Recent advances in the molecular pathology of lung cancer: role of EGFR and KRAS mutation testing in treatment selection

Transcription

1 ASIP 2009 USCAP Companion Meeting Molecular Pathology for the Practicing Pathologist Recent advances in the molecular pathology of lung cancer: role of EGFR and KRAS mutation testing in treatment selection Marc Ladanyi, M.D. Memorial Sloan-Kettering Cancer Center New York, NY Web resources The Sanger Institute (U.K.) maintains the COSMIC database (Catalogue Of Somatic Mutations In Cancer) that includes a listing of all reported EGFR mutations, as well as mutations in KRAS and other kinases: The City of Hope Clinical Molecular Diagnostic Laboratory (Duarte, CA) maintains an EGFR mutation database: Recent reviews and articles Papadopoulos N, Kinzler KW, Vogelstein B. The role of companion diagnostics in the development and use of mutation-targeted cancer therapies. Nat Biotechnol 2006;24: Riely GJ, Politi KA, Miller VA, Pao W. Update on epidermal growth factor receptor mutations in non-small cell lung cancer. Clin Cancer Res 2006;12: Sharma SV, Bell DW, Settleman J, Haber DA. Epidermal growth factor receptor mutations in lung cancer. Nat Rev Cancer 2007;7: Sequist LV, Bell DW, Lynch TJ, Haber DA. Molecular predictors of response to epidermal growth factor receptor antagonists in non-small-cell lung cancer. J Clin Oncol 2007;25: Pao W, Ladanyi M. EGFR mutation testing in lung cancer: searching for the ideal method. Clin Cancer Res 13: , Li AR, Chitale D, Riely GJ, Pao W, Ma Y, Zheng T, Miller VA, Zakowski MF, Kris MG, Ladanyi M. EGFR mutations in lung adenocarcinomas: clinical testing experience and their relationship to EGFR gene copy number and immunohistochemical expression. J Mol Diagn 10: , 2008.

2 Marks JL, Broderick S, Zhou Q, Chitale D, Li AR, Zakowski MF, Kris MG, Rusch VW, Azzoli CG, Venkatraman ES, Ladanyi M, Pao W. Prognostic implications of EGFR and KRAS mutations in resected lung adenocarcinoma. J Thoracic Oncology 3: , Pratilas CA, et al: Genetic predictors of MEK-dependence in non-small cell lung cancer. Cancer Res 68: , Ding L, et al. Somatic mutations affect key pathways in lung adenocarcinoma. Nature 455: , Riely GJ, Kris MG, Rosenbaum D, Marks J, Li AR, Chitale DA, Nafa K, Riedel ER, Hsu M, Pao W, Miller VA, Ladanyi M. Frequency and distinctive spectrum of KRAS mutations in never smokers with lung adenocarcinoma. Clin Cancer Res 14: , Ladanyi M, Pao W. Lung adenocarcinoma: guiding EGFR-targeted therapy and beyond. Mod Pathol 21, S16-S22, Riely GJ, Ladanyi M. KRAS Mutations: an old oncogene becomes a new predictive biomarker. J Mol Diagn 10: , Garcia J, Riely GJ, Nafa K, Ladanyi M. KRAS mutational testing in the selection of patients for EGFR-targeted therapies. Sem Diagn Pathol 25:288-94, 2008.

3 Recent advances in the molecular pathology of lung cancer: role of EGFR and KRAS mutation testing in treatment selection Marc Ladanyi, M.D. Chief, Molecular Diagnostics Service Memorial Sloan-Kettering Cancer Center New York, NY, USA

4 Targeted therapies of cancer with molecular predictive markers the new paradigm

5 Predictive molecular testing for targeted therapy of lung cancers 1. Testing for EGFR and KRAS mutations in Lung Cancer 2. Beyond EGFR and KRAS

6 EGFR EGFR is a receptor tyrosine kinase of the ErbB family 4 closely related receptors ErbB1 EGFR, HER1 ErbB2 HER2/neu ErbB3 HER3 ErbB4 HER4

