The long-term analysis of prognostic role of mutational status in primary resectable gastrointestinal stromal tumors (GIST).

Abstract

e23517 Background: Radical surgery is the primary treatment for GIST, but up to 50% of patients relapse, mainly in form of hepatic and peritoneal metastases. GISTs are a group of tumors with various pathological and molecular features as well as different clinical courses. The aim of the study was the long-term analysis of prognostic role of mutational status in primary GIST after radical resection. Methods: The material consisted of a group of 304 patients with primary GIST diagnosed till 04/2012, with known mutational status and treated surgically with curative intend without adjuvant imatinib. Data were collected prospectively as part of the GIST clinical register. Relapse-free survival (RFS) was calculated from the date of GIST resection to the date of local recurrence, distant metastatic disease or last follow-up. Overall survival (OS) was calculated from the date of resection to the date of last follow-up or death. Survival analyses were performed using the Kaplan-Meier method and log-rank test. Multivariate Cox regression was performed to asses impact of mutational status on RFS and OS. Results: The primary tumor (PT) locations were: gastric (57.6%) vs non-gastric (42.4%). The median of tumor size was 7cm (0.5-33 cm). Mitotic index was ≤5 in 54.3%, > 5 in 45.7%. Mutations in KIT gene exon 11 were as follow: del 557-558 in 22.7% or other (point mutations, other deletions or insertions) in 40.1% with similar rates in gastric and non-gastric GIST. KIT gene exon 9 dup 502-503 (7.2%) were more common in non-gastric vs gastric GIST (16.3% vs 0.6%). PDGFR D842V was observed in 26 patients (8.6%) while other mutations in PDGFRA gene in 17 (5.6%). The median follow-up was 25.1 months. Disease recurrence was observed in 124 cases (40.8%) resulting in median RFS of 84.4 (95%CI 48.3-120.5) months. The negative independent risk factors for RFS in univariate analysis were: primary tumor location, tumor size, mitotic index and KIT exon 9 dup 502-503 mutation (P < 0.05). In the multivariate analysis independent predictive factors for RFS were mitotic index > 5 (HR 4.38, 95%CI 2.82-6.78), PT location (HR 1.56, 95%CI 1.01-2.39). In multivariate analysis in patients with gastric GIST PDGFR D842V mutations were significantly correlated with better RFS (HR 0.16, 95%CI 0.04-0.61), what was not observed in non-gastric GIST. The median OS was 198.1 months. The negative independent risk factors for OS in univariate analysis were: primary tumor location, primary tumor size and KIT gene exon 9 dup 502-503 mutation. Mitotic index was also negative independent risk factors in a multivariate analysis (HR 1.70, 95%CI 1.15 -2.52). Conclusions: In the long-term analysis mutational status of primary resectable GIST is related to risk of relapse especially in gastric location, however it has no independent impact on overall survival. Genotype should be included in modern risk classification of GIST.