Nadine Ezer Clinical Trial Monitor  Roche Site Initiation Meeting – May 23, 2013 Montreal General Hospital, McGill CRP Nadine Ezer Clinical Trial Monitor  Roche

22-Apr-17 JACOB Study (BO25114) A double-blind, placebo-controlled, randomized, multicenter Phase III Study evaluating the efficacy and safety of pertuzumab in combination with trastuzumab and chemotherapy in patients with HER2-positive metastatic gastroesophageal junction and gastric cancer

The global status Study Timelines & Status Completed 1st Site activation (Mar/Apr) Upcoming 780 subjects recruited Study Milestones Data Point 1st Investigator meeting (Mar) FPI (Apr 2013) LPI (May 2016) LSLO 2012 2013 2014 - 2015 2016 2017 2018 - 2022 Final Protocol (27 Sep) 1st HA/IRB approval (Dec) Final Primary Analysis (Jun 2017) Interim Analysis 70% OS event (Sep 2016) 1st Interim Safety Analysis (Oct 2013) 2nd Interim Safety Analysis (Feb 2014)

OUR 2013 SMT (GDT) Goal for the JACOB Study Enrollment of 110 patients of the total 780 patients required by 31 December 2013

Other Teams Supporting the Study Central Laboratory IXRS Central HER2 Testing

Overview of Gastric/ GEJ cancer

Gastric/GEJ Adenocarcinoma Increasing incidence of adenocarcinoma of gastric cardia/GEJ Multifactorial etiology (obesity, gastroesophageal reflux, diets high in salted/smoked foods, FH, H. pylori, tobacco and alcohol use) 65% diagnosed with regional/distant disease3 Median survival < 1 year – better treatments are needed! 4th most common cancer diagnosis (1,000,000 new cases each year) 1 3rd leading cause of cancer death in males globally (700,000 deaths) 2 >70% of new cases and deaths occur in developing countries 2 – GLOBOCAN 2008 Jemal et al., CA CANCER J CLIN 2011;61:69–90 Howlader SEER data 2001-7 6. cardia (including gastroesophageal junction)

Gastric or Stomach Cancer Anatomy of the stomach Esophagus Diaphragm Proximal Lower esophagus sphincter Fundus Gastric or Stomach Cancer GE junction = cardia Lesser curvature Body (Corpus) Pylorus Distal Antrum Small intestine Greater curvature Williams, et al. Gray’s Anatomy 1995

Siewert classification of adenocarcinoma of the GEJ Type II true cardia Esophageal adenocarcinoma Type I distal oesophagus 5 cm 1 cm Z line GEJ adenocarcinoma 2 cm 5 cm Type III subcardia Gastric adenocarcinoma Type Features Topographic anatomical criteria Type I Adenocarcinoma of the distal oesophagus, which usually arises from an area with specialised intestinal metaplasia of the oesophagus (i.e. Barrett oesophagus) and may infiltrate the oesophago-gastric junction from above The centre of the cancer or more than two thirds of identifiable tumour mass is located >1 cm proximal to the gastro-oesophageal junction Type II True carcinoma of the cardia arising immediately at the oesophagogastric junction The centre of the cancer or the tumour mass is located in an area extending 1cm proximal to the gastro-oesophageal junction to 2cm distal to it Type III Subcardial gastric carcinoma that infiltrates the oesophagogastric junction and distal oesophagus from below The centre of the cancer or more than two thirds of identifiable tumour mass is located >2cm below the gastro-oesophageal junction GEJ = gastro-esophageal junction

The incidence of gastric cancer is highest in Asia Mortality Eastern Asia 60 50 40 30 20 10 10 20 30 40 50 60 Central and Eastern Europe Less developed regions World South America More developed regions Central America Southern Europe Western Asia South-Eastern Asia Caribbean Polynesia Melanesia Western Europe Northern Europe Micronesia Australia/New Zealand South-Central Asia Middle Africa Eastern Africa Northern America Western Africa Southern Africa Northern Africa Age-standardised incidence (per 100,000) Male Female Parkin DM, et al. CA Cancer J Clin 2005; 55:74–108.

Gastric cancer clinical presentation/RX Abdominal pain, weight loss, obstruction, GI bleed Limited disease (resectable without distant mets) Surgical resection of primary tumor and LN with perioperative chemotherapy and XRT Locally advanced (unresectable)/metastatic Chemotherapy – differs by geographic region, - usual regimens: ECF, EOF, ECX, TC Biologic therapy (eg. Trastuzumab, Met inhibiting agents) Palliative surgery (e.g. bypass) Palliative radiotherapy

HER2 positive cancer Breast cancer 20% HER2 positive Trastuzumab with chemotherapy improves OS in adjuvant and metastatic settings Gastric cancer 20% HER2 positive Trastuzumab with chemotherapy improves OS in metastatic settings (ToGA) HER2 positive rates may be slightly lower in Asian populations (19 vs. 15%) Different guidelines for interpreting HER2+ in gastric vs. breast cancer Breast cancer methods result in underscoring of HER2 in gastric ca New HER2 targeted therapies Lapatinib: oral TKI of HER1 and HER2 approved for met breast ca TYTAN trial in second line HER2+ (by FISH) GC lapatinib did not prolong OS LOGiC trial in first line HER2+ GC Pertuzumab: IV monoclonal antibody targets HER2 T-DM1: IV antibody drug conjugate targets HER2 Bang et al Lancet 2010 / Hofmann et al Histopath 2008 / Ruschoff et al Mod Path 2012

