Improvement and persistent disparities in completion lymph node dissection: Lessons from the National Cancer Database.

Abstract

Completion lymph node dissection (CLND) is recommended for melanoma patients with positive sentinel lymph node biopsies (SLNB); however, 50% do not undergo CLND. We sought to determine CLND trends over time, and factors contributing to variability. The NCDB was queried for patients undergoing wide local excision (WLE), with or without SLNB and CLND. Cohorts were created based on demographic/socioeconomic variables and era of treatment (Era 1: 2003-07, Era 2: 2008-12). Univariate and multivariate analyses identified factors associated with performance of or trends in CLND. 122 849 underwent WLE with SLNB. Of 24 267 (19.8%) with +SLNB, 13 594 (56.0%) continued to CLND. In multivariate analyses, Medicaid (OR 0.78; P = 0.04) or Medicare (OR 0.79; P < 0.01) in Era 1 and patients without insurance in Era 2 (OR 0.78; P = 0.01) underwent less CLND. In both eras, Blacks (OR 0.45; P < 0.01, OR 0.59; P < 0.01), head/neck lesions (OR 0.72; P < 0.01, OR 0.66; P < 0.01) and lower extremity lesions (OR 0.75; P < 0.01, OR 0.72; P < 0.01) underwent less CLND. However, Blacks experienced greatest increase in CLND usage (+9.2%). CLND usage continues to be low and racial/socioeconomic disparities persist. Until the results of MSLT-2 become available, continued focus on understanding poor adherence to, and improving rates of CLND is necessary.