Early Stage NSCLC Imprimatur of Adjuvant Therapy Overview

Early Stage NSCLC: Imprimatur of Adjuvant Therapy Overview of Recent Data Corey J Langer MD, FACP Professor of Medicine Director of Thoracic Oncology Abramson Cancer Center University of Pennsylvania Philadelphia, PA 19104

Disclosures: Past 10 yrs • Grant/Research Support: – Bristol Myers Squibb, Pfizer, Imclone, Lilly, Schering. Plough Research Institute, Sanofi-Aventis, Amgen, Cell Therapeutics, Ortho. Biotech, Celgene, Vertex, Genentech, OSI, Astra. Zeneca, Pfizer, Medimmune, GSK • Scientific Advisor: – Bristol Myers Squibb, Imclone, Sanofi-Aventis, Pfizer, Glaxo. Smith. Kline, Pharmacyclics, Amgen, Astra. Zeneca, Novartis, Genentech, OSI, Savient, Bayer/Onyx, Abraxis, Clarient, Morphotek, Biodesix, AVEO, Synta • Speakers Bureau: curtailed as of 12/10 – Bristol Myers Squibb, Imclone, Sanofi-Aventis, Lilly, Genentech, OSI

Welcome to Winter in NJ 2 -14 3

Stage is Destiny!

New Staging System (IASLC ’ 07) to be instituted 2009 UICC 6 T/M Descriptor Proposed T/M N 0 N 1 N 2 N 3 T 1 (< 2 cm) T 1 a IA IIIA IIIB T 1 (> 2 -3 cm) T 1 b IA IIIA IIIB T 2 ( 3 to < 5 cm) T 2 a IB IIA IIIB T 2 (>5 -7) T 2 b IIA IIB IIIA IIIB T 2 (> 7 cm) T 3 IIB IIIA IIIB T 3 invasion T 3 IIB IIIA IIIB T 4 (same lobe nodules) T 3 IIB IIIA IIIB T 4 (extension) T 4 IIIA IIIB M 1 (ipsilateral Lung) T 4 IIIA IIIB T 4 (pleural effusion) M 1 a IV IV M 1 (contralateral lung) M 1 a IV IV M 1 (distant) M 1 b IV IV

New Staging System (IASLC ’ 07) Instituted 2009 UICC 6 T/M Descriptor Proposed T/M N 0 N 1 N 2 N 3 T 1 (< 2 cm) T 1 a IA IIIA IIIB T 1 (> 2 -3 cm) T 1 b IA IIIA IIIB T 2 ( 3 to < 5 cm) T 2 a IB IIA IIIB T 2 (>5 -7) T 2 b IIA IIB IIIA IIIB T 2 (> 7 cm) T 3 IIB IIIA IIIB T 3 invasion T 3 IIB IIIA IIIB T 4 (same lobe nodules) T 3 IIB IIIA IIIB T 4 (extension) T 4 IIIA IIIB M 1 (ipsilateral Lung) T 4 IIIA IIIB T 4 (pleural effusion) M 1 a IV IV M 1 (contralateral lung) M 1 a IV IV M 1 (distant) M 1 b IV IV

Limitations of Earlier Adjuvant Trials • Use of regimens with marginal activity in advanced NSCLC • Inclusion of patients with compromised PS and multiple co-morbidities • Difficulty administering systemic therapy in the post-op setting • Inadequate power or overly ambitious survival endpoints

1995 Meta-Analysis Adjuvant Cisplatin Trials n=1394 100 HR 0. 87 p=0. 08 Percentage Survival 80 60 40 Surgery plus Chemotherapy Surgery 20 0 0 12 18 24 36 48 54 60 Time from Randomization (months) BMJ 31: 899 -908, 1995 5% absolute survival benefit at 5 years, NS

Plaitnum-Based Adjuvant Trials in Resected NSCLC Trial BMJ meta IALT ALPI E 3590 BLT BR-10 CALGB 9633 ANITA LACE meta N. of Patients 1394 1867 1209 488 381 482 344 840 4584 HR (95%CI) 0. 87 (0. 74 -1. 02) 0. 86 (0. 76 -0. 98) 0. 96 (0. 81 -1. 13) 0. 93 (0. 74 -1. 18) 1. 02 (0. 77 -1. 35) 0. 70 (0. 53 -0. 92) 0. 63 (0. 60 -1. 07) 0. 79 (0. 66 -0. 95) 0. 89 (0. 82 -0. 96) NEJM 00; JNCI 03; Euro. JTS 04, NEJM 04; NEJM 05; ASCO 04+06; Lancet Oncology 06

Plaitnum-Based Adjuvant Trials in Resected NSCLC: Longterm Results Trial BMJ meta IALT ALPI E 3590 BLT BR-10 CALGB 9633 ANITA LACE meta N. of Patients 1394 1867 1209 488 381 482 344 840 4584 HR (95%CI) 0. 87 (0. 74 -1. 02) 0. 91 (0. 81 -1. 02) 0. 96 (0. 81 -1. 13) 0. 93 (0. 74 -1. 18) 1. 02 (0. 77 -1. 35) 0. 70 (0. 53 -0. 92) 0. 80 (0. 60 -1. 07) 0. 79 (0. 66 -0. 95) 0. 89 (0. 82 -0. 96) NEJM 00; JNCI 03; Euro. JTS 04, NEJM 04; NEJM 05; ASCO 04+06; Lancet Oncology 06

