41
Carsten Denkert German Breast Group and Institute of Pathology Charité Universitätsmedizin Berlin Berlin, Germany RAGMA, Madrid, 19.6.15 Recent advances in immunology and immunotherapy
Carsten Denkert German Breast Group and
Institute of Pathology Charité Universitätsmedizin Berlin
Berlin, Germany
RAGMA, Madrid, 19.6.15
Recent advances in immunology
and immunotherapy
Conflict of interest statement
• I am involved in approaches for standardized diagnostic testing – analytical validation of EndoPredict assay and development of the
ImmunoPredict assay (cofounder and shareholder of Sividon Diagnostics, Cologne)
– standardized image analysis software for immunohistochemistry, including TILs and Ki67 (collaboration with vmScope, Berlin)
Immunological concepts and clinical data the pathologist’s view
Clinical data
• What do we see under the microscope?
• What is the clinical relevance? – for chemotherapy response
– for new therapies
• Biomarkers for checkpoint inhibitor therapy
Immunological concepts
• immunoediting
• the cancer immunity cycle
• chemotherapy-induced immune activation
• mutation-based neoepitopes
The cancer immunoediting concept – 3 phases of cancer immune interaction
Elimination Equilibrium Escape
Dunn, Schreiber et al. Immunity 2004
The cancer immunoediting concept – 3 phases of cancer immune interaction
Elimination Equilibrium Escape
Tumor not clinically relevant, e.g.
some patients with Lynch Syndrome. Clinically relevant tumors
Dunn, Schreiber et al. Immunity 2004
Clinically relevant tumors in equlibrium/ escape
• Is there evidence of immune activation?
• Is is relevant for chemotherapy response and outcome?
• Do we have additional therapeutic options?
• Do we need new biomarkers?
Elimination Equilibrium Escape
The cancer immunoediting concept – 3 phases
Equilibrium phase: Heterogenous immune infiltrate in breast cancer
Lymphocyte predominant breast cancer (LPBC)
Spatial organisation of the tumor-associated immune system
Spatial organisation of the tumor-associated immune system
Tumor
TLS = tertiary
lymphoid
structure
Spatial organisation of the tumor-associated immune system
CD20: B-cells CD3: T-Cells CD21: FDCs Ki67: proliferation
TLS = tertiary
lymphoid
structure
Annals of Oncology, 2015
1% 5%
10% 20%
60% 70%
80% 90%
include area within tumor borders
do not include immune
infiltrate outside of the tumor
TLS
evaluate only TILs in this area
= stromal TILs
do not include TILs in this area
do not include
granulocytes in necrotic
areas
0-10% stromal TILs 20-40% stromal TILs 50-90% stromal TILs
For intermediate
group evaluate
different areas at higher
magnification.
The evaluation of tumor-infiltrating lymphocytes (TILs) in breast cancer: recommendations by an International TIL working group 2014
Lymphocyte-predominant breast cancer (LPBC) „more lymphocytes than tumor cells“ (≥60% TILs)
Denkert C, et al. J Clin. Oncol. 2010,J Clin. Oncol. 2015; SABCS 2013.
Patterns of tumor-infiltrating lymphoytes
intratumoral TILs – direkt contact with tumor cells
Denkert C, et al. J Clin. Oncol. 2010,J Clin. Oncol. 2015; SABCS 2013.
Patterns of tumor-infiltrating lymphoytes
stromal TILs – between the tumor cells
Denkert C, et al. J Clin. Oncol. 2010,J Clin. Oncol. 2015; SABCS 2013.
Zitvogel et al., Cell 2010
Molecular concept 1: Immunogenicity of chemotherapy – dying tumor cells release
antigens that activate the immune system (Laurence Zitvogel)
Chemotherapy
Molecular concept 2: Mutations and Neoepitopes are linked to immune activation
Association of a Neoepitope Signature with a Clinical Benefit from CTLA-4 Blockade in Melanoma
Snyder A et al. N Engl J Med 2014;371:2189-2199
Tumor-associated lymphocytes
Clinical relevance for chemotherapy response and prognosis
pC
R r
ate
(%)
LPBC: n=100 (12%); Multivariat iTuLy: p=0.001 , Denkert et al, JCO, 2010
0
5
10
15
20
25
30
35
40
nolymphocytes
focallymphocytes
Lymphocytepredominantbreast cancer
7,2
15,4
40
p<0.0005
GeparTrio – TILs are linked to response to neoadjuvant
chemotherapy (n=814)
TILs and chemotherapy
response in GeparSixto
n=580
Age
807 0605040302 0
iTu
Ly
6 0
4 0
2 0
0
strLy
1008 06 04 02 00
pCR
no pCR
ypT0_ypN0
Seite 1
intr
atu
mora
l T
ILs
stromal TILs
Her2-pos: Trastuzumab 6(8) mg/kg q3w (for 1 year) +
Lapatinib 750 mg/d 18 wks
TNBC: Bevacizumab 15 mg/kg q3w
Su
rgery
Geparsixto Design
Non-pegylated liposomal doxorubicin 20mg/m²q1w
Paclitaxel
80 mg/m² q1w Carboplatin AUC 1.5* q1w *reduced from AUC 2 at amendment 1 after enrollment of 330 patients
R
N=595
centrally
confirmed
TNBC
or
HER2+ BC
PM
PMCb
Primary clinical objective:
To compare the pCR (ypT0 ypN0)
rates between PM and PMCb.