7

8 Small molecule EGFR tyrosine kinase inhibitors (TKIs) Erlotinib Tarceva (Genentech) Gefitinib Iressa (Astra-Zeneca) Oral drugs with relatively low toxicity (rash, diarrhea) Pharmacokinetics once-daily dosing Expression of EGFR by IHC in large proportion (50-60%) of lung CA provided initial rationale for trials

9 Dramatic response to gefitinib

10 Histologic features of responders to EGFR inhibitors Responders: typically well to moderately differentiated; peripheral pure BAC or adenocarcinoma with BAC component

11 Histologic features of responders to EGFR inhibitors Responders: typically well to moderately differentiated adenocarcinoma BAC components, papillary components peripheral; TTF1 + often with (non-mucinous) BAC component Non-responders: more often poorly differentiated; more often TTF1-negative pure squamous carcinomas (in contrast to adenosquamous CA) mucinous BAC (assoc with KRAS mutations) Histologic overlap between the 2 populations Among lung adenocarcinomas, negative predictive value of morphologic features insufficient to exclude EGFR mutation testing

12 Discovery of activating mutations in EGFR kinase domain in lung cancers (2004)

13 EGFR Mutations Associated with Sensitivity to EGFR-TKIs EGF ligand binding Tyrosine kinase autophos TM K DFG Y Y Y Y GXGXXG K DFG L L Exon: LREA G719A/C deletion L858R L861Q Mutations described in first reports. Lynch et al 04; Paez et al 04; Pao et al 04

14 EGFR Mutations and EGFR TKI Responses Data cumulated by Sequist et al., JCO March 2007 EGFR mutations: Among responders = ~ 80% Among non-responders = 7% Reasons:? Not all EGFR mutations confer sensitivity,? Other mutations (e.g. PTEN loss, PIK3CA mut, etc..) Responses: Among EGFR-mutant cases = ~ 80% Among EGFR-mutation negative cases = 10% Reasons:? undetected EGFR mutations due to limited extent or poor sensitivity of direct sequencing direct sequencing is likely to miss mutations when the tumor cells make up less than 25% of the sample

15 Association between EGFR mutation and EGFR Amplification Difficult to tease out because of their frequent cooccurrence. Reported ranges vary depending on technique, criteria and interobserver variability Among EGFR mutated cases, ~50% show increased EGFR copy number ~75% of cases with increased EGFR copy number show mutations.

16 A CISH - neg CISH - low CISH - high C EGFR mutation 29/60 EGFR amplification by CISH 20/ B IHC 0 IHC EGFR overexpression by IHC 29/60 IHC 2+ IHC 3+

17 Mutation status may be more rational for treatment selection The response rates to EGFR TKIs are high in the EGFR mutated case regardless of the copy number. The EGFR amplified cases that lack mutations have very low response rates in the order on 1-8%. When both alterations are present, the mutated allele is preferentially amplified suggesting that it is the mutation what drives the selection for copy number gains.

18 Phase II trial of erlotinib in 101 pts with BAC or adenocarcinoma, BAC subtype EGFR Status n Response (%) Median PFS (mo) Mut+/CISH Mut+/CISH < Mut-/CISH Mut-/CISH < Miller VA, et al. Molecular Characteristics of Bronchioloalveolar Carcinoma and Adenocarcinoma BAC Subtype Predict Response to Erlotinib. J Clin Oncol 26:1472-8, 2008.

19 Phase II trial of erlotinib in 101 pts with BAC or adenocarcinoma, BAC subtype Waterfall plot of maximal reduction of indicator lesions Miller VA, et al. Molecular Characteristics of Bronchioloalveolar Carcinoma and Adenocarcinoma BAC Subtype Predict Response to Erlotinib. J Clin Oncol 26:1472-8, 2008.

20 Mutant EGFR is biologically linked to ligand independent activation and increased downstream signaling. EGFR mutations more closely linked to known risk factors and patient profile (female, Asian, non-smoker) than is EGFR amplification.

