E6201: Phase III Trial of Gemcitabine (30-minute infusion) vs Gemcitabine (fixed-dose-rate infusion) vs Gemcitabine + Oxaliplatin for Patients with Advanced Pancreatic Cancer Elizabeth Poplin MD, Donna Levy MS, Jordan Berlin MD, Mace Rothenberg MD, Peter O’Dwyer MD, David Cella MD, Edith Mitchell MD, Steve Alberts MD and Al B. Benson III. I thank the organizers for inviting me,on behalf of ECOG and my collegues to present the results of ECOG 6201, a phase III…

Background Median Survival 1 yr Survival Study Gemcitabine (30 minutes) 5.7 months 18% Burris FDR Gemcitabine 7.8 months 24% Tempero Gemcitabine + Oxaliplatin 9.2 months 36% Louvet

E6201: Objectives Primary: To determine whether either FDR GEM or GEMOX increases survival compared with standard 30-min infusion GEM. Secondary: Toxicity, response rates, progression-free survival, and QOL A randomized trial was developed by ECOG In order to assess which of these three treatments was the best. The primary objective was to determine if either FDR Gem or GEMOX oncreased survival compared to standard gem infusion. Secondary endpoints were assessments of toxicity, response rates, progression free survival and quality of life. Poplin, E. ASCO 2006

E6201: Study Design Stratification factors: ECOG PS: 0 –1 vs 2 Locally advanced vs metastatic disease Gem 1000mg/m2/ 30 minutes n=279 Randomization n=832 Gem FDR: 1500 mg/m2/150 minutes n=277 The study design is as follows Pts were stratified by performance status 0-1 vs 2 and locally advanced vs metastatic. The study was initially designed for 666 pts but due to rapid accrual the design was modified for 750 eligible pts to be randomizd to gemcitabine 1000 mg/mw/30 min, gemcitabine by fixed dose rate infusion, 1500 mg.m2/150 min vs Gemcitabine 1000 mg/m2 d1 and oxalipaltin 100 mg./m2 day 2 every 14 d Gem 1000mg/m2/100 minutes, day 1 Oxaliplatin 100mg/m2, day 2 n=276 Poplin, E. ASCO 2006

Statistical Design: Assumes median survival of 6 mo for standard Gemcitabine and 8 mo for FDR Gem and GEMOX Designed to detect >33% difference in survival (hazard ratio <0.75) -81% power -significance level of 0.025 in a two-sided test for each of the two primary comparisons Statistical underpinnings were as follows: We assumed a median survival of 6 mo for gem and 8 months for FDR Gemcitabine and 8 months for GEMOX. The study was designed to detect a 33% of gre ater difearence (hazard ratio of .75, with 81% power and significance level of 0.025 in a two sided comparison for each of the two primary comparisons, gem vs FDR gem and Gem vs GEMOX.

Final Accrual:832 Patients Contributors:ECOG, NCCTG, CTSU Study Conduct Activated March 26, 2003 Terminated March 16, 2005 Final Accrual:832 Patients Contributors:ECOG, NCCTG, CTSU The studied was activiated March 26 2005 and colosed March 16 2005 with the accrual of 832 patients. ECOG NCCTG and participants in the CTSU contributed thru their organizations

E6201: Eligibility Pancreatic adenocarcinoma or poorly differentiated carcinoma, measurable and/or nonmeasurable, locally advanced or metastatic No prior therapy for metastatic disease, prior adjuvant chemotherapy/radiotherapy was permitted No prior GEM or oxaliplatin. ECOG Performance Status of 0-2 4. Adequate organ function: 5. No symptomatic ( >gr2) peripheral neuropathy 6. Other standard eligibility Elibility are as noted. Pancreatic adeno carcinonam or poorly differenntiated catrcinoma of the pancreas with measurable or nonmeasurable disease—locally advanced or metastatic. Patients could not have received prior therapy for metastatic desease but prior adjuvant chemotherapy and/or radiotherapy was permitted. Patients could not have received prior oxaliplatin or gemcitabine. Patients had to have an ECOG performance status of 0-2, adequate hematologic, hepatic and renal function> Patients could not symptomatic neuropathy. Other usual eligibility criteria pertained. Poplin, E. ASCO 2006

