Clinical and histologic involvement of regional lymph nodes in malignant melanoma. Adjuvant vindesine improves survival

Abstract

This report is a study of prognostic factors, including adjuvant chemotherapy, that influence survival of patients with malignant melanoma who have clinical and pathologic involvement of regional lymph nodes. A total of 169 evaluable patients with malignant melanoma metastatic to regional lymph nodes were registered consecutively and prospectively between June 1977 and December 1986 in the computerized data base of the melanoma registry at Westminster Hospital. Eighty-seven of these patients received adjuvant chemotherapy with vindesine after resection of palpable metastatic lymph nodes, and 82 had no systemic treatment after surgery. All were followed up for at least 2 years (median, 8 years) after involvement of regional lymph nodes was noted or until death. Statistical analyses included simple life-table comparisons, unadjusted for covariates. In addition, Breslow's thickness, ulceration of the primary lesion, its anatomical location, number of regional lymph nodes histologically involved, dissection site, patient age and sex, and adjuvant vindesine therapy were included as covariates in Cox regression models. The disease-free interval (P = 0.0001), time to dissemination from lymph node metastases (P < 0.0001), survival time after lymph node dissection (P = 0.0227) and overall survival time after initial diagnosis of malignant melanoma (P = 0.0095, log-rank chi-square test) were superior for the 87 patients who received adjuvant chemotherapy with vindesine. Cox regression analysis confirmed adjuvant vindesine as a highly significant variable influencing all of these outcomes, including overall survival time after first diagnosis (P = 0.003). The apparent effect of adjuvant vindesine on overall survival in this study is large (hazard ratio, 0.52) and highly statistically significant. Adjuvant vindesine therapy merits consideration for malignant melanoma metastatic to regional lymph nodes. However, these results observed in concurrent, but nonrandomized, patients clearly require confirmation.