NCI-Molecular Analysis for Therapy Choice (NCI-MATCH or EAY131) A phase II precision medicine cancer trial Co-developed by the ECOG-ACRIN Cancer Research Group and the National Cancer Institute Version Date: 09/29/2015

NCI-MATCH Hypotheses Primary: Tumors that share common somatic genetic alterations in oncogenes will be variably responsive to therapies targeting the oncogenic pathway based on lineage specific factors.  Secondary: Concomitant somatic genetic alterations will predict responsiveness or resistance. 09/29/2015

Matching patients to therapy on the basis of genetic features in lung cancer: erlotinib in EGFR mutant NSCLC and crizotinib in ALK translocated NSCLC Rizvi N et al. CCR 2011 Camidge R et al. Lancet Oncol 2012 09/29/2015

BRAF inhibitor therapy markedly more effective for V600EBRAF melanoma compared to colon cancer Sosman J et al. NEJM 2012 Kopetz, ASCO 2010 melanoma colorectal 09/29/2015

Decoding the cancer genome: ERBB2 (HER2) Breast cancer Prostate cancer Gastric cancer Chromophobe kidney cancer Bladder cancer Liver cancer Uterine cancer 09/29/2015

Mutation Frequency in TCGA Gene Bladder Breast Cervix Colo-rectal GBM Glioma Head and Neck Lung (Adeno) Lung (Squam) Melanoma Ovarian Pancreas Prostate Stomach Thyroid Uterine AKT 2.4 0.9 0.4 0.7 0.6 1.4 3.5 1.6 BRAF 0.8 2.6 9.9 2.1 1 9.6 4.5 50.2 1.8 4.1 61.3 2.8 EGFR 1.5 26.1 5.9 4.7 14.3 3.9 6.5 2.2 1.2 5 3.2 FGFR1 3.1 0.3 1.3 1.1 1.7 3.6 FGFR2 2.3 9.3 0.2 12.5 FGFR3 14.6 2.9 2 HRAS 4.6 IDH1 4.9 76.1 5.7 0.5 IDH2 4.2 KIT 2.7 6.9 KRAS 43 32.6 57.9 11.4 21.4 NF1 9.2 7.7 6.6 11.8 11.9 13.3 4.4 8.1 NF2 5.1 NRAS 9 30.8 8.5 PIK3CA 20 35.1 23.1 20.1 10.6 20.9 15.7 21.8 53.2 PIK3R1 4 11.3 33.1 PTCH 7.1 4.8 2.5 1.9 PTEN 31.9 7.9 8.2 5.2 6.4 64.9 SMO TSC1 10.7 3.4 TSC2 Compiled by MD Anderson Investigators 09/29/2015

Amplification Frequency in TCGA Gene Bladder Breast Cervix Colo-rectal GBM Glioma Head and Neck Lung (Adeno) Lung (Squam) Melanoma Ovarian pancreas Prostate Stomach Thyroid Uterine HER2 6.3 12.9 2.3 3.1   2.6 2.2 0.6 13 5.5 FGFR1 9.4 12 1.2 3.6 0.4 8.5 3.5 16.9 0.3 4 2 1 2.5 FGFR2 0.5 1.8 0.7 0.9 5.1 FGFR3 0.8 1.3 1.5 7.9 0.2 MET 8.9 1.1 3.9 4.1 PIK3CA 4.9 19.3 2.9 21.1 38.2 28.8 2.8 14.3 Compiled by MD Anderson Investigators 09/29/2015

