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

Objective: Theoretically, completion lymph node dissection (CNLD) should have the lowest benefit in the absence of nonsentinel lymph node (NSLN) metastases. For this reason, substantial research efforts have attempted to define specific criteria that are associated with a low-enough risk of NSLN positivity so that CLND can be deferred. Our objectives were (1) to identify features associated with low risk of NSLN positivity in sentinel lymph node-positive cutaneous melanoma of the head and neck (CMHN) and (2) to analyze the effect of CLND on 5-year disease-specific survival (DSS) among subgroups stratified by risk of NSLN metastasis.

Study Design: Retrospective analysis of population-based data.

Setting: SEER database.

Subjects And Methods: Patients with sentinel lymph node-positive CMHN were categorized according to lymph node treatment following sentinel lymph node biopsy (SLNB): 210 underwent CLND and 140 deferred. Clinicopathologic characteristics and survival were compared between SLNB+CLND and SLNB-only groups. Survival analyses were stratified by age and characteristics associated with NSLN positivity.

Results: Minimal tumor thickness and nonulceration were associated with lowest risk of positive NSLN (P < .025). In the subgroup with the lowest risk of metastasis, patients aged <60 years who underwent CLND+SLNB had markedly better DSS than those receiving SLNB only (>90% vs <25%; P < .0025). Paradoxically, in subgroups with a higher risk of NSLN metastasis, DSS was similar whether CLND was performed or not (P > .25).

Conclusions: Selecting patients for CLND according to risk of NSLN metastasis may be a suboptimal strategy for improving DSS. We believe that CLND should not be withheld on the basis of "low risk" features in CMHN.