Skip to main content
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Jul 17.
Published before final editing as: Int J Radiat Oncol Biol Phys. 2026 Jul 14:S0360-3016(26)03944-1. doi: 10.1016/j.ijrobp.2026.06.3085

Quality of life results of addition of androgen deprivation therapy and pelvic lymph node treatment to prostate bed salvage radiotherapy: NRG Oncology/RTOG 0534 SPPORT

Deborah W Bruner 1, Theodore G Karrison 2,3, Alan Pollack 4, Jeff M Michalski 5, Alexander G Balogh 6, George B Rodrigues 7, Eric M Horwitz 8, Sergio L Faria 9, Andrew S Camarata 10, R Jeffrey Lee 11, Himanshu R Lukka 12, Michael Zelefsky 13, Wendy Seiferheld 2,14, Howard Sandler 15, Benjamin Movsas 16
PMCID: PMC13374713  NIHMSID: NIHMS2192195  PMID: 42448001

Summary

Background:

This is the quality of life (QOL) analysis of the SPPORT trial of men with a detectable prostate specific antigen (PSA) after prostatectomy for prostate cancer. Primary analysis established a benefit of adding pelvic lymph node radiotherapy (PLNRT) and short-term androgen deprivation therapy (STADT) to salvage prostate bed radiotherapy (PBRT) for biochemical progression-free survival.

Methods:

Subjects were randomized to Arm 1, PBRT, Arm 2, 4–6 months of STADT + PBRT, and Arm 3, PLNRT + STADT + PBRT. QOL was assessed at baseline, 6 weeks after the start of RT, 1 and 5 years after the end of RT, including Expanded Prostate Cancer Index Composite (EPIC), Hopkins Symptom Checklist (HSCL-25), EuroQol (EQ-5D), and quality adjusted life years (QALYs). Statistical and clinical significance were calculated. Differences among arms were assessed using pairwise t-tests, Fisher’s exact test, and mixed effects regression modeling, controlling for multiplicity.

Findings:

644 patients consented to QOL; 81% white and 54% <65 years. EPIC, bowel and urinary scores decreased at 6 weeks then increased by years 1 and 5, although not to baseline. For sexual and hormonal domains, there were statistically significant differences between arms at 6 weeks and 1 year with STADT exhibiting poorer QOL scores, but no difference at 5 years. For HSCL-25, statistically significant differences at 6 weeks were not clinically significant. Comparisons of QALYs for overall survival showed no statistical significance. However, freedom from progression QALY means were 5.5, 6.2 and 6.6 years for Arms 1, 2, and 3, respectively, with Arms 3 and 2 indicating statistically significant improvement over Arm 1.

Interpretation:

While QOL generally declined the most among all arms at 6 weeks post RT start, and remained lower than baseline at 5 years, there were no clinically significant differences in QOL among arms at 1 and 5 years. QALYs for freedom from progression favored STADT + PLNRT + PBRT for salvage treatment of prostate cancer.

Introduction

Some of the side effects associated with radiation therapy (RT) and androgen deprivation therapy (ADT) cancer treatments are deleterious to quality of life and may contribute to increased risks for serious health concerns associated with aging. Yet there is sparse quality-of-life data available in the setting of post-prostatectomy salvage therapy. This information is needed to provide well-defined patient-reported side effect profiles for better informing physicians and patients of the far-reaching consequences of ADT and improve the awareness that physicians should incorporate into routine practice strategies for preventing and managing toxicities. Urinary, bowel, and erectile dysfunction are well-known side effects of pelvic RT. Sexual side effects are the most recognized adverse effects from ADT and include loss of libido, erectile dysfunction, and hot flashes. Loss of libido is distressing to many men, and they may not pursue treatments for erectile dysfunction that they may have otherwise pursued after radical prostatectomy or RT. Physiologic effects of ADT, including changes in body composition (weight gain, reduced muscle mass, increase in body fat), and changes in lipids, may lead to an exacerbation of potentially more serious conditions, such as hypertension and coronary artery disease.(Li et al 2022)

Additionally, both the diagnosis of prostate cancer and chemical castration can cause psychological distress and depression (Rice et al 2018). There is robust literature indicating that testosterone plays an important role in the modulation of mood and depression (Hauger et al 2022), and incidence of suicide among men with prostate cancer, especially those treated with ADT, is higher than previously recognized (Dalela et al 2015).

Thus, almost every incremental improvement in survival or progression-free survival comes at a cost. The cost is both financial and experienced in terms of detriments in quality of life. For example, one study found that adding ADT to radiation therapy increased median patient survival by only 0.71 years. Yet, despite ADT-associated costs and long-term side effects, compared with radiation alone, radiation plus ADT was cost-effective at $127,900 per quality adjusted life year (QALY) (Nguyen et al 2019).

These quality of life issues were assessed in the NRG Oncology/RTOG 0534 SPPORT trial in men with a detectable prostate specific antigen (PSA) after prostatectomy for prostate cancer, using salvage radiotherapy and ADT. The main results were published in Lancet (Pollack et al, 2022). Briefly, patients were randomized to prostate bed radiotherapy (PBRT) alone (Arm 1), PBRT + 4–6 months of short-term androgen deprivation therapy (STADT) (Arm 2), or pelvic lymph node radiotherapy (PLNRT) + PBRT + STADT (Arm 3). The 5-year freedom from progression rates were 70.9% (95% CI 67.0–74.9) in Arm 1, 81.3% (78.0–84.6) in Arm 2, and 87.4% (84.7–90.2) in Arm 3. Both PBRT plus STADT and PLNRT plus PBRT plus short-term ADT were significantly different from PBRT alone. The addition of PLNRT to PBRT plus short-term ADT (i.e., Arm 3 vs Arm 2) resulted in a statistically significant increase in 5-year freedom from progression of 6.1%. The results established the benefit of adding STADT to PBRT and were the first randomized findings to demonstrate that extending salvage radiotherapy to treat the pelvic lymph nodes when combined with STADT results in meaningful reductions in disease progression. Toxicity, reported as acute grade 2+ and 3+ adverse events increased significantly from Arm 1 to Arm 2 to Arm 3; however, only significant late grade 2+ blood/bone marrow events were attributable to the use of PLNRT. Overall survival did not differ significantly among the groups.

The aims of this analysis are threefold, first, to assess the degree, duration, and significant differences of disease-specific health related quality-of-life decrements (urinary, bowel, sexual and hormonal domains) among treatment arms. Second, to assess whether depression decreases with the more aggressive therapy if it improves freedom from progression (FFP). Third, an exploratory aim assessed whether an incremental gain in FFP and survival with more aggressive therapy would outweigh decrements in the primary generic domains of health-related quality of life (i.e., mobility, self-care, usual activities, pain/discomfort, and anxiety/depression). This aim is reported as both Quality-Adjusted FFP Years (QAFFPY) and as overall survival Quality-Adjusted Life Years (QALY).

METHODS

Study design and participants

SPPORT was a three-arm, randomized trial designed to determine whether there are incremental gains in outcomes by adding 4–6 months of short-term androgen deprivation therapy (STADT) to PBRT (64.8–70.2 Gy) and the combination of STADT and pelvic lymph node radiotherapy (PLNRT; 45 Gy) to PBRT. The main results including randomization, masking, and treatment procedures were published in Lancet (Pollack et al, 2022). This report presents the analysis of health-related quality of life (QOL) secondary endpoints. Eligible patients for SPPORT were those who after prostatectomy for adenocarcinoma of the prostate had a persistently detectable or initially undetectable and rising PSA of between 0.1 and 2.0 ng/mL. Additionally, for this study, patients had to consent to participate in completing the QOL questionnaires.

QOL Procedures and Measures

All QOL measures were assessed as paper-and-pencil patient self-administered questionnaires at baseline (within 2 weeks of registration), during week 6 of RT, 1 year and 5 years post-RT.

Prostate Cancer-Specific QOL was measured by the Expanded Prostate Cancer Index Composite (EPIC). EPIC is a 50-item robustly validated instrument that measures a broad spectrum of urinary, bowel, sexual, and hormonal symptoms related to radiotherapy and hormonal therapy. (Wei et al 2000) Higher domain scores indicate better quality of life, which requires reverse scoring for some responses. EPIC Sexual Medications/Devices Supplemental questions were assessed to provide context to the EPIC erectile function question.

Depression was measured using the well-validated Hopkins Symptom Checklist (Derogatis et al 1974). The 25-item version of the Hopkins Symptom Checklist (HSCL-25) was used as a screening instrument for depression (Nabbe et al 2018). The HSCL-25 is scored by a four-point Likert scale (1-not at all, 2-a little, 3-quite a bit, and 4-extremely). A higher total score means a worse mood or depression. Using a cutoff of 44 and above for “caseness” of depression, Hough and colleagues (1995) found that the HSCL-25 was comparable or superior to the Center for Epidemiological Studies Depression Scale in detecting psychiatric disorder.

Quality-Adjusted Life Years (QALY) and a study-specific modification, the Quality-Adjusted Freedom From Progression Years (QAFFPY) were assessed using the EQ-5D. The EQ-5D is a method for obtaining valuations (utilities) to be used as an adjustment to survival and/or in cost-utility analysis (Schulz et al 2002). It is a two-part instrument. The first part consists of 5 items covering 5 dimensions including mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. Each dimension is graded on 3-point Likert scale: 1-no problems, 2-moderate problems, and 3-extreme problems. Health states are defined by the combination of the leveled responses to the 5 dimensions (Badia et al 1998). The second part is a visual analog scale (VAS) marked by the patient on a 0 worst to 100 best health line.

