Is adjuvant radiotherapy necessary after positive lymph node dissection inhead and neck melanomas?

Introduction: Postoperative radiotherapy (PR) has been recommended in patie nts with advanced head and neck melanomas to improve regional control. This study examined the incidence of cervical recurrence among patients who did not receive PR after surgical management of node-positive head and neck me lanomas. Methods: A computerized search of a database listing more than 10,000 patie nts with melanoma prospectively acquired between 1971 and 1998 identified 2 17 patients with pathologically positive nodes who had undergone regional l ymph node dissection (RLND). Of these patients, 21 had received PR and 196 had not. Results: Median follow-up after RLND was 20 months for nonsurvivors and 32 months for survivors. The overall incidence of cervical recurrence was 14% (27/196). The 5-year cervical recurrence-free survival rate was 83%. Five-y ear cervical recurrence-free survival rates were 69% vs. 87% for patients w ith vs. without extranodal disease (P = .004), 96% vs. 81% for patients wit h nonpalpable vs, palpable nodes (P = .0761), and 82% vs. 91% for patients with one to three positive nodes vs, more than three positive nodes (P = .2 56). Multivariate analysis, which included the timing of nodal disease pres entation and the effect of systemic adjuvant therapy, identified extranodal disease as the only independent predictor of cervical recurrence (P = .034 ). Cervical recurrence was significantly related to the subsequent occurren ce of distant relapse. Conclusions: The low incidence of cervical recurrence after RLND in patient s with node-positive head and neck melanomas does not justify the routine u se of PR. The only subset of patients who may benefit from PR are those wit h extranodal disease.