Background
Intermediate-risk prostate cancer represents the largest group of prostate cancers with considerable biologic and clinical heterogeneity, and is further subdivided into favorable and unfavorable intermediate-risk (UIR) groups.1 Relative to favorable intermediate-risk (FIR) disease, men with UIR disease have higher rates of biochemical recurrence, metastatic recurrence, and death from prostate cancer.2 NCCN currently recommends 3 definitive treatment options for UIR prostate cancer, including radical prostatectomy with or without pelvic lymph node dissection, external-beam radiation therapy (EBRT) with 4 to 6 months of androgen deprivation therapy (ADT), or combination EBRT with a brachytherapy (BT) boost with or without ADT. The addition of ADT to EBRT is associated with improved biochemical control and is considered standard of care,3,4 but the benefit of ADT in combination with BT is less well defined.5,6
BT is an excellent treatment for patients with localized prostate cancer and is further classified into low-dose rate BT (LDR-BT) or high-dose rate BT (HDR-BT). LDR-BT consists of delivering radiation at a rate of <2 Gy/h via permanent implantation of sealed radioactive sources or seeds into the prostate, whereas HDR-BT is defined as radiation delivery at a rate >12 Gy/h by temporarily implanting radioactive source into the prostate via catheters.7,8 In low-risk and FIR prostate cancer, BT is an option for definitive therapy.1 In unfavorable intermediate- and high-risk prostate cancer, BT is often combined with EBRT as a boost, which permits further dose escalation beyond doses that can be delivered routinely with EBRT. EBRT + BT boost has demonstrated improved biochemical control over EBRT + ADT alone in randomized clinical trials.9 However, there are limited data on the use of definitive BT in UIR prostate cancer, and NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines) for Prostate Cancer do not currently recommend definitive BT as an treatment option for UIR disease.1 Given that BT allows for significant dose-escalation relative to EBRT, we hypothesized that men with UIR prostate cancer in the National Cancer Database (NCDB) undergoing definitive BT ± ADT would have comparable overall survival (OS) to men receiving EBRT ± ADT.
Methods
Patient Cohort and Covariables
Deidentified patients from 2004 and 2015 with histologically diagnosed adenocarcinoma of the prostate were identified in the NCDB. Patients were included who met criteria for UIR prostate cancer, defined as intermediate-risk disease and either primary Gleason 4 disease, diffuse disease (≥50% biopsy cores positive), or 2 of 3 intermediate-risk factors (cT2b–T2c; prostate-specific antigen [PSA] level of 10–20 ng/mL; Gleason 3 + 4 disease). Patients with any of the following criteria were excluded: nodal or metastatic disease; missing information on clinical T stage, Gleason score, or PSA; moderate to severe comorbid disease burden, defined as a Charlson/Deyo comorbidity index (CDCI) scores >1; prior prostate surgery or pelvic radiation; receipt of chemotherapy or immunotherapy; unknown ADT status or ADT start date in relation to diagnosis; ADT initiated >180 days after diagnosis; unknown information on the cumulative EBRT radiation dose, number of fractions, or numbers of days after diagnosis when RT was initiated; or RT initiated >180 days after diagnosis. Selection of the final patient cohort is summarized in supplemental eFigure 1 (available with this article at JNCCN.org). RT treatment included EBRT or definitive BT. EBRT was delivered with conventional fractionation (≥72 Gy in 1.8–2.0 Gy per fraction). Among patients who received BT, treatment included LDR-BT and HDR-BT. Given that the allowed EBRT dose/fractionation in this study was consistent with NCCN Guidelines, we wanted to make sure that the patients who received BT were also treated in a manner consistent with these guidelines. Therefore, we excluded those receiving single-fraction HDR-BT because this treatment is not listed as an acceptable option per NCCN Guidelines and was recently shown to be associated with worse biochemical control.1,10,11 Covariables were selected a priori and included age, race, ethnicity, year of diagnosis, CDCI score, insurance status, educational attainment within the patient’s area of residence (divided into quartiles based on residents in the patient’s postal code who did not graduate high school), median income quartiles (divided into quartiles based on residents in the patient’s postal code), treatment at an academic center, PSA level at diagnosis, Gleason score, and clinical T stage.