ToGA trial design Phase III, randomized, open-label, international, multicenter study Primary endpoint OS 5-FU or capecitabine a + cisplatin (n=290) 3807 patients screened 810 HER2-positive (22.1%) (IHC 3+ or FISH +) HER2-positive advanced GC (n=584) R 5-FU or capecitabine a + cisplatin + trastuzumab (n=294) Stratification factors a Chosen at investigator’s discretion 1 Bang et al; Abstract 4556, ASCO 2009 GEJ, gastroesophageal junction Bang Lancet 2010 Locally Advanced (inoperable) vs. Metastatic GC vs. GEJ Measurable vs. Non-measurable ECOG PS 0-1 vs. 2 Capecitabine vs. 5-FU 13

ToGA (BO18255) The results of the ToGA study (Herceptin®/trastuzumab plus fluoropyrimidine-cisplatin chemotherapy) established a new standard of care for treatment of HER2 positive AGC*: Adding trastuzumab to chemotherapy prolongs survival Median OS duration improved from 11.1 to 13.8 months in the full analysis population Median OS duration improved from 11.8 to 16.0 months in the high HER2 overexpressing population (tumors that are IHC3+ or IHC2+/ISH+) * AGC = Advanced Gastric Cancer Defined as Inoperable, locally advanced or metastatic gastric/GEJ adenocarcinoma

K-M Plot of Overall Survival in ToGA Study (full analysis population) 1.0 Median Events OS F/X + C + H 167 13.8 0.8 F/X + C 182 11.1 0.6 HR = 0.74 95% CI (0.60, 0.91) p value = 0.0046 13.8 Event 11.1 0.4 0.2 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 Time (Months) No.at risk F/X + C + H 294 277 246 209 173 147 113 90 71 56 43 30 21 13 12 6 4 1 F/X + C 290 266 223 185 143 117 90 64 47 32 24 16 14 7 6 5 H, Herceptin / Bang YJ., et al., Lancet 2010; 376: 687-697

K-M Plot of Overall Survival in ToGA Study (High HER2 overexpressors = IHC3+ or IHC2+/FISH+) 1.0 Median Events OS FC + H 120 16.0 0.8 FC 136 11.8 0.6 HR = 0.65 95% CI (0.51, 0.83) 16.0 Event 11.8 0.4 0.2 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 Time (Months) No.at risk FC + H 228 218 196 170 142 122 100 84 65 51 39 28 20 12 11 5 4 1 FC 218 198 170 141 112 96 75 53 39 28 20 13 11 4 3 3 H, Herceptin / Bang YJ., et al., Lancet 2010; 376: 687-697

Evidence that pertuzumab + trastuzumab are more active than either agent alone Pertuzumab has a different mechanism of action and a different binding site on the HER2 receptor than trastuzumab Additive activity in pre-clinical animal models Breast cancer Gastric cancer Activity seen in phase II/III studies in HER2 positive BC BO17929 (phase II – 2nd line MBC) NEOSPHERE (phase II – EBC) CLEOPATRA (phase III – 1st line MBC)

Pertuzumab (Perjeta) An IgG1 (k) humanized monoclonal antibody Produced in CHO (Chinese Hamster Ovary) cell cultures Targets the HER2 receptor extra-cellular domain (sub-domain II) Adams CW et al. Cancer Immunol Immunother 2006

Pertuzumab and trastuzumab bind different epitopes on HER2 Trastuzumab-HER2 complex Pertuzumab-HER2 complex pertuzumab trastuzumab dimerization domain Inhibits ligand-independent HER2 signalling Prevents HER2 ECD shedding Activates ADCC Inhibits ligand-dependent HER2 signalling Inhibits HER2 hetero-dimerization Activates ADCC Non-competitive with trastuzumab due to different binding domain on HER2 receptor Hubbard SR Cancer Cell 2005; Molina MA et al. Cancer Res 2001; Junttila TT et al. Cancer Cell 2009; Scheuer W et al. Cancer Res 2009; Franklin MC et al. Cancer Cell 2004; Agus DB et al. Cancer Cell 2002

CLEOPATRA: Study Design: Registration Trial for Pertuzumab in Metastatic Breast Cancer Pivotal CLEOPATRA trial: Multicenter, randomized, double-blind, placebo-controlled, Phase III study in patients with HER2+ metastatic breast cancer (mBC) Patients with HER2+,* locally recurrent, unresectable, or mBC previously untreated with a biologic or chemotherapy for metastatic disease† (n=808) Docetaxel (q3w) Minimum of 6 cycles Pertuzumab + trastuzumab (q3w) (n=402) 1:1‡ Placebo + trastuzumab (q3w) Docetaxel (q3w) Minimum of 6 cycles Patients treated with pertuzumab and trastuzumab until progression of disease or unmanageable toxicity If docetaxel discontinued, treatment with pertuzumab and trastuzumab may continue (n=406) q3w = every 3 weeks. * HER2 overexpression was defined as IHC 3+ or FISH amplification ratio ≥2.0. IHC 3+ is defined as uniform, intense membrane staining in >10% of tumor cells. † Patients with prior adjuvant or neoadjuvant therapy were required to have a disease-free interval of ≥12 months to be eligible for enrollment in the trial. ‡ Randomization was stratified by prior treatment status (de novo or prior adjuvant/neoadjuvant therapy) and geographic region (Europe, South America, and Asia). 20

