A 3-variable prognostic score (3-PS) for overall survival prediction in metastatic castration-resistant prostate cancer treated with223Radium-dichloride

Objective: In mCRPC patients treated with 223Ra, a major issue is the validation of reliable prognostic and predictive biomarkers to maximize clinical benefit and minimize toxicities and costs. Bearing in mind how changes in tALP did not meet statistical requirements as surrogate marker for survival, aim of this single-center retrospective study was to characterize the prognostic and predictive role of baseline clinical variables associated with overall survival in patients receiving 223Ra treatment. Methods: 92 consecutive CRPC patients with symptomatic bone metastases receiving 223Ra treatment were included. Available baseline clinical data relevant to the survival analysis were retrospectively collected. The primary end-point of the study was overall survival, which was established from the first 223Ra administration until date of death from any cause. Results: Median follow-up time from the first 223Ra administration was 6 months (range 1–31 months). The univariate analysis evaluating the prognostic value of all baseline clinical variables showed that patients’ weight, BMI, ECOG PS, Hb and tALP values were independently associated with OS. On multivariable analysis only baseline Hb value and ECOG PS remained significantly correlated with OS. To determine reliable baseline predictive factors for survival in patients receiving 223Ra treatment, we produced a predictive score. We tried all possible variable combinations, and found that the best score was obtained by combining baseline ECOG PS with Hb < 12 g/dl and PSA ≥ 20 ng/ml. This resulted in a score ranging from 0 to 4, with AUC 78.4% (p < 0.001). Conclusions: We propose a multidimensional clinical evaluation to select those mCRPC subjects suitable to receive the maximum benefit from 223Ra treatment.