Personalized radiotherapy benefit prediction in non-surgically managed prostate adenocarcinoma: a prognostic nomogram for survival risk stratification

Introduction

Prostate adenocarcinoma (PCa) represents the most frequently diagnosed malignancy affecting the male genitourinary tract. Based on global cancer data released by the World Health Organization (WHO) in 2020, an estimated 1.4 million new cases were reported, as the second leading cancer among men, surpassed only by lung cancer (1). Treatment decisions for low- and high-risk PCa patients require balancing tumor control with functional recovery. The National Comprehensive Cancer Network (NCCN) guidelines endorse radical prostatectomy (RP) and radiotherapy as primary options (2).

Radiotherapy, a non-invasive therapeutic modality, has become the preferred option for many patients (3), with approximately 30% undergoing this treatment annually (4). Evidence suggests that radiotherapy yields prognostic outcomes comparable to or surpassing RP in selected PCa patients (5,6).

Additionally, evidence from randomised trials shows that radiotherapy, particularly when combined with androgen deprivation therapy (ADT), achieves durable disease control with survival outcomes comparable to surgery. At the same time, focal therapies such as high-intensity focused ultrasound and cryotherapy are being investigated in carefully selected patients with low-risk disease. Although early studies suggest functional preservation and acceptable short-term cancer control, long-term survival data remain sparse and these approaches are not considered standard of care. Against this backdrop, radiotherapy continues to be recommended as a curative treatment for men with intermediate-risk prostate cancer.

Nonetheless, treatment must be tailored: active surveillance is recommended for low-risk patients with longer life expectancy, whereas palliative care suits elderly patients or those with severe comorbidities to prevent overtreatment. Thus, identifying patient subgroups benefiting from radiotherapy is clinically critical.

Nomograms, as visual tools integrating multiple predictive variables, are widely employed in urology and oncology (7,8). This study was designed to develop a prognostic nomogram for risk stratification, with the objective of identifying subsets of patients with prostate cancer who are most likely to derive survival benefit from radiotherapy, to support more informed clinical decision-making. We present this article in accordance with the TRIPOD reporting checklist (available at https://tau.amegroups.com/article/view/10.21037/tau-2025-434/rc).

Methods

Data source and study population

This retrospective cohort study was based on data from the Surveillance, Epidemiology, and End Results (SEER) program of the U.S. National Cancer Institute (November 2022 submission, covering 17 population-based cancer registries from 2000 to 2020). Records of patients diagnosed with prostate cancer between 2004 and 2015 were extracted using SEER*Stat software (version 8.4.3).

Patients with histologically confirmed adenocarcinoma of the prostate (ICD-O-3 code 8140) diagnosed between 2004 and 2015 were eligible. We included eligible patients in whom prostate cancer was recorded as the first primary malignancy (ICD-O-3/WHO 2008 site code C61.9) and no surgical treatment had been given. Patients were excluded if information on age, race, Gleason score, tumor-node-metastasis (TNM) stage, or prostate-specific antigen (PSA) was missing, if histological confirmation was lacking, if cause of death or survival status was unavailable in SEER, or if survival was less than 6 months. There are 100,155 patients included in the final analysis (Figure S1).

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Variable selection

The variables analyzed in this study encompassed demographic characteristics, tumor features, and treatment information. Demographic characteristics included age (18–65 and >65 years), race (categorized as Black, White, or other), and marital status (married or unmarried). Tumor features comprised TNM staging [based on the American Joint Committee on Cancer (AJCC) 6th edition, classified as T1–T4, N0–N1, and M0–M1], Gleason score (<7, 7, and >7), and PSA levels (<4, 4–<10, 10–<20, and ≥20 ng/mL). Treatment information included chemotherapy (received or not received) and radiotherapy (received or not received). The primary endpoints were overall survival (OS) and cancer-specific survival (CSS), defined respectively as the interval from diagnosis to death from any cause and the duration from diagnosis to death attributable to PCa.

