Effect of Androgen Deprivation Therapy on Other-Cause of Mortality in Elderly Patients with Clinically Localized Prostate Cancer Treated with Modern Radiotherapy: Is There a Negative Impact?

The influence of androgen deprivation therapy (ADT) on other-cause of mortality (OCM) was investigated in patients with localized prostate cancer treated with modern high-dose radiotherapy. A retrospective review was conducted on 1125 patients with localized prostate cancer treated with high-dose radiotherapy, including image-guided, intensity-modulated radiotherapy or brachytherapy with a median follow-up of 80.7 months. Overall survival rate was no different between ADT (+) and ADT (−) group in high-, intermediate-, and low-risk groups. OCM was found in 71 patients, consisting of 4% (10/258) in the ADT (−) group and 7% (61/858) in the ADT (+) group (p = 0.0422). The 10-year OCM-free survival rate (OCMFS), if divided by the duration of ADT (ADT naïve (ADT (−)), ADT <2-year, and ADT ≥2-year groups), showed statistical significance, and was 90.7%, 88.2%, and 78.6% (p = 0.0039) for the ADT (−), ADT <2-year, and ADT ≥2-year groups, respectively. In patients aged ≥75 years, 10-year OCMFS for ADT (−), ADT <2-, and ADT ≥2-year groups was 93.5% (at 115.6 months), 85.6%, and 60.7% (p = 0.0189), respectively, whereas it was 90.7%, 89.9%, and 89.0% (p = 0.4716), respectively, in their younger counterparts. In localized prostate cancer patients, treatment with longer ADT for ≥2 years potentially increases the risk of OCM, especially in patients aged ≥75 years.


Introduction
Prostate cancer is one of the most frequently diagnosed cancers in men in developed countries [1,2]. Its standard treatment options are radical prostatectomy, external beam radiotherapy (EBRT), and interstitial brachytherapy (BT) [3]. Androgen deprivation therapy (ADT) is another important treatment, and a number of well-designed prospective multicenter trials have confirmed the value of ADT in addition to the standard radiation therapy [4,5]. Although modern radiotherapy routinely delivers higher biological equivalent doses (BEDs) in excess of 74 Gy to the prostate, decisions on the use of additional ADT in high-BED modern radiotherapy have proved inconclusive since a randomized trial to confirm the role of ADT therapy was performed with up to 70 Gy of EBRT [3,6]. Accordingly, guidelines recommended additional ADT in high-risk patients; however [3], limited evidence supports these recommendations of higher BED use in modern radiotherapy (including image-guided, intensity-modulated radiotherapy (IG-IMRT) and BT) [7]. Furthermore, ADT has several untoward effects, such as cardiovascular toxicity and diabetes; therefore, fragile patients with multiple comorbidities underwent DT with meticulous caution [8,9] since it is reported to be a potential influencing factor of other-cause mortality (OCM) caused by factors other than prostate cancer [8,9]. As a patient database comprising >1000 BT and IG-IMRT patients had been established for over a decade [10], the merit of ADT was examined in the modern radiotherapy era. In addition, the influence of ADT on the aged population was also investigated because elderly patients risked having a fragile health status with multiple comorbidities [9,10]. Further, this study aimed to examine the role of ADT in men who had been diagnosed with localized prostate cancer and were treated with modern high BED radiotherapy, with a focus on the age factor.