21 Pts with EGFR Mutations Survive Longer on Gefitinib Prospective studies Han et al, JCO 05 Mitsudomi et al, JCO 05 Taron et al, CCR 05 Tokumo et al, CCR 05 Cortes-Funes et al, Ann Oncol 05

22 Iressa Pan Asian Study (ipass) Patients Chemonaïve Age 18 years Adenocarcinoma histology Never or light exsmokers* Life expectancy 12 weeks PS 0-2 Measurable stage IIIB / IV disease Gefitinib (250 mg / day) 1:1 randomization Carboplatin (AUC 5 or 6) / paclitaxel (200 mg / m 2 ) 3 weekly # Endpoints Primary Progression-free survival (non-inferiority) Secondary Objective response rate Overall survival Quality of life Disease-related symptoms Safety and tolerability Exploratory Biomarkers EGFR mutation EGFR-gene-copy number EGFR protein expression *Never smokers, <100 cigarettes in lifetime; light ex-smokers, stopped 15 years ago and smoked 10 pack years; # limited to a maximum of 6 cycles Carboplatin / paclitaxel was offered to gefitinib patients upon progression PS, performance status; EGFR, epidermal growth factor receptor Tony Mok

23 ipass - Study Details 87 centers in 9 countries in Asia China, Hong Kong, Indonesia, Japan, Malaysia, Philippines, Singapore, Taiwan, Thailand 1217 patients randomized Randomisation period: March 2006 to October 2007 Data cut-off: 14 April PFS events observed in ITT population (78% maturity) Mean time on treatment gefitinib, 6.4 months carboplatin / paclitaxel, 3.4 months (median number of cycles # : 6) Final survival data (944 events) expected mid-2010 Myanmar Hong Kong # limited to a maximum of 6 cycles PFS, progression-free survival; ITT, intent-to-treat Tony Mok

24 Iressa Pan Asian Study (ipass) Overall EGFR mutation positive rate = 59.7% (261 / 437)!! Tony Mok

25 ipass - Comparison of PFS by mutation status within treatment arms Probability of PFS Gefitinib EGFR M+ (n=132) Gefitinib EGFR M- (n=91) Carboplatin / paclitaxel EGFR M+ (n=129) Carboplatin / paclitaxel EGFR M- (n=85) Gefitinib, HR=0.19, 95% CI 0.13, 0.26, p< No. events M+ = 97 (73.5%) No. events M- = 88 (96.7%) Carboplatin / paclitaxel, HR=0.78, 95% CI 0.57, 1.06, p= No. events M+ = 111 (86.0%) No. events M- = 70 (82.4%) Time from randomization (months) M+, mutation positive; M-, mutation negative Tony Mok

26 Dramatic responses to EGFR inhibitor gefitinib, but usually not durable 1 month Durable responses (>3 yrs) in some, but most patients progress after 6-12 months. 1 month

27 Resistance to EGFR-TKIs Primary resistance Tumors that are refractory to EGFR TKIs from the outset Genetic predictors at diagnosis KRAS mutations (common); PTEN loss? EGFR mutations not sensitive to EGFR TKIs (rare, ~2%) ex 20 insertion BRAF mutations (rare, ~3%) In most cases, no good negative predictor Secondary resistance Tumors that are initially sensitive to EGFR TKIs but eventually progress while on therapy Most, if not all, initial responders Genetic finding: second EGFR mutation: T790M (most) MET amplification (some) Second generation EGFR TKIs effective against T790M and ex20 insertions (and mutant HER2)

28 KRAS Mutations: A Negative Predictor for Response to EGFR TKIs in lung cancer Riely GJ, Ladanyi M. KRAS Mutations: an old oncogene becomes a new predictive biomarker. J Mol Diagn 10: , BUT, while KRAS mutations are found in about 15-30% of American lung adenocarcinomas, they are found in only <10% of Asian samples

29 Mutations in the EGFR Pathway in lung adenocarcinoma Ligand Ligand-binding domain Mutations in EGFR and KRAS are mutually exclusive. PI3K K K Grb-2 SOS RAS RAF PTEN AKT MEK mtor STAT 3/5 MAPK Survival Proliferation

