Abstract
PURPOSE: The effect of high-dose imatinib (800 mg/day) on survival in the adjuvant treatment of patients with resected KIT exon 9-mutated gastrointestinal stromal tumors (GIST) is not established. Here, the association of dose and other clinicopathologic variables with survival was evaluated in a large multi-institutional European cohort. EXPERIMENTAL DESIGN: Data from 185 patients were retrospectively collected in 23 European GIST reference centers. Propensity score matching (PSM) and inverse-probability of treatment weighting (IPTW) were used to account for confounders. Univariate and multivariate unweighted and weighted Cox proportional hazard regression models were estimated for relapse-free survival (RFS), modified-RFS (mRFS) and imatinib failure-free survival (IFFS). Univariate Cox models were estimated for overall survival. RESULTS: Of the 185 patients, 131 (70.8%) received a starting dose of 400 mg/d and the remaining 54 (29.2%) a dose of 800 mg/d. Baseline characteristics were partially unbalanced, suggesting a potential selection bias. PSM and IPTW analyses showed no advantage of imatinib 800 mg/d. In the weighted multivariate Cox models, high-dose imatinib was not associated with the survival outcomes [RFS: hazard ratio (HR), 1.24; 95% confidence interval (CI), 0.79-1.94; mRFS: HR, 1.69; 95% CI, 0.92-3.10; IFFS: HR, 1.35; 95% CI, 0.79-2.28]. The variables consistently associated with worse survival outcomes were high mitotic index and nongastric tumor location. CONCLUSIONS: In this retrospective series of patients with KIT exon 9-mutated GIST treated with adjuvant imatinib, a daily dose of 800 mg versus 400 mg did not show better results in terms of survival outcomes. Prospective evaluation of the more appropriate adjuvant treatment in this setting is warranted.