
3
1
10
26
4
0
1
1
2
2
2
3
3
9
16Anemia
Fatigue
Neutropenia
Increased AST
Hand-foot syndrome
Dyspnea
Decreased platelet count
Leukopenia
0 25 755075 50 25
Headache
Adverse events ()
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Progression free survival
Robson et al N Engl J Med 2017 377523-533
17783
15946
10
09
08
07
06
05
04
03
02
01
00
Prob
abili
ty o
f pro
gres
sion
-free
surv
ival
Time from randomisation (months)
14121086420 16 18 20 22 24 26 28
234
408
6911
9421
10725
15444
20597
214
111
41
31
21
10
00
214
367
6111
7313
10024
12929
20188
111
111
31
21
10
10
OlaparibTPC
Number of patientrsquos at risk Median PFS was improved by 69 with olaparib treatment compared to standard of care
chemotherapy
Olaparib TPC
n 205 97
Events () 163 (795)
71(732)
Median (m) 70 42
HR = 058 95 CI (043 080)
p=00009
PFS free at 6m()
541 329
PFS free at 12m()
259 150
Olaparib 300 mg bd (N=205)TPC (N=97)
Response rate
Robson et al N Engl J Med 2017 377523-533
Olaparib TPC
Response Evaluable Population n 167 66
ORR n () 100 (599) 19 (288)
Complete Response n () 15 (90) 1 (15)
Partial Response n () 85 (510) 18 (273)
Median Duration of Response months (95CI) 64 (29-97) 71 (32-122)
Median Time to Onset of Responsedays 47 45
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Patient characteristics
Robson et al N Engl J Med 2017 377523-533
Olaparibn=205 n ()
TPCn=97 n ()
Totaln=302 n ()
Full analysis set (randomised ITT) 205 (100) 97 (100) 302 (100)
Patients who received study treatment 205 (100) 91 (938) 296 (980)
Number who received olaparib 205 (100) - 205 (679)
Number who received capecitabine - 41 (423) 41 (136)
Number who received eribulin - 34 (351) 34 (113)
Number who received vinorelbine - 16 (165) 16 ( 53)
Safety
Robson et al N Engl J Med 2017 377523-533
21
17
18
18
12
21
7
15
22
50
23
15
26
35
1
9
11
14
16
16
17
20
21
27
29
30
40
58
0 25 755075 50 25Adverse events ()
NauseaAnemiaVomitingFatigueNeutropeniaDiarrheaHeadacheCough
Decreased appetitePyrexiaIncreased ALTIncreased ASTHand-foot syndrome
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Safety
Robson et al N Engl J Med 2017 377523-533
2
2
3
0
1
3
1
10
26
4
0
1
1
2
2
2
3
3
9
16Anemia
Fatigue
Neutropenia
Increased AST
Hand-foot syndrome
Dyspnea
Decreased platelet count
Leukopenia
0 25 755075 50 25
Headache
Adverse events ()
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Progression free survival
Robson et al N Engl J Med 2017 377523-533
17783
15946
10
09
08
07
06
05
04
03
02
01
00
Prob
abili
ty o
f pro
gres
sion
-free
surv
ival
Time from randomisation (months)
14121086420 16 18 20 22 24 26 28
234
408
6911
9421
10725
15444
20597
214
111
41
31
21
10
00
214
367
6111
7313
10024
12929
20188
111
111
31
21
10
10
OlaparibTPC
Number of patientrsquos at risk Median PFS was improved by 69 with olaparib treatment compared to standard of care
chemotherapy
Olaparib TPC
n 205 97
Events () 163 (795)
71(732)
Median (m) 70 42
HR = 058 95 CI (043 080)
p=00009
PFS free at 6m()
541 329
PFS free at 12m()
259 150
Olaparib 300 mg bd (N=205)TPC (N=97)
Response rate
Robson et al N Engl J Med 2017 377523-533
Olaparib TPC
Response Evaluable Population n 167 66
ORR n () 100 (599) 19 (288)
Complete Response n () 15 (90) 1 (15)
Partial Response n () 85 (510) 18 (273)
Median Duration of Response months (95CI) 64 (29-97) 71 (32-122)
Median Time to Onset of Responsedays 47 45
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Safety
Robson et al N Engl J Med 2017 377523-533
21
17
18
18
12
21
7
15
22
50
23
15
26
35
1
9
11
14
16
16
17
20
21
27
29
30
40
58
0 25 755075 50 25Adverse events ()
NauseaAnemiaVomitingFatigueNeutropeniaDiarrheaHeadacheCough
Decreased appetitePyrexiaIncreased ALTIncreased ASTHand-foot syndrome
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Safety
Robson et al N Engl J Med 2017 377523-533
2
2
3
0
1
3
1
10
26
4
0
1
1
2
2
2
3
3
9
16Anemia
Fatigue
Neutropenia
Increased AST
Hand-foot syndrome
Dyspnea
Decreased platelet count
Leukopenia
0 25 755075 50 25
Headache
Adverse events ()
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Progression free survival
Robson et al N Engl J Med 2017 377523-533
17783
15946
10
09
08
07
06
05
04
03
02
01
00
Prob
abili
ty o
f pro
gres
sion
-free
surv
ival
Time from randomisation (months)
14121086420 16 18 20 22 24 26 28
234
408
6911
9421
10725
15444
20597
214
111
41
31
21
10
00
214
367
6111
7313
10024
12929
20188
111
111
31
21
10
10
OlaparibTPC
Number of patientrsquos at risk Median PFS was improved by 69 with olaparib treatment compared to standard of care
chemotherapy
Olaparib TPC
n 205 97
Events () 163 (795)
71(732)
Median (m) 70 42
HR = 058 95 CI (043 080)
p=00009
PFS free at 6m()
541 329
PFS free at 12m()
259 150
Olaparib 300 mg bd (N=205)TPC (N=97)
Response rate
Robson et al N Engl J Med 2017 377523-533
Olaparib TPC
Response Evaluable Population n 167 66
ORR n () 100 (599) 19 (288)
Complete Response n () 15 (90) 1 (15)
Partial Response n () 85 (510) 18 (273)
Median Duration of Response months (95CI) 64 (29-97) 71 (32-122)
Median Time to Onset of Responsedays 47 45
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Safety
Robson et al N Engl J Med 2017 377523-533
2
2
3
0
1
3
1
10
26
4
0
1
1
2
2
2
3
3
9
16Anemia
Fatigue
Neutropenia
Increased AST
Hand-foot syndrome
Dyspnea
Decreased platelet count
Leukopenia
0 25 755075 50 25
Headache
Adverse events ()
Olaparib 300 mg bd (N=205)Chemotherapy TPC (N=91)
Decreased white blood cells
Progression free survival
Robson et al N Engl J Med 2017 377523-533
17783
15946
10
09
08
07
06
05
04
03
02
01
00
Prob
abili
ty o
f pro
gres
sion
-free
surv
ival
Time from randomisation (months)
14121086420 16 18 20 22 24 26 28
234
408
6911
9421
10725
15444
20597
214
111
41
31
21
10
00
214
367
6111
7313
10024
12929
20188
111
111
31
21
10
10
OlaparibTPC
Number of patientrsquos at risk Median PFS was improved by 69 with olaparib treatment compared to standard of care
chemotherapy
Olaparib TPC
n 205 97
Events () 163 (795)
71(732)
Median (m) 70 42
HR = 058 95 CI (043 080)
p=00009
PFS free at 6m()
541 329
PFS free at 12m()
259 150
Olaparib 300 mg bd (N=205)TPC (N=97)
Response rate
Robson et al N Engl J Med 2017 377523-533
Olaparib TPC
Response Evaluable Population n 167 66
ORR n () 100 (599) 19 (288)
Complete Response n () 15 (90) 1 (15)
Partial Response n () 85 (510) 18 (273)
Median Duration of Response months (95CI) 64 (29-97) 71 (32-122)