OncomineTM Focus Assay Gene List Hotspot genes, n=35 Copy number variants, n=19 Fusion drivers, n=23 AKT1 ALK AR BRAF CDK4 CTNNB 1 DDR2 EGFR ERBB2 ERBB3 ERBB4 ESR1 FGFR2 FGFR3 GNA11 GNAQ HRAS IDH1 IDH2 JAK1 JAK2 JAK3 KIT KRAS MAP2K1 MAP2K2 MET MTOR NRAS PDGFRA PIK3CA RAF1 RET ROS1 SMO FGFR3 FGFR4 KIT KRAS MET MYC MYCN PDGFRA PIK3CA ALK AR BRAF CCND1 CDK4 CDK6 EGFR ERBB2 FGFR1 FGFR2 ABL1 AKT3 ALK AXL BRAF ERBB2 EGFR ERG ETV1 ETV4 ETV5 FGFR1 FGFR2 FGFR3 MET NTRK1 NTRK2 NTRK3 PDGFRA PPARG RAF1 RET ROS1 RNA Panel 52 unique genes DNA Panel 269 amplicons in DNA panel 272 amplicons in RNA panel 09/29/2015

NCI-MATCH Rationale Molecularly targeted therapy has improved outcomes Within individual tumor types imatinib in CML (bcr-abl) imatinib in GIST (CKIT & PDGFRα) erlotinib in NSCLC (EGFR) crizotinib in NSCLC (EML4-ALK) And across tumor types trastuzumab in breast & gastric (HER2) vemurafenib in melanoma, thyroid & NSCLC, but not colon cancer (BRAF) 09/29/2015

NCI-MATCH Objective To understand the relative efficacy of the same therapy applied to oncogene-defined subsets across different tumor histologies, we propose to initiate a broad-based genomic prescreening study to assign patients whose tumors harbor specific molecular abnormalities to relevant targeted treatments, regardless of tumor histology type 09/29/2015

NCI-MATCH Design Features Test many patients to find widely distributed genetic alterations 3,000 to find 1,000 Biopsies needed at time of study entry (cost covered by NCI) Response rate (tumor regression) will be primary efficacy measure Across 22 approved/in-development arms PIs drawn from: 37% ECOG-ACRIN, 30% Alliance, 16% SWOG, 16% NRG More than 150 NCI and NCTN members of 10 subcommittees Advocates involved in trial design and help oversee conduct 09/29/2015

NCI MATCH Eligibility Defined Molecularly Initial tumor biopsy to identify mutations/amplifications/ translocations Patients can be screened with local NGS but results must be confirmed on NCI-MATCH assay Patient assignment to relevant agent(s)/subprotocol Perform tumor biopsies and sequencing at progression to illuminate resistance mechanisms Submit de-identified samples to central labs Conduct whole-exome, mRNA sequencing (research purposes) 09/29/2015

NCI-MATCH Schema 09/29/2015

NCI-MATCH Eligibility Patients with solid tumors or lymphomas whose disease has progressed following at least one line of standard systemic therapy – or with tumors that do not have standard therapy Exclude histologies that had been approved by the FDA or had shown lack of efficacy with an agent Tumor accessible to biopsy and patient willing to undergo biopsy Adults ≥ 18 year of age ECOG performance status ≥ 1 Adequate organ function 09/29/2015

NCI-MATCH Patient Population Considerations Target: at least 25% of total enrollment to be patients who have “rare” tumors “Common” defined as breast, NSCLC, colon, prostate 09/29/2015

Statistical Considerations for Each Molecularly-Defined Arm Primary Endpoint: Overall Response Rate 5% vs 25% Secondary Endpoints: Progression Free Survival (PFS) 6 months 15% (median PFS 2.2 m) vs 35% (median PFS 4 m) TTP Toxicity Biomarker One stage design with 35 evaluable patients per arm 09/29/2015

