The role of anticholinergics in men with lower urinary tract symptoms suggestive of benign prostatic hyperplasia: a systematic review and meta-analysis

Authors

  • Benedict T. Blake-James,

    Corresponding author
    1. Institute of Urology, University College London,
      Benedict Blake-James, c/o Prof C. Fry Group, Institute of Urology, 48 Riding House Street, London W1W 7EY, UK. e-mail: benblakejames@hotmail.com
    Search for more papers by this author
  • Arash Rashidian,

    1. Clinical Effectiveness Unit, Royal College of Surgeons of England, London, UK, and
    2. School of Public Health, Tehran University of Medical Sciences, Iran
    Search for more papers by this author
  • Youko Ikeda,

    1. Institute of Urology, University College London,
    Search for more papers by this author
  • Mark Emberton

    1. Institute of Urology, University College London,
    2. Clinical Effectiveness Unit, Royal College of Surgeons of England, London, UK, and
    Search for more papers by this author

Benedict Blake-James, c/o Prof C. Fry Group, Institute of Urology, 48 Riding House Street, London W1W 7EY, UK. e-mail: benblakejames@hotmail.com

Abstract

Authors from the UK present a systematic review of publications on the safety and efficacy of anticholinergics in men with LUTS; they found them to be safe, but suggested that further studies are required to establish precisely their efficacy.

Authors from the USA studied the correlation between risk factors for vascular disease and the AUA symptom score; there was a possible association between vascular disease and the development and severity of LUTS in men.

OBJECTIVE

To assess the safety and efficacy of anticholinergics in men with lower urinary tract symptoms (LUTS) suggestive of benign prostatic hyperplasia (BPH) by a systematic review of published reports and a meta-analysis of the reported outcomes.

METHODS

We searched Medline, Embase and Cochrane databases (1966–2006), and hand-searched relevant reference lists and conference proceedings, for studies on the use of anticholinergics in men with BPH or bladder outlet obstruction. Eligible studies were assessed for quality and foreign language studies were translated. We collected data on all reported outcomes, conducted meta- analyses on the maximum urinary flow rate (Qmax), postvoid residual urine volume (PVR) and volume at first contraction, and calculated the acute urinary retention (AUR) rate. We used sensitivity analysis to confirm the findings.

RESULTS

We identified five randomized controlled trials (RCTs) and 15 observational studies. Four RCTs incorporating 633 patients were included in the meta-analyses. Anticholinergics did not significantly alter Qmax (0.1 mL/s, 95% confidence interval, CI, 0.6–0.7). The PVR was increased by 11.6 mL (95% CI 4.5–18.6) although there was no significant difference between AUR rates. The total International Prostate Symptom Scores (IPSS) were not significantly different, but there were improvements for IPSS storage subscores in one RCT. The AUR rate was 0.3% at the 12-week follow-up in 365 men in the RCTs and observational studies.

CONCLUSION

Anticholinergic use in men with LUTS suggestive of BPH appears to be safe. Further studies are required to establish efficacy with a suitable precision.

Abbreviations
OAB

overactive bladder

DO

detrusor overactivity

RCT

randomized controlled trial

AUR

acute urinary retention

Qmax

maximum urinary flow rate

PVR

postvoid residual urine volume

AE

adverse event

ITT

intent-to-treat

WMD

weighted-mean difference

VFC

volume at first contraction

MCC

maximum cystometric capacity

QoL

quality of life.

INTRODUCTION

About 30% of men aged 50–80 years have either moderate or severe LUTS [1]; within this group, about half the men will have a pattern of symptoms that overlaps with the syndrome of overactive bladder (OAB) [2]. The ICS defines OAB as a syndrome characterized by urgency with or without urge incontinence, usually with frequency and nocturia [3]. It affects 17% of adults (men and women) in Europe and the USA [4,5]. Like LUTS, the prevalence of OAB increases with age.

Although the conditions of LUTS, BPH and OAB are clearly causally related, the extent to which they are and the nature of the mechanisms linking them are ill-understood. OAB might arise secondarily to BOO (through the mechanism of detrusor instability), although alternative mechanisms in the cause of this condition seem likely, as the incidence of LUTS and OAB in elderly women matches that of men [6,7], a proportion of men with OAB have no BOO, and 30% of men with both OAB and BOO continue to have symptoms despite correcting the BOO [8].

Treatments also differ; OAB is managed with anticholinergic drugs, and recent reviews showed that when used in this way, anticholinergics are both effective and safe [9,10]. LUTS suggestive of BPH tend to be treated with α-blockers and 5α-reductase inhibitors [11]. There are theoretical concerns relating to the use of anticholinergics in men with LUTS suggestive of BPH; these are two-fold, i.e. about precipitating acute urinary retention (AUR) and about increasing the postvoid residual urine volume (PVR). The level of concern is such that anticholinergics are currently contraindicated in those with ‘significant BOO’ or urinary retention [12]. Current clinical practice guidelines (AUA, BAUS, and European Association of Urology) do not recommend the use of anticholinergics in men with LUTS suggestive of BPH. Nevertheless, the use of anticholinergics in such men remains attractive, and is implemented by clinicians in daily practice.

