Patient-Reported Sexual Quality of Life (QOL) after different types of radical prostatectomy (RP) and radiation treatments, analysis of a population-based prospective cohort

Abstract

Background

Prostate cancer patients face many treatment options, but the comparative effects of different surgical and radiotherapy (RT) options on sexual function are unclear.

Methods and Materials

Population-based cohort of 835 men with newly-diagnosed prostate cancer from 2011–2013 were recruited throughout North Carolina in collaboration with the Rapid Case Ascertainment system of the state cancer registry. All men were enrolled prior to treatment and followed prospectively with the validated instrument Prostate Cancer Symptom Indices (PCSI). This analysis compares the Sexual Dysfunction scores of the PCSI among patients who received: external beam RT (EBRT), EBRT with androgen deprivation therapy (ADT), brachytherapy, nerve-sparing radical prostatectomy (RP), and non-nerve sparing RP. Propensity scores were used to balance patient-characteristics across groups, and multiple-imputation for missing data.

Results

EBRT and brachytherapy resulted in similar PCSI scores through 24 months. Compared to EBRT, EBRT with ADT and RP with or without nerve-sparing had worse PCSI scores at all post-treatment time points. Preservation of useful sexual function at 24 months was associated with treatment type, baseline score, and age. Predicted preservation rates were: EBRT (14.1-70.7%), EBRT with ADT (8.4-52.3%), nerve-sparing RP (4.7-45.3%) and non-nerve-sparing RP (4.8-34.5%).

Conclusion

RT alone results in the best sexual function preservation, and brachytherapy provides similar outcomes. RT with ADT and nerve-sparing RP yielded similar outcomes, while non-nerve sparing RP results in the worst sexual function. These results help patients make decisions among the specific types of surgery and RT they face based on each individual’s diagnosis.

Introduction

Prostate cancer is one of the most commonly diagnosed cancers affecting men in developed nations (1). The majority of patients with prostate cancer have localized disease and often have excellent survival outcomes with radical prostatectomy (RP) or radiation therapy (RT) (2). Given that most patients become long-term survivors, considerations regarding quality of life (QOL) is important in the treatment decision-making process. Specifically, many men with prostate cancer consider preserving sexual function to be a high priority (3).

Both RP and RT are known to negatively impact sexual function, and some prior studies have shown that RP has an overall bigger negative impact than RT (4–6). However, nerve-sparing RP is known to improve function preservation compared to the older non-nerve sparing surgery (7); and adding androgen deprivation therapy (ADT) to RT is known to further worsen sexual function. The comparative impact of external beam RT vs brachytherapy on sexual function outcomes is also unknown.

In a population-based cohort of men with newly-diagnosed prostate cancer enrolled before treatment, prospective QOL was collected. The goal of the current study is to compare 2-year patient-reported sexual function outcomes after external beam RT (EBRT), EBRT with ADT, brachytherapy, RP with nerve-sparing and RP without nerve-sparing.

Methods and Materials

Patient Population

Patients were enrolled into the North Carolina Prostate Cancer Comparative Effectiveness & Survivorship Study (NC ProCESS) cohort from 2011–2013 in collaboration with the Rapid Case Ascertainment system of the North Carolina Central Cancer Registry. Rapid Case Ascertainment is a unique part of North Carolina cancer registry where newly-diagnosed cancer patients from across the state are identified for research studies in as little as one week after diagnosis. This allowed all patients of this cohort to be enrolled, and baseline data collected including QOL, before any treatment. Further details of patient identification and enrollment have been described previously (8). To our knowledge, this is the only population-based cohort of prostate cancer patients where all baseline data were collected before treatment. All patients provided written informed consent and this study was approved by the University of North Carolina institutional review board.

Data Collection

Patient-level demographics including race, marital status, insurance, education, and household income were obtained from patient report on baseline survey. Age at diagnosis and treatment details were obtained through cancer registry data supplemented by medical record abstraction. Nerve-sparing status was obtained from the surgeon’s medical/operative reports.

