Patient Reported Urinary Incontinence and Sexual Impotency Following Multimodality Radical Therapy for Prostate Cancer: An Untold Story of Compound Toxicity?

Lay summary : Patients deliberating the choice of primary therapy of unfavorable localized prostate cancer are not aware of the impact of having multimodality radical therapy of both surgery and radiation on functional outcomes. In this study of patient reported outcomes we demonstrated significantly worse urinary control and a threefold risk of a dual impairment of both not being able to control urine and not being sexually potent in patients treated with both surgery and radiation as compared to radiation alone. Abstract Purpose: To evaluate urinary continence and sexual potency following radical prostatectomy and adjuvant radiotherapy. Materials/Methods: Expanded Prostate Cancer Index Composite (EPIC) surveys of patients with localized prostate cancer treated with surgery followed by adjuvant/salvage pelvic radiotherapy (S+RT) were analyzed. A control cohort was primary radiotherapy (RT). Results: Surveys at least 1 year after treatment were available for 130 S+RT and 374 RT patients. For S+RT vs. RT, the mean urinary incontinence score was 68 [6.25 -100] versus 86.4 [CI-95 39.5-100] (p<0.001), confirming 6.5 points of clinically significant difference. The adjusted odds ratio for superior urinary function was 2.67 (1.7-4.1, p<0.001) for primary radiotherapy. The odds ratio of having both poor urinary and sexual performance was 0.29 in RT arm (0.14-0.58, p<0.001) when adjusted to age and ADT use, group risk stratification, co morbidities and smoking status. Conclusion: In this cross sectional study, Surgery with adjuvant/salvage RT was associated with significantly worse patient reported urinary continence outcomes at 1-year post treatment, lower odds of achieving perfect urinary continence and a threefold risk of reverse ‘bifecta’ with inferior urinary continence and sexual performance. Longitudinal studies of evolving toxicity are required to validate these findings.


Introduction
Cancer control , urinary continence and sexual function are the components of Trifecta, an outcome model for localized prostate cancer 1 . Since both overall survival and disease specific survival are comparable following modern surgery or radical radiotherapy 2 , disparate patient reported outcomes of urinary continence and sexual potency , hence "Bifecta" , following primary therapy have been the subject of considerable controversy.
Patients failing primary therapy and requiring local salvage are often susceptible to compound toxicity in these domains. Long term study has shown that in tertiary center a 20-30% of patients treated with prostatectomy will require salvage radiotherapy 3 .Patients with high or very high risk prostate cancer who undergo radical prostatectomy are even more likely to require additional local therapy .Thus for example, using a widely utilized nomogram 4 , a 65 year old patient with T2B Gleason 4+4 with 50% core positivity who undergoes radical prostatectomy is predicted to have a 20% chance of organ confined disease, and 79% chance of extracapsular extension. The nomogram offers an interpretation of this prediction reassuring that half of patients without organ confined disease will have long term control, however it is necessary to point out that the other half will require salvage radiotherapy. Thus, at least forty percent of men electing surgery with the above disease characteristics will be exposed to the compound toxicity of surgery and radiotherapy.
There is a growing body of literature regarding comparative domain specific outcomes of primary surgery versus primary radiotherapy 2 . These comparisons are 5 confounded by multiple factors, including age, inherent differences in the toxicity profile, and differences in technique of both surgery and radiotherapy. For men who undergo surgery with a high probability of requiring radiotherapy, we propose that such comparisons are less meaningful. The composite effects of radical prostatectomy followed by salvage radiotherapy on quality of life are not generally discussed a-priori with the patient during the initial consultation even when the likelihood of requiring radiation is considerable.
In our study we endeavor to provide information regarding bifecta, urinary continence and sexual potency for patients with a high probability of requiring salvage radiation following surgery.
QOL of data routinely collected in an institutional review board (IRB-"Helsinki committee") approved clinical database were utilized for this study. The specific goal of the current study was to evaluate one-year patient-reported sexual and urinary outcomes of patients who underwent radical prostatectomy and salvage/adjuvant radiotherapy. For comparison, we studied a cohort of patients receiving primary radiation therapy. -The study was approved by the Sheba Medical Center Helsinki Ethics Commission -SMC-9132-11

Material and methods
The cohort includes localized prostate cancer patients treated at the Sheba Medical Center, between the years 2002-2018.
Patients were recruited by a research associate (IW) who was present at approximately one out of 4 outpatient genitourinary clinic days per week and asked 6 to complete a paper survey. Patients who experienced difficulty were assisted. The survey included the validated Hebrew translation 5 of the Expanded Prostate Cancer Index Composite short form (EPIC-26) measuring functional outcomes (urinary incontinence, urinary irritation and obstruction, bowel, sexual, and vitality and hormonal function) 6 , epidemiology data, and questions about the use of interventions for sexual dysfunction. For the purpose of the current study, we focused on urinary and sexual outcomes.
Multivariable regression models using SPSS 24th edition, were utilized to compare functional outcomes diagnostic stages and self-reported treatment groups. The study was approved by the institutional Helsinki ethics commission.
In the surgery arm, patients underwent both robot-assisted laparoscopic prostatectomy or open radical prostatectomy. Since both kinds of surgery are associated with similar functional outcomes 7 , they were pooled for this analysis.