Risk stratification

Individual risk scores were calculated for each patient based on the nomogram-derived total points. Restricted cubic spline (RCS) analysis was applied to examine the association between risk score and hazard ratio (HR) for survival. The optimal cutoff point was identified by the spline curve, which maximized separation of survival risk. Patients with scores lower than the cutoff point were classified as the low-risk group, and those with scores over the cutoff point as the high-risk group.

Statistical analysis

Continuous variables conforming to a normal distribution were reported as mean ± standard deviation and analyzed across groups with the independent t-test. Variables exhibiting non-normal distribution were described as median with interquartile range (IQR) and evaluated using the Wilcoxon rank-sum test. Categorical data were expressed as frequencies (percentages) and assessed for differences between groups via the Chi-squared test.

All-subsets regression was used to identify key prognostic variables, selected on the basis of the highest adjusted R². These variables were then entered into multivariable Cox regression models to confirm their independent association with outcomes and to construct a nomogram for survival prediction. Model discrimination, accuracy, and clinical benefit were evaluated using the concordance index (C-index), calibration curves, and decision curve analysis (DCA), respectively. Internal validity was assessed with time-dependent C-indices derived from 200 iterations of 5-fold cross-validation. Survival estimates for OS and CSS were generated using the Kaplan-Meier method, with comparisons between groups assessed by the log-rank test.

To assess the robustness of our findings, a sensitivity analysis was performed. Considering that the inclusion of patients with metastatic disease (M1) could introduce heterogeneity in treatment allocation and prognosis, all analyses were repeated in PCa with non-metastatic disease (M0).

Statistical analyses were conducted using R software (version 4.3.2), with all tests being two-sided and a significance level of α<0.05.

Results

Patient characteristics

This study included 100,155 non-surgically treated PCa patients, with a median age of 67.0 years (IQR, 61.0–73.0 years). Clinical staging revealed that 70.0% (n=70,076) were at T1 stage, 97.0% (n=97,149) at N0 stage, and 94.8% (n=94,955) at M0 stage. Regarding pathological characteristics, 59.5% of patients (n=59,611) had PSA levels of 4–<10 ng/mL, and 43.0% (n=43,086) had a Gleason score <7. In terms of treatment, 1.0% (n=1,001) received chemotherapy, and 61.8% (n=61,914) underwent radiotherapy. Patients receiving radiotherapy were predominantly older (>65 years), White, married, and diagnosed with T1/N0/M0 stage (Table 1). Survival analysis, with a median follow-up of 87.0 months (IQR, 67–109 months), demonstrated 5-year OS and CSS rates of 86.6% [95% confidence interval (CI): 86.4–86.8%] and 94.7% (95% CI: 94.5–94.8%), respectively (Figure 1A,1B).

Figure 1 OS (A) and CSS (B) of PCa patients without surgical treatment. CI, confidence interval; CSS, cancer-specific survival; OS, overall survival; PCa, prostate adenocarcinoma.

Prognostic factor analysis and nomogram construction for OS and CSS

Using all-subsets regression analysis, we identified influential factors for OS and CSS. For OS, the optimal model included age, race, marital status, M stage, PSA level, and Gleason score (adjusted R²=0.16). For CSS, the optimal model comprised T stage, N stage, M stage, PSA level, and Gleason score (adjusted R²=0.28) (Figure S2). Multivariable Cox regression analysis confirmed these variables as independent prognostic factors (Tables S1,S2). Based on these findings, we developed predictive nomograms for OS and CSS (Figure 2A,2B). RCS analysis determined optimal risk stratification cutoffs: OS score >108 (5-year survival rate: 78.9%) and CSS score >47 (5-year survival rate: 86.0%) were classified as high-risk groups (Figure 2C,2D). Model validation revealed C-index values of 0.733 (95% CI: 0.731–0.734) for OS and 0.874 (95% CI: 0.871–0.876) for CSS, with calibration curves confirming high concordance between predicted and observed outcomes (Figure 3A,3B). Decision curve analysis demonstrated that the nomograms provided substantial net clinical benefit across a wide range of threshold probabilities compared to “treat all” or “treat none” strategies (Figure 3C,3D). Additionally, 200 iterations of 5-fold cross-validation confirmed stable predictive performance over 1–5 years, with median C-index ranges of 0.75–0.78 for OS and 0.89–0.92 for CSS (Figure 3E,3F).

