- Top of page
- MATERIALS AND METHODS
- FUNDING SOURCES
- CONFLICT OF INTEREST DISCLOSURES
Although permanent prostate brachytherapy (BT) monotherapy is an established treatment option for low-risk prostate cancer,[1-4] its appropriateness for men with intermediate- or high-risk disease is less well defined. Historically, these patients were not considered suitable candidates given that the higher predicted risk of extraprostatic disease may not adequately be treated with BT alone.
These concerns were in part confirmed by a large, observational cohort study published in 1998. Among the 1872 men treated with either BT monotherapy or radical prostatectomy (RP), biochemical outcomes were similar among men with low-risk disease, but inferior among men who received BT monotherapy for intermediate- or high-risk disease. In 1999, the American Brachytherapy Society (ABS) published its recommendations that BT monotherapy was an appropriate treatment for low-risk patients, but required supplementation with external beam radiotherapy (EBRT; ie, EBRT+BT) if used as definitive treatment for high-risk disease. They recommended that intermediate-risk patients be evaluated on a case-by-case basis. In the 2000 National Comprehensive Cancer Network (NCCN) practice guidelines, BT monotherapy was recommended only for low-risk disease.
Several retrospective studies have since suggested that biochemical and even prostate-specific survival outcomes are not significantly different among men who received BT monotherapy for intermediate-risk disease, as compared with RP or EBRT.[10, 11] Given conflicting retrospective studies and lack of prospective, randomized data to guide physicians on the relative benefit of BT monotherapy, we evaluated within a large US-based registry the impact of the 19997 and 20008 national guidelines on BT monotherapy utilization for definitive treatment of localized prostate cancer.
- Top of page
- MATERIALS AND METHODS
- FUNDING SOURCES
- CONFLICT OF INTEREST DISCLOSURES
In this patterns-of-care study from a large, US-based registry of men with prostate cancer, BT monotherapy for definitive treatment of localized prostate cancer significantly increased from the 1990-1998 to 1999-2011 time periods, in particular relative to EBRT and RP. The temporal trends and disease and demographic characteristics of patients who received BT are consistent with prior studies from the Surveillance, Epidemiology, and End Results (SEER) cancer registry and a 1999 patterns-of-care survey of radiation oncology facilities. When evaluated by risk group, BT monotherapy use increased over time among men with low-risk disease. This finding is not wholly unexpected, because BT monotherapy for low-risk disease is readily recommended by national practice guidelines[3, 7, 8] and is cost-effective.
More provocative is that treatment of high-risk groups increased, despite recommendations to the contrary from 2 national practice guidelines in 19997 and 2000. Of note, this increased utilization was not sustained across all high-risk groups; compared with EBRT and RP, BT monotherapy use among high-risk men decreased linearly within the 1999-2011 time period. Clinical practice guidelines (CPGs), which synthesize the pertinent literature, are created to limit practice variations toward evidence-based treatments. However, CPGs have not always been successfully adopted among physicians. A contemporary, SEER-based study reported increased use of BT for accelerated partial breast irradiation (aPBI) between 2000 and 2007. Based on the American Society of Radiation Oncology (ASTRO) consensus guidelines, 65.8% of women who received aPBI BT were classified as cautionary or unsuitable candidates, and use within these 2 categories increased over time. Deviation from CPGs is likely multifactorial and in part based on the quality of the guidelines and/or supporting evidence, characteristics of target health care professionals and practice settings, patient-related factors, and regulation.
In the case of management of localized prostate cancer, choosing the “best” treatment is complicated by the absence of randomized data comparing the efficacy and toxicity of BT monotherapy with EBRT, EBRT+BT, EBRT+ADT, and/or RP. Although BT monotherapy appears to be at least as efficacious as RP or EBRT for men with low-risk disease,[9, 12] results are mixed but generally poor for high-risk men treated with BT monotherapy. Relapse-free survival at 8 years was 48% in a multi-institutional series and ranged from 60% to 90% at 12 years in other series.[22, 23]
The literature is even more heterogeneous for BT monotherapy treatment of intermediate-risk disease. Poorer outcomes initially seen in some series but not others[9, 11] may in part be secondary to the quality of BT implants, which has been shown to affect biochemical outcome. Furthermore, retrospective, observational data suggest that some but not all intermediate-risk patients may be safely treated with BT monotherapy; the RTOG 0232 trial, which evaluates the role of supplemental EBRT to BT for men with either PSA of 10 to 20 and Gleason score < 7 or Gleason score 7 with PSA < 10, is currently addressing this question in a prospective and randomized setting. Although no consensus definition exists, similar favorable intermediate-risk men have also been identified by expert prostate brachytherapists as potential candidates for BT monotherapy. Therefore, although BT monotherapy use increased among intermediate-risk men, a proportion of these men may comprise the favorable subset that is increasingly being accepted as a standard in practice.
This highlights an important, dynamic role of CPGs. In situations where prospective, randomized data does not exist, CPGs also should evolve to represent existing beliefs or values. The updated 2012 ABS guidelines reflect this in their recommendation that intermediate-risk patients with favorable features may be treated with BT monotherapy. Of note, the 2013 NCCN guidelines still does not endorse use of BT monotherapy for treatment of intermediate-risk disease. In our study, however, increased BT monotherapy use among intermediate-risk men was not exclusively driven by treatment of favorable-intermediate risk disease; use also increased among the unfavorable-intermediate risk group. Furthermore, although the increased use among favorable-intermediate risk men could in part reflect changes in Gleason grading that occurred in the late 1990s and early 2000s, it would not be able to explain the increased use among men with Gleason 4+3 and other intermediate-risk features (ie, unfavorable-intermediate risk group).
