Background: In gastrointestinal stromal tumors (GISTs), rupture is a high-risk feature and an indication for adjuvant treatment; however, the independent impact of rupture on prognosis is uncertain and the term is inconsistently defined. In the present study, a previously proposed definition of 'tumor rupture' was applied on a population-based cohort of gastric GISTs.
Methods: Patients undergoing surgery for non-metastatic gastric GISTs from 2000 to 2015 were identified in the regional sarcoma database of Oslo University Hospital. Tumor rupture included spillage or fracture, piecemeal resection, incisional biopsy, blood-tinged ascites, gastric perforation, and microscopic adjacent infiltration. Minor defects of tumor integrity were not considered rupture, i.e. core needle biopsy, peritoneal tumor penetration, superficial peritoneal rupture, and R1 resection. Risk was assessed according to the modified National Institutes of Health consensus criteria.
Results: Among 242 patients, tumor rupture occurred in 22 patients and minor defects of tumor integrity occurred in 81 patients. Five-year recurrence-free survival (RFS) for patients with tumor rupture, minor defects of tumor integrity, and no defect was 37, 91, and 96%, respectively (p < 0.001). In the high-risk group, 5 year RFS for patients with rupture was 37%, versus 77% without rupture (hazard ratio 3.56, 95% confidence interval 1.57-8.08, p = 0.001). On multivariable analysis, tumor rupture and mitotic index were independently associated with recurrence. Of 13 patients who received adjuvant imatinib after tumor rupture, 11 relapsed.
Conclusions: Tumor rupture according to the present definition was independently associated with recurrence. With tumor rupture, patients relapsed despite adjuvant treatment. Without rupture, prognosis was good, even in the high-risk group.