Intl Adjuvant Lung Cancer Trial (IALT) 1867 pts, I-III • • • Randomized to obs vs. 4 cycles post-op chemo* Radiation therapy optional (~30%) 4. 1% absolute benefit at 5 years, p<0. 03† – Stage III > Stage I benefit • Diminished at 7 yr follow-up • HR 0. 86 [0. 76 -0. 98] 0. 91 [0. 81 -1. 02]; – p = 0. 04 0. 1 after 5 yrs due to non-cancer deaths ^ *cisplatin + etoposide or vinca alkaloid †Arriagado, NEJM 350: 351, 2004, ^ Le Chevalier ASCO 2008

Randomized International Adjuvant Lung Cancer Trial (IALT): Cisplatin-Based Chemotherapy vs No Chemotherapy for Resected NSCLC Stage I-III NSCLC Complete surgical resection within 60 days N=1867 R A N D O M IZ E Arm A* Cisplatin 80 mg/m 2 4 cycles OR Cisplatin 100 mg/m 2 3 -4 cycles OR Cisplatin 120 mg/m 2 3 cycles PLUS Etoposide 100 mg/m 2 3 days/cycle OR Vinorelbine 30 mg/m 2 weekly OR Vinblastine 4 mg/m 2 weekly OR Vindesine 3 mg/m 2 weekly n=935 Arm B Observation ± thoracic radiotherapy 60 Gy† n=932 *Each center selected which chemotherapy it would use. †Optional, but predefined by N stage at each center. Le Chevalier et al. Proc Am Soc Clin Oncol. 2003; 22: 2. Abstract and oral presentation; Arriagada et al. N Engl J Med. 2004; 350: 351.

Adjuvant Chemotherapy IALT • • n=1867 • c. DDP was 80 q 3 weeks X 4 100 q 4 weeks X 3 -4 120 q 4 weeks X 3 56% + Etoposide 100 27% + Vinorelbine 30 11% + Vinblastine 4 6% + Vindesine 3 Le Chevalier, ASCO 2003 abstract 6, NEJM 2004

Adjuvant Chemotherapy IALT (International Adjuvant Lung Trial) n=1867 • • • 1995 -2000 33 countries, initial accrual goal was 3300 80/20 M/F Mean age 59 (all < 75) Squamous 47%, ACAs 40% Chemo to start < 60 days after surgery Median f/u: 56 months Le Chevalier, ASCO 2003 abstract 6, NEJM 1/04

IALT: Cisplatin-Based Chemotherapy vs No Chemotherapy for Resected NSCLC: Overall Survival (Med F/U 56 mos) 100 HR=0. 86 (0. 76 -0. 98) Chemotherapy % Survival 80 Observation ± RT P<0. 03 60 40 20 0 0 12 24 36 48 Months Le Chevalier et al. Proc Am Soc Clin Oncol. 2003; 22: 2. Abstract and oral presentation. NEJM 1/04 60

IALT Trial ARM Number of enrollees Dead of disease progression Tx-related deaths (n) Compliance with RT (%) Median DFS (mo) 2 -yr DFS (%) 5 -yr DFS (%) Median Survival Time (mo) Adjuvant chemo 935 361 Control 14 71 39. 4 61 39 50. 8 2 85 34. 3 55 34 44. 4 932 405

IALT: Cisplatin-Based Chemotherapy vs No Chemotherapy for Resected NSCLC: Overall Survival • Benefit seen across all demographic variables – – Gender type of surgery use of RT geographical location • Greatest benefit in stage III pts (~7. 5%) • In subgroup analyses, survival advantage for stage I and II was not statistically significant – – – Stage III Adj 65. 6 46. 5 37. 4 Control 64. 9 43. 2 29. 9 Rel %↑ 0. 7 3. 3 7. 5

IALT: Cisplatin + a Vinca or Etoposide 2008 Update: 7. 5 -Year Median Follow-Up 100% chemotherapy: 578 deaths - 495 deaths before 5 years - 83 deaths after 5 years 80% 60% HR: 0. 91 (0. 81 -1. 02, P = 0. 10) 40% control 590 deaths 20% - 534 deaths before 5 years - 56 deaths after 5 years 0% 0 1 2 3 4 5 6 7 8 years 935 775 619 520 447 372 282 208 125 932 780 650 550 487 399 300 208 133 Le Chevalier T, et al. J Clin Oncol. 2008(May 20 suppl). Abstract 7507.

Criticisms of IALT • Heterogenous staging, chemo and application of RT (HR favored stage III, not stage I or II) • Study actually closed earlier than planned because of emerging interest in neoadjuvant Tx • Potential Molecular Imbalances: Results of ERCC 1 suggest that one can select a group more likely to benefit; other bio-correlatives still pending • Elderly (> 75) excluded; how do we address this expanding cohort? • Dissipation of survival benefit after 5 years • Why was this trial positive when so many similar trials proved negative?

Recent (-) Trials of Adjuvant CT in Completely Resected NSCLC Study CT Regimen # of Patients Outcome VP 16 -P x 4 462 Negative ALPI/EORTC Italy/Europe MVP x 3 1197 Negative BLT V-P x 4 481 Negative INT 0115 Country USA International on OS

2004: Paradigm Shift

2004: Paradigm Shift Recently Completed (+) Randomized Adjuvant Trials in Early Stage NSCLC Stage No. Intervention CALGB 9633 IB 500 Carboplatin/Paclitaxel NCI-C* IB-II 480 Cisplatin/Vinorelbine

2005: Paradigm Shift Recently Completed (+) Randomized Adjuvant Trials in Early Stage NSCLC Stage No. Intervention CALGB 9633 IB 500 Carboplatin/Paclitaxel NCI-C* IB-II 480 Cisplatin/Vinorelbine ANITA I-IIIA 840

