Studies in NSCLC Based on Gender and Smoking

Differences: Rationale and Outcomes

Silvia Novello University of Turin Thoracic Oncology Unit

www.oncologiapolmonare.it silvia.novello@unito.it

www.womenagainstlungcancer.eusilvia.novello@womenagainstlungcancersilvia.novello@womenagainstlungcancer.eu

Males

Rat

e p

er 1

00,0

00 M

ales

Year of Death

Rat

e p

er 1

00,0

00 F

emal

es

Year of Death1930 1950 1970 1990

Lung & BronchusStomachColon & RectumProstateLiver

1930 1950 1970 1990 2004

LeukemiaPancreas

Females

Annual Age-Adjusted Cancer Death Rates Among Males/Females

for Selected Cancer Types, US, 1930-2004

Colon & RectumStomachOvaryUterusBreastLung & BronchusPancreas

CA Cancer J Clin 2004; 54:9-15

20040

20

40

60

80

100

0

20

40

60

80

100

Lung & Bronchus

Lung & Bronchus

Lopez et al, Tobacco Control 1994

Smoke kills about 50% of smokers: there are 3-4 Smoke kills about 50% of smokers: there are 3-4 decades between prevalence peak and mortality decades between prevalence peak and mortality peak for lung cancerpeak for lung cancer

Model of Smoking Epidemic

Lifetime Probability of Developing Cancer

1997-2001

All sites 1 in 3

Breast 1 in 7

Lung & bronchus 1 in 18

All sites 1 in 2

Prostate 1 in 6

Lung & bronchus1 in 13

Site Risk Site Risk

ACS, 2005

11,3

14,6

9,9

25,5

3,9 3,4

6,78,1

0

10

20

30

Squamouscell

SCLC Adeno-carcinoma

Large cell

Men

Women

Rela

tive R

isk R

ati

oR

ela

tive R

isk R

ati

o

HistologyHistology

Khuder, Lung Cancer, 2001

Meta-Analysis of 28 Lung Cancer Studies*

Relative Risk Ratios for Ever Smokers

Meta-Analysis of 28 Lung Cancer Studies*

Relative Risk Ratios for Ever Smokers

* 27 case-control

Women and Lung Cancer Risk (smokers vs non-smokers)

Author Study Men Women n.cigarettesRisch case- 9.6 27.9 40 packs/yrAm J Epidemiol ’93 control

Zhang case- 11.6 21.4 40 packs/yrJ Natl Cancer Inst ’96 control

Harris case- 24.5 42 40 packs/yrInt J Epidemiol ’93 control

Bach cohort no differenceJ Natl Cancer Inst ’03

Bain cohort no difference J Natl Cancer Inst ‘04

• Female smokers are more likely to develop adenocarcinoma than squamous cell carcinoma Thun MJ et al:JNCI 1997; Fu JB et al: Chest 2005; Patel JD et al: JCO 2005

• Never-smokers with lung cancer have adenocarcinoma and are 2.5 times more likely to be females than males

Wong MP et al: Cancer 2003

• The BAC is two to four times more common among women than among men Thun MJ et al:JNCI 1997; Radzikowska E et al: Ann Oncol 2002; Fu JB et al: Chest 2005

• In some Asian countries never-smokers account for 70% of women with lung cancer Wong MP et al: Cancer 2003

Women and Lung Cancer

• DNA repair capacity

• Gender differences in metabolism of carcinogens

• Differences in proliferation/growth stimulation (GRPR)

• Hormonal interactions

• DNA repair capacity

• Gender differences in metabolism of carcinogens

• Differences in proliferation/growth stimulation (GRPR)

• Hormonal interactions

Some Suggested ExplanationsSome Suggested Explanations

DNA Repair Capacity

Host Cell Reactivation Assay

Benzo[a]pyrene

TestLymphocytes

Acetyl -Chloramphenicol

+ .Acetyl CoA

+ Chloramphenicol

+ .Acetyl CoA

+ Chloramphenicol

DRC and Risk of Lung Cancer

DRC (median)> 8.1 1.0 < 8.1 1.5 (1.2,

1.9)Smoking Status

Never 1.8 (1.0, 3.3) Former 1.4 (1.0, 1.9) Current 1.6 (1.2, 2.3)