30 KRAS patient profile compared to EGFR Smoking history more common No association with BAC histology (except mucinous BAC) Less common in Asians than non- Asians

31 Does it matter? EGFR and KRAS mutations: predictors of survival in resected lung adenoca p = EGFR n = 38 KRAS n = 50 Marks et al JTO 08

32 KRAS patient profile compared to EGFR Riely GJ, et al. Frequency and distinctive spectrum of KRAS mutations in never smokers with lung adenocarcinoma. Clin Cancer Res 14: , 2008.

33 All All Lung Cancers resected at at MSKCC Clinical EGFR/KRAS testing at MSKCC: current flowchart Non-Adenocarcinoma No testing Special request only Adenocarcinoma Reflex testing EGFR mutation testing if negative KRAS mutation testing Adenocarcinoma pts seen by Thoracic Oncologists EGFR initiated late 2004 KRAS initiated late / week ; 750 / year

34 All All Lung Cancers resected at at MSKCC Clinical EGFR/KRAS testing at MSKCC: current flowchart Non-Adenocarcinoma No testing Special request only Adenocarcinoma Reflex testing EGFR mutation testing if negative KRAS mutation testing Adenocarcinoma pts seen by Thoracic Oncologists Feasibility issues Outside pts: in 2007, 700 advanced stage pts through Thoracic Oncology 1/3 come with 15 unstained slides or blocks Operated pts: reflex testing in 1 year period: 500 lung resections 297 adenocas 295 (98%) studied for EGFR/KRAS 58 (20%) EGFR 85 (28%) KRAS

35 All All Lung Cancers resected at at MSKCC Clinical EGFR/KRAS testing at MSKCC: current flowchart Non-Adenocarcinoma No testing Special request only Adenocarcinoma Reflex testing EGFR mutation testing if negative KRAS mutation testing Adenocarcinoma pts seen by Thoracic Oncologists Test ex19 del and L858R only PCR-based mut-specific assays Diagnostic sensitivity: 90% Technical sensitivity: 5-10% Test ex2 / codon 12 & 13 only Direct seq (+/- macrodissection) Diagnostic sensitivity: >95% Technical sensitivity: 25%

36 EGFR ex. 19 deletion assay Lab of Diagnostic Molecular Pathology, MSKCC 9bp deletion - approx. 45% of EGFR mutations - 9, 12, 15, 18, or 24 bp deletions (all preserve reading frame) - most often 15 bp deletion eliminating amino acids ELREA EGFR-Ex19-FWD1 12bp deletion Hotspot for interstitial deletions 15bp deletion EGFR-Ex19-REV1 FAM 18bp deletion

37 EGFR ex. 21 L858R mutation assay Lab of Diagnostic Molecular Pathology, MSKCC - approx. 40% of EGFR mutations - substitution of leucine by arginine at codon 858 (L858R) Undigested sample EGFR-Ex21-FWD1 Sau96I (GGNCC) cctcacagcagggtcttctctgtttcagggcatgaactacttggaggaccgtcgcttggtgcaccgcgacctg G--M--N--Y--L--E--D--R--R--L--V--H--R--D--L- Digested sample No mutation Sau96I (GGNCC) CGGGCC R GCAGCCAGGAACGTACTGGTGAAAACACCGCAGCATGTCAAGATCACAGATTTTGGGCTGGCCAAACTGCTG -A--A--R--N--V--L--V--K--T--P--Q--H--V--K--I--T--D--F--G--L--A--K--L--L- EGFR-Ex21-REV1 Digested GGTGCGGAAGAGAAAGAATACCATGCAGAAGGAGGCAAAgtaaggaggtggctttaggtcagccagcattttcctga sample -G--A--E--E--K--E--Y--H--A--E--G--G--K L858R mutation FAM

38 KRAS mutation assay (PCR-sequencing) Lab of Diagnostic Molecular Pathology, MSKCC F R GGT > GCT G12A

39 Limitations of mutation detection by direct sequencing Sequencing will not detect proportions of tumor cells below the sensitivity level (25%). Microdissection routinely used to increase tumor content (eliminate non-neoplastic areas) Blocks or unstained sections for DNA extraction should be from the most cellular areas with >50% tumor cells. Select sections without excessive inflammatory response.

