Prognostic factors after surgery of primary resectable gastrointestinal stromal tumours

Aims To analyse the prognostic factors of gastrointestinal stromal tumours (GIST) after a complete resection of the primary tumour. Patients and methods Fifty-nine patients who underwent a complete initial resection of a GIST were studied. Peritumoral resections (PTR) were compared to segmental organ resections (SOR). Overall survival (OS) and the disease-free survival (DFS) were calculated using the Kaplan–Meier method. Results Primary sites were: stomach (25), small intestine (22), rectum (7), duodenum (5). Two patients had nodal involvement. The median follow-up was 45 months. Local DFS was significantly better after SOR compared to PTR (median 63 vs. 11 months, respectively, p<0.001). Univariate analysis for OS identified the grade (p=0.005) and size (p=0.02) as prognostic factors. Only a high histologic grade was an independent factor (p=0.02) in the multivariate analysis. Out of 49 patients who relapsed, the first recurrence was local only in 12, local and distant in 10 and distant only in 27; only one had a lymph node failure. Recurrences were accessible to curative surgery in 22 cases. OS of patients submitted to complete resection of their recurrence was significantly better than patients whose recurrence could not be resected (median 52 vs. 12 months, respectively, p<0.001). Conclusion Complete surgery without rupture remains the mainstay of treatment in patients with localized, resectable disease. A peri-tumoral resection confers a high risk of local recurrence and should be avoided. Lymphadenectomy is not systematic. Grade is the main prognostic factor for OS and can be a decision marker for adjuvant treatment with Gleevec.