Current status of biomarker research in NSCLC

1
Current status of biomarker research in NSCLC

• Tony Mok The Chinese University of Hong Kong,
  Hong Kong

2
(No Transcript)
3
Research on prognostic markers
4
EGR-1 and PTEN

• Early Growth Response gene 1 (EGR-1) is linked to
  tumour suppression via the PTEN pathway
• 125 cases of resected NSCLC
• RNA extraction and EGR-1 expression quantified by
  real-time PCR

Ferraro B, et al. J Clin Oncol 20052319216
5
Low EGR-1 is a prognostic marker of poor survival
outcome in resected NSCLC
Overall survival EGR-1
Overall survival PTEN
100 75 50 25 0
100 75 50 25 0
Survival ()
Survival ()
p0.03
p0.04
0 20 40 60 80 100 120
0 20 40 60 80 100 120
Time (months)
Time (months)
Disease-free survival PTEN
Disease-free survival EGR-1
100 75 50 25 0
100 75 50 25 0
Survival ()
Survival ()
p0.05
p0.1
0 20 40 60 80 100 120
0 20 40 60 80 100 120
Time (months)
Time (months)
Ferraro B, et al. J Clin Oncol 20052319216
6
Metagene model in stage IA NSCLC

• Gene expression profile in 89 patients with
  early stage NSCLC
• Dukes Lung Metagene Model
• Microarray assays by Affymetrix GeneChips
  (U133Plus2)
• Validation in two cohorts of patients from two
  clinical trials
• Accuracy in prediction of recurrence 72 and 79

Potti A, et al. N Engl J Med 200635557080
7
Patient selection for adjuvant chemotherapy
Patients with stage IA NSCLC
Stage IA, predicted low riskof recurrence (n47)
100 80 60 40 20 0
Surgery and gene-expressionanalysis
Stage IA (n68)
Application of lung metagene model
Survival ()
Stage IA, predicted high riskof recurrence (n21)
plt0.001
0 25 50 75 100 125 150
Months

• Prospective validation is necessary

Potti A, et al. N Engl J Med 200635557080
8
RRM1 and ERCC1

• Ribonucleotide reductase M1 (RRM1) is a key
  enzyme in DNA synthesis
• low level expression of RRM1 is associated with
  poor survival
• Excision Repair Cross-Complementing Group 1
  (ERCC1) plays a key role in the repair of damaged
  DNA
• high ERCC1 expression is associated with
  cisplatin resistance

9
Automated quantitative assessment of RRM1 and
ERCC1

• Tissue microarray of 187 resected stage I NSCLC
• Immunofluorescence combined with automated
  quantitative analysis (AQUA)
• Also measure PTEN and cytokeratin

Zheng Z, et al. N Engl J Med 20073568008
10
RRM1 as a prognostic marker
100 80 60 40 20 0
High level of RRM1 (gt40.5)
Disease-free survival ()
Low level of RRM1 (lt40.5)
p0.004
0 12 24 36 48 60 72 84 96 108 120
Months
100 80 60 40 20 0
High level of RRM1 (gt40.5)
Overall survival ()
Low level of RRM1 (lt40.5)
p0.02
0 12 24 36 48 60 72 84 96 108 120
Months
Zheng Z, et al. N Engl J Med 20073568008
11
RRM1 and ERCC1 as prognostic markers
Months
Zheng Z, et al. N Engl J Med 20073568008
12
Research on predictive markers chemotherapy
13
ERCC1 as a predictor of response
Control arm docetaxel/ cisplatin
Stage IV NSCLC
21 randomisation
Low ERCC1 docetaxel/cisplatin
Study arm ERCC1 expression analysis
High ERCC1 docetaxel/gemcitabine
Rosell R, et al. J Clin Oncol 200523(Suppl.
16)621s (Abs. 7002)
14
Response according to ERCC1 levels
p0.02
Rosell R, et al. J Clin Oncol 200523(Suppl.
16)621s (Abs. 7002)
15
IALT study adjuvant chemotherapy
100 80 60 40 20 0
Overall survival
Chemotherapy Control
Patients ()
HR0.86 (0.760.98) plt0.03
0 1 2 3 4 5
Years
At risk
The IALT Collaborative Group. N Engl J Med
200435035160
16
Test of interaction between ERCC1 and treatment
plt0.009

• For ERCC1 ve tumours
• 14-month benefit in overall survival for CT vs
  control
• 6-month benefit in overall survival for CT vs
  ERCC1 ve tumours

