Sexual function with localized prostate cancer: active surveillance vs radical therapy

Authors


Roderick van den Bergh, Homeruslaan 24-1, 3581 MH Utrecht, The Netherlands. e-mail: roodvdb@hotmail.com

Abstract

Study Type – Outcomes (cohort sample)

Level of Evidence 2b

What's known on the subject? and What does the study add?

The study compares the sexual function of men with low-risk prostate cancer who chose active surveillance (expectant management) with similar men who received radiation therapy or radical prostatectomy. The first group appeared to be sexually active more frequently and had less erectile dysfunction. The study was non-randomized.

No other studies exist on the effect of active surveillance on sexual function vs other treatment methods.

OBJECTIVE

  • • To compare sexual function of men with localized prostate cancer (PCa) on active surveillance (AS) with similar patients who received radical therapy.

PATIENTS AND METHODS

  • • Two groups of men with screening-detected localized PCa were compared. The first were men on AS within the prospective protocol-based Prostate Cancer Research International: Active Surveillance study. The second were men participating in the European Randomized Study of Screening for Prostate Cancer study who had received radical prostatectomy (RP) or radiation therapy (RT).
  • • Questionnaires were completed at two different timepoints after diagnosis or treatment (6 and 12–18 months). These contained 10 items on sexual function, the mental and physical component summary from the Short-Form 12-item health survey, the Center for Epidemiologic Studies Depression scale depression measure and the State Trait Anxiety Inventory general anxiety measure.
  • • Sexual function was compared between groups, and determinants were analysed in multivariable analysis, adjusting for baseline differences.

RESULTS

  • • A total of 65–68% of men on AS, 35–36% of those who underwent RP, 36–37% of those who underwent RT and 36% of men in the RP and RT groups combined (combined Tx) were sexually active.
  • • A total of 20–30% of men in the AS group, 86–91% of men in the RP group, 56–60% of men in the RT group and 71–76% of men in the combined Tx group were sexually inactive as a result of erectile dysfunction.
  • • A total of 44–51% of men in the AS group, 96% of men in the RP group, 73–76% of men in the RT group and 84–85% of men in the combined Tx group who were sexually active had problems getting or keeping an erection.
  • • In multivariable analysis these differences were significant, except for AS vs RT.

CONCLUSIONS

  • • Men with localized PCa on AS were more often sexually active than similar men who received radical therapy, especially RP. If not sexually active, this was less often attributable to erectile dysfunction for those on AS. If sexually active, this was less often associated with problems getting or keeping an erection for those on AS.
  • • The study was non-randomized; the latest advances in RP and RT might impact results.
Abbreviations
PCa

prostate cancer

AS

active surveillance

RP

radical prostatectomy

RT

radiation therapy

Combined Tx

RP and RT combined

QoL

quality of life

PRIAS

Prostate Cancer Research International: Active Surveillance

ERSPC

European Randomized Study of Screening for Prostate Cancer

SF-12

short-form 12-item health survey

CES-D

Center for Epidemiologic Studies Depression scale

STAI-6

short State Trait Anxiety Inventory

PCS

physical component summary

MCS

mental component summary.

INTRODUCTION

In recent decades, prostate cancer (PCa) has been detected more and more frequently, mainly because of screening. This trend started after the discovery of PSA as a marker for this disease [1,2]. During this same period, mortality rates have remained stable or decreased slightly [3].

Some PCas may never have caused any symptoms had they not been detected through screening [4] and yet almost all patients with PCa undergo surgery or radiation therapy [5]. These radical therapies have an important negative effect on men's urinary, bowel and sexual function [6,7].

Active surveillance (AS) might provide an alternative for managing low-risk PCa [8]. This strategy consists of initially withholding radical treatment after diagnosis and closely monitoring the disease instead [9]. The switch to curative active therapy remains an option and is made at the moment of risk reclassification or when preferred by the patient. AS is assumed to preserve health-related quality of life (QoL), but direct evidence is scarce [10].

In the present study, we compared sexual function of men diagnosed with localized PCa on AS with similar men who received radical therapy.