Both the 5-item index score and the VAS score are transformed into a utility score between 0 “Worst health state” and 1 “Best health state” and can be used in the quality-adjusted survival analysis depending on the health state(s) of interest. For this study we report both the multidimensional and the VAS utilities for comparative purposes between standardized QOL and current health state but only use the multidimensional utilities for the QALY and study-specific QAFFPY that combines QOL with duration of freedom from progression (FFP). The QALY has been modified in a similar manner for different diseases and treatments where survival is not the only outcome (Ontario Health 2023).

Outcomes

Statistical Analysis

Pretreatment characteristics were summarized within each treatment arm using descriptive statistics. All patients consenting to the QOL component of the study were analyzed according to the arm in which they were randomly assigned. QOL compliance was collected for all measures over all collection points. Baseline characteristics of those who were non-compliant at each time point were compared with patients who were compliant using Fisher’s exact test. For certain ordinal outcome variables, Wilcoxon’s rank-sum test was performed.

For the QOL outcomes, scores at each time point (6 weeks, 1 year, and 5 years), as well as change scores, calculated as follow-up score minus baseline score, were compared between the treatment arms using two-sample t-tests. Two pair-wise comparisons (Arm 1 vs. Arm 2 and Arm 1 vs. Arm 3) were conducted. The significance level for the pair-wise comparisons was adjusted by the Bonferroni method to α=0.05/2=0.025 to maintain the overall significance level at α =0.05. As a benchmark for clinical significance, 0.5 standard deviations (SDs) of the baseline scores (pooled over the three treatment arms) for the bowel, urinary, sexual, and hormonal domains were used at each specific time point. Changes from baseline or differences between group means that are both statistically significant and of this magnitude or greater are considered to reflect a clinically meaningful treatment effect. Note that whether the comparison is of the raw (absolute) scores or of the within-patient changes from baseline, p-values and comments regarding whether the magnitude of the difference exceeds the minimal clinically interesting difference (MCID) pertain to the treatment arm contrast (Arm 2 vs. Arm 1 or Arm 3 vs. Arm 1). The required sample size per treatment arm to detect an effect size of 0.5 SDs is 78 for 80% statistical power and 101 for 90% statistical power, respectively, based on a two-sided t-test at the α=0.025 significance level. Because our previous studies indicated the participation rate in QOL assessments would be less than 100%, the sample size for the QOL analysis was increased to n=200 subjects per arm to allow for a response rate as low as 50%.

Linear mixed models were also fit to analyze the full set of longitudinal QOL outcomes. Fixed effects were treatment arm, time (modeled with three dummy variables using baseline as the reference), and treatment-by-time interactions, as well as the four stratification factors used in the randomization (PSA level, pathology, Gleason score, and seminal vesical involvement), age, and race. Random subject effects (intercepts) were included to allow for correlation within patients. Regarding missing data, the mixed effects models are valid under a missing at random (MAR) assumption in which the missingness depends only on observed data, including values of the outcome variable and covariates. While this is a plausible assumption, sensitivity analyses were performed using multiple imputation (Rubin, 1996) and pattern mixture modelling (Hedeker and Gibbons, 1997). For multiple imputation, values of covariates and observed QOL scores were used to predict (impute) values for the missing QOL data. In the pattern mixture models, patients were grouped according to whether they had complete data at all four time points (I1=0) versus incomplete data at one or more time points (I1=1). The variable, I1, together with interaction terms between I1 and treatment arm, I1 and time, and I1, treatment arm, and time (three-way interaction) were added to the mixed-effects model to assess the impact of the missing data on the estimated treatment effects.

To examine trade-offs between the survival/freedom from progression times and QOL, we combined them for each patient into two single measurements: Quality-Adjusted Life Years (QALY) for overall survival and Quality-Adjusted FFP Years (QAFFPY). QALY and QAFFPY are defined by the weighted sum of different time episodes added up to a total quality-adjusted survival time and a total quality-adjusted FFP time, respectively, i.e., QALY=∑i=1kqisi, where qi is the quality (the utility coefficient) of health state i, si is the duration spent health state i, and k is the number of health states. The utility coefficients are the repeated measures of EQ-5D scaled to range between 0 and 1. Estimates of restricted mean QALY and QAFFPY were obtained using the inverse probability of censoring weighted (IPCW) procedure described in Zhao and Tsiatis (2000).

Findings

Between March 31, 2008, and March 22, 2012, a total of 644 patients consented to participate in QOL assessments. QOL endpoints were assessed from 182 sites in the US and Canada, with 208 patients in Arm 1, 216 in Arm 2, and 220 in Arm 3, meeting the targeted sample size for this subset analysis, causing the QOL portion of the study to close accrual 3 years prior to closure of the parent study. The Consort chart is depicted in Figure 1 and baseline characteristics of the 644 patients by treatment arm are shown in Table 1. Randomization produced comparable groups with no statistically significant differences between arms in this patient subset.

Figure 1.

Figure 1.

Consort Chart by Arm and Compliance

ADT=androgen deprivation therapy. PBRT=prostate bed radiotherapy. PLNRT=pelvic lymph node radiotherapy.

Table 1.

Patient and Tumor Characteristics for QOL Population

PBRT Alone
(n=208)
PBRT+STADT
(n=216)
PLNRT+PBRT+STADT
(n=220)
P-value
Age (Years) 0.201
 < 65 103 (49.5%) 117 (54.2%) 128 (58.2%)
 >= 65 105 (50.5%) 99 (45.8%) 92 (41.8%)
Race 0.291
 White 164 (78.8%) 174 (80.6%) 186 (84.5%)
 Other/Unknown 44 (21.2%) 42 (19.4%) 34 (15.5%)
Ethnicity 0.399
 Hispanic or Latino 8 (3.8%) 10 (4.6%) 5 (2.3%)
 Not Hispanic or Latino / Unknown 200 (96.2%) 206 (95.4%) 215 (97.7%)
Zubrod Performance Status 0.491
 0 194 (93.3%) 205 (94.9%) 211 (95.9%)
 1 14 (6.7%) 11 (5.1%) 9 (4.1%)
Pathological Tumor Stage 0.195
 T2 105 (50.5%) 121 (56.0%) 130 (59.1%)
 T3 103 (49.5%) 95 (44.0%) 90 (40.9%)
Baseline PSA (Stratification) 0.832
 >= 0.1 and <= 1.0 ng/ml 183 (88.0%) 194 (89.8%) 197 (89.5%)
 > 1.0 and < 2.0 ng/ml 25 (12.0%) 22 (10.2%) 23 (10.5%)
Gleason 0.936
 4–7 179 (86.1%) 187 (86.6%) 192 (87.3%)
 8–9 29 (13.9%) 29 (13.4%) 28 (12.7%)
Pelvic Lymphadenectomy 0.664
 No 66 (31.7%) 77 (35.6%) 77 (35.0%)
 Yes 142 (68.3%) 139 (64.4%) 143 (65.0%)

P-value from 2-sided Fisher’s exact test.

ADT=androgen deprivation therapy. PBRT=prostate bed radiotherapy. PLNRT=pelvic lymph node radiotherapy. PSA=prostate-specific antigen.

QOL Compliance

EPIC overall compliance (i.e., questionnaires completed within the required time window) was 81% at baseline, 74% for the 6 weeks after RT start, 73% at 1 year, and 45% at 5 years. Excluding patients who died prior to 5 years, the 5-year compliance rate was 47%. Approximately 25% of the patients (n=160–172, depending on the domain) were full completers, i.e., provided data at all four time points; the remaining patients had missing data at one or more visits. To evaluate the potential for bias, baseline characteristics were compared between compliant and non-compliant patients at each time point. Two statistically significant differences were found. At 6 weeks post-RT start, 84.2% of the non-compliant patients had undergone pelvic lymphadenectomy vs. 64.2% of the compliant patients (p=0.026), and at 1 year 11.3% of the non-compliant patients were Zubrod PS 1 vs. 0.9% of the compliant group (p=0.005).

Compliance to the HSCL-25 questionnaire was 75%, 69%, 69%, and 42% at baseline, 6 weeks after RT start, 1 year, and 5 years, respectively and 22% completed the instrument at all four time points. There were no statistically significant differences between compliant and non-compliant patients at any time point with one exception: Zubrod performance status at 1 year was better (score of 0) in those who were compliant versus those who were not (99% vs 90%; p=0.009).

The percentage of patients who completed the EQ-5D questionnaire was 79%, 70%, 68%, and 40% at baseline, 6 weeks, 1 year, and 5 years, respectively, with 22% completing the questionnaire at all four time points. There were no statistically significant differences between compliers and non-compliers at any time point again with the exception of completers having better Zubrod performance status at 1 year similar to the EPIC and HSCL-25.

EPIC

EPIC scores for the bowel, urinary, sexual, and hormonal domains by time point are shown in Table 2. There were no statistically significant differences at baseline among arms, as would be expected given the randomization. The benchmarks (MCID) for clinical significance of 0.5 standard deviations (SDs) of the baseline scores (pooled over the three treatment arms) are 3.9, 6.9, 11.3, and 5.4 for the four main bowel, urinary, sexual, and hormonal domains, respectively

Table 2.

