Impact of c-kit mutations, including codons 557 and/or 558, on the recurrence-free survival after curative surgery in patients with GIST.

Abstract

12 Background: Recently, c-kit exon 11 deletions, including codons 557 and/or 558, have been reported to predict a worse prognosis in GIST patients. However, it is difficult to prove the correlation between genotype and tumor aggressiveness in the imatinib- adjuvant era because exon 11 mutations respond well to imatinib. In this study, we evaluated the impact of c-kit mutational status on recurrence-free survival (RFS) after resection of primary GIST. Methods: Clinical and pathological characteristics of 89 GIST patients in our single institution study were retrospectively analyzed. Tumors were categorized into 4 subgroups based on their mutational locations; A1: mutated codons including neither 557 nor 558, A2: either 557 or 558, B1: only 557 and 558, B2: both 557 and 558. All of the patients underwent curative surgery, and none received adjuvant imatinib. The median duration of follow-up was 49 months. Results: Tumors originated from the stomach (n=75/89, 84%), small intestine (n=10), and colorectum (n=4). Mutation subgroup B was associated with both Fletcher and Miettinen high-risk categories. The 2-year recurrence free survival rate for A1, A2, B1, B2, was 84.9%, 85.7%, 50%, 57%, respectively. Group B2 had a significantly worse RFS than groups A1 (p=0.0004) and A2 (p=0.0014). Multivariate analysis for RFS indicated that only the mutational subgroup was a significant prognostic factor (p=0.03, HR=2.42). Conclusions: C-kit mutations, including both 557 and 558, affected the RFS of GIST patients after curative surgery, but those including either 557 or 558 did not. Our results indicate that the locations of c-kit mutations are associated with PFS, and they may therefore affect the selection of candidates with GIST for adjuvant imatinib. No significant financial relationships to disclose.