Hospital admissions (HA) among chemotherapy-naïve metastatic castration-resistant prostate cancer (mCRPC) patients newly treated with enzalutamide (ENZA) or abiraterone (ABI).

Abstract

e16524 Background: ENZA and ABI are approved to treat mCRPC in chemotherapy-naïve patients; few real-world analyses exist to compare HA pre- and post-treatment. This study characterized and compared incidence and incidence rate ratios (IRRs) of HA pre/post initiation of ENZA or ABI in chemotherapy-naïve mCRPC patients in the US. Methods: Patients aged ≥ 18 years who initiated ENZA or ABI (2 cohorts) from Jan 2014–Dec 2015 were identified retrospectively from MarketScan® Databases. The index date was the initiation of ENZA or ABI after ≥ 6 months of pre-index continuous health plan enrollment. We identified the top 10 HA incidences using ICD-9/10 CM inpatient claims (separate primary and any position) and assessed pre-/post-index IRRs within each cohort. Generalized estimating equations accounting for within-patient correlation and difference-in-differences regression were fitted to compare IRRs between cohorts. Results: A total of 1603 (ENZA, 656; ABI, 947) patients were included. Mean age was 73 years in both cohorts; mean (standard deviation) Deyo-Charlson Comorbidity Index was 7.9 (2.8) for ENZA and 7.8 (2.9) for ABI; the top 10 HA incidences included cardiac disorder, dyspnea, hypokalemia, liver toxicity, pneumonia, renal impairment, skeletal-related events (any, surgery and fracture) and urinary tract infection. IRRs of HA (any position) due to cardiac disorder, dyspnea, hypokalemia, liver toxicity and pneumonia were significantly higher for post- vs pre-ABI-treated patients (IRR [95% CI]: 1.63 [1.22–2.17], 1.90 [1.21–2.98], 3.13 [1.46–6.71], 2.24 [1.36–3.69] and 2.40 [1.41–4.10]). IRRs of HA (any position) due to cardiac disorder, dyspnea and hypokalemia were significantly lower in the ENZA cohort than ABI cohort (1.02 vs 1.63, 0.84 vs 1.90, 0.65 vs 3.13; all P < 0.05). No statistical significance was observed for any other HA comparisons (primary or any position) between or within ENZA and ABI cohorts. Conclusions: Among chemotherapy-naïve mCRPC patients, significant increases in multiple HA IRRs were observed post- vs pre-ABI treatment, and IRRs were significantly lower for ENZA- vs ABI-treated patients for multiple HA.