DRC (median)> 8.1 1.0 < 8.1 1.5 (1.2,

1.9)Smoking Status

Never 1.8 (1.0, 3.3) Former 1.4 (1.0, 1.9) Current 1.6 (1.2, 2.3)

Variable Adjusted OR(95% CI) Variable Adjusted OR(95% CI)

Spitz et al, CEBP, 2003

n = 764 cases and 677 controls

Age (years) < 60 128 7.6 ± 2.8

60 - 69 123 8.2 ± 3.2 0.05 70 65 8.5 ± 3.2

GenderMale 402 8.2 ± 2.8 < 0.001

Female 362 7.5 ± 2.8

Age (years) < 60 128 7.6 ± 2.8

60 - 69 123 8.2 ± 3.2 0.05 70 65 8.5 ± 3.2

GenderMale 402 8.2 ± 2.8 < 0.001

Female 362 7.5 ± 2.8

Variable No Mean % p-valueSD for trend

Variable No Mean % p-valueSD for trend

DNA Repair CapacityLung Cancer Cases

Spitz, CEBP, 2003

Age (yrs)

< 60 109 100.8 ± 94.6 0.05

61+ 112 85.8 ± 83.6

Gender

Male 114 100.1 ± 89.1 NSFemale 107 85.8 ± 89.4

Smoking status Never 21 103.7 ±101.6 NS Former 8587.1 ± 75.0 Current 11595.8 ± 97.0

Age (yrs)

< 60 109 100.8 ± 94.6 0.05

61+ 112 85.8 ± 83.6

Gender

Male 114 100.1 ± 89.1 NSFemale 107 85.8 ± 89.4

Smoking status Never 21 103.7 ±101.6 NS Former 8587.1 ± 75.0 Current 11595.8 ± 97.0

Variable N Mean(%)SD P valueVariable N Mean(%)SD P value

Induced Adduct LevelsLung Cancer Cases

Li et al, Cancer Res 2001Li et al, Cancer Res 2001

Gender differences in metabolism of carcinogens

CYP1A1• CYPA1 codes for an enzyme which activates PAH-

forming DNA adducts. • Significant correlation between CYP1A1 expression

and DNA adduct levels (r= 0.50, p= 0.016)• Female smokers had significantly higher levels of

adducts/pack-year and adducts/cigarette/day than men (1.49 + 1.29 vs. 0.89 + 0.74, P= 0.015)

• Females had higher CYP1A1 levels than males (494 + 334 units vs. 210 + 208 units, P= 0.016)

Mollerup, et al; Cancer Res; 1999: 59: 3317-20

Gender differences in metabolism of carcinogens

Glutathione S-transferase (GST)

Odds ratio for GSTM1 null phenotype and WT

Female (CI)

Male (CI)

Lung cancer2.50

(1.09-5.72)1.40

(0.58 - 3.38)

Lung cancer in smokers

3.03 (1.09 -8.40)

1.42 (.053-4.06)

•GST deactivates carcinogens; the null genotype fails to deactivate carcinogens, resulting in prolonged exposure•GSTM1 null genotype associated with lung cancer (odds ratio: 2.04)

Tang, et al; Carcinogenesis, 19: 1949-1953, 1998

• Gastrin-Releasing Peptide (GRP): plays a role in neoplasia by stimulating cell proliferation. Its effect is mediated mainlythrough the GRPR.

• The gene for GRPR is X-linked, located on chromosome Xp22, near a cluster of genes that escape X-inactivation.

• Women can have two actively transcribed alleles compared with only one in men

• Increased expression of the GRPR gene was noted when human airway cells were exposed to oestrogens.

Differences in proliferation/growth stimulationGRP

Shriver SP 2000, JNCI

Differences in proliferation/growth stimulation

EGFR

CharacteristicsPatients with

mutation/total

In unselected pts

Adenocarcinoma

14/182 (8%)

15/152 (10%)

Women with adenocarcinoma

12/83 (15%)

Women non-smokers with adenocarcinoma

18/34 (53%)

ASCO 2004

Hormonal Interactions

• Early age at menopause (≤ 40 yrs) is associated with reduced risk of lung adenocarcinoma (OR=0.3)

• HRT is associated with risk of adenocarcinoma (OR=1.7)

• Interaction between HRT, smoking and the development of lung adenocarcinoma (OR=32.4)