Outcome Assessment

QOL was assessed using the validated instrument Prostate Cancer Symptom Indices (PCSI) (9) at baseline (pretreatment), and 3, 12, 24 months after treatment. The Sexual Dysfunction domain of the PCSI includes 5 items assessing arousal, erection quality, ability to maintain erection, and ability to orgasm and ejaculate. Responses to these items are used to calculate an overall domain score from 0 to 100, with a higher score indicating more (worse) dysfunction. In addition to numerical QOL scores, which are often difficult to interpret clinically (10), the PCSI instrument is unique in having a scoring algorithm that provides presentation of results using its defined levels of sexual function (normal, intermediate, poor) (11–12). In summary, normal function represents a patient with essentially no dysfunction or distressful symptoms (eg, a patient with no issues with sexual desire or erection) (13). Intermediate function represents a patient with at least 1 distressful symptom but none very distressful (eg, a patient with intermittent issues with arousal but no severe dysfunction). Patients with normal or intermediate sexual function have sufficient function for intercourse. Poor function represents patients with no useful function (13).

Data Analysis

Baseline patient characteristics by treatment type are summarized. The goal of this study is to compare sexual function outcomes among patients who received external beam RT alone (control group), RT with ADT, brachytherapy, RP with nerve-sparing, and RP without nerve-sparing. To balance the distribution of the observed baseline characteristics across groups, propensity scores (PS) weighting was used. Propensity scores were estimated using logistic regression that incorporated age at diagnosis, race, marital status, educational level, health insurance status, and household income level. Four separate binary logistic regression models were run to compute propensity scores; RT alone (control group) was contrasted in each model against one of the other four treatment types. Propensity score odds from the respective regression model were used to assign weights to subjects receiving the other four treatment types whereas subjects receiving RT alone were not weighted (14). Standardized differences were calculated to assess and verify that the balancing was adequate across the different treatment groups (15).

Missing data for post-treatment sexual dysfunction scores and functional levels were multiply imputed (MI) using the fully conditional specification approach as previously described (4). Propensity-weighted sexual dysfunction scores and 95% confidence interval (CI) at baseline, 3, 12, and 24 months were computed for each patient and presented by treatment type. To account for non-independence because of propensity weighting, robust standard errors were used for computing 95% CIs. Post-treatment estimates were pooled over all 100 MI datasets accounting for uncertainty due to imputation.

Further, logistic regression was used to examine the association between treatment modality and preservation of useful sexual function (either normal or intermediate) (13) at 24 months after treatment, adjusting for covariates including baseline sexual dysfunction, race, age at diagnosis, and marital status. First, on each of 100 MI datasets, regression analysis was performed on the entire sample. Then, parameter estimates and odds ratios were pooled across all MI datasets and CI for these statistics were computed accounting for uncertainty due to imputation.

To provide data that can be more useful to patients and physicians, we created figures to show the probability of preserving useful sexual function in surgery and radiation patients at 24 months, based on patients’ baseline (pretreatment) scores. Probabilities computed from the pooled parameter estimates were fitted using penalized cubic-spline.

All tests used a 2-sided p<0.05 for statistical significance. All statistical analyses were performed using SAS version 9.4.

Results

Baseline characteristics

Among the 835 patients, 188 received EBRT (22%), 99 EBRT with ADT (12%), 122 brachytherapy (15%), 84 RP without NS (10%), and 342 received RP with NS (41%) (Table 1). Only six brachytherapy patients received ADT (4.9%). The cohort was sociodemographically diverse with 31% of patients being non-white, 34% having a high school education or less, and 40% having an annual household income less than $40,000. Patients who received RP were younger than patients who received any of the radiation modalities. Differences in baseline demographic characteristics were minimized with propensity weighting. After propensity weighting, median age was 65.8 to 67.4 across all groups.

Table 1:

