Comparative Economic Evaluation of Radical Prostatectomy, Radiation, and Ablative Techniques in the Management of Localized Prostate Cancer

1. Introduction

The current management of localized, low- to intermediate-risk prostate cancer (PCa) includes a spectrum of approaches, ranging from active surveillance to definitive treatments such as radical prostatectomy and radiation therapy [1]. Additionally, minimally invasive ablative therapies could be considered in select patients, particularly within clinical trials or prospective registries, with the primary aim of preserving quality of life [2,3,4]. The current guidelines recommend a multidisciplinary approach before deciding on the management of localized PCa due to the wide range of available options and their potential side effects [5,6]. In addition, treatment cost is a critical factor in the decision-making process for patients, clinical specialists, and policymakers. It has been shown that over 50% of cancer patients are willing to discuss the economic aspects of their treatment; however, less than 33% of such cases receive sufficient counseling on these financial issues [7].

Although clinical and patient-centered outcomes of the aforementioned treatment options have been extensively studied, direct comparisons of treatment costs remain relatively scarce, particularly between conventional options and minimally invasive modalities such as cryotherapy (Cryo), high-intensity focused ultrasound (HIFU), and low-dose brachytherapy (LDBT) [8]. A British cost-effectiveness study compared focal Cryo and focal HIFU to robotic-assisted radical prostatectomy (RALP) and external-beam radiation therapy (EBRT), reporting higher quality-adjusted life year gains for the focal options [9]. Furthermore, a recent German cost-utility analysis compared MRI-guided transurethral ultrasound ablation (TULSA) with active surveillance (AS), RALP, and EBRT, and showed that for a low- to intermediate-risk PCa and considering a willingness-to-pay of approximately €14,000, the probability of being cost effective is similar between AS, RALP, and TULSA, while higher thresholds of willingness-to-pay favored AS [10]. Due to the paucity of contemporary economic evaluations in the U.S. health setting, we aimed to conduct a cost comparison of open retropubic radical prostatectomy (RRP), RALP, intensity-modulated radiation therapy (IMRT), LDBT, stereotactic body radiotherapy (SBRT), Cryo, and HIFU for low to intermediate-risk PCa at a tertiary referral center, from the healthcare system perspective.

2. Materials and Methods

2.1. Patient Population

We conducted a retrospective cohort study at Duke University Hospital to compare institutional costs and charges associated with various primary treatment modalities for eligible consecutive patients with localized PCa treated between January 1, 2018, and December 31, 2019. The study received approval from the institutional review board. Patients were identified by querying financial records for their medical record numbers associated with International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) code “C61” who had at least one registered encounter during the specified study period. These cases were further narrowed using procedure-specific Current Procedural Terminology (CPT) codes (Supplementary Material, Table S1) for surgical procedures (RRP, RALP, Cryo, HIFU), financial charge codes maintained by Duke Finance for radiation modalities (IMRT, SBRT, LDBT), and a manual review to ensure capture of all eligible cases. To maintain data homogeneity, cases were excluded if patients had high-risk disease according to National Comprehensive Cancer Network (NCCN) guidelines, had received prior or repeat PCa treatment, or had incomplete financial data on file. Of note, the selection of treatment modalities involved a shared decision-making process between patients and physicians.

2.2. Study Variables and Definitions

Study variables included patient demographics (age, race), NCCN risk groups, treatment modality, number of treatment sessions, costs, institutional charges, and use of modality-specific procedures such as fiducial placement and hydrogel spacer injection for IMRT and SBRT, prostate volume study, simulation, dosimetry, and seed placement for LDBT, focal versus whole-gland pattern of treatment for ablation modalities. The primary objective of the study was to report and compare the mean and standard deviation (SD) of total and direct institutional costs and total charges. The secondary objective was to assess the correlation between treatment characteristics (e.g., session number, technology-intensiveness) and costs. The exploratory objectives were to report the costs and charges of spacer gel injection for IMRT and SBRT, as well as the costs and charges of post-treatment biopsy and mpMRI for ablation modalities.

In this study, “cost” denotes the actual expense incurred by a provider to deliver a service, whereas “charge” refers to the price assigned by the provider for that service. Direct costs are those associated with patient care, such as procedures, medications, and clinical staff, while indirect costs are necessary but non-clinical expenses, including administrative support and facility operations. Cost data were gathered from the time a treatment modality was selected through to the completion of that treatment. For radiation therapy modalities (SBRT, IMRT, and LDBT), this period extended to the final radiation session, while for RALP, RRP, HIFU, and Cryo treatments, cost analysis was conducted through hospital discharge. Diagnostic services, such as multiparametric MRI, diagnostic prostate biopsy, or potential genomic evaluations, were excluded from cost calculations if performed prior to the assignment of the treatment plan. For radiation therapy, additional procedure-specific expenses related to fiducial marker placement, hydrogel spacer injection, and dosimetry were also included (Table 1). Regarding ablative modalities, the costs and charges of focal and whole-gland techniques were kept unified in the cost comparison analysis due to the similar CPT code. All treatments except RRP were categorized as technology-intensive due to their reliance on specialized equipment and planning protocols. The discount rate was not applied, given the 2-year study period.

