This article is mentioned in:

Introduction

In the same way that benign prostatic hyperplasia (BPH) is a widespread age-related pathological condition that affects men worldwide, prostatism-like symptoms in women, referred to as female lower urinary tract symptoms (LUTS), are also common, mainly including voiding (‘obstructive’) and filling (‘irritative’) symptoms, such as weak stream, hesitancy, intermittency, nocturia, daytime frequency and urgency (1). LUTS are highly prevalent in women, particularly perimenopausal women, but they are rarely reported as they are considered to be part of the aging process and it is assumed that no effective treatment is available. LUTS increase and become more aggravating with advancing age in the majority of individuals (2–6). Voiding symptoms were observed to be more common than filling symptoms in a study of female patients with LUTS who visited a urological clinic. Furthermore, functional bladder outlet obstruction was more prevalent than detrusor underactivity in these female patients (7).

A similarly high prevalence of filling and voiding LUTS in men and women suggests that certain aspects of the underlying etiology may be identical. Adrenergic receptors (adrenoceptors, ARs) were originally categorized into αAR and βAR subgroups. However, the application of molecular biological methods in the last few years has confirmed a total of nine AR subtypes: α1A (formerly named α1c), α1B, α1D, α2A, α2B, α2C, β1, β2 and β3 (8). The α1DAR predominates in the female detrusor and spinal cord (9,10). The α1AAR is expressed at significantly higher levels than other α1AR subtypes are in the female urethra (11,12). These findings suggest that voiding and filling symptoms have a correlation with the expression of these α1AR subtypes, and indicates that α1-adrenoceptor antagonists (α1ARAs) may be used as a potentially novel treatment for female patients with LUTS.

In clinical practice, however, the use of α1ARAs to treat LUTS in women has been adopted based on limited studies, anecdotal case reports and local experience. The specific mechanism by which α1ARAs act in the treatment of female LUTS has not been established; however, some small sample-size clinical trials have confirmed that α1ARAs are able to significantly improve voiding and filling symptoms in female patients with LUTS (13–16).

A meta-analysis was, therefore, carried out to determine the effectiveness of α1ARAs versus placebo for female patients with LUTS.

Materials and methods

Search strategy

Pubmed (1966–2013), Embase (1974–2013), the Cochrane Library (issue 4, 2013), the Chinese Biomedical Literature database (1978–2013), the Chinese Sci-Tech Periodical Full-Text database (1989–2013) and the Chinese Periodical Full-Text database (1994–2013) were searched for randomized controlled trials in which α1ARAs, terazosin, tamsulosin, doxazosin or alfuzosin were compared with placebo in female patients with LUTS. Related references and included studies on the Internet, according to search engines such as Google™, were also searched, and a manual search was used to find key Chinese publications in associated fields. The reference lists of included studies and reviews were searched by hand and experts in the field were contacted; unpublished studies were not sought. No limits based on language were imposed.

This search strategy was used to obtain the titles and abstracts from randomized controlled trials associated with the subject matter of this review. The titles and abstracts were screened independently by two reviewers (Long Cheng and Hao-Han Wang), who discarded studies that were inapplicable. Two reviewers independently assessed the retrieved titles and abstracts of all the identified trials to confirm fulfillment of the inclusion criteria. Disagreements were resolved in consultation with Xiao-Kan Xiong. Data extraction was carried out independently by the same authors using standard data extraction forms.

Inclusion criteria

The patients were women with LUTS, aged 20 to 70 years, with a total I-PSS ≥8. Written informed consent was provided. Excluded cases were pregnancy, breastfeeding, stress incontinence, urinary tract infection, neurological diseases including diabetes mellitus with neuropathy, previous pelvic surgery or radiation, medical conditions mimicking LUTS and concomitant medications affecting the lower urinary tract.

Types of outcome measures

The main outcome measures were total I-PSS and maximum urinary flow rate (MUFR). The I-PSS quality of life (QOL) and average urinary flow rate (AFR) were also observed and analyzed.

Types of intervention

The types of intervention were α1ARAs versus placebo.

Methodological quality assessment and level of evidence

The Cochrane collaboration tool in Review Manager, version 5.2 (Cochrane Collaboration, Copenhagen,. Denmark) was used for assessing the risk of bias in order to evaluate the methodological quality of each randomized controlled trial. The Grades of Recommendation, Assessment, Development and Evaluation (GRADE) approach was applied to assess the level of evidence, and GRADEprofiler software, version 3.6 (Cochrane Collaboration) was used to create the evidence profile.

