Is blind pancreaticoduodenectomy justified for patients with ampullary neoplasms?

Background - Many specialists justify pancreaticoduodenectomy (PD) for pancreatic head neoplasms with suspected but unproven malignance (blind-PD). Our aim in this study was to determine whether blind-PD is also justified for ampullary neoplasms. Methods - We retrospectively reviewed the records of all patients with presumed resectable ampullary neoplasms treated at the National Taiwan University Hospital from 1998 to 2008. Results - Of the 84 patients without a preoperative tissue diagnosis of malignance, 64 had blind-PD and 20 had ampullectomy (AMP) with intraoperative frozen section. Patients with jaundice, gastrointestinal bleeding, imaging findings showing tumor invasion, and larger tumor size were significantly more frequently treated by blind-PD. Final pathological diagnosis was benign in ten of 64 blind-PD-treated patients. Conclusions - Our data support a selective use of blind-PD because (1) a significant portion (65%) of benign ampullary neoplasms can be safely and effectively treated by AMP, (2) blind-PD does not treat ampullary cancer at earlier stage, and (3) blind-PD is associated with significantly more complications and significantly longer hospital stay than AMP. However, blind-PD is strongly recommended for patients with large ampullary neoplasms (>3 cm in diameter), with jaundice, or with malignant endoscopic appearance. ? 2009 The Society for Surgery of the Alimentary Tract.