Patients
Data of 1125 patients who were treated with modern, high-BED radiotherapy between 1995 and 2013 were retrospectively examined. The eligibility criteria were patients who had been treated with high-dose-rate brachytherapy (HDR-BT) monotherapy or low-dose-rate brachytherapy (LDR-BT) with or without EBRT or IG-IMRT with a curative intent; a clinical stage T1-T4 and N0, M0 with histology-proven adenocarcinoma; availability and accessibility of data on determining the National Comprehensive Cancer Network (NCCN) risk classification (pretreatment prostate-specific antigen (initial PSA = iPSA) level, Gleason score sum (GS), and/or T classification); and 1-year minimum follow-up for surviving patients or until death. Among the 1128 included patients, three were excluded due to a lack of follow-up after <1 year or because of missing data. Thus, the final study included 1125 patients as subjects.
A total of 867 ADT (+) patients with stage T1-T4, N0, M0 prostate cancer were treated with radiotherapy as compared to 258 ADT (−) patients. Patients' clinical characteristics are shown in Table 1. Patients were staged according to the NCCN 2015 risk classification as follows: low risk: T1-T2a and GS 2-6 and iPSA level <0 ng/mL; intermediate risk: T2b-T2c or GS 7 or PSA level 10-20 ng/mL; and high risk: T3a-T4 or GS 8-10 or PSA level >20 ng/mL [3]. PSA failure was defined using the Phoenix definition (nadir, +2 ng/mL) or as the start of salvage hormonal therapy. The Common Terminology Criteria for Adverse Events version 4.0 Toxicity was used for toxicity analysis. All patients provided informed written consent. This study was conducted in accordance with the Declaration of Helsinki, and Institutional Review Board permission was obtained from each institution. BT contained low-dose-rate interstitial BT (LDR-BT) with or without external beam radiotherapy (EBRT) and high-dose-rate interstitial BT (HDR-BT) monotherapy.
Low-Dose-Rate Interstitial BT (LDR-BT) with or without EBRT The implant technique has been previously described in detail elsewhere [11,12]. All patients underwent transrectal ultrasound (TRUS) preplanning 3-4 weeks before implantation to determine the number of seeds. Moreover, intraoperative permanent I-125 implantation (The OncoSeed model 6711; General Electric Healthcare, Barrington, IL, USA) was performed using modified peripheral loading method. Our prescription dose for the clinical target volume (prostate) was 145 Gy (ADT (−) only) or 110 Gy (LDR-BT with EBRT). Inter-Plan version 3.4 (Elekta, Stockholm, Sweden) was used as the treatment planning system. A combination therapy was administered for ≤T3 or a GS ≤8, or for a GS of 7 (4 + 3) cases (not for GS of 7 (3 + 4) cases). Our prescription dose for the clinical target volume (CTV) (prostate) was 145 Gy (LDR-BT only) or 110 Gy (LDR-BT with EBRT) (2005-2013). The ADT was used in patients for >6 months before and/or immediately after brachytherapy [11]. In men with prostate volume >40 cc, ADT was selected to reduce the prostate volume.

High-Dose-Rate Interstitial Brachytherapy (HDR-BT) Monotherapy
The detailed technique used has been previously described elsewhere [13,14]. Briefly, a simple, radiography-based treatment planning was used from 1995 to 2007, and the prescription dose point was positioned 5 mm away from a source in the central plane. This method was denoted two-dimensional planning and was used as the initial treatment. The dose point was then shifted from two-to three-dimensional (3D) planning, and the remaining patients were treated with this (i.e., computed tomography (CT)-based planning). For 3D planning, the D90 and D95 or more were used to evaluate adequate coverage of the planning target volume. The CT-based planning with or without magnetic resonance imaging (MRI) assistance was performed by computer optimization (Nucletron an Elekta Company, Veenendaal, The Netherlands, PLATO ® and Oncentra ® brachy, Elekta AB, Stockholm, Sweden) with or without manual modification. The primarily prescribed dose was 45.5 Gy per 7 fractions, 54 Gy per 9 fractions for 5 days, 49 Gy per 7 fractions, and others (36-38 Gy per 4 fractions) [13,14]. The treatment machine used was the microSelectron-HDR ® (Nucletron an Elekta Company, Veenendaal, The Netherlands). For ADT administration, nearly all patients first received both neoadjuvant and adjuvant ADT for two or three years or a lifetime [14]. Second, the duration of ADT for intermediate-risk patients was shortened to 6-12 months, and ADT was administered mainly as neoadjuvant therapy. Third, all patients underwent a total of 12 months of ADT. For our most recent protocol, patients with only one intermediate-risk feature were not administered ADT, whereas all the others underwent 6 months as neoadjuvant but no adjuvant ADT [14].

IG-IMRT
We used helical tomotherapy for IG-IMRT; the details have been described elsewhere [15,16]. Briefly, CT with a slice thickness of 2 mm in a supine position and MRI data (T1w and T2w) were employed in precise radiotherapy planning. The CTV was defined as the prostate and proximal seminal vesicles or prostate only in the low-risk group (Damico's classification: stage, T1c; Gleason score <7; and PSA level <10 ng/mL). We started IG-IMRT using a 2.2 Gy fraction schedule with D95 (95% of pllaning target volume received at least the prescribed dose) of 74.

Statistical Analysis
StatView 5.0 statistical software was used for statistical analyses. Percentages were analyzed using a chi-square test, and student's t-test was used for normally distributed data. The Mann-Whitney U-test for skewed data was used to compare means or medians. The Kaplan-Meier method was used to analyze survival, and comparisons were made using the log-rank test. p < 0.05 was considered as statistically significant.

