Prostate cancer patient characteristics associated with a strong preference to preserve sexual function and receipt of active surveillance

Abstract

BACKGROUND:

Men with early-stage prostate cancer have multiple options which have similar oncologic efficacy, but vary in terms of their impact on quality of life. In low-risk cancer, active surveillance is the option which best preserves patients’ sexual function, but it is unknown if patient preference affects treatment selection. Our objectives were to identify patient characteristics associated with a strong preference to preserve sexual function, and to determine whether patient preference and baseline sexual function level are associated with receipt of active surveillance in low-risk cancer.

METHODS:

In this population-based cohort of men with localized prostate cancer, baseline patient-reported sexual function was assessed using a validated instrument. Patients were also asked whether preservation of sexual function was very, somewhat, or not important. Prostate cancer disease characteristics and treatments received were abstracted from medical records.

RESULTS:

Among 1,194 men, 53% indicated a strong preference for preserving sexual function. Older men were less likely to have a strong preference (adjusted risk ratio [aRR] 0.98 per year, 95% CI 0.97–0.99), while men with normal sexual function were more likely (vs. poor function, aRR 1.59, 95% CI 1.39–1.82). Among 568 men with low-risk cancer, there was no clear association between baseline sexual function or strong preference to preserve function with receipt of active surveillance. However, strong preference may differnetially impact those with intermediate baseline function vs poor function (interaction p=.02).

CONCLUSIONS:

Treatment choice may not always align with patients’ preferences. These findings demonstrate opportunities to improve delivery of patient-centered care in early prostate cancer.

INTRODUCTION:

Prostate cancer is the most common non-skin cancer among men, with more than 180,000 new cases diagnosed each year in the US.[1] Over 90% of newly-diagnosed patients have localized disease and are potentially curable.[2, 3] These patients often have several treatment options available – which can include active surveillance, radical prostatectomy or radiation therapy – which have been shown to result in similar survival[4] but different quality of life effects in terms of sexual, urinary, and bowel function.[5–13]. Radical prostatectomy is associated with worse sexual function and urinary incontinence compared to active surveillance; external beam radiotherapy and brachytherapy with worse short-term urinary obstruction and irritation; and external beam radiotherapy with worse short-term bowel symptoms.[14] Therefore, it is well recognized that treatment decisions for localized prostate cancer are preference-sensitive, and need to account for each patient’s goals and priorities.[15, 16]

Specifically, different treatment options can have different magnitudes of impact on a patient’s sexual function.[13, 14, 17] For many men diagnosed with localized prostate cancer, choosing a treatment option that better allows preservation of sexual function may be prioritized. On the other hand, as the median age of prostate cancer diagnosis is 66 years,[1] some men may already have significant baseline sexual dysfunction and clinically-meaningful function preservation may not be a high priority. It seems logical that part of the treatment decision-making process between the patient and his physicians would be consideration of his baseline sexual function level and the relative priority he places on preserving this function. For men with low risk prostate cancer which is unlikely to be life threatening, active surveillance is the best option to preserve sexual function.

In a population-based, socio-demographically diverse, prospective cohort of newly-diagnosed prostate cancer patients, baseline (pre-treatment) sexual function was assessed using a validated quality of life instrument, and each patient also indicated whether preserving sexual function was of high priority. In this study, we sought to examine patient characteristics (including age, race, marital status, baseline sexual function, and prostate cancer aggressiveness) associated with a high priority for preserving sexual function. In addition, in low-risk patients, we examined whether baseline sexual functional level and patient preference were associated with receipt of active surveillance – to empirically assess whether these factors impacted treatment decisions.