Olaussen KA, et al. N Engl J Med 200635598391
17
Prediction of survival by ERCC1, RRM1 or EGFR
expression
Overall population (treated with
cisplatin/gemcitabine or gemcitabine alone)
1.0 0.8 0.6 0.4 0.2 0
1.0 0.8 0.6 0.4 0.2 0
1.0 0.8 0.6 0.4 0.2 0
ERCC1
RRM1
EGFR
lt4?4
lt7.5?7.5
lt10?10
Cumulative survival
Cumulative survival
Cumulative survival
p0.0032
p0.0390
p0.452
0 10 20 30 40 50 60 70
0 10 20 30 40 50 60 70
0 10 20 30 40 50 60 70
Time (months)
Time (months)
Time (months)
Cisplatin/gemcitabine-treated patients
ERCC1 and RRM1
ERCC1 and RRM1
1.0 0.8 0.6 0.4 0.2 0
1.0 0.8 0.6 0.4 0.2 0
Low expression in both genesOthers
At least one genewith low expression Others
Cumulative survival
Cumulative survival
p0.0023
p0.0216
0 10 20 30 40 50 60 70
0 10 20 30 40 50 60 70
Time (months)
Time (months)
Relative gene expression level versus internal
reference gene (ß-actin) assessed by
quantitative real-time RT-PCR
Ceppi P, et al. Ann Oncol 200617181825
18
Biomarkers for chemotherapy summary

• High RRM1 and ERCC1 expression is associated with
  longer survival after resection of early stage
  NSCLC (prognostic)
• High RRM1 and ERCC1 are predictors of lower
  tumour response rate and shorter survival for
  treatment with gemcitabine and cisplatin
  (predictive)
• Low ERCC1 expression is associated with survival
  benefit from adjuvant chemotherapy for NSCLC
  (predictive)
• These biomarkers have not been prospectively
  validated

19
Research on predictive markers EGFR-targeted
therapy
20
Potential biomarkers for EGFR TKIs
EGFR gene polysomy by FISH
EGFR protein overexpression by IHC
EGFR gene mutation by PCR and sequencing
21
EGFR TK mutations

• Three major types of mutation
• missense mutation in exons 18 and 21
• G719A (4)
• L858R (38)
• deletions in exon 19
• E746A750 (45)

Paez JG, et al. Science 20043041497500
22
Response to gefitinib in Asian patients with EGFR
TK mutations (retrospective studies)
Zhang XT, et al. Ann Oncol 2005161334-42 Shih
JY, et al. Int J Cancer 2006118963-9 Mitsudomi
T, et al. J Clin Oncol 2005232513-20 Han SW,
et al. J Clin Oncol 2005232493-501 Takano T, et
al. J Clin Oncol 2005236829-37 Tokumo M, et
al. Clin Cancer Res 2005111167-73
23
First-line EGFR TKIs in unselected Caucasian
patients with advanced NSCLC
1Reck M, et al. Clin Lung Cancer 2006740611
2Jackman DM, et al. J Clin Oncol
20072576063Giaccone G, et al. Clin Cancer Res
200612604955
24
First-line EGFR TKIs in selected Caucasian
patients
1Paz-Ares L, et al. J Clin Oncol 200624(Suppl.
18)369s (Abs 7020) 2Cappuzzo F, et al. IASLC
3rd International Conference Molecular Targeted
Therapies in Lung Cancer, Taormina, Sicily 2006
25
First-line gefitinib in selected Asian patients
(EGFR TK mutations)
1Inoue A, et al. J Clin Oncol 200624334062Okam
oto I, et al. J Clin Oncol 200624(Suppl. 18)
382s (Abs. 7073)3Sutani A, et al. J Clin Oncol
200624(Suppl. 18) 383s (Abs. 7076) 4Morikawa N,
et al. J Clin Oncol 200624(Suppl. 18) 383s
(Abs. 7077)
26
Survival according to EGFR mutation status BR.21
Wild type EGFR and indeterminate variants
Exon 19 deletions and L858R mutations
100 80 60 40 20 0
100 80 60 40 20 0
Tarceva Placebo
Tarceva Placebo
Log-rank p0.11HR0.75 (95 CI 0.531.07)
Log-rank p0.16HR0.52(95 CI 0.211.31)
Survival ()
Survival ()
0 6 12 18 24
0 6 12 18 24
Time (months)
Time (months)
At riskTarceva Placebo
120 69 38 9 0 57 24 12 6 0
10 6 5 1 0 14 7 5 1 0
Tsao M-S, et al. N Engl J Med 20063545278
p value for interaction 0.45
27
Why?

