Contemporary patterns of local ablative therapies for prostate cancer at United States cancer centers: results from a national registry

To describe the national-level patterns of care for local ablative therapy among men with PCa and identify patient- and hospital-level factors associated with the receipt of these techniques. We retrospectively interrogated the National Cancer Database (NCDB) for men with clinically localized PCa between 2010 and 2017. The main outcome was receipt of local tumor ablation with either cryo- or laser-ablation, and “other method of local tumor destruction including high-intensity focused ultrasound (HIFU)”. Patient level, hospital level, and demographic variables were collected. Mixed effect logistic regression models were fitted to identify separately patient- and hospital-level predictors of receipt of local ablative therapy. Overall, 11,278 patients received ablative therapy, of whom 78.8% had cryotherapy, 15.6% had laser, and 5.7% had another method including HIFU. At the patient level, men with intermediate-risk PCa were more likely to be treated with local ablative therapy (OR 1.05; 95% CI 1.00–1.11; p = 0.05), as were men with Charlson Comorbidity Index > 1 (OR 1.36; 95% CI 1.29–1.43; p < 0.01), men between 71 and 80 years (OR 3.70; 95% CI 3.43–3.99; p < 0.01), men with Medicare insurance (OR 1.38; 95% 1.31–1.46; p < 0.01), and an income < $47,999 (OR 1.16; 95% CI 1.06–1.21; p < 0.01). At the hospital-level, local ablative therapy was less likely to be performed in academic/research facilities (OR 0.45; 95% CI 0.32–0.64; p < 0.01). Local ablative therapy for PCa treatment is more commonly offered among older and comorbid patients. Future studies should investigate the uptake of these technologies in non-hospital-based settings and in light of recent changes in insurance coverage.


Introduction
Historically, the gold standard treatment for localized prostate cancer (PCa) is radical prostatectomy or radiotherapy. While techniques such as cryotherapy and high-intensity focused ultrasound (HIFU) have been available for years, the recent utilization of prostate multiparametric magnetic resonance imaging (MRI) has improved our ability to detect clinically significant PCa and avoid treatment of International Society of Urological Pathology [1]. Additionally, MRI technology has increased the potential and demand for less invasive subtotal gland approaches to treat PCa.
Nevertheless, image-guided local ablative therapy is only as good as the imagining technique's ability to map the tumor location and volume [2].
Local ablative therapies have recently gained popularity in reducing the burden of treatment for men with low-and intermediate-risk PCa [3]. As such, local treatment of the prostate would reduce the common treatment-associated side effects of radical prostatectomy and radiotherapy. Indeed, urinary incontinence and erectile dysfunction are common side effects that directly impact men's quality of life [4][5][6]. Thanks to the effectiveness and safety of local PCa treatment, the American Urological Association (AUA) and the European Association of Urology guidelines now recommend ablative approaches for men at increased risk of surgical morbidity [7,8].
In the context of an aging population, there is a clear interest in local cancer treatment options. Local ablative therapy includes a variety of minimally invasive techniques aimed at destroying tumor cells and sparing normal tissue and adjacent structures such as the external urethral sphincter and neuro-vascular bundles. Thus, the goal is to provide adequate cancer control while minimizing treatment-related toxicities which is of particular interest to men with intermediate-risk PCa warranting active treatment versus men with low-risk PCa who would be eligible for active surveillance. Until now, data supporting the use of local ablative therapies stem from either single surgeon series or from highly controlled research trials that were performed in centralized treatment centers [9][10][11]. Thus, there is a knowledge gap regarding the real-world uptake of PCa local ablative therapy in the United States (US) [12].
In the early period of the adoption of local ablative techniques, understanding the geographic and temporal trends in local ablative therapy utilization will help inform treatment guidelines and reimbursement policies. Herein, we aim to (1) describe the national-level patterns of care for local ablative therapy among men with PCa, and to (2) identify patient-and hospital-level factors associated with the receipt of these techniques. We hypothesize that there will be a national increase in the proportion of local ablative therapy used among men with intermediate-risk PCa.

Data source
Data were retrospectively obtained from the National Cancer Database (NCDB). The NCDB is a comprehensive cancer registry which includes patients seen at one of 1500 participating Commission on Cancer (CoC) accredited hospitals. Data abstractors use a standardized methodology to collect sociodemographic and clinical data. The registry is capture more than 50.8% of PCa cases in the US [13,14].