EPIC domain scores at each timepoint

PBRT Alone
(n=208)
PBRT+STADT
(n=216)
PLNRT+PBRT+STADT
(n=220)
P-value & Difference
Arm 2 vs. 1
P-value & Difference
Arm 3 vs. 1
Bowel Domain
Baseline (n=190) (n=161) (n=160) 0.538 0.179
Mean (SD) 93.3 (8.1) 92.7 (8.1) 94.4 (7.0) −0.5 (−2.2 to 1.2) 1.1 (−0.5 to 2.7)
6 Weeks after RT Start (n=154) (n=152) (n=170) 0.758 0.001
Mean (SD) 83.4 (15.2) 82.9 (15.6) 77.5 (17.6) −0.5 (−4.0 to 2.9) −5.9 (−9.5 to −2.3)
1 Year after RT End (n=149) (n=156) (n=164) 0.361 0.603
Mean (SD) 90.2 (11.1) 88.9 (12.6) 89.5 (11.6) −1.2 (−3.9 to 1.4) −0.7 (−3.2 to 1.9)
5 Years after RT End (n=92) (n=94) (n=98) 0.571 0.762
Mean (SD) 89.9 (10.3) 90.8 (11.4) 90.4 (11.0) 0.9 (−2.2 to 4.1) 0.5 (−2.6 to 3.5)
Urinary Domain
Baseline (n=191) (n=160) (n=160) 0.553 0.905
Mean (SD) 83.0 (13.8) 82.1 (14.3) 83.2 (13.5) −0.9 (−3.8 to 2.1) 0.2 (−2.7 to 3.1)
6 Weeks after RT Start (n=155) (n=149) (n=171) 0.754 0.182
Mean (SD) 76.1 (18.4) 75.4 (18.2) 78.6 (15.7) −0.7 (−4.8 to 3.5) 2.5 (−1.2 to 6.2)
1 Year after RT End (n=148) (n=156) (n=162) 0.898 0.378
Mean (SD) 79.3 (15.1) 79.0 (17.5) 80.8 (15.3) −0.2 (−3.9 to 3.5) 1.5 (−1.9 to 4.9)
5 Years after RT End (n=92) (n=95) (n=96) 0.640 0.532
Mean (SD) 77.5 (16.8) 76.3 (17.8) 78.9 (15.4) −1.2 (−6.2 to 3.8) 1.5 (−3.2 to 6.1)
Sexual Domain
Baseline (n=190) (n=159) (n=162) 0.852 0.914
Mean (SD) 31.4 (23.5) 30.9 (21.4) 31.6 (22.9) −0.5 (−5.2 to 4.3) 0.3 (−4.6 to 5.2)
6 Weeks after RT Start (n=151) (n=147) (n=166) <.001 <.001
Mean (SD) 28.8 (23.4) 19.6 (17.1) 19.9 (17.2) −9.1 (−13.8 to −4.5) −8.9 (−13.4 to −4.4)
1 Year after RT End (n=146) (n=152) (n=161) 0.014 0.064
Mean (SD) 31.3 (24.5) 24.9 (19.9) 26.4 (20.9) −6.4 (−11.5 to −1.3) −4.8 (−9.9 to 0.3)
5 Years after RT End (n=90) (n=91) (n=91) 0.072 0.241
Mean (SD) 28.5 (22.3) 23.0 (19.1) 32.4 (21.5) −5.6 (−11.7 to 0.5) 3.8 (−2.6 to 10.3)
Hormonal Domain
Baseline (n=188) (n=161) (n=162) 0.689 0.505
Mean (SD) 90.1 (10.8) 90.5 (10.9) 90.8 (10,8) 0.5 (−1.8 to 2.8) 0.8 (−1.5 to 3.1)
6 Weeks after RT Start (n=149) (n=151) (n=172) <.001 <.001
Mean (SD) 87.6 (12.0) 72.9 (14.9) 74.4 (14.4) −14.7 (−17.8 to −11.7) −13.2 (−16.1 to −10.3)
1 Year after RT End (n=144) (n=156) (n=166) 0.012 0.046
Mean (SD) 88.5 (11.4) 84.7 (14.6) 85.6 (14.0) −3.8 (−6.8 to −0.8) −2.9 (−5.8 to −0.1)
5 Years after RT End (n=91) (n=91) (n=94) 0.848 0.194
Mean (SD) 87.5 (14.3) 87.9 (13.2) 90.0 (11.2) 0.4 (−3.6 to 4.4) 2.5 (−1.3 to 6.2)

ADT=androgen deprivation therapy. PBRT=prostate bed radiotherapy. PLNRT=pelvic

lymph node radiotherapy. PSA=prostate-specific antigen.

Arm 1: PBRT Alone; Arm 2: PBRT+NC-STADT; Arm 3: PLNRT+PBRT+NC-STADT

Higher scores are better

P-value from 2-sided pairwise t-tests with pooled variances.

Figure 2 presents visual plots of the means of the EPIC domains over time, while Supplementary Table 1 provides detailed summary statistics and group comparisons for all of the domains, including both raw scores and change from baseline scores, as well as “bother,” “function,” “incontinence,” and “irritative” subdomains.

Figure 2a.

Figure 2a.

EPIC Bowel Domain Score

* Statistically significant and ≥ MCID relative to PBRT Alone

Bowel Domain

For the bowel domain, scores decreased at 6 weeks post-RT start and then increased by years 1 and 5, although not quite to baseline levels. The one statistically significant treatment effect was the Arm 3 vs. Arm 1 difference at 6 weeks (Figure 2a). Arm 3 scores were 5.9 points lower in Arm 3 compared to Arm 1 at 6 weeks, and the change from baseline was more than 7-points lower compared to the change from baseline in Arm 1, where a difference of >3.9 indicates a clinically significant difference. Both bowel bother and function were statistically significantly decreased on Arm 3 but again only at 6 weeks. For the bowel domain and subdomains there were no statistically significant group differences at the 1 and 5 year time points. The results of mixed effects modeling, which included adjustment for covariates, are consistent with the analyses presented above. Estimated mean bowel domain scores decreased at 6 weeks post-RT start and then increased by years 1 and 5 toward baseline levels. The only statistically significant treatment effect occurred between Arm 3 and Arm 1 at 6 weeks, a 7.0 (SE 1.4) point greater decrease in Arm 3 (p<0.001). There were no statistically significant covariate effects. Pattern-mixture models and multiple imputation modeling for the missing data yielded similar results.

Urinary Domain

For the urinary domain (Figure 2b) and urinary irritative/obstructive subdomain (Figure 2d) scores also decreased at 6 weeks post-RT start and remained below baseline at 1 and 5 years, but there were no statistically significant differences between the treatment arms. The urinary incontinence subdomain score (Figure 2c) increased in Arm 3 at 6 weeks then declined concordant with Arms 1 and 2, and there were no statistically significant between arm differences at any timepoint. As further demonstrated in Supplementary Table 1 for urinary domain and subdomain scores, there were no statistically significant differences among arms in raw scores or change from baseline at any time point. The pattern-mixture models did detect a significant difference between Arm 2 and Arm 1 at the end of RT, but the effect was in the opposite direction of that expected (higher score/better urinary function in Arm 2). Older patients had slightly worse urinary scores in the imputation model, i.e., 2.5 points lower (SE 1.3) at any given time point (p=0.048).

Figure 2b.

Figure 2b.

EPIC Urinary Domain Score

Figure 2d.

Figure 2d.

EPIC Urinary Irritative/Obstructive Subscale Score

Figure 2c.

Figure 2c.

EPIC Urinary Incontinence Subscale Score

Hormonal Domain

For the hormonal domain (Figure 2e), there were statistically significant differences among arms at the six week and one year time points for both raw scores and changes from baseline and the magnitude of the differences exceeded the MCID at 6 weeks. At both six weeks and one year, Arm 1 without STADT had only a 2.2 and 2.1 point decline in the domain score, whereas at 6 weeks Arm 2 declined by 17.7 points and Arm 3 declined by 17.2 points (a 5.4 point change is considered clinically significant) and by 6.8 and 5.3 points, respectively, at one year. Hormonal function scores were statistically significantly decreased at 6 weeks and 1 year and for both arms with STADT (Arms 2 and 3) relative to Arm 1. Hormonal bother was statistically significantly worse only at 6 weeks and on both arms with STADT compared to Arm 1. No significant differences at 5 years were detected. Mixed effects regression results for the hormonal EPIC domain identified statistically significant detrimental treatment effects after 6 weeks and at 1 year in both Arms 2 and 3. Here, the sensitivity analyses support the 6-week findings, but the 1-year differences are no longer significant, particularly the Arm 3 vs. Arm 1 contrast.

Figure 2e.

Figure 2e.

EPIC Hormonal Domain Score

* Statistically significant and > MCID relative to PBRT Alone

Sexual Domain and EPIC Sexual Medications/Devices Supplemental Questions

For the sexual domain scores (Figure 2f), there were statistically significant differences between the treatment arms at 6 weeks in terms of both the raw scores and the changes from baseline with patients receiving STADT exhibiting poorer sexual QOL scores, and the magnitude of the differences was shy of 0.5 SDs. At six weeks, Arm 1 without STADT had only a 1.5 point decline from baseline, whereas Arm 2 and Arm 3 with STADT declined by 12.5 points and 11.8 points respectively (an 11.3 point change is considered clinically significant). Within sexual domain subdomains, the decrease from baseline in sexual function at six weeks was statistically significantly larger for both arms with STADT (Arms 2 and 3) compared to Arm 1. However, the decrease from baseline in sexual bother at 6 weeks was not statistically significantly different from Arm 1 in either arm. At one year, the raw sexual domain and function scores were statistically significantly different between Arms 2 and 1, and for sexual function the difference persisted at 5 years. The 1- and 5-year differences were not clinically significant.

Figure 2f.

Figure 2f.

EPIC Sexual Domain Score

Mixed effects findings for sexual QOL show that scores in Arm 1 changed little over time. Scores were significantly worse, however, in Arms 2 and 3 at 6 weeks post-RT start; the magnitude of the deficit was near 0.5 SDs. By years 1 and 5, the differences were no longer significant. Older patients (>=65) exhibited worse scores. In the pattern-mixture models, the 6-week Arm 2 vs. Arm 1 and Arm 3 vs. Arm 1 contrasts were a little smaller in magnitude (and the latter did not reach statistical significance), but not enough to alter the conclusions. Similarly, the results after multiple imputation were similar to those based on the MAR assumption.

