Second Line Treatment for Gastroesophageal Cancers: Are We Helping People Feel Better and Live Longer? Johanna Bendell, MD Director, GI Oncology Research Associate Director, Drug Development Unit Sarah Cannon Research Institute Nashville, TN

Second Line Therapy for Gastroesophageal Cancers – A Difficult Situation After progression on first line therapy, only roughly 20% of patients receive second-line therapy with a median overall survival of 5.6 months (Chau, GI ASCO 2004) Patients are sick –Symptomatic – weight loss, anorexia, pain, difficulties eating (with or without prior local intervention), N/V Do we have data to show second line therapy can help?

Randomized Second Line Trials to Date StudyTreatmentNOS Treatment Group OS Control Group QoL AIO Thuss- Patience 2011 Irinotecan vs. BSC mo HR 0.48 [0.25,0.92] 2.8 mo50% vs. 7% improvement in tumor- related symptoms Park 2011 Irinotecan or docetaxel vs. BSC mo HR 0.63 [0.47,0.86] 3.8 moN/A WJOG4007 Ueda 2012 Irinotecan vs. Paclitaxel mo (paclitaxel) 8.4 mo (irinotecan) N/A GRANITE-1 Van Cutsem 2012 Everolimus vs. BSC mo HR 0.90 [ ] 4.34 moN/A

Trial Design Adenocarcinoma of esophagus, esophagus-gastric junction or stomach refractory to platinum and fluoropyrimide Arm A (n=84): Docetaxel 75mg/m 2 IV every 3 weeks for up to 6 cycles + ASC Arm B (n=84): Active symptom control May include: Radiotherapy, analgesia, anti-emetics, steroids Assess every 3 weeks for 18 weeks, then every 6 weeks RANDOMISE 1:1 n=168 Stratified by: 1.Disease status (Locally advanced vs metastatic); 2. Site of disease (Esophagus vs GEJ vs Stomach); 3. Time to progression after previous chemotherapy ( 0 vs 0-3 vs 3-6 months); 4. ECOG PS ( 0/1 vs 2)

COUGAR-02 Well-designed with good stratification factors Screening vs. number enrolled show how difficult these trials are to accrue –Many are ineligible (sick population) –Refusal of study (vs. BSC) Restaging at 3 and 6 cycles for docetaxel patients, as indicated for BSC patients (no TTP endpoint) Included QUALITY OF LIFE

This is a difficult population to treat These patients are sick Most do not complete treatment plan DocetaxelBSC Completed 18 weeks 23%36% Reason off Death15%38% PD40%2% Tox31%N/A TreatmentN/A14%

Toxicity Toxicity rates relatively low for q 3 week docetaxel Only significant differences are neutropenia and febrile neutropenia Relatively low neuropathy and thrombocytopenia Due to low amount of chemotherapy given? Median number of cycles of docetaxel given was 3

Overall survival Median survival: 5.2 months (95% CI ) for Docetaxel 3.6 months (95% CI ) for ASC Hazard ratio 0.67 (95% CI ), p=0.01

Who really received benefit??? Patients who had disease progression 3-6 months after first line therapy –Longest disease free interval –Selection of patients who respond to therapy better Patients with ECOG 0

We Help Them Live Longer, But Do We Help Them Feel Better? QoL forms (EORTC QLQ-C30, STO22, EQ5D) planned in both arms q 3 weeks for 18 weeks, then q 6 weeks Number of QoL forms expected/returned lower in the control arm – potential source of bias, but common in these types of studies Arm A (Docetaxel)Arm B (ASC) QLQ-C30 Form Number Received n(%)* Number Expected Number died/not expecting QOl n(%)~ Number Received n(%)* Number Expected Number died/not expecting QOl n(%)~ Baseline82(98)84076 (90)840 3 weeks62 (81)777 (8)50 (68)7311 (13) 6 weeks49 (68)7212 (14)38 (57)6717 (20) 9 weeks45 (68)6618 (21)25 (45)5529 (35) 12 weeks34 (55)6222 (26)29 (64)4539 (46) 18 weeks29 (60)4836 (43)13 (45)2955 (65) 24 weeks17 (52)3351 (61)11 (65)1767 (80) Total318 (72) (65)370