In the present systematic review and meta-analysis of published reports, we assessed the extent to which anticholinergic medication is safe and effective when used in men with LUTS suggestive of BPH.

METHODS

Studies that involved at least one study arm using an anticholinergic drug, and that included men with evidence of BPH or BOO, were selected for review. Randomized controlled trials (RCTs) were identified for data abstraction and meta-analysis; observational studies were included for establishing long-term adverse events (AEs).

Anticholinergic drugs selected included the muscarinic antagonists darifenacin, solifenacin, tolterodine and trospium chloride, and spasmolytic agents with significant antimuscarinic action, including oxybutynin chloride, propiverine and flavoxate. Terodiline, emepronium and anisotopine were excluded, as these drugs have been withdrawn from the market or are no longer readily available. Participants included men with clinical evidence of BPH or urodynamic BOO; the target population was men with symptoms of OAB and, as a consequence of their LUTS or benign prostatic enlargement, a suspicion of BOO.

A sensitive search strategy was devised and run on Medline (1966 to February 2006), Embase (1980 to February 2006) and Cochrane Collaboration databases (CENTRAL 2005 Issue 3). The general format for the strategy was [BPH or BOO] AND [anticholinergics] AND [clinical trial] (see Appendix 1 for full search strategy). To identify clinical trials we used a validated sensitive search filter [13]. In addition, we hand-searched reference lists from relevant publications, and conference proceedings from the ICS, AUA, European Association of Urology and BAUS between 1997 and 2005.

Potentially relevant articles were identified according to inclusion and exclusion criteria (Appendix 2), first by title then by abstract, before retrieving the article in full text. Where articles were not available from international library sources (British Library, National Library of Medicine) authors were contacted directly requesting a manuscript copy. Relevant foreign articles were obtained in full and translated.

The following data was extracted from included trials: study design, population characteristics (including sex, age, ethnicity, diagnosis); eligibility and exclusion criteria; interventions (dose and duration); comparisons; numbers screened, eligible, enrolled, and lost to follow-up. Data on all reported outcomes was collected, including AEs.

We assessed the quality of the trials according to predefined criteria (Appendix 3 and 4) [14]. We rated the internal validity of RCTs on the methods used for randomization, allocation concealment, and blinding; the similarity of compared groups at baseline; the similarity of protocol that each group was subjected to; adequate reporting of attrition; and the use of intent-to-treat (ITT) analysis. The external validity of trials was assessed based on the general applicability of the study to the target population and standard practice. Similarly, observational studies were rated according to criteria considered pertinent to the validity of reported AEs. A scoring system developed previously, determined by quality indicators, was used; these included the study objective, the definition of main outcomes, inclusion/exclusion criteria, patient characteristics and representativeness of sampling. Articles were not included in the analysis if they were flawed in more than two quality indicators, or followed patients for <1 month. Where any doubt remained as to the eligibility of a study, it was included subject to a sensitivity analysis, whereby outcomes with and without the study were assessed, to determine whether the inclusion of the study had a significant effect upon overall conclusions.

Meta-analyses were conducted where sufficient data with error calculations were available. Means and sds were used to derive a weighted-mean difference (WMD) and the inverse-variance fixed-effect model used for calculating summary statistics with 95% CI and the generation of Forest plots (StataTM software, StataCorp LP, Texas). All reported AEs were recorded from RCTs and observational studies, and incidence rates were calculated. Studies were divided according to follow-up duration and treated both separately and together. Studies not reporting a particular AE were not included in the analysis of that outcome.

RESULTS

Database searches yielded 1817 references after duplicate deletion. Of these, 157 were retrieved in full text, and 14 articles were found which fulfilled our inclusion/exclusion criteria (Fig. 1). Of the 14 selected articles, three referred to medication withdrawn or not in use [15–17].

Figure 1.

The trial flow diagram.

Hand-searching retrieved a further 19 articles that met the inclusion criteria. Of these, one article referred to a withdrawn anticholinergic [18], and the remaining articles were all either in abstract form from conference proceedings or from foreign-language journals not listed in the searched databases. Seven of these articles [19–25], and two from the database searches [26,27], were in abstract form and referred to studies already included, and were therefore also excluded. Consequently, 20 studies were included with nine obtained from database searches and 11 from hand-searches (Fig. 1).

Five of the selected trials were RCTs [28–32], the remainder being observational studies [33–47]. Eight studies were published in foreign languages [31,36,38,43–47]. A summary of the trial characteristics is shown in Table 1[19–47]. Of the five RCTs, three compared anticholinergics combined with α-adrenergic antagonists (α-blockers) to α-blockers alone, whereas two used anticholinergics alone and were placebo-controlled (Table 1).

