Treatment outcomes of 125I low-dose-rate brachytherapy vs radical prostatectomy for patients with intermediate-risk prostate cancer

Background: Radical prostatectomy (RP) and low-dose-rate brachytherapy (LDR) are two widely used treatment options for patients with intermediate-risk prostate cancer (IRPC). However, which one is better remains controversial. Therefore, the purpose of this study was to compare the ecacy of RP vs LDR for patients with IRPC. Methods: A retrospective analysis was performed on 361 IRPC patients who underwent treatment from January 2010 and August 2017. 160 underwent RP and 201 underwent LDR using Iodine-125. Biochemical failure for RP was dened as prostate-specic antigen (PSA) levels > 0.2 ng/ml, and for LDR as nadir PSA level + 2 ng/ml. The log-rank test compared biochemical relapse-free survival (bRFS) between the two modalities, and Cox regression identied factors associated with bRFS. Results: Median follow-up was 54 months for RP and 69 months for LDR. The 5-and 8-year bRFS rates were 70.2% and 63.1% in the RP group, and 83.2% and 68.9% in the LDR group, respectively, P=0.003. There were no signicant differences in terms of clinical relapse-free survival (cRFS), cancer-specic survival (CSS) or overall survival (OS) between the two groups. Conclusion: LDR is a reasonable treatment option for IRPC patients, yielding improved bRFS and equivalent rates of cRFS, CSS and OS compared with RP. our of both Goy’s study. Goy's the present study, the 5- and 8-year bRFS were 70.2% 63.1% in the RP 83.2% in the LDR group, respectively. The log-rank test showed the RP using Our study


Background
Prostate cancer (Pca) is the second most common cause of cancer-related death in U.S. men in 2018 (1).
Although the incidence of Pca in China is lower than in western countries, it keeps growing in recent years due to aging and the improvements in healthcare (2). Radical prostatectomy (RP), low-dose-rate brachytherapy (LDR), and external beam radiation therapy (EBRT) are the three major de nitive treatment options for localized prostate cancer. Established clinical guidelines recommend that treatment decisions should be made based on tumor features, baseline prostate speci c antigen (PSA) levels, patient age, comorbidity, life expectancy, and quality of life (3)(4)(5). However, which of these treatments is the best for patients with localized prostate cancer remains a subject of debate.
LDR has shown impressive 10-year results (6,7). However, most of these patients were low risk, for which there is no signi cant bene t to local therapy (8). The National Comprehensive Cancer Network (NCCN) guidelines recommend local therapy for intermediate-risk prostate cancer (IRPC) (5). Randomized trials of IRPC patients treated with RP vs LDR have not been done. Thus, there is no clear evidence to inform clinical decisions regarding adequate treatment for IRPC patients. Our aim was to perform a retrospective study of IRPC, comparing the long term of outcomes of RP vs LDR at a single institution in China. We sought to identify variables that may predict differences in biochemical control, including treatment modality as a variable, using the most recent consensus de nitions of biochemical failure (BF). The following information was evaluated for all patients: medical history, physical examination, digital rectal examination and iPSA (serum PSA prior to treatment). Clinical staging was based on digital rectal examination and, when clinically indicated, chest radiography, bone scintigraphy, CT-scan and/or magnetic resonance imaging of the pelvis.

Treatments
RP was performed by a pure laparoscopic prostatectomy, with the extent of pelvic lymph node dissection being based upon the risk category of the patient. The procedure was performed according to the technique described by Walsh (10). The vesico-urethral anastomosis was made with a running suture with Y604 (Ethieon, USA).
LDR was planned so that the prostate and proximal seminal vesicles received 145 Gy with a 5-mm margin laterally, anteriorly, and inferiorly (11). No margin was planned superiorly (bladder) and posteriorly (rectum). 125 I seeds were accurately introduced into preplanned positions by a brachytherapy stepping unit MICK200 (Computerized Medical Systems, Inc, St. Louis, MO, USA) using a standard 0.5 cm brachytherapy template placed over the perineum. One week after implantation, dosimetric analysis was performed by CT scan, and the D90 (de ned as the minimum dose covering 90% of the prostate) was obtained for each patient.
3 Follow-up and study endpoints Time zero was the day of RP or LDR. Patients were followed up monthly during the rst 3 months and at 3-month intervals thereafter. If PSA level was stable, routine follow-up was scheduled every 6 months after 2 years of treatment. The treatment outcomes were assessed in terms of biochemical relapse-free survival (bRFS), clinical relapse-free survival (cRFS), and cancer-speci c survival (CSS), and overall survival (OS). BF was de ned as a PSA value ≥ 0.2 ng/mL for patients who underwent RP (12) and an increase of 2 ng/mL or > nadir PSA value (Phoenix de nition) (13) for patients receiving LDR. If a patient received salvage radiotherapy or endocrine therapy, the patient was counted as having experienced a BF.
Clinical relapse was de ned as metastases identi ed by medical imaging, with or without localizing symptoms, or as biopsy-proven local recurrence. Cancer-speci c mortality was de ned as mortality due to Pca, as noted on the death certi cate alongside the biochemical and clinical information, or the presence of uncontrolled metastatic disease at the time the patient succumbed.

