Dr. Ofer Purim Chief Gastrointestinal Malignancy Service Assuta Samson Hospital

1. Molecular profiling-selected treatment in metastatic Gastroesophageal cancer: can we change our standard of practice? Dr. Ofer Purim Chief Gastrointestinal Malignancy Service Assuta Samson Hospital

2. Disclosure Caris Life Sciences - consultant fees

3. Gastric & Esophageal cancer • Gastric cancer is the fifth most common malignancy and the third leading cause of cancer death worldwide. • Esophageal cancer is the eighth most common malignancy and the sixth leading cause of cancer death worldwide. In 2012, there were approximately, 1.4 million new cases of gastric/esophageal cancers and 1.1 million deaths from these cancer types.

4. At present, there is no internationally-accepted consensus regarding standard-of-care in the 1st-line metastatic setting in gastric/esophageal cancer, except for 10-25% of patients with gastric/gastroesophageal junction (GEJ) cancer whose tumors are human epidermal growth factor receptor 2 (HER2)-positive.

5. Chemotherapy regimens used in gastric cancer treatment cisplatin-based oxaliplatin-based CF/FP XP ECF ECX FLP EOX EOF CAPOX/XELOX FLO FOLFOX taxane-based DCF mDCF FLOT paclitaxel docetaxel TPC FOLFIRI irinotecan + cisplatin 5-FU/LV Xeloda S-1 irinotecan monotherapy irinotecan-based

6. The ToGA trial: A phase III study of trastuzumab added to standard chemotherapy in first-line human epidermal growth factor receptor 2 (HER2)-positive advanced gastric cancer

7. OS in IHC2+/FISH+ or IHC3+ (exploratory analysis) MedianOS 16.011.8 1.0 Event Events 120136 HR 0.65 95% CI 0.51, 0.83 0.9 FC + T 0.8 0.7 FC 0.6 0.5 0.4 0.3 0.2 11.8 16.0 0.1 0.0 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 Time (months) No. at risk 228 218 218 198 196 170 170 141 142 112 12296 100 75 84 53 65 39 51 28 39 20 28 13 20 11 12 4 11 3 5 3 4 0 1 0 0 0

8. Change Clinical Practice in Gastroesophageal cancer with Caris Multiplatform Molecular Profiling

9. The concept of Precision Medicine • Identify biomarkers that predict treatment outcomes • Biomarker guided treatments lead to patient benefit Young woman with hedgehog mutation: complete remission under vismodegib Von Hoff, NEJM 2009

10. Caris Molecular Intelligence

11. Precision – high quality IHC is key success factor RUO antibodies do not perform Western blots of two batches of commercially available anti-TLE3 antibodies Tissue microarray on every slide to verify staining CMI is approved by NYS-DOH, CAP, CLIA, ISO15189:2012, State Licenses and Fully Reimbursed in the US and some other countries

12. Precision Oncology Alliance Manifesto • Develop standards, best practices, and guidelines • Develop value based reimbursement paradigm • Optimize clinical decisions to improve patient care • Generate and publish clinical outcomes data • Conduct research studies • Educate physicians on the latest developments in molecular profiling • Create awareness within the patient community • Partner with payors, regulators, and guideline committees

13. Caris Molecular Intelligence – How it Works Tissue Sample Sent to Caris Biomarker Testing Informatics Analysis Tumor ProfileGenerated Physician Reviews Report A tissue sample of a tumor is obtained through biopsy. A pathology laboratory packages and ships to Caris. Caris analyzes tumor tissue through a series of advanced pathology and panomic assays to identify biomarkers (DNA, RNA and Proteins). The Caris expert team interprets the data using more than 120,000 clinical papers to match a tumor’s biomarkers to potential treatment options. Results are organized into a report that includes relevant findings and treatment options. The report is sent to ordering physician for review. Physician may suggest a treatment based on the report, availability of drugs, medical history and other clinical factors.

14. Traditional Molecular Profiling Caris Next Generation Molecular Profiling Pharma-Driven Companion Diagnostic Orientation AI-Driven Personalized Cancer Signature for Each Patient • One-to-one association of specific biomarker assay to drug • Large clinical trials required to demonstrate association • Limited to lineages included in clinical trials • Challenges with assay & drug access, especially outside U.S. • Changes in assay technology quickly drive obsolescence of older tests • Individual drug recommendations don’t match real world, multi-drug treatment regimens • Based on multi-platform molecular assessment of tens of thousands of patients with real clinical outcomes • Use of Big Data and machine learning techniques • Identify molecular signatures based on drug response • Recommendations for existing drug regimens – not just new drugs

15. PFS is getting shorter in subsequent lines of therapy 87% 79% 70% 51% Adapted from Bailey et al. (2012) Progression-free survival decreases with each subsequent therapy in patients presenting for Phase I Clinical Trials. Journal of Cancer 2012 3:7-13

16. Bisgrove Study Primary Endpoint Compare progression free survival (PFS) for therapy selected by molecular profiling with PFS for the last line of therapy on which the patient progressed. PFS: length of time during and after treatment in which a patient is living with a disease that does not get worse. PFS Last prior therapy PFS Selected by MP Period A Period B If PFSb/PFSa ratio was > 1.3, MP-selected therapy was defined as having benefit for patient. Temple, R. Clinical Measurement in Drug Evaluation. Ningano W. Thicker GT, eds. John Wiley and Sons Ltd: 1995; Von Hoff, D.D. 1999; Dhani et al. Clinical Cancer Research. 2009; 15: 118-123.