• Lack of survival benefit in EGFR
  mutation-positive patients in BR.21 is probably
  due to the small sample size (10 vs 14 patients)
• Other hypothetical factors
• mutation type and presence of T790M mutation
• mutations in related genes
• ethnicity/polymorphisms
• gene expression levels

28
EGFR gene copy number by FISH BR.21
FISH ve
FISH ve
Tsao M-S, et al. N Engl J Med 200535313344
29
Survival according to EGFR gene copy number BR.21
EGFR FISH ve
EGFR FISH ve
100 80 60 40 20 0
100 80 60 40 20 0
HR0.85 (0.481.51) p0.59
HR0.44 (0.230.82) p0.008
Survival ()
Survival ()
0 6 12 18 24 30
0 6 12 18 24 30
Time (months)
Time (months)
Treatment x FISH status p0.10 (not
significant) Cox regression analysis
Tsao M-S, et al. N Engl J Med 200535313344
Log-rank test
30
Survival according to EGFR protein expression
(IHC) BR.21
EGFR ve
EGFR ve
100 80 60 40 20 0
100 80 60 40 20 0
Tarceva Placebo
Tarceva Placebo
Log-rank p0.02HR0.68 (0.49, 0.95)
Log-rank p0.70HR0.93 (0.63, 1.36)
Percentage
Percentage
0 6 12 18 24 30
0 6 12 18 24 30
Time (months)
Time (months)
At riskTarceva 117 71 43 5 5
0 Placebo 67 23 12 5 0 0
At riskTarceva 93 42 22 8 3 0
Placebo 48 24 14 3 0 0
p value for interaction 0.25
Tsao M-S, et al. N Engl J Med 200535313344
31
My simple view
EGFR mutation
EGFR wild type
Tarceva
Dramatic tumour response
PR, MR or SD
Survival not captured
Survival benefit as per BR.21
EGFR overexpression Gene amplification
32
Biomarker analyses in clinical trials
33
Selecting patients for adjuvant therapy by
biomarker RADIANT
Tarceva 150mg p.o. once daily for 2 years
Stage IBIIIA EGFR ve Complete resection No
radiotherapy
Four cycles of standard platinum-based chemothera
py (optional)
Stratified by country adjuvant chemotherapy
histology stage smoking status EGFR status
R
Placebo

• Primary endpoint disease-free survival (all
  patients, EGFR IHC ve and/or EGFR FISH ve)
• Co-primary disease-free survival in FISH ve
  (USA) to be determined from the outcome of
  SATURN for Europe
• Planned n945

34
Marker identification trial (MERIT)
n250 recruited, second line after standard
first-line treatment
Mandatorysamples Tumour biopsy Tumour block(if
available) RNA blood sample
Biopsy
Tarceva 150mg/day
PD
SecondRNA bloodsample
Day 1
Day 28 Day 1
Week 6
Follow-up every 12 weeks
Clinical assessment at screening and every 6
weeks until PD

• Primary endpoint differentially expressed genes
  (prediction of clinical benefit)
• Secondary endpoints EGFR mutation analysis,
  molecular (explorative) assessment of putative
  alterations of downstream targets of EGFR

35
SATURN Sequential Tarceva in unresectable NSCLC
Stratify by EGFR protein expression (IHC) results
Tumour samples (mandatory)
Tarceva 150mg/day
PD
Off study
Chemotherapy naïve stage IIIB/IV NSCLC Planned
n1,700
Four cycles of first-line standard platinum-based
doublet
Non-PD
11
(n850)
Placebo
PD
Off study
PD

• Primary endpoint PFS (25 increase in all and
  30 increase in EGFR IHC ve)
• Secondary endpoints OS, PFS (EGFR ve, TTP,
  safety)

TITAN or off study
36
Biomarkers for EGFR TKIs summary

• EGFR mutation seems to be a powerful predictor of
  tumour response to EGFR TKIs
• not established as a predictive marker (small
  sample size in BR.21)
• EGFR gene copy number (FISH) and EGFR protein
  expression (IHC) may be predictors of survival
• Prospective validation of these markers, as well
  as the assays, is required before they can be
  used in routine clinical practice
• this is being performed in the Roche-led SATURN
  and RADIANT studies of Tarceva

37
(No Transcript)
38
12th World Conference on Lung CancerSeptember
2-6, 2007 Seoul, Korea
See you in beautiful and historic Seoul
www.iaslc.org www.2007worldlungcancer.org