Statistical analysis

The data were analyzed using Stata (version 12.0; Stata Corporation, College. Station, TX, USA) and data were extracted and pooled for summary estimates. Results are expressed for continuous outcomes as weighted mean difference or standardized mean difference, and dichotomous outcomes as relative risk with 95﹪ confidence intervals (CI). The χ2 statistical test was used to assess heterogeneity between trials and the Ι2 statistical test was used to assess the extent of inconsistency. A fixed effect model was used for calculations of summary estimates and their 95﹪ CI, unless there was significant heterogeneity, in which case results were confirmed using a random effects statistical model. Subgroup analyses were designed to explore important clinical differences that may be expected to alter the magnitude of treatment effects.

Results

Search results

The selection process is depicted in Fig. 1. Seven potentially eligible trials were identified and five trials were subsequently excluded for the following reasons: Two were not controlled; one failed to extract necessary data such as the mean and standard deviations of the outcome index; one did not include a placebo group; and one was an unpublished conference article that contained no detailed data. Two randomized controlled trials totaling 213 patients [133 in the study by Pummangura et al (16) and 80 in the study by Low et al (15)] were included. The two trials reported two types of outcome including total I-PSS and MUFR. Patients were matched for age and the severity of urinary symptoms. Co-morbidities that clinically indicated symptoms similar to LUTS were not sought.

Figure 1. Flowchart for selecting randomized controlled trials (RCTs) for the meta-analysis.

Assessing the risk of bias, and characteristics of included studies

The main characteristics of the two included studies are shown in Table I. The new ‘Risk of bias’ tool in Review Manager 5.2 was used to assess risk of bias (Table II).

Table I. Characteristics of the two studies.

Table I.

Characteristics of the two studies.

			No.		Intervention			TI-PSSa			MUFRa			I-PSS QOLa			AFRa
Study (ref.)	Gender	Age(years)	T	C	T	C	Pharmacologyb	T	C	P-value	T	C	P-value	T	C	P-value	T	C	P-value
15	Female	20–70	40	40	Terazosin	Placebo	α1A=α1B=α1D	15.6	14.38	>0.05	21.73	20.13	>0.05	4.78	5.03	>0.05		–
16	Female	27–69	65	68	Tamsulosin	Placebo	α1A=α1D>α1B	18.2	22.50	0.001c	18.00	18.80	>0.05		–		7.00	7.70	>0.05

a Mean baseline value.

b Effects of α1ARAs blocking three α1AR subtypes.

c Scatter plot and Spearman's rank correlation showed that the difference in baseline TI-PSS between the two treatment groups had no significant effect on the difference in mean change from baseline in TI-PSS between groups. T, treatment group; C, controlplacebo group; MUFR, maximum urinary flow rate; TI-PSS, Total International Prostate Symptom Score; QOL, quality of life; AR, adrenergic receptor; ARA, andrenergic receptor antagonist.

Table II. Assessment of the risk of bias in the studies.

Table II.

Assessment of the risk of bias in the studies.

Study (ref.)	Random sequence generation	Allocation concealment	Blinding of participants and personnel	Blinding of outcome assesment	Incomplete outcome data	Selective reporting	Other bias
15	Low riska	Unclear riska	Low riskb	Low riskc	Low riskd	Low riske	Low riskf
16	Low riskg	Low riskg	Low riskh	Low riskc	Low riski	Low riskj	Low riskf

a Randomized controlled trial

b terazosin and placebo were used similarly when given to subjects

c double blind

d results of intent-to-treat analysis and per-protocol analysis are similar

e research plan and outcomes reported in paper

f no significant other bias found

g block randomization size of four was carried out by computer-generated random numbers

h study medications were packaged in a concealed card to maintain blinding in the pharmacy unit

i no effect of missing subjects on the trial outcome

j there is sufficient reason to verify that the statistically significant difference in baseline International Prostate Symptom Score between the two group (P=0.001) did not affect the trial outcome.

Meta-analysis results

Results of the analysis are depicted in Fig. 2. Following almost 4 weeks of treatment, the MUFR post-treatment [standardized mean difference (SMD, −0.05; 95% CI, −0.32 to 0.22] between the experimental and control groups was found not to differ. The improvement of AFR was not significant (SMD, 0.09; 95% CI, −0.25 to 0.43) in the study conducted by Pummangura et al (16). The meta-analysis indicates that the two groups are inhomogeneous, due to a statistically significant difference (P=0.001) in the baseline I-PSS between experimental and control groups (16). In general, the total I-PSS was decreased in both groups following treatment compared with the baseline. In patients receiving α1ARAs and placebo, the total I-PSS following treatment was lower than that prior to treatment, but the total I-PSS was significantly lower in females treated with α1ARAs than in females treated with placebo (SMD, −0.67; 95% CI, −0.94 to −0.39). In addition, the I-PSS QOL post-treatment was lower in the α1ARA group compared with that in the placebo-treated group (SMD, −0.86; 95% CI, −1.32 to −0.40) (15).