Figure 2 Development of OS and CSS prognostic nomograms (A,B) and risk stratification based on RCS analysis (C,D). CI, confidence interval; CSS, cancer-specific survival; HR, hazard ratio; M, metastasis; N, node; OS, overall survival; PSA, prostate-specific antigen; RCS, restricted cubic spline; T, tumor.

Figure 3 Calibration plots (A,B), decision curve analysis (C,D), and 200 times 5-fold cross-internal validation of the C-indexes for the OS (E) and CSS (F) nomograms. C-index, Concordance index; CSS, cancer-specific survival; OS, overall survival.

Survival analyses

Survival analysis revealed that radiotherapy significantly improved OS (5-year OS: 82.8% vs. 89.0%, HR =0.74, P<0.001) and CSS (5-year CSS: 91.3% vs. 96.7%, HR =0.46, P=0.08) in the whole cohort (Table 2, Figure 4A,4B). Subgroup analysis further showed that radiotherapy conferred significant OS (66.1% vs. 84.5%, HR =0.49, P<0.001) and CSS (77.6% vs. 95.0%, HR =0.25, P<0.001) benefits in the high-risk group. However, in the low-risk group, radiotherapy did not improve CSS (99.4% vs. 99.4%, HR =0.93, P=0.41) and was associated with a slight reduction in OS (94.7% vs. 94.6%, HR =1.07, P=0.01) (Table 2, Figure 5A-5D).

Figure 4 Kaplan-Meier curves showing the benefits of radiotherapy for OS (A) and CSS (B) in PCa patients. CSS, cancer-specific survival; OS, overall survival; PCa, prostate adenocarcinoma.

Figure 5 OS and CSS benefits of radiotherapy in PCa patients in the low- and high-risk groups. (A,B) OS in low- and high-risk patients; (C,D) CSS in low- and high-risk patients. CSS, cancer-specific survival; OS, overall survival; PCa, prostate adenocarcinoma.

Among patients with non-metastatic disease (n=94,955), radiotherapy was associated with improved outcomes overall. In the whole M0 cohort, radiotherapy was linked to a 22% reduction in cancer-specific mortality (CSS: HR 0.78, 95% CI: 0.74–0.83, P<0.001) and a more modest 7% reduction in overall mortality (OS: HR 0.93, 95% CI: 0.90–0.96, P<0.001). When stratified by risk group, in the low-risk subgroup, radiotherapy conferred no advantage in CSS (HR 1.45, 95% CI: 0.11–18.9, P=0.80), and OS was slightly worse among those treated (HR 1.08, 95% CI: 1.02–1.14, P=0.006). In contrast, high-risk patients derived substantial benefit. Radiotherapy was associated with a 23% reduction in cancer-specific mortality (CSS: HR 0.77, 95% CI: 0.73–0.82, P<0.001) and a 38% reduction in overall mortality (OS HR 0.62, 95% CI: 0.60–0.64, P<0.001). In the M0-only cohort, models performance are similar, with a C-index of 0.80 (95% CI: 0.79–0.81) for CSS and 0.69 (95% CI: 0.69–0.69) for OS. The full set of results for the M0 subgroup, including nomogram performance and stratified treatment effects, are presented in the Supplementary file (Table S3, Figures S3-S6).

Discussion

According to the 2025 American Cancer Statistics, PCa accounted for 313,780 new cases, representing approximately 30% of all new cancer diagnoses in the United States, with 35,770 associated deaths. This disease has emerged as the second leading cause of cancer-related mortality among men (9). Localized PCa is primarily treated with RP or radiotherapy. However, existing clinical trials predominantly focus on surgical options and prognostic outcomes in RP patients (10,11), leaving a gap in systematic research on PCa patients suitable for radiotherapy and their prognostic characteristics. To address this, our study utilizes large-scale retrospective cohort data from the 2004–2015 SEER database to construct a risk stratification model. This model aims to objectively evaluate the survival benefits of radiotherapy and accurately identify specific patient subgroups most likely to benefit from it, providing evidence-based support for individualized clinical treatment decisions.

