Cabazitaxel (microtubule-inhibitor taxane)–based “cabascore” stratifies mCRPC survival and predicts benefit vs 2nd ARPI

CABASTY post hoc (n=194) derived a 6-variable Cox-model “cabascore” (ECOG PS 0 vs 1–2, <2 vs ≥2 metastatic sites, PSA, LDH, ALP, hemoglobin) and split patients on cabazitaxel into favorable/intermediate/poor groups with median OS 21.6 vs 12.5 vs 6.2 months (intermediate vs favorable HR 2.14, p<0.001; poor vs favorable HR 5.41, p<0.001). External validation in PROSELICA cabazitaxel-treated patients (n=1175) and CARD cabazitaxel-exposed patients (n=126) preserved the OS separation (both p<0.001). Importantly, in the full CARD population (n=250) cabazitaxel improved OS vs a second ARPI only in the cabascore-favorable subgroup (p=0.003)—a notable nuance given CARD’s overall OS benefit for cabazitaxel vs abiraterone/enzalutamide (median 13.6 vs 11.0 months; HR 0.64).