Frozen section analysis of SLNs in trunk and extremity melanoma has a high false negative rate but can spare some patients a second operation

Objectives: The purpose of this study was to evaluate the accuracy of frozen section (FS) analysis of sentinel lymph nodes (SLN) in melanoma. Methods: Five hundred seventy-one patients underwent FS analysis of SLN between 1/2000 and12/2010. Surgical and pathological characteristics, recurrence, and survival were analyzed. Comparisons were made using χ² and Fisher's exact t-test. Results: One hundred thirty-three (23%) patients were SLN positive of which 63 (47.4%) were identified on FS. 16/70 SLN metastases not identified on FS (23%) were seen only on immunohistochemistry. FS analysis detected 84% of SLN metastasis >2 mm. SLN FS false negative rate was 53%, positive predictive value 100%, negative predictive value 88%, and overall accuracy 89%. Among patients with a FS positive SLN, 17/63 (27%) had additional positive nodes on CLND, versus 1 of 70 (1.4%) with a positive SLN identified only on permanent section pathology (P < 0.0001). The nodal recurrence rate following a negative SLN biopsy was 5%. Conclusions: FS analysis for SLNs spared approximately half of patients a second operation. Patients with a positive SLN detected on FS were more likely to have further nodal involvement. In our experience intraoperative pathologic analysis of melanoma SLNs does not impair our ability to detect SLN metastasis or lead to a high rate of false positive results or nodal recurrences. J. Surg. Oncol. 2016;114:879–883.