Indirect treatment comparison of cabozantinib vs sunitinib in the treatment of pancreatic neuroendocrine tumors.

Abstract

e16327 Background: Pancreatic neuroendocrine tumors vary widely in grade, functionality, growth rate and aggressiveness. Patients are often diagnosed at a locally advanced or metastatic stage. Treatment options include surgery, somatostatin analogs, targeted therapies, radionuclide therapy, and chemotherapy. Though there is no universally accepted optimal therapy sequence, tyrosine kinase inhibitors are increasingly used for disease stabilization. Sunitinib has been shown to improve progression free survival, overall survival and objective response rate in a placebo-controlled trial and more recently cabozantinib has also been shown to improve progression free survival. No direct head-to-head comparison between these two agents has been performed in a trial setting. Methods: Data pertaining to the CABINET (cabozantinib vs. placebo) and NCT00428597/SUN1111 (sunitinib vs. placebo) clinical trials was analyzed; Hazard ratios for progression free survival for these trials were extracted along with 95% confidence intervals. CABINET was a phase 3 randomized double-blind placebo controlled clinical trial with 2:1 randomization with 2 cohorts - extrapancreatic neuroendocrine tumors and pancreatic neuroendocrine tumors. NCT00428597 was a phase 3 randomized double-blind placebo controlled clinical trial with 1:1 randomization. For the CABINET trial, only data pertaining to the pancreatic neuroendocrine tumor cohort was included. The placebo arm was used as the common comparator. Baseline patient population characteristics were compared along with treatment emergent adverse events. Hazard ratios for progression free survival were compared using the Bucher method for indirect comparison. Upper and lower limits of the 95% confidence intervals were also calculated. Results: Indirect treatment comparison of the progression free survival between cabozantinib and sunitinib yields a HR of 0.55 indicating superior efficacy of cabozantinib. However, this result is not statistically significant (95% CI 0.25 to 1.25). Both medications have significant toxicities; although cabozantinib has been noted to have more frequent grade 3+ toxicities, common dose reductions, and discontinuation, it appears to have better efficacy. This result is pertinent as cabozantinib appears to be active in patients who are previously treated with sunitinib. Adjustment for baseline characteristics of the state populations did not significantly alter the results. Conclusions: Indirect treatment comparison of cabozantinib and sunitinib in the treatment of pancreatic neuroendocrine tumors appears to show superior efficacy of cabozantinib. Given the significant side effect profile of cabozantinib, it may be worthwhile to consider a head-to-head study between sunitinib and a lower dose of cabozantinib to establish the efficacy of cabozantinib with the possibility of a lower side effect profile.