Prognostic Significance of a Positive Nonsentinel Lymph Node in Cutaneous Melanoma

Abstract

Sentinel lymph node (SLN) biopsy provides important prognostic information for patients with cutaneous melanoma. There may be additional prognostic significance to melanoma spreading from the SLN to nonsentinel lymph nodes (NSLN). We examined the implications of a positive NSLN for overall and distant disease-free survival. Using a prospectively maintained, Institutional Review Board-approved melanoma database we studied patients who had a cutaneous melanoma, a positive SLN, and a completion lymph node dissection (CLND). Survival was determined using a combination of hospital records and the Social Security Death Index (SSDI). Univariate and multivariate Cox regression analysis was performed to further characterize predictors of overall and distant disease-free survival. Kaplan-Meier analysis was used to generate survival curves. A total of 429 patients with positive SLN biopsies were identified, with at least one positive NSLN identified in 71 (17%). Median follow-up time was 36.8 months. Presence of a positive NSLN was significantly associated with poor outcome, although long-term survival was possible. Presence of ulceration, high mitotic rate, angiolymphatic invasion, total number of positive nodes, and volume of disease>1% in the SLN were significant predictors of survival on univariate analysis, but lost significance on multivariate. Multivariate Cox analysis revealed several predictors of overall survival: increasing age [hazard ratio (HR) 1.04, P<0.01], Breslow depth (HR 1.76, P<0.01), presence of extracapsular extension in the SLN (HR 2.39, P<0.01), and positive NSLN (HR 1.92, P<0.01). Among node-positive melanoma patients, presence of a positive NSLN is a highly significant poor prognostic sign, even after considering the total number of positive nodes and volume of disease in the SLN. CLND after a positive SLN provides this important prognostic information.