CALGB 9633 - 344 pts, stage IB • 4 cycles carboplatin/paclitaxel vs. observation • Radiation therapy not allowed • 2004 - HR 0. 62; p = 0. 028 • 2006 - HR 0. 80; p = 0. 10 Strauss Proc ASCO 2004 abs 7019, 2006 abs 7007, JCO 2009

CALGB 9633 Overall Survival by Tumor Size Tumor ≥ 4 cm Tumor <4 cm Observation Paclitaxel + Carboplatin n=97 n=99 HR = 0. 66; 90% CI, 0. 45 -0. 97; P=0. 04 Probability Observation Paclitaxel + Carboplatin n=74 HR = 1. 02; 90% CI, 0. 67 -1. 55; P=0. 51 Years Strauss Proc ASCO 2004 abs 7019, 2006 abs 7007, JCO 2008 Years

JBR. 10 ELIGIBLE: Resected IB and II T 2 N 0, T 1 N 1, or T 2 N 1 R A N D O M I Z E Observation N = 240 Vinorelbine 25 mg/m 2 weekly x 16 plus Cisplatin 50 mg/m 2 day 1 & 8 q 4 wk x 4 cycles No prior chemo N = Stratification: No RT 242 • Nodal status – N 0 vs Primary endpoint = overall survival N 1 • RAS status Winton T, et al. N Engl J Med. 2005; 352: 2589 -2597.

NCIC-CTG JBR. 10 482 pts, stage IB-II • • Stage IB-II only 4 cycles of post-op cis/vin vs. observation No radiation therapy 15% survival advantage at 5 years – 54% vs 69% alive (NEJM paper) • HR 0. 69, p=. 012 in 482 pts • Benefit only in stage II in subset analysis – Although post-hoc analysis shows a benefit in IB > 4 cm Winton, Proc ASCO 2004 Abs: 7018, 2004, NEJM 2005

Overall Survival by Treatment Arm Fig. 1 All Patients 5 yr: 67% vs 56% MST 94 m vs 72 m HR 0. 69 5 yr: 69% vs 54% MST 94 m vs 73 m Absolute improvement in 5 yr OS = 15% (69% vs 54%) Winton et al. NEJM 2005 Absolute improvement in 5 yr OS = 11% (67% vs 56%); benefit persists at 9+ yrs Vincent, Butts et al, 2009, ASCO -7501

Cumulative Incidence Plots for Disease and Non-disease Related Deaths

Stage IB Analysis T < 4 cm T ≥ 4 cm HR OS p CALGB 9633 1. 02 . 51 0. 66 . 04 JBR. 10 1. 73 . 07 0. 66 . 13 No Chemo Benefit Strauss JCO 2008 Potential Chemo Benefit

ANITA - 840 pts, stage IB-IIIA • 1: 1 Randomization post-resection – Vinorelbine (30 mg/m 2) Q wk X 16 + Cisplatin 100 mg/m 2 Q 4 wks X 4 vs – Observation • Radiation therapy allowed • 8. 6% survival advantage at 5 years (persists at 7 yr) – 42. 6% vs 51. 2% alive • HR 0. 7 [0. 66 -0. 95], p =. 013 • Benefit only in stage II and IIIA Douillard Lancet Oncol. 7: 719, 2006

ANITA: Rand Phase III Trial of Vinorelbine and Cisplatin vs Obs in Resected stage I-III NSCLC: Demographics • • • 840 pts accrued from 12/94 through 12/00 Median F/U > 70 mos Each arm well balanced Median age 59 (18 – 75) 86% male 95% PS 0 -1 59% Squamous ca 37% pneumonectomy Stage – – – I – 35% II – 30% III – 35% Douillard Lancet Oncol. 7: 719, 2006

Abstract #7013: ANITA Trial Overall survival - ITT population OBS. Survival Distribution Function 1. 00 Median months NVB + CDDP 43. 8 65. 8 P-value 0. 75 0. 013 Hazard Ratio 0. 79 [0. 66 - 0. 95] 0. 50 Obs 0. 25 NVB + CDDP 0 0 20 Douillard Lancet Oncol. 7: 719, 2006 40 60 months 80 100 120

ANITA: Randomized Phase III Trial of Vinorelbine and Cisplatin vs Observation in resected stage I-III NSCLC Arm Observation Adjuvant No 433 407 RFS (mo) 21 36 Median Surv (mo)* 44 66 2 yr OS 63 68 5 yr OS 43 51 7 yr OS 37 48 Stage I 62 63 Stage II 39 52 Stage III 26 42 5 yr OS * P =0. 002, HR 0. 76 Douillard Lancet Oncol. 7: 719, 2006

ANITA: Randomized Phase III Trial of Vinorelbine and Cisplatin vs Observation in resected stage I-III NSCLC Arm Observation Adjuvant No 433 407 RFS (mo) 21 36 Median Surv (mo)* 44 66 2 yr OS 63 68 5 yr OS 43 51 7 yr OS 37 48 Stage I 62 63 Stage III 39 26 52 42 5 yr OS * P =0. 002, HR 0. 76 Douillard Lancet Oncol. 7: 719, 2006

ANITA: Randomized Phase III Trial of Vinorelbine and Cisplatin vs Obs in resected stage I-III NSCLC Adjuvant Toxicities Neutropenia Gr 3+4 86% Febrile Neutropenia 12. 5% Nausea-Vomiting Gr 3+4 27% Aesthenia 28% Constipation 5% Peripheral Neuropathy 3% Drug-Related Fatality 1% Douillard et al ASCO 2005, A-7013, p 624