EPIC Question 17b specifically asks about ability to have an erection. Among all study participants, the percentage reporting “poor to very poor to none” at baseline is 79% with no difference among arms. At 6 weeks, the percentage reporting “poor to very poor to none” is 80% in Arm 1, 88% in Arm 2 and 93% in Arm 3, and while Arms 2 and 3 (both containing STADT) are higher compared to Arm 1 (without STADT), only Arm 3 is significantly so at the 0.025 cutoff (Fisher’s exact p<0.001). Among the subset who did not use erectile medications or device aids at 6 weeks (n=75, 87, and 91 in Arms 1, 2, and 3, respectively), these percentages were similar (81%, 89%, and 95%), with the Arm 3 vs. Arm 1 difference again statistically significant (p=0.013). Among the subset who did indicate use of medications or device aids (n=74, 60, and 74, respectively), the percentage reporting erectile function as “poor to very poor to none” was 78%, 88%, and 92%, again higher in Arms 2 and 3 although not reaching statistical significance in either comparison. For all patients at 1 and 5 years after RT, 82% and 83% of men reported “poor to very poor to none” erectile function, respectively, with no statistically significant differences among arms.

Approximately 61% of patients reported using sexual medications or devices (USMD) to improve erections at baseline. This percentage dropped to 50%, 42%, and 44% in Arms 1, 2, and 3, respectively at 6 weeks post RT; 52%, 48%, and 47% at 1 year; and 52%, 34%, and 43% at 5 years. The difference between Arms 1 and 2 at 5 years was statistically significant (p=0.007). A little over half (52%) of Arm 1 patients reported trying Viagra at 5 years but found it not helpful in Arm 1 compared to 36% and 45% in Arms 2 and 3, respectively (not statistically significant). Details regarding USMD are provided in Supplementary Table 2.

Study participants were also asked about the quality of their erections without assistance of medicine or devices. At 6 weeks the percentage of participants reporting “none at all” or “not firm enough for any sexual activity” was 70% in Arm 1, 89% in Arm 2, and 91% in Arm 3. The differences between the arms with STADT vs without (between Arm 2 and Arm 1 and between Arm 3 and Arm 1) were highly significant (p<0.001 for both contrasts).

To provide greater insight into the QOL decrements, Table 3 breaks out individual EPIC domains and gives the percentage of patients reporting specific levels of distress or dysfunction. In terms of the highest percentage of patients reporting symptoms, sexual dysfunction was the greatest with 75–94% of patients consistently reporting poor erections from baseline to 5 years, with the highest number (94%) reporting poor erections at 6 weeks on Arm 3. Of note is the high percentage of patients (about 80% on all three Arms) reporting poor erections at baseline. Among the other symptoms, >10% of patients consistently reported overall urinary problems, urinary frequency and urinary incontinence across all arms at all tine points. The highest percentage of patients reporting urinary frequency was at 6 weeks on Arms 1, 2 and 3 (28%, 29% and 23% respectively) and then decreasing but never returning to baseline. Urinary leaking >1/day and pad use had >25% of patients (range 24% to 45%) consistently reporting these problems across arms at all time points, with similar percentages among arms at the various time points. For hormonal changes, hot flashes were notable in Arms 2 and 3 but only at 6 weeks with 35% and 29% respectively reporting this symptom. Lack of energy was reported consistently across arms and time points in >10% of patients, with week 6 having the highest percentage of patients reporting this symptom at 20% for Arm 1, 28% for Arm 2 and 26% for Arm 3, declining over time but not returning to baseline levels.

Table 3.

Percentage of Patients Reporting Specific Levels of Function or Bother for Specific EPIC Items

PBRT Alone PBRT+STADT PLNRT+PBRT+STADT
Domain and Questionnaire Item Baseline 6 Wk 1 Yr 5 Yr Baseline 6 Wk 1 Yr 5 Yr Baseline 6 Wk 1 Yr 5 Yr
Urinary Domain (n) (191) (155) (148) (92) (160) (149) (156) (95) (160) (171) (162) (96)
Irritation or obstruction†
 Dysuria <1 7 3 0 1 7 1 1 <1 6 2 1
 Hematuria 0 0 1 2 <1 1 <1 2 0 <1 0 1
 Weak stream 5 5 5 12 5 9 7 7 4 5 4 5
 Urinary frequency 12 28 14 18 14 29 14 20 11 23 15 14
Incontinence
 Leaking >1/day 31 28 34 35 28 30 31 31 28 24 28 37
 Frequent dribbling 11 15 18 21 14 17 17 18 9 10 13 23
 Pad use 33 35 38 45 28 32 33 44 33 26 34 45
 Leaking problem† 12 16 18 22 13 14 17 18 13 10 14 17
Overall urinary problem† 11 22 17 20 12 17 18 18 9 14 12 14
Bowel Domain† (n) (190) (154) (149) (92) (161) (152) (156) (94) (160) (170) (164) (98)
Rectal urgency 2 7 7 7 6 10 6 9 3 16 5 5
Bowel frequency 1 8 3 5 2 9 6 4 <1 22 5 3
Fecal incontinence 1 4 3 3 0 4 3 3 <1 6 2 0
Bloody stools 0 0 1 0 0 1 0 0 0 2 1 1
Rectal pain 2 6 2 3 1 7 2 0 <1 12 <1 4
Overall bowel problem 1 9 5 4 3 10 7 5 0 15 4 4
Sexual Domain (n) (190) (151) (146) (90) (159) (147) (152) (91) (162) (166) (161) (91)
Poor erections 80 80 75 85 82 89 86a 88 76 94 85b 77
Orgasm difficulty 53 63b 60 65 58a 83 69a 75 58 90a 71c 64
Erections not firm‡ 81 85 81 88 83 93 88 91 83 93 87 87
Erections not reliable‡ 74 75 74 79 78 90 86 92 74 93 83 78
Poor sexual function 76 77 75 83 77 90 82 84 75 93 86 75
Overall sexual problem† 50 47 49 41 46 44 49 45 47 45 51 34
Vitality/Hormonal Domain† (n) (188) (149) (144) (91) (161) (151) (156) (91) (162) (172) (166) (94)
Hot flashes 0 2 2 4 1 35 4 4 <1 29 6 2
Breast problems 0 1 0 3 0 1 6 1 <1 <1 4 0
Depression 7 5 4 7 8 8 4 7 7 8 8 4
Lack of energy 10 20 13 12 9 28 17 10 10 26 14 12
Weight change 4 1 5 7 2 10 8 4 2 5 3 3

The sample size was occasionally less than the n shown due to missing data for an individual question. Percentages are based on the actual number of respondents.

For all counts the discrepancy was <5, except where indicated: a5 fewer, b6 fewer, c8 fewer.

†

Responses for these items were dichotomized based on the response that the quality-of-life concern was ‘a moderate or big problem.’

‡

If a respondent reported having erections that were not ‘firm enough for intercourse,’ the response was categorized as ‘erections not firm.’ If a respondent reported not achieving erections ‘more than half the time when desired,’ the response was categorized as ‘erections not reliable.’

Tables 4 gives the percentage of patients exceeding MCID thresholds (i.e., showing “deterioration”) for the four main domains at the 6-week, 1-year, and 5-year time points. Deterioration is defined as a negative change from baseline (timepoint - baseline) with absolute value ≥ MCID. Using this definition, all domains of QOL deteriorated at 5 years in between 22 and 42% of patients across arms with little difference among arms at 5 years. There were a few significant group differences at earlier time points, i.e., bowel domain at 6 weeks (Arm 3 vs 1), sexual and hormonal domains at 6 weeks (Arms 2 and 3 vs 1), and hormonal domain at 1 year (Arm 2 vs 1).

Table 4.

EPIC Domains Deterioration Status

Tool (MCID) PBRT Alone
(n=208)
PBRT+STADT
(n=216)
PLNRT+PBRT+STADT
(n=220)
P-value
Arm 2 vs. 1
P-value
Arm 3 vs. 1
Bowel Domain (3.9)
6 Weeks after RT Start (n=147) (n=124) (n=138) 0.460 0.002
 No Deterioration 60 (40.8%) 57 (46.0%) 33 (23.9%)
 Deterioration 87 (59.2%) 67 (54.0%) 105 (76.1%)
1 Year after RT End (n=138) (n=126) (n=125) 1.000 0.203
 No Deterioration 91 (65.9%) 83 (65.9%) 72 (57.6%)
 Deterioration 47 (34.1%) 43 (34.1%) 53 (42.4%)
5 Years after RT End (n=86) (n=73) (n=74) 0.323 0.321
 No Deterioration 52 (60.5%) 50 (68.5%) 51 (68.9%)
 Deterioration 34 (39.5%) 23 (31.5%) 23 (31.1%)
Urinary Domain (6.9)
6 Weeks after RT Start (n=149) (n=121) (n=138) 0.218 0.813
 No Deterioration 81 (54.4%) 75 (62.0%) 77 (55.8%)
 Deterioration 68 (45.6%) 46 (38.0%) 61 (44.2%)
1 Year after RT End (n=136) (n=123) (n=122) 0.293 0.434
 No Deterioration 86 (63.2%) 86 (69.9%) 83 (68.0%)
 Deterioration 50 (36.8%) 37 (30.1%) 39 (32.0%)
5 Years after RT End (n=84) (n=74) (n=74) 0.873 0.422
 No Deterioration 48 (57.1%) 41 (55.4%) 47 (63.5%)
 Deterioration 36 (42.9%) 33 (44.6%) 27 (36.5%)
Sexual Domain (11.3)
6 Weeks after RT Start (n=145) (n=118) (n=137) <.001 <.001
 No Deterioration 112 (77.2%) 61 (51.7%) 78 (56.9%)
 Deterioration 33 (22.8%) 57 (48.3%) 59 (43.1%)
1 Year after RT End (n=137) (n=119) (n=123) 0.028 0.163
 No Deterioration 105 (76.6%) 76 (63.9%) 84 (68.3%)
 Deterioration 32 (23.4%) 43 (36.1%) 39 (31.7%)
5 Years after RT End (n=82) (n=70) (n=69) 0.075 0.850
 No Deterioration 63 (76.8%) 44 (62.9%) 52 (75.4%)
 Deterioration 19 (23.2%) 26 (37.1%) 17 (24.6%)
Hormonal Domain (5.4)
6 Weeks after RT Start (n=142) (n=124) (n=140) <.001 <.001
 No Deterioration 92 (64.8%) 21 (16.9%) 23 (16.4%)
 Deterioration 50 (35.2%) 103 (83.1%) 117 (83.6%)
1 Year after RT End (n=133) (n=125) (n=127) 0.006 0.026
 No Deterioration 97 (72.9%) 70 (56.0%) 76 (59.8%)
 Deterioration 36 (27.1%) 55 (44.0%) 51 (40.2%)
5 Years after RT End (n=84) (n=70) (n=72) 0.864 0.730
 No Deterioration 57 (67.9%) 46 (65.7%) 51 (70.8%)
 Deterioration 27 (32.1%) 24 (34.3%) 21 (29.2%)

