Total Medical and Pharmacy Costs for Patients With Metastatic Castration- Sensitive Prostate Cancer Initiating Oral Androgen Receptor Pathway Inhibitors in the United States

J Clin Pathways. 2025;11(2):10-21. doi:10.25270/jcp.2025.02.02

Prostate cancer (PC) is the second leading cause of cancer death and the most com­mon cancer type among men in the US, with an estimated 32 350 deaths and 299 010 new cases expected in 2024.¹ Metastatic castration-sensitive PC (mCSPC) accounts for approximately 4% of prostate tumors.² Although 5-year survival rates for local­ized PC remain near 100%, survival among patients with distant disease, including mCSPC, decreases to approximately 37%.³

Androgen deprivation therapy (ADT) via surgical or medi­cal castration has long been the standard of care for mCSPC.^4,5 Initial prostate-specific antigen (PSA)-based response rates to ADT exceed 80%,^5,6 however, nearly all patients with mC­SPC eventually progress to metastatic castration-resistant PC (mCRPC) within 2 to 3 years.^7,8 More recently, androgen re­ceptor pathway inhibitors (ARPIs), used in combination with ADT, have emerged as important additions to the therapeutic landscape for mCSPC.⁹ The ARPIs approved by the US Food and Drug Administration (FDA) include apalutamide (ap­proved on September 17, 2019, for mCSPC),^10-12 enzalutamide (approved on December 16, 2019, for mCSPC),^13-15 abiraterone acetate plus prednisone (approved on February 7, 2018, for high-risk mCSPC),^16,17 and most recently darolutamide plus docetaxel (approved on August 5, 2022, for mCSPC).^18,19 To­gether with ADT, ARPIs improve survival outcomes and help to slow the progression of mCSPC to mCRPC.^9,20

Although real-world evidence suggests apalutamide may offer superior clinical outcomes in patients with mCSPC com­pared with other ARPIs,^21,22 recent health care resource use (HRU) and cost data, which provide important information for assessing these agents’ utility in mCSPC, are lacking. Previous studies have established the high economic burden of mCSPC²³ and increased HRU and costs associated with progression from mCSPC to mCRPC,^24,25 but have not evaluated the costs as­sociated with different ARPIs.

The economic burden of mCSPC may be reduced by con­sidering advanced treatment options and slowing progression from mCSPC to mCRPC. In addition, evaluating annual total cost of care and the relationship of medical costs to nonmedi­cal costs may be one of the means of assessing the therapeutic value of treatment alternatives in the real-world setting, with suboptimal adherence to the prescribed regimen and increas­ing HRU or higher medical costs serving as indicators of lower therapeutic value for specific pharmaceutical agents.^26,27

With the approval of these newer therapeutic agents for the treatment of mCSPC, it is important to study their impact on health care costs, particularly in a Medicare-insured popula­tion, to prepare for future health care needs of older adults in the US. Thus, this study aimed to evaluate the total cost of care and components of HRU and costs in ARPI-naïve Medicare-insured patients with mCSPC initiating their first ARPI.

Data Sources

Clinical data from more than 90 US private community-based urology practices (Precision Point Specialty [PPS] Analytics) were linked with administrative claims (Komodo Research Database [KRD]; January 1, 2016, to September 30, 2022) to select Medicare-insured patients with mCSPC who newly ini­tiated apalutamide, enzalutamide, or abiraterone acetate. Prac­tices included in PPS, which are distributed nationally across the US, represent 4000 active community urology providers, who have treated approximately 25 000 patients with mCSPC. The database includes patient demographics and clinical vari­ables such as laboratory testing results, PC-related medical pro­cedures, and dispensation information for ARPIs.

