A phase II precision medicine cancer trial NCI-Molecular Analysis for Therapy Choice (NCI-MATCH / EAY131) Overview A phase II precision medicine cancer trial Co-developed by the ECOG-ACRIN Cancer Research Group and the National Cancer Institute Version Date: 10/13/2017

On Behalf of the NCI-MATCH Study Leadership Keith T. Flaherty1, Alice P. Chen2, Peter J. O'Dwyer3, Barbara A. Conley2, Lyndsay N. Harris2, James V. Tricoli2, Stanley R. Hamilton4, Mickey Williams5, Robert J. Gray6, Shuli Li6, Lisa M. McShane6, Lawrence V. Rubinstein2, Susanna I. Lee1, Constantine A. Gatsonis7, Lalitha K. Shankar8, Paolo F. Caimi9, Shaji Kumar10, Edith P. Mitchell11, James A. Zwiebel2, A. John Iafrate1, Jeffrey Sklar12, Carlos L. Arteaga13 1Massachusetts General Hospital, Boston, MA; 2National Cancer Institute (NCI), Division of Cancer Treatment and Diagnosis, Bethesda, MD; 3University of Pennsylvania, Philadelphia, PA; 4MD Anderson Cancer Center, Houston, TX; 5 NCI Frederick National Laboratory for Cancer Research, Frederick, MD; 6Dana-Farber Cancer Institute, Boston, MA; 7Brown University, Providence, RI; 8NCI Cancer Imaging Program, Rockville, MD; 9Case Western Reserve University, Cleveland, OH; 10Mayo Clinic, Rochester, MN; 11Thomas Jefferson University, Philadelphia, PA; 12Yale University, New Haven, CT; 13Vanderbilt University, Nashville, TN

Central screening cohort New process to identify patients Contents Trial introduction Brief history Central screening cohort New process to identify patients Important details Future plans

Precision Medicine BRAFV600E melanoma responds to BRAF inhibitors or BRAF inhibitors combined with MEK inhibitors; but colorectal cancers with the same mutations appear not to respond Various tumors with NTRK fusions appear to respond to NTRK inhibitors (ASCO 2017) We need to know more about which tumors will respond to agents targeted to “driver” mutations Most driver mutations are relatively rare

What is NCI-MATCH?

NCI-MATCH Eligibility Overview Adults ≥ 18 years of age Solid tumor, lymphoma, or myeloma Progressed following at least one line of standard systemic therapy OR type of cancer for which no therapy exists that has been shown to prolong overall survival Patients who cannot receive other standard therapy due to medical issues may be eligible Rare diseases for which there is no standard therapy

NCI-MATCH Objective To determine whether matching certain drugs or drug combinations in adults whose tumors have specific gene abnormalities will effectively treat their cancer, regardless of the cancer type A signal-finding trial Treatments that show promise can advance to larger, more definitive trials

Levels of Evidence for Target Selection in NCI-MATCH Level 1: Gene variant credentialed for selection of an approved drug Level 2: Variant is eligibility criterion for an ongoing clinical trial for that drug OR variant identified in an N of one response 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 preclinical model

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

NCI-MATCH’s ~1100 Trial Locations Nationwide In every state, the District of Columbia, and Puerto Rico

Central screening cohort New process to identify patients Contents Trial introduction Brief history Central screening cohort New process to identify patients Important details Future plans

Brief History of NCI-MATCH Opened on August 12, 2015, with 10 treatment arms and a goal to have 3000 patients submit tumor samples for central testing 795 patients were screened in the first three months Reached >100/week by the end of this period Far surpassed the original estimate of 50 screens/month Paused enrollment of new patients in November 2015, for a planned interim analysis This resulted in a more accurate estimation of the prevalence rates for the targeted mutations

Brief History of NCI-MATCH (cont’d) Resumed enrollment of new patients in May 2016, with 24 treatment arms and more laboratory capacity to handle rapid pace of enrollment Increased central screening goal to 6000 patients in late 2016 Added myeloma patients in late 2016 Expanded to 30 treatment arms in March 2017 Completed central screening of ~6000 patients in July 2017, nearly two years ahead of schedule