CLEOPATRA: Primary Endpoint1 Progression-Free Survival (PFS): A hazard ratio (HR) of 0.62 describes 38% reduction in risk of disease progression or death HR = 0.62* 95% CI (0.51, 0.75) p value = <0.0001 100 Pertuzumab + trastuzumab + docetaxel 80 60 18.5 Progression-free survival (%) 40 12.4 20 Placebo + trastuzumab + docetaxel 5 10 15 20 25 30 35 40 Time (Months) Patients at risk Pertuzumab+T+D 402 345 267 139 83 32 10 Placebo+T+D 406 311 209 93 42 17 7 * Stratified by prior treatment status and geographic region. Source: 1. PERJETA™ (pertuzumab) full prescribing information. Genentech, Inc., June 2012.

CLEOPATRA: Confirmatory interim analysis of overall survival (OS IA2) Median follow-up: 30.0 months, n=267 OS events 100 Pertuzumab + T + D: 113 events; median not reached 80 HR = 0.66 95% CI (0.52−0.84) p value = 0.0008 60 Overall survival (%) 40 Placebo + T + D: 154 events; median 37.6 months 20 402 406 5 387 383 10 371 350 15 342 324 20 317 285 25 230 198 30 143 128 35 84 67 40 33 22 45 9 4 50 55 Time (Months) Patients at risk Ptz + T + D Pla + T + D Stopping boundary for concluding statistical significance at this interim analysis was p≤0.0138 D, docetaxel; OS, overall survival; Pla, placebo; Ptz, pertuzumab; T, trastuzumab S Swain, San Antonio Breast Cancer Symposium (SABCS), 2012 (Abstract #P5-18-26)

CLEOPATRA: Adverse events (all grades) ≥25% incidence or ≥5% difference between arms Adverse event, n (%) Placebo + trastuzumab + docetaxel (n=397) Pertuzumab + trastuzumab + docetaxel (n=407) Diarrhea 184 (46.3) 272 (66.8) Alopecia 240 (60.5) 248 (60.9) Neutropenia 197 (49.6) 215 (52.8) Nausea 165 (41.6) 172 (42.3) Fatigue 146 (36.8) 153 (37.6) Rash 96 (24.2) 137 (33.7) Decreased appetite 105 (26.4) 119 (29.2) Mucosal inflammation 79 (19.9) 113 (27.8) Asthenia 120 (30.2) 106 (26.0) Peripheral edema 119 (30.0) 94 (23.1) Constipation 99 (24.9) 61 (15.0) Febrile neutropenia 30 (7.6) 56 (13.8) Dry skin 17 (4.3) 43 (10.6)

Perjeta/Pertuzumab 1st Line MBC Approvals as of 16 May 2013 Country Approval United States 08 Jun 2012 Switzerland 14 Aug 2012 Mexico 26 Sep 2012 Macau 03 Dec 2012 Uruguay 26 Dec 2012 Kosovo 20 Jan 2013 Georgia 21 Jan 2013 Aruba 31 Jan 2013 Kuwait 6 Feb 2013 Country Approval Israel 12 Feb 2013 European Union 4 Mar 2013 Iceland Mar 2013 Canada 14 Apr 2013 Russia 16 Apr 2013 Argentina 25 Apr 2013 Norway Apr 2013 Luxemburg Australia 5 May 2013

Rationale for a Phase III study of pertuzumab in HER2-positive gastric cancer Preclinical data show that pertuzumab increased the activity of trastuzumab in HER2-positive gastric cancer xenograft models HER2-positive gastric/GEJ cancer (ToGA): clinical evidence that HER2-targeted therapy extends survival In HER2-positive mBC: CLEOPATRA study confirmed improved efficacy of combined HER2-directed therapy with docetaxel Large safety database showing good tolerability across multiple indications , includes data with capecitabine and platinum Plan → Phase III trial in advanced gastric/GEJ cancer. But what dose Pertuzumab in aGC ?

JOSHUA BP27836: Phase IIa study rationale Determine pertuzumab dose in HER2 positive advanced GEJ/gastric cancer Achieve PK clinical target for minimum pertuzumab concentration of 20 g/ml in >90% of patients Same clinical target concentration as in HER2-positive breast cancer Move to Phase III with dose modeled from Phase IIa data

JOSHUA: Study design, Phase IIa trial HER2-positive inoperable locally advanced or recurrent and/or metastatic GEJ/GC (IHC3+ or IHC2+ and FISH+; ECOG PS 0 or 1) Arm A Pertuzumab 840 mg, then 420 mg + X + cisplatin x6 q3w + trastuzumab n=15 Dose selection for Phase III study based on PK modelling Arm B Pertuzumab 840 mg + X + cisplatin x6 q3w + trastuzumab n=15 Primary objectives: Estimate minimum (trough) serum pertuzumab concentration to identify a steady state of ≥20 g/ml in 90% of patients Safety and tolerability ECOG PS, Eastern Cooperative Oncology Group performance status; FISH, fluorescence in situ hybridisation; GC, gastric cancer; GEJ, gastroesophageal junction; IHC, immunohistochemistry; PD, progressive disease; pK, pharmacokinetics; X, capecitabine 27