Median Time to Onset of Responsedays 47 45
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Progression free survival
Robson et al N Engl J Med 2017 377523-533
17783
15946
10
09
08
07
06
05
04
03
02
01
00
Prob
abili
ty o
f pro
gres
sion
-free
surv
ival
Time from randomisation (months)
14121086420 16 18 20 22 24 26 28
234
408
6911
9421
10725
15444
20597
214
111
41
31
21
10
00
214
367
6111
7313
10024
12929
20188
111
111
31
21
10
10
OlaparibTPC
Number of patientrsquos at risk Median PFS was improved by 69 with olaparib treatment compared to standard of care
chemotherapy
Olaparib TPC
n 205 97
Events () 163 (795)
71(732)
Median (m) 70 42
HR = 058 95 CI (043 080)
p=00009
PFS free at 6m()
541 329
PFS free at 12m()
259 150
Olaparib 300 mg bd (N=205)TPC (N=97)
Response rate
Robson et al N Engl J Med 2017 377523-533
Olaparib TPC
Response Evaluable Population n 167 66
ORR n () 100 (599) 19 (288)
Complete Response n () 15 (90) 1 (15)
Partial Response n () 85 (510) 18 (273)
Median Duration of Response months (95CI) 64 (29-97) 71 (32-122)
Median Time to Onset of Responsedays 47 45
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Response rate
Robson et al N Engl J Med 2017 377523-533
Olaparib TPC
Response Evaluable Population n 167 66
ORR n () 100 (599) 19 (288)
Complete Response n () 15 (90) 1 (15)
Partial Response n () 85 (510) 18 (273)
Median Duration of Response months (95CI) 64 (29-97) 71 (32-122)
Median Time to Onset of Responsedays 47 45
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Summary of efficacy data
Robson et al N Engl J Med 2017 377523-533
ndash OlympiAD showed a statistically significant improvement in PFS (HR=058 (95 CI 043-080) p=00009) with a median PFS of 70 months in the olaparib arm compared to 42 months in the TPC arm
ndash Treatment with olaparib increased PF2 demonstrating a continued and consistent clinical benefit beyond progression with olaparib (HR 057 95 CI 040-083 p=00033)
ndash Median OS in the olaparib arm was 193 months compared to 171 months in the TPC arm (HR= 09 (95 CI 066-123)) the difference did not reach statistical significance p=051
ndash Patients treated with olaparib had a significantly better HRQoL
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Talazoparib
BRCA=breast cancer susceptibility gene CR=complete response ECOG=Eastern Cooperative Oncology Group HA=alternate hypothesis H0=null hypothesis ORR=objective response rate PR=partial response QD=once daily RECIST=Response Evaluation Criteria in Solid Tumors 3L=third-line
Turner NC et al Abstract 1007 ASCO 2017
Patients with advanced breast cancer with germline BRCAmutation
bull ECOG le 1bull Measurable
disease by RECIST v11
Cohort 1 (Prior Platinum)n=49
PR or CR to last platinum-containing regimen for metastatic disease with disease
progression gt8 weeks following the last dose of platinum
Cohort 2 (3L+ No Prior Platinum)n=35
3 or more prior cytotoxic regimens for metastatic disease
No prior platinum for metastatic disease
Phase 2 2-Stage 2-Cohort Study
Talazoparib
Primary EndpointbullConfirmed ORR by central independent radiology facility using RECIST v11Secondary Endpoints bullClinical benefit rate lasting ge24 weeksbullDuration of responsebullProgression-free survivalbullOverall survivalbullSafety
2-stage design for each cohort separatelybull Stage 1 enrollment of 35 patients in each
cohortbull If ge 5 responses seen in 1 cohort Stage 2
enrollment of 35 patientsbull Designed to compare HA 30 versus H0 15
response rate (90 power alpha=005)
Note enrollment discontinued at 84 pts
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Talazoparib
Turner NC et al Abstract 1007 ASCO 2017
Cohort 1 Cohort 2
BRCA1+BRCA2+Unknown
BRCA1+BRCA2+
Overall ORR for BRCA1=23 and for BRCA2=33
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
EMBRACA Study Design
Phase 3 international open-label study randomized 431 patients in 16 countries and 145 sites
Patients with locally advanced or metastatic HER2-negative breast cancer and a germline BRCA1 or BRCA2mutationdagger
Stratification factors
bull Number of prior chemo regimens (0 or ge1)
bull TNBC or hormone receptor positive (HR+)
bull History of CNS mets or no CNS mets
Talazoparib1 mg PO daily
Physicians choice of therapy (PCT)Dagger
capecitabineeribulin
gemcitabineor vinorelbine
R21
Additional inclusion criteria included no more than 3 prior cytotoxic chemotherapy regimens for locally advanced or metastatic disease prior treatment with a taxane andor anthracycline unless medically contraindicated daggerHER2-positive disease is excluded DaggerPhysicians choice of therapy must be determined prior to randomizationCNS=central nervous system EORTC=European Organisation for Research and Treatment of Cancer HER2=human epidermal growth factor receptor 2 mets=metastases PO= by mouth QLQ-BR23=Quality of Life Questionnaire breast cancer module QLQ-C30=Quality of Life Questionnaire Core 30 R=randomized RECIST=Response Evaluation Criteria In Solid Tumors version 11 TNBC=triple-negative breast cancer
Litton J et al N Engl J Med 2018 379753-763
Primary endpointbull Progression-free
survival by RECIST by blinded central review
Key secondary efficacy endpointsbull Overall survival (OS)bull Objective response rate
(ORR) by investigatorbull Safety
Exploratory endpointsbull Duration of response
(DOR) for objective responders
bull Quality of life (QoL EORTC QLQ-C30 QLQ-BR23)
Treatment (21-day cycles) continues until
progression or unacceptable toxicity
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
TALA(n=287)
Overall PCT(n=144)
Age median (range) y 45 (270-840) 50 (240-880)
lt50 y no 182 (634) 67 (465)
Gender female 986 979
ECOG = 0 1 2 530 440 20 580 400 10
Measurable disease by investigator no () 219 (763) 114 (792)
History of CNS metastasis no () 43 (150) 20 (139)
Visceral disease no () 200 (697) 103 (715)
Hormone receptor status no ()
TNBC 130 (453) 60 (417)
HR+ 157 (547) 84 (583)
BRCA status no ()
BRCA1+ 133 (463) 63 (438)
BRCA2+ 154 (537) 81 (563)
Disease free interval (initial diagnosis to aBC) lt12 months 108 (376) 42 (292)
Patient characteristics
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
TALA(n=287)
Overall PCT(n=144)
Patients no ()
Prior adjuvantneoadjuvant therapy 238 (829) 121 (840)
Prior hormonal therapy 161 (561) 77 (535)
Prior platinum therapy 46 (160) 30 (210)