LACE: Trials and patients • 5 trials including 4, 584 patients • Median follow-up: 5. 1 years (3. 1 – 5. 9) • 80% male • Median age 59 years, 9% > 70 years old • Pathological Stage: IA: 8%, IB: 30%, II: 35%, III: 27% • Surgery: 31% pneumonectomy • Histology: 49% squamous cell, 39% adenocarcinoma, 12% other Pignon Proc ASCO 2006 abs 7008

Survival curve

CT effect & stage CT may be detrimental for stage IA, but stage IA patients were generally not given the potentially best combination cisplatin+vinorelbine (13% of stage IA patients versus ~43% for other stages)

CT effect & stage CT may be detrimental for stage IA, but stage IA patients were generally not given the potentially best combination cisplatin+vinorelbine (13% of stage IA patients versus ~43% for other stages)

Stage-Specific Hazard Ratios Recent Adjuvant Trials Trial I < 4 cm IALT BR-10 I > 4 cm 1. 73 ANITA 1. 10 0. 79 0. 59 N/A 0. 71 0. 66 IIIA 0. 93 0. 95 II 0. 69 CALGB 1. 02 0. 66 N/A LACE 1. 41* 0. 91* 0. 83 Negative Positive Indeterminate Not studied * 3 cm as cut point

Therapeutic Implications • Short course adjuvant, platinum-based therapy has emerged as standard practice in resected stage Ib-IIIa NSCLC • Ongoing controversies re: – – – – Molecular Selection Influence of Age on Outcome Ideal plating agent: carbo vs cisplatin Choice of partner agent Impact of Stage Role of targeted agents Utility of RT in IIIA (N 2)

Disease Free Patients (%) Potential Benefit from Adjuvant Systemic Therapy 100 Patients with residual micrometastases resistant to adjuvant therapy 80 Patients with residual micrometastases sensitive to adjuvant therapy Predictive Markers ? 60 40 Patients cured with local regional therapy Prognostic Markers ? 20 0 0 2 4 6 Years 8 10

Influence of Age on Outcome

Elderly Specific Analyses: BR 10 Pepe et al ASCO ’ 06, A-7009 • 65: designated cut-off – 327 younger pts (68 %) – 155 older pts (32 %) • Baseline demographics similar except for histology and PS Cohort Adenoca Squamous PS 0 Younger Older P value 58% 32% 53% 43% 59% 42% 0. 001 0. 01

JBR-10 Outcomes by Age • Worse PS in older; fewer PS 0 >65 (53% vs 41%, = 0. 01) • adeno>squam in younger, squam>adeno in older pts. • Patients >65 received significantly less chemo – no significant diff. in toxicity, or growth factor support – more elderly patients refused treatment • OS 46% vs. 66% for obs. vs. chemo in pts >65 – Overall Survival HR 0. 61 [0. 38 -0. 98], p =. 04 in elderly • OS 58% vs. 70% for obs. vs. chemo in pts <65 – Overall Survival HR 0. 77 [0. 54 -10. 9], p = 0. 14 in young • Older patients (>75) – Worse survival regardless of Rx, but same when corrected for disease-specific survival – Benefit from chemo not seen in pts >75 (? harmful) – However, patient numbers are too small to answer clearly Pepe C, et al. Adjuvant vinorelbine and cisplatin in elderly patients: National Cancer Institute of Canada and Intergroup Study JBR. 10. J Clin Oncol. 2007 Apr 20; 25(12): 1553 -61

BR 10: Overall Survival by Age Group 1. 0 0 -65 N = 327 66 -70 N = 84 71 -75 N = 48 >75 N = 23 Probability 0. 8 H-R =2. 38 0. 8 Log-Rank, p =0. 0006 0. 6 63% 0. 4 75 N = 459 >75 N = 23 0. 2 26% 0. 0 0 2 4 6 8 10 12 0 2 4 6 8 10 Time (Years) Pepe C, et al Adjuvant vinorelbine and cisplatin in elderly patients: National Cancer Institute of Canada and Intergroup Study JBR. 10. J Clin Oncol. 2007 Apr 20; 25(12): 1553 -61 12

Ideal Plating Agent

Argument Favoring Carboplatin • The best results obtained in stage IB have been observed with Cb. Pac (not with DDPbased regimens) – Subset analysis in > 4 cm tumors demonstrates a survival benefit • Cb. Pac has not been tested in stage II/IIIA in the adjuvant setting – Absence of data does not prove absence of benefit (…. absence of proof is not proof of absence…. ) • Finally, a substantial percentage of adj pts are poor candidates for cisplatin-based therapy due to age, co-morbidities, etc

Which Agents Partner Best with Platinum

Randomized phase 2 Trial on Refinement of Early stage NSCLC Adjuvant chemotherapy with cisplatin and pemetrexed (CPx) versus cisplatin and vinorelbine (CVb) - TREAT M. Kreuter, J. Vansteenkiste, J. Fischer, W. Eberhardt, H. Zabeck, J. Kollmeier, M. Serke, N. Frickhofen, M. Reck, W. Engel-Riedel, S. Neumann, M. Thomeer, C. Schumann, P. De Leyn, T. Graeter, G. Stamatis, I. Zuna, F. Griesinger and M. Thomas on behalf of the TREAT investigators

