Prostate-specific antigen response in black and white patients treated with abiraterone acetate for metastatic castrate-resistant prostate cancer.

PURPOSE: Evidence suggests differences in androgen receptor AR signaling between black (B) and white (W) patients with prostate cancer, but pivotal trials of abiraterone acetate (AA) for patients with metastatic castration-resistant prostate cancer (mCRPC) enrolled few black patients, a population with a higher mortality from prostate cancer. Our primary objective was to determine differences in response to AA between B and W patients. METHODS: We performed a retrospective case-control study of B vs. W patients treated with AA between May 1, 2008 and June 16, 2015 at Duke University Medical Center. Patients were identified (W control patients were matched 2:1 to B patients stratified based on previous docetaxel exposure) through pharmacy records and were eligible if treated with AA for metastatic castration-resistant prostate cancer. Patients with previous enzalutamide use were excluded. The primary objective was to compare the rate of≥90% prostate-specific antigen (PSA) decline from baseline between B vs. W patients. Secondary outcomes included comparing time on therapy, time to PSA progression, and overall survival among groups. RESULTS: Baseline characteristics among patients (n = 45 B, n = 90 W) were identified; these included Karnofsky performance status, PSA, Gleason score, alkaline phosphatase, albumin, hemoglobin, lactate dehydrogenase, opiate use for pain, and metastatic sites. Baseline characteristics among groups were similar except for median hemoglobin (B = 11.4g/dl, W = 12.3g/dl). The proportion of B patients achieving a≥90% PSA level decline was 37.8% vs. 28.9% for W patients (P = 0.296). Statistically significant differences were found in the proportion of patients achieving a≥50% PSA level decline (B = 68.9%, W = 48.9% [P = 0.028]) and≥30% PSA level decline (B = 77.8%, W = 54.4% [P = 0.008]). Rates of primary abiraterone-refractory disease (PSA increase as best response) trended higher in W (31.1%) than in B (15.6%) patients (P = 0.052). Median treatment duration (B = 9.4 mo, W = 8.3 mo) did not differ (Wilcoxon P = 0.444). Median overall survival (B = 27.3 mo [95% CI: 13.9, not estimable], W = 24.8 mo [95% CI: 19, 31.6] [P = 0.669]) and median time to PSA progression (B = 11.0 mo [95% CI: 4.3, 18.0], W = 9.4 mo [95% CI: 6.2, 13.0] [P = 0.917]) did not differ. CONCLUSIONS: Black patients may have a higher PSA response to AA than white patients. An ongoing prospective clinical study (NCT01940276) is evaluating outcomes between black and white patients treated with AA.

Duke Scholars

Author Yuan Wu Biostatistics & Bioinformatics, Division of Biostatistics

Author Michael Roger Harrison Medicine, Medical Oncology

Author Andrew John Armstrong Medicine, Medical Oncology

Author Daniel James George Medicine, Medical Oncology

Author Tian Zhang Medicine, Medical Oncology