Background: Stage III melanoma is currently stratified by number of lymph nodes (LNs) involved. However, given the variability of LN retrieval counts we hypothesize that lymph node ratio (LNR) may also provide prognostic information.
Methods: Retrospective cohort study of 411 patients with stage III melanoma were divided into two groups based on LNR (<0.15, n = 291 and ≥0.15, n = 120).
Results: In multivariate analysis N stage (N3 vs. N1, hazard ratio [HR] = 2.13, p < 0.001), extranodal extension (HR = 1.92, p = 0.002), macrometastasis (HR = 1.70, p = 0.005), non-SLN involvement (HR = 1.65, p = 0.005), risk of N2 disease exceeding 35 % (HR = 1.51, p = 0.03), and LNR ≥0.15 (HR = 1.46, p = 0.03) were associated with overall survival (OS). LNR failed to further stratify stage III melanoma; however, the number of LNs examined was an independent prognostic factor. Patients who had >8 inguinal, >15 axillary, or >20 cervical LNs examined had fewer same nodal basin recurrences (26 [8 %] vs. 20 [20 %], p = 0.0009) and for N1 patients an improved OS (3-year OS 84 % vs. 76 %, 10-year OS 53 % vs. 34 %, p = 0.06) compared with N1 patients who had fewer LNs examined.
Conclusions: LNR is an important prognostic factor in stage III melanoma; however, it was not independent over the current AJCC TNM staging system. Diligence by the surgeon and pathologist to retrieve and examine >8 inguinal, >15 axillary, or >20 cervical LNs is associated with fewer same nodal basin recurrences and improved survival and is critical to reliable prognostication.