Rationale: Dose delivery: Adjuvant CTX LACE-Metaanalysis NCIC-JBR. 10 9% 4. 5% Treatment incomplete 24 % (≤ 2 cycles) 50% (< 4 cycles) • early death or progression 9% 5% • toxicity 34% 13% • patient refusal 35% 29% No treatment Therapy delay 55% Dose reductions 77% TREAT Rationale: • Adjuvant CTX: mainly Cisplatin / Vinorelbine • Need: reduction of toxicity, improvement of dose delivery & compliance • Cisplatin / Pemetrexed in thoracic malignancies: high dose intensity, low toxicities Pignon, JCO, 2008 ; Winton, N Engl J Med, 2005; Alam, Lung Cancer, 2005; Vogelzang, JCO, 2003; Scagliotti, JCO, 2008; Schmid-Bindert, ASCO, 2009

TREAT: Design Inclusion • NSCLC stages IB, IIA, IIB, T 3 N 1 M 0 • ≤ 42 Tage postoperatively, R 0, systematic LN-dissection • ECOG 0, 1 • amenable to Cisplatin treatment Stratification • Center • Nodal status (N 0 versus N 1) • Surgical procedure (lobectomy vs pneumonectomy) Cisplatin / Vinorelbine (CVrb) 50 mg/m 2 d 1+8 / 25 mg/m 2 d 1, 8, 15, 22 q d 29 x 4 R 0 Winton et al. , N Engl J Med (2005) 352: 258 Cisplatin / Pemetrexed (CPx) 75 mg/m 2 d 1 / 500 mg/m 2 d 1 q d 22 x 4

TREAT: Conduct / endpoints Study conduct • Study concept 2005, Inclusion 10/2006 -12/2009 (16 sites, 132 patients) • Treatment until 2/2010, primary endpoint analysis 12/2010 Primary endpoint Clinical Feasibility • No death due to cancer, toxicity, comorbidity • No Non-acceptance by patients leading to premature withdrawal • No observation of DLT ØNeutropenia ØThrombocytopenia ØNon-hematologic toxicity grade 4 >7 d grade 3/4 with fever/infection grade 4 >7 d any grade with bleeding grade 3/4 related to CTX Secondary endpoints Dose delivery, safety, TTTF, RFS, OS, DMFS, LRFS, site of relapse Kreuter et al. , BMC Cancer, 2007

TREAT: Conduct / endpoints Study conduct • Study concept 2005, Inclusion 10/2006 -12/2009 (16 sites, 132 patients) • Treatment until 2/2010, primary endpoint analysis 12/2010 Primary endpoint Clinical Feasibility [considered promising if > 80%] • No death due to cancer, toxicity, comorbidity • No Non-acceptance by patients leading to premature withdrawal • No observation of DLT ØNeutropenia ØThrombocytopenia ØNon-hematologic toxicity grade 4 >7 d grade 3/4 with fever/infection grade 4 >7 d any grade with bleeding grade 3/4 related to CTX Secondary endpoints Dose delivery, safety, TTTF, RFS, OS, DMFS, LRFS, site of relapse Kreuter et al. , BMC Cancer, 2007

TREAT: Characteristics Age (years [range]) Gender (%) • male • female Smoking status (%) • Smoker • Ex-smoker • Non-smoker • Not available CPx CVb Total (n=67) (n=65) (n=132) 58 [40 -73] 60 [38 -74] 59 [38 -74] 72 28 77 23 74 26 33 61 6 0 26 71 1. 5 29 66 4 1 37 12 46 5 38 8 48 6 38 10 47 5 Stage (%) IB IIA IIB T 3 N 1

TREAT : Characteristics CPx CVb Total (n=67) (n=65) (n=132) Lobectomy 84 82 83 Pneumonectomy 12 15 14 Complex resections 4 3 3 Squamous cell carcinoma 45 42 43 Non-squamous 55 58 57 • Adenocarcinoma • Large cell carcinoma • Mixed cell carcinoma 37 9 9 44 9 5 41 9 7 Surgical procedures (%) Histology (%)

Results: Primary endpoint - feasibility CPx 95. 5 Feasibility rate (%) CVb p = 0. 0010 75. 4 (CI 87. 5 -99. 1) Death (%) (CI 63. 1 -85. 2) 1. 5 3. 1 0 6. 2 3. 0 15. 4 patients (n=2) patients (n=10) G 4 neutropenia >7 d 0 4 G 4 thrombocytopenia >7 d G 3/4 febrile neutropenia Thrombocytopenia with bleeding G 3/4 non-hematologic toxicity 0 1 0 2 0 5 0 1 Withdrawal of consent (%) DLT (%) Reasons for DLT (events) * * multiple reasons possible

Results: End of therapy EOT CPx CVb Regular EOT (%) 77. 6 36. 9 Earlier termination of therapy (%) 22. 4 63. 1 patients (n=15) patients (n=41) • Unacceptable toxicity according to protocol** 4 19 • Unacceptable toxicity perceived by patient 6 7 • Relapse of disease 0 2 • Withdrawal of consent 0 4 1 (0) 2 (0) • Non-compliance to protocol 0 2 • Medical decision by investigator 4 5 • Major protocol violation 0 1 • Other reasons 0 4 Reasons for earlier termination (events)* • Death (therapy related) *multiple reasons possible **delay >2 weeks due to toxicity or in case of G 3/4 non-hem toxicity

TREAT: Toxicity Mean Number (AE / SAE) CPx CVb 6. 8 / 0. 3 6. 9 / 0. 2 Hematologic Toxicity G 3/4 (%) 10. 5 p<0. 0001 76. 5 Non-hematologic Toxicity G 3/4 (%) 33 p=0. 7988 31 Hematologic Toxicity (%) G 3/4 Anemia 0 1. 5 Thrombocytopenia 0 0 Neutropenia 9 69 1. 5 6 Febrile Neutropenia