No of prior cytotoxic regimens for aBC
0 111 (387) 54 (375)
1 107 (373) 54 (375)
2 57 (199) 28 (194)
ge3 12 (42) 8 (56)
aBC=advanced breast cancer PCT=physicianrsquos choice of therapy TALA=talazoparib
Patient characteristics
Litton J et al N Engl J Med 2018 379753-763
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade
Grade3
Grade 4
All Grade
Grade 3
Grade 4
No of patients with ge1 AE no ()
194 (678)
140
(490)
17
(59)
63 (500)
29
(230)
19
(151)
Anemia151
(528)110
(385)2 (07)
23 (183)
5 (40) 1 (08)
Neutropenia99
(346)51
(178)9 (31)
54 (429)
25 (198)
19 (151)
Thrombocytopenia77
(269)32
(112)10
(35)9 (71) 2 (16) 0
Lymphopenia21
(73)9 (31) 0 4 (32) 0 1 (08)
Febrile neutropenia 1 (03) 0 1 (03) 1 (08) 0 1 (08)
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Safety
Litton J et al N Engl J Med 2018 379753-763
TALA(n=286)
Overall PCT(n=126)
All Grade Grade 3 Grade 4 All
Grade Grade 3 Grade 4
No of patients with ge1 nonhematologic AE no ()
282 (986) 91 (318) 123
(976) 48 (381)
Fatigue 144 (503) 5 (17) 0 54 (429) 4 (32) 0
Nausea 139 (486) 1 (03) 0 59 (468) 2 (16) 0
Headache 93 (325) 5 (17) 0 28 (222) 1 (08) 0
Alopecia 72 (252) ‒ ‒ 35 (278) ‒ ‒
Vomiting 71 (248) 7 (24) 0 29 (230) 2 (16) 0
Diarrhea 63 (220) 2 (07) 0 33 (262) 7 (56) 0
Constipation 63 (220) 1 (03) 0 27 (214) 0 0
Decreased appetite 61 (213) 1 (03) 0 28 (222) 1 (08) 0
Back pain 60 (210) 7 (24) 0 20 (159) 2 (16) 0
Dyspnea 50 (17 5) 7 (24) 0 19
(15 1) 3 (24) 0
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Progression free survival
Litton J et al N Engl J Med 2018 379753-763
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Overall survival
Litton J et al N Engl J Med 2018 379753-763
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Response rate
Litton J et al N Engl J Med 2018 379753-763
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Summary of efficacy datandash Talazoparib resulted in prolonged progression-free survival
vs physicianrsquos choice of therapy by blinded central review HR 054 (95 CI 041 071) Plt00001
ndash Overall survival is immature (51 of projected events) HR 076 (95 CI 054 106) P=0105
ndash Global Health StatusQuality of Life showed overall improvement from baseline
ndash Talazoparib was generally well tolerated with minimal nonhematologic toxicity and few adverse events resulting in treatment discontinuation
Litton J et al N Engl J Med 2018 379753-763
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Cisplatin in basal like 1
32
RANDOMISE(11)
Docetaxel (D)100mgm2 q3w 6 cycles
Carboplatin (C)AUC 6 q3w 6 cycles
ER- PgR-unknown amp HER2- or known BRCA12Metastatic or recurrent locally advanced
Exclusions includebull Adjuvant taxane in le12 monthsbull Previous platinum treatmentbull Non-anthracyclines for MBC
A Priori subgroup analysesbull BRCA12 mutationbull Basal-like subgroups (PAM50 and IHC) bull Biomarkers of HRD
On progression crossover if appropriate
On progression crossover if appropriate
Carboplatin (C)AUC 6 q3w 6 cycles
Docetaxel (D)100mgm2 q3w 6 cyclesn-376
BRCA12 = 912
Tutt A et al 2104
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Cisplatin in basal like 1
59188 (314)
67188(356)
0 20 40 60 80 100
Carboplatin
Docetaxel
with OR at cycle 3 or 6 (95 CI)
33
Absolute difference (C-D)-42 (95 CI -137 to 53)
Exact p = 044
Randomised treatment - all
patients (N=376)
2192(228)
2390(256)
0 20 40 60 80 100
Carboplatin(Crossover=Docetaxel)
Docetaxel(Crossover=Carboplatin)
with OR at cycle 3 or 6 (95 CI)
Absolute difference (D-C)-28 (95 CI -152 to 96)
Exact p = 073
Crossover treatment - all patients (N=182)
Denominator excludes those with no first progression and those not starting crossover treatment
Tutt SABCS 2014
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Cisplatin in basal like 1
34
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
prog
ress
ion
free
Months from randomisation
Median PFSCarboplatin 31 mths (95 CI = 25 to 42)Docetaxel 45 mths (95 CI = 41 to 52)
Restricted mean survival to 15 mthsCarboplatin 48 mthsDocetaxel 52 mthsAbsolute difference
-04 (95 CI -11 to 03)p = 029
Carboplatin = 181 188
Docetaxel = 182 188
C 0188 9098 4056 3222 913 58 07
D 0188 57130 6069 4820 713 65 23
Number of eventsat risk
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Cisplatin in basal like 1
35
C 0188 23165 18141 24114 2289 1471 2244
D 0188 11176 20151 35110 1985 2358 1639
Number of eventsat risk
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts
aliv
e
Months from randomisation
Restricted mean survival to 15 mths
Carboplatin 107 mthsDocetaxel 108 mthsAbsolute difference
-02 (95 CI -11 to 08)p = 031
Median OSCarboplatin 124 mths(95 CI = 104 to 153)Docetaxel 123 mths
(95 CI = 105 to 136)
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Cisplatin in basal like 1
1725(680)
618(333)
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
36
Absolute difference (C-D)347 (95 CI 63 to 631)
Exact p = 003
Germline BRCA 12 Mutation (n=43)
Interaction randomised treatment amp BRCA 12 status p = 001
0 20 40 60 80 100
Carboplatin
Docetaxel
Percentage with OR at cycle 3 or 6 (95 CI)
No Germline BRCA 12
Mutation (n=273)
Absolute difference (C-D)-85 (95 CI -196 to 26)
Exact p = 016
36128(281)
53145(366)
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Cisplatin in basal like 1
0
10
20
30
40
50
60
70
80
90
100
0 3 6 9 12 15 18
p
atie
nts p
rogr
essi
on fr
ee
Months from randomisation
Carboplatin + BRCA12 mutated
Carboplatin + BRCA12 not mutated
Tutt SABCS 2014
Median PFSC + BRCA 12 mutated
68mnths (95 CI = 44 to 81)C + BRCA12 not mutated
31mnths (95 CI = 24 to 42)
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Basal like 2Growth factor signalling
Lisa A Carey et al JCO 2012302615-2623
The combination of cetuximab plusCarboplatin in metastatic TNBC producedresponses in fewer than 20 of patientsEGFR pathway analysis showed that mostTNBCs involved activation
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Subsets of triple-negative breast cancer
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Evidence from clinical trials
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Pembrolizumab in TNBC
bull PD-L1 positivity 58 of all patients screened had PD-L1-positive tumorsbull Treatment 10 mgkg IV Q2Wbull Response assessment Performed every 8 weeks per RECIST v11