2.3. Statistical Analysis

Continuous variables, such as direct and total costs, were summarized using means and SDs, while categorical variables, including treatment type and risk classification, were described as frequencies and percentages. Economic evaluations mostly use mean and standard deviation. However, due to the sample size considerations, we reported both the mean and the median for better illustration of the data spread. The Chi-squared test was used to compare the race and risk groups across treatment modalities. To compare age differences across treatment groups, we used a One-way ANOVA test with Tukey post hoc comparisons. The Mann-Whitney U test with Bonferroni correction was used for pair-wise cost comparisons. Relationships between treatment costs and factors such as the number of treatment sessions or use of technology-intensive modalities were evaluated using Spearman's correlation. All statistical analyses were performed using SPSS version 26 (IBM Corp., Armonk, NY), with statistical significance set at a p-value of < 0.05, unless corrected with Bonferroni, which is mentioned in the results when applicable.

3. Results

We identified 633 cases of low- to intermediate-risk PCa patients who underwent primary treatment at our center. Of these, 83 cases were excluded due to missing financial data. Of note, 73 of 83 excluded patients belonged to the IMRT cohort, mostly because they received at least one part of the treatment (fiducial placement or some radiation sessions) outside our center.

A total of 550 patients met the inclusion criteria during the study period. Of these, 372 (68%) were White, 149 (27%) patients were Black, 14 (2%) belonged to other ethnic backgrounds, and 15 (3%) did not declare their race. Table 2 presents details on age, race, NCCN risk groups, and the number of sessions for each modality. The mean (SD) age of all participants was 65.0±7.2 years. Pairwise Mann-Whitney comparisons with Bonferroni corrections showed that participants in Cryo (69.43 years) and IMRT groups (69.27 years) were statistically significantly older (p<0.0024) than those in the LDBT (64.54 years), RALP (62.91 years), and RRP groups (62.83 years). NCCN low-, favorable-intermediate-, and unfavorable-intermediate-risk categories were identified in 85 (13%), 218 (41%), and 249 (46%) patients, respectively. RALP, RRP, Cryo, and HIFU were single-session interventions, whereas IMRT, SBRT, and LDBT were delivered over multiple sessions. The number of radiation delivery sessions is provided in Table 2. Of note, an additional 2-3 visits for fiducial placement, simulation, and dosimetry should also be considered for IMRT and SBRT.

Table 3 outlines the breakdown of mean total costs, direct costs, and overall institutional charges by treatment type. IMRT and SBRT were the two most resource-intensive modalities, with mean total costs of $28.5 K (±6.0 K) and $20.3 K (±1.7 K), and charges averaging $136.2 K (±10.0 K) and $121.3 K (±6.9 K), respectively. RALP was the third most costly modality, with mean total costs of $14.0 K (±3.3 K) and overall charges averaging $40.3 K (±6.6 K). On the other hand, RRP was the least costly treatment overall, followed by Cryo, LDBT, and HIFU. Among focal therapy modalities, Cryo had an average total cost of $10.5 K (±1.4 K), and HIFU showed a mean total cost of $12.9 K (±1.5 K).

Supplementary Tables S2-S4 present the detailed costs for LDBT, IMRT, and SBRT modalities, respectively. Regarding LDBT, seed placement was the most expensive step, followed by the prostate volume study, simulation, and post-implant dosimetry. We could not separate the IMRT and SBRT costs similar to LDBT, as several patients received simulation and fiducial placement in a single session.

Supplementary Table S3a and S4 break down the costs and charges for patients who received hydrogel spacer injections compared to those who did not. As anticipated, spacer injections incurred significantly higher charges; nevertheless, higher total and direct costs were observed for patients without spacer injection for both SBRT and IMRT cohorts. This unexpected finding prompted further investigation. As shown in Table S3 b, the location of service caused the unexpected difference in costs; All spacer gel injections were delivered in the interventional radiology clinic, in a community hospital, while the fiducial placements without a spacer were done in the academic center, either in the clinic or the ambulatory surgery facility’s operative room. Table S5 provides costs and charges for Cryo and HIFU modalities, stratified by ablation pattern used (focal versus whole-gland). The Mann-Whitney test did not show a significant difference between the focal and whole-gland subgroups for each modality.

Statistical analysis revealed a significant positive correlation between the number of treatment sessions and total cost (Spearman’s ρ = 0.55, p<0.001), indicating that multi-session therapies tend to incur higher expenses. Similarly, classification as a technology-intensive modality was also significantly correlated with overall cost (ρ=0.58, p<0.001).

Among the ablation cohort (68 cases overall), 38 patients underwent follow-up multi-parametric MRI, and 17 underwent biopsy. The costs and charges for these cases were not included in the cost comparison analysis per our inclusion criteria, but are reported separately as complementary information (Table S6). Surveillance MRI and biopsy added $0.9 K (±0.2 K) and $4.0 K (±1.0 K) to the costs and $6.5 K (±1.7 K) and $9.0 K (±4.5 K) to the charges, respectively.