Data Analysis
Epic-26 a well-established validated tool for assessment of quality of life function in prostate cancer 6 .The four domains of assessment urinary, sexual, bowel, and hormonal domains were calculated as a score between 0-100 through a coding conversion to a SPSS analysis. Baseline patient's characteristics was collected by medical chart. Baseline characteristics were compared across treatments using Ttest or Mann Whitney test for continuous parametric and a-parametric variables respectively. Categorical variables were analyzed using X2 test.
Each domain in the EPIC questionnaire was calculated for a score mean, median, range and standard deviation for each group of treatment. Scores distribution was analyzed for normalization criteria using histogram, kurtosis and skewness tests.
None of the domains had normal distribution, so comparison was done by nonparametric analysis using Mann-Whitney test.
We also used distribution based approaches to establish minimally important (i.e., clinically-relevant) differences for each EPIC-26 domain. Previous studies have found that half of a standard deviation is appropriate choices for a distribution-based minimally important difference cut-point 8 .
A perfect urinary score (above 90) was then evaluated as a dichotomy variable. A Multivariable logistic regression was performed for treatment type, age, group risk, use of ADT, diabetes and smoking status. Adjust Odds ratio (OR)and CI 95% were calculated for each variable.
For the purpose of this quality of life study we have defined 'reverse bifecta' as sexual impotency and urinary incontinence with a score lower than 60 in both sexual and urinary domains of the EPIC 9 . This definition while new, correlates with previous studies that notice that a score below 60 is associated with worse QOL outcomes 10 .
We calculated the odds of 'reverse bifecta' using multivariable logistic regression.

Patient characteristics
The clinical database includes 2023 patients with prostate cancer treated with radiotherapy between 2002-2018, 1500 received radiotherapy (RT) and 523 surgery Patients epidemiologic and clinical data are reported in Table 1.
The mean age for RT was older compare with S+RT with a mean age was 69.8 ±6.6 vs. 64±6.6 (P<0.001) respectively. There was no difference in smoking status diabetes, peripheral vascular, respiratory, or gastrointestinal diseases. There was a statistical significant difference for ischemic heart disease (19.7% vs 11.8%. p=0.042) and chronic renal disease (8.5% vs 3.1%, p=0.042). 9 The National Comprehensive Cancer consortium group risk classification between both cohorts was statistically different. High, intermediate and low risk were 49.5%, 29.3% and 11% in the RT group respectively and 60%, 36% and 4% in the S+RT group respectively. (P=0.021). There was more Androgen deprivation therapy (ADT) use in the RT group (58.7%)as compare to the RP+RT cohort (p<0.001).

EPIC domain scores
The mean urinary incontinence score was 68±32 [6.25 -100] for S+RT versus 86.4±19 [CI-95 39.5-100] for RT (p<0.001). (Table 2). Using the distribution approach with a 0.5 SD this result confirms a 6.5 points of clinically significant difference. The urinary irritation domain was not statistical different between the groups, but the overall urinary function was higher in the RT group (82.6 vs 74.8, p=0.002).
There was no statistical difference in sexual domains between the two cohorts. For the S+RT the mean score was 31.3±29.1 and for the RT group the mean score was 28.5±25 (p=0.576) (Table 3) The hormonal score was significantly worse in the primary radiotherapy cohort with a mean score of 83.3±20.6 compared with 89.5±16 in the surgery arm (p<0.001).
Moreover, the bowel function score was also significantly lower in the primary radiotherapy relatively to surgery arm with a score of 87.1±18.1 and 91.1±16 respectively (p=0.009).

Multivariable analysis
Multivariable regression analysis was performed for superior urinary score Adjustment for age (as a dichotomy variable), risk group (high risk group for reference), ADT use and treatment type (RT for reference) were performed. The adjusted Odds ratio for superior urinary function was 2.67 (1.7-4.1, p<0.001) for definitive radiotherapy. (Table 4.) Multivariable regression analysis was also performed for the 'reverse bifecta' as a dichotomy variable. The odds ratio of having both poor urinary and sexual performance was 0.29 (0.14-0.58, P<0.001) for primary radiotherapy when adjust to the same variables. (Table 4).

