Can RECIST response predict success in phase 3 trials in men with metastatic castration-resistant prostate cancer?

Intermediate endpoints are needed in early phase studies of men with metastatic castration-resistant prostate cancer (mCRPC) that can reliably predict success in phase 3 trials. Among men with measurable disease, objective response may provide information as to whether a treatment is likely to be successful.

We conducted a systematic review of systemic agents that have proceeded to phase 3 trials in men with mCRPC and examined the relationship between improvements in measurable disease response in phase 2 trials and successful phase 3 trials leading to regulatory approval. Only trials that included men with radiographically measurable disease were included.

We examined 31 eligible mCRPC phase 3 trials between 1992 and 2017 and 29 of the preceding phase 2 trials for RECIST responses. Measurable tumor responses in phase 2 trials were higher for successful therapies in phase 3 trials in chemotherapy-naive men with mCRPC, but were less correlated with success in trials investigating docetaxel combination regimens or the post chemotherapy mCRPC setting. Many failed agents did not produce higher than expected response rates over control arms; however, several agents such as anti-angiogenic therapies or orteronel produced higher than expected responses without survival benefit.

Objective responses in men with mCRPC may be associated with prolonged survival, but this association is mechanism dependent and inconsistent across trials or disease states. These data support considering RECIST response as a supportive but not sole endpoint in phase 2 trials to support launching phase 3 trials.