Suciu, Stefan

8526 Background: Many randomised trials have evaluated the role of adjuvant interferon-a (IFN) in high-risk melanoma, some suggesting benefit and others not. To assess the totality of current evidence, an individual patient data (IPD) meta-analysis of all available trials was performed. Methods: Standard IPD meta-analysis methods were used to assess event-free (EFS) and overall survival (OS), with odds ratios (OR) and 95% confidence intervals (CI) calculated. Trials were divided by dose of IFN - high (20 MU/m2), intermediate (5–10 MU), low (3 MU) and very low (1 MU). Subgroup analyses by patient age, gender and disease characteristics were also performed. Results: IPD was provided for 10 of 13 reported trials of IFN vs. no IFN (for the other 3 trials published data were used). 6067 patients (IPD available for 85%) were included in the analysis, with over 3,700 and 3,000 events for EFS and OS. There was statistically significant benefit for IFN for both EFS (OR=0.87, CI=0.81–0.93, p=0.00006) and OS (0.9, 0.84–0.97, p=0.008). There was no evidence of differences according to dose (Table 1; trend p>0.1) or duration of IFN. This proportional survival advantage translates into an absolute benefit of about 3% (CI 1%-5%) at 5 years. The effect of IFN did not differ with age, gender, tumor site, Breslow thickness, clinical nodes or disease stage. Only for ulceration was there some evidence of an interaction (p=0.03); patients with ulcerated tumors had greater benefit from IFN (EFS: OR=0.76, OS: OR=0.77) than those with no ulceration (EFS: OR=0.94, OS: OR=0.98). Conclusions: This meta-analysis provides evidence that adjuvant IFN significantly reduces the risk of relapse and improves overall survival, although the absolute survival benefit is relatively small. This analysis does not however, clarify the optimal (high, intermediate or low) dose of IFN. Given the large number of subgroup analyses performed, the apparent increased benefit in patients with ulceration requires confirmation. No significant financial relationships to disclose.