Patients with melanoma have experienced durable therapeutic benefits with immunotherapy, and high-dose (HD) interferon alfa-2b (IFN2b) for 1 year and pegylated (PEG) IFN2b for 5 years have received regulatory approval as adjuvant therapy. IFNhas been administered in low-, intermediate-, and high-dose regimens for patients with intermediateand high-risk melanoma (corresponding to the American Joint Committee stages II, III, or IV). A durable impact on both relapse-free (RFS) and overall survival (OS) has only been demonstrated with HD IFN2b as tested in Eastern Cooperative Oncology Group (ECOG) and Intergroup trials E1684, E1690, and E1694. All three trials demonstrated significant reduction in relapse rate, whereas significant improvement in overall survival (OS) was observed in two of these trials. Specifically, in E1684, median relapse-free survival (RFS) was 1.72 versus 0.98 years (P .0023; hazard ratio [HR], 0.61; P .0013). Median OS was 3.82 versus 2.78 years, P .0237 (HR 0.67, P .01) in favor of IFN treatment. Likewise, in E1694, improvement in RFS with IFN (HR, 0.68; P .0015) and OS (HR, 0.66; P .009) were observed, both in the eligible population and in the intent-to-treat analysis (ITT) for RFS (HR, 0.67) and OS (HR, 0.72). The one trial that did not show OS benefit was conducted partly before and partly after the US Food and Drug Administration approval of HD IFN2b and did not require removal of regional lymph nodes, unlike the pivotal E1684. Thus, patients assigned to observation in E1690 frequently experienced lymph node relapse and uniformly crossed over to receive HD IFN2b at relapse in regional lymph nodes. Retrospective analysis has shown that those who crossed over to HD IFN2b at nodal relapse (n 38) also demonstrated a large benefit, which seemed to confound the outcome such that an RFS benefit was seen but no OS benefit was. A pooled analysis of the two observation-controlled trials (E1684 and E1690) as updated through April 2001 showed that HD IFN2b maintained significant benefits in relapse prevention out to intervals of 20 years. However, this analysis excluded the E1694 trial in which GM2 conjugated to keyhole limpet hemocyanin and administered with QS-21 (GMK) vaccine served as control. Given that the larger of the two observation-controlled trials (E1690) did not show an OS benefit for HD IFN2b, the pooled analysis did not show compelling evidence of OS benefit. In addition, the median 12.6-year follow-up in E1684 introduces the strong possibility that competing causes of death led to the erosion of the OS benefits originally seen in this trial at the mature median follow-up of 6.9 years. Three large meta-analyses examined the survival benefits of adjuvant IFN, pooling data from randomized controlled trials testing IFN. The first meta-analysis of 12 randomized controlled trials estimated significant RFS benefit for IFN compared with observation (HR, 0.83; 95% CI, 0.77 to 0.90; P .001) and showed OS benefits that were less significant (HR, 0.93; 95% CI, 0.85 to 1.02; P .1) suggesting increased benefit with increasing dosage. The second meta-analysis (using individual cooperative group patient data from 13 randomized controlled trials) demonstrated significant benefits for IFNin lowering the risk of recurrence (odds ratio [OR], 0.87; 95% CI, 0.81 to 0.93; P .001) and of death (OR 0.90; 95% CI, 0.84 to 0.97; P .008) versus observation or vaccination. The latest meta-analysis included 14 randomized controlled trials and estimated significant reductions in the risk of recurrence (HR, 0.82; 95% CI, 0.77 to 0.87; P .001) and death (HR, 0.89; 95% CI, 0.83 to 0.96; P .002). In the article that accompanies this editorial, Eggermont et al present mature results of European Organisation for Research and Treatment of Cancer (EORTC) 18991. This randomized phase III trial was conducted in patients with resected stage III melanoma who were assigned to observation or PEG IFN2b treatment for 5 years, testing the hypothesis that long-term therapy with PEG IFN2b is necessary to optimize the effects of IFNon the basis of previous EORTC and French studies in which the authors concluded that RFS benefit is lost after discontinuation of treatment. These investigators also hypothesized that weekly self-administration of PEG IFN2b might improve the benefit–toxicity ratio to allow this prolonged therapy. Distant metastasis-free survival (DMFS) was originally the primary end point, but RFS was later adopted as the primary regulatory end point. Mirroring E1684, PEG IFN2b was administered in two phases: induction for 8 weeks followed by maintenance for up to 5 years. The trial was first reported in 2007 (at 3.8 years median follow-up), demonstrating significant improvement in RFS in favor of PEG IFN2b versus observation (HR, 0.82; P .011); no significant benefit to either DMFS or OS were observed. Subgroup analyses showed a lack of benefit in patients with gross nodal disease (N2) across all end points: RFS, DMFS, and OS. PEG IFN2b significantly improved RFS and DMFS (but not OS) in patients with microscopic nodal involvement (N1). JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 30 NUMBER 31 NOVEMBER 1 2012