Central screening cohort New process to identify patients Future plans Contents Trial introduction Brief history Central screening cohort New process to identify patients Future plans

NCI-MATCH Central Screening Summary # Registered for Screening # With Samples Submitted # With Testing Complete # Assigned to Rx # Enrolled on Rx 6397 5962 5559 992 689 as of 09/17/2017

Estimated Central Screening Rates in NCI-MATCH Cases with samples/# enrolled 93% Assay successful/# with samples 94% Assigned to Rx/Assay successful 19% Enrolled on Rx/Assigned 70% as of 09/17/2017

NCI-MATCH Wait Times for the ~6000 Tested Patients Processing Step Median Calendar Days Tumor sample submission from sites to EA central lab at MD Anderson Cancer Center 7 Completion of tumor testing by lab network and return of results to site 16 Further eligibility evaluation for patients assigned to a treatment arm 14 as of 09/17/2017

NCI-MATCH Screening Accrual as of 09/17/2017 Screened % of Total Less Common Cancers 3476 62.5% Common Cancers* 2083 37.5% Total 5559 Far exceeded the goal of 25% Common cancers screened: Colorectal 15.3%, Breast 12.4%, NSCLC 7.3%, Prostate 2.5%

NCI-MATCH Less Common Cancers as of 09/17/2017 Disease Site Screened % of Total Screened (N=4694) Assigned to Rx % of Screened Ovarian 530 9.5 75 14.2 Uterine 345 6.2 90 26.1 Pancreas 341 6.1 21 Cholangiocarcinoma 154 2.8 40 26.0 Liver and Hepatobiliary other than Cholangio. 104 1.9 20 19.2 Sarcoma 255 4.6 31 12.2 Head and Neck 215 3.9 47 21.9 Neuroendocrine 186 3.3 26 14.0 Gastroesophageal 177 3.2 49 27.7 Central Nervous System 94 1.7 35 37.2 Bladder/Urinary Tract 1.6 32 35.6 Cervical 88 25 28.4 Small Cell Lung 80 1.4 4 5.0 Melanoma 77 27.3 Kidney 68 1.2 10 14.7 Anal 46 0.8 11 23.9 Mesothelioma 9 19.6 Lymphoma 0.7 2 Myeloma -- Other 540 9.7 100 18.5 Less Common Cancers 3476 62.5 648 18.6

NCI-MATCH Central Screening What is the biggest change that occurred once the trial reached its 6000-patient central screening goal? Patients do not need a new biopsy to determine eligibility for the trial because the trial laboratory network no longer collects and tests specimens centrally NCI ended reimbursement for these procedures

NCI-MATCH Central Screening 40% of the treatment arms (12/30) reached the accrual needed for the primary outcome analysis Some arms targeting more common gene variants were expanded to accommodate the higher numbers of patients with matches Several arms did not complete accrual within the ~6000 patients given the low frequency of some of the gene variants Final complete data is pending publication

NCI-MATCH Important Discovery In the ~6000 patients tested, the tumor gene variants we are studying occurred less frequently than expected in this study population, ranging from 3.47 percent to zero

NCI-MATCH Major Change How did the discovery that gene abnormalities are rarer than expected change the study? It shed light on the fact that in order to fill the ‘rare variant’ arms, we need to look at tens of thousands of patients It required setting a new goal for the trial: Broaden the base of screened patients to complete those arms that target rare tumor gene abnormalities Rare Variant: Estimated frequency rate of ≤ 1.5% in central screening cohort

NCI-MATCH 19 ‘Rare Variant’ Treatment Arms, By Gene Abnormality Prevalence Rate of ≤ 1.5% Z1C CDK4 or CDK6 1.36 S2 GNAQ/GNA11 0.16 M TSC1 or TSC2 1.11 E EGFR T790M 0.11 Z1B CCND1/2/3 0.84 V cKIT R BRAF fusions 0.80 L mTOR 0.31 Y AKT 0.77 G ROS1 0.05 H BRAF V600 E/K 0.69 F ALK 0.03 U NF2 loss Z1E NTRK 0.10 C2 MET exon 14 sk 0.61 A EGFR activating C1 MET amplif. 0.51 X DDR2 0.00 T SMO/PTCH1 0.42

