Pancreatic Frozen Section Guides Operative Management With Few Deferrals and Errors: Five-Year Experience at a Large Academic Institution

* Context.--Pancreatic adenocarcinoma is the third leading cause of cancer death in the United States. Surgery remains the mainstay of treatment, and frozen section analysis is used to confirm diagnosis and determine resectability and margin status. Objective.--To evaluate use and accuracy of frozen section and how diagnosis impacts surgical procedure. Design.--We reviewed patients with planned pancreatic resections between January 2014 and March 2019 with at least 1 frozen section. Pathology reports including frozen sections, preoperative cytology, and operative notes were reviewed. Frozen sections were categorized by margin, primary pancreatic diagnosis, metastasis, or vascular resectability. The deferral and error rates and surgeons' response were noted. Results.--We identified 898 planned pancreatic resections and 221 frozen sections that were performed on 152 cases for 102 margins, 94 metastatic lesions, 20 primary diagnoses, and 5 to confirm vascular resectability. The diagnosis was deferred to permanent sections in 13 of 152 cases (8.6%) on 16 of 221 frozen sections (7.2%): 6 for metastasis, 8 for margins, and 2 for primary diagnosis. Discrepancies/errors were identified in 4 of 152 cases (2.6%) and 4 of 221 frozen sections (1.8%). Surgeons' responses were different than expected in 8 of 221 frozen sections (3.6%), but their actions were explained by other intraoperative findings in 6 of 8. Conclusions.--Frozen section remains an important diagnostic tool used primarily for evaluation of margins and metastasis during pancreatectomy. In most cases, a definitive diagnosis is rendered, with occasional deferrals and few errors. Intraoperative findings explain most cases where surgeons act differently than expected based on frozen section diagnosis.