The JACOB Study (BO25114): Rationale of Study Design and Protocol Overview

The JACOB Study (BO25114) Design Study Treatment ~ 6 treatment cycles (21 day cycle) Study Treatment * HER2 targeted therapy continues until PD or unacceptable toxicity Randomization (1:1) n = 780 (approximately 390 per treatment arm) Primary Endpoint OS (Events = 502) Secondary Endpoints PFS, ORR, DoR, CBR, Safety, PK, QOL Treatment Arm A Capecitabine or 5FU + Cisplatin * Trastuzumab + Pertuzumab 840 mg IV Q3W Follow-Up # Treatment Arm B Capecitabine or 5FU + Cisplatin * Trastuzumab + Pertuzumab placebo IV Q3W Key Eligibility Criteria: HER2-positive Metastatic gastric/GEJ adenocarcinoma IHC 3+ or IHC 2+/ISH+ (central testing required) ECOG PS 0 or 1 * After Cycle 6, continuation of chemotherapy is at physician’s discretion Stratification factors: Geographic Region (Japan, N. America/W. Europe, Asia (not including Japan), Latin America/E. Europe) Prior gastrectomy (yes/no) HER2 IHC3+ vs. IHC2+/ISH+ # If patients permanently stop study treatment (HER2-directed), they continue to be followed for side effects and cardiac monitoring

JACOB: Study endpoints Primary Overall survival (15.0 vs. 19.3 mos, HR 0.78) Secondary PFS, ORR, duration of objective response, clinical benefit rate Safety PK QoL Exploratory Biomarkers Tumor tissue: HER receptor status, ERCC1, PTEN, PI3K, MET Blood: HER2 ECD, HER ligands, anti-therapeutic antibody ORR, objective response rate; PFS, progression-free survival; pK, pharmacokinetics; QoL, quality of life

JACOB: Biomarker Analyses Overview Biomarker analysis and sampling strategy in JACOB Patient selection (Block/partial block*) Exploratory BM Research (Block/partial block*) RCR (Roche clinical repository, optional) Central confirmation of HER2 (prospectively) IHC ISH c-MET ERCC1 PTEN PIK3CA mutations HER2/3 mRNA HER3 protein Serum Plasma Whole blood sample for DNA extraction * Minimum of 15 slides to be sent if country or site regulations do not allow shipment of a tumor block

JACOB: Statistical assumptions Hazard ratio 0.78 Number of patients 780 Number of events for 80% power 502 Median OS: control vs. experimental arm, months 15 vs. 19.3 (Δ=4.3) Minimum detectable difference at 80% power 0.837 2.9 months Enrollment*, months 38 Expected availability of OS data†, months 55 ~ Q4 2017 * Assuming a 5% withdrawal rate within 55 months † Clinical cut-off from EAST plus 2 months OS, overall survival

JACOB: Interim Analyses Safety: Monthly review of AEs first six months Regular safety monitoring reviews (approximately every 6 months) Two formal interim safety analyses After 50 patients have been enrolled and followed up for at least 2 months After 100 patients have been enrolled and followed up for at least 4 months Efficacy: One interim analysis of OS after 351 deaths have occurred

JACOB: Target Study Population Approximately 780 patients with histologically confirmed metastatic GEJ/GC who have received no prior treatment for metastatic disease Measurable or non-measurable evaluable (RECIST 1.1) ECOG PS 0 or 1 Baseline LVEF ≥ 55% (MUGA or echocardiogram) Life expectancy ≥ 3 months HER2 positive (IHC3+ or IHC2+/ISH+) tumor by central laboratory testing

Definition of HER2 positivity in JACOB HER2 positivity is defined as: IHC 3+ (regardless of ISH results) IHC 2+ if confirmed by a positive ISH result (cut-off for ISH ratio: 2.0 HER2/CEP17) For IHC testing, the cut-off as approved by FDA in context of ToGA will apply Screening population IHC test IHC 0 and 1+ IHC 2+ IHC 3+ Not eligible Reflex test by ISH eligible ISH- ISH+ Not eligible eligible

Tissue requirements for HER2 central assessment Tumor blocks preferred (submission of slides should be an exception!)  Cutting at central lab ensures higher quality and higher level of standardization and saves tissue! If blocks not available a minimum of 15 unstained and freshly cut slides can be submitted for HER2 testing and exploratory Biomarker analyses (maximum 48hrs from slicing to shipment) Biopsy or surgical specimen of either primary or recurrent/metastatic tumor are allowed for HER2 assessment For biopsies, ideally 6 to 8 biopsies taken from different metastatic lesions should be submitted and assessed Fine needle aspirates CANNOT be accepted The fixative allowed is neutral buffered formalin (10%)