Presented by: Carsten Denkert
pCR rates in GeparSixto: LPBC vs non-LPBC
San Antonio Breast Cancer Symposium - Cancer Therapy and Research Center at UT Health Science Center – December 10-14, 2013
pC
R r
ate
(%)
40% 37%
44%
34% 34% 34%
60%
45%
75%
0,00
10,00
20,00
30,00
40,00
50,00
60,00
70,00
80,00
both Tx groups StandardTx CarboplatinTx all patients PM - therapy PMCarbo - therapy
n= 580 438 142 290 217 73 290 221 69
Denkert et al, JCO 2015
P<0.0005
P<0.0005
P=0.09
test for interaction p=0.002
This presentation is the intellectual property of the author/presenter. Contact them at [email protected] for permission to reprint and/or distribute.
STEPP analysis – pCR rate in GeparSixto
0
10
20
30
40
50
60
70
80
90
0 50
0
10
20
30
40
50
60
70
80
90
0 50
PM
% stromal TILs % stromal TILs
pC
R r
ate
(%
)
pC
R r
ate
(%
)
all patients PM vs PMC therapy PMC
Denkert et al. JCO 2015
TNBC – increased lymphocytic infiltrate defines a good prognosis group
BIG2-98
(TNBC n=256)
Loi et al, JCO 2013
ECOG E2197+E1199
(TNBC n=481)
Adams et al, JCO 2014
2 French trials
(TNBC n=199)
Dieci et al. Ann Oncol 2015
Tumor-associated lymphocytes – options for
clinical utility
• Conclusions for clinical practice – immune signals are strong and easily detectable – but there is no clear clinical utility so far
• Option 1: neoadjuvant carboplatin in TNBC – high complete response rates in GeparSixto with increased TILs – might be an additional factor for therapy decisions – validation in other Platin trials needed – GeparOcto: dose-dense conventional vs. dose-dense carboplatin
• Option 2: HER2 positive BC – trastuzumab effect dependent on TILs (Finher) – not observed in N9831 (Perez, SABCS 2015) – other validations pending, e.g. BCIRG-006
• Option 3: immune therapies ... prediction of response
Melanoma Triple-negative breast cancer
Concepts for immunotherapy in breast cancer – can we learn from melanoma?
Chen, Mellman, Immunology 2013
Anti-PD-L1
Anti-PD-1
The cancer immunity cycle - therapeutic options
Immunmodulation with PD1 and PDL1 Blockade (immune checkpoint inhibitors)
Wolchok, ASCO 2014
Nivolumab (anti-PD1) and Ipilimumab (anti-CTLA-4) in malignant Melanoma
Nanda, SABCS 2014
anti-PD1 antibody
Emens LA, et al. SABCS, 2014.
30
One patient who discontinued the study without any
post-baseline tumor assessment is not included.
Emens et al, SABCS 2014
anti-PD-L1
Molecular signatures of immune infiltrate
morphological classification molecular characterization
Hypothesis:
Lymphocyte-predominant breast cancer (LPBC)
non-LPBC
Immunosuppressive
regulators:
PD1, PDL1,
CTLA4, IDO1, FOXP3
Immune activation:
T-Cells: CD8A, CCL5
B-Cells: IGKC, CD21, CD80
Chemoattractants:
CXCL9, CXCL13
Denkert et al, JCO 2015
1
19
50
0
10
20
30
40
50
60
Immune-A Immune-B Immune-C
p<0.00001
PDL1
IDO1
PD1
CTLA4
CXCL9
CD8A
CCL5
CXCL13
IGKC
CD21
FOXP3
CD80
HR+
HR-
strLy (%)
LPBC
Immune-type A C B
Presented by: Carsten Denkert
Type A Type C
Three different immune subtypes: correlation with TIL morphology
LP
BC
ca
se
s (
%)
24
37
56
0
10
20
30
40
50
60
Immune-A Immune-B Immune-C
pC
R r
ate
(%
)
p<0.00001
PDL1
IDO1
PD1
CTLA4
CXCL9
CD8A
CCL5
CXCL13
IGKC
CD21
FOXP3
CD80
HR+
HR-
strLy (%)
LPBC
Immune-type A C B
Presented by: Carsten Denkert
Three different immune subtypes: correlation with response rate
Immune markers were significantly linked to increased pCR rates –
all cases (n=481)
0
2
4
6
8
10
12
14
0,6 0,8 1,0 1,2 1,4 1,6 1,8 2,0 2,2 2,4
Stromal TILs: OR per 10% change, mRNA markers: OR per 1 dCt value (l≈ doubling of mRNA)
stromal TILs 1.26 (1.16-1.36) 0.00000001 0.0000005 0.007
CCL5 1.41 (1.23-1.62) 0.000001 0.00001 0.002
CXCL9 1.25 (1.14-1.38) 0.000006 0.003 0.09