Statistical analysis
Factors considered to in uence the endpoint were recorded for baseline analysis. Student's t-test was used to evaluate differences in the mean of continuous variables. A χ 2 test was performed to compare ratios and Mann-Whitney U test to compare medians. Differences between two survival curves were evaluated by log-rank tests. Cox proportional-hazard models were constructed to identify factors associated with bRFS. Statistical analyses were performed using SPSS version 25.0 (SPSS Inc., Chicago, IL, USA). P <0.05 was considered to be statistically signi cant in this study.

Patient Characteristics
A total of 361 patients were included in the present study, comprising 201 LDR patients (55.7%) and 160 RP patients (44.3%). The median follow-up for RP and LDR was 54 and 69 months, respectively. Patients treated with LDR were older, experienced a longer follow-up time and had a higher preponderance of combined ADT treatment and a lower proportion of clinical stage T2b-c. Table 1 presents a full comparison of pretreatment characteristics between the LDR and PR groups. The mean D90 for the LDR group was 144 Gy (1 standard deviation = 20.58Gy). For RP, positive margins were found in 55 (34.4%) patients. Extracapsular extension was found in 31 (19.3%), seminal vesicle invasion in 18 (11.3%), resulting in 42 (26.3%) RP patients upstaged to pathologic T3. Neoadjuvant or adjuvant androgen deprivation therapy (ADT) was given 31.2% for RP vs 94.5% for LDR, P<0.001. Supplemental EBRT were used in 15 patients (9.4%) in the RP group and 5 patients (2.5%) in the LDR group, respectively.

Main Outcomes
The 5-and 8-year bRFS rates were 70.2% and 63.1% in the RP group, and 83.2% and 68.9% in the LDR group, respectively (Figure-1A). The log-rank test indicated that bRFS rates for the RP group were lower compared with the LDR group (P=0.003, (Figure- bRFS curves between LDR and RP Comparison of bRFS curves between LDR and RP groups under the condition of different variables is shown in Table 2. Log-rank test was used to compare the bRFS curves between LDR and RP in terms of different variables according to pretreatment characteristics. Risk of BF was signi cantly higher with RP compared with LDR in the patients with GS 7 (3+4: P=0.027; 4+3: P=0.001), prostate volume >30ml (P<0.001), iPSA ≤ 10 ng/ml (P=0.007), clinical T stage with T1c-T2a(P=0.016), any PPBC(≤50%: P=0.043; >50%: P=0.003) or unfavorable IRPC(P=0.003). However, survival of LDR patients was not statistically different from RP patients at any age, GS 6, prostate volume≤30ml, iPSA > 10 ng/ml, clinical T stage with T2b-T2c or favorable IRPC.

Prognostic Factors
Cox proportional-hazard models were constructed to identify factors associated with bRFS, and results are shown in Table 3