The KRD is a deidentified database sourced from a variety of payers and health care organizations in the US. The Komodo database contains over 320 million US patients across Medicaid, commercial, and Medicare insurers, and consists of information on insurance coverage, diagnosis and procedures received in in­patient and outpatient settings, along with prescription fills, and mortality data. Only closed claims data were used in HRU and cost analyses. Patients in PPS and KRD were linked by Data­vant, which employed their patent-pending deidentification technology using machine learning validation to substitute pri­vate patient data with an encrypted token that cannot be reverse- engineered to reveal any personal information. This study was considered exempt research under 45 CFR § 46.104(d)(4) as it involved only the secondary use of data that were deidentified in compliance with the Health Insurance Portability and Account­ability Act (HIPAA), specifically, 45 CFR § 164.514.

Study Design

A retrospective longitudinal cohort design was used to evalu­ate HRU and costs. The study design scheme is depicted in Figure 1. The index date was defined as the date of the first dispensation (from PPS) or paid pharmacy claim (from KRD) for apalutamide, enzalutamide, or abiraterone acetate. The start of the index treatment identification period was based on the FDA approval for each treatment for mCSPC (ie, apalutamide: September 17, 2019; enzalutamide: December 16, 2019; abi­raterone acetate: February 7, 2018, for high-risk mCSPC). The baseline period was defined as the 12-month period preceding the index date. The follow-up period spanned from the index date until the earliest of progression to castration resistance (ie, mCRPC), discontinuation of the index ARPI (ie, using a 90-day gap in days of supply), initiation of a different ARPI or a radiopharmaceutical agent, end of continuous insurance eligibility, or end of data availability (September 30, 2022). No minimum follow-up time was imposed to reduce the potential impact of survivor bias. Concurrent ADT use was not required for inclusion in this study. Concurrent prednisone use was not required for inclusion in the abiraterone acetate cohort.

Study Population

Patients with mCSPC who newly initiated an ARPI (ie, apalutamide, enzalutamide, or abiraterone acetate) with Medicare insurance were of interest in this study. The combination of darolutamide and ADT was not in­cluded in the analyses because it had not been approved for use in the US during the time frame of this study. Pa­tients were required to meet the following criteria: (1) ≥ 1 paid pharmacy claim (from KRD) or dispensation (from PPS) for an ARPI (ie, apalutamide, enzalutamide, or abi­raterone acetate) on or after their respective FDA approvals; (2) no prescription for another ARPI observed before the in­dex date; (3) had mCSPC on the index date; (4) ≥ 18 years of age on the index date; (5) ≥ 12 months of continuous closed insurance enrollment prior to the index date; and (6) Medi­care insurance coverage on the index date. The exclusion cri­terion was no use of radiopharmaceutical therapy before or on the index date.

Patients were considered to have mCSPC if they had evi­dence of metastases observed any time before or on the index date, in the absence of castration resistance observed any time before or on the index date. Metastatic disease was defined based on bone, nodal, or visceral metastasis identified through derived variables from PPS or diagnosis codes from PPS and KRD. Castration resistance was assessed using a previously published algorithm.²⁸

Study Measures

Patient demographics and clinical characteristics (ie, Quan- Charlson Comorbidity Index [Quan-CCI], comorbidities) were evaluated during the baseline period (ie, 12 months pre-index) using data from both PPS and KRD, as applicable.

All-cause and PC-related HRU and cost outcomes were evaluated separately during the baseline and follow-up peri­ods based on medical and pharmacy claims from KRD. HRU categories included the number of inpatient admissions, num­ber of inpatient days, number of days with emergency depart­ment visits, number of days with outpatient visits, number of days with pharmacy claims, and number of days with other services. PC-related HRU and costs were identified based on medical claims with diagnosis code for PC (International Clas­sification of Diseases, 10th Revision, Clinical Modification: C61) or claims with a procedure code for ADT or other therapies for metastatic PC (ie, ARPIs, chemotherapies, estrogens, im­munotherapies, poly ADP-ribose polymerase inhibitors, or radiopharmaceuticals).

