Vemurafenib plus cobimetinib (VC) for the treatment of metastatic melanoma was requested to be included in the National Formulary in Uruguay. The standard of care for metastatic melanoma in Uruguay is dacarbazine. There is no published head-to-head trial assessing the effects of VC versus dacarbazine. The objective of this study was to perform an indirect comparison of the effects of dacarbazine, compared with VC, based on the results of trials that included both treatments versus the same comparator (vemurafenib alone).

We searched Pubmed and The Cochrane Library for trials comparing either VC or dacarbazine with vemurafenib. Trials were assessed in terms of risk of bias, similarity of interventions and inclusion and exclusion criteria, and comparability of characteristics of patients in the vemurafenib arm. We performed an indirect comparison using the Bucher method.

From the literature search we retrieved two studies that met the inclusion criteria: a randomized clinical trial that assessed VC versus vemurafenib or placebo and another assessing dacarbazine versus vemurafenib. Both studies were similar in terms of methodological quality, inclusion and exclusion criteria, and comparability of the vemurafenib arms. However, the comparison of overall survival and progression-free survival curves for the vemurafenib arms were quite different between the two trials. At 9 months, overall survival was eighty-one percent and fifty-five percent and progression-free survival was thirty percent and fifteen percent, respectively. The indirect comparison provided the following hazard ratios: 0.24 (95% confidence interval [CI]: 0.14–0.48) for overall survival; 0.13 (95% CI: 0.09–0.19) for progression-free survival; and 0.15 (95% CI: 0.02–1.29) for grade 4 adverse events.

Treatment with VC increased overall survival and progression-free survival, compared with dacarbazine. Severe adverse events were less frequent with the combined therapy. However, the differences in the vemurafenib survival curves increases doubts about the accuracy of the indirect estimators of overall survival and progression-free survival.