In the absence of strong scientific evidence, treatment of localized prostate cancer has been categorized as preference-sensitive care in which a patient's values of the perceived benefits and harms may drive treatment selection. Therefore, patient preference may have contributed to the temporal trends seen with BT monotherapy. Patients may prefer a single treatment and/or the perceived side-effect profile. Although patient convenience, measured in some studies as distance to a radiotherapy facility, has explained some variations in treatment received such as use of mastectomy versus breast-conserving surgery and radiotherapy, this has not necessarily been the case for BT utilization. In the study by Hattangadi and colleagues, location outside of a metropolitan area was associated with decreasing (rather than increasing) odds of receiving aPBI BT for breast cancer; similarly, urban residence (versus nonurban) was a predictor of BT and BT+EBRT use among men being treated for localized prostate cancer.
Finally, physician and possibly clinic preferences may drive BT monotherapy use. The odds of BT monotherapy utilization compared with all other treatments increased with age (OR = 1.11). This could be due to physicians' perceptions that older patients may not tolerate a long EBRT treatment course or a surgical procedure such as RP. At the same time, we observed considerable site-to-site variability for treatment of unfavorable-intermediate and high-risk patients (Fig. 2). Although the factors that contribute to this are unknown, we hypothesized that a site's BT experience (assessed using the proportion of all patients treated with any BT) could influence treatment decisions. Although exploratory in nature, clinics that treated a larger proportion of all patients with prostate cancer with any BT also treated a higher proportion of unfavorable-intermediate and high-risk patients with BT monotherapy. These observations require confirmation in other large, population-based registries.
Aligning treatment practices with those of CPGs is an active area of research. Basing CPGs on better quality data (ie, prospective, randomized) would likely promote concordance among CPGs and provide more weight with recommendations. When mature, data from RTOG 023214 will help clarify who may be an appropriate BT monotherapy candidate. Few studies have evaluated how practicing physicians including urologists and radiation oncologists view and utilize national guidelines. In a survey-based study of Canadian oncologists, 41% reported using guidelines routinely or most of the time, and 20% felt that they were too rigid to apply to individual patients. These results cannot be extrapolated to our US-based study, but they are consistent with our findings that physicians may not be using guidelines in their treatment of men with high-risk prostate cancer. Finally, patients can be better educated with decision aids, which are designed to help them understand likely outcomes for each treatment option and the level of scientific uncertainty. A systematic review of trials demonstrates that patient decision aids improved both a patient's realistic perception of benefits and harms and the agreement between a patient's values and the option chosen. Together, these interventions may reduce the variability in BT monotherapy used, in particular among patients who may not be the most appropriate candidates.
Although BT monotherapy use increased from the 1990-1998 to 1999-2011 time periods across all risk groups, it is important to note that there is some evidence of a decrease in BT monotherapy use across years within the 1999-2011 period, especially in the higher risk groups. This may reflect a slow acceptance of CPGs into clinical practice; prior studies have observed that on average, it takes 17 years for results of a randomized controlled trial to be implemented in the community. However, this decline also coincides with the timing of Medicare reimbursement for intensity-modulated radiotherapy and the rise of robotic-assisted RP. Therefore, the relative decline in BT monotherapy during this period may also be secondary to increased utilization of newer technology,[31, 32] although our dataset does not provide the information to make such inferences.
Our study has several strengths. The CaPSURE registry provides detailed data on a large, representative sample of US men with prostate cancer and is drawn from various site types including VA, academic, and community practices. A few points require further consideration. Physicians may consider other disease characteristics such as percent of positive biopsies and presence of perineural invasion when deciding if a patient is a suitable candidate for BT monotherapy, but this is not reflected in the risk groups that we used for our analysis. Nonetheless, using the risk groups defined by the ABS in 1999 and NCCN in 2000 permitted us to directly compare whether BT monotherapy use was concordant with the aforementioned guidelines. We subdivided the intermediate-risk group into favorable and unfavorable categories based on general consensus among expert brachytherapists and the eligibility requirements for the ongoing RTOG 0232 trial. However, these definitions have not been validated, and future studies are required to determine whether these subcategories are meaningful in predicting long-term biochemical outcome or prostate cancer–specific mortality. Finally, although we looked at whether BT monotherapy treatment patterns were concordant to practice guidelines, arguably a more important endpoint is to evaluate whether treatment patterns are associated with improved outcomes such as disease control, overall survival, and lower toxicity. These endpoints were not evaluated in our study and deserve future study.
Within a large, US-based cohort, brachytherapy monotherapy for treatment of localized prostate cancer increased from the 1990-1998 to 1999-2011 periods across low, intermediate, and high-risk groups, despite 2 national guidelines published in 1999 and 2000 recommending BT monotherapy preferentially for low-risk patients and advising against it for high-risk disease. These results suggest that clinical practice guidelines did not significantly change treatment decisions, in particular for the treatment of men with high-risk disease. They also highlight the importance of having prospective, randomized data to help define practice guidelines and minimize treatment variation. However, follow-up studies are required to evaluate other potential factors that influence deviations from guidelines including patient preference and site-to-site variability.