PATIENTS AND METHODS

We analysed and compared the sexual function scores of two groups of patients with localized PCa; the first on AS, the second after radical therapy. Patients were all Dutch, had screening-detected PCa, and were mainly Caucasian. The study was non-randomized.

The first study group consisted of patients with low-risk PCa, participating in a QoL side study of the prospective protocol-based active surveillance study Prostate Cancer Research International: Active Surveillance (PRIAS) [11]. Only men with Gleason scores of 6, PSA ≤10 ng/mL, PSA density <0.2 ng/mL/mL, T-stage ≤ T2 and <3 positive biopsies are eligible for inclusion in PRIAS. Participants are followed according to a strict protocol using PSA, DRE and standard repeat biopsies, and are advised to switch to radical treatment in case of risk reclassification during follow-up.

The second study group consisted of patients participating in a QoL side study of the European Randomized Study of Screening for Prostate Cancer (ERSPC) [12,13]. All were diagnosed with PCa through PSA screening and received either radical prostatectomy (RP) or radiation therapy (RT), without (neo-)adjuvant hormonal therapy. The exact methods of surgery (e.g. open vs laparoscopic or non-nerve-sparing vs nerve-sparing) or the radiation dose were not incorporated in the present study analysis. Patients with ≥T3 disease in the radical therapy group were excluded from analysis; patients with Gleason scores >6 were included.

Both patient groups completed QoL questionnaires at two different timepoints after diagnosis (AS) or start of treatment (RP or RT). The AS group received the first questionnaire within 6 months (t = 1) and the second at 18 months after diagnosis (t = 2). The RP/RT group received a questionnaire at 6 months (t = 1) and a questionnaire at 12 months after start of treatment (t = 2). Information was not available for either group on the use of adjuvant medication or aids to achieve erection.

In addition to biographical information (educational level and marital status), the questionnaires (both at t = 1 and t = 2) included the following measures. Sexual function was measured with 10 single items derived from Korfage et al. [13,14] and Slob et al. [15]. These items and their multiple choice answers are shown in Tables 1 and 2. Some items were only completed if applicable (item 3 depending on item 2; item 7 or item 8/9/10 depending on item 6). Furthermore, both questionnaires contained the Short-Form 12-item health-survey (SF-12), which allows for scoring a physical component summary (PCS) and a mental component summary (MCS) [16], the Center for Epidemiologic Studies Depression scale (CES-D) [17] and the short State Trait Anxiety Inventory (STAI-6) [18]. Medical information (age, comorbidity, PSA, T-stage and Gleason score) was obtained from the PRIAS and ERSPC databases.

Table 1.  Sexual function scores and group comparisons at first measurement (t = 1)
QuestionResponse optionsAS %, N= 129Combined Tx P *
RP %, N= 67RT %, N= 70AS vs RPAS vs RTAS vs combined Tx
 1. During the last week, how much interest in sex (e.g. sexual thoughts or desires) did you have?1 – none at all1114330.056 <0.001 <0.001
2 – only a little bit375046
3 – a moderate amount402818
4 – a lot1283
5 – a tremendous amount100
 2. During the last week how often have you awakened in the morning or night with a spontaneous erection: i.e. a stiff penis (partly stiff or very stiff) but not during sexual activity?1 – never (skip question 3)178859 <0.001 <0.001 <0.001
2 – about once in the last week35920
3 – more than once in the last week38315
4 – about once a day905
5 – more than once a day202
  %, N= 107%, N= 8%, N= 29 
 3. How would you describe the quality of your spontaneous erections (i.e stiff penis, not during sexual activity)?1 – totally limp0130 0.004 0.280 0.031
21130
3 – half limp/half stiff455062
4422532
5 – very stiff1208
 4. How important is sex for you?1 – very unimportant0360.609 <0.001 0.002
2 – unimportant9822
3 – slightly important343443
4 – important484625
5 – very important993
 5. Overall, how satisfied are you with your current sex-life?1 – very satisfied938 <0.001 0.015 <0.001
2 – satisfied39831
3 – slightly satisfied352222
4 – dissatisfied154427
5 – very dissatisfied22312
 6. Were you sexually active (e.g. masturbation, sexual intercourse) during the last 2 weeks?– Yes (skip question 7)683537 <0.001 <0.001 <0.001
– No326563
  %, N= 41%, N= 44%, N= 44 
 7. I have not been sexually active during the past 2 weeks because … (you may give more than one reason): (you may now skip questions 8, 9, and 10)– I did not feel like sex441434 <0.001 0.001 <0.001
– I do not have a partner827
– My partner did not feel like sex28212
– I could not get an erection (stiff penis)209160
– Other reasons, namely101417
  %, N= 88%, N= 23%, N= 26 
 8. In the last two weeks, have you had difficulty in getting an erection (stiff penis)?1 – Yes, (nearly always)78320 <0.001 0.005 <0.001
2 – Yes, occasionally281344
3 – No, (almost never)65436
  %, N= 88%, N= 23%, N= 26 
 9. In the last two weeks, have you had difficulty in maintaining an erection (stiff penis)?1 – Yes, (nearly always)98232 <0.001 0.005 <0.001
2 – Yes, occasionally31932
3 – No, (almost never)59936
  %, N= 88%, N= 23%, N= 26 
  • *