TREAT: Time to treatment failure Time from surgery to withdrawal due to • AE • progression / relapse / death • failure to return to therapy • refusal of treatment / withdrawal of consent Withdrawal probability Tt. TF: p<0. 001

TREAT: Conclusions • CPx safe and feasible Ø less toxicity compared to CVb Ø superior dose delivery compared to CVb Ø high dose density (mg/m 2/week) • Dose delivery failure in CVb mostly due to Vb (delivery d 15, d 22) • Efficacy: longer follow up to be awaited

Molecular Selection

Immunohistochemical (IHC) Staining of the Excision Repair Cross-Complementing 1 (ERCC 1) Protein as Predictor of Benefit from adjuvant chemotherapy (CT) in the International Lung Cancer Trial (IALT) · 761 pts. (28 centers, 14 countries) evaluable for ERCC 1 expression · ERCC 1 repairs cisplatin-DNA adducts, so expression indicates platinum resistance · ERCC 1 a “double-edged sword”; worse prognosis of NSCLC if low expression, but more responsive to platinum Soria, Proc ASCO 2006 A#7010; Olaussen K et al. N Engl J Med 2006; 355: 983 -991

ERCC 1 -Negative: Overall Survival ERCC 1 -Positive: Overall Survival 47% 46% 39% 40% Adjusted HR = 1. 14, 95% CI [0. 84 -1. 55], p = 0. 40 Adjusted HR = 0. 65 95% PI [0. 50 -0. 86], p = 0. 002 Soria, Proc ASCO 2006 A#7010; Olaussen K et al. N Engl J Med 2006; 355: 983 -991

IFCT-0801 TASTE TAilored post-Surgical Therapy in Early stage NSCLC Principal Investigator Jean-Charles SORIA Institut Gustave Roussy - Villejuif Biological Coordinator Marie Wislez Hôpital Tenon - Paris

TASTE design Control Arm CDDP - pemetrexed EGFR mutated Experimental Arm Customized Erlotinib ERCC 1+ Observation ERCC 1 - CDDP-Pemetrexed EGFR wt Non-SCC NSCLC stage II and IIIA (non-N 2)

TASTE Results • • • 150 pts randomized between May 2009 and July 2012, 74 in arm A (PEM/DDP) and 76 in arm B (Selected) Most pts were male (61%), > 60 years (51%), and smokers (91%) Pathological stage was IIA in 69 pt, IIB in 48 pt and IIIA in 32 pt. ERCC 1 was positive in 38 pts (19 in each arm) – only 25%, not 44% expected • EGFR mutation was identified in 10 pts (3 in arm A, 7 in arm B). – Arm A, all pts received CP. – Arm B, • 7 pts received erlotinib, • 53 pts received CP • 16 were observed • Median exposure time to erlotinib was 276 days (10 -365). • Out of 127 pts allocated to CP, 82% received the expected 4 cycles with a very good tolerability profile (no febrile neutropenia). • Success rate was 80% (120 out of 150 pts): appropriate Tx A-7505, ASCO ‘ 13 Soria J-C et al assignment 68

TASTE Conclusions • Met its primary end point for the phase II component, demonstrating feasibility of a national biology-driven trial in the adjuvant setting. • Nevertheless, the phase III was canceled due to the unexpected unreliability of the ERCC 1 IHC read-out. – Current commercial antibodies are unable to distinguish all isoforms of ERCC 1 TASTE Implications • TASTE Needs to be redone with accurate ERCC 1 IHC Ab • May have been responsible for the (-) Made. IT phase III trial in advanced NSCLC (relied on PCR m. RNA probes and AQUA) 69 Soria J-C et al A-7505, ASCO ‘ 13

Role of Targeted Therapy

ECOG 1505: Adjuvant Bevacizumab ELIGIBLE: Resected IB^-IIIA ³Lobectomy No prior chemo No planned XRT No h/o CVA/TIA No ATE w/in 1 yr STRATIFIED: Stage Histology Gender Chemo regimen* R A N D O M I Z E Chemotherapy X 4 cycles Chemotherapy x 4 cycles Plus Bevacizumab X 1 year *Investigator Choice of 4 chemo regimens ^ Revised to exclude IB < 4 cm N >1500; closed to accrual summer 2013

ECOG 4599: Overall Survival Proportion Surviving 1. 0 Median Survival 10. 3 mo PC 1 -year BV/PC survival HR=0. 80; 12. 3 mo 51% vs 44% P=0. 013 2 -year survival 23% vs 15% 0. 8 0. 6 0. 4 0. 2 0. 0 0 6 12 18 24 30 36 Months Sandler, et al. NEJM. 355; 24. Dec 14 2006. 42 48

Chemotherapy Regimens • Therapy to start 6 -12 weeks post-operatively – Investigator Choice of Chemo - 4 cycles (12 wks) • Cisplatin/Vinorelbine – Cis 75 mg/m 2 d 1, Vin 25 mg/m 2 d 1, 8 q 21 d • Cisplatin/Docetaxel – Cis 75 mg/m 2 d 1, Docetaxel 75 mg/m 2 d 1 q 21 d • Cisplatin/Gemcitabine – Cis 75 mg/m 2 d 1, Gem 1250 mg/m 2 d 1, 8 q 21 d • +/- Bevacizumab 15 mg/kg q 21 days x 12 mos