Endpoints and Statistical Analysis
OS was the primary endpoint. Chi-square and Student t tests were used to detect significant differences among categorical and continuous variables, respectively. OS was estimated using the Kaplan-Meier method and log-rank tests were used to compare treatment arms. Multivariable analysis (MVA) using Cox regression modeling was used to compare OS hazard ratios (HRs). The inverse probability of treatment weighting (IPTW) was used to adjust for covariable imbalance. The probability of receiving a treatment was estimated using a binomial logistic regression model to generate propensity scores that included age, race, ethnicity, CDCI score, PSA level, clinical T stage, Gleason score, year of diagnosis, treatment at an academic center, insurance status, educational attainment within the patient’s area of residence, and median household income within the patient’s area of residence. Unstabilized inverse propensity weights were generated, with truncation of the most extreme weights as previously described (α=0.0001),10 and a pseudo-sample population in which measured baseline covariables were balanced among treatment groups was generated. Acceptable covariable balance among treatment groups was verified using the standardized mean difference, with a standardized mean difference <0.1 (10%) considered a negligible difference among treatment populations.12 The standardized mean difference between BT and EBRT treatment groups for each covariable after adjustment was as follows: age (0.03), race (0.009), Hispanic ethnicity (0.009), CDCI (0.003), treatment at an academic center (0.03), insurance status (0.06), educational attainment (0.01), median income (0.001), PSA level at diagnosis (0.02), clinical T stage (0.004), Gleason score (0.004), period of diagnosis (0.07), and ADT treatment (0.006). MVA was performed with weights applied to the time-dependent Cox proportional hazard model to compare the effects of BT versus EBRT alone. IPTW-weighted Kaplan-Meier curves were generated as the weighted product limit estimator. All analysis was performed in R software (R Foundation for Statistical Computing). All tests were 2-sided, and P values <.05 were considered statistically significant. Data analyses were perfomed using R version 4.0.2.
Results
The study cohort included 31,783 patients with UIR prostate cancer treated with EBRT (n=12,985), EBRT + ADT (n=12,960), BT alone (n=4,535), or BT + ADT (n=1,303). Radiation dose and fractionation regimens for EBRT and BT are summarized in supplemental eTable 1. Among patients receiving BT, 3,559 (61%) received LDR-BT, 432 (7%) received HDR-BT, and 1,847 (32%) received BT not otherwise specified. Baseline characteristics among men stratified by treatment with EBRT or BT are summarized in Table 1. The average age of men receiving EBRT versus BT was 70 and 68 years, respectively. The group treated with EBRT had a slightly higher proportion of men with a CDCI score of 0 relative to the BT arms.
Baseline Patient Characteristics
Unadjusted MVA of each treatment group is summarized in Table 2. Relative to patients treated with EBRT alone, improved OS was associated with EBRT + ADT (HR, 0.92; 95% CI, 0.87–0.97; P =.002), BT alone (HR, 0.90; 95% CI, 0.83–0.98; P =.01), and BT + ADT (HR, 0.78; 95% CI, 0.69–0.88; P =.00006). BT was associated with better OS relative to EBRT (P<.0001; Figure 1A). In men treated in the absence of hormones, BT correlated with improved OS relative to EBRT (HR, 0.92; 95% CI, 0.84–0.99; P =.03). At 10-year follow-up, 55% of men receiving EBRT and 60% of those receiving BT were alive (P <.0001; Figure 1B). In men treated with ADT, BT correlated with improved OS relative to EBRT (HR, 0.84; 95% CI, 0.75–0.95; P =.004). At 10-year follow-up, 56% of men receiving EBRT and 63% of those receiving BT were alive (P =.0002; Figure 1C).
Unweighted Kaplan-Meier curves showing overall survival stratified by treatment with either definitive BT or EBRT for (A) all men with UIR prostate cancer, (B) the subgroup of men treated without ADT, and (C) the subgroup of men treated with ADT. (D) Inverse probability of treatment weighting-adjusted Kaplan-Meier curves stratified by treatment with either BT or EBRT for all men with UIR prostate cancer.
Abbreviations: ADT, androgen deprivation therapy; BT, brachytherapy; EBRT, external-beam radiotherapy; UIR, unfavorable intermediate-risk.
Citation: Journal of the National Comprehensive Cancer Network 20, 4; 10.6004/jnccn.2021.7061
Unweighted Kaplan-Meier curves showing overall survival stratified by treatment with either definitive BT or EBRT for (A) all men with UIR prostate cancer, (B) the subgroup of men treated without ADT, and (C) the subgroup of men treated with ADT. (D) Inverse probability of treatment weighting-adjusted Kaplan-Meier curves stratified by treatment with either BT or EBRT for all men with UIR prostate cancer.
