RESUMO
PURPOSE: The prognosis for stage III melanoma patients is mixed, and there is need for new prognostic factors to be incorporated into a revision of melanoma TNM staging. We analyzed the possible role of the timing of lymph node involvement as an important prognostic factor. METHODS: Among 249 melanoma patients who underwent ilioinguinal lymphadenectomy, a group of 185 patients with a thick (>4mm) melanoma and full clinical data available was analyzed. The mean depth of invasion was 5.85 mm; the tumor was ulcerated in 67 cases (36.2%); and Clark V was diagnosed in 82 patients (44.3%). The median interval between primary excision and the time of lymphadenectomy was 11.1 months. RESULTS: Recurrent disease was reported in 150 of 185 patients. The first sites of recurrence were the skin in 15.7%, lymph nodes in 13.5%, and distant metastases in 28.7%; the remaining 43 patients (23.2%) had multifocal recurrences. In all, 35 patients (18.9%) were disease-free. Skip metastases (positive iliac and negative inguinal lymph nodes) were found in 26 patients (14%). Multivariate Cox analysis showed that only the time between the first surgery and lymphadenectomy and the number of involved nodes were significant predictors of survival. Relative risk of death was 5.2 times higher for patients who had simultaneously undergone lymphadenectomy (compared to lymph dissection performed >1 year after primary excision) and about 2.7 times higher for those with more regionally advanced disease (pN3 vs. pN1). CONCLUSIONS: The long disease-free interval before the development of lymph node metastases and before node dissection is a favorable prognostic factor independent of other well known parameters.