E6201:Treatment Arms ARM A: GEM 1000 mg/m2/ 30 min q wk x 7 of 8 weeks,  GEM 1000 mg/m2/ 30 min q wk x 3 of 4 weeks(GEM) ARM B: GEM 1500 mg/m2/150 min q wk x 3 of 4 weeks (FDR) ARM C: GEM 1000 mg/m2/100 min day 1, and Oxaliplatin 100 mg/m2/120 min day 2 repeated every 14 days. (GEMOX) The treatment arms are as follows Gemcitabine 1000 mg/m2/30 min weeks one thru 7 followed by a weeks’ rest, then gem 1000 mg/m2 weely 3 weeks of every 4. Arm b gemcitabine 1500 mg/m2/150 min weekly x 3 every 4 weekly. Finally Arm C Gemcitabine 1000 mg/m2/100 min day 1 and oxaliplatin 100 mg/m2m120 min day2 repeated every 14 days Patients remained on treatment til progression or withdrawal or death. Poplin, E. ASCO 2006

Patient Entry Characteristics: Total: 832 88% PS 0-1 12% PS 2 88% metastatic 12% locally advanced 832 patients were entered 88% had PS 0-1 and 12 % PS 2 88% had metastatic disease 12% had locally advanced disease. Patients were randomized to gem, FDR Gem or GEMOX. The median F/U is 12.2 months. The study is mature, 725pts- 87% having expired. However, as the final analysis was just done last month, the only data presented today will be the primary endpoint of survival. GEM:279 FDR:277 GEMOX:276 Median follow-up 12.2 mo 725 patients expired @5-25-06 Poplin, E. ASCO 2006

E6201: Patient Characteristics GEM FDR GEMOX Gender (% M/F) 56/44 58/42 46/54*[P=.03] Age, median 63 61 6 month weight loss (% ): <10% 62 59 >20% 9 10 12 Prior Surgery (%) 16 11 Non-measurable (%) 5 4 6 CA19-9, median 2009 1147 1151 Patient characteristics are well-balalnced, though the male :female ratio for GEMOX is inverted, and is statistically different. The median age is early 60s, The majority of patients had a less than 10% weight loss. 11-16% had prior surgery. Few had non measurable disease. The CA19-9 ranged from 1100-2000, but the differences were not-statistically different. Poplin, E. ASCO 2006

E6201: Toxicity: Grade 3/4 GEM FDR GEM GEMOX % Gr 3 Gr 4 ANC 18 15 29 30 10* 12* PLT 12 2 3 10 1 HGB 8 5 Nausea 7 <1 Vomiting 6 11 Febr Neutrop Fatigue 17 Neuropathy 9 Compared to Gemcitabine, FDR Gem had more myelosuppression, more neutropenia-59%;more thrombocytopenia 32% and more anemia 18% All other toxicities were equivalent. Compared to gemcitabine, GEMOX had statistically less neutropenia, but more nausea( 16%),and vomiting 12% and neuropathy 9%. There were few toxic deaths, though there was 1 infection on each of the first two arms and embolus on arm 3. Poplin, E. ASCO 2006

Off-treatment reasons (prelim) % Gem FDR GEMOX PROGRESS 49 51 40 TOXICITY 15 20 24 WITHDRAWAL 4 8 12 DEATH 7 6 other 25 18 Though this study closed 15 months ago, data other than survival data should be considered somewhat preliminary. Thus, Preliminary data only is available on the reasons for patients coming off study. The most common reason was disease progression. Fewer patients, however, come off GEMOX for progression However patients also came off for toxicity or withdrew on their own> This was more common in in both of the experimental arms :20 and 24% of patients came off for toxicity on the FDR GEM and GEMOX arms respectively. 8 and 12 % withdrew from FDR Gem and GEMOX, respectively Poplin, E. ASCO 2006