Firstly, multivariate analysis in this study identified age, race, marital status, M stage, PSA level, and Gleason score as independent prognostic factors for OS in non-surgical PCa patients. In contrast, T stage, N stage, M stage, PSA level, and Gleason score significantly influenced CSS. Specifically, older patients exhibited poorer prognosis, aligning with the clinical consensus that age is a critical prognostic indicator (12,13). Racial disparities revealed black patients with the worst prognosis, followed by white patients, while other races demonstrated the best outcomes, consistent with previous reports (14,15). Notably, married patients showed a significant survival advantage over unmarried individuals (including single, separated, divorced, and widowed), potentially attributable to systemic spousal support and neuroendocrine homeostasis sustained by a regular lifestyle (16). Stage at diagnosis is an important part of interpreting these findings. Only a small proportion of patients with metastatic disease were included, and their presence did not change the overall pattern of results. The model is therefore most informative for patients with localized disease, while still reflecting the mix of treatment approaches seen in large population datasets.

In PCa patients, elevated serum PSA levels typically indicate heightened risks of tumor progression, recurrence, or metastasis (17,18), explaining the poorer prognosis observed among high-PSA patients in this study. The Gleason score, a key measure of prostate cancer differentiation, demonstrated that higher scores correlate with poorer differentiation, increased invasiveness, and worse prognosis—a significant negative correlation widely validated by prior studies and corroborated by our findings (19,20). Additionally, TNM staging analysis revealed that patients with extensive primary tumor invasion or distant metastasis face a markedly elevated mortality risk, consistent with international cancer staging prognostic standards (21).

Subsequently, we developed risk stratification models based on key prognostic factors. Survival analysis results showed that radiotherapy significantly improved survival for high-risk patients, while low-risk patients did not benefit from radiotherapy due to the indolent nature of their tumors (5-year OS of 94.6% and CSS of 99.2%). This finding suggests that although radiotherapy is an effective curative treatment, considering its treatment-related toxicities (including gastrointestinal reactions such as rectal pain, rectal bleeding, enteritis, and urinary system complications like urethral stricture, hematuria, urinary incontinence), the benefit-risk ratio of radiotherapy for low-risk patients may not be advantageous. These findings align with prior studies. Martinez et al. (22) confirmed that high-risk patients treated with external beam radiotherapy (EBRT) combined with high-dose-rate brachytherapy (HDR-BT) significantly improved the 10-year biochemical recurrence rate and clinical failure rate, and the prognosis was positively correlated with the radiotherapy dose. Yaxley et al.’s cohort study of 507 intermediate- and high-risk PCa patients showed that the high-risk group had a 5-year biochemical failure-free survival rate of 74.2%, and 10-year CSS and OS rates of 90.8% and 86.7%, respectively (23). In addition, two large clinical trials, PROTECT (24) and PIVOT (25), both showed that there was no statistical difference in long-term survival outcomes between active surveillance and radical treatment for low-risk patients. However, Galalae et al.’s study obtained different results from this study, reporting that low-risk patients had better 5-year biochemical control rates after radiotherapy (87.2% by ASTRO definition, 88.2% by Phoenix definition) than high-risk patients (26). This discrepancy may stem from methodological differences: this study included more clinical and pathological parameters based on a large sample size and used nomogram scoring for risk stratification, while Galalae et al. stratified based on T stage, initial PSA level, and tumor grade; in addition, due to data missingness, this study only analyzed the overall efficacy of radiotherapy, while Galalae et al. specifically evaluated the treatment effect of HDR-BT. These differences may lead to heterogeneity in risk population classification and efficacy evaluation results.

Thus, clinical decision-making requires careful integration of histological features, biochemical markers, clinical stage, and other relevant parameters to guide individualized treatment strategies. The nomogram developed in this study quantifies the contribution of each prognostic factor, enabling more precise survival risk prediction. Compared with previous work, our approach places stronger emphasis on individualized risk assessment and provides an objective basis for radiotherapy decision-making. However, while these findings suggest potential translational value, the model should be interpreted alongside established guideline-based risk systems, and its role will need confirmation through external validation before being applied in routine clinical practice.