Study population, endpoints, and covariates
Men with a new diagnosis of clinically localized PCa between 2010 and 2017 were selected. For each patient, age at diagnosis, gender, race/ethnicity, insurance status, family income, and tumor stage were assessed. Men were categorized accordingly to the D'Amico risk-class classification [15]. Exclusion criteria were age at diagnosis ≤ 49 years and presence of local or distant metastasis at diagnosis.
The main outcome was receipt of local ablative therapy as the primary treatment, defined as local tumor destruction including "cryoprostatectomy, laser ablation, and other method of local tumor destruction including high-intensity focused ultrasound (HIFU)". Hyperthermia was excluded due to low number of participants of this treatment modality. Men who underwent local tumor excision using transurethral resection of the prostate were also excluded. The different included modalities of local ablative therapy are categorized as in Supplementary Table 1. Because the NCDB does not distinguish between the focality of ablative therapies performed for local therapy destruction (e.g., between focal, hemiablation, hockeystick or whole gland) we categorized all forms of ablative therapy for local tumor destruction as local ablative therapies in the current study.
Sociodemographic covariates comprised age at diagnosis, race (White, Black, other), primary insurance type, and family income. Clinical comorbidities were estimated using the Charlson-Deyo modification of the Charlson Comorbidity Index score (CCI; 0, 1, > 2) [16]. Hospital-level covariates included facility type, geographic location, location in the US, hospital annual caseload year by year 2010-2017.

Statistical analysis
First, baseline characteristics were compared using t-test and chi-squared test for continuous and categorical variables, respectively. Second, mixed effect logistic regression models were fitted to identify, in separate models, patient-level and hospital-level predictors of receipt of local ablative therapy. Third, Cochran-Armitage test was used to test for trends in receipt of local ablative therapy between 2010 and 2017. Significance was set at a two-sided p < 0.05. All statistical analyses were performed using Stata v.17.0 (StataCorp, College Station, TX, USA).

Hospital-level predictors of receipt of local ablative therapy
At the hospital-level, patients treated in academic/research programs were less likely to receive local ablative therapy (OR 0.45; 95% CI 0.32-0.64; p < 0.01) ( Table 2). Hospitals with a high annual caseload were more likely to offer local ablative therapy (2nd vs. 1st quartile, OR 1.23; 95% CI 1.12-1.35; p < 0.01), as well as patients living in urban and rural areas compared to metropolitan areas (rural: OR 1.33; 96% CI 1.15-1.52; p < 0.01).

Trends of local ablative therapy
Cochran-Armitage test showed a decrease of local ablative therapy use among PCa patients captured in NCDB from 2% in 2010 to 1% in 2017 (p < 0.01 for trend).