aPD-L1 expression was assessed in archival tumor samples using a prototype IHC assay and the 22C3 antibody Only patients with PD-L1 staining in the stroma or in ge1 of tumor cells were eligible for enrollmentbIf clinically stable patients are permitted to remain on pembrolizumab until progressive disease is confirmed on a second scan performed ge4 weeks later If progressive disease is confirmed pembrolizumab is discontinued An exception may be granted for patients with clinical stability or improvement after consultation with the sponsor
bull Recurrent or metastatic ER-PR-HER2- breast cancer
bull ECOG PS 0-1bull PD-L1+ tumourbull No systemic steroid therapybull No autoimmune disease
(active or history of)bull No active brain metastases
CR Discontinuation permitted
PRSD Treat for 24 mo or until PD or toxicity
ConfirmedPD Discontinue
Pembro10 mgkg
Q2W
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Pembrolizumab in TNBCn =32
Confirmed complete response (nodal disease)Confirmed partial responseStable diseaseProgressive disease
Objective response rate 185Stable disease 259
Nanda SABCS 2015
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Pembrolizumab in TNBCCohort A (N = 170) Previously Treated
Regardless of PD-L1 Expression
Cohort B (N = 52) Previously Untreated
PD-L1 Positive
0
4
8
12
16
20
24
28
32
36
40
OR
R
47
0
4
8
12
16
20
24
28
32
36
40
OR
R
231
Complete response
Partial response
Adams S et al ASCO 2017
4847
Total PD-L1Positive
PD-L1Negative
Total(All PD-L1 Positive)
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Anti PD1 and anti PDL1 in TNBC
Schmid P et al AACR 2017 Adams S et al ASCO 2017
26
11
2L+1L
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
23
47
2L+1L
Pembrolizumab(n=222)
Atezolizumab(n=115)
CRPRCRPR
Keynote-086 Cohort B
Keynote-086 Cohort A
Anti-PD-L1PD-1 single agent in mTNBC ge1L PDL1+-
No clear relationship with PD-L1 positivity
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Role of TILs in TNBC
19
9
TIL high2
Obj
ectiv
e R
espo
nse
Rat
e (
)
10
20
30
0
64
19
Pembrolizumab(Cohort A gt2nd line)
Atezolizumab
TIL low
391
87
TIL lowTIL high
Pembrolizumab(Cohort B 1st line)
4
TIL high2TIL low
Schmid P et al AACR 2017 Adams S et al ASCO 2017 Loi ESMO 2017
Different levels by source of sample (archival vs new)
and organ site sampled LNgtlunggtliver
Metastatic breast cancer is a low TIL disease
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Best OverallResponse
1L(n = 9)
2L(n = 8)
3L+(n = 7)
All PatientsN = 24
Confirmed ORR(95 CI)a
667 (299 925)
25 (32 651)
286(37 710)
417(221 634)
ORR (95 CI)b
889(517 997)
750(349 968)
429(99 816)
708(489 874)
CR 111 0 0 42
PR 778 750 429 667
SD 111 250 286 208
PD 0 0 286 83
a Confirmed ORR defined as at least 2 consecutive assessments of complete or partial response b Including investigator-assessed unconfirmed responsesEfficacy-evaluable patients were dosed by June 1 2015 and were evaluable for response by RECIST v11 Minimum efficacy follow up was ge 3 months
Response rates were higher for patients who received atezolizumabnab-paclitaxel treatment as 1L therapy compared to 2L+
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Phase Ib Study of Atezolizumab and Nab-Paclitaxel in mTNBC
Adams S et al SABCS 2015 [abstract 850477]
Including investigator-assessed unconfirmed responses
bull 11 of 17 responses (65) continued on treatment at time of data cut off
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
bull Co-primary endpoints were PFS and OS in the ITT and PD-L1+ populationsd
ndash Key secondary efficacy endpoints (ORR and DOR) and safety were also evaluated
IMpassion130 study design
Schmid P et al IMpassion130 ESMO 2018 (abstract 2056)
IC tumour-infiltrating immune cell TFI treatment-free interval a ClinicalTrialsgov NCT02425891 b Locally evaluated per ASCOndashCollege of American Pathologists (CAP) guidelines c Centrally evaluated per VENTANA SP142 IHC assay (double blinded for PD-L1 status) d Radiological endpoints were investigator assessed (per RECIST v11)
Key IMpassion130 eligibility criteriaabull Metastatic or inoperable locally advanced
TNBC‒ Histologically documentedb
bull No prior therapy for advanced TNBC‒ Prior chemo in the curative setting
including taxanes allowed if TFI ge 12 mo
bull ECOG PS 0-1
Stratification factorsbull Prior taxane use (yes vs no)bull Liver metastases (yes vs no)bull PD-L1 status on IC (positive [ge 1] vs
negative [lt 1])c
Atezo + nab-P armAtezolizumab 840 mg IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Plac + nab-P armPlacebo IV
‒ On days 1 and 15 of 28-day cycle
+ Nab-paclitaxel 100 mgm2 IV‒ On days 1 8 and 15 of 28-
day cycle
Double blind no crossover permittedRECIST
v11 PD or toxicity
R11
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
DRUG EFFECT ON IMMUNE SYSTEM
Doxorubicin
Induces immunogenic cell deathIncreases proliferation of CD8 T cellsStimulates antigen presentation by DCsStimulates MCP1 and M6PR
Cyclophosphamide
Induces immunogenic cell deathSuppressed Treg inhibitory functions and restoration of the proliferative capacity of effector T cells and NK cell cytotoxicity
TaxanesEnhance T-cell and NK-cell functionIncrease recruitment of TILIncrease efficacy of immunostimulatory agents
GemcitabineReduce the number of myeloid suppressor cells Increase the antitumor activity of CD8(+) T cells and activated NK cells
Oxaliplatin Induces immunogenic cell deathIncreases MHC I complexInhibits PD-L2
Is nabpaclitaxel the ideal partner
DC dendritic cells MHC major histocompatibility complex NK natural killer
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
bull Primary PFS analysis (PFS tested in ITT and PD-L1+ populations)
bull First interim OS analysis (OS tested in ITT population then if significant in PD-L1+ population)
IMpassion130 statistical testing
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)a α recycled if PFSORR testing is significant Hazard ratio (HR)P valuendashstopping boundaries are dependent on the OS analysis timing
Atezo + nab-P vs Plac + nab-P
α = 005
PFS (primary)α = 001
OSa
bull Interimbull Primary (α ge
004)
OS in ITT population
OS in PD-L1+ population
1 PFS in ITT population
α = 0005
3 ORR in ITT population
α = 0001
4 ORR in PD-L1+ population
α = 0001
2 PFS in PD-L1+
populationα = 0005
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
IMpassion130 baseline characteristics