Taioli and Wynder:JNCI 1994

• Late menopause and short menstrual cycles were associated with increased risk of lung cancer

Siegfried JM: The Lancet Oncology 2001

risk of lung carcinoma in women with family history of reproductive cancer

Sellers: Genetic Epidem 1991

β-estradiol inducesproliferation in NSCLC cells and anti-estrogens block this effect

Kiuper, Endocrinology 1997

Oestrogens may be involved in lung tumorigenesis at different levels:• As ER ligands activating cell proliferation

• Via ERs in the plasma membrane causing interactions between ERs and growth factors such as EGF and IGF (in EGFR and ER+ cells)

• Oestrogens may alter metabolic activation of carcinogens (modulations of CY1A1, CY1B1) Stabile: Cancer Res 2005

Hormonal Interactions

Cross-Talk Cascade of ER Activation IGFR

PlasmaMembrane

Cytoplasm

Akt

SOS

P

P

P

P

MAPK

MEK

Nucleus

p90RSK

p160ERER CBP

BasalTranscription

Machinery

ERE

P P P

ER Target Gene Transcription

P

P

P

P

RAS

RAFPI3-K

PP

P

Growth factors

Estrogen

ER

EGFR / HER2

AI

MoAb

TKI

FTI

CCI

CellGrowth

Cell Survival

Tamoxifen

SERD

Johnston, S. 2004

Decreased Cell Proliferation in Lung Tumors Treated with Gefitinib and

Fulvestrant

Rela

tive K

i67 E

xp

ressio

n

0

20

Treatments

control fulvestrant fulvestrant +gefitinib

gefitinib

40

60

80

100

120

*

**

**

Stabile, et al. Cancer Res, 2005

P-value compared to control: *<0.05, **<0.005

UW/UPitt Pilot Study of Gefitinib + Faslodex in Post-menopausal Women with Advanced

Recurrent NSCLC

• Eligibility: 2 or more prior chemo regimens• Treatment: 250 mg gefitinib + 250 mg

Faslodex IM monthly• Objectives: response rate, TTP, survival• Laboratory Objectives:

- ER and EGFr- CYP3A polymorphisms

Traynor AM, Schiller J, JCO 2005

Hormone Replacement Therapy and Lung Cancer Risk

Some positive

Type of Study

NRisk of lung Ca

with HRT95% CI

Taioli, JNCI, 1994

Case control

180 1.7 1.0-2.8

•All studies derived from secondary data or exploratory analysis

Hormone Replacement Therapy and Lung Cancer Risk

Some show no increase in risk

Type of Study NRisk of lung Ca with

HRT95% CI

Adams,Int J Cancer, 1989

Cohort study 23,244 1.3 0.9-1.7

Wu, Cancer Res, 1998

Case control 336 1.3 0.71-1.53

Women's Health Initiative, JAMA, 2002

Cohort 16,690 1.04 0.71-1.53

Blackman, Pharmacoepidemiol Drug Saf, 2002

Case Control 662 1.0

All studies derived from secondary data or exploratory analysis

Hormone Replacement Therapy and Lung Cancer Risk

Some show REDUCED risk of lung cancer with HRT

Type of study

NRisk of lung Ca

with HRT95% CI

Ettinger, Ob Gynecol, 1996

454 0.78 0.04-1.15

Krenzer, 1993

Case control

1723 0.83 0.64-1.09

Olson, Ob Gyne, 1993

Cohort 29,508 0.24 0.08-0.76

Schabath, Clin Ca Res, 2004

Case Control

1008 0.66 0.51 - 0.89

All studies derived from secondary data or exploratory analysis

Sex as a predictive factor

Study N Unfavorable in

Multivariate Analysis

Radzikowska (1995-98) 20,561 Male, poor PS, advancedPolish population, all stage, non surgical

treatment, >50 yrs, SCLC

Ouellete (1988-90) 208 Male, advanced stageFrench population, cohort

Moore (1974-98) 7,613 Male, age >65, advancedSingle institution, all stage, large cell, no surgery

Visbal (1997-2000) 4,618 Male, older age, high grade,Single institution, advanced stage, treatment,consecutive cohort adenocarcinoma**smoking status/dose, comorbidy interaction with cause of death, not significant

Relative risk (RR) of death for MEN = 1.15 (p=0.001)

