Overall survival with everolimus, sunitinib, and placebo for advanced pancreatic neuroendocrine tumors: A matching-adjusted indirect comparison of randomized trials.

237 Background: Everolimus and sunitinib are approved in the US and EU for treating advanced pancreatic neuroendocrine tumors (pNET). No randomized trial has directly compared these treatments for pNET. This study indirectly compared overall survival (OS) with everolimus vs. 1) placebo and 2) sunitinib. Methods: Individual patient data were used from the RADIANT-3 trial of everolimus 10 mg/day vs. placebo (cutoff: April 4, 2010); published summary data were used from the A6181111 trial of sunitinib 37.5 mg/day vs. placebo (cutoff: June 1, 2010). To adjust for cross-trial differences, RADIANT-3 patients were subjected to exclusion criteria used in A6181111 and then re-weighted to match A6181111 in terms of baseline demographics, performance status, time since diagnosis, disease sites, distant metastases and prior therapy. After matching, OS was compared between balanced trial populations for everolimus in RADIANT-3 vs. 1) the placebo arm in A6181111, which had access to sunitinib after early stopping, and 2) the sunitinib arm in A6181111. Progression-free survival (PFS) was also assessed, but was confounded by early stopping of A6181111. Results: Analyses included 394 patients from RADIANT-3 (after excluding 15 with worse performance status than allowed in A6181111 and 1 with missing baseline data) and all 171 patients in A6181111. Before matching, patients in RADIANT-3 had better performance status, more prior use of somatostatin analogues and less prior use of systemic chemotherapy. After matching, these and all other baseline characteristics were well-balanced between trials. Everolimus was associated with prolonged OS vs. placebo in A6181111 (hazard ratio [HR] for death = 0.61, 95% CI = 0.38-0.98, p=0.04). No statistically significant difference was observed with everolimus vs. sunitinib in OS (HR for death = 0.81, 95%CI=0.49-1.31) or in PFS (HR for progression = 0.84, 95% CI=0.46-1.53). Conclusions: After adjusting for cross-trial differences, treatment of advanced pNET with everolimus was associated with prolonged OS compared to treatment with placebo.