Predictors of overall survival among Black South African men treated with androgen-deprivation therapy for metastatic prostate cancer.

Abstract

5046 Background: Men in sub-Saharan Africa (SSA) are disproportionately affected by prostate cancer (PCa), and many have metastatic disease (mPCA) at presentation. In SSA, androgen deprivation therapy (ADT) is the first-line treatment for mPCa, and often the only available therapy. Treatment failure and death is common. We identified predictors of overall survival (OS) in Black South African (SA) men with mPCa on ADT. Methods: We performed a retrospective analysis of prospectively gathered data from men diagnosed with mPCA (3/22/2016 - 10/30/2020) at Chris Hani Baragwanath Hospital in Johannesburg, which was also a study site for the concurrent Men of African Descent and Carcinoma of the Prostate study. We included men with mPCA treated with ADT (received at least 1 dose of luteinizing hormone-releasing hormone agonist and/or had surgical castration), who had ≥1 PSA level drawn ≥12 weeks after ADT start. OS was defined from ADT start to death. PSA progression (PSA-P) definition was adapted from PCWG 3. Cox regression models were used to identify predictors of OS. PSA-P was treated as a time-dependent covariate. Results: Of 200 men with mPCa, we excluded 6 who did not receive ADT and 41 without sufficient data for PSA-P analysis. Of 153 men, 26.8% were <65 years old and 12% had a family history of PCa. Median PSA at diagnosis was 71.5 ng/mL (interquartile range (IQR) 20.7-432.6), median alkaline phosphatase level (ALP) 108 IU/L (79-224) and median hemoglobin (Hb) 13 g/dL (IQR 10-15). Median PSA nadir was 2.8 ng/mL (IQR 0.55-17.93). The rate of PSA-P at 1- and 2-years was 12.1% [95%CI 5.9-17.8] and 37.5% [95%CI 26.1-47.2]. The median follow-up was 2.75 years, and the 3-year OS was 61.9% [95%CI 52.7-72.6]. Cox proportional hazard ratio (HR) models of risk factors for OS are shown in Table 1. PSA-P was a strong predictor of OS. Men with PSA nadir >4ng/mL after ADT start had a HR for death of 3.77 [1.86-7.62]. Men with ALP >150 IU/L and those with Hb <13.5g/dL at diagnosis were also at higher risk for death (HR 3.09 [1.64-5.83] and HR 2.00 [1.28-6.56] respectively). Conclusions: Among Black men in SA treated with ADT for mPCA, PSA-P strongly predicts OS. In this cohort, high ALP and anemia at diagnosis, and PSA nadir >4ng/mL after ADT start are associated with higher risk for death. These factors can be used identify high risk men with mPCA, for whom early treatment escalation to chemotherapy should be considered. [Table: see text]