NO differences in overall mean survival

BUT women lived longer at each stageOverall median survival 12.4 mo vs 10.3mo (p<.001) and survival advantage for all stages

RR for MEN= 1.2

NSCLC: completely resected diseases

NSCLC: completely resected diseases

Author # Women Survival

MinamiR (#1242 consec.1984-1998)

[adavntage also for st I &III,adenoca & > 60yrs]

FergusonR (#451)

1980-1998

AlexiouP (#833)

UK 1990-2000

337

186

252

5yr 69% p<.005

st I OS 109vs50 mo p .008

5yr 48%vs36.5% p<0.01

OS p 0.0006

NSCLC: completely resected diseases

Bouchardy, et al. Cancer, 1999

EVEN CONSIDERING OTHER DEATH

CAUSES

2.1

ALPI Multivariate AnalysisALPI Multivariate Analysis

Variable HR 95% CI P Value

Age (5-yr interval)

1.06 1.00-1.12

.051

Stage II III

2.003.15

1.6-2.52.6-3.9

.0001

.0001

Histology-Squamous

0.86 0.7-1.0 .094

Gender – Male(86%)

1.32 1.0-1.7 .034

Complete Dissection

0.89 0.8-1.1 .164

Chemotherapy 0.96 0.8-1.1 .566

Scagliotti GVS, JNCI 2003Similar Data from UFT, Kato et al.

NSCLC: locally advanced disease

Werner WasikR (RTOG)- #1,999- 9 studies

OS 11.4 vs 9.9 mo

AlbainP (SWOG)- #126- st IIIA, IIIB

OS 21 vs 12 mo (p0.08)

AlbainP (RTOG 93-09)-#429 (35%)- st IIIA

> OS in women(p=0.051)

NSCLC: advanced disease*

Favourable factors inGroup N Stage multivariate analysis

MemorialR 378 IIIB/IV Women, normal LDH ,(O’Connell et al) good PS, no bone mts, ≤ 2 sites mts

ECOGR 893 IV Women, PS 0, no bone mts,

(Finkelstein et al) no liver mts,

7 trialsno weight loss,

non-large cell istol.

SWOGR 2,290 IV Women, PS 0-1, (Albain et al) “cisplatin-based”

therapy,13 trials age ≤ 70 yrs

male female P Value

MULTIVARIATE

median survival 8.8 mo 12.4 mo 0.001

*first and second generation chemotherapy

male female P ValueMULTIVARIAT

E

1 yr survival rate

16% 26% 0.005

Women is a strong indipendent factor for improved survival

ECOG 1594

Study # PtsSt. IV,

%ORR,%

MST (mos.)

G4ANC,

%

G4Plts,

%

G> 3Neuro,

%

ECOG 1594DDP-TAXDDP-GEMDDP-TXTCb-TAX

292288293290

9121.32117.3

15.3

8.18.17.48.3

57 394943

22912

59510

HA Wakelee JTO 2006

Patient outcomes for ECOG 1594

N=1157Women (433)

Men (726) P value

Response rate (%)Median PFS (mo)Median survival time (mo)

193.8

9.2 mo

193.57.3

0.990.0220.004

- Separately for each arm, trend for improved survival mantained; statistical significance was LOST

HA Wakelee JTO 2006

Toxicity all grade in ECOG 1594

N=1157Women (431)

Men (726)

P value

Nausea (%)Vomiting (%)Alopecia (%)Neurosensory (%)Neuropsychiatric (%)Cardiac Toxicity ≥ g3 (%)

83656449224.1

70525342147.6

<0.0001<0.00010.00030.020.0010.02

HA Wakelee JTO 2006

Bevacizumab in Advanced NSCLC: Efficacy by Gender

Brahmer et al. J Clin Oncol. 2006;24(No 18S):373s. Abstract 7036.

