Background
Prostate cancer survival is high in developed countries due to the availability of prostate specific antigen testing and effective treatment options. The five-year relative survival for men with prostate cancer in Australia is approximately 96% [
1]. Given the high probability of long-term survival from prostate cancer, maintaining quality of life should also be central to clinical practice [
2]. However, men diagnosed with prostate cancer are faced with challenging decisions about treatment choices, due to the substantial risk of adverse effects with most recommended treatments. These can include post-treatment impact on urinary, bowel and sexual functioning [
3‐
8]. Understanding the negative effects of the treatments and how the treatments affect quality of life of men has become crucial for decision making [
6].
It is increasingly recognised that selecting appropriate treatments needs to be tailored to each man’s circumstances to minimize impacts on physical functioning and mental wellbeing. To this end, greater emphasis has been placed on collecting information from patients about their physical functioning (i.e., patient-reported outcome measures (PROMs)) to audit and evaluate various treatment outcomes [
2]. Because they are derived from the patient’s perspective, self-reported measures of functional outcomes such as erectile function and urinary continence are very important [
9]. Functional outcomes are often under-reported by physicians [
10] and PROMs can guide clinical practice to be more responsive to individual patients’ needs and inform ways in which patients can self-manage their condition [
2].
The Prostate Cancer Outcomes Registry Australia and New Zealand (PCOR-ANZ), which was established primarily to monitor and improve outcomes for men with prostate cancer [
11], collects and reports post-treatment PROMs data. However, this collection does not include baseline PROMs, due to logistical difficulties identifying men before they commenced treatment. Baseline PROMs assessment is important for interpreting post-treatment impacts, given the differing age and comorbidity profiles among those who undergo different treatment options. In addition, baseline PROMs are useful in identifying patients at risk of impaired function, to facilitate treatment decision-making by patients and clinicians [
12] and to avoid patients’ regret due to unexpected physical outcomes of treatment [
13].
Currently, Australian data on baseline functioning among men with prostate cancer are sparse. However, the long-standing South Australian Prostate Cancer Clinical Outcomes Collaborative (SA-PCCOC) database (which contributes core data to PCOR-ANZ) has collected PROMs data at baseline and sequential time points after diagnosis from enrolled participants over the past two decades. The aim of this study was to describe PROMs at baseline (pre-treatment) and after 12 months of follow-up, and to compare the extent to which functional outcomes are impacted by four common prostate cancer treatments— radical prostatectomy (RP), external beam radiation therapy (EBRT), brachytherapy and active surveillance—using PROMs data collected by SA-PCCOC.
Discussion
Comparing the adverse side effects of treatment options for prostate cancer is a high research priority. In this study, patient-reported functional outcomes and related bother at baseline and at 12 months post-treatment are described for a contemporary Australian cohort and functional outcomes at 12 months compared across different treatment options after accounting for baseline function, clinical and sociodemographic factors.
Our results indicate that baseline functioning differs between treatment groups, most notably in relation to levels of sexual function. The baseline sexual function score for EBRT in our cohort was very low compared with other cohorts [
3–
5]. This may be due to referral patterns with older more comorbid patients more likely to receive EBRT in our cohort. The largest change in function at 12 months, across all treatment groups, was seen for sexual functioning. Results of multivariable models confirmed RP’s greater impact on sexual function compared with other treatments. Unsurprisingly, the level of sexual bother was also much higher after RP compared with EBRT, brachytherapy and active surveillance, with 43% of RP patients reporting a moderate-to-big sexual problem after 12 months of treatment. This may be partly related to age, given the median age was much lower in the RP than the EBRT group, and younger men often have high sexual functioning expectations [
21].
Psycho-social factors relating to having cancer which are known to affect men’s ability to achieve or maintain sexual activity, may be contributing to declined sexual function and increased sexual bother across all four treatment groups [
22]. Providing sexual assessment, counselling, information, and emotional support following a diagnosis could improve sexual function and expectations among some men with prostate cancer. Interestingly, we found that men on active surveillance experienced a decline in sexual function 12 months after diagnosis, with an adjusted mean change score of − 22.1 points. More men on active surveillance were also bothered by their sexual dysfunction, with the proportion reporting sexual bother increasing from 20% at baseline to 37% at 12 months. Smith et al. (2009) reported a decline in sexual function in men on active surveillance compared to age-matched controls without prostate cancer [
3]. Further studies are required to determine the extent to which men with low-risk prostate cancer experience sexual dysfunction while on active surveillance and to identify factors that may account for the decline.