Baseline Patient Characteristics

	Before Propensity Weighting, N (%)						After Propensity Weighting, %
	All Patients N=835	EBRT N=188	EBRT + ADT N=99	BT N=122	RP, No NS N=84	RP + NS N=342	EBRT	EBRT + ADT	BT	RP, No NS	RP + NS
Age at Diagnosis (mean, years)	63.7	66.4	67.5	64.6	63.8	60.8	66.4	66.6	67.4	65.9	65.8
Race  White  AA/Other	572 (69%) 263 (31%)	123 (65%) 65 (35%)	47 (47%) 52 (53%)	88 (72%) 34 (28%)	57 (68%) 27 (32%)	257 (75%) 85 (25%)	65% 35%	66% 34%	67% 33%	68% 32%	69% 31%
Married	649 (78%)	136 (72%)	69 (70%)	104 (85%)	58 (69%)	282 (82%)	72%	69%	70%	65%	69%
Education  College Grad  Some College  HS or less	309 (37%) 239 (29%) 287 (34%)	66 (35%) 52 (28%) 70 (37%)	22 (22%) 28 (28%) 49 (50%)	40 (33%) 37 (30%) 45 (37%)	27 (32%) 24 (29%) 33 (39%)	154 (45%) 98 (29%) 90 (26%)	35% 28% 37%	35% 27% 38%	39% 27% 34%	35% 24% 41%	38% 27% 35%
Insurance  Private  Medicare  Other	332 (40%) 340 (41%) 163 (19%)	51 (27%) 97 (52%) 40 (21%)	19 (19%) 56 (57%) 24 (24%)	43 (35%) 59 (48%) 20 (17%)	30 (36%) 37 (44%) 17 (20%)	189 (55%) 91 (27%) 62 (18%)	27% 52% 21%	28% 49% 23%	25% 55% 20%	27% 53% 20%	26% 55% 19%
Income  >90,000  70,001–90,000  40,001–70,000  ≤40,000	169 (20%) 102 (12%) 228 (27%) 336 (40%)	31 (17%) 19 (10%) 47 (25%) 91 (48%)	6 (6%) 6 (6%) 24 (24%) 63 (64%)	14 (11%) 14 (11%) 41 (34%) 53 (44%)	11 (13%) 12 (14%) 23 (28%) 38 (45%)	107 (31%) 51 (15%) 93 (27%) 91 (27%)	17% 10% 25% 48%	13% 13% 26% 48%	18% 8% 26% 48%	18% 8% 23% 51%	22% 9% 22% 47%
Risk Group  Low-Risk  Intermediate  High-Risk	308 (37%) 381 (46%) 140 (17%)	67 (36%) 105 (56%) 14 (8%)	11 (11%) 35 (36%) 52 (53%)	83 (69%) 34 (28%) 3 (3%)	24 (29%) 34 (41%) 25 (30%)	123 (36%) 173 (51%) 46 (13%)	36% 56% 8%	14% 37% 49%	62% 33% 5%	28% 40% 32%	36% 46% 18%
Sexual Dysfunction Score (mean)	42.4	50.0	55.1	45.0	46.8	32.5	49.9	50.2	51.2	47.6	46.3

Comparison groups are:

EBRT: external beam radiotherapy

EBRT + ADT: external beam radiotherapy with androgen deprivation therapy

BT: brachytherapy

RP, no NS: radical prostatectomy without nerve sparing

RP + NS: radical prostatectomy with nerve sparing

Abbreviations: AA: African American; HS: High School.

Sexual Dysfunction

Propensity score-weighted mean sexual dysfunction scores for the 5 groups at baseline and 3, 12, and 24 months post-treatment are shown in Table 2. At baseline and 3, 12, and 24 months post-treatment, sexual dysfunction scores were present for 835/835 (100%), 712/835 (85%), 662/835 (79%) and 587/835 (70%) of patients. For patients who received EBRT alone (control group), sexual dysfunction scores increased (worsened) after treatment. The difference in sexual dysfunction score at 24 months compared to baseline was 9.5 points.

Table 2:

Propensity-Weighted Mean Sexual Dysfunction Scores (higher score = worse)

	Sexual Dysfunction Score Mean Score (95% CI)
Treatment	Baseline	3 months		12 months		24 months
		Score	Mean Difference vs EBRT	Score	Mean Difference vs EBRT	Score	Mean Difference vs EBRT
EBRT	49.9 (44.4–55.5)	59.5 (53.6–65.3)	---	54.4 (48.5–60.4)	---	59.4 (53.2–65.5)	---
EBRT + ADT	50.2 (40.7–59.7)	78.2 (69.7–86.8)	18.8* (8.4–29.1)	78.3 (70.4–86.1)	23.8* (14.1–33.5)	74.1 (65.4–82.8)	14.7* (4.2–25.2)
Brachytherapy	51.2 (41.3–61.1)	62.7 (53.0–72.5)	3.3 (−8.1–14.6)	64.7 (57.0–72.5)	10.3* (0.8–19.8)	66.0 (57.8–74.3)	6.7 (−3.4–16.7)
RP + NS	46.3 (40.3–52.3)	80.8 (75.8–85.9)	21.4* (13.6–29.1)	76.5 (72.0–81.1)	22.1* (14.6–29.5)	74.4 (69.4–79.4)	15.0* (7.2–22.8)
RP: No NS	47.6 (37.9–57.2)	83.9 (77.2–90.5)	24.4* (15.6–33.2)	77.6 (69.9–85.3)	23.1* (13.5–32.8)	77.8 (69.9–85.6)	18.4* (8.7–28.2)

