A phase IB/randomized phase II study of gemcitabine (G) plus placebo (P) or vismodegib (V), a hedgehog (Hh) pathway inhibitor, in patients (pts) with metastatic pancreatic cancer (PC): Interim analysis of a University of Chicago phase II consortium study

4022 Background: Sonic Hh (SHh), the ligand for the Hh pathway, is over-expressed in >80% of PC. V, a small molecule antagonist of the Hh signaling pathway, has activity in preclinical PC models. Methods: We conducted a multi-center, placebo-controlled, phase IB/randomized phase II trial of GV or GP. Eligible pts, KPS 80-100, had previously untreated metastatic PC, or had completed adjuvant therapy > 6 months (mo) prior. Primary endpoint: progression-free survival (PFS). Correlatives: serial SHh serum levels; serial contrast perfusion CT imaging. To allow for early stopping due to lack of efficacy, a planned interim analysis was performed after approximately 50% of the expected number of PFS events occurred. The protocol stipulated that the trial would be terminated if the probability of rejecting the null hypothesis were the trial to continue (conditional power) was 2 over 30 minutes, days (D) 1, 8, 15, Q28D. Pts, stratified by KPS (80 v 90/100), and disease status (newly-diagnosed/recurrent), were randomized to V (150 mg PO daily) or P. For pts on P, cross-over was allowed at progression. Results: 70 evaluable pts (V/P 35/35) enrolled at 12 sites 2/10-6/11. Pt characteristics: median age 63/63 (range 49-79/48-82); KPS 80: 8/8; 90: 12/14; 100: 15/13. Grade 3/4 toxicity (%pts): neutropenia 20/26; hyponatremia 3/11; fatigue 9/6; hyperglycemia 14/6; elevated alkaline phosphatase 9/11. Response (%): complete 0/3, partial 0/11, stable disease 49/31. Median PFS: 3.7/2.4 mo (95% CI: 2.4-4.6/1.9-3.7; adjusted HR 0.92 [0.52-1.64]). Upon progression/unblinding, 23 GP pts crossed over to GV. Median overall survival (OS): 6.3/5.4 mo (95% CI:4.9-7.8/4.2-8.0, adjusted HR 0.97, [0.47-2.01]). 1-year survival (%): 24/24. Laboratory and radiological correlatives will be presented. Conclusions: GV has an acceptable toxicity profile. This trial did not meet criteria for futility at this interim analysis. The study is expected to complete accrual of 112 pts in February 2012. The final analysis will be reported after 90 events. Funded by NCI N01-CM-62201.