Table 1. Clinical trials included in study
Trial referenceYearTrial designNo. of subjectsAnticholinergic agent, mg, ×/dayCombined withBOO criteriaSource
  1. Combined with, drug, if any, combined with anticholinergic agent in the intervention arm of the study; DB, databases; HS, hand-searches; α-B, α-adrenoceptor antagonist; BPE, benign prostatic enlargement; phyto, phytotherapeutic agent.

[23–26,28]2006RCT221tolterodine 2, × 2 urodynamicDB
[19,29]2005RCT228propiverine 20, × 1α-BurodynamicDB
[27,30]2003RCT 50tolterodine 2, × 2α-BurodynamicDB
[31]1999RCT134propiverine 20 × 1α-BclinicalHS
[32]1995RCT 70flavoxate 400, × 3 clinicalDB
[20,33]2005case series 43tolterodine 4, × 1 clinicalDB
[34]2005case series 82propiverine 20, × 1α-BclinicalHS
[35]2005case series 89tolterodine or oxybutyninα-BclinicalHS
[36]2005case series  4propiverine urodynamicDB
[21,37]2004case series144tolterodine 2, × 2α-BurodynamicDB
[38]2004case series 35propiverine 20, × 1α-BBPEHS
[22,39]2004case series 35propiverineα-BclinicalHS
[40]2004case series 46propiverine 10, × 1α-BclinicalHS
[41]1990case series 25flavoxate 400, × 3 urodynamicDB
[42]1987case series 85flavoxate 200, × 3 BPEDB
[43]1982case series 20flavoxate 200–400, × 3 clinicalHS
[44]1981case series  8flavoxate 200, × 3 clinicalHS
[45]1981case series 48flavoxate 200, × 3 clinicalHS
[46]1981case series 22flavoxate 200–400, × 3phytoclinicalHS
[47]1981case series 20flavoxate 200–400, × 3phytoclinicalHS

The study by Abrams et al.[28] describes the urodynamic findings in 221 men with urodynamically confirmed BOO taking tolterodine (2 mg twice daily) or placebo. After 12 weeks, there was a significant improvement in volume at first contraction (VFC) and bladder capacity (MCC). Although there was also a significant deterioration in both bladder contractility index and PVR (+ 25 mL) this was not considered clinically significant. This study appears to be well-designed, unlikely to give rise to significant bias, and reflects the target population well; these data were included in the analysis.

Lee et al.[29] reported on 228 men aged 50–80 years with urodynamically confirmed BOO and symptoms of OAB. The treatment arm was given propiverine and doxazosin, and compared with a doxazosin-only arm. At 8 weeks, the treatment arm had significantly better improvements in frequency and urgency; however, there was a significant improvement in the IPSS only when restricted to storage symptoms alone. On quality assessment there was a question as to the effectiveness of blinding, given the lack of a reported combined preparation or placebo, and although ITT analysis was declared, the efficacy analysis only included 93% of the men randomized to the study. Overall the study was of sound design and outcomes from this study were included in the analysis.

Athanasopoulos et al.[30] included 50 men in their study; participants had urodynamically confirmed moderate or mild BOO and detrusor overactivity (DO), and were given either the combination of tolterodine 2 mg twice daily and tamsulosin 400 µg, or tamsulosin alone. After 12 weeks, they repeated the urodynamic evaluation and issued a quality-of-life (QoL) questionnaire; LUTS were not evaluated. The authors reported a significant improvement between groups in MCC, VFC and maximum unstable contraction pressure. The Qmax and PVR were not significantly different between groups. QoL was improved only in the combined-treatment group. Overall this study had good internal validity, although by excluding men with severe BOO the safety might be overestimated. Points of concern on quality assessment were that no mention was made of the treatment withdrawals, and ITT analysis was not declared, although as 92% of men completed the study, this effect is not expected to be large. Again, no mention was made of a combined preparation or use of a placebo, so the blinding was also suspect, and in common with all other studies there is no mention of techniques used to conceal treatment allocation. However, the overall quality assessment was considered acceptable and data from this study were included in the analysis. On conversion of the reported sem there was a very low sd for the number of participants in his study, but when these data was excluded from the analysis this did not alter the conclusion.

Saitoh et al.[31] recruited 134 patients, including men with BPH and LUTS or urinary frequency, although excluding men with severe voiding difficulty. Patients were given either tamsulosin (200 µg) and propiverine (20 mg) or tamsulosin alone. After 4 weeks, patients completed a questionnaire (including the IPSS) and had uroflowmetry. Saitoh et al.[31] found a significant improvement over tamsulosin alone only in the mean (sd) rate of nocturia, at − 1.32 (1.3) vs − 0.65 (0.9). Objective measurements of voided volume, flow rate and PVR were not significantly different between the groups. Quality assessment showed that the inclusion criteria were not well defined; the trial was quasi-randomized, using odd vs even case-note numbers; no attempt at blinding was documented; ITT analysis was not used and there was a variable response rate to all outcomes, which at best in the treatment group was 73%, and was as low as 59% for estimating PVR. Data from this study were therefore used after sensitivity analysis.