NCI-MATCH Accrual Per Arm is Posted Publicly ecog-acrin.org/trials/nci-match-eay131

Central screening cohort New process to identify patients Contents Trial introduction Brief history Central screening cohort New process to identify patients Important details Future plans

New Process for Identifying NCI-MATCH Patients Effective May 11, 2017 (Addendum #10): Began a process of finding potentially eligible patients through routine genomic testing done by designated commercial and academic labs that have the capacity to test large numbers of patients Now, the only pathway for new patients to enroll is through routine testing by a designated lab Applies to patients at ~1100 participating trial sites Labs seeking matches to the 19 ‘rare’ variant arms

Commercial Labs Collaborating with NCI-MATCH Caris Life Sciences® and Foundation Medicine, Inc. Will send a written referral email or letter to physicians at any of the ~1100 participating sites Notifying them that the genomic test they ordered to guide clinical care for their cancer patient has found a gene mutation that could make the patient eligible for one of the treatment arms

Selection Criteria for NCI-MATCH Designated Labs Trial leaders vetted labs for their capabilities to: Test a high volume of patients Identify rare variant subsets using a comprehensive and highly-validated assay that considers the specific inclusion and exclusion criteria for each treatment arm Provide results in a format that can be uploaded into MATCHbox, the trial’s informatics system that generates treatment arm assignments for patients

How Patients Move Thru the Trial Process Physician orders routine test to guide patient’s clinical care (test is not specifically for NCI-MATCH) Lab runs its standard test and sends results to physician Lab flags case because patient has tumor gene abnormality and disease characteristics that possibly qualify them for the trial Lab sends written referral email or letter to physician Physician considers referral Physician informs patient of possible trial match that must be confirmed by enrolling in NCI-MATCH for screening (Step 0) Patient decides to pursue and consents to screening (Step 0); physician completes eligibility checklist to confirm they are a candidate for screening Cont’d next slide

How Patients Move Thru the Trial Process (cont’d) 8. Physician enrolls patient to Step 0 with required lab referral Trial team completes Step 0 and informs physician whether patient will or will not be assigned to a treatment arm Patient receives assignment, understanding their physician must evaluate them for treatment arm eligibility before they can enroll Patient decides to pursue and consents to treatment (Step 1); physician evaluates patient for the arm, and enrolls if eligible Once enrolled, patient will be treated for as long as their tumor shrinks or remains stable

Central screening cohort New process to identify patients Contents Trial introduction Brief history Central screening cohort New process to identify patients Important details Future plans

More Details About Testing Physicians do not order tests specifically for NCI-MATCH This is standard testing done outside of the trial and prior to any involvement in the study The patient does not need a fresh biopsy for this trial The labs are collaborating to help identify patients with rare tumor gene abnormalities Because these tests are performed as part of routine care planning, the NCI does not reimburse for them

More Details About the Lab Referral Due to the specific inclusion and exclusion criteria for each treatment arm in NCI-MATCH: The information in the standard testing report is not sufficient to determine trial eligibility, regardless of its content Physicians must have the written lab referral to consider the trial as an option for their patient The lab will send the written referral email or letter to a physician separate from the standard report

Questions About Commercial Lab Referrals Physicians should contact the designated lab with any questions on the referrals they receive Physicians may also reach out to the lab about the potential availability for a particular patient http://ecog-acrin.org/nci-match-eay131-designated-labs

Academic Labs Collaborating with NCI-MATCH Two clinical laboratories are designated to test their own patients using their institutional gene profiling assays done in the setting of clinical care The University of Texas MD Anderson Cancer Center Memorial Sloan Kettering Cancer Center These labs also do not specifically test for NCI-MATCH Again, the NCI does not reimburse for testing