Arm 1: PBRT Alone; Arm 2: PBRT+NC-STADT; Arm 3: PLNRT+PBRT+NC-STADT

MCID = Minimal clinically important difference

P-value from 2-sided Fisher’s exact test.

Deterioration is defined as a negative change from baseline (timepoint - baseline) with absolute value ≥ MCID.

HSCL-25

HSCL-25 total scores assessing depression are shown in Table 5 and plotted in Figure 2g. Scores are statistically significantly higher (worse) in Arms 2 and 3 at 6 weeks, but the differences are less than the baseline 0.5 SD of 3.9. One- and 5-year scores are similar across the three treatment arms. Change score analyses yielded similar findings, as did mixed effects regression modelling. However, while effect estimates with pattern mixture modelling were about the same, after multiple imputation for missing data the 6-week treatment contrasts were smaller and not statistically significant. The percentage of patients with scores >=44 (“caseness” for depression) at 6 weeks is 13%, at 1 year is 8%, and at 5 years is 13%, with no statistically significant differences among arms at any timepoint.

Table 5.

HSCL-25 total scores at each timepoint

PBRT Alone
(n=208)
PBRT+STADT
(n=216)
PLNRT+PBRT+STADT
(n=220)
P-value
Arm 2 vs. 1
P-value
Arm 3 vs. 1
Baseline (n=185) (n=155) (n=145) 0.263 0.461
 Mean (SD) 33.4 (8.2) 32.4 (7.0) 32.7 (8.2)
6 Weeks after RT Start (n=143) (n=143) (n=159) 0.018 0.003
 Mean (SD) 32.5 (8.1) 34.8 (8.0) 35.2 (7.6)
1 Year after RT End (n=140) (n=151) (n=155) 0.410 0.937
 Mean (SD) 32.9 (8.3) 32.1 (7.3) 32.9 (8.4)
5 Years after RT End (n=86) (n=92) (n=94) 0.878 0.641
 Mean (SD) 33.1 (9.7) 32.9 (8.8) 32.5 (8.3)

ADT=androgen deprivation therapy. PBRT=prostate bed radiotherapy. PLNRT=pelvic

lymph node radiotherapy. PSA=prostate-specific antigen.

Arm 1: PBRT Alone; Arm 2: PBRT+NC-STADT; Arm 3: PLNRT+PBRT+NC-STADT

Lower scores are better

P-value from 2-sided t-test with pooled variances.

Figure 2g.

Figure 2g.

HSCL-25 Total Score

EQ-5D

Mean scores for both indices are summarized in Table 6 and Figures 2h and 2i. Of note, for the Index Score, 0.5 SD at baseline=0.05, while for the VAS scale, 0.5 SD=7.1. EQ-5D utility scores range between 0 (death or worse possible health state) and 1 (best possible health state) with EQ-5D Index utilities in this study ranging from 0.87 to 0.91, with no statistically significant differences in the mean scores. VAS scores, with a possible range of 0–100 with the same directionality as the Index score, ranged from a low of 79.2 to a high of 83.1, again, with no statistically significant differences in the mean scores. In the change-from-baseline analyses, a statistically significant but not clinically significant difference was found between Arms 3 and 1 in the VAS at 6 weeks (−3.0 vs. +1.2, p=0.009). The results of mixed effects regression models assuming MAR for the Index Scores found no significant effects; however, the pattern mixture modelling detected significant differences and interactions related to missingness. The model detected a significant effect of race with whites having a 0.035 higher average score than non-whites (p=0.004), although this is not clinically significant. The model reveals that among completers, there was a significant difference between Arm 2 and Arm 1 at the end of RT (−0.079, SE 0.028, p=0.005); however, the difference was essentially zero among those with incomplete data (−0.006). There was also a statistically significant Arm 2 vs. Arm 1 difference at 5 years among completers (−0.070, SE 0.028, p=0.011) but not among the non-completers −0.007).

Table 6.

EQ5D Index and VAS scores at each timepoint

PBRT Alone
(n=208)
PBRT+STADT
(n=216)
PLNRT+PBRT+STADT
(n=220)
P-value
Arm 2 vs. 1
P-value
Arm 3 vs. 1
Baseline
EQ5D Index (n=184) (n=160) (n=162) 0.298 0.221
 Mean (SD) 0.89 (0.1) 0.91 (0.1) 0.91 (0.1)
VAS (n=180) (n=158) (n=163) 0.679 0.222
 Mean (SD) 80.7 (14.7) 81.3 (13.7) 82.6 (14.2)
 6 Weeks after RT Start
 EQ5D Index (n=138) (n=137) (n=161) 0.218 0.915
 Mean (SD) 0.89 (0.1) 0.87 (0.2) 0.89 (0.1)
 VAS (n=140) (n=144) (n=162) 0.183 0.218
 Mean (SD) 81.8 (14.7) 79.5 (14.5) 79.7 (14.6)
 1 Year after RT End
 EQ5D Index (n=134) (n=145) (n=155) 0.557 0.940
 Mean (SD) 0.90 (0.1) 0.91 (0.1) 0.90 (0.1)
 VAS (n=134) (n=147) (n=154) 0.140 0.989
 Mean (SD) 83.1 (13.5) 80.6 (13.9) 83.0 (11.6)
 5 Years after RT End
 EQ5D Index (n=83) (n=82) (n=90) 0.722 0.940
 Mean (SD) 0.89 (0.1) 0.89 (0.1) 0.89 (0.1)
 VAS (n=83) (n=80) (n=88) 0.813 0.267
 Mean (SD) 79.8 (17.0) 79.2 (17.7) 82.4 (13.2)

ADT=androgen deprivation therapy. PBRT=prostate bed radiotherapy. PLNRT=pelvic

lymph node radiotherapy. PSA=prostate-specific antigen.

Arm 1: PBRT Alone; Arm 2: PBRT+NC-STADT; Arm 3: PLNRT+PBRT+NC-STADT

Higher scores are better

P-value from 2-sided t-test with pooled variances.

Figure 2h.

Figure 2h.

EQ-5D Index Score

Figure 2i.

Figure 2i.

EQ-5D VAS

The mixed effects models for VAS detected the same significant difference between Arms 3 and 1 at 6 weeks after RT start as the change score analysis. The estimated difference was of similar magnitude but not statistically significant under pattern mixture modelling; however, this was primarily due to increased standard error. Following multiple imputation, this difference was smaller and not statistically significant.

Quality Adjusted Life Years (QALYs)

Table 7 provides statistical comparisons of QALYs for overall survival QAFFPY years using the EQ-5D score as health state weights. Estimated means are restricted to 8 years. While there were no statistically significant differences among arms in QALYs, Arm 3 and Arm 2 had statistically significantly higher QAFFPY than Arm 1. Figure 3a depicts the overlap among arms of QALYs and the differences among arms of QAFFPYs (figure 3b), favoring Arm 3.

Table 7.

Quality-Adjusted Survival and Quality-Adjusted Freedom From Progression at each timepoint

PBRT Alone
(n=166)
PBRT+STADT
(n=143)
PLNRT+PBRT+STADT
(n=148)
Difference (SE), P-value
Arm 2 vs. 1
Difference (SE), P-value
Arm 3 vs. 1
Difference (SE), P-value
Arm 3 vs. 2
 Quality-Adjusted Overall Survival (Q-OS) Restricted Mean (SE) 6.817 (0.116) 6.943 (0.107) 6.859 (0.110) 0.127 (0.158) p=0.80 0.042 (0.160) p=0.79 −0.084 (0.154) p=0.58
 Quality-Adjusted Freedom From Progression (Q-FFP) Restricted Mean (SE) 5.538 (0.190) 6.160 (0.178) 6.560 (0.137) 0.622 (0.261) p=0.017 1.062 (0.234) p<0.0001 0.439 (0.224) p=0.050

ADT=androgen deprivation therapy. PBRT=prostate bed radiotherapy. PLNRT=pelvic

lymph node radiotherapy. PSA=prostate-specific antigen.

Arm 1: PBRT Alone; Arm 2: PBRT+NC-STADT; Arm 3: PLNRT+PBRT+NC-STADT

Higher scores are better

Since the last QOL assessment was at 5 years, the QALY estimates were restricted to τ=8 years.

