Completion Node Dissection Based on Risk of Nonsentinel Metastasis in Head and Neck Cutaneous Melanoma

Abstract

Objectives: Data describing the clinical implications of withholding completion lymph node dissection (CLND) from low-risk patients with sentinel lymph node (SLN)-positive melanoma are virtually non-existent, possibly due to the intuitiveness of the notion that only patients with positive non-SLNs will benefit from CLND. Furthermore, none of the published risk stratification models have exclusively analyzed patients with cutaneous melanoma of the head and neck (CMHN). Our objectives were: 1) Identify features associated with a low risk of non-SLN positivity in CMHN, and 2) Analyze the effect of CLND on disease-specific survival (DSS) among patient subgroups stratified by risk of non-SLN metastasis. Methods: Using the Surveillance Epidemiology and End Results database, we identified 350 patients with SLN-positive CMHN. Of these, 210 underwent CLND and were classified based on non-SLN status. Clinicopathologic characteristics were compared between non-SLN positive and non-SLN negative subgroups. DSS was compared between SLN biopsy (SLNB)-only and SLNB+CLND groups by Kaplan-Meier analyses and the log-rank test. Results: Minimal tumor thickness and non-ulceration were associated with the lowest risk of non-SLN positivity ( P < 0.025). Only 1 of 19 melanomas ≤1mm had non-SLN metastasis; however, patients <age 60 in this subgroup who underwent SLNB-only had a 5-year DSS rate of <25% vs. >90% among similar patients who underwent CLND ( P < 0.0025). DSS was similar in the SLNB-only and SLNB+CLND groups when melanomas were ulcerated or >2mm thick ( P > 0.25). Conclusions: Selecting patients for CLND based on risk of NSLN metastasis may be an inappropriate strategy. In CMHN, CLND should not be withheld based on low risk features.