Dahm et al.[32] reported a RCT including 70 men with a diagnosis of BPH, a Qmax of <10 mL/s, and nocturia of at least twice. The treatment arm was given flavoxate and compared against placebo. At 12 weeks, subjects completed a questionnaire on LUTS and had uroflowmetry. Dahm et al. found no significant difference between the arms in any the outcome measures. Concerns on quality assessment included the absence of ITT analysis, and with only 71% included in their analysis there was a significant discontinuation rate. Nocturia was used as the criterion for inclusion, which does not match that of OAB and might include alternative causes such as nocturnal polyuria; furthermore, flavoxate is not commonly used in clinical practice, although in general the findings were considered applicable to the target group of this review. The results from this study were to be included subject to sensitivity analysis, although the absence of error values precluded this.

For objective measures the main outcomes are summarized in Table 2[28–32]. Qmax values were recorded as an outcome in all RCTs. There was no significant difference between the intervention and control arm in any of the studies. A Forest plot of the synthesized data showed a WMD of 0.07 mL/s (Fig. 2A[29–31]). A fixed-effect technique was used in the meta-analysis; the results were also tested with a random-effect technique, but were not significantly different.

Table 2. A summary of the outcomes of the five RCTs
Outcome[28][29][30][31][32]
  1. Change is intervention arm vs change in control; statistical significance between groups P < 0.05; NS, not significant.

Number of subjects2212285013470
Agenttolterodinepropiverine + doxazosintolterodine + tamsulosinpropiverine + tamsulosinflavoxate
Qmax, mL/s−0.3 vs +0.5+1.0 vs 1.7+1.32 vs + 1.16+0.5 vs + 2.9−0.1 vs + 0.1
 PNSNSNSNSNS
PVR, mL/s+25 vs + 0+20.8 vs −4.7−4.2 vs − 8.2+24 vs − 9.5−2 vs − 6
 Change, Pup, 0.004up, 0.002NSNSNS
VFC, mL+ 59 vs − 31+100.4 vs +30.4
 Change, Pup, 0.003up, <0.001
MCC, mL+67 vs − 8.0+36.4 vs +0.8
 Change, Pup, <0.001up, 0.002
Frequency episodes−1.9 vs − 0.9−1.74 vs − 1.87
 Change, Pdown, 0.004NS
Nocturia episodes−0.7 vs − 0.6−1.32 vs − 0.65−0.6 vs − 0.8
 Change, PNSdown, 0.004NS
Urge, IPSS−1.2 vs −0.7−0.62 vs – 0.55
 Change, P down, 0.02NS
IPSS total−7.4 vs −7.3−5.01 vs − 5.51
 PNSNS
IPSS storage−3.8 vs −2.9−2.99 vs − 2.22
 Change, P down, 0.03   
Figure 2.

Forest plots for: A, Qmax; B, PVR; and C, VFC.

The PVR was also estimated in all RCTs (Table 2); there was a tendency towards an increase in PVR with anticholinergic use. The meta-analysis of the data showed a WMD of 11.6 mL (95% CI 4.5–18.6; Fig. 2B[29–31]). Again, a fixed-effects technique was used, and results did not differ with a random-effect technique.

The urodynamic variable VFC was measured in both the study by Abrams et al.[28] and that by Athanasopoulos et al.[30]. In both cases increases in VFC after anticholinergic use were significantly greater than in the control arm. Abrams et al.[28] found an increase of 59 mL (95% CI 19–100, P < 0.003), whilst Athanasopoulos et al.[30] reported that VFC increased from 193 (9) to 294 (14) mL (P < 0.001; Fig. 2C[28,30]).

These findings are supported by observations of increases in both MCC and mean voided volume. As with VFC, both Abrams et al.[28] and Athanasopoulos et al.[30], from urodynamic measurements, reported increases in MCC significantly higher than in the control arm. Abrams et al.[28] reported an increase of 67 mL (95% CI 35–103, P < 0.001) and Athanasopoulos et al.[30] found that the MCC increased from 499 (17) to 536 (11) mL (+ 36 vs + 1 mL, P < 0.05) in the intervention group. The mean voided volume after anticholinergic use was significantly higher in the studies of Lee et al.[29], at + 55 vs + 31 mL (P < 0.05), Okada et al.[39] (+32 mL, P < 0.05), and Suzuki et al.[40] (P < 0.01), although Saitoh et al.[31] and Sugiyama et al.[38] found the increase to be smaller and not significant.

For the subjective outcomes, Lee et al.[29], Saitoh et al.[31] and Dahm et al.[32] reported LUTS, although the only validated scoring system was the IPSS, used by Lee et al. and Saitoh et al. The overall IPSS was not significantly improved in either of these studies. However Lee et al.[29] reported a significant improvement in the storage subscore (sum of frequency, urgency and nocturia components: − 3.8 vs − 2.9, P < 0.05). Saitoh et al.[31] also found greater improvements in each of these variables in the anticholinergic arm (storage subscore − 3.0 vs − 2.2), although it was significant only in the rate of nocturia. Dahm et al.[32] used unvalidated questions evaluating LUTS but found no significant difference between treatment groups.