Patients' Characteristics
The median follow-up for the entire cohort was 80.7 (ranging from 5 to 241) months, with a minimum of 1 year in surviving patients or until death.
A comparison of the background characteristics between ADT (+) and ADT (−) group is shown in Table 1. ADT (+) patients included those with advanced disease (higher T category, higher initial PSA level, higher Gleason score sum, and higher risk group in NCCN risk classification), and also underwent a higher proportion of IG-IMRT than BT. Elderly patients underwent more ADT with borderline significance.
Bold values indicate statistically significance, ADT = androgen deprivation therapy, NA; not available.
OCM was found in 71 patients, consisting of 4% (10/258) in ADT (−) group and 7% (61/858) in ADT (+) group (p = 0.0422) ( Table 3). Table 3 shows the association between OCM and influential factors. Age, advanced disease (higher T category, higher initial PSA level, higher risk group in NCCN classification), ADT use, and modalities used were the significant influencing factors for OCM.

OS and OCM between ADT (−) and ADT (+) Patients According to Age
A comparison of background patient characteristics between elderly patients and their younger counterparts is displayed in Table 5. Elderly patients developed advanced diseases and required more ADT treatment by external beam radiotherapy. It is natural that patients aged ≥75 years showed poorer OS than their younger counterparts (10-year OS rate = 88.9% vs. 82.8%, p < 0.0001). This is further enhanced when comparing the OS in three ADT groups divided by the duration of ADT (ADT naïve, ADT <2 years and ADT ≥2 years). The 10-year OS rate (at 115.6 months) were 93.5%, 85.6%, and 58.2% (p = 0.005) for ADT (−), ADT <2 years, and ADT ≥2 years groups, respectively, in the elderly population. In contrast, no statistically significant difference was found in the younger counterpart, showing 90.7%, 89.2%, and 84.5%, respectively (p = 0.1744). Bold values indicate statistically significance, NA; not available, ADT = androgen deprivation therapy, HDR-BT; high-dose-rate brachytherapy, LDR-BT; low-dose-rate brachytherapy, iPSA = initial PSA, IG-IMRT; image guided intensity modulated radiotherapy, BT = HDR-BT + LDR-BT, NCCN = The National Comprehensive Cancer Network, OCM = other cause of mortality.
The 10-year OCMFS were 89.8% and 90.7% for the ADT (−) and ADT (+) groups (p = 0.5688), respectively, in the younger population, whereas they were 93.5% at 115.6 months and 81.5% for the ADT (−) and ADT (+) groups (p = 0.2451), respectively, in the elderly population ( Figure 2). These differences were enhanced by dividing the analysis into three ADT groups (ADT (−), ADT <2-year, and ADT ≥2-year groups), which were 93.5% (at 115.6 months), 85.6%, and 60.7% (p = 0.0189) in the elderly group, and 90.7%, 89.9%, and 89.0% (p = 0.4716) for ADT (−), ADT <2-year, and ADT ≥2-year groups in their younger counterpart, respectively.  Table 6 depicts the cause of OCM. Other cancer forms are a major cause of OCM. In the ADT (−) group, no cardiovascular death was recorded, whereas 0.4% and 1.5% mortality (due to cardiovascular death) was recorded in the ADT <2-year and ADT ≥2-year groups, respectively, although no statistically significant difference was observed. Unknown cases included three cases of sudden death that did not exclude cardiovascular death.   Table 6 depicts the cause of OCM. Other cancer forms are a major cause of OCM. In the ADT (−) group, no cardiovascular death was recorded, whereas 0.4% and 1.5% mortality (due to cardiovascular death) was recorded in the ADT <2-year and ADT ≥2-year groups, respectively, although no statistically significant difference was observed. Unknown cases included three cases of sudden death that did not exclude cardiovascular death.

