Background: The duodenum is a rare site of primary gastrointestinal stromal tumor (GIST). Overall (OS) and disease-free survival (DFS) after limited resection (LR) versus pancreaticoduodenectomy (PD) were studied.
Methods: All patients who underwent surgery for primary, localized duodenal GIST between 2000 and 2011 were identified from four prospective institutional databases. OS and DFS were calculated by Kaplan-Meier method. Univariate analysis was performed.
Results: Eighty-four patients (median follow-up 42 months) underwent LR (n = 56, 67%) or PD (n = 28, 33%). Patients in the PD group had a larger median tumor size (7 cm vs. 5 cm, p = 0.024) and higher mitotic count (39% vs. 19% >5/50 high-power fields, p = 0.05). Complications were observed in five patients (9%) in the LR group and ten patients (36%) in the PD group. OS and DFS for the entire cohort were 89% and 64% at 5 years, respectively. No difference in outcome between LR and PD were observed. Eleven patients were treated with preoperative IM. A major RECIST response was obtained in nine (80%), whereas two had stable disease. Twenty-three patients received postoperative Imatinib (IM). A trend toward a better OS in IM-treated patients could be detected only in the high-risk group.
Conclusions: Type of duodenal resection does not impact outcome. The choice should be determined by duodenal site of origin and tumor size. IM may be considered in cases at high risk of recurrence; in neoadjuvant setting, IM might facilitate resection and possibly increase the chance of preserving normal biliary and pancreatic anatomy.