Quality of life (EORTC QLQ-C30) Standardised AUC analysis Comparison using O’Brien global rank procedure No significant differences in function or global health scale Symptom score for pain significantly better in chemotherapy arm CT ASC Global health Pain p=0.53 p=0.0008

COUGAR-2 Chemotherapy benefits patients in the second line setting –We now see docetaxel and irinotecan work –But overall benefit is still around 1.5 mo But need to think about who will really benefit –Can we get better at selecting the appropriate patient? –Longer PFI, better PS –Patients who respond better and feel better live longer

COG ASCO GI 24 th Jan 2013 Gefitinib in advanced esophageal cancer progressing after chemotherapy Patients progressing following chemotherapy Planned: 18 months to recruit 450 patients Primary endpoint: Overall survival - powered to detect an increase in 1 year survival from 10 to 18%, 82.5% power, 5% significance level. Secondary endpoints: PFS, toxicity & PROs 13 Gefitinib 500mg od (n=225) Gefitinib 500mg od (n=225) Placebo (n=225) Simple randomisation Multi-centre Double-blind – patients, clinicians and trial office staff blinded to trial treatment Treated until progression Regular CT scans

COG No benefit for overall survival –3.73 vs. 3.6 mo, HR 0.90 [0.74,1.09] PFS –1.60 vs mo, HR [0.657,0.962] –Was there a subgroup who had benefit? Is there a biomarker? Studies ongoing… Did anyone feel better? –EORTC QLQ-C30 –EORTC QLQ-OG25

Quality of Life Planned assessments at baseline, 4, 8, and 12 weeks then until progression Prespecified PRO: global QoL, dysphagia, difficulty eating, odynophagia Primary evaluation at 4 weeks –Not surprisingly for this patient population, only 70% alive and progression free at 4 weeks (82.5% placebo vs. 74.7% geftinib compliance)

Odynophagia is improved in the gefitinib arm at 4 weeks, and stays consistently better with time out to 12 weeks This points to a group of patients who appear to benefit from gefitinib There was a group with tumor shrinkage at 4 weeks –RR 3.1 vs 0.4% –DCR (8 weeks) 26 vs. 16% (p = 0.014) Can we identify who they are?

It always comes down to the biomarker question… But what are the biomarkers? REAL-3 –Mutations/pathway dysregulation not common –EXPAND biomarkers pending Though we do see EGFR overexpression (50-70%) Squamous vs. adenocarcinoma –Different responses to EGFR inhibitors? –SCC head and neck respond –Lordick, et al, R ph II SCC esophagus, RR 19 vs. 13%, PFS 5.7 vs. 3.6 mo, OS 9.5 vs. 5.5 mo Chau 2011

Randomized Second Line Trials to Date StudyTreatmentNOS Treatment Group OS Control Group QoL AIO Thuss- Patience 2011 Irinotecan vs. BSC mo HR 0.48 [0.25,0.92] 2.8 mo50% vs. 7% improvement in tumor-related symptoms Park 2011 Irinotecan or docetaxel vs. BSC mo HR 0.63 [0.47,0.86] 3.8 moN/A WJOG4007 Ueda 2012 Irinotecan vs. Paclitaxel mo (paclitaxel) 8.4 mo (irinotecan) N/A COUGAR-2 Ford 2013 Docetaxel vs. BSC mo HR 0.67 [0.49,0.92] 3.6 moNo diff global QoL but pain improvement GRANITE-1 Van Cutsem 2012 Everolimus vs. BSC mo HR 0.90 [ ] 4.34 moN/A COG Dutton 2013 Gefitinib vs. BSC mo HR 0.90 [0.74,1.09] 3.6 moImprovement in odynophagia, social function

MOVING FORWARD…

T-DM1 structure T-DM1 is a novel ADC Average drug: antibody ratio ≅ 3.5:1 Highly potent cytotoxic agent Monoclonal antibody: Trastuzumab Systemically stable Target expression : HER2 Cytotoxic agent: DM1 Linker: MCC T-DM1 Trastuzumab