Chemotherapy Regimens: amended 2010 • Therapy to start 6 -12 weeks post-operatively – Investigator Choice of Chemo - 4 cycles (12 wks) • Cisplatin/Pemetrexed – Cis 75 mg/m 2 d 1, Pemetrexed 500 mg/m 2 d 1 • Cisplatin/Vinorelbine – Cis 75 mg/m 2 d 1, Vin 25 mg/m 2 d 1, 8 q 21 d • Cisplatin/Docetaxel – Cis 75 mg/m 2 d 1, Docetaxel 75 mg/m 2 d 1 q 21 d • Cisplatin/Gemcitabine – Cis 75 mg/m 2 d 1, Gem 1250 mg/m 2 d 1, 8 q 21 d • +/- Bevacizumab 15 mg/kg q 21 days x 12 mos

RADIANT Trial: Adjuvant Trial of Erlotinib in NSCLC Stage IB, II, or IIIA NSCLC* Complete surgical resection And subsequent adjuvant chemo No prior or concurrent neoadjuvant or adjuvant N=1654 *Enriched Population: FISH and/or IHC (+) R A 2 N Arm A D Erlotinib qd 2 years O M I 1 Arm B Z Placebo qd 2 years E ٭

RADIANT Trial: Adjuvant Trial of Erlotinib in NSCLC Stage IB, II, or IIIA NSCLC* Complete surgical resection And subsequent adjuvant chemo No prior or concurrent neoadjuvant or adjuvant N=1654 R A 2 N Arm A D Erlotinib qd 2 years O M I 1 Arm B Z Placebo qd 2 years E ٭ ted le mp Co al ru 10 cc A 20 il, pr A *Enriched Population: FISH and/or IHC (+)

IPASS: Progression-free survival in EGFR mutation positive and negative patients EGFR mutation positive EGFR mutation negative Probability of progression-free survival 1. 0 0. 8 0. 6 0. 4 0. 2 0. 0 Gefitinib (n=91) Carboplatin/paclitaxel (n=85) HR (95% CI) = 2. 85 (2. 05, 3. 98) p<0. 0001 No. events gefitinib , 88 (96. 7%) No. events C/P, 70 (82. 4%) Median PFS G, 1. 5 months Median PFS C/P, 5. 5 months Probability of progression-free survival Gefitinib (n=132) Carboplatin/paclitaxel (n=129) HR (95% CI) = 0. 48 (0. 36, 0. 64) p<0. 0001 No. events gefitinib, 97 (73. 5%) No. events C/P, 111 (86. 0%) Median PFS G, 9. 5 months Median PFS C/P, 6. 3 months 0. 2 0 Patients at risk : Gefitinib 132 C/P 129 4 8 12 16 20 24 0. 0 0 4 8 Months 108 103 71 37 31 7 12 16 20 24 1 0 0 0 Months 11 2 3 1 0 0 91 85 21 58 4 14 2 1 Treatment by subgroup interaction test, p<0. 0001 Cox analysis with covariates; HR <1 implies a lower risk of progression on gefitinib; ITT population Fukuoka et al, abstract 8006, ASCO 2009; Mok et al NEJM 2010

Phase II trial SELECT trial: adjuvant erlotinib in resected, early stage NSCLC pts with EGFR mutations PI L Sequist SCREENING PHASE Resected stage IA-IIIA NSCLC Screen tumor for activating EGFR mutations (del 19, L 858 R) N=100 Primary Endpoint: Two year disease-free survival TREATMENT PHASE +/Adjuvant erlotinib adjuvant x 2 years chemo Surveillance CT scans and CTC analysis q 6 mos x 3 years then q 12 mos x 2 years surveillance Biopsy at recurrence, sequence EGFR gene for new mutations, FISH for EGFR and MET copy number Enroll if: screen + or documented EGFR mutation positive –and. No evidence recurrence on baseline CT scans

SELECT Trial: Adjuvant Erlotinib in Resected NSCLC Disease Free Survival

BR 19: Adjuvant Trial of Gefitinib in NSCLC Stage IB, II, or IIIA NSCLC Complete surgical resection No prior or concurrent neoadjuvant or adjuvant N=1242 R A N D O M I Z E ٭ Arm A Gefitinib qd 2 years Arm B Placebo qd 2 years ٭ Modified 2004 to allow adjuvant chemotherapy CAN-NCIC-BR 19, CTSU, ECOG-CAN-NCIC-BR 19, SWOG-CAN-NCIC-BR 19 Protocol. Clinical Trials (PDQ®). At: www. cancer. gov. Commenced October 2002.

* Trend toward impaired outcome for EGFR mutation pts receiving Gef [HR of OS: 1. 51] Goss ASCO 2010 Abstr LBA 7005

ALCHEMIST E 4512 A 081105 A 151216 Target ALK+ EGFRmut Registry Prevalence ~5% ~10% All comers n 336 410 6000 -8000 DFS-OS OS -- Power 80% 85% -- One-sided α 0. 025 0. 05 -- HR 0. 67 -- Peripheral screening for ALK; RTPCR to identify fusion partners Targeted sequence and kinome analysis; PRO and QOL Extended sequencing for additional targets; correlation with local testing Primary Endpt Adjunct

Vaccines

MAGE-A 3 Antigen (melanoma antigen family A, 3) • Truly tumor-specific –Not expressed on normal cells (RT-PCR) –Expressed by various tumor types • Lung • Bladder • Head & Neck • Melanoma 35 -50% 35% 49% 74% • Associated with poorer prognosis (Bolli et al. , 2002; Gure et al. , 2005) • Member of a large family of genes (portfolio)

