Accuracy of Intraoperative Frozen Section Analysis of Lymph Nodes in Women Undergoing Axillary Sampling for Treatment of Breast Cancer: Single Institution Audit

Background: Axillary lymph node (LN) positivity is an important prognostic factor in breast cancer. Almost 30% clinically node negative (cN0) early breast cancers have positive nodes on pathology, wherein an axillary dissection is done as a second stage surgery. Intra operative frozen section (FS) potentially avoids redo surgery. Materials and Methods: We performed a retrospective audit for the false negative rate of intraoperative FS, from 2014 to 2018. All cN0 women undergoing upfront surgery, who underwent low axillary sampling (LAS) with FS were included. Results: Of 22,854 breast cancer cases, 2230 underwent LAS, of which 877 were node positive. Intraoperative FS was negative in 1423/2230 (63.81%) cases, of which 71/1423 (4.98%) were false negative, and came positive on final histopathology report (HPR). These 71 women had a median of 5 nodes (mean 4.85) dissected on FS (range 1-12) with a median 1 (mean 1.3) node positive (range 1-6) on HPR. The sensitivity of FS was 91.89% (95% CI, 89.89-93.62), with a negative predictive value of 95.01% (95% CI, 93.84-95.97), accuracy of 96.73% (95% CI, 95.90-97.43) and false negative rate 4.98%. On logistic regression analysis, micrometastasis (Odds ratio (OR) 7.76, 95% CI, 3.49-17.25, P < .001) lobular histology (OR 2.50, 95% CI, 1.007-6.223, P = .04) and nodes dissected (OR 1.18, 95% CI, 1.07-1.30, P = .001) were associated with higher false negative FS, and extra nodal extension (OR 0.32, 95% CI, 0.18-0.57, P <= .001) with lower false negative FS. Conclusion: The high concordance between intraoperative FS and definitive histology, suggests it's routine use for Sentinel lymph node/LAS LN can help avoid a second surgery.