17. Bisgrove study

18. Cohort Definitions for Monotherapies and Doublet Combinations Based on Predictive Association with Biomarker Results Marshall et al. (2015) Panomics validation of time to next treatment (TNT) as a surrogate outcome measure in 4729 cancer patients. J Clin Oncol 34, 2016 (suppl; abstr 11521)

19. Overall Survival is higher in Matched Cohort • Matched Cohort (n=534) • Median OS 1068 days • Median 3.2 treatments after profiling • Unmatched Cohort (n=493) • Median OS 646 days • Median 4.2 treatment after profiling 646 days 1068 days Spetzler D et al. (2015) Multi-platform molecular profiling of 1,180 patients increases median overall survival and influences treatment decision in 53% of cases. Presented at ECCO 2015

20. Median Matched TNT is 15% (or 33 days) Longer in Caris POA Data • Matched (n=3,011) • Median TNT = 248 days • Unmatched (n=1,718) • Median TNT = 215 days • HR = 0.85 (p=0.00018) Marshall et al. (2015) Panomics validation of time to next treatment (TNT) as a surrogate outcome measure in 4729 cancer patients. J Clin Oncol 34, 2016 (suppl; abstr 11521)

21. CMI-Guided treatment leads to gain of PFS

22. 78% of Profiled Patients Are Treated In Line with CMI Report 78% 69%* 79% 95% 71% 100% 61% 55% 79% 79% 89% Metastatic Adenoid Cystic Carcinoma 9Popovtzer et al. Refractory Solid Cancers 13Seeber et al. Refractory Solid Cancers 14El Ahmadie et al. Refractory Gastric Cancers 10Purim et al. Refractory Pancreatic Cancers 6Ramanathan et al. Refractory Solid Cancers 16Chahine et al. Refractory Solid Cancers 8Dean et al. Refractory Breast Cancers 7Jameson et al. Refractory Pancreatic Cancers 5Epelbaum et al. Refractory Solid Cancers 2Von Hoff et al. * Still enrolling

23. 43% of CMI-Guided Treatments result in Clinical Benefit (improving trend) 55%* 65% 27% 38% 44% 60% 41% 34% 72% 29% Refractory Solid Cancers PFS Ratio ≥ 1.3 2Von Hoff et al. Refractory Solid Cancers PFS Ratio ≥ 1.3 13Seeber et al. Refractory Solid Cancers RECIST Response 14El Ahmadie et al. Refractory Gastric Cancers PFS Ratio ≥ 1.3 10Purim et al. Refractory Pancreatic Cancers OS > 6 months 6Ramanathan et al. Refractory Solid Cancers PFS Ratio ≥ 1 16Chahine et al. Refractory Solid Cancers PFS Ratio ≥ 1.3 8Dean et al. Refractory Breast Cancers PFS Ratio ≥ 1.3 7Jameson et al. Refractory Pancreatic Cancers PFS Ratio ≥ 1.3 5Epelbaum et al. Metastatic Adenoid Cystic Carcinoma CR/PR/SD>6 months 9Popovtzer et al. 43% * Still enrolling

25. MethodsStudy design and patient population • Study period between January 2010 and March 2014 • The study was approved by the institutional review boards of the participating institutions: • Rabin Medical Center • Sourasky Medical Center • Rambam Health Care Campus • Hadassah Hebrew University Medical Center • Kaplan Medical Center • Wolfson Medical Center

26. Results:Patients characteristics Deomgraphics

27. Results:Disease characteristics

28. Results:Treatment regimens received prior to molecular profiling

29. Results:Actionable biomarkers Only markers that were tested in samples of at least 35 patients are included in the table. ERCC1, excision repair cross-complementation 1; MGMT, O-6-methylguanine-DNA methyltransferase; RRM1, ribonucleotide reductase M1 subunit; TOPO1, topoisomerase 1; TOP2A, topoisomerase IIA; TS, thymidylate synthase.

30. Results:Patients treated according to MP

31. Results:Patients achieving the PFS ratio endpoint

33. Conclusions In summary, this study shows in real-life clinical practice that implementing MP is feasible and provides clinical benefit therapy (PFS ratio of ≥1.3) for a close to a third of patients with metastatic gastric/esophageal cancer. Prospective studies are warranted.

34. I thank the patients and the families. My co-authors And Teva and Caris THANK YOU