    Independent samples T-test.

10. How would you describe the quality of your erections during sexual activity in the last 2 weeks?1 – totally limp0484 <0.001 0.001 <0.001
21264
3 – half limp/half stiff262244
449444
5 – very stiff2404
Table 2.  Sexual function scores and group comparisons at second measurement (t = 2)
QuestionResponse optionsAS %, N= 78Combined Tx P *
RP %, N= 67RT %, N= 70AS vs RPAS vs RTAS vs Combined Tx
 1. During the last week, how much interest in sex (e.g. sexual thoughts or desires) did you have?1 – none at all101931 0.021 <0.001 <0.001
2 – only a little bit314252
3 – a moderate amount403112
4 – a lot1985
5 – a tremendous amount020
 2. During the last week how often have you awakened in the morning or night with a spontaneous erection: i.e. a stiff penis (partly stiff or very stiff) but not during sexual activity?1 – never (skip question 3)268656 <0.001 0.003 <0.001
2 – about once in the last week38629
3 – more than once in the last week3389
4 – about once a day303
5 – more than once a day103
  %, N= 58%, N= 9%, N= 31 
 3. How would you describe the quality of your spontaneous erections (i.e stiff penis, not during sexual activity)?1 – totally limp0040.118 0.020 0.011
222211
3 – half limp/half stiff453354
4414425
5 – very stiff1307
 4. How important is sex for you?1 – very unimportant4330.237 0.001 .184
2 – unimportant12335
3 – slightly important353939
4 – important444620
5 – very important693
 5. Overall, how satisfied are you with your current sex-life?1 – very satisfied826 <0.001 0.042 <0.001
2 – satisfied45929
3 – slightly satisfied252929
4 – dissatisfied183828
5 – very dissatisfied4238
 6. Were you sexually active (e.g. masturbation, sexual intercourse) during the last 2 weeks?- Yes (skip question 7)653636 <0.001 <0.001 <0.001
- No356464
  %, N= 27%, N= 43%, N= 45 
 7. I have not been sexually active during the past 2 weeks because … (you may give more than one reason): (you may now skip questions 8, 9, and 10)- I did not feel like sex331034 <0.001 0.030 <0.001
- I do not have a partner1127
- My partner did not feel like sex19107
- I could not get an erection (stiff penis)308656
- Other reasons, namely151412
  %, (N= 51)%, (N= 24)%, (N= 25) 
 8. In the last two weeks, have you had difficulty in getting an erection (stiff penis)?1 – Yes, (nearly always)108323 <0.001 0.036 <0.001
2 – Yes, occasionally33945
3 – No, (almost never)57932
  %, (N= 51)%, (N= 24)%, (N= 25) 
 9. In the last two weeks, have you had difficulty in maintaining an erection (stiff penis)?1 – Yes, (nearly always)148736 <0.001 0.061 <0.001
2 – Yes, occasionally35927
3 – No, (almost never)51436
  %, N= 51%, N= 24%, N= 25 
  • *

    Independent samples T-test.

