Surgery for gastrointestinal stromal tumors (GISTs) of the stomach and small bowel: short- and long-term outcomes over three decades

Abstract

Background: Many studies on gastrointestinal stromal tumors (GISTs) derive from tertiary referral centers, but few examine strictly population-based cohorts. Thus, we evaluated the clinical features, surgical treatments, clinical outcomes, and factors predicting the survival of patients with GISTs in a population-based series.

Methods: Patients with GISTs diagnosed at Stavanger University Hospital over three decades (1980-2012) were analyzed. Data were retrieved from hospital records. Descriptive statistics and survival analyses (Kaplan-Meier) are presented. A limited number of colorectal GISTs (n = 6) restricted most analyses to those with a gastric or small bowel location.

Results: Among 66 patients surgically treated for GISTs, 60 patients (91 %) had either a gastric or a small bowel localization. Females comprised 61 %. The median age at diagnosis was 63 (range, 15-88) years. Clinical symptoms were recorded in 43 patients (65 %). Complete tumor resection was achieved in 85 % of the patients. During follow-up, 6 patients were surgically treated for local recurrence or metastatic disease. The median follow-up time was 6.1 years. At last follow-up, 30 patients (46 %) were deceased, 10 of whom died from GISTs. The median overall survival was 10.4 years. For GISTs with a gastric or small bowel location, a 1- and 5-year disease-specific survival of 100 and 96 %, and a relapse-free survival of 96 and 78 % were observed. Male gender, incidental diagnosis, smaller tumor size, a low mitotic rate, an intact pseudocapsule, low-risk categorization, and an early stage were significantly associated with improved outcomes.

Conclusion: Surgery in a low-volume, population-based setting yields enhanced long-term disease and recurrence-free survival for patients with GISTs of the stomach or small bowel. Incidental diagnosis, complete tumor resection, and low-risk categorization are good predictors of long-term prognosis.