Discussion
We presented here that ADT did not always improve outcomes after high-BED radiotherapy for localized prostate cancer patients. In fact, long-term ADT ≥2-year may have a negative impact on OCM, especially in elderly patients aged ≥75.
ADT has played an important role in the management of prostate cancer. In the early 1940s, Huggins and Hodges [17,18] established how castration arrested the growth of prostate cancer cells and suppressed serum prostate phosphatases in metastatic prostate cancer cells. Following several randomized controlled trials, simultaneous radiotherapy with ADT has been established as a standard treatment for high-risk prostate cancer with radiotherapy of up to 70 Gy [3,4]. Bolla et al. reported that in patients with locally advanced disease, the use of goserelin simultaneously with external beam radiotherapy improved the 5-year OS in comparison to external beam RT alone (79% vs. 62%, p = 0.001) [5].
ADT is usually used for advanced prostate cancer, including locally advanced and metastatic cancers. However, in Japan, ADT is likely to be accepted even for localized disease [10], because Japanese patients consider mild to moderate toxicity (i.e., sexual dysfunction) acceptable as long as they received curative treatment for lethal cancer, which may reflect the social and philosophical situation in Asian countries [19]. However, the use of primary ADT in localized disease is not a recommended treatment in guidelines [3] because of its possible harmful effect and the lack of benefits for survival [20].
Dose escalation in radiotherapy improved outcomes in several randomized controlled trials and high-BED radiotherapy is recognized as the standard treatment for localized prostate cancer [3,6]. In EBRT, IG-IMRT is now the standard treatment of choice because of its superior dose distribution, which makes it possible to administer high-dose radiotherapy to the lesion without elevating toxicity [3,6]. BT (LDR-BT and HDR-BT) as a "boost" or as monotherapy has also been incorporated as the standard radiotherapy for these excellent dose distributions [21]. These modalities combined with ADT are regarded as the standard treatments for locally advanced and/or intermediate-to high-risk disease; however, there is no high-level evidence that supports higher-dose radiotherapy of ≥74 Gy and/or BT. In addition, the optimum duration of ADT with higher dose RT is yet to be determined [22].
Adverse effects of ADT include decreased bone mineral density; metabolic changes such as weight gain; decreased muscle mass and increased insulin resistance; decreased libido and sexual dysfunction; hot flashes; gynecomastia; reduced testicle size; anemia; and fatigue [4,8]. Several observational studies suggest an increased risk of diabetes and cardiovascular events (myocardial infarction (MI) and sudden cardiac death) [23], although most published studies reported that ADT is not linked to greater cardiovascular mortality, and a meta-analysis of randomized trials in men assigned to ADT vs. no ADT did not record much/excessive discrepancy in cardiovascular mortality between the groups [8]. Studies have also suggested that there may be other harmful effects of ADT, including cerebrovascular diseases [24], kidney injury [25], thromboembolic events [26], and diabetes [23], which may all contribute to excess OCM. ADT significantly increases fat mass and fasting insulin levels and decreases insulin sensitivity [8,23]. Treatment-related changes in serum lipoproteins and arterial stiffness, as well as possible QT interval prolongation, may also contribute to the association between ADT and cardiovascular and cerebrovascular outcomes [23,24].
Unexpectedly, we did not find that ADT had a beneficial effect on survival; furthermore, a negative impact was observed on OCM in the elderly population. In addition, we found several previous retrospective/population-based studies concerning the negative impact of ADT [27][28][29].
Beyer et al. reported that a 10-year OS rate decreased from 44% of hormone naïve case into 20% with ADT [27] in prostate cancer patients who were receiving LDR-BT, with the leading causes of death being cardiovascular, prostate, and other cancers with no obvious discrepancy between the two groups. Abdollah et al. reported that treatment with medical ADT may increase the risk of OCM in 137,524 patients with non-metastatic prostate cancer who were treated between 1995 and 2009; this study was extracted from the Surveillance Epidemiology and End Results Medicare-linked database [28]. Nanda reported that neoadjuvant ADT is significantly associated with an increased risk of all-cause mortality among men with a history of coronary artery disease (CAD)-induced chronic heret falure or MI, but not among men with no comorbidity or a single CAD risk factor, which exceeds death after the LDR-BT of using ADT [29].
Our findings are in line with these findings, and the strength of this study is that we examined the role of ADT on OCM in patients treated with modern high BED radiotherapy focused on age. Morgans et al. speculated that the risk of incident diabetes mellitus (DM) or cardiovascular disease in men exposed to prolonged ADT ≥2 years increases with age at diagnosis and occurred 5-10 years later [30]. They reported that younger men are not at increased risk for incident DM or cardiovascular disease even when treated with ADT ≥2, whereas older men exposed to prolonged ADT are at an increased risk of these illnesses. This finding supports our result that DM and/or cardiovascular disease could increase OCM risk. Recently, the outcomes of prostate cancer treatment has improved and reached nearly a 100% survival rate; therefore, simultaneous importance of OCM has increased. The use of ADT in elderly patients should be performed with meticulous care.
This study has several limitations. First, as it was a retrospective study carried out in a few institutes dealing with a rather small number of patients, a longer follow-up with larger number of patients is needed before making concrete conclusions since there is a chance of an inherent bias due to inhomogeneity remaining. Next, comorbidity analysis is lacking. Body mass index, adrenal dysfunction, metabolic syndrome, and comorbidities, such as DM and cardiovascular disease, are confirmed as important influencing factors for OCM [8,[23][24][25]. In addition, lack of serum testosterone measurements is a problem since many of the older men will have a prolonged recovery time after two years of ADT. Lastly, the reason for OCM could not be specified; it is not always the cardiovascular system, i.e., mainly other cancers. This retrospective study may influence the uncorrected bias. At present, no single likely explanation can be provided for the excess deaths. Therefore, our results did not reduce the importance of ADT usage in situations to improve survival in randomized clinical trials.
In conclusion, treatment with ADT is correlated with the risk of mortality due to causes other than prostate cancer, especially in localized prostate cancer patients aged ≥75 years. Whether this is a simple association or a cause-and-effect relationship is unknown and warrants further prospective studies.