Comparison groups are:

EBRT: external beam radiotherapy

EBRT + ADT: external beam radiotherapy with androgen deprivation therapy

BT: brachytherapy

RP, no NS: radical prostatectomy without nerve sparing

RP + NS: radical prostatectomy with nerve sparing

^*Statistically significantly different compared to EBRT alone at the same time point.

Differences were observed between EBRT alone vs other treatment groups at all post-treatment time points. EBRT+ADT patients had worse sexual dysfunction scores compared to EBRT patients at 3 (between-group difference vs EBRT, 18.8 points [95% CI 8.4–29.1]), 12 (23.8 [95% CI 14.1–33.5]), and 24 month time points (14.7 [95% CI 4.2–25.2]).

Score differences between brachytherapy and EBRT alone were statistically significant at the 12 month time point only. RP with NS and RP without NS patients had statistically significantly worse scores compared to EBRT patients at 3, 12 and 24 months.

Multivariable logistic regression analysis confirmed these findings (Table 3). At 24 months compared with EBRT alone, EBRT+ADT, RP with NS and RP without NS were all associated with decreased odds of preserving useful (normal or intermediate) sexual function. In addition, older age and baseline sexual dysfunction score were also significantly associated with 24 month sexual function.

Table 3:

Multivariable Analysis of Factors Associated with Preserved Sexual Function

	OR (95% CI)	P-value
Intercept
Treatment (ref: EBRT)
EBRT + ADT	0.47 (0.23–0.94)	.03
Brachytherapy	0.84 (0.45–1.59)	.60
RP + NS	0.28 (0.17–0.48)	<.001
RP, no NS	0.23 (0.10–0.51)	<.001
Race (ref: white)
Non-white	0.94 (0.60–1.45)	.76
Age at diagnosis (per year)	0.94 (0.92–0.97)	<.001
Marital status (ref: not married)
Married	0.87 (0.53–1.43)	.58
Baseline Sexual Dysfunction Score (per point)	0.98 (0.97–0.98)	<.001

Comparison groups are:

EBRT: external beam radiotherapy

EBRT + ADT: external beam radiotherapy with androgen deprivation therapy

BT: brachytherapy

RP, no NS: radical prostatectomy without nerve sparing

RP + NS: radical prostatectomy with nerve sparing

Estimates are pooled over all multiply imputed data sets

We show the fitted probability of a patient preserving useful sexual function at 24 months after EBRT with/without ADT (Figure 1) or RP with/without NS (Figure 2) based on the patient’s baseline (pretreatment) score. These plots included patients who were ≤ 71 years of age because of the better prediction model fit in this younger subgroup. For patients who received EBRT alone, the probability of preserving useful sexual function varies from 14.1% to 70.7% depending on baseline sexual dysfunction score (Figure 1A). For patients who received EBRT with ADT, the probability of preserving useful sexual function varies from 8.4% to 52.3% (Figure 1B). Additionally, for patients who received RP with NS or RP without NS, the probabilities of preserving useful sexual function varies from 4.7% to 45.3% and 4.8% to 34.5%, respectively (Figure 2A & B).

Figure 1:

Probability of Preserving Useful Sexual Function given Baseline Function. External Beam Radiation Therapy Patients (Age ≤71) A) Model of EBRT Patients (No ADT) B) Model of EBRT + ADT Patients

Figure 2:

Probability of Preserving Useful Sexual Function given Baseline Function. Radical Prostatectomy Patients (Age ≤71) A) Model of RP + NS Patients B) Model of RP Patients (No NS)

Discussion

Standard treatment options for localized prostate cancer result in similar long-term oncologic outcomes but differ in their impact on quality of life. It is well known that for many patients, preserving quality of life is of high importance in the treatment decision-making process; specifically, prior studies have shown that over 50% of men consider preserving sexual function to be important (3, 16–17). However, the impact of different types of surgery and radiation treatments on sexual function has not been well-studied. To fulfill this knowledge gap, we analyzed data from a diverse, prospective population-based cohort of patients.