* Dose selection based on PK/safety/efficacy ** Investigator’s choice between paclitaxel 80 mg/m 2 /wk and docetaxel 75 mg/m 2 q 3 wk Phase II: 3 arm; 2:2:1 randomization; endpoints: safety, PK, PFS, ORR; n=100 2L Her2 positive mGC PS: 0 -1 IHC 3+ or IHC 2+/ISH+ Prior Ctx + prior HER2 N=412 Chemotherapy ** T-DM1 2.4 mg/kg/wk T-DM1 3.6 mg/kg q3 wk Phase II n= Trastuzumab Emtansine: Phase II Study of 2L treatment for HER2+ Metastatic Gastric Cancer Stratified by: region, PS, prior gastrectomy, prior HER2-targeted tx

REGARD: Randomized Phase III Trial 2 nd Line Ramicirumab vs. Placebo 1:1 Second line metastatic gastric and GEJ adenocarcinoma R Ramucirumab IV q 2 weeks Placebo q 2 weeks 22 Primary EP: OS N = 355 Press release 10/12: met primary endpoint of OS and secondary endpoint of PFS Press release 1/23/13: OS 5.2 vs. 2.6 mo PFS 2.1 vs. 1.3 mo

RAINBOW: Randomized Phase III Trial 2 nd Line Paclitaxel +/- Ramicirumab 1:1 Second line metastatic gastric and GEJ adenocarcinoma R Paclitaxel 80 mg/m2 d1, 8, 15 + Ramucirumab IV q 2 weeks Paclitaxel 80 mg/m2 d1, 8, 15 + Placebo q 2 weeks 23 Primary EP: OS N = 665

Second Line Treatment of Gastroesophageal Cancers Consistent trials showing some benefit to chemotherapy COUGAR trial shows docetaxel is an appropriate chemotherapy choice COG trial shows gefitinib overall does not improve survival endpoints However, both trials show suggestion of subpopulations that may benefit more –COUGAR – longer PFI, better PS –COG – subpopulation with improved QoL factors Drug development in this setting needs to be more targeted to the right population –We are doing this –TDM1 for HER2 positive patients –Inclusion of biomarker studies and QoL in trials for this population

AIO: Randomized Phase III Trial 2 nd Line Irinotecan vs. BSC 1:1 Second line metastatic gastric and GEJ adenocarcinoma R Irinotecan 250 mg/m 2 IV q 3 weeks (could be increased to 350 mg/m 2 ) Best supportive care 25 Thuss-Patience, Eur J Cancer 2011 Primary endpoint: OS Stratification: Time to PD after first line therapy ECOG PS First line therapy

AIO – Second Line Study Thuss-Patience, Eur J Cancer 2011 Irinotecan (n = 19) BSC (n = 21) OS (months) HR 0.48 [0.25,0.92] P = Improvement of tumor- related symptoms 50%7% First randomized trial to look at second line therapy for gastric cancer Study closed early with 40 patients secondary to poor accrual Still showed overall survival benefit and symptom improvement

Next 2 nd line Chemotherapy (SLC) RCT Refused RCT, but prefer SLC Willing to participate RCT Screening & consent for RCT Refused RCT, but prefer BSC SLC BSC 2:1 randomization RCT RCT + PPT Docetaxel or irinotecan ClinicalTrials.gov, NCT RCT: randomized controlled trial PPT: patient-preference trial N = 202 Park ASCO 2011

Survival Survival Probability Months SLC + BSC 5.1 mo BSC alone 3.8 mo Median 95% CI Log-rank P=0.009 Median f/u (95% CI): 17 mo (16-18 mo) Park ASCO 2011 HR 0.63 [0.47,0.86]

WJOG4007 Trial – Second line chemotherapy for metastatic gastric cancer RANDOMIZATION Stratified by Institution, PS 0-1/2, target lesion -/+ IRI 150 mg/m 2 d1, 15 q4w weekly Paclitaxel 80 mg/m 2 d1, 8, 15 q4w AGC refractory to prior FP confirmed by imaging Age 20-75, PS 0-2, No history of CPT-11 or Taxane Ueda ASCO 2012

Overall Survival Probability (%) ( Months ) wPTX IRI Number at risk 0101 IRI wPTX n Median 8.4M 9.5M P 0.38 HR (95% CI) 1.13 ( ) Log-rank test Udea ASCO 2012