Long-term Results of Hypofractionated Radiotherapy for Malignant Melanoma of the Head and Neck

Abstract:Purpose/Objective: To report long-term local control and patterns of failure after high-dose per fraction radiotherapy (RT) given on days 0, 7, and 21 for patients with malignant melanoma of the head and neck (HN). Materials/Methods: Between 1990 and 2001, 71 patients with melanoma originating in and confined to the HN received three fractions of 700-800cGy each on days 0, 7, and 21. Median age at start of RT was 67 years (range 18-88 years); the male-to-female ratio was 2.3:1. Thirty-one patients (44%) had cutaneous melanoma not otherwise specified; 15 patients (21%) had desmoplastic/spindle cell histology; and 13 (18%) had mucosal melanoma. Sixty-two per cent (n=44) were treated for recurrent disease, and 38% (n=27) received RT at the time of initial disease presentation. Sixty-six patients (93%) had undergone complete resection of their disease prior to RT; the remaining five had gross disease at the time of RT. Of 49 patients who had surgical resection of their primary or in-transit sites of disease, the resection margin status was positive in 20 (41%); close (≤5mm) in 15 (31%) and negative in 14 (28%). Forty patients (56%) underwent neck dissection as part (or all) of their surgery. The median time from surgery to initiation of RT was 6.4 weeks (range 3-16 weeks). The treatment portals included the primary site (n=24), in-transit site (n=2), neck (n=13), or a combination thereof (n=32). Fraction size was 800cGy in 47 patients (66%) and 700cGy in 24 (34%). In general, the lower fraction size was employed when the optic chiasm or brainstem were in the field. Median follow-up for surviving patients after completion of RT was 41 months (range 3-153 months). Results: Seventeen patients (24%) failed in the treatment portal after completion of RT; 12 of the 17 failed within the first year after treatment. The 5-year actuarial control within the RT portal was 76%. Five-year portal control for patients with grossly resected disease was 83%, versus 0% for patients with gross disease within the RT portal (p<0.001). Median survival was 61 months for patients who had undergone surgical resection prior to RT, versus 8.5 months for patients with gross disease at time of RT (p<0.001). The 5-year portal control rates for patients with non-mucosal melanomas versus mucosal melanomas were 79% and 50%, respectively (p=0.03). The 5-year neck control and distant metastases-free survival rates were 69% and 51%, respectively. Of 42 patients who developed disease recurrence, the most common site of first failure was distant (41%). Overall survival at two and five years was 62% and 51%, respectively. Neither age, gender, pathologically positive nodes, resection of 3 or more positive nodes prior to RT, extracapsular nodal extension, resection margin status, perineural invasion, nor primary versus recurrent disease presentation had any impact upon portal control, disease-free survival or overall survival. There were no grade 3 or 4 acute complications; late grade 3 or 4 toxicity was observed in 2 patients. Conclusions: Excellent long-term control within the irradiated portal can be achieved utilizing hypofractionated radiotherapy for patients with high risk of microscopic residual disease after surgery for non-mucosal malignant melanoma of the head and neck. This treatment regimen was ineffective for patients with gross residual disease. Patients with mucosal melanomas experienced significantly inferior control within the irradiated portal compared to those with non-mucosal melanomas. Local tumor control remains an important objective given the prolonged survival observed in a significant number of patients for whom the cosmetic and functional effects of uncontrolled local disease in the head and neck would be devastating.

C17-372

Long-term Results of Hypofractionated Radiotherapy for Malignant Melanoma of the Head and Neck