METHODS:

PATIENT POPULATION:

The North Carolina Prostate Cancer Comparative Effectiveness & Survivorship Study (NC ProCESS) is a population-based cohort of prostate cancer patients, enrolled prior to treatment and followed prospectively. This study was approved by the University of North Carolina institutional review board. Details of patient identification and enrollment were described previously.[18] Briefly, patients with newly-diagnosed prostate cancer were identified through Rapid Case Ascertainment of the North Carolina Central Cancer Registry from all NC counties between 2011 and 2013. Upon identification, the patient was mailed an introductory letter describing the study. This was followed by a phone call from the study team; if a patient agreed to participate, a baseline survey was conducted at the same time (detailed below). The median time from date of diagnosis to study enrollment and baseline survey was 5 weeks (range: 3.1–18.9 weeks). Patients who had already initiated any prostate cancer treatment were ineligible. Among all eligible participants, 57% agree to participate and were enrolled.

DATA COLLECTION:

A baseline/pre-treatment survey was conducted prior to treatment for all enrolled patients. Demographic information was collected by patient report, and included date of birth, race, marital status, educational attainment, and health insurance.

Baseline sexual function was assessed using the validated Prostate Cancer Symptom Indices (PCSI).[5, 6, 17, 19–21] The Sexual Dysfunction domain of the PCSI includes 5 questions which assess patient erectile function (firmness, difficulty getting and keeping erections), ejaculation, and orgasm. Answers to these questions are transformed to an overall scale score from 0 (no dysfunction) to 100 (maximum dysfunction). In addition to a numerical score, the PCSI has also previously validated categorizing patients into 3 levels of function: normal, intermediate, or poor.[6, 17] As defined by the instrument, normal function describes a patient with essentially no sexual dysfunction or distress. Intermediate function describes patients with abnormal but useful sexual function.[6] Patients with poor sexual function have no useful function or have had no sexual activity within the past 4 weeks.

In addition, patients were asked the question “Preserving your ability to have sex. Would you say this is…” – and response choices were very important, somewhat important, or not important. For purposes of data analysis, we considered patients who responded “very important” to have a strong preference for preserving sexual function, and those who indicated “somewhat” or “not important” to have a lower preference for function preservation.

Prostate cancer diagnosis information (prostate-specific antigen [PSA] level, Gleason score, and clinical stage) and first course of treatment received within 1 year of diagnosis were abstracted from medical records. Only patients with early stage (non-metastatic) prostate cancer were included for analysis. Patients who pursued initial active surveillance but subsequently received treatment were counted as “active surveillance,” as were patients who did not receive any treatment within 1 year of diagnosis. Prostate cancer risk group was categorized using definitions from the National Comprehensive Cancer Network (NCCN).[22]

STATISTICAL ANALYSIS:

One goal of this study was to examine characteristics associated with patients who indicated a strong preference to preserve sexual function in their treatment decision-making. We compared the baseline sexual function scores of patients who indicated strong preference vs. lower preference for preserving sexual function. In addition, multivariable regression was used to examine additional factors associated with a strong preference for sexual function preservation. Covariates examined included baseline sexual function level, age, marital status, race, NCCN risk group, and education. Specifically, modified Poisson regression model with robust standard errors was used to compute adjusted risk ratio (aRR) estimates.

Another study goal was to empirically assess whether patient preference impacted treatment received. Specifically, in low-risk prostate cancer patients, active surveillance is a standard option and best preserves sexual function.[23] Therefore, we hypothesized that low-risk patients with a strong preference for preserving function, and those with normal baseline sexual function, would be most likely to receive active surveillance. Stratifying patients into 6 groups based on baseline sexual function level (normal, intermediate, poor) and patient preference to preserve function (strong preference vs. lower preference), we examined whether use of active surveillance differed by group. Our hypothesis was formally tested using multivariable modified Poisson regression analysis, which examined whether patient preference and baseline sexual function level were associated with receipt of active surveillance and aRR was reported. Other covariates included marital status, age, and race.

All statistical analysis was performed using SAS version 9.4 (Cary, NC), and P < 0.05 was considered statistically significant.

RESULTS:

PATIENT CHARACTERISTICS AND PREFERENCE FOR PRESERVATION OF SEXUAL FUNCTION:

This population-based cohort of 1194 patients is socio-demographically diverse, with 28% overall being non-white, and 32% with high school or less education. A total of 628 men (53%) indicated a strong preference for preserving sexual function, and 566 patients (47%) indicated a lower preference (Table 1).