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
Data cutoff 17 April 2018 a Race was unknown in 12 patients in the Atezo + nab-P arm and 15 in the Plac + nab-P arm b Of n = 450 in each arm c ECOG PS before start of treatment was 2 in 1 patient per arm d Of n = 450 in the Atezo + nab-P arm and n = 449 in the Plac + nab-P arm arm
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Median age (range) y 55 (20-82) 56 (26-86)
Female n () 448 (99) 450 (100)
Race n ()a
White 308 (68) 301 (67)
Asian 85 (19) 76 (17)BlackAfrican
American 26 (6) 33 (7)
Othermultiple 20 (4) 26 (6)
ECOG PS n ()bc
0 256 (57) 270 (60)
1 193 (43) 179 (40)Prior (neo)adjuvant treatment n () 284 (63) 286 (63)
Prior taxane 231 (51) 230 (51)
Prior anthracycline 243 (54) 242 (54)
Characteristic
Atezo +nab-P
(N = 451)Plac + nab-P
(N = 451)Metastatic disease n () 404 (90) 408 (91)
No of sites n ()d
0-3 332 (74) 341 (76)
ge 4 118 (26) 108 (24)
Site of metastatic disease n ()
Lung 226 (50) 242 (54)
Bone 145 (32) 141 (31)
Liver 126 (28) 118 (26)
Brain 30 (7) 31 (7)
Lymph node onlyd 33 (7) 23 (5)
PD-L1+ (IC) n () 185 (41) 184 (41)
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
bull BRCA status unknownbull 43 previously untreatedbull PD-L1 on immune cells 41 positive
Patient population
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
bull 1 grade 5 AESI per arm (both treatment related)
ndash Atezo + nab-P autoimmune hepatitis
ndash Plac + nab-P hepatic failure
bull All hypothyroidism AESIs were grade 1-2 none led to discontinuation
ndash Atezo + nab-P 17 ndash Plac + nab-P 4
bull Pneumonitis was infrequent with only 1 grade 3-4 event in the Atezo + nab-P arm
ndash Atezo + nab-P 3ndash Plac + nab-P lt 1
bull Hepatitis rates were balanced
Safety
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)
AESI adverse event of special interest Data cutoff 17 April 2018 a Baskets of preferred terms according to medical concepts b All events of photophobia c Includes all AESIs occurring in ge 1 of patients in either arm
AESI n ()a
Atezo + nab-P (n = 452)
Plac + nab-P (n = 438)
Any Grade Grade 3-4 Any Grade Grade 3-4All 259 (57) 34 (8) 183 (42) 19 (4)Important AESIs
Hepatitis (all) 69 (15) 23 (5) 62 (14) 13 (3)Hepatitis (diagnosis) 10 (2) 6 (1) 7 (2) 1 (lt 1)Hepatitis (lab abnormalities) 62 (14) 17 (4) 58 (13) 12 (3)
Hypothyroidism 78 (17) 0 19 (4) 0Hyperthyroidism 20 (4) 1 (lt 1) 6 (1) 0Pneumonitis 14 (3) 1 (lt 1) 1 (lt 1) 0Meningoencephalitisb 5 (1) 0 2 (lt 1) 0Colitis 5 (1) 1 (lt 1) 3 (1) 1 (lt 1)Adrenal insufficiency 4 (1) 1 (lt 1) 0 0Pancreatitis 2 (lt 1) 1 (lt 1) 0 0Diabetes mellitus 1 (lt 1) 1 (lt 1) 2 (lt 1) 1 (lt 1)Nephritis 1 (lt 1) 0 0 0
Other AESIsc
Rash 154 (34) 4 (1) 114 (26) 2 (lt 1)Infusion-related
reactions 5 (1) 0 5 (1) 0
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Primary PFS analysis ITT population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)NE not estimable Data cutoff 17 April 2018 Median PFS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 months
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 451 360 226 164 77 34 20 11 6 1 NE NEPlac + nab-P 451 327 183 130 57 29 13 5 1 NE NE NE
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)PFS events
n 358 378
1-year PFS(95 CI)
24(20 28)
18(14 21)
72 mo(56 75)
55 mo(53 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 080(95 CI 069 092)
P = 00025
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Primary PFS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018
0 3 6 9 12 15 18 21 24 27 30 33Months
No at riskAtezo +
nab-P 185 146 104 75 38 19 10 6 2 1 NE NEPlac + nab-P 184 127 62 44 22 11 5 5 1 NE NE NE
75 mo(67 92)
50 mo(38 56)
100
80
60
40
20
0
Prog
ress
ion-
free
surv
ival
Stratified HR = 062(95 CI 049 078)
P lt 00001
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)PFS events
n 138 157
1-year PFS(95 CI)
29(22 36)
16(11 22)
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Interim OS analysis ITT populationa
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot Median follow-up (ITT) 129 monthsa For the interim OS analysis 59 of death events had occurred b Significance boundary was not crossed
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 451 426 389 337 271 146 82 48 26 15 6 NE NEPlac + nab-P 451 419 375 328 246 145 89 52 27 12 3 1 NE
213 mo(173 234)
176 mo(159 200)
100
80
60
40
20
0
Ove
rall
surv
ival
Stratified HR = 084(95 CI 069 102)
P = 00840b
Atezo + nab-P
(N = 451)Plac + nab-P
(N = 451)OS events
n 181 208
2-year OS(95 CI)
42(34 50)
40(33 46)
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Interim OS analysis PD-L1+ population
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Median OS durations (and 95 CI) are indicated on the plot a Not formally tested
250 mo(226 NE)
155 mo(131 194)
100
80
60
40
20
0
Ove
rall
surv
ival
0 3 6 9 12 15 18 21 24 27 30 33 36Months
No at riskAtezo +
nab-P 185 177 160 142 113 61 36 22 15 9 5 NE NEPlac + nab-P 184 170 147 129 89 44 27 19 13 6 NE NE NE
Stratified HR = 062 (95 CI 045 086)a
Atezo + nab-P
(n = 185)Plac + nab-P
(n = 184)OS events
n 64 88
2-year OS(95 CI)
54(42 65)
37(26 47)
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
aThe 31 figure is for all tumors the ORR was 37 in nonsquamous tumors and 12 in squamous cases In PDL-1 negative cases ORR was 14 in nonsquamoustumors and 0 in squamous tumors bThis study concerned squamous cell carcinomas only cThese authors also used the anti-PD-1 monoclonal M3 in their immunohistochemical analysis dIHC score 3 ge10 TIICs positive IHC score 2ndash3 ge5 TIICs positive IHC score 1ndash2ndash3 ge1 TIICs positive
PD-L1 Expression (ICH) As a Predictive Biomarker
Kerr KM et al J Thorac Oncol 201510(7)985-989 IHC immunohistochemistry Rx treatment TIICs tumor infiltrating immune cells
DrugBiomarkerAntibody
Rx Line
Definition of ldquoPositiverdquo ()
N Positive ()
Positive Predictive Outcome
ORR IHC Positive
Cases
ORR IHC Negative
Cases
Nivolumab Dako 28-8 1st ge5 in gt100 cells 59 Yes 31a 10
Nivolumab Dako 28-8 ge2nd ge5 ge1 49 56 No 15 13 14 17
Nivolumab + Ipilimumab Dako 28-8 1st ge5 in gt100 cells 42 No 19 14
Nivolumab Dako 28-8 ge2nd ge5 33b Yes 24 14
Nivolumab 5H1c ge2nd ge5 alsostudied TIICs 67 Yes No data for
lungNo data for
lung
Pembrolizumab Dako 22C3 Any ldquoStrongrdquo ge50
ldquoWeakrdquo 1-49 25 70 Yes Yes 37 17 9
Pembrolizumab Dako 22C3 1st ge5 ge1 Yes 47 26