Discussion
Currently the guidelines of the American Urological Association, the European Association of Urology and the NCCN endorse RP, LDR and EBRT therapy as appropriate treatment options for IRPC. No differences have been observed in OS or CSS among the three approaches in the recent studies (14,15). However, treatment options for Pca are diverse, and therapeutic decisions are largely based on the condition of each medical institution, the preference of doctors and the cognition of patients (14). The comparison of the oncological outcomes of RP and LDR treatments remains a challenge, due to differential de nitions for recurrence and methodological biases arising from the differences in baseline characteristics, including age, comorbidity and cancer risk features (14,(16)(17)(18). A randomized controlled trial is the ideal approach for comparing competing treatment modalities (19). However, most of the previously reported literatures comparing the e cacy of LDR and RP are retrospective studies (14,20). Compared with candidates for RP, patients who are offered LDR generally tend to be older and have higher comorbidity scores, therefore, a random trial is impractical (17,18).
We retrospectively analyzed data from 361 IRPC patients treated in our hospital. Our results showed there were no statistically signi cant differences in the cRFS, CSS and OS between the two therapeutic groups. This result was consistent with recent publications in the literature, though they were not supported by randomized prospective studies (21-23). Hamdy et al reported 10-year outcomes after monitoring, surgery, or radiotherapy for localized prostate cancer, and there were no signi cant difference in cRFS, CSS and OS between treatments of surgery and radiotherapy (15). However, his study included patients of all risk categories, and thus we cannot make any conclusions regarding IRPC. Goy et al statistically analyzed 1503 IRPC patients who underwent treatment from 2004 to 2007. There were no signi cant differences in cRFS or CSS. The 10-year bRFS was 80.2% for LDR vs 57.1% for RP vs 57.0% for EBRT, which showed improved bRFS with LDR, P = 0.0003 (24). But LDR patients had a signi cantly smaller proportion number compared with RP (7.3% Vs 54.5%) in his study, which might cause some bias in the results. In our study, the proportion of patients in both groups was more balanced than that in Goy's study. Consistent with Goy's results, in the present study, the 5-and 8-year bRFS rates were 70.2% and 63.1% in the RP group, and 83.2% and 68.9% in the LDR group, respectively. The log-rank test showed the bRFS for RP was signi cantly lower than LDR. However, in terms of the pretreatment characteristics, patients treated with LDR were older, experienced longer follow-up time, and had a higher preponderance of combined ADT treatment. In previous studies using models adjusted for risk, ADT was not shown to be an independent predictor (18,25). Our study supports the use of LDR as a reasonable treatment option for IRPC.
Neither treatment modality has proven superior to the other with respect to RP and LDR; therefore, the optimal treatment for different risk categories in Pca remains a matter of debate. Although our median follow-up time of 54 months was su cient to capture a considerable number of systemic failure events, it may still be too short to achieve mortality results. Therefore, we chose bRFS as the main evaluation criterion of curative effect. We observed that bRFS was signi cantly higher with LDR than RP in patients with GS 7, prostate volume >30ml, iPSA ≤ 10 ng/ml, clinical T stage with T1c-T2a, any PPBC or unfavorable IRPC. Supported by the results of our study, LDR may be a better option than RP in patients with the above conditions. In our experience, a subgroup analysis of unfavorable IRPC had improved bRFS with LDR compared to RP. These results showed that LDR was also a good option for those with unfavorable IRPC. Taussky et al reported RP and LDR treatment did not result in signi cantly different outcomes at four years post-treatment in patients with low-and low-intermediate-risk Pca (26). However, Ferreira et al found the 5-year bRFS of patients with early Pca who had undergone LDR was signi cantly higher than those who had undergone surgery (27). Furthermore, Ciezki et al reported high-risk Pca treated with EBRT, LDR, or RP yields e cacy with better bRFS for LDR and EBRT compared with RP (21). It should be noted that these studies were obviously heterogeneous, because different centers adopted different LDR technologies, and different methods were used to compare the results between RP and LDR (28).
As the previously studies reported, there were many factors affecting the prognosis of Pca, including general situation, tumor stage, tumor grade, iPSA, age and bone scintigraphy (29)(30)(31). Ciezki et al found clinical stage T3, GS 8 to 10, higher iPSA and more frequent PSA testing after therapy were all associated with a signi cantly worse bRFS (21). Taussky et al reported younger age, higher percentage of positive biopsies, and PSA at diagnosis were predictive of BF (26). On univariate analysis of the entire cohort in the present study, treatment with RP, age <70 years old clinical T stage with T2b-T2c, prostate volume ≤30ml and PPBC>50% were associated with signi cantly worse bRFS. Previous studies had reported that clinical stage was the most dangerous factor in uencing Pca prognosis (32,33). On multivariate analysis of the entire cohort, treatment with RP, prostate volume ≤30ml and PPBC>50% were associated with signi cantly worse bRFS. Therefore, they were important independent prognostic factors.
The limitations of this study include the following: (1) The baseline characteristics of the two groups did not completely match, which was inevitable due to random grouping in a retrospective study. Therefore, a prospective study comparing eligible patients is needed to make a more accurate conclusion. (2) The aim of our study was to provide a guide to aid clinical decision making at diagnosis. Thus, duration of ADT following initial treatment, which may contribute to survival, was not adjusted. However, in previous studies using models adjusted for risk, ADT was not shown to be an independent predictor (18,25). (3) The de nition of BF is different for the RP versus LDR groups. Although this de nition is commonly used in the world (15,23,24), it could cause some bias in interpretation of the results. (4) There were fewer deaths in this study; therefore, whether higher bRFS rates observed in patients could translate into superior oncological endpoints is still undetermined. A longer observational period is needed for a meaningful comparison of overall survival.

Conclusions
In summary, LDR, with or without androgen deprivation, is as good a treatment option for IRPC patients, producing equivalent rates of cRFS, CSS and OS compared with RP. Despite the difference in BF de nitions, LDR improved bRFS compared with RP. Treatment with RP (P=0.029), prostate volume ≤30ml (P<0.001) and PPBC>50% (P<0.001) were independent predictors for worse bRFS. A longer follow-up may be necessary to detect a difference in OS between these two treatments.

Consent for publication
Not applicable.