Athanasopoulos et al.[30] used the Urolife BPH QoL questionnaire before and after treatment, and found a significant improvement in QoL, but only in the group taking anticholinergics. Lee et al.[29] found a significantly higher satisfaction amongst the intervention group using an overall satisfaction questionnaire, and Saitoh et al.[31] reported a better, although insignificant, overall impression from patients on anticholinergics. However, Dahm et al.[32] found that most patients reported no change in a global evaluation questionnaire after treatment with flavoxate.

OBSERVATIONAL STUDIES

The present review identified 15 studies of case-series design. Quality assessment found significant flaws in six studies [42–47], five of which were published in Japanese language journals. The remaining nine studies included 362 subjects.

Kaplan et al.[33] prescribed tolterodine as monotherapy to 43 men in whom treatment with α-blockers had failed. At 6 months both frequency and nocturia were significantly improved, and there were no episodes of AUR.

Lee et al.[37] described a prospective study of 144 men with BOO who had urodynamics and were divided into those with BOO or BOO + DO. The 60 men who did not show symptomatic improvement with an α-blocker were then given combined treatment with α-blocker plus anticholinergic; 63% had a subsequent improvement, as indicated by a decrease in the IPSS of ≥ 3 points. The greatest improvement was amongst men with BOO + DO (72%). Two men given an anticholinergic developed AUR.

Okada et al.[39] selected 35 men with predominantly irritative LUTS and BPH who had not responded to α-blockers. The men were given propiverine in addition and after 12 weeks had a significantly lower IPSS and higher QoL score, with no significant increase in PVR or occurrence of AUR.

Lim et al.[35] recruited 89 men, 68 of whom completed the study. All men had LUTS suggesting BOO and OAB, and had persistent storage symptoms despite α-blockade. They were given an anticholinergic for ≥ 6 months and re-assessed; 88% had an improvement, with significant improvements in IPSS, QoL, uroflow and nocturia, with no increase in PVR or AUR.

Suzuki et al.[40] studied 46 men with evidence of BOO and OAB who had responded to α-blockers but had persistent symptoms. After propiverine, the IPSS and QoL Index were significantly improved, and most men elected to continue using anticholinergics.

AEs

All RCTs reported on AEs; of the nine observational studies, two did not record AEs [36,38], and another did not follow-up patients for >2 weeks [40] and so was considered unsuitable for AE reporting. Therefore, 847 subjects from 11 studies were available for AE reporting (Table 3[28–35,37,39,40]). Subjects were followed for a mean (range) of 12.2 (4–26) weeks.

Table 3. A summary of the AEs, as the number of adverse events (% of subjects in trials which reported AEs)
Trial (weeks of follow-up)InterventionControl
NDry mouthAUROther urinary AENDry mouthAUROther urinary AE
  1. N, number of participants; NR, not reported; NA, not applicable.

[28] (12)221 36 (16.3)0 9 (4.1)149 1 (0.7)1 (0.7)3 (2.0)
[29] (8)142 26 (18.3)0 6 (4.2) 69 4 (5.8)02 (2.9)
[30] (12) 25  2 (8)0 0 25 000
[31] (4) 67  5 (7.5)2 (3) 9 (13.4) 46 000
[32] (12) 37  8 (21.6)0 0 33 7 (21.2)1 (3.0)0
[33] (26) 43 16 (37.2)2 (4.7)NRNANANANA
[34] (12) 82  6 (7.3)1 (1.2)NRNANANANA
[35] (26) 89NR0NRNANANANA
[37] (8) 60 16 (26.7)2NRNANANANA
[39] (12) 35NR0NRNANANANA
[40] (4) 46  00NRNANANANA
Total (4) 113  5 (4.4)2 (1.8) 9 (13.4) 46 000
Total (8)202 42 (20.8)2 (1.0) 6 (4.2) 69 4 (5.8)02 (2.9)
Total (12)365 52 (14.2)1 (0.3) 9 (3.2)207 8 (3.9)2 (1.0)3 (1.4)
Total (26)132 16 (37.2)2 (1.5)NRNANANANA
Total (all)847 115 (15.9)7 (0.8)24 (4.9)32212 (3.7)2 (0.6)5 (1.6)

The most common AE reported was xerostomia, or dry mouth, with an incidence of 15.9% vs 3.7% amongst controls. This symptom was frequently described as mild, although in 14 cases (2%) was not tolerated and resulted in withdrawal from the trial.

Instances of AUR were seldom reported, and rates between intervention and control groups were comparable (0.8% vs 0.6%). More frequently reported were cases of difficulty in micturition, or a significantly raised PVR, in all reported in 24 patients (4.9%), 14 of whom were withdrawn from the trial as a result.

Other AEs typically recorded included gastrointestinal symptoms, dizziness and blurred vision, although these symptoms were seldom associated with withdrawal from the trial.

DISCUSSION

We identified 20 studies addressing the use of anticholinergics in men with BOO, i.e. five RCTs and 15 observational studies. These studies show reasonable homogeneity in their findings. Whilst the efficacy of the anticholinergic is usually moderate, it is often significantly greater than the control, and although associated with AEs, these are frequently minor and well tolerated.