10. How would you describe the quality of your erections during sexual activity in the last 2 weeks?1 – totally limp2355 <0.001 0.146 <0.001
20220
3 – half limp/half stiff353050
4451336
5 – very stiff1809

First, baseline characteristics and QoL measures at t = 1 and t = 2 were stratified and compared between treatment groups (AS vs RP or RT). Comparisons were also made between AS and the radical therapy, i.e. RP/RT combined (combined Tx), group. Second, sexual function scores at t = 1 and also at t = 2 were assessed and compared between groups. Third, multivariable analyses were performed with the three most important sexual function outcomes as dependent variables. A first model corrected for patient-related factors (age, education, comorbidity, baseline or follow-up QoL measures), a second model corrected for tumour-related factors (T-stage, Gleason score), and a third model corrected for both patient- and tumour-related factors.

Independent sample T-tests and linear regression analyses were applied using SPSS version 16.0.

All patients signed the PRIAS or ERSPC informed consent forms. The medical ethical committee of the Erasmus University Medical Centre and of participating local hospitals approved the PRIAS, ERSPC and QoL side studies (Dutch Trial Registry: PRIAS: NTR156, ERSPC: NTR1718).

RESULTS

A total of 266 patients with localized PCa were included in the present study (AS: n= 129, RP: n= 67, RT: n= 70). Table 3 presents baseline study cohort characteristics, stratified per treatment group, and comparisons between treatment groups. Significant differences were observed in age at diagnosis, education level, number of comorbidities, T-stage, Gleason score, and QoL scores. The strict inclusion criteria in the AS group resulted in more favourable characteristics in that group than in the radical treatment groups.

Table 3.  Baseline study cohort characteristics, stratified per treatment group, and comparisons between groups (N= 266)
 ASCombined Tx P *
RPRTAS vs RPAS vs RTAS vs Combined Tx
  • SF, short-form.

  • *

    Independent samples T-test.

  • Score range 0–100 (best health), mean score is 50 in general population.

  • Score range 0–100 (best health), mean score is 50 in general population.

  • §

    Score range 0–60 (maximum depression), threshold for clinically depressive: ≥16.

  • Score range 20–80 (maximum anxiety), threshold for high anxiety: >44.

No. of patients1296770
Mean age at diagnosis, years64.962.168.10.001<0.0010.715
Education (%)   0.1940.001<0.001
 Low313643
 Mid365149
 High33127
Marital status (%)   0.6420.1360.071
 Married928786
 Other81314
Comorbidities (%)   <0.0010.5170.009
 None336037
 ≥1674063
Mean PSA at diagnosis (ng/mL)5.75.57.40.6980.0510.154
T-stage (%)   <0.001<0.001<0.001
 T1712220
 T2297880
Gleason score (%)      
 610084670.001<0.001<0.001
 701529(6 vs >6)(6 vs >6)(6 vs >6)
 8014   
Mean SF-PCS (t = 1) scores50.351.948.00.1510.1380.797
Mean SF-MCS (t = 1) scores54.253.852.50.7870.2880.395
Mean CES-D (t = 1) scores5.87.58.60.1330.0120.010
Mean STAI (t = 1) scores35.830.333.7<0.0010.1910.002
Mean SF-PCS (t = 2) scores51.251.247.30.9940.0100.110
Mean SF-MCS (t = 2) scores53.955.354.80.3970.5910.418
Mean CES-D§ (t = 2) scores5.47.38.40.1380.0170.026
Mean STAI (t = 2) scores34.832.032.80.0950.2610.090

Table 1 presents sexual function scores and comparisons between treatment groups at t = 1. Significant differences were found between the AS and the combined Tx group on all 10 items, in favour of AS. When comparing AS with RP or RT only, men in the AS group showed better scores on eight and nine (different) items, respectively. Sexual activity in the previous 2 weeks was 68% in the AS group, vs 35%, 37% and 36% for the RP, RT and combined Tx groups, respectively.

Of 129 men on AS completing the first questionnaires, 78 (60%) also completed the second. Other patients had stopped AS, most (>90%) as a result of tumour risk reclassification during follow-up.

