Melanoma does not respond meaningfully to systemic chemotherapy. No significant improvement in overall survival has been observed with any therapy in the setting of advanced disease, nor in the adjuvant setting. The most active single drugs achieve 14% to 22% response rates in larger phase II series, and drug combinations have not in general improved true response rates by 15% or more. Experience with recombinant interferon alpha-2 (rIFN alpha 2) administered by a variety of schedules and routes has demonstrated tumor response rates of 12% to 22% in advanced melanoma. Administered combined with chemotherapy, rIFN alpha 2 has improved response rates in some, but not all trials to date. On the basis of increased responsiveness noted to be inversely related to tumor size, IFN alpha has been explored over the past 7 years as an adjuvant to definitive surgery for melanoma. The North Central Cancer Treatment Group (NCCTG) and Eastern Cooperative Oncology Group (ECOG) have recently completed trials of 3 and 12 months' duration of maximal dosages of rIFN alpha 2 in subjects with high-risk resected deep primary or lymph node metastatic melanoma. These randomized controlled studies completed accrual in 1990. The ECOG trial EST 1684 has tested the effect of rIFN alpha 2b, given intravenously daily for 5 days per week for 4 weeks at 20 x 10(6) IU/m2/d, then subcutaneously three times a week at 10 x 10(6) IU/m2/d for 11 months. A series of three analyses is planned for the ECOG trial, and the first of these interim analyses was reported to the ECOG in June, 1990, revealing an encouraging divergence of survival curves that does not achieve statistical significance at this early time. A randomized controlled study of 3 months' duration of rIFN alpha 2a at 12 x 10(6) IU/m2/d intramuscularly three times a week performed by the NCCTG is also in too early a stage to allow definitive conclusions yet. The toxicity of IFN alpha administered at maximum tolerated doses has been significant in terms of constitutional symptoms and organ dysfunction. There have been two instances of fatal toxicity observed in the ECOG study (286 subjects). Consideration of lower, more tolerable dosage regimens has been derived from the desire to evaluate longer periods of IFN alpha 2 therapy, and long-term maintenance treatment with IFN alpha 2 in adjuvant melanoma therapy. In addition, immunologic studies suggest that dosages of 3 x 10(6) IU/m2/d may be more immunologically active in vivo.(ABSTRACT TRUNCATED AT 400 WORDS)