A risk scoring system for African-American (AA) patients (pts) with metastatic castration-resistant prostate cancer (mCRPC) treated with first-line abiraterone (ABA) or enzalutamide (ENZ).

Abstract

51 Background: AA pts represented < 3% of the COU-AA-302 and PREVAIL trial cohorts that led to the approval of ABA and ENZ in the 1st line treatment of mCRPC. We characterized the clinical outcomes (CO) and developed a risk score for AA pts with mCRPC on these agents. Methods: We retrospectively reviewed 77 AA pts with mCRPC treated with 1st line ABA or ENZ at Grady Memorial Hospital from 2015-2018. The CO included median overall survival (mOS), progression-free survival (mPFS) and PSA response (PSAr) as defined by a ≥ 50% drop in PSA over the 1st 12 weeks of treatment. Cox proportional hazard model and Kaplan-Meier method were used for association with OS and PFS and logistic regression model for PSAr. The risk score was built by regression coefficient-based scoring system using OS as the 1° outcome. Covariates included grade group (GG), baseline (bl) PSA, albumin, and BMI, ECOG status, and age. Results: Median age was 60 years with median follow-up of 11.5 months (mos). 50 pts received ABA; 27 received ENZ. The overall mOS was 45.7 mos, mPFS was 12.9 mos, and PSAr was 84.4%. CO did not differ significantly for ABA vs. ENZ with a 24-month OS of 66.6% vs. 57.7% and PFS of 34.3% vs. 45%. 1 point was assigned for each of the following: GG > 3, bl PSA ≥ 108, or bl albumin ≤ 4.2. The total was classified into low (0-1), intermediate (2), and high (3) risk and associated with CO via univariate (UVA) and multivariate (MVA) analyses (Table). Conclusions: We present the efficacy of ABA and ENZ in a cohort of AA pts with mCRPC. Risk grouping using bl PSA, bl albumin and GG may predict CO in this population. These results should be validated in a larger, prospective study.[Table: see text]