MAGE A 3 ASCI* randomized phase II • Stage p. IB or p. II: double-blind, randomly assigned 2: 1 to postoperative MAGE-A 3 vaccination or placebo. • Vaccination was started >6 weeks after surgery, with 5 vaccinations at 3 -week intervals, followed by 8 vaccinations every 3 months. • Other anti-cancer adjuvant therapy was not allowed. • Primary endpoint was time-to-recurrence, other endpoints were recurrence rates at different times, and survival. * antigen-specific cancer immune therapeutic Vansteenkiste et al, ASCO 2006, abstract 7019

Safety Status • 182 patients / 1214 MAGE-A 3 doses administered • Well tolerated • Mild grade 1 or 2 toxicities • Local or systemic reactions, 48 hours • 29 grade 3 or 4 adverse events in 21 patients • Three grade 3 events, possibly related to treatment • Leading to withdrawal of 2 patients – (local pain, COPD exacerbation)

Disease-Free Interval HR=0. 73 p=0. 107 (95% CI = 0. 44 - 1. 2) 10% one-sided Vansteenkiste et al, ASCO 2006, abstract 7019

Efficacy Endpoints Overview Final analysis 05 Oct, 2006 Cox regression model (95% confidence interval) P-value from Cox regr. model adjusted for stratification covariates HR with a 10% one-sided 0. 73 Disease-Free Interval HR = 0. 73 (0. 44 -1. 20) 0. 73 Disease-Free Survival HR = 0. 73 (0. 45 -1. 16) 0. 66 Overall Survival 0. 00 2. 00 0. 50 « MAGE-A 3 » better HR = 0. 66 (0. 36 -1. 20) 1. 00 1. 50 Hazard ratio « Control » better P=0. 107

MAGE Trial Design Resectable NSCLC Surgery Pathological stage IB, IIIA Chemotherapy (up to 4 cycles platinum based chemotherapy) No chemotherapy R MAGE-A 3 +AS 15 Placebo MAGE-A 3 +AS 15 MAGRIT Trial R Placebo

MAGRIT: Phase III • • • Largest lung cancer study EVER Began in October 2007 Goal: 2270 patients from 400 centers in 33 countries in Europe, North and South America, Asia, Australia • 2289 ultimately enrolled

Role of Adjuvant RT in Stage II and Stage IIIA

Should the pt receive adj RT? 1) 2) 3) Yes No Maybe

Adjuvant Radiotherapy: Meta-analysis 1998 • Individual data from 9 randomized trials including 2128 patients • Treatment details (staging, surgery, RT) highly variable among series • PORT: better local control: 29% fewer local recurrences - 195 LR vs 276 LR for no RT • Overall HR = 1. 21 (1. 08 -1. 34) ~ survival decrement of 7 % at two years (55% vs 48%) • Increase risk greater for early stage patients(Stage I/II vs. III) Lancet 25 July 1998

PORT Meta-analysis Survival Curves Stewart et al Lancet 1998

PORT - Heterogeneity of Hazard • No increased risk for patients with N 2 disease • Patients with the least to gain have the most to lose Stewart et al Lancet 1998

PORT Meta-analysis Methodologic Flaws • Variable and unspecified staging • Variable and unspecified interval between resection and PORT • Inadequate RT – Suboptimal doses; large fields – Poor treatment planning – Outmoded techniques (e. g. : use of low-energy photons or 60 Co for a substantial proportion of patients) • • • Inclusion of N 0 patients Unpublished data (2 of 9 studies) Relatively short F/U (< 4 yrs) Stewart et al Lancet 1998

Risks of PORT with Modern Technology • Retrospective review – 202 patients treated with surgery and PORT for Stage II and III disease – Median dose 55 Gy – Actuarial rate of death from intercurrent disease was 13. 5% compared to expected rate of 10% Machtay et al JCO 2001

ANITA TRIAL: N 2 Disease – Influence of RT

ANITA TRIAL: N 2 Disease – Influence of RT RT Effect? Or Serendipity?

ANITA - PORT Evaluation • • • PORT: 33% on obs, 22% on chemo For all Chemo > XRT = chemo/XRT > 0 For N 2 Chemo/XRT > chemo > XRT > 0 XRT Chemo No No No Yes Yes All pts MST 26 mo 93 mo 50 mo 46 mo N 2 MST 13 mo 24 mo 23 mo 47 mo Rosell, IASLC 11, Abs Pr 3, 2005

Plot of overall survival for N 2 patients stratified by postoperative radiotherapy (PORT) use – SEER data Lally, B. E. et al. J Clin Oncol; 24: 2998 -3006 2006

PORT Conclusions • PORT has no role in N 0 or N 1 disease • Role of PORT in N 2 is controversial – Recent subset and retrospective analyses hint at benefit – Ongoing “Lung ART” trial in France • • • 700 pts with resected N 2 randomized to PORT or not Adjuvant chemo allowed 1 st Accrual sluggish

“Lung ART” P. I. Dr Cécile Le Pechoux Completely resected N 2 NSCLC Primary end-point: DFS (Sample size: 700 patients) S U R G E R Y Conformal RT 54 Gy/27 -30 fxs No post-op RT Pre or post-op chemotherapy allowed Concomitant chemo not allowed Sponsors: FNCLCC, IFCT, LARS-G, EORTC

Conclusions: Adjuvant Therapy • Adjuvant Platinum-based Chemotherapy is the Standard of Care for Resected Stage II-IIIA NSCLC – Improves OS 5%-15% at 5 years with newer drugs • Fit elderly patients (< 75 yrs) benefit as much as younger patients • Ongoing trials with molecularly determined Tx, erlotinib, bevacizumab, vaccines • Controversies – – Benefit in IB Neoadjuvant vs adjuvant therapy Which chemotherapy to use PORT