QOL assessments (EQ-5D) were obtained at baseline, 6 weeks after RT, 1 year after RT, and 5 years after RT.

P-value from 2-sided z-test.

For each individual, the QOL scores were treated as a step function with change points at the assessment times.

Figure 3a.

Figure 3a

Figure 3b.

Figure 3b

DISCUSSION

The first aim, to assess the degree, duration, and significant differences of disease-specific health-related quality-of-life decrements among treatment arms, demonstrated that among 644 patients consented to QOL on the SPPORT trial, bowel and urinary domain scores as measured by EPIC, decreased at 6 weeks post-RT start then increased by years 1 and 5, although not to baseline levels. One clinically significant difference in bowel scores was on STADT + PLNRT + PBRT at 6 weeks but not at future timepoints, and the resolution of this issue may further support the addition of PLNRT. Assessing differences among treatments for bowel and bladder patient reported function was essentially similar over time leaving treatment options open. Older patients had slightly worse urinary QOL among arms. However, because these outcomes were similar across arms does not indicate acceptable levels of dysfunction for patients as deterioration in all EPIC QOL domains over time was high with 22% to 42% of patients reporting worsening of QOL at 5 years. Specifically, the 5-year global continence scores decreased by a mean of −9.1, and median −2.3 points on STADT + PLNRT + PBRT and pad use increased from baseline to last follow-up across all treatment arms, from roughly 30% to 45%. In contrast to our findings, a non-randomized single institution U.S. study by Melotek et al (2015) in 122 men of similar age and racial profiles, treated with similar PBRT dose and a slightly higher PLNRT median dose of 50.4Gy, (56% had STAD), showed global continence scores were unchanged over time and pad use went from 39% at baseline (prior to post-prostatectomy IMRT) to 27% at their last timepoint at 4 years following salvage therapy. The randomized multi-institutional design of the current study vs the Melotek study, and other differences such as the current study had a longer follow up (5 vs 4 years) and better QOL compliance (45% at 5 years in this study vs 33% at 4 years in Melotek study), may explain some of the contrasts, but continence and pad use still require further study. Improvement of these symptoms is critical since studies have demonstrated that urinary symptoms related to incontinence induce significantly more psychological distress than either sexual or bowel dysfunction (Maggie et al, 2019).

For the sexual and hormonal domains measured by the same instrument, there were statistically significant differences between arms at 6 weeks and 1 year with patients receiving STADT exhibiting poorer QOL scores. By year 5 the differences were no longer significant. However, while the studies show no overall domain differences among treatment arms at 5 years, individual symptoms items such as sexual function and hormonal bother were worse at 6 weeks and sexual function was worse at one year on the STAD arms compared to the arm without STADs. Other studies have similar findings, that sexual functioning was decreased postprostatectomy and during STADT, but returned to a level not statistically significantly different from controls after stopping STADT (Black et al 2007 Sexuality remains important for many older adults, yet they consistently report poor responsiveness from clinicians in addressing these issues due to clinician-related embarrassment, negative attitudes and seeming disinterest (Bauer et al 2015). Even when patients are referred for sexual aids, our data indicates that sexual aids provide little difference in function among arms. Therefore, continued clinical attention and further research is needed to assist patients with sexual function regardless of therapy chosen for this disease.

In comparison to the well-cited study by Sanda et al (2008) of prostate cancer QOL using the EPIC instrument in men treated with radical prostatectomy, brachytherapy, or external-beam radiotherapy with or without hormones and prior to any salvage therapy, patients in this study across treatment arms had similar frequency of urinary and bowel problems regardless of treatment. Notable differences in this study are that about 80% of the patients across arms reported “poor to very poor to no erections” at baseline, compared to only 30–37% of those on the radiotherapy arms of the Sanda et al study. Yet, the men in this study are on average 6 years younger compared to the men in the Sanda study. At one year, patients on this study also had higher rates of erectile dysfunction with 75% on the arm without STAD, 85–86% on the arms with STAD compared to 48–57% with this level of dysfunction at the same time point on the Sanda study.

For the second aim, depression as measured by the HSCL-25 showed statistically but not clinically significant differences at 6 weeks only. This is contrary to increasing evidence that androgen deprivation therapy for prostate cancer may be associated with depression (Yiannopoulou et al 2020). For example, a recent review of the literature (Alwhaibi et al (2022) found 23 studies confirming the occurrence and worsening of depressive symptoms compared to 10 studies which indicated no change in depressive symptoms in ADT-treated patients. However, there is little evidence in the literature for duration of ADT and depressive symptoms. It is possible that depressive symptoms related to ADT may manifest with longer courses than used in the current study, but more research is needed. Further, the HSCL-25 is a screening tool, not a diagnostic tool, which may account for some differences in rates of depression reported, however, screening rates are usually higher than diagnostic rates of depression. Regardless, clinical awareness of the association of depression with STADT should be incorporated into assessments or referrals for patients on this therapy.

The third aim assessed whether an incremental gain in FFP and overall survival with more aggressive therapy outweighs decrements in the primary generic domains of health-related quality of life (i.e., mobility, self-care, usual activities, pain/discomfort, and anxiety/depression). This report is unique in the literature, reporting both Quality-Adjusted Life Years (QALY) and Quality-Adjusted FFP Years (QAFFPY) in this patient population and among treatments.

In contrast to men in this study reporting worse sexual QOL at baseline and following treatment, the EQ-5 Index of overall QOL/utilities indicated better overall QOL/utilities at baseline and throughout the follow up timepoints compared to population norms. This appears to be the first study in the literature reporting QALYs for salvage radiotherapy treatments of men with prostate cancer. For reference, a study of population norms among men and women over the life span in the U.S. reported mean EQ-5D scores for those age 65–74 years to be 0.824 (0.217). In this study, the QALYs (accounting for both the duration of life and quality of life) are similar among arms, and with any of the treatments tested we find close to a 7-year restricted mean (over an 8-year time horizon) of relatively satisfactory health. Regarding freedom from progression, QAFFPYs demonstrated a significant benefit to the STADT + PLNRT + PBRT arm. These generally reflect the differences in the FFP curves themselves (see Pollack et al, 2022, Figure 2a) and therefore the QOL decrements did not erase the time-to-event differences. One quality-adjusted survival analysis using data from RTOG 85–13 of RT alone compared to RT plus androgen deprivation in men with locally advances prostate cancer also found the arm with androgen deprivation to have slightly improved QALYs compared to RT alone (Ahn et al 2016).

Limitations

QOL compliance dropped significantly at the 5-year timepoint to 42–47% depending on the assessment. Only 22–25% of the patients were fully compliant, i.e., providing data at all four time points. Accrual was doubled to allow for statistical power with a 50% non-compliance rate, however, the 5-year time point still slightly underperformed. Statistical methods were used to adjust for missing data. While these methods are helpful, they cannot eliminate the possibility of attrition and survivorship bias. For example, if attrition is related to previous outcomes or other observable subject characteristics, the mixed effects models provide valid inferences, but there is no guarantee that this is the case. Multiple imputation assumes that all factors related to missingness have been captured by the predictor variables in the imputation model. Regarding survivorship bias, the high 5-year survival rates (over 93% in all three treatment arms) and lack of significant differences in overall survival argues against such a bias. One could also argue that interest lies only in the surviving patients.

Interpretation

Results suggest that while QOL generally declined among all arms at 6 weeks post RT start, there were no clinically significant differences in QOL among arms at 5 years, either in terms of the mean scores or the percentage of patients with decrements >MCID. QOL did not return to baseline in any of the arms, although the mean decrements at 5 years were generally modest and <MCID (with the exception of bowel bother in Arm 3, urinary incontinence in all arms, and sexual bother in Arm 3). The percentage of patients with decrements >MCID at 5 years was not negligible, and within each arm ranged from 31% to 40% for bowel, 31% to 42% for urinary, 22% to 37% for sexual, and 29% to 34% for hormonal symptoms. Similarly, “caseness” for depression at 6 weeks and 5 years was 13% across arms. Thus, management of the decrements in bowel, bladder, hormonal, sexual function and depression require clinical attention regardless of treatment choice. QALYs for overall survival was not statistically significantly different among arms; however, for freedom from progression, the QAFFPYs favored STADT + PLNRT + PBRT for salvage treatment of prostate cancer.

Supplementary Material

Supp Table 1
Supp Table 2

Funding Statement:

Research reported in this publication was supported by the National Cancer Institute of the National Institutes of Health under award number U10CA180868 (NRG Oncology Operations), U10CA180822 (NRG Oncology SDMC), UG1CA189867 (NCORP), CTEP and DCP. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Conflicts of Interest:

Drs Balough, Camarata, Horowitz, Lee, Rodrigues, Zelefsky have nothing to disclose.

Dr Bruner declares in the last 36 months Since the initial planning of the work all support for the present manuscript from NRG Oncology with Specifications/Comments of Studies funded by the NIH.

Dr Faria declares in the last 36 months Since the initial planning of the work all support for the present manuscript from RTOG.

Dr Karrison declares in the last 36 months Since the initial planning of the work all support for the present manuscript, from U10CA180822 grant to NRG Oncology Statistical and Data Management Center with Specifications/Comments of Salary support paid to institution.

Dr Lukka declares in the last 36 months Consulting fees received from Tersera, Bayer, Tolmar, Knight Pharma with Specifications/Comments of Honoraria recd; Payment or honoraria for lectures, presentations, speakers, bureaus, manuscript writing or educational events from Tersera, Bayer, Tolmar, Knight Pharma with Specifications/Comments of Honoraria recd; Participation on a Data Safety Monitoring Board or Advisory Board for Tersera, Bayer, Tolmar, Knight Pharma with Specifications/Comments of Honoraria recd.