E6201: Response (preliminary) % GEM FDR GEMOX CR/PR 5 10 9 Stable 29 Progress 26 24 UnE/UnK 41 37 38 Best Response was graded by RESIST criteria with initial assessment at 8 weeks. Complete and partial responses were seen in 5 on Gem, 10 on FDR Gem and 9% on GEMOX Stable was noted in 29% of patients. One quarter progressed. RESIST criteria do not account for symptomatic progression , an all too common occurrence in pancreatic cancer. Many of patients currently counted as Unevaluable or unknown- approximately 40% of patients-may fall into the category of symptomatic progression. We will be making an effort to sort out these patients currently considered unevaluable and unknown over the next few months. Thus, at this time , I will not be discussing progresion-free survival. Poplin, E. ASCO 2006

Survival Months,Median (CI) %1-year(SE) GEM 4.9 (4.5-5.6) 17%(2) FDR GEM 6.0 (5.4-6.9) 21%(3) GEMOX 5.9 (5.1-6.8) 21%(3) The median survival for Gem is 4.9 mo with a 17% I year survival;, for FDR Gem, 6 mo and 21% one year survival and for GEMOX 5.9 mo with a 21 % one year survival. Please note that the confidence intervals overlap for all three arms. Similarly the standard error overlap for all three arms. Poplin, E. ASCO 2006

E6201: Overall Survival By Treatment GEM 4.9 (4.5-5.6) 17% FDR GEM 6.0 (5.4-6.9) 21% GEMOX 5.9 (5.1-6.8) 21% To look at it another way, here are the survival curves:Gem in black, FDR gem in blue and GEMOX in red. By inspection the curves track closely or overlap for the first four months. GEM FDR GEMOX Poplin, ASCO, 2006

E6201: Survival Hazard Ratios HR 95% CI P FDR Gem vs Gem 0.83 0.69, 1.00 0.05 GemOX vs Gem 0.88 0.73, 1.05 0.16 The hazard ratios for the two experimental arms compared to the control gemcitabine arm are as follows FDR Gem vs Gem 0.83 with a confidence interval that includes 1. P=.05 For gemox vs Gem the hazard ratio is .88 , confidence interval of .73-1.05 and a stratified log rank of .16. Neither arm meets the threshold for statistiscal significance What of the p=.05 for FDR GEM? If you remember, the p value for each of these two survival comparisons, gem vs FDR Gem and Gem vs GEMOX was set at .025, splitting the usual p.05 between the two experimental arms, done to offset the problem of multiple comparisons increasing the likelihood of spurious findings). As such, the p=.05 of FDR Gem does not constitute a statistically significant improvement in survival though perhaps there is a trend towards improved outcome Neither experimental treatment succeeded in achieving the hazard ratio goal of <0.75 or the required p=.025. Poplin, ASCO, 2006

Survival by disease extent Locally advanced 9.1 mo Locally advanced 9.1 mo P=.0001 Disease extent clearly impacted on survival with patients with locally advanced disease having a risk of dying 60% that of patients with metastatic disease. However, there was no differing impact of the GEM. FDR GEM or GEMOX on outcome of patients be they locally advanced or metastatic. Again with regard to survival, There was no impact of gender. With regard to Performance status, median Survival for patient with PS 0-1 was 5.8 mo and for PS 2 3.9 mo. Progression-free survival, a secondary endpoint, will not be discussed at present. Metastatic 5.4 mo mo P+.00 Poplin, ASCO, 2006

Conclusions: Neither FDR GEM nor GEMOX proved statistically significantly better than standard GEM. Both had approximately one month longer median survival than standard GEM. The survival outcome for standard GEM is consistent, though slightly shorter than, other large studies showing survival of 5-6 months. Neither FDR nor GEMOX had survival durations as long as previously reported in smaller studies. Poplin, ASCO,2006

Conclusions: Both experimental arms had increased toxicity: increased myelosuppression with FDR GEM and increased N/V and neuropathy with GEMOX. Thus, E6201 adds to multiple prior studies that demonstrate either no or modest overall benefit with the addition of a second chemotherapy agent to gemcitabine. FDR GEM, with higher dose and longer duration, is of limited benefit. Poplin, ASCO,2006