JACOB: Exclusion criteria Cancer-Related Exclusion Criteria Previous systemic cytotoxic chemotherapy for advanced (inoperable or metastatic) disease History of exposure to the following cumulative doses of anthracyclines: Epirubicin > 720 mg/m2 Doxorubicin or liposomal doxorubicin > 360 mg/m2 Mitoxantrone > 120 mg/m2 and idarubicin > 90 mg/m2 Other (e.g., other anthracycline greater than the equivalent of 360 mg/m2 of doxorubicin) If more than one anthracycline has been used, then the cumulative dose must not exceed the equivalent of 360 mg/m2 of doxorubicin Evidence of disease progression documented within 6 months after completion of prior neoadjuvant and/or adjuvant cytotoxic chemotherapy or radiotherapy for gastric/GEJ adenocarcinoma   Previous treatment with any HER2-directed therapy, at any time, for any duration Previous exposure to any investigational treatment within 30 days before the first dose of study treatment Radiotherapy within 30 days before the first dose of study treatment (within 2 weeks if given as palliation to peripheral bone metastases, if recovered from all toxicities) History or evidence of brain metastasis Clinically significant active GI bleeding (Grade ≥ 2 according to CTCAE v4.03) Residual toxicity resulting from previous therapy, e.g., hematologic, cardiovascular, or neurologic toxicity that is Grade ≥ 2 (according to CTCAE v4.03). Alopecia is permitted Other malignancy (in addition to GC) occurring within 5 years before enrollment, except for carcinoma in situ of the uterine cervix or squamous or basal cell carcinoma of the skin that has been previously treated with curative intent Previous systemic cytotoxic chemotherapy for advanced (inoperable or metastatic) disease History of exposure to the following cumulative doses of anthracyclines: Epirubicin > 720 mg/m2 Doxorubicin or liposomal doxorubicin > 360 mg/m2 Mitoxantrone > 120 mg/m2 and idarubicin > 90 mg/m2 Other (e.g., other anthracycline greater than the equivalent of 360 mg/m2 of doxorubicin) If more than one anthracycline has been used, then the cumulative dose must not exceed the equivalent of 360 mg/m2 of doxorubicin Evidence of disease progression documented within 6 months after completion of prior neoadjuvant and/or adjuvant cytotoxic chemotherapy or radiotherapy for gastric/GEJ adenocarcinoma   Previous treatment with any HER2-directed therapy, at any time, for any duration

JACOB: Exclusion criteria Clinical Laboratory Exclusion Criteria (must be confirmed within 7 days before first dose of study treatment) Absolute neutrophil count (ANC) < 1.5  109/L Platelet count < 75  109/L Hemoglobin < 9.0 g/dL Creatinine clearance < 60 mL/min/1.73 m² Serum bilirubin (total) > 1.5  ULN of laboratory normal range; in case of known Gilbert’s disease a total bilirubin of 2  ULN is permitted AST, ALT, and alkaline phosphatase (ALP) parameters: In patients with no liver and no bone metastases AST or ALT > 1.5 × ULN, and ALP > 2.5 × ULN AST or ALT > 2.5 × ULN In patients with liver metastases and no bone metastases AST or ALT > 5 × ULN, and ALP > 2.5 × ULN In patients with liver metastases and bone metastases AST or ALT > 5 × ULN, and ALP > 10 × ULN In patients with bone metastases and no liver metastases AST or ALT > 1.5 × ULN, and ALP > 10 × ULN Serum albumin < 25 g/L Serum pregnancy test positive in a female patient of childbearing potential

JACOB: Exclusion criteria General Exclusion Criteria Documented history of congestive heart failure of any New York Heart Association (NYHA) criteria Angina pectoris requiring treatment Myocardial infarction within the past 6 months before the first dose of study treatment Clinically significant valvular heart disease or uncontrollable high-risk cardiac arrhythmia, i.e., atrial tachycardia with a heart rate > 100/min at rest, significant ventricular arrhythmia (ventricular tachycardia) or higher-grade AV-block (second degree AV-block Type 2 [Mobitz 2] or third degree AV-block) History or evidence of poorly controlled arterial hypertension (systolic blood pressure > 180 mmHg or diastolic blood pressure > 100 mmHg) Baseline LVEF value < 55%, assessed by echocardiogram [ECHO], MUGA scan, or cardiac MRI

JACOB: Exclusion criteria Dyspnea at rest due to complications of advanced malignancy or other disease or requirement of supportive oxygen therapy Any significant uncontrolled intercurrent systemic illness (e.g., active infection, poorly controlled diabetes mellitus) Previous major surgery within 30 days before the first dose of study treatment, unless completely recovered Known infection with HIV, hepatitis B virus, or hepatitis C virus that requires active treatment Ongoing chronic treatment or high-dose treatment with corticosteroids. Inhaled steroids and clinically indicated short courses of oral steroids are permitted Known dihydropyrimidine dehydrogenase (DPD) deficiency Known hypersensitivity to any component of study treatment Current use of antiviral drug sorivudine or its chemically related analogues, such as brivudine Lactating female patient Any patient unwilling or unable to use adequate contraceptive measures