Rules for Treatment Selection in NCI-MATCH Three assay components are equal (SNV/indel; CNV; fusion) Select actionable mutation of interest (aMOI) winner in each component Level of evidence 1,2 > 3 If > 1 LOE 2,3, assign to arm with fewer patients SNV/indels: LOI 1,2 > 3; IF > 1 LOE 1,2 If difference between VAF > 15%, choose greater If difference < 15%, choose arm with fewer patients If 2 or more aMOI meet LOE 1,2, choose arm with fewer patients Compare winner of each component LOE 1,2 > 3 If > 1 LOE 1,2 select arm with fewer patients 09/29/2015

aMOIs in NCI MATCH and Estimated Prevalence 09/29/2015

NCI-MATCH Subprotocols - Activated August 2015 Molecular Target Estimated % Prevalence Agent(s) for Subprotocol ID EGFR activating mutations 1 - 4 Afatinib EAY131-A HER2 activating mutations 2 - 5 EAY131-B EGFR T790M mutations and rare activating mutations of EGFR 1 - 2 AZD9291 EAY131-E ALK translocations 4 Crizotinib EAY131-F ROS1 translocations 5 EAY131-G BRAF V600E or V600K mutations 7 Dabrafenib and trametinib EAY131-H HER2 amplification Ado-trastuzumab emtansine EAY131-Q BRAF fusions, or non-V600E, non-600K BRAF mutations 2.8 Trametinib EAY131-R NF2 loss 2 Defactinib EAY131-U cKIT mutations Sunitinib EAY131-V 09/29/2015

NCI-MATCH Subprotocols – Expected December 2015 Molecular Target Estimated % Prevalence Agent for Molecular Target Subprotocol ID PIK3CA mutations or amplifications, but without RAS mutations or PTEN loss 17-18 Taselisib EAY131-I mTOR mutations 5 MLN0128 EAY131-L TSC1 or TSC2 mutations 2.6-3.5 EAY131-M PTEN mutations or deletions, with PTEN expression on IHC 11 GSK2636771 EAY131-N PTEN loss by IHC EAY131-P NF1 mutations 7.7 Trametinib EAY131-S1 GNAQ or GNA11 mutations 2 and 1.6 EAY131-S2 Smoothened (SMO) or patched 1 (PTCH1) mutations 2.6 - 3.8 Vismodegib EAY131-T DDR2 mutations 2 Dasatinib EAY131-X 09/29/2015

NCI-MATCH Subprotocols – Expected April 2016 Molecular Target Estimated % Prevalence Agent for Molecular Target Subprotocol ID MET amplifications 4 Crizotinib EAY131-C Aberrations in FGFR pathway 5 AZD4547 EAY131-W AKT mutations 1 - 10 AZD5363 EAY131-Y 09/29/2015

Levels of Evidence for Drugs in NCI-MATCH Level 1: FDA approved for any indication for that target Level 2: Agent met a clinical endpoint (objective response, PFS, or OS) with evidence of target inhibition Level 3: Agent demonstrated evidence of clinical activity with evidence of target inhibition at some level 09/29/2015

Levels of Evidence for Target Selection in NCI-MATCH Level 1: Gene variant credentialed for selection of an approved drug Level 2a: Variant is eligibility criteria for an ongoing clinical trial for that drug Level 2b: Variant identified in an N of 1 response(s) Level 3: Preclinical inferential data Models with variant respond; without variant do not Gain of function mutation demonstrated in preclinical model Loss of function (tumor suppressor genes or pathway inhibitor e.g. NF1); stop codon or demonstrated loss of function in pre-clinical model 09/29/2015

Requirements for New NCI-MATCH Subprotocol Proposals Rationale: Why proposed treatment should be effective in a particular molecular subgroup – NEED evidence in patients Preliminary data: Levels of Evidence for drug(s) AND target At least 2% estimated prevalence of aMOI across refractory solid tumors, lymphomas Proposal for measurement of the molecular eligibility criterion(a) Recommended phase 2 dose Willingness of company 09/29/2015

Genetic Platform and Laboratory Network in NCI-MATCH Ion Personal Genome Machine® (PGMTM) System and Ion TorrentTM Server Ion AmpliseqTM custom DNA panel About 200 genes SNV, indel, CNV, targeted translocations Selected IHC, FISH Network of 4 CLIA-approved molecular diagnostics laboratories provides capacity NCI Molecular Characterization Laboratory (Dr. Mickey Williams) Plus competitively chosen lab sites: MD Anderson (Dr. Stanley Hamilton), Massachusetts General (Dr. John Iafrate), Yale (Dr. Jeffrey Sklar) Validation within and across all four labs: same SOP 09/29/2015