This study has several limitations that should be acknowledged. First, its retrospective design and reliance on registry data introduce the potential for selection bias and unmeasured confounding. Although the SEER database provides robust demographic and tumor-related variables, it does not capture important clinical parameters such as comorbidity status, performance status, ADT use, or details of radiotherapy dose, fractionation, and technique. These unmeasured factors may have influenced treatment allocation and survival outcomes. Second, although internal validation was performed using cross-validation, external validation with an independent dataset was not feasible and should be pursued in future studies to strengthen generalizability. Third, treatment comparisons were limited to radiotherapy versus non-radiotherapy; surgical cohorts were excluded, and therefore the nomogram cannot be directly compared with outcomes after RP. Fourth, we were unable to benchmark our model against standard guideline-based systems such as the NCCN risk groups, since SEER lacks several key variables required for their definition, including biopsy core involvement, PSA density, and more granular staging distinctions (e.g., T3a vs. T3b). Finally, follow-up duration, while adequate for estimating medium-term outcomes, may be insufficient to fully capture long-term survival differences, particularly among low-risk patients with indolent disease. Despite these limitations, the large sample size, rigorous statistical methodology, and consistency of sensitivity analyses support the reliability and clinical relevance of our findings.

Conclusions

This study developed a prognostic nomogram for non-surgically managed prostate cancer using multiple clinicopathological parameters. The model showed good predictive performance and clinical utility, allowing accurate risk stratification to identify patients most likely to benefit from radiotherapy. Our findings indicate that radiotherapy confers significant survival benefit in high-risk patients, whereas low-risk patients derive little clinical advantage. These results provide an evidence-based foundation for more personalized treatment strategies in prostate cancer and highlight the value of incorporating risk stratification into therapeutic decision-making.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://tau.amegroups.com/article/view/10.21037/tau-2025-434/rc

