Prognosis in patients with sentinel node-positive melanoma without immediate completion lymph node dissection

Abstract

Abstract Background The therapeutic value of immediate completion lymph node dissection (CLND) for sentinel node (SN)-positive melanoma is unknown. The aim of this study was to evaluate the impact of immediate CLND on the outcome of patients with SN-positive melanoma. Methods Patients with SN metastases treated between 1993 and 2008 at ten cancer centres from the European Organization for Research and Treatment of Cancer Melanoma Group were included in this retrospective study. Maximum tumour size, intranodal location and penetrative depth of SN metastases were measured. Outcome in those who had CLND was compared with that in patients who did not undergo completion lymphadenectomy. Results Of 1174 patients with SN-positive melanoma, 1113 (94·8 per cent) underwent CLND and 61 (5·2 per cent) did not. Median follow-up for the two groups was 34 and 48 months respectively. In univariable survival analysis, CLND did not significantly influence disease-specific survival (hazard ratio (HR) 0·89, 95 per cent confidence interval 0·58 to 1·37; P = 0·600). However, patients who did not undergo CLND had more favourable prognostic factors. Matched-pair analysis, with matching for age, Breslow thickness, tumour ulceration and SN tumour burden, showed that CLND had no influence on survival (HR 0·86, 0·46 to 1·61; P = 0·640). After adjusting for prognostic factors in multivariable survival analyses, no difference in survival was found. Conclusion In these two cohorts of patients with SN-positive melanoma and prognostic heterogeneity, outcome was not influenced by CLND.