Discussion
In this retrospective study of men treated for localized prostate cancer between 2010 and 2017, we analyzed patientlevel clinical and demographic patterns in use of local ablative therapy at CoC accredited centers in the US. Within the subset of analyzed hospitals, we found that the use of local ablative therapy decreased across the study period with cryotherapy being the most common type of focal therapy. Intermediate-risk PCa was more likely to be treated with local ablative therapy. At the patient-level, we identified a family income < $47,999, Medicare insurance, Black race, age over 61 years, and a CCI ≥ 1 as factors associated with an increased likelihood of receiving local ablative therapy. Finally, academic and research centers were less likely to use local ablative therapy. Taken in context of growing interest in local ablative therapies, our findings highlight an important research gap regarding the use of these minimally invasive technologies. We found that older patients and patients with more comorbidities were more likely to receive local ablative therapy. The mean age in our study was 69.81 years which is comparable with the mean ages in prior studies on HIFU and cryotherapy [17,18]. Nevertheless, local ablative therapy was developed on the premise to provide PCa treatment by minimizing treatment-associated toxicities and preserving quality of life. The higher odds of receiving local ablative therapy among elderly men might be due to their comorbidities which puts them at higher surgical and anesthesia risk for adverse events if they were to receive radical prostatectomy [19,20]. In fact, the AUA guidelines recommend considering cryotherapy for men with low-and intermediate-risk PCa who are "not suitable for either radical prostatectomy or radiotherapy due to comorbidities and yet have > 10 year life expectancy" [21]. In contrast, several specialized centers in Europe have prioritized techniques to optimize patient selection for focal ablation based on disease characteristics including MRI features, tumor size, and location, rather than based on age and patient comorbidities [22]. This might explain why more recent European cohorts might have younger mean age and a lower proportion of patients with comorbidities [11].
Men with lower income and with Medicare were less likely to receive local ablative therapy. Insurance coverage of local ablative therapy in the US has been delayed, and even today coverage is often on case-by-case basis. Therefore, we expected a greater utilization of local ablative therapy among men in higher income. However, this was not the case and may warrant further investigation. Possible explanations include enrollment of lower income patients in clinical trials or the fact that higher income individuals may be more likely to receive local ablative therapy in cash pay settings which are not captured in the NCDB. Black patients were more likely to receive local ablative therapy than their White counterparts. Black men in the US are 1.6 times more likely than other men to get PCa, and twice as likely to die from it [23]. Black-White difference in access to focal therapy is multifactorial. First, focal therapy was mostly offered at cancer community programs which are strategically located in the outreach setting. Thus, minoritized populations could have had increased geographic accessibility. Second, treatment-associated toxicities of radical prostatectomy (e.g., urinary incontinence) were shown to be a significant factor in the treatment decision-making among Black men due to occupational and/or cultural factors [24]. Third, it is possible that clinical trial accrual for focal therapy has particularly attracted minoritized populations who are often underinsured or uninsured. However, it remains unclear whether local ablative therapy is an approach to overcome the racial disparities in PCa treatment. Thus, future qualitative studies should investigate the access of participants for local ablative therapy.
In line with the AUA recommendations, we found that men with intermediate-risk PCa were more likely to receive local ablative therapy than men with low-or high-risk PCa [21]. For men with low-risk PCa, active surveillance is the preferred and safer option while local ablative therapy could be an option for those who progress on active surveillance [25].
Although local ablative therapy has been included in the AUA guidelines within the study period, we saw a decrease in the proportion of local ablative therapy in our patient sample. This is possibly because cryotherapy including older versions of whole gland and hemiablation were included within the definition of local ablative treatment. Cryotherapy was a popular treatment option in the early 2000s, but has declined as more recent forms of local ablative therapy were introduced [26]. Additionally, while HIFU was approved by the Food and Drug Administration in 2015, Medicare and insurance coverage only started in 2019, after the study period.
Furthermore, we showed that academic and research centers are less likely to provide local ablative therapy than community cancer program. The CoC's community cancer programs often have a majority of private practice physicians, but thanks to their CoC association, would offer clinical trials and new treatments, including focal therapy [27]. This may be different than in Europe, where the majority of focal therapy is offered at academic and research centers [9,11,12]. Theoretically, cases identified from academic and research hospitals should reflect all cases including uninsured and self-payers. This is a strength of the NCDB relative to claims-based data. However, there is a disconnect between the growing use of MRI for PCa diagnosis [28], the growing interested in focal therapy [9,12,29], and the declining proportion of ablative therapy in this study. These technologies might be more frequently performed at outpatient surgical centers or in community centers than at the hospitals that report data to the NCDB [13].
This study is not devoid of limitations. First, we were not able to identify a specific variable for HIFU, or to distinguish between local and whole gland ablation as this is not captured in NCDB. Due limitations inherent to the registry, there remains unanswered questions regarding the payer type for local ablative treatment. Since insurance coverage for HIFU started in 2019 (after the study period), we might expect an increase in the proportion of focal or as a whole gland therapy. The NCDB does not capture procedures performed in outpatient surgical centers, which might limit the generalizability about the "true current state" of which type of facility is more likely to offer local ablative therapy in the US. Nevertheless, we identified patterns in the use of local ablative therapy within the US population captured in the NCDB. While most contemporary guidelines focus on cancer staging, we identified several patient characteristics that should be considered in upcoming policy implementations. We identified a trend towards older patients and patients with comorbidities receiving local ablative therapy as a primary approach for PCa treatment. Our results provide one of the first national analyses on trends in local ablative therapy over the past decade; however, there remains several unanswered questions especially regarding cost-effectiveness. While financial toxicities of radical prostatectomy and radiation therapy for PCa are well known, the financial burden associated with local ablative therapy in the US at the expense of reduced treatment-associated toxicities is attractive to a value-based health care system that prioritizes outcomes that matter to patients [30]. In the context of growing interest in local ablative therapies for PCa treatment, our findings highlight an important research gap regarding the trends and predictors of receipt of these technologies.

Conclusions
We conducted one of the first US national analyses on trends of local ablative therapy for PCa. Older and multimorbid patients were more likely to receive local ablative therapy as a primary treatment for PCa. Patients treated in academic and research programs were less likely to receive local ablative therapy than those treated in community cancer programs. Future studies should investigate the trend in use of local ablative therapy in non-hospital-based settings and further tease out the insurance type most likely to cover these procedures.

Material and/or code availability
Code is available upon request from the corresponding author.