The most pragmatic assessment of efficacy is to measure symptom scores and QoL indicators, as these usually represent the ultimate treatment goal in clinical practice. Overall the IPSS was not significantly improved by the use of anticholinergics, although this might simply reflect the low effect of these agents on voiding symptoms. When storage symptoms were sub-analysed they were lower in those taking anticholinergics, and storage symptoms are associated with the most bother and impact on QoL. That QoL and overall satisfaction questionnaires generally showed improvements amongst those taking anticholinergics is perhaps a testament to this.

The results from case-series tend to support these findings, with improvements in storage symptoms, but not voiding symptoms where sub-analysed. QoL indicators were improved in all studies that measured them.

Urodynamic variables can be used as surrogate markers of efficacy, and are better suited to objective measurement. The increase in VFC and MCC might be expected to correlate with a reduction in frequency. However, there might be sources of bias common to these studies. The significant placebo effect seen in symptoms of OAB is well-reported [10]. The relatively few trials using a placebo comparison means the effect is less apparent in this review, although the absence of combined preparations or placebo amongst the control arms of trials using an α-blocker comparison might result in an overestimate of the intervention efficacy.

Safety concerns amongst men with BOO are the primary reason for caution in using anticholinergics in this group. The most significant urinary AE is AUR; this is readily recorded and well reported in the studies included in this review, although of low incidence, and in similar numbers amongst the control arms of trials. However, other urinary AEs are reported and the finding of a significantly raised PVR or difficulty in micturition might precede frank AUR. Nevertheless, the incidence of such events necessitating discontinuation of the anticholinergic remains relatively low (2.8%) and appears to be readily reversible on stopping the drug. Urodynamic variables indicative of a urinary adverse outcome are reductions in detrusor contractility, flow rate or voiding efficiency, resulting in a raised PVR. Whilst modest reductions were noted in indicators of detrusor contractility, Qmax was essentially unaltered and the increase in PVR, whilst significant, was small and unlikely to be of clinical significance. However, the median and mode duration of follow-up for AEs was 12 weeks; this might be insufficient to determine disease progression and the rate of AUR might thus be underestimated.

These findings correlate well with the opinion cited in reviews of the subject published to date. Reynard [48] and Gonzalez and Te [49] reviewed the studies of Abrams [24] and Athanasopoulos et al.[30]. Reynard concluded that further studies were required, but that the preliminary evidence supported the use of anticholinergic drugs in men with BOO; Gonzalez and Te also supported anticholinergic use in men with BPH, but proposed an algorithm that excluded men with significant BOO or PVR. Ruggieri et al.[50] included the work of Abrams [24], Athanasopoulos et al.[30], Lee et al.[37] and Kaplan et al.[20], and reviewed the mechanism of action of α-blockers and anticholinergics, relating this to the physiology of micturition in OAB. They concluded that there was greater theoretical benefit from combined therapy than from either agent alone.

The present systematic review therefore represents an objective synthesis of evidence published on the subject to date. The database search was designed to be sensitive and identified all published articles cited in other reviews. Hand-searching revealed a further 19 studies, all but one [18] of which were not referenced in the databases searched. The article not identified by the database search was in Japanese and referred to terodiline, a drug now withdrawn from the market and not included in this review; thus search terms were validated and not modified. The remaining articles were all either in abstract form from conference proceedings, or from foreign language journals; a significant proportion of the abstract articles referred to studies already included. Whilst the exclusion of articles in languages other than English can be justified [51], we elected to translate foreign articles, as there was a significant proportion of articles in Japanese, the exclusion of which might have skewed the findings.

Although the target population was men with clinical evidence of BPH and symptoms of OAB we included studies, the selection criteria for which included the presence of BOO and/or DO. Only a proportion of men with evidence of BPH are expected to have demonstrable BOO; we included both groups of men, as whilst this distinction is not expected to alter the efficacy, it might overstate the risk. Similarly, we made no distinction in recruitment between the symptoms of OAB and urodynamically confirmed DO, as these two groups have been found to correlate well [52].

In this review, we did not differentiate among anticholinergic agents, including all in current usage, and at all doses. Whilst this generates a degree of heterogeneity amongst studies, we justify this approach by aiming to determine the action of this class of drug, rather than any particular drug or dose per se. It was not anticipated that this approach would overestimate efficacy, and it might be apparent that all efficacy data are derived from the use of only two agents, tolterodine and propiverine. However, safety data included all agents, including those considered less ‘uroselective’, which might be expected to overestimate AEs.

