Targeted Therapy for Lung Cancer: The Greater Impact in the New Era

Targeted Therapy for Lung Cancer: The Greater Impact in the New Era Fadlo Raja Khuri, MD Professor and Roberto C. Goizueta Chair Department of Hematology & Medical Oncology Deputy Director Winship Cancer Institute Georgia Cancer Coalition Professor Debates and Didactics in Hematology and Medical Oncology Sea Island, Georgia, July 23, 2015

Targeted Therapies vs Immunotherapy for Lung Cancers Opportunities to specifically target drivers of oncogene addicted lung cancers Much higher response rates with more durable benefits Readily available and reliable biomarker in genomic testing Far better cost/benefit ratio Better understood mechanisms of resistance

Therapeutic Plateau in Metastatic NSCLC? ECOG 1594 Patient Survival (%) 1.0 0.8 0.6 0.4 0.2 Cisplatin/Paclitaxel Cisplatin/Gemcitabine Cisplatin/Docetaxel Carboplatin/Paclitaxel 0.0 0 5 10 15 20 25 30 Time (mos) ECOG=Eastern Cooperative Oncology Group. Schiller JH, et al. N Engl J Med. 2002;346:92-98.

Positions of Mutations Detected in EGFR Tyrosine Kinase Domain in NSCLC EGF ligand binding Tyrosine kinase Autophosphorylation TM K DFG Y Y Y Y 718 745 776 835 858 861 869 964 Exon: Paez: GXGXXG K R H DFG L L Y 18 19 20 21 22 23 24 747-750 719 757-750 858 Lynch: G719 L858 Pao: Tumor with point mutation (amino acid substitution) Tumor with in-frame deletion Activation loop EGF=endothelial growth factor; TM=transmembrane. Adapted from Pao et al. Proc Natl Acad Sci U S A. 2004;101:13306. Lynch et al. N Engl J Med. 2004;350:2129. Paez et al. Science. 2004;304:1497.

IPASS: PFS in EGFR Mutation Positive and Negative Patients EGFR mutation positive EGFR mutation negative Probability of progression-free survival At risk : Gefitinib C / P 1.0 0.8 0.6 0.4 0.2 0.0 Gefitinib (n=132) Carboplatin / paclitaxel (n=129) HR (95% CI) = 0.48 (0.36, 0.64) P <.0001 No. events gefitinib, 97 (73.5%) No. events C / P, 111 (86.0%) RR- 71%; 47% 0 4 8 12 16 20 24 Months 132 108 71 31 11 3 0 129 103 37 7 2 1 0 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 <.0001 No. events gefitinib, 88 (96.7%) No. events C / P, 70 (82.4%) RR- 1.1%; 23% 0 4 8 12 16 20 24 Months 91 21 4 2 1 0 0 85 58 14 1 0 0 0 ITT population Cox analysis with covariates Mok TS, et al. N Eng J Med. 2009;361:947-957. Treatment by subgroup interaction test, P <.0001

First- Line TKI Therapy in EGFR Mutated NSCLC Randomized Phase III Trials Study IPASS (Mok TS, et al. N Engl J Med. 2009;361:947-957.) WJTOG3405 (Mitsudomi T, et al. Lancet Oncol. 2010; 11:121-128.) Maemondo M, et al. N Engl J Med. 2010; 362:2380-2388. OPTIMAL (Zhou C, et al. Lancet Oncol. 2011;12:735-742.) EURTAC (Rosell R, et al. Lancet Oncol. Jan 25, 2012 [Epub ahead of print].) N Treatment Arm Control Arm 1217 Gefitinib Carboplatin/ Placitaxel 177 (M+) Gefitinib Cisplatin, Docetaxel 230 (M+) Gefitinib Carboplatin, Paclitaxel 165 (M+) Erlotinib Carboplatin/ Gemcitabine 153 (M+) Erlotinib Platinum-based chemotherapy Stage IIIB/IV IIIB/IV Median PFS 5.7 vs 5.8 months (HR for EGFR mutated pts 0.48; HR for nonmutated pts 2.84) 9.2 vs 6.3 months (P < 0.001) IIIB/IV 10.8 vs 5.4 months (HR 0.3, P < 0.0001) IIIB/IV 13.6 vs 4.6 months (HR 0.16, P < 0.0001) IIIB/IV 9.4 vs 5.2 months (HR, 0.42, P < 0.0001) Median OS 18.6 vs 17.3 months (P = NS) 30.5 vs 23.6 months (P = NS) 22.9 vs 18.8 months (P = 0.42) PFS is superior, no survival advantage for first line TKI- effect of cross-over Indication First-line First-line First-line First-line First-line