All-cause and PC-related health care costs, including medi­cal costs (ie, sum of inpatient, emergency department, outpa­tient, other costs), pharmacy costs, and total health care costs (sum of medical and pharmacy costs) were reported. All HRU and cost outcomes were reported per patient per year (PPPY), and costs were reported in 2022 US dollars and adjusted for in­flation using the medical care component of the US Consumer Price Index. Costs were reported from the payer’s perspective.

Statistical Analyses

All-cause and PC-related HRU and health care costs were sum­marized using descriptive statistics. For categorical variables (eg, proportion of patients with an outpatient visit), frequencies and proportions were reported. The means, medians, and SD were reported for count variables (eg, number of outpatient visits) and continuous variables (eg, pharmacy costs). HRU and costs were reported separately for each index ARPI (ie, apalutamide, enzalutamide, and abiraterone acetate). This study was descrip­tive, and no comparisons of HRU and cost outcomes were made between index ARPIs (eg, no P values were generated).

Demographic and Clinical Characteristics

The study included a total of 152 patients who initiated apalu­tamide, 189 patients who initiated enzalutamide, and 387 pa­tients who initiated abiraterone acetate (Figure 2). The follow­ing values were similar across cohorts: mean age (apalutamide: 76.5 years; enzalutamide: 76.2 years; abiraterone acetate: 76.5 years); time between initial diagnosis and index date (apalu­tamide: 39.7 months; enzalutamide: 43.4 months; abiraterone acetate: 39.7 months); and Quan-CCI score (apalutamide: 9.1; enzalutamide: 9.6; abiraterone acetate: 9.8) (Table 1). Most patients were White (apalutamide: 64.5%; enzalutamide: 63.0%; abiraterone acetate: 71.6%) and had evidence of bone metastasis (apalutamide: 73.0%; enzalutamide: 76.7%; abiraterone acetate: 79.8%). Almost all patients (95.3%) who initiated abiraterone acetate had concurrent prednisone use. Patients were followed while on treatment for a mean of 7.3 months in the apalutamide cohort, 7.7 months in the enzalu­tamide cohort, and 9.4 months in the abiraterone acetate cohort.

Health Care Resource Utilization

During the baseline period, 12.5% of the apalutamide cohort, 20.6% of the enzalutamide cohort, and 18.9% of the abiraterone acetate cohort had an all-cause inpatient admission (Table 2). The mean number of all-cause admissions and days spent in an inpatient setting PPPY were 0.16 admissions and 1.01 days for the apalutamide cohort, 0.35 admissions and 2.92 days for the enzalutamide cohort, and 0.28 admissions and 1.65 days for the abiraterone acetate cohort. During the same time pe­riod, all-cause emergency department visits occurred in 43.4% of the apalutamide cohort (mean 1.09 visits PPPY), 54.0% of the enzalutamide cohort (1.32 visits PPPY), and 54.3% of the abiraterone acetate cohort (1.27 visits PPPY), and all-cause outpatient visits occurred in 100% of the apalutamide cohort (mean 25.18 visits PPPY), 99.5% of the enzalutamide cohort (33.46 visits PPPY), and 99.5% of the abiraterone acetate co­hort (36.34 visits PPPY).

During the follow-up period, 10.5% of the apalutamide co­hort, 9.0% of the enzalutamide cohort, and 15.2% of the abi­raterone acetate cohort had an all-cause inpatient admission (Table 2). The mean number of all-cause admissions and days spent in an inpatient setting (PPPY) were 0.36 admissions and 2.14 days for the apalutamide cohort, 0.36 admissions and 3.67 days for the enzalutamide cohort, and 0.65 admissions and 4.30 days for the abiraterone acetate cohort. Similarly, PC-related in­patient admissions occurred in 7.9% of the apalutamide cohort (mean 0.28 admissions and 1.71 days PPPY), 6.3% of the enzalu­tamide cohort (mean 0.30 admissions and 2.01 days PPPY), and 13.2% of the abiraterone acetate cohort (mean 0.55 admissions and 3.95 days PPPY). All-cause emergency department visits occurred in 24.3% of the apalutamide cohort (mean 1.89 visits PPPY), 32.3% of the enzalutamide cohort (1.68 visits PPPY), and 36.4% of the abiraterone acetate cohort (1.95 visits PPPY) and all-cause outpatient visits occurred in 94.1% of the apalu­tamide cohort (mean 32.31 visits PPPY), 93.7% of the enzalu­tamide cohort (38.34 visits PPPY), and 95.6% of the abiraterone acetate cohort (43.77 visits PPPY). PC-related emergency de­partment visits occurred in 10.5% of the apalutamide cohort (mean 0.36 visits PPPY), 14.8% of the enzalutamide cohort (0.61 visits PPPY), and 19.9% of the abiraterone acetate cohort (0.90 visits PPPY) and PC-related outpatient visits occurred in 91.4% of the apalutamide cohort (mean 20.46 visits PPPY), 87.3% of the enzalutamide cohort (21.07 visits PPPY), and 92.8% of the abiraterone acetate cohort (26.03 visits PPPY).

