Frailty and overall survival in metastatic castration-resistant prostate cancer treated with abiraterone or enzalutamide.

Abstract

71 Background: Abiraterone (ABI) and Enzalutamide (ENZ) are both treatment options for metastatic castration-resistant prostate cancer (mCRPC). However, both drugs have differing pharmacodynamics, so patient factors may selectively impact drug efficacy. Age-associated health deficits make older adults variably susceptible to disease under a syndrome termed frailty. The average age of initial prostate cancer diagnosis is age 66, and the average time to mCRPC progression is 11.7 years. Thus, frailty is an important patient factor to consider in mCRPC, and it may play a role in personalizing drug selection. Methods: Patients treated with ABI or ENZ for mCRPC from September 2014 to June 2017 were identified in the VA and followed until April 2020. Frailty was assessed using the VA frailty index, which looks at ICD-10, CPT, and HCPCS codes to assign a standardized frailty score between 0-1. Patients were categorized as frail if their VA-FI scores were > 0.2. The primary outcome was overall survival (OS) between the ABI and ENZ treatment groups. Cox regression models included age, race, body mass index, PSA, and total Charlson Comorbidity Index (CCI). For frail patients, a propensity score-matched cohort was also created by matching age, race, CCI, and total VA-FI between ABI and ENZ treatment groups. Results: Of the 5,822 patients identified, 57% were initially treated with ABI and 43% with ENZ. The ABI group had a lower mean VA-FI score of 0.192 compared to the ENZ group of 0.203 (p=0.004). Frail patients were older: mean 74.9 years vs. 76.1 years for non-frail (p<0.001). Frail patients had shorter survival than non-frail (18.5 vs. 26.6 months, p<0.001). In multivariable models, frailty was associated with increased risk of death (aHR 1.32, 95% CI 1.23-1.42). Among non-frail patients (n=3508, 60.3%), there was no significant difference in OS between ABI and ENZ treatment. Among frail patients (n=2314, 39.7%), patients treated with ENZ had a 3.5 month greater median OS than ABI (20.7 vs. 17.2 months, p = 0.0003 by log-rank). In frail patients, ENZ was associated with reduced risk of death (aHR 0.85, 95% CI 0.78-0.94). In a propensity matched cohort of 2070 patients, the enzalutamide treatment group had a 3.2 months greater median OS in the enzalutamide group than the abiraterone group (24.1 months vs. 20.9 months, p < 0.001). Conclusions: ENZ was associated with improved OS compared to ABI among frail veterans, but not among non-frail veterans. The improved survival remained significant in a propensity score-matched cohort and models including important covariates. Further assessment of the value of frailty in personalizing drug treatment is warranted to improve the care of older adults with advanced prostate cancer.