Prospective evaluation of supportive care with or without CVD chemotherapy as a second-line treatment in advanced melanoma by patient's choice: a multicentre Dermatologic Cooperative Oncology Group trial.

This prospective, nonrandomized multicentre, phase III study compared best supportive care (BSC) alone with cisplatin, vindesine and dacabazine-based (CVD) chemotherapy and BSC in patients with advanced melanoma. A total of 117 pretreated patients with metastatic melanoma were evaluated, 34 patients in arm A (BSC) and 83 in arm B (BSC and CVD). Primary endpoint was overall survival and secondary endpoints were disease control rate and quality of life (European Organisation for Research and Treatment of Cancer QLQ-C30). Owing to sparse recruitment of patients for randomization, the protocol has been changed based on patients' choice. Baseline characteristics were imbalanced with respect to the Karnofsky Performance Index (P=0.001), the existence of brain metastases (P=0.035) and earlier application of chemoimmunotherapy (P=0.038). Disease control was observed in 8.8% of patients in arm A and in 28.9% of patients in arm B (P=0.028). Median overall survival time was 137 days in arm A and 229 days in arm B (P=0.014). Multivariate analyses could not ascribe this prognostic benefit to CVD treatment. No significant difference in the quality of life could be found. This study could not detect clear survival benefits for polychemotherapy with CVD compared with BSC alone in patients with advanced metastatic melanoma. Interestingly, having the choice of chemotherapy or BSC alone in a second-line situation, more than 70% of patients chose polychemotherapy.