EGFR-TKIs: Secondary Resistance 37 patients re-biopsied at the time of progression 6/18 patients with T790M had other molecular abnormalities 5 patients had SCLC phenotype Sequist L, et al. Sci Transl Med. 2011;3:75ra26.

T790M in Acquired Resistance Acquired exon 20 mutation found in >50% of patients with acquired resistance to TKI Increases relative affinity of mutant EGFR for ATP, may also cause steric hindrance to erlotinib T790M More likely to show progression in lungs/pleura Less commonly detected in CNS May have better prognosis than non-t790m Oxnard GR, et al. Clin Cancer Res. 2011;17:1616-1622.

Phase I Dose Escalation / Expansion Global Study Design First-line cohorts objective assessment of the safety and tolerability of AZD9291 when given orally as first-line therapy to patients who are treatment-naïve for locally advanced or metastatic EGFRm positive NSCLC Escalation Rolling six design Cohort 1 Cohort 2 Cohort 3 Cohort 4 Cohort 5 20 mg 40 mg 80 mg 160 mg 240 mg Expansion Enrollment by local identification of EGFR mutation followed by central laboratory confirmation (cobas EGFR Mutation Test) or by central laboratory testing alone Positive Positive Positive Positive Positive Negative Negative Negative First-line EGFRm* 80mg Biopsy Tablet First-line EGFRm* 160mg Biopsy T790M cohorts *Prior therapy for advanced NSCLC not permissible in this cohort Data cut-off April 15, 2015 Data from cohorts in grayed out boxes are not included in the analyses reported here Cytology Ramalingam SS, et al. J Clin Oncol. 2015;33(suppl): Abstract 8000.

AZD9291: T790M+ (central test; N=138) 20 mg 40 mg 80 mg 160 mg 240 mg Confirmed ORR in patients with centrally tested T790M+ tumours was 61% (78/127; 95% CI 52%, 70%) Disease control rate (CR+PR+SD) was 95% (121/127; 95% CI 90%, 98%) CI, confidence interval; ORR, overall response rate; CR, complete response; PR, partial response; SD, stable disease; D, discontinued *Imputed values for patients who died within 14 weeks (98 days) of start of treatment and had no evaluable target lesion assessments. Patients are eligible for confirmed response if they have two post-baseline RECIST assessments or patients who withdraw/die prior to the second RECIST assessment. Data cut-off 1 Aug 2014 Janne et al, N Engl J Med, 2015

Response Rate in First-line Cohorts by Dose 50 40 30 20 10 0-10 -20-30 -40-50 -60-70 -80-90 -100 D D D D D D Best percentage change from baseline in target lesion D D 80 mg N=30 160 mg N=30 D D Total N=60 80 mg 160 mg * D Objective response rate # 63% (95% CI 44, 80) 83% (95% CI 65, 94) 73% (95% CI 60, 84) Disease control rate 93% (95% CI, 78, 99) 100% (95% CI 88, 100) 97% (95% CI 89, 100) Best objective response Complete response # Partial response # Stable disease Progressive disease 0 19 9 2 1 24 5 0 1 43 14 2 Population: evaluable for response, data cut-off April 15, 2015 Response Evaluation Criteria In Solid Tumors version 1.1 (RECIST 1.1), programmatically calculated from investigator-recorded tumor measurement *Considered a complete response because target lesion <10mm (lymph node) with complete response in non-target-lesions and no new lesions #Confirmed responses only CI, confidence interval; CR, complete response; D, discontinued; DCR, disease control rate; PR, partial response; SD, stable disease Ramalingam SS, et al. J Clin Oncol. 2015;33(suppl): Abstract 8000.