Contraception For women of childbearing potential and male participants with partners of childbearing potential: Agreement to use one highly effective form of contraception or two effective forms of non-hormonal contraception by the patient and/or partner while on study treatment and for 6 months after stopping study treatment: Highly effective (any one) True abstinence Male sterilization Tubal ligation Combined hormonal contraceptives Effective (any two) IUD/IUS Condom with spermicidal foam/gel/film/cream/suppository Occlusive cap (diaphragm or cervical/vault cap) with spermicidal foam/gel/film/cream/suppository

Screening Assessments Day -28 to -1 Informed consent HER2 tumor testing Serum sample for levels of HER2 shed extracellular domain (ECD) Demographics and medical history Physical exam ECOG PS Tumor assessments ECG & LVEF assessments Tumor tissue block for biomarkers Concomitant therapy Day -7 to -1 Randomization Hematology, serum chemistries Creatinine clearance Urinalysis Pregnancy test Concomitant therapy

JACOB: Study treatment All agents will be given on the same day in the following order (pertuzumab/placebo → trastuzumab → chemotherapy) Chemotherapy backbone will be administered on a 3-weekly schedule Pertuzumab 840 mg or placebo q 3 weeks Trastuzumab 8 mg/kg initial dose, followed by 6 mg/kg q3w with Capecitabine 1000 mg/m2 taken orally twice daily for 14 days then 7 days off q3w OR 5-Fluorouracil 800 mg/m2/24 hours given intravenously by continuous infusion for 120 hours (Days 1−5) q3w with Cisplatin 80 mg/m2 intravenously (Day 1) q3w

Administration guidelines for pertuzumab and trastuzumab Administration Time (minutes) Observation Time (minutes) First dose Pertuzumab/placebo 60 Trastuzumab 90 All later doses 30 Incomplete dose: If patient receives 50-75% of dose, the remainder should be given during the same treatment cycle, preferably before day 15 (Section 4.3.1.4.1 for details)

The JACOB Study (BO25114): Safety Monitoring

Overview of Safety Monitoring Upfront discussion of diarrhea management and anti-diarrheal medicines prescribed C1 Day 7 safety check (by phone) ** please document in source ! AE grading using CTCAE version 4.03 Progressive disease is not an AE

Diarrhea management and dose reductions Table 4 in protocol Grade 2 or greater: STOP capecitabine Supportive care with loperamide Cannot restart until grade ≤1 and last loperamide given 24 hrs beforehand If controlled within 2 days, can restart at 100% If > 2 days, medical evaluation required

Diarrhea grades CTCAE Grade (v. 4.03) 1 Increase of < 4 stools per day over baseline; mild increase in ostomy output compared to baseline 2 Increase of 4 – 6 stools per day over baseline; moderate increase in ostomy output compared to baseline 3 Increase of >= 7 stools per day over baseline; incontinence; hospitalization indicated; severe increase in ostomy output compared to baseline; limiting self care ADL 4 Life-threatening consequences; urgent intervention indicated

Dose Modifications for pertuzumab and trastuzumab No dose reduction permitted for toxicity, for either pertuzumab or trastuzumab Treatment with either pertuzumab or trastuzumab (or both) may be delayed to assess or treat AEs Dosing should be resumed as soon as appropriate after the AE has improved (see protocol section 4.3.1.4.1) Patients should be re-assessed at least weekly to determine if treatment can be resumed If trastuzumab dosing is delayed and ≥ 6 weeks have elapsed since the last dose, a re-loading dose of 8 mg/kg should be given as the next dose, then resume the 6 mg/kg Q3W dosing schedule Pertuzumab, capecitabine, and cisplatin dosing should be re-synchronized to the new schedule

Dose Modifications for Hematologic Toxicity Due to FP Chemotherapy Table 2 in protocol ANC (x 109/L) Platelets (x 109/L) Fluoropyrimidine/Cisplatin ≥ 1.5 and ≥ 75 100% of original doses, without delay ≥ 1.0 – < 1.5 75% of original doses, without delay < 1.0 and/or Delay for up to 3 weeks, then resume treatment if ANC ≥ 1.0 and platelets ≥ 75. If ANC is ≥ 1.0 – < 1.5, administer 75% of original doses. If ANC is ≥ 1.5, administer 100% of original doses. If ANC < 1.0 or platelet count is < 75, permanently discontinue fluoropyrimidine and cisplatin for unacceptable toxicity. If recovery has not occurred after a delay of 3 weeks (i.e., 6 weeks since previous chemotherapy doses), the patient should be permanently discontinued from fluoropyrimidine and cisplatin.