Atezolizumab
Roche Ventana SP142 ge2nd ge10d ge5 ge1
TIICs 13 28 56 Yes 83 46 31 18 18 20
Durvalumab Roche Ventana SP263 ge2nd Data not
available 41 Yes 25 3
Dedicated
Different
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
bull Numerically higher and more durable responses were seen in the Atezo + nab-P arm
ndash Differences were not significant based on α level = 01 (ITT P = 00021 PD-L1+ P = 00016)
bull The CR rate was higher in the Atezo+ nab-P arm vs the Plac + nab-P arm
ndash ITT population 7 vs 2 ndash PD-L1+ patients 10 vs 1
Secondary efficacy endpoints
Schmid P et al IMpassion130
ESMO 2018 (abstract 2056)Data cutoff 17 April 2018 Objective responsendashevaluable patients a 450 in Atezo + nab-P arm and 449 in Plac + nab-P arm b 185 in Atezo + nab-P arm and 183 in Plac + nab-P arm c No death or PD
0
10
20
30
40
50
60
70
ITT A-nabPx
ITT P-nabPx
PD-L1+A-
nabPx
PD-L1+P-
nabPx
OR
R (
)ITTa PD-L1+b
56
46
59
43
49
44
49
42
12
7 10CR
PR
Atezo+ nab-
P
Plac + nab-P
Atezo + nab-P
Plac + nab-P
DOR median(95 CI) mo
74(69 90)
56(55 69)
85(73 97)
55(37 71)
No of ongoing responses n ()c
78 (31)
52 (25)
39 (36)
19 (24)
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
0
5
10
15
20
25
30
35
0
5
10
15
20
25
30
1st line 2nd line 1L Keynote 086Cohort B
2L Keynote 086Cohort A
Atezolizumab (n=115) Pembrolizumab (n=222)
Schmid P et al AACR 2017 Abstract 2986 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1088 Adams S et al J Clin Oncol 201735(Suppl 4)Abstract 1008 Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
26
11
23
5
Response to Immunotherapy AloneAnti-PD1PDL1 single agent in TNBC PDL1 +- TILs +-
TNBC Triple negative breast cancer
ORR
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
bull Data cutoff date Nov 10 2016
0
10
20
30
40
50
60
70
OR
R
6419
Cohort A Cohort B Combined Cohorts
0
10
20
30
40
50
60
70
OR
R
12617
0
10
20
30
40
50
60
70
OR
R
391
87
sTIL level
n
ge5
94
lt5
53
ge175
23
lt175
23
ge5
135
lt5
58
Responders 6 1 9 2 17 1
Ongoing responses
3 1 5 2 10 1
Loi S et al Ann Oncol 201728(Suppl 5)Abstract LBA13
ORR by sTIL Level geMedian vs ltMedian
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
bull A positive study for IO in TNBC Immunotherapy is ldquotransformativerdquo in TNBC
bull Subset data for PDL1 expression seem compelling which means we have to figure out how best to test tumors
bull Single agent vs combination therapy with chemotherapy in highly immunogenic selected patients best chemotherapy to be combined with IO
bull Endpoints should we be more focused on OS than on PFS
bull There is a ldquomissing armrdquo in the study atezolizumabalone Might that be a good option for certain subset
Conclusions
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Immunotherapy in TNBC
Pembrolizumab (Merck)Humanized IgG4 anti-
PD-1 antibody
Atezolizumab(Genentech)
engineered human IgG1 anti-PD-L1 antibody
Nivolumab(BMS)
Human IgG4 anti-PD-1 antibody
Durvalumab(AZ)
Human IgG1 anti-PD-L1 antibody
Tremelimumab(AZ)
Human IgG2Anti-CTLA-4 antibody
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Biop
sy
Biop
syBi
opsy
Biop
syBi
opsy
Biop
sy
Radiation3 x 8 Gy
Doxorubicin15 mg
x2
Cyclophos-phamide
50 mgdaily
Cisplatin 40 mgm2
x2
No treatment
Niv
olum
ab
2 weeks
R
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Adaptive Phase II Randomized Non-comparative Trial of Nivolumab After Induction Treatment in Triple-negative Breast Cancer (TNBC) Patients
TONIC-trial (The Netherlands Cancer Institute)
Kok M et al Ann Oncol 201735(suppl) Abstract LBA14
Total (n = 50)Best ORR (CR + PR) iRECIST 24CBR (CR + PR + SD) 26
CR 1 (2)PR 11 (22)SD ge24 weeks 1 (2)
ORR RECIST 11 22Median PFS [95 CI] 34 months [25-37]Median time to response [range] 21 months [05-35]Median duration of response [95 CI] 90 months [55-NA]
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
PARP inhibitors and IO
Patients with OC or TNBC Patients with OC (target n = 48) or TNBC (target n = 48) RP2D
Dose level 1Niraparib 200 mg + pembrolizumab 200 mg
Dose level 2Niraparib 300 mg + pembrolizumab 200 mg
Endpoint assessment
Endpoint assessment
Phase I Phase II
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
PARP inhibitors and IO
Konstantinopoulos PA et al Ann Oncol 201728(Suppl 5) Abstract 1143PD
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Phase 1b Study of docetaxel + PF-
03084014 in Triple-negative Breast Cancer
PF-03084014
Notch pathway
Notch pathway
M Locatelli et al Oncotarget 2016
Notch pathway
M Locatelli et al Oncotarget 2016
Clinical Heterogeneity of TNBC
Subtype Gene expression profile ClinicalBasal-like 1 high Ki-67 DNA damage response BRCA-associatedBasal-like 2 GF pathways Higher pCRImmunomodulatory Immune genesMesenchymal Cell motility Lower DDFSMesenchymal stem-like Cell motility claudin-lowLuminal androgen receptor Steroid pathways Apocrine features
higher LRF PI3KmutLehman BD et al J Clin Invest 20111212750-67
Luminal Androgen Receptor Bicalutamide
Bicalutamide 150mg dailyERPR(-)
(IHC le10)LABCMBC
AR+ DAKO Ab gt
10
bull Primary endpoint = CBR24 (CR + PR + SD gt 24 weeks)
Gucalp et al CCR 2013
Screened patients 12 AR+ (mostly TNBC)
Clinical Benefit Rate = 19 (95 CI 7-39)
All SD
Luminal Androgen Receptor Abiraterone
bull MBC ERPR le10 bull 138 screened 38 AR+
(ge10)
bull Primary Endpoint = CBR24
bull N = 30 evaluable patientsbull ~ 25 prior lines Rxbull ~ 50 visceral mets
bull Most common related AEs bull fatigue (18)bull HTN (12)bull hypokalemia (9)bull nausea (6)
Bonnefoi et al Ann Onc 2016
Median PFS 28m
CBR24 = 20 (95CI 8-39)
1 confirmed CR
Luminal Androgen Receptor Enzalutamide
Enzalutamide 160 mg dailyERPRHER2 (-) LABCMBC
AR+ Ventana
gt 0
bull Primary endpoint = CBR16 (CR + PR + SD gt 16) weeks)Screened patients 79 AR+ (55 by 10 cutoff)
bull Median 1 prior Rx
Evaluable (n=75 AR gt 10)CBR16 35 (24-46)CBR24 29 (20-41)RR 8SAE 29
0 8 16 24 33 41 49 6164
100
80
60
40
20
0
PFS
()
PFS 147 weeks95 CI 81 193
Luminal Androgen Receptor
64
PREDICT ARminus
0 8 16 24 32 40 52Time (weeks)
PREDICT ARminus PREDICT AR+Total n () 62 (53) 56 (47)CBR16
(95 CI)n
11(5 21)n = 7
39(27 53)n = 22
CBR24 (95 CI)n
6(2 16)n = 4
36(24 49)n = 20
CR or PR n
3n = 2
9n = 5
ActiveConfirmed CR or PR
0ndash1 Prior Lines2+ Prior Lines
PREDICT AR+
0 8 16 24 32 40 52 64Time (weeks)
Traina et al ASCO 2015