More About NCI-MATCH Eligibility Testing A few treatment arms require patients to have additional testing to find certain gene abnormalities, beyond the NGS testing done by the designated lab Example: immunohistochemistry (IHC) testing If a patient needs additional testing for eligibility, the study team will notify the trial site to submit archived tissue to the central trial lab

Confirmation Testing Using the NCI-MATCH Assay The patient’s matching tumor gene abnormality on the outside assay will be confirmed on the NCI-MATCH assay by the central trial lab Using archived tissue Patients do not need a fresh biopsy Patient treatment will not be delayed for confirmation testing

Archival Tissue for Confirmation Testing Preferably from the same sample that was tested by the designated lab Trial site should submit archival tissue for confirmation after receipt of a formal treatment assignment If the trial team requested archival tissue earlier to determine eligibility (e.g. IHC testing), that tissue will also be used for confirmation testing and sites will not need to submit additional tissue unless requested

NCI-MATCH Outside Assay Demonstration Project Trial leaders are testing out this new approach in a demonstration project with the 19 rare variant arms and the four initial labs The project is helping to study the feasibility of this approach and to determine the potential for it to become a model for use in other precision medicine trials

Central screening cohort New process to identify patients Contents Trial introduction Brief history Central screening cohort New process to identify patients Important details Future plans

Plans to Expand NCI-MATCH to Additional Labs An upcoming protocol addendum, if approved, will allow more laboratories to be designated for testing of their own patients, using their institutional assays Massachusetts General Hospital NCI Molecular Characterization (MoCha) Laboratory Yale University Cancer Center

Plans to Expand NCI-MATCH to Add’l Labs (cont’d) NCI is seeking letters of interest from CLIA-certified/ accredited labs that test tumor specimens from patients utilizing NGS assays Federal Register Notice, Vol. 82, No. 147. Pgs. 35976-35978 Posted: August 2, 2017 Deadline: January 31, 2018

NCI-MATCH Outcomes Reporting Reporting of ORR (primary endpoint), PFS, TTP, OS, and toxicity on individual arms will occur once complete response and toxicity data are available for at least 31 patients per arm If the ORR is ≥ 5/31 (16%), the agent will be worthy of further study as defined by the master protocol 35 patients need to be accrued per arm to obtain 31 for that arm (assuming a 10% ineligibility rate) Need ~eight months of follow-up after accrual is complete Secondary analyses will examine response by disease histology, individual variants, and other factors

NCI-MATCH Effect of Outside Testing on Analysis Patients confirmed by the NCI-MATCH assay will be included in the primary outcome analysis Patients without confirmation may continue treatment, but will not be included in the primary outcome analysis

Prevalence Rate in NCI-MATCH NCI-MATCH New Treatment Arms in Development, for Possible Activation in Late 2017* Mutation Prevalence Rate in NCI-MATCH Arm Drug PIK3CA 3.47% Z1F Follows arm I copanlisib PTEN loss by IHC 1.93% Z1G Follows arm P PTEN mutation and expression by IHC 1.75% Z1H Follows arm N FGFR amplification 1.86% K1 Follows arm W erdafitinib FGFR mutation or fusion 1.00% K2 TP53 mutation and MYC amplification unavailable Z1J AZ1775 * Patient entry via designated labs

Prevalence Rate in NCI-MATCH NCI-MATCH New Treatment Arms in Development, for Possible Activation in Spring 2018* Mutation Prevalence Rate in NCI-MATCH Arm Drug AKT mutations 0.77% Z1K Follows arm Y ipatasertib Non-V600 BRAF mutations 0.80% Z1L Follows arm R ulixertinib (BVD-523) * Patient entry via designated labs

NCI-MATCH Resources Main Webpages: ecog-acrin.org/nci-match-eay131 cancer.gov/nci-match Protocol Documents: ctsu.org (password required) Spanish: cancer.gov/espanol/nci-match Email Inquiries: match@jimmy.harvard.edu NCI Contact Center: 1-800-4-CANCER and cancer.gov/contact