Health care costs

During the baseline period, the mean all-cause total costs PPPY were $18 216 for the apalutamide cohort, $23 633 for the enzalutamide cohort, and $27 595 for the abiraterone acetate cohort (Figure 3). Medical costs, which accounted for most of the baseline all-cause total costs, were $15 530 PPPY for the apalutamide cohort (85.3% of total costs), $21 113 PPPY for the enzalutamide cohort (89.3% of total costs), and $24 009 PPPY for the abiraterone acetate cohort (87.0% of total costs).

During the follow-up period, mean all-cause total costs PPPY were $128 187 for the apalutamide cohort, $145 382 for the enzalutamide cohort, and $114 959 for the abiraterone acetate cohort (Figure 3). Medical costs, which accounted for a smaller proportion of the follow-up all-cause total costs compared with the baseline period, were $22 902 PPPY for the apalutamide cohort, $30 933 PPPY for the enzalutamide cohort, and $35 625 PPPY for the abiraterone acetate cohort. All-cause medical costs were largely driven by outpatient costs, which were $16 696 PPPY for the apalutamide cohort (72.9% of medical costs), $20 148 PPPY for the enzalutamide cohort (65.1% of medical costs), and $24 559 PPPY for the abiraterone acetate cohort (69.0% of medical costs). All-cause inpatient and emergency department costs combined were $5922 PPPY for the apalutamide cohort, $10 388 PPPY for the enzalutamide cohort, and $10 832 PPPY for the abiraterone acetate cohort. Mean all-cause pharmacy costs PPPY were $105 285 for the apalutamide cohort, $114 450 for the enzalutamide cohort, and $79 334 for the abiraterone acetate cohort. Mean PC-re­lated total costs PPPY were $116 416 for the apalutamide co­hort, $132 236 for the enzalutamide cohort, and $100 338 for the abiraterone acetate cohort (Figure 4). Similarly, PC-related medical costs PPPY were $15 067 for the apalutamide cohort, $20 968 for the enzalutamide cohort, and $24 760 for the abi­raterone acetate cohort.

In this real-world descriptive analysis of Medicare-insured pa­tients with mCSPC treated with an ARPI, annual all-cause and PC-related medical costs were similar between those treated with enzalutamide and abiraterone acetate, and numerically lower among those treated with apalutamide. Given similarities in baseline PC-related medical costs before initiating an ARPI, descriptively higher medical costs for patients after initiation of a cancer medication may indicate suboptimal medical manage­ment due to adverse events or insufficient treatment response, reflected in variations in HRU, including hospitalizations and emergency department visits.