Our findings have a plausible scientific rationale, particularly as we gain some understanding of the pathophysiology of micturition [53], although this remains to be fully elucidated. It is nevertheless becoming clear that the action of anticholinergics is not simply to reduce the parasympathetic autonomic outflow at micturition, as supported by the findings of changes to MCC and VFC. It is likely that anticholinergics act both at afferent and efferent pathways and both pre- and post-synaptically. Should anticholinergics suppress afferent signals, yet have little effect in the face of the flood of parasympathetic outflow at micturition, we could expect a reduction of frequency and urgency whilst leaving voiding contractility relatively unaffected, thus conferring both efficacy and safety. It is perhaps not surprising therefore, that Qmax is not significantly altered by anticholinergics, although it is intriguing that Abrams et al.[28] found a significant reduction in BOO after anticholinergic use. It is conceivable that anticholinergics are acting directly to relax the prostate or urethra, and that the small decrease in bladder contractility might be offset by this, to leave urinary flow unaltered.

Many men with uncomplicated LUTS might be successfully managed with education, reassurance and monitoring alone, and even when intervention becomes necessary, lifestyle measures are recommended before medical therapy. Nevertheless, a significant proportion of men will require medical therapy, and treatment options are currently largely restricted to the use of 5α-reductase inhibitors, α-adrenoceptor blockers, the combination of these two, or phytotherapeutic agents. However, there remains a proportion of men whose symptoms persist despite medical therapy, or for whom this therapy is not tolerable. The use of anticholinergics might provide a therapeutic solution and appears to be both safe and effective in men with BPH, even in the presence of known BOO. However, most studies reported used the combination of α-blocker and anticholinergic; adding an α-blocker might ameliorate the BOO, thus contributing both to safety and efficacy, although where an anticholinergic was used alone this was not associated with higher rates of AUR. Several studies also excluded men with evidence of significantly impaired voiding efficiency, by using a PVR threshold of 30–40% or 50–150 mL [28,29,34,37,40], or by excluding men with a high grade of BOO [30,31,37]. It would therefore appear prudent to check the PVR before starting anticholinergics, and to continue α-blockade if this carries a symptomatic improvement for the patient. Patients should be warned of the possibility of difficulty in micturition, and should be advised to discontinue the drug if this occurs.

However, further studies are required to establish clearly the role that this treatment regimen has for men with BPH, with or with no urodynamic evidence of BOO and DO. It remains to be seen whether this is a safe option for men with high grades of BOO, and what level of PVR is acceptable. RCTs with attention to blinding procedures and inclusive selection criteria will help to establish this. A long-term follow-up might be required to exclude an increased risk of AUR.

In conclusion, the use of anticholinergics in men with BPH was investigated in both RCTs and observational studies; they show that, as in the absence of BOO, anticholinergics offer a modest but significant advantage, and result in an improvement in QoL indicators and patient satisfaction. The concerns that anticholinergics might be associated with impaired voiding and AUR do not appear to be supported by the evidence from the studies assessed. Anticholinergics are associated with a small rise in PVR, but not an increased rate of AUR.

ACKNOWLEDGEMENTS

We thank the Prostate Research Campaign UK for their support in providing a research grant for Mr B. Blake-James.

CONFLICT OF INTEREST

None declared.

Appendices

APPENDIX 1

Search strategy used for Medline (online version via Ovid)

1. (prostat$ or BPH or bladder out$ obstruction or BOO or bladder neck or infravesical).mp.

2. exp Prostatic Hyperplasia/or exp Bladder Neck Obstruction/or exp Prostate/

3. ((urin$ adj2 (retention or retain$)) or (residual adj2 volume) or (postvoid adj2 residual) or urin$ cathete$ or ((difficulty or inability or unable) adj3 (mictur$ or urin$)) or (bladder adj1 (obstruction or emptying or distension))).mp.

4. exp urinary retention/or urinary catheterization/or urination/de or urination disorder/de or urodynamics/

5. (urodynamic$ or cystometr$).mp.

6. 1 or 2 or 3 or 4 or 5 [BOO or urinary retention]

7. ($cholinergic$ or $muscarinic$ or parasympatholytic$).mp.

8. (anticholinergic or anticholinergic or anti cholinergic or anticholinergics or anticholinergics or anti cholinergics or cholinergic or cholinergics or antimuscarinic or antimuscarinic or anti muscarinic or antimuscarinics or antimuscarinics or anti muscarinics or muscarinic or muscarinics or parasympatholytic or para-sympatholytic or para sympatholytic or parasympatholytics or para-sympatholytics or para sympatholytics).mp.

9. (tolterodine or ox?butynin or trospium or propiverine or darifenacin or solifenacin).mp.

10. exp cholinergic antagonists/or exp muscarinic antagonists/or exp parasympatholytics/

11. Benzhydryl Compounds/ae, tu, ct or Phenylpropanolamine/ae, tu, ct or Benzofurans/ae, tu, ct or Pyrrolidines/ae, tu, ct or Cresols/ae, tu, ct or Nortropanes/ae, tu, ct or Mandelic acids/ae, tu, ct or Tartrates/ae, tu, ct or Quinuclidines/ae, tu, ct or Tetrahydroisoquinolines/ae, tu, ct

12. Receptor, Muscarinic M1/ai or Receptor, Muscarinic M2/ai or Receptor, Muscarinic M3/ai

13. 7 or 8 or 9 or 10 or 11 [anticholinergics]

14. 6 and 13 [anticholinergics + BOO]

15. limit 14 to (randomized-controlled-trial or meta-analysis or controlled-clinical-trial or clinical-trial)

16. (meta-anal$ or metaanaly$ or meta analy$).mp.