Dose Modifications for Life-Threatening Hematologic Toxicity Occurring during FP Administration Table 3 in protocol Toxicity Fluoropyrimidine and Cisplatin in Subsequent Cycles Grade 4 neutrophil count decrease > 5 days in duration 75% of original doses Grade 4 platelet count decrease 50% of original doses Grade 3 febrile neutropenia ANC < 1000/mm3 with a single temperature of > 38.3⁰C (101⁰F) or a sustained temperature of ≥ 38⁰C (100.4⁰F) for more than 1 hour Grade 4 febrile neutropenia ANC < 1000/mm3 with a single temperature of > 38.3⁰C (101⁰F) or a sustained temperature of ≥ 38⁰C (100.4⁰F) for more than 1 hour and life-threatening consequences 50% of previous original, at investigator’s discretion Interrupt chemotherapy Dose reduce as directed when starting next cycle Dose reductions are permanent

Cardiac Safety Monitoring HER2 directed therapy has been implicated in cardiac toxicity Pertuzumab and trastuzumab cardiac safety must be monitored Currently, there is no evidence for increased cardiac toxicity with pertuzumab-containing combination regimens Regularly scheduled ECG and LVEF assessments are mandatory Every 9 weeks while on chemotherapy treatment Every 12 weeks while on antibody alone treatment Every 6 months for 1 year after stopping study drug Then annually during survival follow up for up to 5 years

Algorithm for Asymptomatic decline in LVEF Figure 3 of Protocol LVEF drop from baseline LVEF ≥ 50% LVEF < 50% LVEF drop ≤ 20% LVEF drop > 20% LVEF drop < 10% LVEF drop ≥ 10% CONTINUE Treatment CONTINUE Treatment and repeat LVEF in 3 weeks HOLD Treatment and repeat LVEF in 3 weeks Not Confirmed (LVEF drop < 10pts or LVEF ≥ 50%) LVEF drop CONFIRMED (LVEF drop ≥ 10pts and LVEF < 50%) RESUME Treatment STOP Treatment

Pertuzumab/Trastuzumab Cardiac Toxicity: Reporting AE/SAE Table 7 in protocol Observation How to Report Term to be Reported Grading Asymptomatic decline in LVEF < 10% points from baseline or to an LVEF ≥ 50% No additional reporting required, LVEF results to be reported on eCRF N/A Asymptomatic decline in LVEF ≥ 10% points from baseline to an LVEF < 50% AE (eCRF) “Ejection fraction decreased” NCI CTCAE for “ejection fraction decreased” Asymptomatic decline in LVEF requiring treatment or leading to discontinuation of pertuzumab/ placebo and trastuzumab Congestive heart failure (symptomatic LVSD) SAE (eCRF) “Congestive heart failure” NCI CTCAE for “heart failure” and NYHA Class AE = adverse events; CTCAE = Common Terminology Criteria for Adverse Events; eCRF = electronic Case Report Form; LVEF = left ventricular ejection fraction; LVSD – left ventricular systolic dysfunction; N/A = not applicable; NCI = National Cancer Institute; NYHA = New York Heart Association; SAE = serious adverse event. Notes: Any symptomatic LVSD event must be reported as “congestive heart failure”.

LVSD Dysfunction eCRF If a patient has a decrease in LVEF by ≥10 percentage points from baseline to an absolute value below 50% recorded in the LVEF assessment CRF, the Principal Investigator will then complete the following CRF: Does the subject have difficulty breathing? If yes, select all that apply:  While climbing less than 8 stair steps  While getting dressed  At rest  While lying flat  Other Does the subject have any clinical signs that may be due to heart failure? If yes, select all that apply:  Bilateral ankle edema  Pulmonary rales  Weight gain  Jugular venous distention   Other

Pertuzumab/Trastuzumab Cardiac Toxicity: Reporting AE/SAE Table 7 in protocol Observation How to Report Term to be Reported Grading Asymptomatic decline in LVEF < 10% points from baseline or to an LVEF ≥ 50% No additional reporting required, LVEF results to be reported on eCRF N/A Asymptomatic decline in LVEF ≥ 10% points from baseline to an LVEF < 50% AE (eCRF) “Ejection fraction decreased” NCI CTCAE for “ejection fraction decreased” Asymptomatic decline in LVEF requiring treatment or leading to discontinuation of pertuzumab/ placebo and trastuzumab Congestive heart failure (symptomatic LVSD) SAE (eCRF) “Congestive heart failure” NCI CTCAE for “heart failure” and NYHA Class AE = adverse events; CTCAE = Common Terminology Criteria for Adverse Events; eCRF = electronic Case Report Form; LVEF = left ventricular ejection fraction; LVSD – left ventricular systolic dysfunction; N/A = not applicable; NCI = National Cancer Institute; NYHA = New York Heart Association; SAE = serious adverse event. Notes: Any symptomatic LVSD event must be reported as “congestive heart failure”.

Algorithm for Asymptomatic decline in LVEF Figure 3 of Protocol LVEF drop from baseline LVEF ≥ 50% LVEF < 50% LVEF drop ≤ 20% LVEF drop > 20% LVEF drop < 10% LVEF drop ≥ 10% CONTINUE Treatment CONTINUE Treatment and repeat LVEF in 3 weeks HOLD Treatment and repeat LVEF in 3 weeks Not Confirmed (LVEF drop < 10pts or LVEF ≥ 50%) LVEF drop CONFIRMED (LVEF drop ≥ 10pts and LVEF < 50%) RESUME Treatment STOP Treatment

Tumor Assessments Measurable or non-measurable evaluable tumor (RECIST 1.1) Baseline assessment at screening, then after every 9 weeks until progressive disease (PD) is documented Measurable disease should be imaged using CT or MRI – use the same method throughout the study for a given patient Report measurements of target and non-target lesions for measurable disease (per RECIST) Report objective response (CR/PR, SD, PD) for both measurable and non-measurable evaluable disease If PD occurs off cycle, an unscheduled Tumor Assessment should be performed