Table 1.

Demographic and clinical characteristics of study participants

Variable	Importance of preserving sexual function		P-value
	Strong Preference (N=628)	Lower Preference (N=566)
Age, Median (Range), years	63 (42-80)	67 (41-81)	< 0.001
NCCN risk category, N (%)			0.2
Low	318 (51)	262 (47)
Intermediate	226 (37)	207 (37)
High	76 (12)	86 (16)
Race, N (%)			< 0.001
White	407 (65)	453 (80)
African-American	202 (32)	97 (17)
Other	19 (3)	15 (3)
Married, N (%)	483 (77)	468 (83)	0.02
Education, N (%)			0.004
High school or less	228 (36)	155 (27)
Some college	168 (27)	176 (31)
College graduate	232 (37)	235 (42)
Insurance, N (%)			< 0.001
Medicare	274 (44)	342 (60)
Medicaid	28 (4)	10 (2)
VA	43(7)	15 (3)
Private	256 (41)	185 (33)
None	27 (4)	14 (2)
Baseline sexual dysfunction score, Mean (SD)	34.5 (36)	53.5 (40)	< 0.001
Baseline sexual function category, N (%)
Normal	227 (36)	116 (21)	< 0.001
Intermediate	184 (30)	130 (23)
Poor	213 (34)	317 (56)

Abbreviations: NCCN, National Comprehensive Cancer Network; SD, standard deviation; VA, Veteran’s Affairs

Patients who indicated a strong preference for sexual function preservation had significantly better baseline function (lower sexual dysfunction score, mean 34.5) than patients who indicated a lower preference for function preservation (mean 53.5, P < 0.001). In addition, examining baseline sexual functional levels in the two patient groups revealed a more nuanced picture. Among patients who indicated a strong preference to preserve sexual function, 36% had normal baseline function but 34% actually had poor function. Conversely, 21% of patients who indicated a lower preference for function preservation had normal baseline function.

Multivariable analysis confirmed the finding that poor baseline sexual function was associated with lower likelihood of patients indicating a strong preference for function preservation (Table 2). Older patients were also less likely to indicate a strong preference to preserve sexual function (aRR = 0.98 per year; P < 0.001), as were white patients (aRR = 0.77; P < 0.001) and men with high-risk prostate cancer (compared to low-risk prostate cancer; aRR = 0.84; P = 0.06). Results specifically for the low-risk patient subgroup were similar (Appendix Table 1 and 2). Sensitivity analysis re-categorizing patients as “very important” and “somewhat important” vs “not important” also showed age and baseline sexual function to be significantly associated with a stronger preference to preserve function (Appendix Tables 3 and 4).

Table 2.

Multivariable regression model assessing factors associated with a strong preference to preserve sexual function

Variable	Adjusted Risk Ratios (95% CI)	P-value
Age (years)	0.98 (0.97-0.99)	<0.001
Baseline sexual function
Normal vs. poor function	1.59 (1.39-1.82)	<0.001
Intermediate vs. poor function	1.39 (1.21-1.59)	<0.001
Married vs. unmarried	0.89 (0.79-1.00)	0.06
Race (White vs. non-white)	0.77 (0.68-0.86)	<0.001
NCCN risk group
Intermediate vs. low	0.94 (0.84-1.04)	0.23
High vs. low	0.84 (0.71-1.00)	0.06
Education
High school or less vs. college graduate	1.22 (1.07-1.39)	0.01
Some college vs. college graduate	0.96 (0.84-1.10)	0.59

Abbreviations: CI, Confidence Interval; NCCN, National Comprehensive Cancer Network

TREATMENT PATTERNS

Table 3 summarizes the treatment patterns of 568 patients with low-risk prostate cancer, stratified by baseline sexual function level and preference for function preservation. Overall, 43% of low-risk patients received active surveillance. However, we found no evidence to support our hypothesis of a higher use of active surveillance in patients with a strong preference to preserve their sexual function. Among patients with normal baseline sexual function and a strong preference to preserve this function, 39% received active surveillance; the proportions of patients who received active surveillance was actually higher for patients who had normal baseline sexual function but indicated lower preference to preserve function (49%). A similar finding was seen for patients with poor baseline sexual function, where again a numerically higher proportion of patients who indicated a lower preference to preserve sexual function received active surveillance.

