Conventional dose versus dose escalated radiotherapy including high-dose-rate brachytherapy boost for patients with Gleason score 9–10 clinical localized prostate cancer

As several recent researches focus on the importance of Gleason 9–10, we examine the role of radiotherapy dose escalation in those patients. We analyzed 476 patients with Gleason score 9–10 prostate cancer treated with radiotherapy. Of them, 127 patients were treated with conventional-dose external beam radiotherapy (Conv RT) and 349 patients were treated with high-dose radiotherapy (HDRT; 249 patients received high-dose-rate brachytherapy boost + external beam radiotherapy [HDR boost] and 100 patients received intensity-modulated radiotherapy [IMRT]). We compared these treatment groups using multi-institutional retrospective data. The patients had a median follow-up period of 66.3 months. HDRT showed superior biochemical disease-free survival (bDFS) rate (85.2%; HDR boost 84.7% and IMRT 86.6%) to Conv RT (71.1%, p < 0.0001) at 5 years, with a hazard ratio of 0.448. There were borderline difference in prostate cancer-specific mortality (PCSM; 4.3% and 2.75%, p = 0.0581), and distant metastasis-free survival (DMFS; 94.4% and 89.6%, p = 0.0916) rates at 5-years between Conv RT and HDRT group. Dose escalated radiotherapy showed better bDFS, borderline improvement in PCSM, and equivocal outcome in DMFS in with clinically localized Gleason 9–10 prostate cancer.

). The patient eligibility criteria were as follows: treatment with EBRT + BT or EBRT alone, clinical tumor-node-metastasis stage T1-T4, N0M0 with Gleason score 9-10, histology-proven adenocarcinoma, and availability and accessibility of pretreatment data (initial prostate-specific antigen [PSA] level, Gleason score sum, and T classification). We defined PSA failure according to the Phoenix definition (nadir + 2 ng/mL). The Common Terminology Criteria for Adverse Events version 4.0 was used for the toxicity analysis. Toxic effects occurring within 90 days after radiotherapy completion were considered acute, and toxic effects occurring after that 90-day period were considered late. All patients in the study by the Uji-Takeda group provided written informed consent, and patients in the public data source provided informed consent during the process of building public data. This study was conducted in accordance with the Declaration of Helsinki and received institutional review board approval (Kyoto Prefectural University of Medicine Institute, approval no. ERB-C-1403).
Treatment planning. HDR BT boost. Multi-institution data were obtained from an open data source 6 , and the detailed method of applicator implantation has been described elsewhere 8 . Table 2 shows the detailed schedules of the combination of HDR boost and EBRT. HDR boost used 31.5 Gy (10.5-31.5 Gy) and EBRT used 30 Gy (30-51 Gy) as the median dose. The median fraction size of HDR boost was 6.3 Gy (5-11 Gy), and that of EBRT was 3 Gy (2-3 Gy).
EBRT. Table 2 shows the detailed schedule of Conv RT and HDRT group, including conventional two-dimensional treatment planning, three-dimensional conformal radiotherapy planning, and IMRT planning. Some EBRT data were obtained from a freely accessible dataset (n = 155) 6 , and 72 image-guided IMRT procedures using helical tomotherapy were performed at the Department of Radiology of Uji Takeda Hospital. The technique of image-guided IMRT using helical tomotherapy has been described elsewhere 7

Results
Patient and tumor characteristics. A total of 476 patients with Gleason 9-10 (very high-risk) clinically localized prostate cancer were treated with HDR boost (n = 249) or EBRT (n = 227). The median patient age was 71 years (range 60-89 years). The clinical characteristics of the patients are presented in Table 1. The median follow-up duration of the entire cohort was 66.3 months (range 2-158 months), with a minimum of 1 year for surviving patients or until death. Table 1 compares the background patient characteristics between the Conv RT and HDRT groups. Supplemental Table 2 shows the patient characteristics among the Conv RT, HDR boost, and IMRT groups.
As shown in Table 3, the predictors of biochemical control in univariate analysis included the treatment group, T classification, and baseline PSA level. In multivariate Cox regression analysis, the HDRT group showed superior outcomes to those of the Conc RT group (hazard ratio 0.448, 95% CI 0.283-0.7081, p = 0.0006) and advanced T classification 3 ≤ showed statistically significant influence. Among three groups, both HDR boost (HR 0.382) and IMRT (HR 0.375) showed statistically significant improvement than conv RT in univariate analysis ( Table 2).
Late toxicity. Table 4 shows comparison of late toxicity between Conv and HDRT group. Equivocal gastrointestinal toxicity and higher genitourinary toxicity were found in HDRT group. In detailed analysis (Supplemental Table 2), HDR boost showed highest genitourinary toxicity and lowest gastrointestinal toxicity.