Abbreviations: ADT, androgen deprivation therapy; BT, brachytherapy; EBRT, external-beam radiotherapy; UIR, unfavorable intermediate-risk.
Citation: Journal of the National Comprehensive Cancer Network 20, 4; 10.6004/jnccn.2021.7061
Unweighted Kaplan-Meier curves showing overall survival stratified by treatment with either definitive BT or EBRT for (A) all men with UIR prostate cancer, (B) the subgroup of men treated without ADT, and (C) the subgroup of men treated with ADT. (D) Inverse probability of treatment weighting-adjusted Kaplan-Meier curves stratified by treatment with either BT or EBRT for all men with UIR prostate cancer.
Abbreviations: ADT, androgen deprivation therapy; BT, brachytherapy; EBRT, external-beam radiotherapy; UIR, unfavorable intermediate-risk.
Citation: Journal of the National Comprehensive Cancer Network 20, 4; 10.6004/jnccn.2021.7061
Unweighted Multivariable Cox Regression Analysis of Overall Survival Associated With EBRT + BT, BT Alone, and BT + ADT Relative to EBRT Alone
Given the significant potential selection bias for EBRT versus BT, IPTW was used to balance covariables that influenced both treatment allocation and outcomes, and weight-adjusted Cox regression was used to determine the average treatment effect of definitive BT on groups with balanced cofounders. Relative to EBRT, definitive BT was associated with improved OS on weight-adjusted Kaplan-Meier analysis (Figure 1D; P<.0001) and weight-adjusted MVA (HR, 0.91; 95% CI, 0.84–0.98; P =.009) (Table 3). The addition of ADT was associated with a comparable improvement in OS (HR, 0.90; 95% CI, 0.83–0.97; P =.004). Advanced age (HR, 1.05; 95% CI, 1.05–1.06; P<2 × 10−16), higher CDCI score (1.35; 95% CI, 1.22–1.50; P =4.7 × 10−9), and higher PSA level (HR, 1.03; 95% CI, 1.02–1.04; P = 9.1 × 10−7) correlated with reduced OS, with higher clinical stage trending toward significantly reduced OS (HR, 1.09; 95% CI, 0.99–1.20; P =.08).
IPTW-Adjusted Multivariable Cox Regression Analysis of Average Treatment Effect of Definitive BT vs EBRT
We next compared whether adding ADT improves survival in patients undergoing BT. Relative to BT alone, BT + ADT was associated with improved OS on propensity-weighted MVA (HR, 0.84; 95% CI, 0.73–0.96; P =.01) (supplemental eTable 2). Additional analysis was performed to address potential confounders. Because OS can be confounded by death from other causes, we restricted analysis to patients without any reported underlying comorbidities (CDCI score = 0), and propensity-weighted MVA showed that both ADT (HR, 0.90; 95% CI, 0.83–0.97; P =.009) and BT (HR, 0.92; 95% CI, 0.85–0.99; P =.04) remained associated with significantly improved OS (supplemental eTable 3).
Discussion
In this study of men with UIR prostate cancer, definitive BT was associated with a statistically significant improvement in OS compared with EBRT. A survival benefit was observed for BT versus EBRT alone and BT + ADT versus EBRT + ADT, after adjusting for measured confounders. We observed the following trends among the 4 treatment groups. Men treated with EBRT without ADT had higher mortality rates relative to all other treatment groups. Relative to men treated with EBRT without ADT, those receiving BT without ADT or receiving EBRT + ADT had better survival rates. The best outcomes were seen in men receiving BT + ADT. IPTW analysis showed that ADT and BT were independently associated with improved OS, suggesting that treatment with both may be associated with the best outcomes. To our knowledge, this is the first large retrospective study comparing single-modality EBRT versus BT in men with UIR prostate cancer, and this study adds to the existing literature on the comparative effectiveness of definitive BT for UIR disease.