Table 3.

Treatment received among low-risk patients based on patient preference and baseline sexual function (N=568)

Baseline Sexual Function	Preference to preserve sexual function	AS	RT	RT + ADT	RP	RP + ADT	ADT alone	Other
Normal	Strong (N=120)	39%	21%	3%	36%	0%	2%	0%
	Lower (N=63)	49%	21%	2%	27%	2%	0%	0%
Intermediate	Strong (N=97)	51%	15%	1%	33%	0%	0%	0%
	Lower (N=52)	33%	25%	2%	38%	0%	2%	0%
Poor	Strong (N=94)	37%	34%	6%	22%	0%	0%	0%
	Lower (N=142)	49%	22%	4%	22%	0%	1%	2%

Abbreviations: ADT, Androgen Deprivation Therapy; AS, Active Surveillance; NCCN, National Comprehensive Cancer Network; RP, Radical Prostatectomy; RT, Radiotherapy (includes external beam radiotherapy and brachytherapy)

We formally tested our hypothesis using multivariable analysis, examining factors associated with active surveillance in low-risk patients (Table 4). Older patients were significantly more likely to undergo active surveillance (aRR = 1.04 per year; P < 0.001). Interaction terms examined whether patient preference differentially impacted patients with different baseline sexual function levels in their “risk” of receiving active surveillance. Results suggest that patients with a strong preference to preserve sexual function are more likely to receive active surveillance if they have intermediate baseline function than poor baseline function. However, preference to preserve sexual function did not differentially impact patients with baseline normal vs those with poor function.

Table 4.

Multivariable regression model assessing factors associated with active surveillance in patients with low-risk disease

Variable	Adjusted Risk Ratios (95% CI)	P-value
Importance of preserving sexual function (Strong vs. lower preference)	0.80 (0.58-1.10)	0.17
Baseline sexual function
Normal vs. poor function	1.18 (0.87-1.60)	0.29
Intermediate vs. poor function	0.76 (0.50-1.17)	0.22
Married vs. unmarried	1.13 (0.87-1.46)	0.35
Age (years)	1.04 (1.02-1.06)	<0.001
Race (White vs. non-white)	0.77 (0.61-0.97)	0.03
Preference × normal vs poor	1.08 (0.69-1.71)	0.74
Preference × intermediate vs poor	1.88 (1.10-3.21)	0.02

Abbreviations: CI, Confidence Interval

DISCUSSION:

In this population-based and diverse cohort of men with localized prostate cancer, we report several notable findings which add to the published literature. First, more than half of patients with newly-diagnosed prostate cancer had a strong preference to preserve sexual function. Second, not surprisingly, younger patients and those with better baseline sexual function were more likely to indicate a strong preference for preserving sexual function. Third, while 43% of low-risk patients received active surveillance, there was not a clear pattern of patient’s baseline function or preference for sexual function preservation being associated with receipt of active surveillance. To the best of our knowledge, this study is the first to assess whether patient preference for sexual function preservation matched treatment received in a large, population-based and diverse cohort of prostate cancer patients. Prior studies [24–27] along with ours provide insights into the prostate cancer treatment-making process.

Prostate cancer is unique in that many patients with localized disease have multiple options which have similar oncologic efficacy but differ significantly in terms of their impact on quality of life.[5–13] Thus, decision-making for men with early prostate cancer is a preference sensitive process, and should taking into account a patient’s priorities. For low risk patients, active surveillance results in similar overall survival compared to radiation and surgery, and better preserves sexual function.[4, 23] In light of this, physicians should inform their low-risk patients that active surveillance would best meet their preference for preserving sexual function without compromising life expectancy. However, it is unknown whether patient preference is elicited by their physicians or taken into account in final treatment selection. With this background, there are several findings which warrant further discussion.