Functional scores for urinary continence had also declined from relatively similar levels at baseline among all treatment groups. Results from adjusted models indicate that post-treatment urinary incontinence was worse after RP compared with the other three treatments. Conversely, men who received RP showed clinically significant positive change (improvement) in urinary irritative/obstructive symptoms after 12 months. This finding is in line with previous research [
5,
7] in which authors hypothesised that, for men with coexisting hyperplasia of the prostate gland, removing the prostate improves urinary obstructive symptoms.
Our findings indicate that a clinically significant decline in bowel function was observed among men who had EBRT compared with RP. However, we did not observe any clinically meaningful change in bowel function 12 months after receiving brachytherapy. Bowel related complications from radiotherapy for prostate cancer have been observed previously [
3,
5,
6], and are reportedly worse for EBRT compared with brachytherapy [
23], though dose and technique can have a considerable impact [
24].
Australia’s national prostate cancer registry, PCOR-ANZ, does not collect baseline PROMs, as is the case for other international prostate cancer registries [
11]. Our models showed that baseline functional measures were positively associated with their respective 12 months functional measure across all domains, in line with previous studies [
4,
5]. Our findings with respect to baseline measures provide a reference for future PCOR-ANZ studies reporting 12-month functional outcomes. Accurate assessment of baseline PROMs is required to properly estimate the harms attributed to prostate cancer treatment over time and failure to adjust for baseline differences could result in over-estimation of treatment-related harms and thereby erroneously shift the balance of benefits and harms [
25]. In our study, without adjusting for baseline sexual function and other covariates, men who had received EBRT had lower scores in 12-months sexual function than RP (mean difference = − 9.7 points; 95% CI, − 13.6 to − 5.8) (Supplementary Table S5) but after adjusting for baseline characteristics, men in the EBRT group had higher scores (mean difference = 6.4 points; 95% CI, 0.9–12.0) in post-treatment sexual function compared with RP.
There have been advancements in radical therapy techniques that have led to an improvement in functional and oncologic outcomes. For example, different nerve-sparing surgical techniques [
26‐
28] and more accurate delivery of radiotherapies such as Intensity-modulated Radiation Therapy (IMRT) and Image Guided Radiation Therapy (IGRT) has led to an improvement in functioning after treatment [
29,
30]. The scope of our study was to compare the impacts between the main prostate cancer treatment groups not to compare specific treatment techniques.
The findings of this study should be interpreted in light of the following limitations. First, SA-PCCOC database captures the majority of prostate cancer cases in South Australia, and a significant proportion of men in the database had missing outcome responses for baseline and/or 12 months PROMs. We have used inverse probability weighting to minimise this bias. This cohort represents patients managed in both large tertiary centres (with associated government funded prostate cancer nurses) as well as single clinician private practices meaning that pre-treatment counselling may vary, which in turn may influence patient expectations, and thus, their responses. Second, our study did not assess further functionality changes overtime. While some studies showed that the changes over time (especially after 2 years) may be little and may not be significantly different between treatment categories [
5,
6], other studies [
31,
32] reported significant differences in functional outcome between treatments after 5 years. Third, the use of secondary treatment regimens (adjuvant or salvage therapy with radical therapies or conversion from active surveillance to treatment) was not addressed. The significant decline in post-treatment bowel and sexual function in the active surveillance group could be due to some men converting to radical therapies. Fourth, although active surveillance and watchful waiting are two different management options, some misclassification is likely within population-wide registries [
11] which may have impacted findings with respect to the surveillance group. Also, our brachytherapy group included both those receiving high-dose and low-dose rate brachytherapy, with the majority (70%) having undergone low-dose rate brachytherapy. In South Australia, high-dose rate brachytherapy is generally combined with EBRT and hence likely to have a different side effect profile to that of low-dose rate brachytherapy. Moreover, the type of radical treatment technique was not separately compared which may have an effect in the level of function decline. Fifth, although the final models were adjusted for sociodemographic and baseline characteristics, there may be residual confounding due to unmeasured covariates, for example previous medical and surgical histories, which are likely to differ between treatment groups. Finally, while most of our findings are in line with previous population-based studies [
3‐
6,
33,
34], the lack of consistency in the treatment approaches compared, PROMs measurement tools used, differences in follow-up periods, and the outcomes studied limits our ability to compare our results
directly with other PROMs research [
7,
8]. It should be highlighted that our findings are based on ‘real world data’, and as such, reflect the actual experiences of men during their prostate cancer journey.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.