EBRT is the principal radiation treatment modality used for definitive treatment of localized prostate cancer. In men with UIR prostate cancer, 30% will eventually develop recurrent disease, and EBRT treatment is often intensified by adding ADT and/or dose-escalating with a BT boost. Prior studies have overwhelmingly shown that adding 4 to 6 months of ADT to EBRT is associated with reduced biochemical recurrence, and have even shown a cancer-specific survival benefit.3,4,13 The ASCENDE-RT trial, which randomly assigned men with UIR and high-risk disease to dose-escalated EBRT alone or EBRT + BT boost, showed a biochemical-free survival benefit in men receiving dose-escalation with BT boost.9 Although this study used a mixed population of UIR and high-risk patients, a recent retrospective study showed that EBRT + BT boost improved biochemical recurrence-free survival in men with UIR disease.14 Consistent with these prior studies, NCCN currently recommends that men with UIR disease be treated with either EBRT and short-term ADT or EBRT + BT boost with or without ADT.1
Although BT monotherapy is an accepted treatment option for men with FIR disease in the NCCN Guidelines, NCCN does not currently recommend BT monotherapy for men with UIR disease.1 This recommendation is based on limited data. Although there are ample data on the use of a BT boost in combination with EBRT for UIR disease, there are fewer data on definitive BT for this patient cohort, because the distinction between FIR and UIR disease is a relatively recent convention. Prior retrospective studies comparing treatment-related outcomes of men with FIR and UIR prostate cancer treated with BT monotherapy have arrived at different conclusions. In Berlin et al,15 outcomes after BT monotherapy were evaluated in 258 patients with intermediate-risk disease, which failed to show a statistical difference in biochemical recurrence or distant metastases between FIR and UIR patients. In contrast, a recent retrospective analysis of 1,200 men with intermediate-risk prostate cancer showed that LDR-BT among UIR disease had a much higher risk of biochemical failure and distant metastasis compared with FIR disease.16 Similar to patients treated with EBRT, this would suggest that men treated with BT may benefit from treatment escalation with short-term ADT or EBRT. However, studies addressing treatment escalation in patients receiving BT are sparse and limited to men with primarily FIR disease, in whom they have shown no benefit to treatment escalation with ADT or EBRT.16–18 In contrast, our study shows that the addition of ADT to BT monotherapy is associated with a survival benefit.
Although studies have reported increased recurrence risk in men with UIR, few studies have compared definitive BT with EBRT in the UIR cohort. In this large retrospective study, we show that BT achieved comparable survival outcomes to EBRT + ADT. The addition of ADT to BT monotherapy provided a further survival advantage over BT alone. Given that EBRT + ADT is currently recommended for treatment of UIR, our results provide additional data to support the use of definitive BT with or without ADT for the treatment of UIR disease, because patients receiving BT monotherapy did as well or better than those receiving EBRT + ADT. This finding is consistent with prior biologic studies suggesting a benefit to dose-escalation with BT monotherapy in prostate cancer. Interestingly, 34% of men received prophylactic pelvic lymph node RT in the EBRT arms, whereas no men received nodal RT in the BT monotherapy arms. Thus, men treated with BT monotherapy appeared to do better, despite the fact that 34% of men receiving EBRT were treated with larger radiation fields that could reduce the risk of nodal relapse.
Our study has several limitations, including unmeasured confounders and selection biases inherent to retrospective studies. Information on ADT duration, BT dose, and BT dosimetry are not recorded in the NCDB. In addition, information on the type of BT (LDR-BT vs HDR-BT) and total fractions was missing for a proportion of patients. Although the comorbidity index was equally balanced between groups, CDCI scores are a surrogate for performance status and not a true measure of overall performance. However, prior surgical studies have reported that adding ASA/ECOG performance scores to models already containing CDCI scores to address confounding yielded no improvements in risk adjustment models for comparative assessment of cancer outcomes.19 MVA and IPTW were used to adjust for measured confounders; however, these techniques cannot address unmeasured confounders (ie, performance status, BT dose, and plan quality), which may have influenced treatment decision or treatment outcomes. Last, our analysis is limited by the fact that NCDB does not code for prostate-specific mortality or recurrence, which represent better endpoints for assessing the relationship between treatment and the natural disease course.
Conclusions
This large retrospective analysis of >30,000 men demonstrates that definitive BT is associated with improved OS compared with EBRT in men with UIR prostate cancer. Given that EBRT + ADT is a recommended treatment option for men with UIR prostate cancer,1 and BT + ADT achieved better results than EBRT + ADT, our results provide evidence that argues in favor of incorporating definitive BT + ADT into the NCCN Guidelines. These results should be considered hypothesis-generating and need further validation in randomized clinical trials. NRG GU-010 is a soon-to-open randomized cooperative group trial for men with UIR that will allow definitive BT, but does not randomize or stratify by EBRT versus BT to answer this question directly. Definitive BT is already an appealing option for patients because it is a highly conformal treatment with the lowest integral dose to normal tissues and is associated with less erectile dysfunction,9,17 greater patient convenience relative to standard or moderately hypofractionated RT, and is more cost-effective compared with some EBRT dose/fractionation schemes.20 Our finding of similar or better efficacy for BT relative to EBRT further adds to the treatment’s appeal. Although our results require validation in randomized trials, definitive BT represents a viable alternative to EBRT with or without ADT for men with UIR disease.