First, while we found that better baseline sexual function was associated with a higher likelihood of men indicating a strong preference for function preservation, there are important exceptions. As Table 1 detailed, one-third of men with normal baseline sexual function indicated a lower preference for functional preservation, while 40% of men with poor baseline function indicated a strong preference. This finding suggests that simply ascertaining a patient’s baseline sexual function level is insufficient for determining patient treatment goals, because physicians cannot assume that patients with good sexual function will always place a high priority on preserving this function. To allow patients to make an informed decision on treatment, physicians need to directly ask patients about their priorities during the decision-making process; including preferences for sexual function, other quality of life impacts (e.g., urinary function), and aggressiveness of cancer treatment. Our multivariable analysis showed that additional factors associated with patient preference included patient age and also cancer aggressiveness.

This study also showed a disconnect between patients’ preference for sexual function preservation and treatment received. One possible explanation is that physicians may not routinely elicit patient preferences in making treatment recommendations.[28–30] Prior studies have shown that prostate cancer treatment decisions are associated with physician specialty,[31, 32] and patients who are younger and healthier are more likely to receive surgery, while older patients more likely to receive radiotherapy or no treatment. Our findings are consistent with this, and showed that older age was significantly associated with active surveillance. Whereas age is a factor which is relatively easy to ascertain during a clinic visit, thoroughly assessing patients’ baseline sexual function level and preferences is more time consuming, and there may also be other cultural barriers to accurately attaining this information in the clinical setting. The development and use of decision aids in prostate cancer can help systematically clarify patient priorities as well as educate patients about treatments that align with their goals. Several analyses found that decision aids increase patient knowledge and involvement throughout the decision-making process,[33] thus facilitating the delivery of high-quality, patient-centered care. Indeed, patient knowledge about active surveillance and being fully informed about expectations are important factors in patient treatment selection.[34]

This study has several strengths and potential limitations. A strength is the population-based design of modern patients, which yielded a cohort that is sociodemographically diverse and more representative than institutional reports. However, this study was conducted in one East Coast state and may not be representative of other states or regions, and the 57% response rate in this study could further limit generalizability of results. This study is also unique in assessing patient-reported sexual function prior to treatment, as well as patients’ preference to preserve this function. We are unware of other large population-based cohorts that contain this information. However, the latter question was created specifically for this study, as no validated assessment was available in the published literature. Furthermore we did not quantify patients’ preferences or their understanding of their preferences directly – and therefore a definitive conclusion about lack of congurency in patient preferences with treatment selection cannot be made. It is also possible that the expressed preference for maintaining sexual function captured in this study could reflect physician-patient interaction about treatment options and consequences. Another potential limitation is that while a patient may have indicated a strong preference to preserve sexual function, he may have an even stronger desire for aggressive treatment or have anxiety related to active surveillance – and we did not capture the latter information in this study. However, low risk prostate cancer is unlikely to be life threatening. For patients with these early diagnoses who express concurrent preferences regarding maximizing survival and preserving sexual function, and if indeed these preferences were ascertained by their physicians, informed decision-making should include clarification that choosing active surviellance would be concordant with patient preferences without compromising survival outcomes. Finally, this study did not survey physicians so we assessed the patients’ treatment selection but are unable to directly make conclusions about physician recommendations. The physician plays an important role in treatment selection and recommendations vary considerably by specialty.[32] Furthermore, certain specialties may be more or less attentive to the preferences of their patients for preservation of sexual function.

In this population-based cohort of 1,194 men with newly-diagnosed localized prostate cancer, more than half (53%) indicated a strong preference to preserve sexual function in their treatment decision-making process. However, among men with a low-risk diagnosis, there was no clear association between patient preference or baseline sexual function level with receipt of active surveillance. Although many men with low risk disease report preserving sexual function is a high priority, many do not receive active surveillance which is the treatment most aligned with this preference. These findings call for improved efforts for delivering high-quality, patient-centered care in this common disease.