Table 2 shows sexual function scores and comparisons between treatment groups at t = 2. For sexual function, scores were similar at the t = 2 measurement to those at the t = 1 measurement. The AS group showed better results than the combined Tx group on nine items and better results than both the RP and RT groups on eight items. Sexual activity in the previous 2 weeks was 65% in the AS group, vs 36% for the RP, RT and combined Tx groups.

At t = 1 it was found that, of the men who had not been sexually active during the previous 2 weeks, this was less often attributable to erectile problems for the AS group (20%) than it was for the RP (91%), RT (60%), or combined Tx (76%) groups. This was also found at t = 2 (30% vs 86%, 56%, and 71%, respectively; all P values significant).

When combining sexual activity with items on problems in getting or keeping an erection (items 6, 8 and 9) at t = 1, it was found that, of the men who had been sexually active during the previous 2 weeks, for the AS group this less often involved problems getting or keeping an erection (44%) than it did for the RP (96%), RT (76%), or combined Tx (85%) groups. Similar results were found at t = 2 (51% vs 96%, 73%, and 84%, respectively; all P values were significant except for the AS vs the RT group at t = 2; data not shown).

Table 4 shows the third model of the multivariable analysis, correcting for patient as well as tumour-related characteristics (age, education, comorbidity, CES-D scores, STAI scores, PCS scores, T-stage and Gleason scores). The three outcomes were: sexual activity in the previous 2 weeks, sexual inactivity owing to erectile dysfunction, and sexual activity but with problems getting or keeping an erection. Significant differences were found between the AS and the RP and combined Tx groups, but not between the AS and RT groups; however, in the first and second multivariable model (correcting for patient-specific factors or tumour-specific factors only), the AS group also scored significantly better than the RT group at t = 1. Age and, to a lesser extent, T-stage and education were important confounders.

Table 4.  Multivariable analysis (linear regression) of the effect of treatment group on the most important sexual function outcome measures, corrected for age, education, comorbidities, CES-D scores, STAI scores, PCS scores, T-stage and Gleason scores (3rd model)
 T = 1T = 2
AS vs RPAS vs RTAS vs Combined TxAS vs RPAS vs RTAS vs Combined Tx
  1. B; standardized coefficient beta; measure of effect magnitude.

Sexually active in the last 2 weeks. B 0.401B 0.066B 0.250B 0.376B 0.034B 0.232
P < 0.001 P= 0.172 P= 0.001 P < 0.001 P= 0.541 P= 0.008
Not sexually active in the last 2 weeks owing to erectile problems. B 0.770B 0.126B 0.494B 0.702B 0.134B 0.525
P < 0.001 P= 0.101 P < 0.001 P < 0.001 P= 0.216 P < 0.001
Sexually active but with problems getting or keeping an erection. B 0.484B 0.087B 0.318B 0.467B 0.051B 0.314
P= 0.001 P= 0.208 P= 0.004 P= 0.003 P= 0.514 P= 0.010

Figure 1 shows the differences between treatment groups on the three most important sexual function outcome measures.

Figure 1.

Differences between treatment groups on the three most important sexual function outcomes measures. (N indicates group size.)

DISCUSSION

The present study found a significant difference in sexual function for men diagnosed with localized PCa in favour of AS when compared with similar men who received radical therapy (RP or RT), as measured at two different timepoints after starting AS or RP/RT. Men on AS were more frequently sexually active, if not sexually active this was less often owing to erectile problems, and if sexually active they had fewer problems getting or keeping an erection. This difference was more apparent when comparing AS with RP than when comparing AS with RT. Only minor differences in sexual function were found over time in all treatment groups. Outcomes are in line with a generally assumed advantage of AS, i.e. one of sparing physical domains. The larger observed difference between AS and RP than between AS and RT is in line with other studies that found that RP has a greater impact on sexual function than RT [7].

Of the 129 men on AS completing the first questionnaire, only 78 (60%) also completed the second questionnaire. The main reason for this relatively low percentage was not non-compliance, but the fact that strict protocol reasons (repeat PSA testing and repeat biopsy) led to the advice to quit AS for many participants. Men who had stopped AS did not receive the second questionnaire. The 2-year treatment-free survival in the PRIAS study (73%) is in line with other prospective AS studies, when allowing for patient selection and follow-up criteria, and the prospective or more retrospective nature of these studies [11,19,20].

