RESUMEN
BACKGROUND: Standard of care is to perform a complete lymph node dissection (CLND) in melanoma patients with positive sentinel lymph nodes (SLNs). However, less than 20% will have metastases in non-SLNs. The S classification was described to predict the non-SLN status, hoping to identify a subset of patients who can be spared the CLND. We tried to validate the feasibility and usefulness of this classification. MATERIALS AND METHODS: We performed a retrospective chart review. All melanoma cases between 1996 and 2006 were included, and 359 patients with SLN biopsies were identified. All pathology slides were reviewed with an emphasis on the S classification. RESULTS: There were 365 SLN biopsies performed. A total of 82 patients (22.8%) had positive SLNs, while 277 patients (77.2%) had negative SLNs. There were 22 patients classified as SI, 18 as SII, 37 as SIII, and 5 were unclassified. On CLND, only 10 patients (12.2%) had positive non-SLNs. None of these were classified as SI while 2 patients (11%) were classified as SII and 8 (22%) as SIII. The S category was found to be a predictor of non-SLN status, and this reached statistical significance (P = 0.044). On univariate analysis, only an increasing Breslow depth and ulceration were predictive of a non-SI status. CONCLUSION: Our results suggest that the S classification is easily feasible and predicts the status of non-SLNs. No patient with SI status was found to have additional non-SLN positive nodes. A larger-scale, prospective trial should be done to confirm these results and possibly spare patients the morbidity of CLND with a positive SLN.