CO-1686: A third-generation EGFR Inhibitor Presented By Lecia Sequist at 2014 ASCO Annual Meeting

Best Response to Rociletinib (All Doses) in 256 Centrally Confirmed Tissue T790M+ Patients 100 SLD Change from Baseline (%) 80 60 40 20 0 20 40 500mg 625mg 750mg 1000mg Total N 50 124 78 4 256 ORR (%) 60 54 46 75 53 DCR (%) 90 84 82 100 85 ORR, objective response rate; DCR, disease control rate 500mg BID HBr 625mg BID HBr 750mg BID HBr 1000mg BID HBr + Ongoing 60 80 100 SLD, sum of longest diameters *3 patients currently have no evaluable baseline lesions per database and are omitted from this analysis 14

Acquired Resistance to Rociletinib is Heterogeneous Longitudinal quantitative analyses of EGFR T790M and activating mutations in plasma MGH/Stanford analysis of primary/acquired* rociletinib resistance (N=13) T790-WT SCLC (n=2) T790M-positive NSCLC (n=4) T790-WT NSCLC (n=4) T790Mpositive NSCLC (EGFR- *Acquired resistance = progression after initial tumor shrinkage; primary resistance = PD at first assessment scan Piotrowska Z, et al. Cancer Disc. 2015;5(7):713-722.l

ALK Fusion Gene Potent oncogenic activity Present in approximately 4-5% of NSCLC More common in Never-smokers Adenocarcinoma Signet ring morphology Soda et al, Nature, 2007; Shaw et al, J Clin Oncol, 2009.

Clinical and Demographic Features of Patients With ALK-positive NSCLC Median age Is younger 30% smokers N=149 N = 136 Median (range) age, years 52 (21 86) 52 (29-82) Gender, male/female 49/51 47/53 Race, n (%) White 95 (64) 87 (64) Asian 41 (28) 43 (32) Others 13 (9) 6 (4) Smoking history, n (%) Never smoker 106 (71) 92 (68) Former smoker 42(28) 39 (29) Current smoker 1 (1) 5 (4) Histology, n (%) Adenocarcinoma 144 (97) 130 (96) Squamous 2 (1) 0 (0) Prior treatment regimens, n (%) Other 3 (2) 6 (4) 0 24 (6) 1 47 (33) 2 31 (18) 3 47 (41) 1. Camidge DR, et al. Lancet Oncol. 2012;13:1011-1019. 2. Crino D, et al. J Clin Oncol. 2011;29(suppl. Abstract 7514).

Crizotinib: Best Percent Change from Baseline in Target Lesions* % Decrease or increase from baseline 100 80 60 40 20 0 20 40 60 80 100 Progressive disease Stable disease Partial response Complete response Objective response details (all evaluable patients) N=116** ORR (95% CI) 61% (52, 70) Median response duration Median time to response 48 weeks 8 weeks Disease control rate at 8, 16 weeks 79%, 67% *excludes patients with early death and indeterminate response (n=106) **includes patients with early death and indeterminate response (n=116) Camidge et al, ASCO 2011

Mechanisms of Crizotinib Resistance A B Doebele et al. CCR, 2012

Best % change from baseline 100 Marked Activity of LDK378 in Patients with Advanced ALK+ NSCLC 80 60 40 20 0 20 40 60 Best % change from baseline in target lesions LDK378 400 750 mg/day Prior crizotinib Crizotinib- naïve 80 PFS event 100 ORR: 60% in crizotinib-naïve patients and 57% in crizotinib-treated patients

Other ALK Inhibitors Alectinib 1 Response rate of > 90% in Crizotinib-naïve patients Activity demonstrated in crizotinib-resistant patients with RR of ~ 50% AP26113 Dual ALK and EGFR inhibitor Active in patients with secondary resistance to ALK 1. Gadgeel et al, WCLC 2013, Abstract O16.06