Schedule of Assessment • Until Post-treatment Monitoring Visit 1 (1) Screening/ Baseline Cycle Post-treatment Monitoring Visit 1 b (28 days post last dose) 1 2 3 4 5 6 7 8 9+ Day -28 to -1 -7 to -1 22 (± 7) 43 64 85 106 127 148 169+ (± 7) a Informed consent ● Randomization c ● c Demographic data Medical history Physical examination ● d Height Weight Vital signs e ECOG PS Hematology f Biochemistry f Creatinine clearance g ● h Urinalysis As clinically indicated Pregnancy test i ● i HER2 status (central testing) j ● j Tumor assessment k ● k PK sampling l ● l

Schedule of Assessment • Until Post-treatment Monitoring Visit 1 (2) Screening/ Baseline Cycle Post-treatment Monitoring Visit 1 b 1 2 3 4 5 6 7 8 9+ Day -28 to -1 -7 to -1 22 (± 7) 43 64 85 106 127 148 169+ (± 7) a Serum for HER2/neu ECD (shed ECD) and HER ligands m ● ECG n ● n ● n,o,p LVEF n (MUGA/ECHO/MRI) ● n,p Safety monitoring q AEs and SAEs will be monitored from the time of enrollment Concomitant therapy ● r Capecitabine (28 doses, Days 1-15/ cycle) OR 5-FU (120-hour CIV, Days 1-5/ cycle) s Cisplatin (Day 1/cycle) s Pertuzumab or matching placebo (Day 1/cycle) t Trastuzumab (Day 1/cycle) t Patient-reported outcomes u Antitherapeutic antibodies ● v Tumor tissue block for mandatory biomarker analyses (FFPE) w Serum for RCR (optional) ● x ● y Plasma for RCR (optional) Whole blood for RCR DNA (6 mL) (optional)

Schedule for Pharmacokinetic Sampling for Pertuzumab and Trastuzumab Appendix 2 in protocol Cycle 1 Cycle 2 Cycle 3 Cycle 4 Cycle 6 Cycle 8 a Post-treatment Monitoring Visit 1 b (28 days post last dose) Post-treatment Monitoring Visit 2 c (60-90 days post last dose) Predose d ● Postdose e a Sample collected when on biologic therapy alone and last chemotherapy was ≥6 weeks ago. This timepoint may not coincide with Cycle 8 for all patients b Sample collected 28 days after last administration of study c Sample collected 60-90 days after last administration of study treatment d Sample can be collected up to 6 hours prior to administration of the specified drug e Sample can be collected up to 30 minutes after end of drug infusion of specified drug

Table 1-2 Follow-up Assessments until Death Post-treatment Phase Monitoring d Survival Follow-up Timepoint Post-treatment Monitoring Visit 2 60-90 days after last study treatment 24 wk (±2 wk) Every 12 weeks a (±2 wk) ATA ● PK sampling Cardiac monitoring: LVEF b (MUGA/ECHO/MRI), ECG including cardiac medication (start and end dates, changes in dosing) Drug-related SAEs c Survival assessment d Post-treatment phase anticancer therapy Information about any subsequent (post-study treatment) anticancer therapy will be collected, including the names of all components of treatments and start and end dates of dosing

Recruitment Challenge: High Screen Failure Rate It is assumed that screening ineligibility rate may be ~85% Primarily driven by requirement for HER2 positivity ~ 22% of patients with advanced gastric cancer are HER2+ But… it’s not impossible: ToGA trial recruited successfully with a ~80% screen failure rate Improve overall screening rates by implementing process of pre-identifying and pre-screening patients This will help reduce screening failure rate not related to HER2 positivity If local pre-screening HER2 result is borderline or equivocal, please still send for central HER2 testing Reviewing pre-screening information will help identify why potentially eligible patients are not considered for screening

Synergy with MetGastric study JACOB study accepts HER2 test result from Targos and vice versa In those MetGastric and JACOB overlapping sites (25%), screen for MetGastric study first Where patient screened in MetGastric study has HER2 positive status: Obtain Informed Consent for JACOB Investigator/study nurse registers MetGastric study patient screen number in the JACOB’s IXRS screening module Send results transfer form with detailed patient information to Targos → Tissue sample will be transferred within Targos If site is not a JACOB/MetGastric overlapping site, a referral network is recommended

Role and responsibility of CRC (Cardiac Review Committee) Patients with potential cardiac events (e.g. symptomatic LVSD) per investigator assessment will be reviewed by independent external Cardiac Review Committee (CRC) CRC provides independent blinded central determination of symptomatic declines in LVEF and definite and probable cardiac deaths for reporting to IDMC CRCC (Duke Clinical Research Institution) is the vendor who will: Raise queries using RAVE (10 day turnaround) Request/collect source documents via email Ensure relevant variables are clean Prepare event packets Record CRC adjudications Coordinate CRC adjudication meetings

Potential source documents to be collected Physician progress or consultation notes Discharge summary ECG tracing and report ECHO/MUGA/cardiac MRI report Chest X-ray report Any other documents that CRCC deems necessary for case adjudication, e.g. cardiac catheterization report CRCC will be responsible for document translation if in language other than English

Q & A