Dr Michalski declares in the last 36 months Since the initial planning of the work all support for the present manuscript from US NCI: 5U10CA180833 (Lead Academic Site grant).

Dr Movsas declares in the last 36 months Grants or contracts from Varian, Inc and Siemens Inc with Specifications/Comments of to institution; Consulting fees received from Varian with Specifications/Comments of to me; Payment or honoraria for lectures, presentations, speakers, bureaus, manuscript writing or educational events from Varian with Specifications/Comments of to institution for educational event; Leadership or fiduciary role in other board, society, committee, or advocacy group, paid or unpaid from ASTRO BOard about 2 years ago with Specifications/Comments of travel covered; Stock or stock options received from United therapeutics, metlife, acrivon, mineralys, abbvie, biolife with Specifications/Comments of stocks owned by me.

Dr Pollack declares in the last 36 months Since the initial planning of the work all support for the present manuscript (e.g., funding,provision of study materials, medical writing, article processing charges, etc.) from support from a University of Miami School of Medicine, Sylvester Comprehensive Cancer Center NCI P30 grant (P30CA240113) and a University of Miami Sylvester Professorship with Specifications/Comments of P30 grant Payments made to institution. Covered some effort for research; Grants or contracts from NCI grant funding. Varian grant funding. Nothing related to this manuscript with Specifications/Comments of No conflicts; Payment or honoraria for lectures, presentations, speakers, bureaus, manuscript writing or educational events from No conflict. Visiting Professor honorarium at an academic institution with Specifications/Comments of No conflicts. Payments made to me; Support for attending meetings and/or travel for being a visiting professor, as above with Specifications/Comments of Payments made to me; Participation on a Data Safety Monitoring Board or Advisory Board for External Advisory Board at two institutions who are submitted NCI K12 Calabresi grants with Specifications/Comments of No payments have been made; Leadership or fiduciary role in other board, society, committee, or advocacy group, paid or unpaid from NRG Oncology Translational Research Program co-Chair emeritus with Specifications/Comments of No payments.. No conflicts.

Dr Sandler declares since the initial planning of the work all support for the present manuscript from ACR-NRG Oncology with Specifications/Comment of Support for my role as chair of the NRG Oncology GU Oncology Committee; Leadership or fiduciary role in other board, society, committee, or advocacy group, paid or unpaid from ASTRO with Specifications/Comments of Member, Board of Directors.

Dr Seiferheld declares in the last 36 months Since the initial planning of the work all support for the present manuscript from NRG Oncology SDMC with Specifications/Comments of Payments made to my institution from the NCI NRG Oncology SDMC grant.

Appendix

Dr. Bruner and NRG Oncology would like to thank the following accrual institutions who have been part of our mission to improve the lives of patients with cancer by conducting practice-changing multi-institutional clinical and translational research.

21st Century Oncology-Palatka

Academic Urology Prostate Center

Adventist Health Cancer Care Center Chico

Allan Blair Cancer Centre

Anne Arundel Medical Center

Appleton Medical Center

Arizona Center for Cancer Care

Arizona Oncology Associates-West Orange Grove

Arizona Oncology Services Foundation

Asante Rogue Regional Medical Center

Aspirus Regional Cancer Center

Auora West Allis Medical Center

Aurora St. Luke’s Medical Center

Backus Hospital

Beaumont Hospital

Bellin Memorial Hospital

Billings Clinic

Boston Medical Center

Brooke Army Medical Center

Cancer Treatment Center

Cape Radiation Oncology

Carolina East Medical Center

Carolina Regional Cancer Center

Carolinas Medical Center/Levine Cancer Institute

Carteret Hospital/Coleman Comprehensive Cancer Center

Case Western Reserve/University Hospital Seidman Cancer Center

Cedars-Sinai Medical Center

CentraState Medical Center

CHUM Hopital Notre Dame

City of Hope

Cleveland Clinic Foundation

Cleveland Clinic Wooster Specialty Center

Coastal Carolina Radiation Oncology Center

Columbia St. Mary’s Hospital Ozaukee

Columbia St. Mary’s Water Tower Medical Commons

Cooper Cancer Institute-Voorhees

Cooper Health System University Hospital

Cross Cancer Institute

Danville Regional Medical Center

Dartmouth-Hitchcock Medical Center/Norris Cotton Cancer Center

Decatur Memorial Hospital

Delaware Christiana Care Community Clinical Oncology Program

Dixie Regional Medical Center

Dr H Bliss Murphy Cancer Centre

Duke Cancer Institute Macon Pond

Emory University

Emory University Hospital Midtown

Fairview Hospital

Flower Hospital

Fox Chase Cancer Center

Geisinger Medical Center

Greenville Health System-Eastside

Gundersen Lutheran Medical Center

Henry Ford Health System

Hunter Holmes McGuire Veterans Administration Medical Center

Inova Alexandria Hospital

Intermountain Medical Center

IU Health Goshen Center for Cancer Care

James A. Haley Veterans Hospital

John H Stroger Jr Hospital of Cook County

Juravinski Cancer Center

Kaiser Permanente

Kaiser Permanente Cancer Treatment Center (Rad Onc)

LAC+USC Medical Center

Lake University Seidman Cancer Center

Lankenau Medical Center

Legacy Mt. Hood Medical Center

Lewis Cancer & Research Pavilion at St. Joseph’s/Candler

London Health Sciences Centre/Western University

Lowell General Hospital

Loyola University Medical Center

Mary Bird Perkins Cancer Center

Massachusetts General Hospital Cancer Center

Mayo Clinic

McGill University Health Centre

McKay Dee Hospital

McKee Medical Center

McLaren Flint

Medical College of Wisconsin

Medical University of South Carolina Hollings Cancer Center

Memorial Medical Center

Memorial Medical Center

Memorial Sloan-Kettering Cancer Center

Mercy San Juan Medical Center

Minnesota Oncology Hematology PA-Maplewood

MSKCC at Commack

MSKCC at Rockville Centre

MSKCC at Sleepy Hollow

Naval Medical Center

New Mexico Oncology Hematology

New York Oncology Hematology PC-Albany Patroon

Norris Cotton Cancer Center-North

North Star Lodge Cancer Center at Yakima Valley Memorial Hospital

Northeast Radiation Oncology Center

Northern Rockes Radiation Oncology Center

Northwestern Ontario Regional Cancer Centre

Ochsner Clinic

Ohio State University

Our Lady of Lourdes Memorial Hospital

Piedmont Hospital

Pomona Valley Hospital Medical Center

Presbyterian-Weill Medical College

Princeton Medical Center

Radiological Associates of Sacramento

Rapid City Regional Hospital

Reading Hospital

REX Cancer Center

Richard A Roudebush VA Hospital VA Medical Center

Riverside Methodist Hospital

Robert Wood Johnson University Hospital

Rochester General Hospital

Rohnert Park Cancer Center (Rad Onc)

Roseville ROC

Roswell Park Cancer Institute

Saint Agnes Hospital/Agnesian Cancer Center

Saint John Regional Hospital

Saint Marys Hospital and Regional Medical Center

Sanford Medical Center Fargo

Seattle Cancer Care Alliance

Sentara Norfolk General Hospital

Shields Radiation Oncology Center- Mansfield

Somerset Medical Center

Southwest General Hospital

Spartanburg Regional Healthcare System

St. Alphonsus Regional Medical Center

St. Anne’s Hospital

St. Johns Hospital Cancer Institute

St. Luke’s Hospital

St. Luke’s Hospital & Health Network

St. Luke’s-Roosevelt Hospital Center

Texas Cancer Center-Sherman

Texas Oncology Cancer Center-Sugar Land

The Center for Cancer Prevention and Treatment (CCCPT) at St. Joseph Hospital

The Klabzuba Cancer Center

The Nebraska Medical Center

The New York Methodist Hospital

The Permanente Medical Group Roseville Radiation Oncology Center (Rad Onc)

The Urology Center of Colorado

Thomas Jefferson University Hospital

Tom Baker Cancer Centre

Tulane University

Unity Point Health Des Moines/Iowa Methodist Medical Center

University of Alabama at Birmingham Medical Center

University of California, Davis

University of California, San Diego (UCSD)

University of California, San Francisco (UCSF)

University of Colorado

University of Florida Health Science Center

University of Kentucky

University of Maryland Medical Center

University of Miami

University of Michigan Medical Center

University of New Mexico Cancer Center/New Mexico Cancer Care Alliance

University of Oklahoma Health Sciences Center

University of Rochester

University of Texas Medical Branch

University of Utah/Huntsman Cancer Institute

University of Vermont Medical Center

University of Washington Medical Center

UT Southwestern Simmons Cancer Center

Utah Cancer Specialists

UW Cancer Center Johnson Creek

VA Long Beach Healthcare System

Veteran Affairs New York Harbor

Via Christi

Virginia Mason CCOP

Wake Forest University Health Sciences

Washington Cancer Institute at MedStar Washington Hospital Center

Washington University School of Medicine

Waukesha Memorial

Wayne State University/Karmanos Cancer Institute

WellSpan Health-York Hospital

Wenatchee Valley Medical Center

West Michigan Cancer Center

West Virginia University Charleston

Wheaton Franciscan Center Care-All Saints

Wilford Hall USAF Medical Center

Zablocki VA Medical Center

Footnotes

Authorship:

Conceptualization: DWB, EMH, HRL, BM, JMM, AP, HS

Methodology: DWB, TGK, HRL, JMM

Software: TGK, WS

Validation: AP

Formal analysis: DWB, TGK, JMM, WS

Investigation: AGB, ASC, EMH, AP, HS

Resources: AGB, SLF, EMH, GBR, HS, MJZ

Data Curation: AGB, ASC, RJL, AP, WS

Writing – Original Draft: DWB, TGK, WS

Writing — Review & Editing: AGB, DWB, ASC, EMH, TGK, HRL, RJL, BM, JMM, AP, GBR, HS

Visualization: TGK, HRL, WS

Supervision: DWB, BM, HS, MJZ

Project Administration: RJL

Funding Acquisition: DWB

Agree to be accountable for all aspects of the work, which includes ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved: AGB, DWB, ASC, SLF, EMH, TGK, HRL, RJL, BM, JMM, AP, GBR, HS, WS, MJZ

NOTICE:

This research was supported by the National Institutes of Health (NIH). In accordance with the NIH Public Access Policy, the final peer-reviewed manuscript will be submitted to PubMed Central upon acceptance for publication and made publicly available immediately upon publication.