17. random$.mp. [mp = title, original title, abstract, name of substance word, subject heading word]

18. ((doubl$ or singl$) and blind$).mp. [mp = title, original title, abstract, name of substance word, subject heading word]

19. exp Clinical Trials/

20. crossover.ti. or crossover.ab. or crossover.me.

21. clin$ trial$.mp. [mp = title, original title, abstract, name of substance word, subject heading word]

22. (control$ and (trial$ or stud$)).mp. [mp = title, original title, abstract, name of substance word, subject heading word]

23. ((singl$ or doubl$ or tripl$ or trebl$) and (blind$ or mask$)).mp. [mp = title, original title, abstract, name of substance word, subject heading word]

24. placebo$.mp. [mp = title, original title, abstract, name of substance word, subject heading word]

25. Research Design/

26. Comparative Study/

27. 15 or 16 or 17 or 18 or 19 or 20 or 21 or 22 or 23 or 24 or 25 or 26 [validated sensitive clinical trial filter. ref [13]]

28. 14 and 27

The search strategy terms were translated as necessary for the Cochrane database search (online via Wiley Interscience); for Embase (online via WebSPIRS) subject headings were translated according to Emtree term advice, and check tag substitution according to WebSPIRS technical support staff.

APPENDIX 2

Inclusion criteria:

Documented BOO or prostatic enlargement in men.

Recorded use of anticholinergic medication.

Clinical trials of any design.

Exclusion criteria:

Female subjects.

Subjects aged <18 years.

Neurogenic bladder or subjects with a history of spinal cord injury, multiple sclerosis or cerebral infarction.

Subjects using indwelling or intermittent catheterization.

Parenteral administration of anticholinergic agent, e.g. inhaled, intravesical.

Concurrent use of alternative intervention acting upon detrusor, e.g. botulinum toxin.

Previous reconstructive lower urinary tract surgery, e.g. enterocystoplasty.

Previous lower urinary tract surgery for BOO, e.g. prostatectomy.

Review articles.

Intervention drug not possessing primarily anticholinergic mode of action.

APPENDIX 3

RCT Quality Assessment form.

1.1. Does the study clearly and appropriately address the guideline question? (Yes/No and comment)

1.2. Was the assignment of subjects to treatment groups randomized? (Yes/No/NotClear and comment)

1.3. Was an adequate concealment method used? (Yes/No/Not clear and comment)

1.4. Were subjects, clinicians and investigators kept ‘blind’ about treatment allocation? (answer Yes/No for each of the three categories and comment)

1.5. Were the treatment and control groups similar at the start of the trial? (comments on imbalances)

1.6. Apart from the treatment under investigation, were the groups treated equally? (Yes/No; comments on the way patients have been treated)

1.7. Are all relevant outcomes measured in a standard, valid and reliable way? (Comments on concerns about outcomes and their measurement)

1.8. What percentage of the individuals or clusters recruited into the study are included in the analysis?

1.9. Were all the subjects analysed in the groups to which they were randomly allocated? (Comments and ITT analysis: Yes/No)

1.10. Are the results homogeneous between sites? (comments on any concerns)

2.1. How well was the study done to minimize the bias? (summarize the comments raised while answering previous questions)

2.2. What is the likely direction in which the bias might effect the study results? (any thoughts on summaries in 2.1)

2.3. Overall (based on the methods and statistical power) are the observed effects due to the study intervention? (comments)

2.4. Are the results of this study directly applicable (generalisable) to the patient group targeted by this guidelines? (comments on similarities and differences between the study sample, and its setting with the guideline’s target population, etc.)

APPENDIX 4

Case Series Quality Assessment form.

1. Is the aim/objective of the study clearly described?

2. Are the main outcomes to be measured clearly described in the Introduction or Methods section?

3. Are the inclusion/exclusion criteria explicit?

4. Are the characteristics of the patients included in the study clearly described?

5. Were the subjects asked to participate in the study representative of the entire population from which they were recruited?

6. Did all individuals enter the study at a similar point in their disease progression?

7. Was patient selection consecutive?

8. Was data collection prospective?

9. Was the recruitment period clearly defined?

10. Does the study provide estimates of the random variability in the data for the main outcomes?

11. Were the main outcome measures used accurate (valid and reliable)?

12. Have all important adverse events that may be a consequence of the intervention been reported?

13. Are the main findings of the study clearly described?

14. Were the statistical methods used to assess the main outcomes appropriate?

15. Do the analyses adjust for different lengths of follow-up of patients?

16. Have the characteristics of patients lost to follow-up been described and have these losses been taken into account?

17. If comparisons of subseries are being made, was there sufficient description of the series and the distribution of prognostic factors?

18. Was the funding source made explicit if applicable?

Ancillary