Discussion
In this multivariable analysis, one-year urinary domain quality of life scores after prostatectomy with salvage/adjuvant radiotherapy were significantly worse compared to patients undergoing primary radiotherapy. An estimate of the clinical significance of this discrepancy using the standard deviation and distribution approach revealed a 6.5-point difference in the urinary domain which has been previously shown to be a clinically important difference 8 .
While the overall urinary function was indeed worse in the surgery, when dividing this domain into incontinence and irritate/obstruction domains the results were different. In the incontinence domain as well documented, primary radiotherapy was significantly better, as radiation is more forgiving regarding sphincter function than surgical reconstruction. In the obstruction and irritation domain, there was no difference between the cohorts as radiation was delivered in both cohorts. While surgery alone is associated with less urinary irritation than radiotherapy, patients 11 who undergo both modalities will more likely suffer from both incontinence and irritation [11][12][13] .
Despite the fact that the cohort of primary radiotherapy was older and the use of ADT was higher as compare to the surgery arm, there was no significantly difference in sexual performance. Furthermore, most importantly for patient education prior to choice of treatment for high risk prostate cancer, the odds of achieving both worse sexual and urinary function were threefold for the multimodality radical surgery and radiation cohort versus radiotherapy alone. This finding was sustained after adjusting for age, diabetes, and risk group. This is despite a significantly worse hormonal score in the primary radiotherapy cohort due to a higher rate of ADT.
Bowel function score was marginally worse in the primary radiotherapy arm, possibly due to a higher radiation dose to the rectum in primary radiotherapy arm compared to salvage radiotherapy. This difference was lower than the distribution-based minimally important difference.
Few studies comprehensively address urinary function following post prostatectomy radiotherapy. A literature search, reveals retrospective studies that compared surgery alone to post prostatectomy radiation. A quality of life decrement of 10% was reported when radiation was added to prostatectomy, which is further reduced by androgen deprivation therapy 14 . The authors suggest that patients with high risk disease "should be counseled before RP on the potential net impairment of functional outcomes arising from multimodal treatment." An odds ratio of 1.6 for urinary incontinence has been reported for the combination compared to surgery alone 15 . A limitation of most studies is that urinary incontinence was reported by the number of pads used in 24 hours rather than a comprehensive patient reported domain for urinary incontinence [14][15][16] . In particular, no other study compared quality of life outcomes of surgery with adjuvant/salvage radiation versus primary radiation therapy with or without androgen deprivation therapy.
There are a number of limitations to the current study. Firstly, pre-treatment baseline QOL measurement was not collected in the surgical cohort, thus precluding a report of longitudinal quality of life outcomes or a comparison with a surgery alone cohort. However, the one year mean scores for EPIC in our study were comparable with data from Sanda et al. 17 with identical population mean scores at one year follow up for primary radiotherapy for urinary incontinence and the sexual domains, thus supporting the generalizability of our findings. The historical urinary incontinence score of the surgery alone cohort was 78 in that study versus 68 for the combined surgery and radiotherapy cohort 17 and 86.4 for the radiation only cohort in this study, thus supporting a clinically significant difference that we detected. We are currently working to improve the pre-treatment capture of patient reported outcome measures. Secondly, it may be argued that a one-year time point following radiotherapy does not capture all late effects. However one year status of urinary outcomes is fairly predictive of later outcomes and very late onset of incontinence in patients with good outcomes at one year is uncommon 18 . Thirdly, only one quarter of patients in the database were requested to complete surveys selected by the presence of the research associate on one of four clinic days. Thus while it may be argued that a capture rate of approximately one quarter of the patients may introduce a bias, this was not due to a patient selection bias as 95% of patients asked to fill surveys complied. Furthermore, the capture rate was similar in both the study 13 group and the control arm, thus confirming an absence of selection bias and the risk of a skewed view of outcomes.
This study provides important information for both patients and physicians asked to guide choice of treatment since it has a number of important strengths; Firstly, these are patient reported outcomes and not physician reported toxicity, in a sizeable cohort of real life subjects without exclusion criteria. Second, the statistical methods used in this paper were multivariable logistic regression with robust consideration of multiple confounders. We utilized clinical significant difference metrics, and while this approach is more common in a longitudinal study, it supports a significant difference in population based mean scores. Finally, it is becoming increasingly important to report patient reported outcomes specific to an institution as radiotherapy volumes and techniques vary considerably, particularly in the post prostatectomy setting. Patient reported outcomes from a specific institution can provide a realistic expectation based on the local clinical expertise and thus is more pertinent than outcomes reported from other centers 19 .
In conclusion, surgery with adjuvant/salvage RT was associated with significantly worse patient reported urinary continence outcomes at one-year post treatment, lower odds of achieving perfect urinary continence and higher odds of both unfavorable urinary and sexual performance. Patients referred for surgery with a high probability of requiring adjuvant or salvage radiotherapy should be informed regarding the impact on urinary and sexual performance due to the composite effect of two radical treatments, surgery and radiotherapy. These patients should always 14 be referred to an oncologist to discuss the option of primary radiotherapy with or without ADT, which was associated with more favorable 'bifecta' in this study.
These findings required further investigation with longer follow-up and longitudinal quality of life outcomes.    Definitive radiotherapy 0.29 (0.14-0.58) <0.001*