Predictive Effectiveness of the Glasgow Prognostic Score for Gastrointestinal Stromal Tumors

Abstract

The aim of this study was to evaluate the significance of the Glasgow prognostic score (GPS) in patients with resected gastrointestinal stromal tumors (GISTs). Forty-six GIST patients who underwent radical resection between January 2004 and December 2011 were enrolled in this retrospective study. The clinicopathological parameters examined included predictors of recurrence-free survival (RFS). Univariate and multivariate analysis of prognostic factors related to RFS were calculated using Cox proportional hazards model. The GPS classification system revealed 37 (80.4%), 6 (13.1%), and 3 (6.5%) patients with a GPS of 0, 1, and 2, respectively. Patients with GPS 1/2 had a significantly shorter RFS compared to those with GPS 0 (P = 0.01). The 3- and 5-year RFS rates for patients with GPS 0 were 94.0% and 90.9%, respectively, compared to 66.7% and 53.3%, respectively, for patients with GPS 1/2. Univariate analyses indicated that tumor size (P < 0.01), mitotic rate (P < 0.01), higher GPS (P < 0.01), and platelet count (P = 0.04) were prognostic factors for RFS; tumor size (P = 0.01) and GPS (P = 0.04) were independent prognostic factors in multivariate analysis. Preoperative high GPS were predictors of long-term prognosis in patients with resected GISTs.