Results from this study are consistent with the prior literature, including the finding that adding ADT worsened sexual function outcomes in RT patients (18), and non-sparing RP was worse than nerve-sparing RP (19–24). However, to the best of our knowledge, direct comparison of sexual function outcomes between the five contemporary treatment groups included in this study has not yet been reported in the literature. In light of this, several notable findings from this study add to the existing literature.

First, this study shows that radiotherapy alone without ADT can worsen sexual function in some patients in the first 2 years after treatment. Perhaps more meaningful than the PCSI Sexual Dysfunction score over time, we report the predicted probability of having useful sexual function at 2 years after RT. Not surprisingly, this is dependent on the patient’s baseline PCSI score. Patients and physicians can use Figure 1, assess baseline sexual function, and determine the probability of sexual function preservation 2 years later. For example, in patients with excellent baseline function, there is a 70% probability of preserving useful sexual function. This study also showed similar results after EBRT and brachytherapy (Table 2) through the first 2 years, which is consistent with a prior study (25).

Second, while published studies have generally reported better sexual function after radiation therapy than RP, most prior studies including the ProtecT trial have not differentiated RT patients who received vs did not receive ADT, and RP patients by nerve-sparing vs not. Sexual function outcomes of neve-sparing RP vs RT with ADT is an especially relevant clinical question. This study (Table 3 and Figures 1–2) shows that RT with ADT results in similar, or slightly better, sexual function outcomes compared to nerve-sparing RP. The probability of preserving useful sexual function at 2 years is higher for RT with ADT patients by 3.7–5.0% (Figure 1–2). This comparison may be relevant for patients with unfavorable intermediate-risk prostate cancer who would likely receive ADT with RT, and for whom nerve-sparing surgery is often possible.

On the other hand, for patients with locally-advanced cancer for whom nerve-sparing surgery is not possible, non-nerve sparing RP results in the worst sexual function outcomes among the treatment options examined in this study. Predicted probability of useful sexual function at 2 years is only 4.8% to 34.5%, depending on baseline function.

There are several strengths and limitations of this study. Given its population-based design, the cohort was sociodemographically diverse and provides more generalizable results than other study designs including institutional studies and some randomized trials. Similar to other population-based studies, a slightly higher percentage of African American patients underwent RT compared to RP (5, 12, 18); however, this difference was not associated with differences in preservation of sexual function (Table 3). While this study is not randomized, and therefore residual confounding is possible despite propensity score methods, it is unlikely that a randomized trial will be conducted to compare these 5 treatments. In addition, observational studies comparing RT to RP patients often bias in favor of surgical patients due to their younger age and better baseline health; despite this, this study found better sexual function outcomes for RT patients. An additional strength is that, to our best knowledge, this is the only population-based prostate cancer patient cohort where all data were collected prospectively, including all baseline QOL data collected before treatment. This is important to minimize potential recall bias which could result if baseline data were collected after treatment. Another limitation is that the study did not provide sexual function outcomes beyond 24 months. However, based on the ProtecT trial, sexual function after RT or RP changed little between 2 and 6 years of follow-up (5). The median duration of ADT was 6 months in our study (59% of patients were on ADT for 6 months or less and 33% on ADT longer than 12 months). Since sexual function effects were not stratified by duration of ADT, this is an additional potential limitation. Detailed information on RT and brachytherapy dose/source are also lacking; however, it is unlikely that there was much deviation from standard of care. Finally, it is important to note that our study compares active treatment options, and active surveillance is another option for many patients with lower-risk disease to delay the onset of treatment-related dysfunction. As demonstrated by the ProtecT trial and Chen et al, sexual dysfunction also gradually worsens over time with active surveillance, and differences are not clinically meaningfully different compared to active treatment at 2 years (4–5).

In conclusion, in a population-based patient cohort, this study showed that RT alone results in the best sexual function preservation, and brachytherapy provides similar outcomes through 2 years after treatment. RT with ADT and nerve-sparing RP yielded similar outcomes, while non-nerve sparing RP results in the worst sexual function. These results provide tangible new information to help patients make decisions among the specific types of surgery and RT they face based on each individual’s diagnosis.