Data Availability Statement:

The data from the present publication will be made available by request from the NCTN/NCORP Data Archive: https://nctn-data-archive.nci.nih.gov.

References:

  1. EPIC Expanded Prostate Cancer Index Composite, Accessed 12-29-2023 https://medicine.umich.edu/dept/urology/research/epic [Google Scholar]
  2. Li JR, Wang SS, Chen CS, Cheng CL, Hung SC, Lin CH, Chiu KY. Conventional androgen deprivation therapy is associated with an increased risk of cardiovascular disease in advanced prostate cancer, a nationwide population-based study. PLoS One. 2022. Jun 28;17(6):e0270292. doi: 10.1371/journal.pone.0270292. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Rice SM, Oliffe JL, Kelly MT, Cormie P, Chambers S, Ogrodniczuk JS, Kealy D. Depression and Prostate Cancer: Examining Comorbidity and Male-Specific Symptoms. Am J Mens Health. 2018. Nov;12(6):1864–1872. doi: 10.1177/1557988318784395. Epub 2018 Jun 29. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Hauger RL, Saelzler UG, Pagadala MS, Panizzon MS. The role of testosterone, the androgen receptor, and hypothalamic-pituitary-gonadal axis in depression in ageing Men. Rev Endocr Metab Disord. 2022. Dec;23(6):1259–1273. doi: 10.1007/s11154-022-09767-0. Epub 2022 Nov 22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Dalela D, Krishna N, Okwara J, Preston MA, Abdollah F, Choueiri TK, Reznor G, Sammon JD, Schmid M, Kibel AS, Nguyen PL, Menon M, Trinh QD. Suicide and accidental deaths among patients with non-metastatic prostate cancer. BJU Int. 2016. Aug;118(2):286–97. doi: 10.1111/bju.13257. Epub 2015 Sep 19. [DOI] [PubMed] [Google Scholar]
  6. Nguyen C, Lairson DR, Swartz MD, Du XL. Cost-Effectiveness of Adding Androgen Deprivation Therapy to Radiation Therapy for Men with Advanced Prostate Cancer from a U.S. Payer’s Perspective. J Manag Care Spec Pharm. 2019. Feb;25(2):225–234. doi: 10.18553/jmcp.2019.25.2.225. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Pollack A, Karrison T, et al. The addition of androgen deprivation therapy and pelvic lymph node treatment added to prostate bed Salvage radiotherapy (NRG Oncology/RTOG 0534 SPPORT): an international, multicenter, randomised phase 3 trial, Lancet 2022; 399:1886–1901. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Wei JT, Dunn RL, Litwin MS, Sandler HM, Sanda MG. Development and validation of the expanded prostate cancer index composite (EPIC) for comprehensive assessment of health-related quality of life in men with prostate cancer. Urology. 2000. Dec 20;56(6):899–905. doi: 10.1016/s0090-4295(00)00858-x. [DOI] [PubMed] [Google Scholar]
  9. Derogatis LR, Lipman RS, Rickels K, Uhlenhuth EH, Covi L. The Hopkins Symptom Checklist (HSCL): a self-report symptom inventory. Behav Sci. 1974. Jan;19(1):1–15. doi: 10.1002/bs.3830190102. [DOI] [PubMed] [Google Scholar]
  10. Nabbe P, Le Reste JY, Guillou-Landreat M, Beck-Robert E, Assenova R, Lazic D, Czachowski S, Stojanović-Špehar S, Hasanagic M, Lingner H, Clavería A, Fernandez San Martin MI, Sowinska A, Argyriadou S, Lygidakis C, Le Floch B, Doerr C, Montier T, Van Marwijk H, Van Royen P. One consensual depression diagnosis tool to serve many countries: a challenge! A RAND/UCLA methodology. BMC Res Notes. 2018. Jan 3;11(1):4. doi: 10.1186/s13104-017-3111-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Hough R, Landsverk J, Stone J. Comparison of psychiatric screening questionnaires for primary care patients. Final report for NIMH Contract No. (DB) 122:913–922, 1995. [Google Scholar]
  12. Schulz MW, Chen J, Woo HH, Keech M, Watson ME, Davey PJ. A comparison of techniques for eliciting patient preferences in patients with benign prostatic hyperplasia. J Urol. 2002. Jul;168(1):155–9. [PubMed] [Google Scholar]
  13. Badia X, Herdman M, Kind P. The influence of ill-health experience on the valuation of health. Pharmacoeconomics. 1998. Jun;13(6):687–96. doi: 10.2165/00019053-199813060-00005. [DOI] [PubMed] [Google Scholar]
  14. Ontario Health (Quality). Homologous Recombination Deficiency Testing to Inform Patient Decisions About Niraparib Maintenance Therapy for High-Grade Serous or Endometrioid Epithelial Ovarian Cancer: A Health Technology Assessment. Ont Health Technol Assess Ser. 2023. Aug 10;23(5):1–188. [PMC free article] [PubMed] [Google Scholar]
  15. Rubin DB. Multiple imputation after 18+ years. Journal of American Statistical Association 1996;91:473–489. [Google Scholar]
  16. Hedeker D, Gibbons RD (1997). Application of random-effects pattern-mixture models for missing data in longitudinal studies. Psychological Methods 2:64–78. [Google Scholar]
  17. Zhao H, Tsiatis AA (2000). Estimating mean quality adjusted lifetime with censored data. The Indian Journal of Statistics, Series B, 62:175–188. [Google Scholar]
  18. Melotek JM, Liao C, Liauw SL. Quality of Life after Post-Prostatectomy Intensity Modulated Radiation Therapy: Pelvic Nodal Irradiation Is Not Associated with Worse Bladder, Bowel, or Sexual Outcomes. PLoS One. 2015. Oct 29;10(10):e0141639. doi: 10.1371/journal.pone.0141639. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Maggi M, Gentilucci A, Salciccia S, Gatto A, Gentile V, Colarieti A, Von Heland M, Busetto GM, Del Giudice F, Sciarra A. Psychological impact of different primary treatments for prostate cancer: A critical analysis. Andrologia. 2019. Feb;51(1):e13157. doi: 10.1111/and.13157. Epub 2018 Oct 3. [DOI] [PubMed] [Google Scholar]
  20. Black PC, Basen-Engquist K, Wang X, Swartz RJ, Eddings T, Matin SF, Swanson D, Wood CG, Pisters LL, Babaian RJ, Troncoso P, Pettaway CA. A randomized prospective trial evaluating testosterone, haemoglobin kinetics and quality of life, during and after 12 months of androgen deprivation after prostatectomy: results from the Postoperative Adjuvant Androgen Deprivation trial. BJU Int. 2007. Jul;100(1):63–9. doi: 10.1111/j.1464-410X.2007.06846.x. [DOI] [PubMed] [Google Scholar]
  21. Bauer M, Haesler E, Fetherstonhaugh D. Let’s talk about sex: older people’s views on the recognition of sexuality and sexual health in the health-care setting. Health Expect. 2016. Dec;19(6):1237–1250. doi: 10.1111/hex.12418. Epub 2015 Oct 8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Sanda MG, Dunn RL, Michalski J, Sandler HM, Northouse L, Hembroff L, Lin X, Greenfield TK, Litwin MS, Saigal CS, Mahadevan A, Klein E, Kibel A, Pisters LL, Kuban D, Kaplan I, Wood D, Ciezki J, Shah N, Wei JT. Quality of life and satisfaction with outcome among prostate-cancer survivors. N Engl J Med. 2008. Mar 20;358(12):1250–61. doi: 10.1056/NEJMoa074311. [DOI] [PubMed] [Google Scholar]
  23. Yiannopoulou KG, Anastasiou AI, Kontoangelos K, Papageorgiou C, Anastasiou IP. Cognitive and Psychological Impacts of Different Treatment Options for Prostate Cancer: A Critical Analysis. Curr Urol. 2020. Dec;14(4):169–177. doi: 10.1159/000499242. Epub 2020 Dec 18. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Alwhaibi A, Alsanea S, Almadi B, Al-sabhan J, & Alosaimi FD (2022). Androgen deprivation therapy and depression in the prostate cancer patients: review of risk and pharmacological management. The Aging Male, 25(1), 101–124. 10.1080/13685538.2022.2053954 [DOI] [PubMed] [Google Scholar]
  25. Ahn S, Lee M, Jeong CW. Comparative quality-adjusted survival analysis between radiation therapy alone and radiation with androgen deprivation therapy in patients with locally advanced prostate cancer: a secondary analysis of Radiation Therapy Oncology Group 85–31 with novel decision analysis methods. Prostate Int. 2018. Dec;6(4):140–144. doi: 10.1016/j.prnil.2018.01.002. Epub 2018 Feb 2. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supp Table 1
Supp Table 2

Data Availability Statement

The data from the present publication will be made available by request from the NCTN/NCORP Data Archive: https://nctn-data-archive.nci.nih.gov.

RESOURCES