Local ampullary resection with careful intraoperative frozen section evaluation for presumed benign ampullary neoplasms.

BACKGROUND: Frozen section evaluation has been reported to be inaccurate in detecting foci of adenocarcinoma within adenomas of the ampulla of Vater, leading many authors to advocate pancreaticoduodenectomy as the method of treatment for these neoplasms. The authors hypothesized that (1) ampullary resection is less morbid than pancreaticoduodenectomy, and (2) frozen section evaluation following ampullary resection is accurate and allows for a selective application of pancreaticoduodenectomy to those with carcinoma or benign lesions too large to be locally resected.
METHODS: A retrospective review of a single-surgeon experience was conducted. Thirty-eight patients who underwent ampullary resection and pancreaticoduodenectomy (39 procedures) for benign and malignant ampullary neoplasms were identified. Our technique of step-frozen section analysis is described.
RESULTS: Twenty-one ampullary resections were performed for preoperative diagnoses of benign (16) and malignant (5) ampullary neoplasms. Frozen section evaluation accurately predicted the final histology in all patients undergoing ampullary resection. Ampullary resection (vs pancreaticoduodenectomy) was associated with a statistically lower operative time (169 minutes vs 268 minutes), estimated blood loss (192 mL vs 727 mL), mean length of stay (10 days vs 25 days), and overall morbidity (29% vs 78%).
CONCLUSIONS: Frozen section evaluation of ampullary neoplasms is accurate. Because ampullary resection is less morbid than pancreaticoduodenectomy and frozen section evaluation is accurate, ampullary resection with frozen section evaluation is our current approach to the treatment of small benign ampullary neoplasms.