Similar efficacy for phase I trials in comparison with DTIC for advanced malignant melanoma: an analysis of melanoma outcomes in CTEP-sponsored phase I trials 1995-2011.

After ipilimumab, vemurafenib, and interleukin-2, standard of care chemotherapy for melanoma remains dacarbazine (response rate ∼9%). Despite this, many physicians hesitate to refer patients to phase I protocols given a perceived lack of clinical benefit and potential for harm. To better understand the validity of these perceptions, the experience of all patients with melanoma treated on phase I trials sponsored by the National Cancer Institute-Cancer Therapy Evaluation Program (NCI-CTEP) from 1995 to 2011 were analyzed and compared with the pooled results of six contemporary phase III trials of dacarbazine. A total of 937 patients with melanoma were treated in 148 CTEP phase I trials. The majority were men with a median of two prior therapies (46% receiving prior dacarbazine). Response and clinical benefit rates in these trials were not clinically different from those of dacarbazine (phase I: 6.3 and 26.8% vs. dacarbazine: 8.8 and 27.9%) although grades 3 and 4 toxicity was significantly higher (54 vs. 28%). Efficacy and toxicity were generally consistent within phase I subgroups (targeted agents, immunotherapies, or chemotherapeutics) though targeted therapy was associated with a lower response rate, immunotherapy with lower clinical benefit rate, and chemotherapy with higher incidence of grade 4 toxicity. Thus, the perception of limited efficacy of phase I trials for patients with melanoma was disproven, whereas the perception of toxicity was observed. However, this difference in toxicity may have been largely because of the nature of phase I vs. phase III trials (i.e. more heavily pretreated) and because of the phase I trials often being multiagent as opposed to dacarbazine alone.