RESUMO
BACKGROUND: The utility of endoscopic ampullectomy versus surgical ampullectomy remains a topic of debate, particularly for the treatment of malignant tumors. The goal of this study was to prospectively establish the outcomes of endoscopic ampullectomy, with focus on carcinoma. METHODS: From February 2002 to December 2008, 61 patients underwent endoscopic ampullectomy. The ampulloma was discovered fortuitously in the majority of cases (43 patients, 70%). All patients had had an echoendoscopy revealing a T1N0 lesion without invasion of the orifice. RESULTS: Forty-three patients (70%) underwent a monobloc resection. Histological analysis revealed a low-grade dysplastic adenoma in 21 patients (35%), a high-grade dysplastic adenoma in 11 patients (18%), no dysplasia in 16 patients (26%), an adenocarcinoma in 10 patients (16%), and a well-differentiated endocrine carcinoma in 3 patients (5%). Among the ten patients with adenocarcinoma, four with adenocarcinoma of poor prognosis were treated by pancreaticoduodenectomy (PD) with R0 resection, of whom one patient had no more lesion. Two intramucosal adenocarcinomas were cured by endoscopic ampullectomy without any recurrence. Four patients received palliative care after endoscopic ampullectomy due to cephalic pancreatico-duodenectomy contraindication. For the three patients with well-differentiated endocrine carcinomas, one was treated by PD with R0 resection and two were treated solely by endoscopic ampullectomy, without recurrence. Eleven patients (18%) presented with complications. The complication rate was 30.5% for carcinomas versus 14.5% for benign tumors (p < 0.05). CONCLUSION: Endoscopic ampullectomy allows for the oncologic resection of well-differentiated intramucosal carcinomas with negative margins. The risk of complications is greater for papillary carcinomas.