Ampullectomy versus pancreaticoduodenectomy for ampullary tumors.

BACKGROUND Considerable controversy surrounds the treatment of ampullary neoplasms. This report describes the authors' experiences regarding the choice of either ampullectomy or pancreaticoduodenectomy for treatment of ampullary tumors. METHODS Demographics, statistical findings concerning diagnosis, surgical risks including morbidity and mortality, and outcomes were evaluated and compared between the ampullectomy and pancreaticoduodenectomy groups for ampullary tumors retrieved from a prospectively collected computer database of 992 periampullary tumors resected during the period from 1965 to 2013. RESULTS A total of 377 patients with ampullary tumors were included; 15 underwent ampullectomy and 362 underwent pancreaticoduodenectomy. The overall false-negative rate for diagnosis of ampullary malignancy was 11.2%, specificity was 50.0%, positive predictive value was 98.3%, negative predictive value was 12.2%, and the overall accuracy was 87.6% (77.5% by preoperative endoscopic biopsy and 83.9% by intraoperative frozen-section biopsy). Ampullectomy was associated with shorter postoperative stays and lower surgical morbidity. There was no statistical difference observed between the two groups regarding surgical mortality, pancreatic leakage, or gastric atonia. The tumor recurrence rate was lower after pancreaticoduodenectomy, but the difference between the groups was not significant. Overall, there was no difference in survival observed between the two groups. CONCLUSION Because biopsy is not routinely reliable, pancreaticoduodenectomy is preferable to ampullectomy for an ampullary tumor of uncertain diagnosis. Ampullectomy is associated with lower surgical morbidity and should therefore remain in the armamentarium of the pancreatic surgeon when comorbidity precludes major surgery.