ParameterMales Females

PC (n=230)

PCB (n=191)

PC(n=162)

PCB(n=190)

OS, mo 8.7 11.7* 13.1 13.3

PFS, mo 4.3 6.3* 5.3 6.2*

RR, % 16 29* 14 41*

No survival benefit for females despite 4-fold increase in RR and statistically significant difference for PFS

A number of potential explanations (eg, statistical chance, imbalance of unmeasured prognostic factors, or a true difference)

*Statistically significant

Survival by Treatment - Males

0.0

0.2

0.4

0.6

0.8

1.0

Months

Pro

ba

bili

ty

0 6 12 18 24 30 36 42

PC (219 events/ 253 cases)PCB (158 events/ 210 cases)

P = 0.0010

Medians: 8.7, 11.7

J Brahmer ASCO 2006

Survival by Treatment - Females

J Brahmer ASCO 2006

0.0

0.2

0.4

0.6

0.8

1.0

Months

Pro

ba

bili

ty

0 6 12 18 24 30 36 42

PC (125 events/ 180 cases)PCB (147 events/ 207 cases)

P = 0.87

Medians: 13.1, 13.3

Vandetanib 200 mgVandetanib 200 mgPaclitaxel 200 mg/mPaclitaxel 200 mg/m22

Carboplatin AUC 6 mg/mL·minCarboplatin AUC 6 mg/mL·min(n=15)(n=15)

Randomized PhaseRun-In Phase

Vandetanib 300 mgVandetanib 300 mgPaclitaxel 200 mg/mPaclitaxel 200 mg/m22

Carboplatin AUC 6 mg/mL·minCarboplatin AUC 6 mg/mL·min(n=10)(n=10)

Vandetanib 300 mgVandetanib 300 mg(n=73) (n=73) [discontinued][discontinued]

Vandetanib 300 mgVandetanib 300 mgPaclitaxel 200 mg/mPaclitaxel 200 mg/m22

Carboplatin AUC 6 mg/mL·minCarboplatin AUC 6 mg/mL·min(n=56)(n=56)

PlaceboPlaceboPaclitaxel 200 mg/mPaclitaxel 200 mg/m22

Carboplatin AUC 6 mg/mL·minCarboplatin AUC 6 mg/mL·min(n=52)(n=52)

Vandetanib (ZD6474) With Carboplatin and Paclitaxel

as First-Line NSCLC Therapy: Schema

Objectives:Objectives:Appropriate Dose in Combination Therapy, Appropriate Dose in Combination Therapy, Pharmacokinetics, SurvivalPharmacokinetics, Survival

Fir

st-L

ine

NS

CL

CF

irst

-Lin

e N

SC

LC

Fir

st-L

ine

NS

CL

CF

irst

-Lin

e N

SC

LC

Heymach et al., IASLC 2005, Abstract P-497Heymach et al., ASCO 2007, Abstract 7544

Vandetanib With Carboplatin and Paclitaxel (CP) as First-Line NSCLC Therapy: Efficacy

Vandetanib/CP

n = 56

Placebo/CP

n = 52

HR P Value

ORR 32% 25% - -

Median OS 10.2 months 11.9 months 1.07 .595

Median PFS 24.0 weeks 23.1 weeks 0.76 .098*

Exploratory Analysis (females)

n = 17 n = 15

Median PFS 28.6 weeks 11.7 weeks 0.47 .037

Median OS ≥ 8.6 months 5.8 months 0.52 .113

Heymach et al., ASCO 2007, Abstract 7544*Met prespecified significance level of P < .2

No significant differences in PFS or OS based on histology were observed.

ZD6474 plus docetaxel vs docetaxel in previously treated NSCLC

A randomized, double-blind, two-part, multicenter study

2nd

-lin

e N

SC

LC

ZD6474 100 mgDocetaxel 75 mg/m2

n=4

Randomized phase*Run-in phase

2nd

-lin

e N

SC

LC

ZD6474 100 mgDocetaxel 75 mg/m2

n=42

PlaceboDocetaxel 75 mg/m2

n=41

ZD6474 300 mgDocetaxel 75 mg/m2

n=11

ZD6474 300 mgDocetaxel 75 mg/m2

n=44

JV Heymach, ASCO 2006

0.25 0.5 1.0 2.0 4.0

Time to progression

Time to death

Males vs females: exploratory analysis

Hazard ratios and 95% confidence intervals

0.25 0.5 1.0 2.0 4.0

Males

Females

All patients

ZD6474 100 mg + doc (n=42) vs placebo + doc (n=41)

ZD6474 300 mg + doc (n=44) vs placebo + doc (n=41)

ZD6474 100 mg + doc (n=42) vs placebo + doc (n=41)