References
- 2.↑
Zumsteg ZS, Spratt DE, Pei I, et al. A new risk classification system for therapeutic decision making with intermediate-risk prostate cancer patients undergoing dose-escalated external-beam radiation therapy. Eur Urol 2013;64:895–902.
- 3.↑
D’Amico AV, Chen MH, Renshaw AA, et al. Androgen suppression and radiation vs radiation alone for prostate cancer: a randomized trial. JAMA 2008;299:289–295.
- 4.↑
Jones CU, Hunt D, McGowan DG, et al. Radiotherapy and short-term androgen deprivation for localized prostate cancer. N Engl J Med 2011;365:107–118.
- 5.↑
Pickles T, Morris WJ, Keyes M. High-intermediate prostate cancer treated with low-dose-rate brachytherapy with or without androgen deprivation therapy. Brachytherapy 2017;16:1101–1105.
- 6.↑
Pickles T, Tyldesley S, Hamm J, et al. Brachytherapy for intermediate-risk prostate cancer, androgen deprivation, and the risk of death. Int J Radiat Oncol Biol Phys 2018;100:45–52.
- 7.↑
Davis BJ, Horwitz EM, Lee WR, et al. American Brachytherapy Society consensus guidelines for transrectal ultrasound-guided permanent prostate brachytherapy. Brachytherapy 2012;11:6–19.
- 8.↑
Yamada Y, Rogers L, Demanes DJ, et al. American Brachytherapy Society consensus guidelines for high-dose-rate prostate brachytherapy. Brachytherapy 2012;11:20–32.
- 9.↑
Morris WJ, Tyldesley S, Rodda S, et al. Androgen Suppression Combined with Elective Nodal and Dose Escalated Radiation Therapy (the ASCENDE-RT Trial). Int J Radiat Oncol Biol Phys 2017;98:275–285.
- 10.↑
Cole SR, Hernán MA. Constructing inverse probability weights for marginal structural models. Am J Epidemiol 2008;168:656–664.
- 11.↑
Siddiqui ZA, Gustafson GS, Ye H, et al. Five-year outcomes of a single-institution prospective trial of 19-Gy single-fraction high-dose-rate brachytherapy for low- and intermediate-risk prostate cancer. Int J Radiat Oncol Biol Phys 2019;104:1038–1044.
- 12.↑
Austin PC, Stuart EA. Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies. Stat Med 2015;34:3661–3679.
- 13.↑
Bolla M, Maingon P, Carrie C, et al. Short androgen suppression and radiation dose escalation for intermediate- and high-risk localized prostate cancer: results of EORTC trial 22991. J Clin Oncol 2016;34:1748–1756.
- 14.↑
Abugharib AE, Dess RT, Soni PD, et al. External beam radiation therapy with or without low-dose-rate brachytherapy: analysis of favorable and unfavorable intermediate-risk prostate cancer patients. Brachytherapy 2017;16:782–789.
- 15.↑
Berlin A, Moraes FY, Sanmamed N, et al. International multicenter validation of an intermediate risk subclassification of prostate cancer managed with radical treatment without hormone therapy. J Urol 2019;201:284–291.
- 16.↑
Tom MC, Reddy CA, Smile TD, et al. Validation of the NCCN prostate cancer favorable- and unfavorable-intermediate risk groups among men treated with I-125 low dose rate brachytherapy monotherapy. Brachytherapy 2020;19:43–50.
- 17.↑
Bruner DW, Moughan J, Prestidge BR, et al. Patient reported outcomes of NRG Oncology/RTOG 0232. Int J Radiat Oncol Biol Phys 2018;102:S2–3.
- 18.↑
Keyes M, Merrick G, Frank SJ, et al. American Brachytherapy Society Task Group Report: use of androgen deprivation therapy with prostate brachytherapy—a systematic literature review. Brachytherapy 2017;16:245–265.
- 19.↑
Dobbins TA, Badgery-Parker T, Currow DC, et al. Assessing measures of comorbidity and functional status for risk adjustment to compare hospital performance for colorectal cancer surgery. BMC Med Inform Decis Mak 2015;15:55.
- 20.↑
Shah C, Lanni TB Jr, Ghilezan MI, et al. Brachytherapy provides comparable outcomes and improved cost-effectiveness in the treatment of low/intermediate prostate cancer. Brachytherapy 2012;11:441–445.