ROS1 Rearrangements in NSCLC TPM3-ROS1 SLC34A2-ROS1 SDC4-ROS1 CD74-ROS1 Present in ~1% of NSCLC cases (also found in some GBMs and cholangiocarcinomas) Enriched in younger never or light smokers with adenocarcinoma histology No overlap with other oncogenic drivers LRIG3-ROS1 EZR-ROS1 ROS1 Bergethon et al., JCO 30(8): 863-70, 2012; Takeuchi et al., Nat Med 18(3): 378-81, 2012

Summary of Tumor Responses in Patients with Advanced ROS1+ NSCLC (N=14*) 100 Decrease or Increase From Baseline (%) 80 60 40 20 0 20 40 60 80 100 15+ 16+ 18+ 4+ 12+ PD SD PR CR 8+ 22+ 18 44+ 20+ 35+ 48+

BRF113928: Study Design Single arm, phase 2, open label NSCLC BRAF V600E mutation N = 20 Stage 1 N = 20 Stage 2 1 line prior tx Dabrafenib 150 mg twice daily Primary objective: Investigator-assessed ORR Secondary objectives: PFS, duration of response, overall survival (OS), safety, tolerability, and population pharmacokinetics Green-Dahlberg 2-stage: H(0) ORR 10% versus H(1) ORR 30% (primary cohort)

Maximum Percent Reduction at Time of Best Disease Assessment Maximum Reduction of Sum of Lesion Diameters by Best Confirmed Response for First 20 Patients 50 40 30 20 10 0 10 20 30 40 50 60 70 80 90 100 *** *** *** Best Confirmed Response Partial response Stable disease Progressive disease * ** ** ** Smoking History * ** *** ** Nonsmoker Smoker, 40 pack years Smoker, > 40 pack years * ** * ** * ** ** *

BRF113928: Study Design Cohort A (monotherapy) n = 60 Stage IV NSCLC BRAF V600E ECOG 0-2 At least 1 platinum-based chemotherapy Dabrafenib 150mg BID Stage 1 N = 20 Stage 2 N = 20 Expansion N = 20 COMPLETE Reported at ESMO 2014* Cohort A (combination D+T) n = 40 Interim futility analyses Stage IV NSCLC BRAF V600E ECOG 0-2 1-3 tx lines only (at least 1 platinum-based chemotherapy) Dabrafenib 150mg BID Trametinib 2mg once daily Stage 1 N = 20 Stage 2 N = 20 Recruitment stopped if insufficient clinical activity is observed in first 20 subjects ONGOING > Passed futility LSFV 14JAN2015 *Planchard D, et al. Ann Oncol 2014;25(suppl 4):abstract LBA38_PR.

Maximum Reduction of Sum of Lesion Diameters By Best Confirmed Response in 2nd Line (N = 24) Maximum Percent Reduction at Time of Best Disease Assessment 20 10 0-10 -20-30 -40-50 -60-70 -80-90 -100 Best Confirmed Response PR SD PD The median duration of response was not reached

Duration of Treatment for All Enrolled Patients in the Interim Analysis (n = 33) * Best Unconfirmed Response Complete Response Partial Response Stable Disease Progressive Disease Not Evaluable Not Available First complete response First partial response Disease progressed Still on study treatment 0 1 2 3 4 5 6 7 8 9 *1 st -line patient (protocol deviation) Treatment Duration (Months)

Nivolumab Overall Survival 100 90 80 mos mo, (95% CI) Nivolumab n = 135 9.2 (7.3, 13.3) Docetaxel n = 137 6.0 (5.1, 7.3) 70 # events 86 113 OS (%) 60 50 40 30 1-yr OS rate = 42% HR = 0.59 (95% CI: 0.44, 0.79), P = 0.00025 Nivolumab 20 10 0 0 3 6 9 1-yr OS rate = 24% 12 15 18 Docetaxel 21 24 Time (months) Number of Patients at Risk Nivolumab Docetaxel Symbols represent censored observations 135 113 86 69 52 31 15 7 0 137 103 68 45 30 14 7 2 0

Pembrolizumab Potent, humanized monoclonal antibody against PD-1 that prevents PD-1 from binding to its ligands, PD-L1 and PD-L2 Robust antitumor activity and manageable toxicity in multiple tumor types As monotherapy for previously treated and treatment-naive advanced NSCLC in KEYNOTE-001 (n = 495) 1 19.4% ORR, 3.7-month median PFS, 12.0-month median OS 45.2% ORR in patients with PD-L1 expression in 50% of tumor cells 1. Garon EB et al. N Engl J Med. 2015;372:2018-20.

