RESUMO
BACKGROUND: Management of duodenal neuroendocrine tumors (DNETs) is not standardized, with smaller lesions (< 1-2 cm) generally treated by endoscopic mucosal resection (EMR) and larger DNETs by surgical resection (SR). This study reviewed how patients were selected for treatment and compared outcomes. PATIENTS AND METHODS: Patients with DNETs undergoing resection were identified through institutional databases, and clinicopathologic data recorded. χ2 and Wilcoxon tests compared variables. Survival was determined by Kaplan-Meier, and Cox regression tested association with survival. RESULTS: Among 104 patients, 64 underwent EMR and 40 had SR. Patients selected for SR had larger tumor size, younger age, and higher T, N, and M stage. There was no difference in progression-free (PFS) or overall survival (OS) between SR and EMR. In 1-2 cm DNETs, there was no difference in PFS between SR and EMR [median not reached (NR), P = 0.1]; however, longer OS was seen in SR (median NR versus 112 months, P = 0.03). In 1-2 cm DNETs, SR patients were more likely to be node-positive and younger. After adjustment for age, resection method did not correlate with survival. Comparison of surgically resected DNETs versus jejunoileal NETs revealed longer PFS (median NR versus 73 months, P < 0.001) and OS (median NR versus 119 months, P = 0.004) DISCUSSION: In 1-2 cm DNETs, there was no difference in survival between EMR and SR after adjustment for age. Recurrences could be salvaged, suggesting that EMR is a reasonable strategy. Compared with jejunoileal NETs, DNETs treated by SR had improved PFS and OS.