Tumor Biopsy in NCI-MATCH Upon entry to initial screening (Step 0) a biopsy (4 cores) FFPE, shipped to MDACC Adjacent section H&E stained, examined by pathologist for tumor content, % necrosis, inflammation, and scanned into high resolution image database RNA and DNA extracted from the same tissue section Planned research assays: If sufficient DNA available, whole-exome sequencing performed for research RNA will be used for research grade gene expression profiling by either whole- transcriptome or miRNA microarray analysis Biopsy and sequencing on progression 09/29/2015

NCI-MATCH Assay Workflow: 10-14 Day Turn-Around Time Biopsy Received Tissue Fixation Path Review Tumor content >70% 1 DAY Nucleic Acid Extraction DNA/RNA yields > 20 ng 1 DAY Library/Template Prep Library yield > 20 pM Test fragments Total read Reads per BC Coverage NTC, Positive, Negative Controls 1 DAY Sequencing , QC Checks Centralized Data Analysis 1 DAY aMOIs identified 3 DAYS Clinical Laboratory aMOI Verification 10-14 days Rules Engine Treatment Selection 09/29/2015

NCI-MATCH Logistics Master protocol with multiple ph 2 subprotocols with molecularly defined eligibility Activated on CTSU on August 12, 2015 with 10 subprotocols Expected to expand to 19 by December 2015 and to 22 by April 2016 IND for protocol template Subprotocols can be opened or closed without affecting others Single agents or combinations - recommended phase II dose is known FDA approved (for a different indication) or investigational agents/combinations Central IRB required as the IRB of record US-based sites across NCTN and NCORPs CLIA lab network: Validated assay 09/29/2015

Team Approach to NCI-MATCH NCI-MATCH Steering Committee Agent & Gene Selection Committee Vetting of actionable genetic alterations and most robust agents Informatics Committee Imaging Committee Specimen/Assay Committee Developed and validated next-generation sequencing platform Additional IHC or other assays developed at ECOG-ACRIN Central Biorepository and Pathology Facility at MDACC 09/29/2015

NCI-MATCH Steering Committee Clinical Study Chairs: Drs. Alice Chen, Keith Flaherty, Peter O'Dwyer Scientific Chairs: Drs. Barbara Conley, Stanley Hamilton, Mickey Williams, Carlos Arteaga Statistical Chairs: Drs. Robert Gray, Shuli Li, Lisa McShane, Larry Rubenstein Safety Chairs: Drs. Edith Mitchell, James Zwiebel Informatics Chairs: Warren Kibbe, Jose Galvez, Susan Lemont, Mark Routbot 09/29/2015

Additional Expertise for NCI-MATCH Leadership for each individual subprotocol From all NCTN groups From outside NCTN groups Young to mid career investigators 09/29/2015

NCI-MATCH Overview NCI-MATCH is a signal-finding trial Largest, most rigorous precision oncology trial in history Conducted by ECOG-ACRIN but has PIs on subprotocols from across NCTN Target selection based on levels of evidence Coordinated sample collection/pre-analytics Validated, standardized gene sequencing through MATCH Box Using NCI-MATCH to inform other trials 09/29/2015

Resources for NCI-MATCH Main Webpages: cancer.gov/nci-match ecog-acrin.org/nci-match-eay131 Protocol Documents: ctsu.org (password required) Spanish: cancer.gov/espanol/nci-match Email Inquiries: match@jimmy.harvard.edu Patient Brochure: EA website (above) Site Process Brochure: Coming soon NCI’s Cancer Information Service: 1-800-4-CANCER and cancer/gov/contact This slide presentation is updated regularly. For the latest version, visit ecog-acrin.org. 09/29/2015

Questions? 09/29/2015