40 L29T: 35G A H358: 34G T H1734: 37G T 100% 100% 25% 25% 100% 25% Sequencing will not reliably detect proportions of tumor cells below 25%. 12.5% 12.5%? 12.5%? 6.25% 6.25% X 6.25% X Forward Reverse Forward

41 Use of EGFR and KRAS mutation status to select EGFR-TKI therapy Scenario EGFR Mutation KRAS Mutation Treatment Gefitinib or Erlotinib Trial of drug? Alternative agents Extremely rare, if real

42 The Lung Adenocarcinoma Oncogenome Unknown EGFR (2004) KRAS (1987) ALK fusions (2007) ERBB2 (2004) BRAF (2002) PIK3CA (2004) Mutually exclusive mutations: EGFR, KRAS, BRAF, ERBB2

43 Mutations in the EGFR Pathway in lung adenocarcinoma Ligand Ligand-binding domain Mutations in EGFR, KRAS, BRAF, and HER2 are mutually exclusive. PI3K K K Grb-2 SOS RAS RAF PTEN AKT MEK mtor STAT 3/5 MAPK Survival Proliferation

44 BRAF MUTATIONS in Lung CA BRAF mutations identified in 2-3 % of lung adenocarcinoma Most common mutation is V600E, Exon 15 predicts resistance to EGFR inhibition predicts benefit from MEK selective inhibition (Pratilas et al., Cancer Res 68: , 2008)

45 Predictive molecular testing for targeted therapy of lung cancers 1. Testing for EGFR and KRAS mutations in Lung Cancer 2. Beyond EGFR and KRAS

46 Predictive Testing for Targeted Therapies: impact on clinical labs New emphasis on mutation detection Methods need to work on DNA extracted from archival pathology paraffin blocks Methods need to detect mutations even when tumor cells are <10% of tumor biopsy sample Increasing need for higher-throughput molecular diagnostic technologies

47 EGFR mutations in lung adenocarcinoma: beyond exon 19 deletions and exon 21 L858R exon 20 insertions major secondary resistance mut. exon 19 deletions Sharma et al. Nature Reviews Cancer 7, (March 2007)

48 A New Way to Look at Lung Cancer KRAS mutated Squamous Large Cell Adenocarcinoma Non-Small Cell Lung Cancer (Classical Morphologic Categories) 180,000 USA cases in 2008 EGFR mutated BRAF mutated PI3KCA mutated HER2 mutated Adenocarcinoma (New Molecular Categories) 121,000 USA cases in 2008 EGFR testing alone is not enough >5 genes x multiple different mutations/gene = lots of assays 1 mutation / assay no longer sustainable for clinical labs

49 Nature Oct 23, 2008

50 Next generation molecular diagnostics: screening for multiple mutations more efficiently What s needed: high throughput / high sensitivity Multiplexing (multiple mutations tested / assay) Automated liquid handling (reduce labor costs) need to detect mutations even when tumor cells are only 5% (or less) of tumor biopsy sample Semi-automated mutation calling would be nice New technology in place at MSKCC since 2006: Sequenom: instrument for highly multiplexed mass spectrometry-based nucleic acid assays

51 Sequenom system Mass spectrometry-based nucleic acid assays Assay Design Genotyping Assay Design 3 primers specific to each SNP or mutation: 2 PCR Primers 1 Extend primer 3 or 5 adjacent to the SNP point mutation detected by different mass of single base extension product

52 Sequenom assays for lung kinase mutations H358 cell line: KRAS 34 G>T Mutation 1 Mut : 1 WT WT Peak G measure presence of >20 mutations per assay MUT Peak T automated detection of mutations; followed by manual review sensitivity: 5-10% (if manual review)

