Sir—Natale Cascinelli and colleagues (Sept 15, p 866) 1 report that low-dose interferon is inactive adjuvant therapy for high-risk melanoma, which contradicts their earlier results in WHO trial 16. 2 The report unfortunately completely omits mention of a large significantly positive study of interferon, intergroup trial E1694, 3 which was unblinded by the external safety committee because of survival benefits judged as unethical to ignore. High-dose interferon was first reported to significantly improve survival of high-risk melanoma in the Eastern Cooperative Group trial E1684, published at 6· 9 years’ median followup. 4 Updated results at 12· 6 years5 showed a continued survival advantage for high-dose interferon, with 13· 8 months median survival gained. The sustained survival benefit of high-dose interferon in that trial stands in striking contrast to low-dosage regimens. All relapses reported in the E1684 study were distant, because of entry lymphadenectomy requirements. The significant distant disease-free survival for E1684 (p= 0· 005 at 12· 6 years) suggests a curative effect. In E1694, 3 high-dose interferon had significant survival benefits over the vaccine GMK, which was selected as the most promising vaccine for trial in melanoma by the US intergroup. The relapse-free and overall survival analyses of E1694 show significantly increased hazard ratios for relapse and death among patients not given high-dose interferon (1· 49 [p= 0· 0004] and 1· 38 [p= 0· 023], respectively). This benefit reflects a 33% reduction in the likelihood of relapse and a 28%