All available radical therapies for PCa, either RT or surgery, have an important negative impact on urinary, bowel and sexual function, which influences satisfaction with treatment outcomes [6,7]. Although patients tend to evaluate these side effects as less bothersome because of a response shift after being treated for and surviving cancer, only those patients who really need surgery or RT should be subjected to it [21]. The aim of the recently developed strategy of AS is to avoid, or at least delay, this negative impact on the health-related QoL in a selected group of patients. A more definitive solution for the problem of overdiagnosis and the resulting overtreatment, however, should be sought in new biomarkers or in the improvement of existing screening methods [22,23].

To our knowledge, this is the first study to primarily focus on the sexual function of men on AS and to compare it with men who received radical therapy. Thong et al. [24] studied the health-related QoL and symptom burden of men diagnosed with PCa managed expectantly when compared with similar men who received RT. It was found that patients managed expectantly at initial diagnosis (AS) had a similar health-related QoL and a lower symptom burden than patients treated with RT up to 10 years after the diagnosis. Thong et al. also found patients treated with RT reported significantly more problems related to getting and maintaining an erection in the last 2 weeks than the AS group and the normative population. Van den Bergh et al. [25] studied levels of anxiety and distress in men on AS and found favourably low scores when compared with reference scores and scores reported in the literature.

The percentages of men on AS who were sexually active (65–68%) and who were sexually inactive owing to erectile dysfunction (20–30%, or 6–10% of total) were similar to a group of 3810 men without PCa participating in the ERSPC (69% sexually active, 10.1% sexually inactive owing to ED) [15]. As found in the present study, percentages for men after RP or RT were less favourable.

A number of comments on the present findings should be made. First, the study was not randomized. It could be hypothesized that men with an a priori good sexual function prefer AS over radical therapy and therefore show a better sexual function after starting AS than the men receiving RP or RT. In relation to this point, the present analysis lacks a baseline measurement of sexual function before starting AS or receiving RP or RT. Second, the lack of randomization led to differences between groups, partly caused by the strict selection of men considered suitable for AS. Although we aimed to correct for these differences by a multivariable analysis, these may have led to a selection bias. Third, both RP and RT techniques are constantly being improved. Nerve-sparing robot-assisted laparoscopic RP and RT sparing surrounding tissues may decrease the damage of these therapies and the resulting negative side effects, thus decreasing the difference in sexual function scores between AS and RP/RT [6]. Fourth, our findings are based on measurements a relatively short time after starting AS or radical therapy. Although erectile dysfunction 1 year post-treatment can be considered permanent after RP or RT [7], some factors besides aging alone may have a negative effect on the development of sexual function during AS. Fujita et al. [26] found a worsening sexual function for patients on AS in the long-term owing to the effect of serial prostate biopsies which are incorporated in all existing AS protocols. Comparisons of sexual function between these groups in the longer term may therefore show different results. Fifth, as previously mentioned, no information was available on the use of adjuvant medication or aids to achieve erection. Finally, the t = 2 measurement was not synchronized between study groups.

Many questions remain unanswered regarding the optimum management of men diagnosed with (screening-detected) low-risk PCa. AS might prove to be a (temporary) alternative to surgery or RT, but outcomes of observational studies should be awaited. QoL side studies within ongoing trials randomizing men with low-risk PCa between AS and radical therapy will provide a more definitive answer to the questions on the effect of this strategy on physical domains and QoL.

In summary, the present study found that men with localized PCa on AS were sexually active more often, showed less erectile dysfunction, and had fewer problems getting or keeping an erection when compared with similar men who had received radical therapy at two different timepoints in the first years after diagnosis. This effect was more pronounced when AS was compared with RP than with RT.

The present retrospective non-randomized study suggests that AS may preserve sexual function of men with low-risk PCa by delaying or avoiding the side effects of radical therapy and may have a positive effect on these men's QoL in general. Observational studies and randomized controlled trials incorporating QoL issues are needed to confirm these results.

ACKNOWLEDGEMENTS

Sources of support: Prostate Cancer Research Foundation (SWOP), Rotterdam, The Netherlands.

CONFLICT OF INTEREST

None declared.

Ancillary