Does adjuvant vaccine therapy really have activity in malignant melanoma?

interferon (IFN) versus low-dose IFN plus allogeneic melanoma lysate vaccine as representing “the first time a randomized large-scale trial of a vaccine regimen has demonstrate d... activity” in the adjuvant melanoma setting, and suggest that “low-dose IFN was probably not solely responsible” for any activity observed. However, since publication, an individual patient data meta-analysis presented recently at the 2007 Annual Meeting of the American Society of Clinical Oncology (ASCO; Chicago, IL) 2 provides no evidence that low-dose IFN is any less effective than high-dose. The hazard ratio (HR) for high-dose IFN is 0.90 (95% CI, 0.80 to 1.02;P .1), and the HR for low-dose IFN is also 0.90 (95% CI, 0.80 to 1.01; P .07). Nor is this finding new. While Mitchell et al quote two qualitative systematic reviews in their discussion, they fail to mention a quantitative published data meta-analysis from 2003, 3 which also showed no dose effect on overall survival (high-dose HR 0.90, low-dose HR 0.90). Thus, it is entirely plausible that the “indistinguishable” survival of patients in the highdose IFN and low-dose IFN plus vaccine arms of their study is due to the similar efficacy of low-dose and high-dose IFN, with the vaccine making no contribution. Furthermore, at the 2007 Annual Meeting of ASCO, a trial of a vaccine in stage III melanoma patients reported adverse survival in the vaccine arm (P .04). 4 We therefore consider that caution should be exercised in interpreting the trial by Mitchell and colleagues as providing any evidence of benefit for adjuvant vaccine therapy in melanoma. Also, as the authors point out, the trial was not large enough to demonstrate equivalence, but no idea (ie, HR with CI) of the range of uncertainty of the treatment effect is presented. We calculate them to