Peer Review File: Available at https://tau.amegroups.com/article/view/10.21037/tau-2025-434/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tau.amegroups.com/article/view/10.21037/tau-2025-434/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021;71:209-49. [Crossref] [PubMed]
2. Schaeffer E, Srinivas S, Antonarakis ES, et al. NCCN Guidelines Insights: Prostate Cancer, Version 1.2021. J Natl Compr Canc Netw 2021;19:134-43. [Crossref] [PubMed]
3. Holmes JA, Wang AZ, Hoffman KE, et al. Is primary prostate cancer treatment influenced by likelihood of extraprostatic disease? A surveillance, epidemiology and end results patterns of care study. Int J Radiat Oncol Biol Phys 2012;84:88-94. [Crossref] [PubMed]
4. Cancer Research UK. Prostate cancer statistics. Available online: https://www.cancerresearchuk.org/health-professional/cancer-statistics/statistics-by-cancer-type/prostate-cancer (accessed June 14, 2025).
5. Guo B, Liu M. Evaluation of Cancer Specific Mortality with Surgery versus Radiation as Primary Therapy for Localized High Grade Prostate Cancer in Men Younger Than 60 Years. J Urol 2021;205:1530-1. Letter. [Crossref] [PubMed]
6. Cozzi G, Musi G, Bianchi R, et al. Meta-analysis of studies comparing oncologic outcomes of radical prostatectomy and brachytherapy for localized prostate cancer. Ther Adv Urol 2017;9:241-50. [Crossref] [PubMed]
7. Yang C, Liu Z, Fang Y, et al. Development and validation of a clinic machine-learning nomogram for the prediction of risk stratifications of prostate cancer based on functional subsets of peripheral lymphocyte. J Transl Med 2023;21:465. [Crossref] [PubMed]
8. Iasonos A, Schrag D, Raj GV, et al. How to build and interpret a nomogram for cancer prognosis. J Clin Oncol 2008;26:1364-70. [Crossref] [PubMed]
9. Siegel RL, Kratzer TB, Giaquinto AN, et al. Cancer statistics, 2025. CA Cancer J Clin 2025;75:10-45. [Crossref] [PubMed]
10. Robert G, Blin P, Bladou F, et al. Comparative effectiveness of robot-assisted vs. open prostatectomy: a real-life nationwide study. World J Urol 2025;43:367. [Crossref] [PubMed]
11. Eastham JA, Heller G, Halabi S, et al. Cancer and Leukemia Group B 90203 (Alliance): Radical Prostatectomy With or Without Neoadjuvant Chemohormonal Therapy in Localized, High-Risk Prostate Cancer. J Clin Oncol 2020;38:3042-50. [Crossref] [PubMed]
12. Eguchi T, Bains S, Lee MC, et al. Impact of Increasing Age on Cause-Specific Mortality and Morbidity in Patients With Stage I Non-Small-Cell Lung Cancer: A Competing Risks Analysis. J Clin Oncol 2017;35:281-90. [Crossref] [PubMed]
13. Wray CJ, Phatak UR, Robinson EK, et al. The effect of age on race-related breast cancer survival disparities. Ann Surg Oncol 2013;20:2541-7. [Crossref] [PubMed]
14. Murphy A, Cottrell-Daniels CC, Awasthi S, et al. Understanding and Addressing Prostate Cancer Disparities in Diagnosis, Treatment, and Outcomes Among Black Men. Cancer Control 2024;31:10732748241275389. [Crossref] [PubMed]
15. Zhang Z, Cai Q, Wang J, et al. Development and validation of a nomogram to predict cancer-specific survival in nonsurgically treated elderly patients with prostate cancer. Sci Rep 2023;13:17719. [Crossref] [PubMed]
16. Liberles SD. Mammalian pheromones. Annu Rev Physiol 2014;76:151-75. [Crossref] [PubMed]
17. Van Poppel H, Albreht T, Basu P, et al. Serum PSA-based early detection of prostate cancer in Europe and globally: past, present and future. Nat Rev Urol 2022;19:562-72. [Crossref] [PubMed]
18. Corres-Mendizabal J, Zacchi F, Martín-Martín N, et al. Metastatic hormone-naïve prostate cancer: a distinct biological entity. Trends Cancer 2024;10:825-41. [Crossref] [PubMed]
19. Ham WS, Chalfin HJ, Feng Z, et al. New Prostate Cancer Grading System Predicts Long-term Survival Following Surgery for Gleason Score 8-10 Prostate Cancer. Eur Urol 2017;71:907-12. [Crossref] [PubMed]
20. Chaoying L, Chao M, Xiangrui Y, et al. Risk factors of bone metastasis in patients with newly diagnosed prostate cancer. Eur Rev Med Pharmacol Sci 2022;26:391-8. [Crossref] [PubMed]
21. van der Poel H, van der Kwast T, Aben K, et al. Imaging and T Category for Prostate Cancer in the 8th Edition of the Union for International Cancer Control TNM Classification. Eur Urol Oncol 2020;3:563-4.
22. Martinez AA, Gonzalez J, Ye H, et al. Dose escalation improves cancer-related events at 10 years for intermediate- and high-risk prostate cancer patients treated with hypofractionated high-dose-rate boost and external beam radiotherapy. Int J Radiat Oncol Biol Phys 2011;79:363-70. [Crossref] [PubMed]
23. Yaxley JW, Lah K, Yaxley JP, et al. Long-term outcomes of high-dose-rate brachytherapy for intermediate- and high-risk prostate cancer with a median follow-up of 10 years. BJU Int 2017;120:56-60. [Crossref] [PubMed]
24. Hamdy FC, Donovan JL, Lane JA, et al. 10-Year Outcomes after Monitoring, Surgery, or Radiotherapy for Localized Prostate Cancer. N Engl J Med 2016;375:1415-24. [Crossref] [PubMed]
25. Wilt TJ, Brawer MK. The Prostate Cancer Intervention Versus Observation Trial: a randomized trial comparing radical prostatectomy versus expectant management for the treatment of clinically localized prostate cancer. J Urol 1994;152:1910-4. [Crossref] [PubMed]
26. Galalae RM, Zakikhany NH, Geiger F, et al. The 15-year outcomes of high-dose-rate brachytherapy for radical dose escalation in patients with prostate cancer - a benchmark for high-tech external beam radiotherapy alone? Brachytherapy 2014;13:117-22. [Crossref] [PubMed]