Discussion
Donald Gleason proposed a grading system for prostate cancer half a century ago, and the Gleason scoring system still has diagnostic importance or may even have a more central role at present 10 . Gleason identified five histological patterns (from the most well differentiated [Gleason pattern 1] to the least differentiated [Gleason pattern 5]), and this system, when combined with stage, has been shown to be prognostic for OS 10 . Many trials have confirmed the importance of this grading system, and several recent studies have focused on the importance of Gleason score 9-10 1-5 . It is already well established that dose escalation improves bDFS. Many randomized controlled trials and meta-analysis studies have demonstrated the superiority of treatment with increased prescribed dose for localized prostate cancer [11][12][13][14] . Pollack et al. confirmed the superiority of the 78-Gy dose to the 70-Gy dose (the bDFS rates for the 70-and 78-Gy arms at 6 years were 64% and 70%, respectively) 13 . According to these notions, the Comprehensive Cancer Network Clinical Practice Guidelines in Oncology (2019) stated that a dose of 70 Gy in conventional fractions is not appropriate for patients with localized prostate cancer 1 . Therefore, we compared the outcomes of Conv RT using a prescribed dose of 70-72 Gy with higher-dose EBRT with IMRT with a prescribed dose of ≥ 74 Gy and HDR boost. In this study, we presented evidence that dose escalation, including HDR boost and IMRT, improves the biochemical control rate even in Gleason 9-10 prostate cancer based on a population of > 400 patients, which may be in line with the results of previous studies for the entire high-risk group [11][12][13][14] . Our findings may be beneficial for counseling individual patients with Gleason score 9-10 prostate cancer with respect to their treatment and prognosis.
BT has several merits that enable the delivery of higher doses of radiation to the target lesion while avoiding unnecessary higher irradiation to adjacent organs at risk, and is therefore considered one of the best radiotherapy techniques 15 . Additionally, the low α/β ratio of prostate adenocarcinoma cells implies higher sensitivity to large radiation doses per fraction than most other malignancies 15,16 . Therefore, better outcomes could be expected with dose escalation using hypofractionated schedules with HDR boost 16 . A few prospective studies and several retrospective studies have reported the merits of HDR boost 1,15-17 . These trials focused on low-and intermediaterisk prostate cancer. Therefore, little prospective data have been accumulated in high-risk groups, especially in patients with very high-risk prostate cancer, such as those with Gleason score 9-10. Thus, our study could provide useful information for making daily clinical decisions for very high-risk patients. We also investigated www.nature.com/scientificreports/ HDR monotherapy and reported a 5-year bDFS rate of 91.5% in Gleason 9-10 disease (n = 48) 18 , indicating that HDR monotherapy is also a promising procedure with good outcomes compared with Conv RT. Some authors reported the superiority of HDR boost not only to Conv RT but also to high-dose EBRT (e.g., IMRT) in terms of the bDFS rate 19,20 . Spratt et al. reported the superior bDFS outcome of HDR boost in patients with intermediate-risk prostate cancer compared with high-dose IMRT alone (even at a dose of 86.4 Gy), but not in the high-risk group 20 . Furthermore, several studies observed improvement not only in terms of bDFS but also PCSM with dose escalation using BT [20][21][22] . Kishan et al. reported that EBRT + BT was associated with significantly lower PCSM rate (3%) than either radical prostatectomy (12%) or EBRT (13%) in Gleason 9-10 disease even after propensity adjustment 5 . In contrast, our data indicated that BT boost and IMRT did not translate into improved PCSM (HDR boost 3.81%, IMRT 0%). Muralidhar et al. also reported equivocal results between BT boost and radical prostatectomy in the Surveillance, Epidemiology and End Results cohort, with no difference in the 5-year PCSM (radical prostatectomy 6.0% vs. BT boost 5.7%) 23 . Although our shorter follow-up period did not allow concluding that HDRT could improve PCSM better than Conv RT, high-dose IMRT showed equivocal or superior outcome to HDR boost, which does not concur with previous data 5, [20][21][22] .
BT has been facing a slow but progressive decline over the past decades. To overcome this problem, specific strategic interventions must be carried out in the field of national guidelines, education, research, and communication with patients and colleagues of other specialties in an interdisciplinary setting 24 .
For toxicity analysis, higher dose did not always elevate toxicity in gastrointestinal tract. IMRT and HDR boost could avoid higher dose to gastrointestinal organ, resulting in non-inferior toxicity profile to Conv RT group. Hydrogel spacer also could reduce GI toxicity not only in fresh case but also for reirradiation even though with ulcerative colitis 25 . However, higher dose was inevitably irradiated to genitourinary organs including urethra  www.nature.com/scientificreports/ which is located inside of prostate, therefore higher toxicity was found in HDRT group especially in HDR boost group. Androgen deprivation therapy (ADT) plays an important role in the treatment of high-risk prostate cancer. Zapatero et al. showed an improvement in 5-year bDFS with an additional 2-year adjuvant ADT from 81 to 90% after 6-82 Gy of EBRT in high-risk patients treated with three-dimensional conformal radiotherapy as neoadjuvant therapy 26 . We used long-term ADT, which could be one of the reasons for our good outcomes compared with previous studies. Furthermore, the good efficiency of ADT has been demonstrated in Japanese men, which can be attributed to the Japanese-specific high sensitivity to hormonal therapy 27 . However, a recent meta-analysis of trials of RT and ADT suggested that patients with Gleason score 9-10 prostate cancer had the greatest benefit from lifelong ADT, whereas the optimal treatment for those with Gleason score 8 prostate cancer might be long-term (but not lifelong) ADT 28 . We also observed that longer-term ADT use by > 2 years increases the occurrence of other causes of mortality in patients aged > 75 years 29 . Meticulous patient selection should be considered to maximize the efficacy of ADT without toxicity.
The present study had several limitations. The retrospective nature of the study confers limitations related to follow-up time. Moreover, the small sample size cannot reflect the entire population of patients with prostate cancer, which may limit the application of our findings.

Conclusions
This study shows that dose-escalated radiotherapy results in improved bDFS, borderline improvement in PCSM, and equivocal outcomes in terms of DMFS in patients with Gleason 9-10 prostate cancer.

Data availability
The data of HDR-BT and part of EBRT for this manuscript can be obtained from the public data base 6 and other EBRT was can be obtained from the author upon reasonable request.