ABSTRACT: Individuals affected by melanoma with thick primary tumours or regional node involvement have a poor outlook, with only 30-50% alive at 5 years. High-dose and low-dose interferon alfa have been assessed for the treatment of these patients, with the former having considerable toxicity and a consistent effect on disease free survival, but not on overall survival, and the latter no consistent effect on either. Our aim was, therefore, to assess the effect of two regimens of interferon of intermediate dose versus observation alone on distant metastasis-free interval (DMFI) and overall survival in such patients.
We did a randomised controlled trial in 1388 patients who had had a thick primary tumour (thickness > or = 4 mm) resected (stage IIb) or regional lymph node metastases dissected (stage III) and had been assigned to 13-months (n=553) or 25 months (n=556) of treatment with subcutaneous interferon alfa 2b, or observation (n=279). Treatment comprised 4 weeks of 10 million units (MU) of interferon alfa (5 days per week) followed by either 10 MU three times a week for 1 year or 5 MU three times a week for 2 years, to a total dose of 1760 MU. Our primary endpoint was DMFI. Analyses were by intent to treat.
After a median follow-up of 4.65 years, we had recorded 760 distant metastases and 681 deaths. At 4.5 years, the 25-month interferon group showed a 7.2% increase in rate of DMFI (hazard ratio 0.83, 97.5% CI 0.66-1.03) and a 5.4% improvement in overall survival. The 13-month interferon group showed a 3.2% increase in rate of DMFI at 4.5 years (0.93, 0.75-1.16) and no extension of overall survival. Toxicity was acceptable, with 18% (195 of 1076) of patients going off study because of toxicity or as a result of refusal of treatment because of side-effects.
Interferon alfa as used in the regimens studied does not improve outcome for patients with stage IIb/III melanomas, and cannot be recommended. With respect to efficacy of the drug, duration of treatment seemed more important than dose, and should be assessed in future trials.
The Lancet 10/2005; 366(9492):1189-96. · 38.28 Impact Factor