Gene expression classifier created = ldquoPredictARrdquo(Basal- apocrine+ etc to identify LAR)
Luminal Androgen Receptor
78
5662
5355
4946
4537
4227
4024
3213
156
116
32
Data cutoff 1Jul2015ITT = intent to treat mOS = median survival CI = confidence interval
Patients at riskPREDICT AR+PREDICT ARminus
0
80
40
20
n = 118
PREDICT ARminusmOS 323 weeks(95 CI 207 483)
PREDICT AR+mOS 756 weeks(95 CI 516 914)
0 8 16 24 33 41 49 61 64Weeks
100
60
Ove
rall
Surv
ival
()
85
ITT Population
NCT01889238
PREDICT AR+ mOS 180 monthsPREDICT AR ndash mOS 75 months
Courtesy of J Cortes ECCO 2015
Metastatic triple negative BC
Metastatic triple negative BC
bull Chemotherapy is still a standard for many patientsbull PARP inhibitors standard in BRCA carriers (will work
in somatically inactivated)bull AR inhibitors may be in very well selected
populationbull I-O is it transformative Eventually in more
immunogenic tumors (enrichment by PD-L1 and TILs)
Conclusions
bull Select the right partner and validate studies with the same backbone
bull Demonstrate bioactivity and not MTDbull Metastatic breast cancer is not always the right
settingbull Neoadjuvant and Post-neoadjuvant can be more
informative
Conclusions
Luminal Androgen Receptor Bicalutamide
Bicalutamide 150mg dailyERPR(-)
(IHC le10)LABCMBC
AR+ DAKO Ab gt
10
bull Primary endpoint = CBR24 (CR + PR + SD gt 24 weeks)
Gucalp et al CCR 2013
Screened patients 12 AR+ (mostly TNBC)
Clinical Benefit Rate = 19 (95 CI 7-39)
All SD
Luminal Androgen Receptor Abiraterone
bull MBC ERPR le10 bull 138 screened 38 AR+
(ge10)
bull Primary Endpoint = CBR24
bull N = 30 evaluable patientsbull ~ 25 prior lines Rxbull ~ 50 visceral mets
bull Most common related AEs bull fatigue (18)bull HTN (12)bull hypokalemia (9)bull nausea (6)
Bonnefoi et al Ann Onc 2016
Median PFS 28m
CBR24 = 20 (95CI 8-39)
1 confirmed CR
Luminal Androgen Receptor Enzalutamide
Enzalutamide 160 mg dailyERPRHER2 (-) LABCMBC
AR+ Ventana
gt 0
bull Primary endpoint = CBR16 (CR + PR + SD gt 16) weeks)Screened patients 79 AR+ (55 by 10 cutoff)
bull Median 1 prior Rx
Evaluable (n=75 AR gt 10)CBR16 35 (24-46)CBR24 29 (20-41)RR 8SAE 29
0 8 16 24 33 41 49 6164
100
80
60
40
20
0
PFS
()
PFS 147 weeks95 CI 81 193
Luminal Androgen Receptor
64
PREDICT ARminus
0 8 16 24 32 40 52Time (weeks)
PREDICT ARminus PREDICT AR+Total n () 62 (53) 56 (47)CBR16
(95 CI)n
11(5 21)n = 7
39(27 53)n = 22
CBR24 (95 CI)n
6(2 16)n = 4
36(24 49)n = 20
CR or PR n
3n = 2
9n = 5
ActiveConfirmed CR or PR
0ndash1 Prior Lines2+ Prior Lines
PREDICT AR+
0 8 16 24 32 40 52 64Time (weeks)
Traina et al ASCO 2015
Gene expression classifier created = ldquoPredictARrdquo(Basal- apocrine+ etc to identify LAR)
Luminal Androgen Receptor
78
5662
5355
4946
4537
4227
4024
3213
156
116
32
Data cutoff 1Jul2015ITT = intent to treat mOS = median survival CI = confidence interval
Patients at riskPREDICT AR+PREDICT ARminus
0
80
40
20
n = 118
PREDICT ARminusmOS 323 weeks(95 CI 207 483)
PREDICT AR+mOS 756 weeks(95 CI 516 914)
0 8 16 24 33 41 49 61 64Weeks
100
60
Ove
rall
Surv
ival
()
85
ITT Population
NCT01889238
PREDICT AR+ mOS 180 monthsPREDICT AR ndash mOS 75 months
Courtesy of J Cortes ECCO 2015
Metastatic triple negative BC
Metastatic triple negative BC
bull Chemotherapy is still a standard for many patientsbull PARP inhibitors standard in BRCA carriers (will work
in somatically inactivated)bull AR inhibitors may be in very well selected
populationbull I-O is it transformative Eventually in more
immunogenic tumors (enrichment by PD-L1 and TILs)
Conclusions
bull Select the right partner and validate studies with the same backbone
bull Demonstrate bioactivity and not MTDbull Metastatic breast cancer is not always the right
settingbull Neoadjuvant and Post-neoadjuvant can be more
informative
Conclusions
Luminal Androgen Receptor Abiraterone
bull MBC ERPR le10 bull 138 screened 38 AR+
(ge10)
bull Primary Endpoint = CBR24
bull N = 30 evaluable patientsbull ~ 25 prior lines Rxbull ~ 50 visceral mets
bull Most common related AEs bull fatigue (18)bull HTN (12)bull hypokalemia (9)bull nausea (6)
Bonnefoi et al Ann Onc 2016
Median PFS 28m
CBR24 = 20 (95CI 8-39)
1 confirmed CR
Luminal Androgen Receptor Enzalutamide
Enzalutamide 160 mg dailyERPRHER2 (-) LABCMBC
AR+ Ventana
gt 0
bull Primary endpoint = CBR16 (CR + PR + SD gt 16) weeks)Screened patients 79 AR+ (55 by 10 cutoff)
bull Median 1 prior Rx
Evaluable (n=75 AR gt 10)CBR16 35 (24-46)CBR24 29 (20-41)RR 8SAE 29
0 8 16 24 33 41 49 6164
100
80
60
40
20
0
PFS
()
PFS 147 weeks95 CI 81 193
Luminal Androgen Receptor
64
PREDICT ARminus
0 8 16 24 32 40 52Time (weeks)
PREDICT ARminus PREDICT AR+Total n () 62 (53) 56 (47)CBR16
(95 CI)n
11(5 21)n = 7
39(27 53)n = 22
CBR24 (95 CI)n
6(2 16)n = 4
36(24 49)n = 20
CR or PR n
3n = 2
9n = 5
ActiveConfirmed CR or PR
0ndash1 Prior Lines2+ Prior Lines
PREDICT AR+
0 8 16 24 32 40 52 64Time (weeks)
Traina et al ASCO 2015
Gene expression classifier created = ldquoPredictARrdquo(Basal- apocrine+ etc to identify LAR)
Luminal Androgen Receptor
78
5662
5355
4946
4537
4227
4024
3213
156
116
32
Data cutoff 1Jul2015ITT = intent to treat mOS = median survival CI = confidence interval
Patients at riskPREDICT AR+PREDICT ARminus
0
80
40
20
n = 118
PREDICT ARminusmOS 323 weeks(95 CI 207 483)
PREDICT AR+mOS 756 weeks(95 CI 516 914)
0 8 16 24 33 41 49 61 64Weeks
100
60
Ove
rall
Surv
ival
()
85
ITT Population
NCT01889238
PREDICT AR+ mOS 180 monthsPREDICT AR ndash mOS 75 months
Courtesy of J Cortes ECCO 2015
Metastatic triple negative BC
Metastatic triple negative BC
bull Chemotherapy is still a standard for many patientsbull PARP inhibitors standard in BRCA carriers (will work
in somatically inactivated)bull AR inhibitors may be in very well selected
populationbull I-O is it transformative Eventually in more
immunogenic tumors (enrichment by PD-L1 and TILs)
Conclusions
bull Select the right partner and validate studies with the same backbone
bull Demonstrate bioactivity and not MTDbull Metastatic breast cancer is not always the right
settingbull Neoadjuvant and Post-neoadjuvant can be more
informative
Conclusions
Luminal Androgen Receptor Enzalutamide
Enzalutamide 160 mg dailyERPRHER2 (-) LABCMBC
AR+ Ventana
gt 0
bull Primary endpoint = CBR16 (CR + PR + SD gt 16) weeks)Screened patients 79 AR+ (55 by 10 cutoff)
bull Median 1 prior Rx
Evaluable (n=75 AR gt 10)CBR16 35 (24-46)CBR24 29 (20-41)RR 8SAE 29
