Evolution of phase 1 trials for patients with advanced pancreatic cancer: An update on the experience from MD Anderson Cancer Center.

Abstract

320 Background: Pancreatic cancer is a disease that generally presents in advanced stage and is nearly fatal in all cases. In 2011, we reported our experience with pancreatic cancer patients enrolled on phase 1 clinical trials from 2004-2009. At the time, gemcitabine and erlotinib were the only US FDA approved drugs for pancreatic cancer. Median overall survival from presentation in the phase 1 clinic was 5 months. After an additional 5 years of progress, we queried the impact of novel cancer therapeutics, evolving molecular profiling, and targeted therapy on pancreatic cancer patient outcomes in the phase 1 setting. Methods: A retrospective review of advanced pancreatic adenocarcinoma patients from the Department of Investigational Cancer Therapeutics at MD Anderson Cancer Center, was conducted for patients treated from 1/2009 to 1/2014. Statistical analyses utilized the Kaplan-Meier method. Results: A total of 90 patients were identified in 50 trials reviewed. Median age was 62 years (40-84), 57% were male, and 40% had stage IV disease at presentation. Median documented PS was 1 (range 0-3). The median time from diagnosis to treatment on a phase I protocol was 15 months (0.2-119). Patients were treated with an average of 4 prior regimens prior to Phase 1 referral (0-10) with 8% having undergone >5 treatments. The median overall survival from the first phase I treatment was 5.2 months, 95% CI = (4.4-6.3) and the 1 year overall survival from the first phase 1 treatment was 15% (9%- 25%). The median duration on regimen with best phase I response was 1.9 months (0.2-21.3). Of 88 evaluable, the best responses were PR in 1 patient and SD (> 6 months) in 5 patients. Although 47 patients had biomarker profiling performed and 61 patients were treated with targeted therapy alone or in combination with cytotoxic therapy, only 6 patients (5 PD, 1 SD currently on protocol) were placed on trials based on biomarker testing results. Conclusions: Pancreatic cancer remains a difficult disease to treat with poor outcome. Referral to phase 1 occurs late in the disease course. Biomarker based therapies may be more successful with more stringent patient selection and when referred earlier or used prior to cytotoxic treatment.