A systematic review and meta-analysis comparing pancreaticoduodenectomy versus limited resection for duodenal gastrointestinal stromal tumors.

Purpose: This study was designed to compare the clinical outcomes of patients who underwent limited resection (LR) versus pancreaticoduodenectomy (PD) for duodenal gastrointestinal stromal tumors (GISTs).

Methods: A systematic review of the literature was performed to identify studies analyzing the clinical outcomes of LR and PD for duodenal GISTs.

Results: Eleven studies were included, of which 7 that compared 162 patients who underwent LR versus 98 patients who underwent PD were suitable for meta-analysis. Patients who underwent PD were more likely to have tumors which were large (≥ 5 cm) [76.0 vs. 36.6 %, odds ratio (OR) 5.49, 95 % confidence interval (CI) 1.8-16.76], with high mitotic count ≥5/50 high-power field (HPF) (33.7 vs. 18.5 %, OR 2.23, 95 % CI 1.22-4.08), classified as high risk (60.3 vs. 32.0 %, OR 3.23, 95 % CI 1.65-6.34), and which were located at D2 (80.5 vs. 28.6 %, OR 10.33, 95 % CI 5.22-20.47) compared with LR. PD was associated with a higher postoperative morbidity rate than LR [48.3 vs. 20.7 %, relative risk (RR) 2.34, 95 % CI 1.61-3.42]. LR was not associated with an increased local recurrence rate, had a better DFS [hazard ratio (HR) 2.07, 95 % CI 1.07-4.01], and lower rate of distant metastasis (8.9 vs. 25.8 %, OR 0.28, 95 % CI 0.13-0.59) compared with PD.

Conclusions: LR should be the procedure of choice for duodenal GIST whenever technically feasible, because it is associated with good oncologic outcomes and lower morbidity compared with PD. The oncologic outcome of GIST is more likely to be dependent on tumor biology rather that the type of surgical resection. The use of Imatinib in patients with duodenal GIST may potentially allow a proportion of patients who would otherwise require a PD to undergo LR instead.