0 8 16 24 33 41 49 6164
100
80
60
40
20
0
PFS
()
PFS 147 weeks95 CI 81 193
Luminal Androgen Receptor
64
PREDICT ARminus
0 8 16 24 32 40 52Time (weeks)
PREDICT ARminus PREDICT AR+Total n () 62 (53) 56 (47)CBR16
(95 CI)n
11(5 21)n = 7
39(27 53)n = 22
CBR24 (95 CI)n
6(2 16)n = 4
36(24 49)n = 20
CR or PR n
3n = 2
9n = 5
ActiveConfirmed CR or PR
0ndash1 Prior Lines2+ Prior Lines
PREDICT AR+
0 8 16 24 32 40 52 64Time (weeks)
Traina et al ASCO 2015
Gene expression classifier created = ldquoPredictARrdquo(Basal- apocrine+ etc to identify LAR)
Luminal Androgen Receptor
78
5662
5355
4946
4537
4227
4024
3213
156
116
32
Data cutoff 1Jul2015ITT = intent to treat mOS = median survival CI = confidence interval
Patients at riskPREDICT AR+PREDICT ARminus
0
80
40
20
n = 118
PREDICT ARminusmOS 323 weeks(95 CI 207 483)
PREDICT AR+mOS 756 weeks(95 CI 516 914)
0 8 16 24 33 41 49 61 64Weeks
100
60
Ove
rall
Surv
ival
()
85
ITT Population
NCT01889238
PREDICT AR+ mOS 180 monthsPREDICT AR ndash mOS 75 months
Courtesy of J Cortes ECCO 2015
Metastatic triple negative BC
Metastatic triple negative BC
bull Chemotherapy is still a standard for many patientsbull PARP inhibitors standard in BRCA carriers (will work
in somatically inactivated)bull AR inhibitors may be in very well selected
populationbull I-O is it transformative Eventually in more
immunogenic tumors (enrichment by PD-L1 and TILs)
Conclusions
bull Select the right partner and validate studies with the same backbone
bull Demonstrate bioactivity and not MTDbull Metastatic breast cancer is not always the right
settingbull Neoadjuvant and Post-neoadjuvant can be more
informative
Conclusions
Luminal Androgen Receptor
64
PREDICT ARminus
0 8 16 24 32 40 52Time (weeks)
PREDICT ARminus PREDICT AR+Total n () 62 (53) 56 (47)CBR16
(95 CI)n
11(5 21)n = 7
39(27 53)n = 22
CBR24 (95 CI)n
6(2 16)n = 4
36(24 49)n = 20
CR or PR n
3n = 2
9n = 5
ActiveConfirmed CR or PR
0ndash1 Prior Lines2+ Prior Lines
PREDICT AR+
0 8 16 24 32 40 52 64Time (weeks)
Traina et al ASCO 2015
Gene expression classifier created = ldquoPredictARrdquo(Basal- apocrine+ etc to identify LAR)
Luminal Androgen Receptor
78
5662
5355
4946
4537
4227
4024
3213
156
116
32
Data cutoff 1Jul2015ITT = intent to treat mOS = median survival CI = confidence interval
Patients at riskPREDICT AR+PREDICT ARminus
0
80
40
20
n = 118
PREDICT ARminusmOS 323 weeks(95 CI 207 483)
PREDICT AR+mOS 756 weeks(95 CI 516 914)
0 8 16 24 33 41 49 61 64Weeks
100
60
Ove
rall
Surv
ival
()
85
ITT Population
NCT01889238
PREDICT AR+ mOS 180 monthsPREDICT AR ndash mOS 75 months
Courtesy of J Cortes ECCO 2015
Metastatic triple negative BC
Metastatic triple negative BC
bull Chemotherapy is still a standard for many patientsbull PARP inhibitors standard in BRCA carriers (will work
in somatically inactivated)bull AR inhibitors may be in very well selected
populationbull I-O is it transformative Eventually in more
immunogenic tumors (enrichment by PD-L1 and TILs)
Conclusions
bull Select the right partner and validate studies with the same backbone
bull Demonstrate bioactivity and not MTDbull Metastatic breast cancer is not always the right
settingbull Neoadjuvant and Post-neoadjuvant can be more
informative
Conclusions
Luminal Androgen Receptor
78
5662
5355
4946
4537
4227
4024
3213
156
116
32
Data cutoff 1Jul2015ITT = intent to treat mOS = median survival CI = confidence interval
Patients at riskPREDICT AR+PREDICT ARminus
0
80
40
20
n = 118
PREDICT ARminusmOS 323 weeks(95 CI 207 483)
PREDICT AR+mOS 756 weeks(95 CI 516 914)
0 8 16 24 33 41 49 61 64Weeks
100
60
Ove
rall
Surv
ival
()
85
ITT Population
NCT01889238
PREDICT AR+ mOS 180 monthsPREDICT AR ndash mOS 75 months
Courtesy of J Cortes ECCO 2015
Metastatic triple negative BC
Metastatic triple negative BC
bull Chemotherapy is still a standard for many patientsbull PARP inhibitors standard in BRCA carriers (will work
in somatically inactivated)bull AR inhibitors may be in very well selected
populationbull I-O is it transformative Eventually in more
immunogenic tumors (enrichment by PD-L1 and TILs)
Conclusions
bull Select the right partner and validate studies with the same backbone
bull Demonstrate bioactivity and not MTDbull Metastatic breast cancer is not always the right
settingbull Neoadjuvant and Post-neoadjuvant can be more
informative
Conclusions
Metastatic triple negative BC
Metastatic triple negative BC
bull Chemotherapy is still a standard for many patientsbull PARP inhibitors standard in BRCA carriers (will work
in somatically inactivated)bull AR inhibitors may be in very well selected
populationbull I-O is it transformative Eventually in more
immunogenic tumors (enrichment by PD-L1 and TILs)
Conclusions
bull Select the right partner and validate studies with the same backbone
bull Demonstrate bioactivity and not MTDbull Metastatic breast cancer is not always the right
settingbull Neoadjuvant and Post-neoadjuvant can be more
informative
Conclusions
Metastatic triple negative BC
bull Chemotherapy is still a standard for many patientsbull PARP inhibitors standard in BRCA carriers (will work
in somatically inactivated)bull AR inhibitors may be in very well selected
populationbull I-O is it transformative Eventually in more
immunogenic tumors (enrichment by PD-L1 and TILs)
Conclusions
bull Select the right partner and validate studies with the same backbone
bull Demonstrate bioactivity and not MTDbull Metastatic breast cancer is not always the right
settingbull Neoadjuvant and Post-neoadjuvant can be more
informative
Conclusions
bull Chemotherapy is still a standard for many patientsbull PARP inhibitors standard in BRCA carriers (will work
in somatically inactivated)bull AR inhibitors may be in very well selected
populationbull I-O is it transformative Eventually in more
immunogenic tumors (enrichment by PD-L1 and TILs)
Conclusions
bull Select the right partner and validate studies with the same backbone
bull Demonstrate bioactivity and not MTDbull Metastatic breast cancer is not always the right
settingbull Neoadjuvant and Post-neoadjuvant can be more
informative
Conclusions
bull Select the right partner and validate studies with the same backbone
bull Demonstrate bioactivity and not MTDbull Metastatic breast cancer is not always the right
settingbull Neoadjuvant and Post-neoadjuvant can be more
informative
Conclusions