Rationale For Anti PD-1 + Anti CTLA-4 Combination Therapy Complementary mechanisms of action CTLA-4 functions at the activation stage of the anticancer immune response PD-1 functions at the effector stage of the anticancer immune response Preclinical data: synergistic anticancer activity shown in multiple models 1-3 Clinical data: nivolumab (anti PD-1) + ipilimumab has shown significant efficacy and manageable toxicity in advanced melanoma 4,5 and NSCLC 6 1. Curran MA et al. Proc Natl Acad Sci U S A 2010;107:4275-80; 2. Duraiswamy J et al. Cancer Res. 2013;73:3591-603; 3. Selby M et al. J Clin Oncol 2013;31:Abstr 3061; 4. Wolchok JD e al. NEJM. 2013;369:122-133; 5. Postow MA et al. N Engl J Med. 2015;doi: 10.1056/NEJMoa1414428; 6. Antonia SJ et al. J Clin Oncol 2014;32:Abstr 8023.

Adverse Events of Special Interest (Any Frequency) Adverse Event, n (%) Total N = 18 Any 6 (33) Adrenal insufficiency (grade 3) 1 (6) Autoimmune thyroiditis (grade 2) 1 (6) Hypothyroidism (grade 1) 1 (6) Hyperthyroidism (grade 2) 1 (6) Maculopapular rash (grade 3) 1 (6) Myasthenia gravis (grade 2) 1 (6) Myocarditis (grade 2) 1 (6) Pneumonitis (grade 1) 1 (6) Thyroiditis (grade 2) 1 (6) Uveitis (grade 2) 1 (6) Analysis cutoff date: March 31, 2015.

Best Overall Response (RECIST v1.1, Investigator Review) ORR, n (%) [95% CI] DCR, n (%) [95% CI] Pembro 10 mg/kg Q3W + IPI 1 or 3 mg/kg N = 6 3 (50) [12-88] 6 (100) [54-100] Pembro 2 mg/kg Q3W + IPI 1 mg/kg N = 12 4 (33) [10 65] 9 (75) [43 94] Total N = 18 7 (39) [17 64] 15 (83) [59 96] Best overall response, n (%) CR 1 (17) 0 (0) 1 (6) PR 2 (33) 4 (33) 6 (33) SD 6 weeks 3 (50) 5 (42) 8 (44) PD 0 (0) 3 (25) 3 (17) Analysis cutoff date: March 31, 2015.

Change From Baseline in Tumor Size (RECIST v1.1, Investigator Review) Change From Baseline, % 100 80 60 40 20 0-20 -40-60 -80-100 -120 Pembrolizumab 10 mg/kg + ipilimumab 1 or 3 mg/kg Pembrolizumab 2 mg/kg + ipilimumab 1 mg/kg 71% of patients showed decrease in target lesion burden Change from baseline was evaluated in patients with 1 postbaseline tumor assessment. Analysis cutoff date: March 31, 2015.

Treatment Exposure and Response Duration * * * * * * * CR as best response PR as best response PD as best response Last pembrolizumab dose Treatment ongoing The length of the bars is equivalent to the time to the last disease assessment by the investigator. Analysis cutoff date: March 31, 2015. * 0 6 12 18 24 30 36 42 48 Time, weeks * * All responses ongoing at the time of data cutoff Response duration: 6+ to 39+ weeks Pembrolizumab 10 mg/kg + ipilimumab 1 or 3 mg/kg Pembrolizumab 2 mg/kg + ipilimumab 1 mg/kg

Targeted Therapies vs Immunotherapy for Lung Cancers Opportunities to specifically target drivers of oncogene addicted lung cancers Much higher response rates with more durable benefits Readily available and reliable biomarker in genomic testing Far better cost/benefit ratio Better understood mechanisms of resistance Far lower toxicity than with immunotherapy, especially combined CTLA-4/PD-1 blockade