A systematic review of post first-line treatments for advanced gastrointestinal stromal tumor (GIST): Direct pairwise meta-analyses and indirect comparisons.

Abstract

195 Background: Treatment for metatastatic or unresectable GIST is with imatinib. The majority of patients eventually acquire imatinib resistance, prompting the development of a growing number of agents as post-first-line treatment. Currently no studies directly compare these treatments. Methods: A systematic review was performed through MEDLINE, EMBASE, CENTRAL, and ASCO meeting abstracts up to July 2014 to identify randomized controlled trials that included GIST patients who were previously treated with a first-line chemotherapy for advanced disease. Progression-free survival (PFS) and overall survival (OS) with 95% credible regions were extracted using the Parmar method. Direct pairwise meta-analyses and indirect comparisons using the Bucher method were performed. Results: Four studies were identified for the systematic review. 1 study (n=312) showed that sunitinib in the second-line setting (vs. placebo) improved PFS but not OS. 3 studies (n=528) examined the third-line setting (imatinib resumption vs. placebo; regorafenib vs. placebo; nilotinib vs. best supportive care with or without imatinib or sunitinib). Of the 3 third-line studies, there was significant heterogeneity between placebo-controlled trials and the non-placebo controlled trial (I2= 98%). Direct pairwise meta-analysis using random-effects of the 2 placebo-controlled studies showed that the PFS hazard ratio (HR) was 0.63 (0.22-0.61, p=0.0001), whereas the PFS HR for the non-placebo controlled study was 0.90 (0.65-1.26, p=0.56). These 2 HRs are different statistically (interaction: p=0.002). Indirect comparisons of imatinib resumption vs. regorafenib suggested that the PFS HR was 0.59 (0.31-1.11, p=0.10), trending in favor of regorafenib. OS HRs were not significant for direct or indirect comparison in the third-line setting. Conclusions: The number and size of completed phase III studies and the lack of a standard comparator arm limit the use of direct and indirect comparison methods to determine the best therapeutic option for patients who have progressed on imatinib. At this time, clinicians should interpret the available evidence at an individual level.