This study adds to the small but growing body of litera­ture demonstrating the high economic burden associated with mCSPC, particularly in the setting of progression from local­ized to metastatic disease.^29,30 In a large retrospective cohort study using administrative claims data from patients insured through commercial, Medicare Advantage, and Medicare fee-for-service health plans, Ryan and colleagues found that mean all-cause health plan-paid costs were four to five times higher after the onset of metastases in patients with castration-sensi­tive PC.²⁹ In a separate study using commercial insurance and Medicare claims data, disease progression from localized to mCSPC resulted in between two- and four-fold increases in total direct all-cause health care costs over an approximately 15-month mean follow-up period.³⁰ Importantly, HRU and costs are shown to further increase when mCSPC progresses to mCRPC.24,^31-33 In a longitudinal cohort study of patients who progressed to mCRPC, Kaye and colleagues found substantial increases in all-cause mean (SD) costs per patient per month (PPPM) from $4166 ($7548) to $8278 ($18 510), primarily driven by PC-related costs, such as pharmacy, outpatient, and imaging costs.25 Among Medicare-insured patients in the study, mean (SD) all-cause and PC-related costs PPPM increased from $3457 ($5412) to $6672 ($18 062) and $2222 ($3937) to $5455 ($16 163), respectively, following progression to mCRPC. These results suggest that effective interventions for patients with mCSPC that delay progression may help to alleviate the economic burden associated with mCRPC.

Medical costs, when used alongside clinical evidence, may help to provide a more comprehensive understanding of a medication’s therapeutic value by reflecting both its real-world economic impact and its effectiveness in clinical practice.²⁶ In this study, medical costs were numerically lower for patients treated with apalutamide relative to those treated with enzalu­tamide and abiraterone acetate, which may suggest sufficient treatment response and reduced HRU during treatment in pa­tients treated with apalutamide. This finding further supports the clinical benefit of apalutamide, which has been established in previous comparative real-world studies that used clini­cal data from community urology practices in the US. In one study, patients treated with apalutamide were 53% more likely to achieve ≥ 90% reduction in PSA (PSA90) by 6 months after initiating treatment than patients treated with abiraterone ac­etate (P = .016).²² Similarly, a separate comparative study found that patients with mCSPC treated with apalutamide were 56% more likely to achieve PSA90 than patients treated with enzalutamide (P = .014).²¹

Results from this study should be considered in the context of certain limitations. Given the descriptive nature of this study and the small sample size, no adjustments for potential con­founders, formal comparisons, or significance testing were per­formed. Differences in measured baseline characteristics, such as comorbidities (ie, Quan-CCI), prevalence of visceral metas­tasis, and resource utilization observed, may explain postindex medical costs for these cohorts. As with all observational studies that use real-world data sources (eg, electronic medical record, claims databases), coding inaccuracies or omissions may exist in the data. Specifically, the selection of patients with mCSPC re­lied on the provided electronic medical record variables, which may have led to misclassification depending on the inaccura­cies or omissions contained in the database. Lastly, given earlier mCSPC indication approval for abiraterone acetate, patients in this cohort were observed with longer follow-up than those treated with apalutamide or enzalutamide.

In this descriptive study of patients with mCSPC, annual all-cause and PC-related medical costs were similar between enzalutamide and abiraterone acetate cohorts, and descriptive­ly lower in the apalutamide cohort. Variations in medical and pharmacy costs observed among the three ARPI agents inves­tigated may reflect how well these agents manage PC. Future work leveraging more recent data and enhanced cohorts will allow for formal comparative analyses evaluating HRU and costs across different ARPIs in this area, which may facilitate a framework for health care value that may be of interest to population health decision makers.

Affiliation:

¹Johnson & Johnson, Horsham, PA, USA; 2Analysis Group, Inc, Montréal, QC, Canada

Funders:

Johnson & Johnson sponsored the study.

Correspondence:

Carmine Rossi, PhD

Manager, Analysis Group, Inc.

1190 Av. des Canadiens-de-Montréal, Suite 1500

Montreal, QC H3B 0G7, Canada

Phone: 514-871-4233

Email: carmine.rossi@analysisgroup.com

Disclosures:

C.R., J.K., L.D., B.M., and D.P., are employed by Analysis Group, Inc, a consulting company that has provided paid consulting support to Johnson & Johnson, which sponsored the study. I.K., S.D., and L.E. are employed by Johnson & Johnson.