4. Discussion

To our knowledge, this is the first cost-comparison study to report on seven treatment modalities for PCa. We identified significant variation in associated costs across treatment modalities. While the majority of patients were White and classified within intermediate-risk NCCN categories, notable differences in age distribution were observed, with older patients more likely to receive Cryo or IMRT. Treatment modality played a key role in cost differentiation, with IMRT and SBRT were the most expensive options, both in total costs and institutional charges, followed by RALP, HIFU, LDBT, Cryo, and RRP. Multi-session and technology-intensive interventions were strongly correlated with higher costs.

Extracting and evaluating IMRT charges are challenging and vary significantly between institutions; Agrawal et al. evaluated 52 hospitals of National Cancer Institute-designated cancer centers and showed IMRT “delivery-only” charges could vary by a 21.7-fold difference between the least and most expensive hospitals [11].

Among radiation modalities, the higher number of sessions in the IMRT cohort resulted in higher costs compared to SBRT. When interpreted in the context of the PACE-B trial, which demonstrated comparable oncological efficacy and toxicity profiles between the two radiation modalities, SBRT appears to be a more cost-effective option [12,13]. However, IMRT and SBRT were more than twice as expensive as LDBT, which is comparable to the findings from the national Medicare data analysis conducted by Tang and colleagues [14].

Our study found that RALP costs were almost 1.5 times higher, and its charges were nearly twice than that of RRP. Although post-operative costs were not included in our analysis, previous research has shown that the higher upfront costs of RALP could be offset within the first year due to increased healthcare use following open RP [15]. Notably, a recent meta-analysis by Bejrananda et al., evaluating the economic outcomes of RALP versus open RP in localized PCa revealed that both techniques are comparable in terms of cost-effectiveness. However, RALP appears to offer a more favorable long-term economic profile, while open RP remains a financially viable option in both high- and middle-income countries, particularly in the latter [16].

Focal ablative therapies for intermediate-risk PCa are relatively new alternatives for the definitive management of PCa, with promising trials showing favorable mid- and long-term oncological and functional outcomes [3,17,18,19]. The American Urological Association and European Association in Urology guidelines recommend offering focal ablative therapies to highly selected patients within clinical trials or prospective registries [3,17]. A meta-analysis in 2023 on the primary whole-gland Cryo and HIFU ablation included 29 studies with a minimum sample size of 50 patients, showing that cancer-specific, overall, and metastasis-free survival rates at 10 years were 96%, 63%, and 84%, respectively [20]. Multiple multidisciplinary consensus statements recommended post-ablation imaging for all focal therapy patients, and biopsy was mostly recommended to be performed “for cause” rather than “per protocol” [21,22]. The MRI cost was not included in the main cost comparison for ablation modalities because our cohort included both focal and whole-gland ablations, and not all patients undergo MRI; however, a separate comparison accounting for MRI cost was provided in the supplementary material. The costs and charges of whole-gland treatment did not statistically differ from those of focal ablation. Biopsy costs were also not included in the comparison. The rationale was that a for-cause biopsy, similar to the workup for biochemical recurrence after definitive treatment, necessitates additional evaluation, such as biopsy following ablation and further imaging after definitive management.

Conducting economic evaluations for contemporary, rapidly evolving PCa treatment modalities presents several challenges. These complexities arise from the wide range of available treatment options, which may be delivered in different settings (e.g., inpatient vs. outpatient, academic vs. non-academic institutions, etc) and across varying delivery sessions (e.g., single vs. multiple). The heterogeneity of the localized PCa patient population, such as differences in cancer characteristics and comorbidities, further complicates such analyses. Additional factors influencing outcomes include the choice of time horizon (short- vs. long-term), the selection of cost components (e.g., diagnostics, consultations, readmissions, lost workdays), and whether the focus is on oncological or functional outcomes. These challenges are compounded by the broader complexities inherent to economic evaluations, such as determining the appropriate analytical perspective (e.g., payer, healthcare system, societal) and selecting the most suitable framework—whether cost-comparison, cost-effectiveness, or cost-benefit analysis.

Our results are limited by the following factors. First, we did not include post-treatment costs, as some patients followed up with their local physicians, resulting in missing data on post-operative emergency room visits, re-admissions, complication management, and possible adjuvant therapy. Second, the study period was short; however, this was to avoid confounding effects from the COVID-19 pandemic. Third, while the study adopted a healthcare system perspective, we could not account for out-of-pocket costs due to the heterogeneity of data and variability of insurance coverage across patients and treatment modalities.

5. Conclusions

Significant variation is observed in associated costs across treatment modalities for localized PCa, with the highest cost seen in IMRT and SBRT, followed by RALP, HIFU, LDBT, Cryo, and RRP. The number of sessions and technological complexity are significantly correlated with increased costs, highlighting the need to weigh economic considerations in treatment planning. Studies that evaluate costs alongside functional and oncological outcomes across multiple modalities are necessary to draw more definitive conclusions.