Hospitalizations and common infections among veterans treated for metastatic prostate cancer.

Abstract

92 Background: Abiraterone (AA) and Enzalutamide (ENZ) are both hormone therapies used in the treatment of metastatic castrate resistant prostate cancer (mCRPC). Due to a lack of large comparative studies, they are currently used interchangeably, but may have different adverse outcomes in patients with comorbid conditions. In this study, we aim to identify adverse events that lead to hospitalization in patients treated with ENZ versus AA. Methods: Patients treated with AA or ENZ between September 10, 2014 and June 3, 2017 were identified in the Veterans Health Administration and followed until April 2020. We obtained ICD 9/10 codes using the Veterans Health Administration Informatics and Computing Infrastructure (VINCI) platform among VA hospitalizations. We used the top 3 ICD 9/10 codes at discharge to ascertain causation of hospitalization. Infections were defined by ICD 9 and ICD 10 codes falling into 3 categories: pneumonia (ICD9: 480.x-486.x, ICD10: J13.x-J18.x), urinary tract infection (ICD9: 599.0, ICD10: N39.0), and sepsis (ICD9: 995.91, 995.92, ICD10: A40.x-A41.x, R65.20, R65.21). Results: 5,822 patients were identified for the cohort. The mean age of the patients was 75.3 years old, with a mean Charlson comorbidity index of 4.2. Patients first treated with enzalutamide were older (75.8 vs 75.0 years, p = 0.002) with a higher mean Charlson comorbidity score (4.4 vs 4.1, p < 0.001). There were no significant differences in time from initial diagnosis of PCa to treatment with AA or ENZ. Of the 5,822 patients, 2504 (43.0%) were initially treated with ENZ, and 3,318 (57.0%) with AA. Total hospitalization rate in events/person-years was 1.52 and 1.29 for the abiraterone and enzalutamide cohorts respectively, with an incidence rate difference (IDR) of 0.23 (p < 0.0001, CI: 0.14-0.32), indicating a statistically significant increase in hospitalizations among the abiraterone cohort. Of the total hospitalizations, 1/4th of them were caused by infections defined as pneumonia, sepsis, or UTI. The incidence in events/person-years for the combined infections in the AA and ENZ cohort was 0.99 and 0.86 (IDR = 0.13, p = 0.06) respectively, 0.97 and 0.86 (IDR = 0.11, p = 0.26) for UTI; 0.76 and 0.74 (IDR = 0.021, p = 0.82) for sepsis; and 0.82 and 0.71 (IDR = 0.12, p = 0.22) for pneumonia. This shows no statistical difference in infection rate between the two treatments. Conclusions: The increase in incidence rate of hospitalizations of 0.23 events/person-years among the abiraterone group, despite greater age and comorbidities in the enzalutamide group, suggests an increased risk of adverse events requiring hospitalization in patients started on abiraterone initially. However, no difference was seen in incidence of the three most common infections between cohorts. This suggests that the difference in mechanism of action and use of prednisone with AA may not be of clinical significance with regards to risk for infection.