53 Clinical Genes in Sequenom Panel example: KRAS & EGFR Total # of mutations in COSMIC % of mutations for gene coding sequence mutation amino acid change Cases Gene KRAS c.34g>t p.g12c KRAS c.34g>c p.g12r KRAS c.34g>a p.g12s KRAS c.35g>c p.g12a KRAS c.35g>a p.g12d KRAS c.35g>t p.g12v KRAS c.37g>t p.g13c KRAS c.37g>c p.g13r KRAS c.37g>a p.g13s KRAS c.38g>c p.g13a KRAS c.38g>a p.g13d KRAS c.38g>t p.g13v KRAS c.181c>g p.q61e KRAS c.181c>a p.q61k KRAS c.182a>t p.q61l KRAS c.182a>c p.q61p KRAS c.182a>g p.q61r KRAS c.183a>c p.q61h KRAS c.183a>t p.q61h KRAS c.436g>c p.a146p KRAS c.436g>a p.a146t Total # of mutations in COSMIC % of mutations for gene coding sequence mutation amino acid change Cases Gene EGFR c.2155g>t p.g719c EGFR c.2155g>a p.g719s EGFR c.2156g>c p.g719a EGFR c.2156g>a p.g719d 2791 >1000 >36% EGFR c.2235 deletion EGFR c.2236 deletion EGFR c.2237 deletion EGFR c.2238 deletion EGFR c.2239 deletion EGFR c.2240 deletion EGFR c.2241 deletion EGFR c.2281g>t p.d761y EGFR c.2303g>t p.s768i EGFR c.2369c>t p.t790m ? EGFR c.2560a>g p.t854a EGFR c.2572c>a p.l858m EGFR c.2573t>g p.l858r EGFR c.2582t>a p.l861q EGFR c.2582t>g p.l861q

54 Research Genes in Sequenom Panel example: BRAF & PIK3CA Total # of mutations in COSMIC % of mutations for gene coding sequence mutation amino acid change Cases Gene BRAF c.1406g>c p.g469a BRAF c.1406g>a p.g469e BRAF c.1406g>t p.g469v BRAF c.1781a>g p.d594g BRAF c.1781a>t p.d594v BRAF c.1798g>a p.v600m BRAF c.1799t>c p.v600a BRAF c.1799t>a p.v600e BRAF c.1799t>g p.v600g Total # of mutations in COSMIC Cases % of mutations for gene Gene coding sequence mutation PIK3CA c.263g>a p.r88q amino acid change PIK3CA c.1258t>c p.c420r PIK3CA c.1624g>a p.e542k PIK3CA c.1624g>c p.e542q PIK3CA c.1634a>c p.e545a PIK3CA c.1634a>g p.e545g PIK3CA c.1635g>t p.e545d PIK3CA c.1633g>a p.e545k PIK3CA c.3129g>t p.m1043i PIK3CA c.3140a>t p.h1047l PIK3CA c.3140a>g p.h1047r PIK3CA c.3139c>t p.h1047y

55 Clinical / Research tumor mutation profiling Present platform: Sequenom mass-spec genotyping Great for recurrent point mutations in oncogenes Sensitivity: 5-10% Good for poor quality DNA Applicable to known gene fusions at cdna level only Poor for insertions, deletions Poor for tumor suppressor genes (P53, PTEN, NF1, STK11) because point mutations less recurrent, more scattered Not good for previously unknown mutations

56 Predictive Testing for Targeted Therapies: impact on clinical trials Phase 1: safety and dosage Phase 2: efficacy and safety Phase 3: randomized controlled study

57 Predictive Testing for Targeted Therapies: impact on practice Nature Biotechnology, August 2006

58 Predictive molecular testing for targeted therapy of lung cancers 1. Testing for EGFR and KRAS mutations in Lung Cancer 2. Beyond EGFR and KRAS

59 Acknowledgements Some slides graciously provided by: William Pao, MD PhD Maria Arcila, MD