CXCL13 1.16 (1.06-1.26) 0.001 0.006 ns
CD8A 1.29 (1.13-1.48) 0.0002 0.001 0.01
PD1 1.43 (1.24-1.66) 0.000001 0.00002 0.02
PDL1 1.57 (1.34-1.86) 0.00000003 0.000001 0.09
CTLA4 1.38 (1.19-1.60) 0.00001 0.0001 0.06
FOXP3 1.23 (1.003-1.50) 0.05 0.02 ns
IDO1 1.25 (1.14-1.36) 0.0000005 0.00003 0.03
IGKC 1.15 (1.06-1.24) 0.0004 0.002 ns
CD80 1.59 (1.26-2.01) 0.0001 0.0002 ns
CD21 1.11 (1.02-1.21) 0.01 ns ns
less pCR more pCR
univariate multivariate interacti. w. CbTx
OR (95% CI) p-value p-value p-value
Presented by: Carsten Denkert
Immune markers were significantly linked to increased pCR rates –
all cases (n=481)
0
2
4
6
8
10
12
14
0,6 0,8 1,0 1,2 1,4 1,6 1,8 2,0 2,2 2,4
Stromal TILs: OR per 10% change, mRNA markers: OR per 1 dCt value (l≈ doubling of mRNA)
stromal TILs 1.26 (1.16-1.36) 0.00000001 0.0000005 0.007
CCL5 1.41 (1.23-1.62) 0.000001 0.00001 0.002
CXCL9 1.25 (1.14-1.38) 0.000006 0.003 0.09
CXCL13 1.16 (1.06-1.26) 0.001 0.006 ns
CD8A 1.29 (1.13-1.48) 0.0002 0.001 0.01
PD1 1.43 (1.24-1.66) 0.000001 0.00002 0.02
PDL1 1.57 (1.34-1.86) 0.00000003 0.000001 0.09
CTLA4 1.38 (1.19-1.60) 0.00001 0.0001 0.06
FOXP3 1.23 (1.003-1.50) 0.05 0.02 ns
IDO1 1.25 (1.14-1.36) 0.0000005 0.00003 0.03
IGKC 1.15 (1.06-1.24) 0.0004 0.002 ns
CD80 1.59 (1.26-2.01) 0.0001 0.0002 ns
CD21 1.11 (1.02-1.21) 0.01 ns ns
less pCR more pCR
univariate multivariate interacti. w. CbTx
OR (95% CI) p-value p-value p-value
Presented by: Carsten Denkert
immune
suppressive
markers
Immunomodulatory subgroup of TNBC Lehmann et al. JCI, 2011
The ImmunoPredict approach for evaluation of
immune activation in breast cancer - strong immune reaction, but tumor still growing
- good prognosis with chemotherapy
immunogenic effects of chemotherapy present
can they be enhanced by checkpoint inhibition?
- no evidence of
immune
activation
- immune therapy
approaches not
useful?
- partial immune activation
- immune heterogeneity
- immune escape?
- enhancement of
response by immune
therapy / checkpoint
inhibition?
Denkert et al, JCO 2015
International TIL ring trial – 32 pathologists – ongoing June 2015
Take-home message for tumor diagnosis
• pathologists should include TILs in pathology reports – stromal TILs are the best parameter
– use standardized methodology, e.g. Salgado et al., Ann. Oncol. 2015
• TILs are a predictive marker for response to neoadjuvant therapy – several studies with >2000 patients.
• TILs are prognostic in TNBC (n>900, three studies).
• currently no basis for therapy decisions
– validation needed: Carboplatin therapy and HER2 therapy
Take-home message for tumor therapy and clinical studies
• there is evidence that a subgroup of breast cancer is immunogenic
• phase 1 trials are suggesting clinical activity for immune checkpoint inhibitors in breast cancer
• combinations of immune checkpoint inhibitors and conventional chemotherapy should be considered
• well-designed biomarker studies are needed in breast cancer immune trials
• possible biomarkers include: – TILs – the therapeutic targets: PD1, PDL1 – immune gene expression signatures, e.g. ImmunoPredict – mutational load and neoepitope signatures
Charité
Britta Beyer
Jan Budczies
Silvia Darb-Esfahani
Sylwia Handzik
Frederick Klauschen
Ines Koch
Berit Pfitzner
Judith Prinzler
Bruno Sinn
Wolfgang Schmitt
Petra Wachs
Stephan Wienert
Manfred Dietel
GBG
Gunter von Minckwitz
Sibylle Loibl
Valentina Nekljudova
Keyur Mehta
Stephan Gade
Christiane Rothhaar
Translational Subboard of GBG
Neoadjuvant Subboard of GBG
RESPONSIFY partners
Sherene Loi
Christos Sotiriou
Fabrice André
We would like to thank all patients, clinicians, and pathologists participating in
the clinical studies and the biomaterial collection.
EU FP7 No 278659