ZD6474 300 mg + doc (n=44) vs placebo + doc (n=41)

JV Heymach, ASCO 2006

• Work better in women than men

• Gefitinib and erlotinib studies: females RR, more symptom improvement, longer OS (univariate), but no difference in benefit phase III study erlotinib

• Biologic basis for difference by sex complex: interactions with frequency of activating mutations, EGFR+ (by IHC/amplification), adenocarcinoma or BAC histology, non-smoking status

EGFR Tyrosine Kinase Inhibitors Second and Third Line Therapy

EGFR Tyrosine Kinase Inhibitors

• Female sex is predictive of response: - Symptoms improvement: 50% vs 31% (p

0.006) - Radiographic regression: 19% vs 3%

(p=0.001) with 82% of responses among women

Kris MG, JAMA 2003

BR.21 Trial: Overall Survival by Gender

SUMMARY STATISTICS:Log-Rank test for equality of groups: p=0.0025

Female/OSI-774 Female/PlaceboMale/OSI-774 Male/Placebo

Perc

enta

ge

0

20

40

60

80

100

Time (months) # At Risk(Female/OSI-774) # At Risk(Female/Placebo) # At Risk(Male/OSI-774) # At Risk(Male/Placebo)

0.017383

315160

10.0752711332

20.03292

30.00000

_____ Erlotinib Female

_____ Placebo Female

_____ Erlotinib Male

_____ Placebo Male

Months

Interaction p = n.s.

Shepherd, NEJM, 2005

Is there an interaction between the ER and EGFr pathways?

•EGFR protein expression is down-regulated in response to estrogen

•EGFR protein expression is up-regulated when estrogen is depleted

•Suggests “cross-talk” between these two pathways

Paclitaxel Poliglumex (PPX)

• Macromolecule that combines paclitaxel with poly-L- glutamic acid

• Into the cell broken in its active form by cathepsin B (regulated by estrogen) minimize systemic exposure & prolong exposure to active drug

ELIGIBILITY CRITERIA

chemonaive

advanced NSCLC

PS2

STRATIFIED BY

disease stage

sex

brain mts

geography

STELLAR 3

PPX 210 mg/mq + CBDCA AUC6 Q3w

versus

Paclitaxel 225 mg/mq + CBDCA AUC 6 Q3w

STELLAR 4

PPX 175 mg/mq Q3w

versus

GMC 1200 mg/mq d1,8,15 Q4w OR NVB 30 mg/mq d 1,8,15 Q3w

Ross H, ASCO 2006

-The composite analysis of STELLAR 3 and 4 shows a statistically significant survival benefit for women receiving PPX (p=0.03)

-The presence of estrogen appear to make PPX a more effective drug

-PIONEER (Paclitaxel Poliglumex Investigating Outcomes in NSCLC: Establishig Estrogen Response) study is ongoing

Favourable factors in Group N Multivariate analysis

Danish 443 Cox: women, good PS,

(Osterlind et al) normal sodium e uric acid

CALGB R 1745 Cox: women, good PS, (Spiegelman et al) age <60 yrs

SWOG R 1,316 Cox: women, PS 0-1, caucasian,(Albain et al) conc. CT/RT, LDH

1,137 RPA: women, LDH nn,

no pleural effusion,age<70 yrs

SCLC: limited disease

Albain (10 SWOG trials from 1976 to1988)

male female

MST

LD

ED

17.7mo

no

24.4mo

differences

Singh, S. et al. J Clin Oncol; 23:850-856 2005

Small Cell Lung Cancer Survival by Sex: retrospective review (4 trials)

SCLC: Toxicity and outcomes by sex

• No difference in treatment delivered or toxic deaths, but greater treatment delays in females, with more toxicity (anemia, neutropenia, stomatitis, emesis)

• Greater toxicity females significant in multivariate model

• Females RR higher and overall survival (p<.0001)

1006 Patients (648 m, 358 f)

on 4 trials of similar chemotherapy

Singh, S. et al. J Clin Oncol; 23:850-856 2005

Conclusion• A better understating of the genetic, metabolic, and

hormonal factors in women represents a research priority

• Evidence suggests that the development of lung cancer is different in women compared with men

• Women with lung cancer live longer than men with

lung cancer, regardless of therapy and stage

• Sex as stratification factor in prospective clinical trials