Figure 2. Forest plots for the results of the meta-analysis: α1-adrenergic receptor antagonists versus placebo for female lower urinary tract symptoms. MUFR, maximum urinary flow rate; AFR, average urinary flow rate; I-PSS, International Prostate Symptom Score; QOL, quality of life; SMD, standardized mean difference; CI, confidence interval; B.Y.L, Low et al (15); N.P., Pummangura and Kochakarn (16).

GRADE profile of evidence

The quality of evidence in the included studies, as determined by the GRADE approach, is shown in Table III. Two critical outcome measures: Total I-PSS and I-PSS QOL, and two important outcomes: MUFR and AFR, were judged to indicate high-quality evidence.

Table III. GRADE profile of evidence of the included studies.

Table III.

GRADE profile of evidence of the included studies.

			Quality assessment					No. of patients
Outcome	No. of study	Design	Risk of bias	Inconsistency	Indirectness	Imprecision	Other considerations	α1−AR antagonists	Placebo	Absolute effect, SMD (95% CI)	Quality	Importance
TI-PSS	2	RCT	Not serious	Not serious	Not serious	Not serious	None	105	108	−0.67 (−0.94 to −0.39)	High	Critical
MUFR	2	RCT	Not serious	Not serious	Not serious	Not serious	None	105	108	−0.05 (−0.32 to 0.22)	High	Important
I-PSS QOL	1	RCT	Not serious	Not serious	Not serious	Not serious	None	40	40	−0.86 (−1.32 to −0.4)	High	Critical
AFR	1	RCT	Not serious	Not serious	Not serious	Not serious	None	65	68	0.09 (−0.25 to 0.43)	High	Important

[i] Study 2, Low et al (15); Study 1, Pummangura et al (16); AFR, average urinary flow rate. AR, adrenergic receptor; CI, confidence interval; I-PSS, International Prostate Symptom Score MUFR, maximum urinary flow rate; QOL, quality of life; RCT, randomized controlled trial; SMD, standardized mean difference; TI-PSS, Total International Prostate Symptom Score.

Discussion

A randomized double-blind study of 29 women in New York reported by Lepor and Theune (17) in 1995 indicated that terazosin was not effective for relieving prostatism-like symptoms. Another open non-randomized trial that used doxazosin, demonstrated that α1ARAs were at least as effective as the anticholinergic drug hyoscyamine in reducing the total I-PSS (13). These studies were excluded due to non-conformity of the inclusion criteria. In the study by Chang et al (14), which comprised 97 female patients and was not placebo controlled, the outcome suggested that tamsulosin improved voiding symptoms and urodynamic parameters in nearly one-third of women with voiding difficulty, and comparable good therapeutic response rates were observed between patients with bladder outlet obstruction and detrusor underactivity (14). This disagreement among studies of whether α1ARAs are more effective than placebo in the treatment of female LUTS was perplexing; hence, the collection and evaluation of studies concerning the efficacy of α1ARAs in the treatment of female LUTS in the present meta-analysis was important and it may provide clinicians with a temporary guideline for selecting α1ARA treatments.

Although the number of studies included in the present meta-analysis was small, the quality of the two studies and the reliability of the outcome measures were high. Due to the low number of participants it was not possible to reach a reliable conclusion; however, a low risk of bias and a high GRADE quality of evidence suggested that the outcome measures were reliable, at least in terms of the two included studies. The two studies mentioned randomized, double-blind, allocation concealment, and one of them included intent-analysis. The majority of participants were Asian, so any differences due to ethnicity could not be observed. In clinical trials conducted in the future, it may be appropriate to expand the sample size and select participants of different nationalities.

The bladder, bladder neck and urethra are responsible for urine storage and voiding in females. During the storage phase of the micturition cycle, the bladder relaxes to accommodate increasing volumes of urine at acceptable pressures, and the bladder neck and urethra contract to provide resistance to prevent involuntary leakage. During the micturition phase, the bladder neck and urethral muscle relax to allow the detrusor to contract and expel urine without major resistance. The α1DAR predominates in the female detrusor and the α1AAR is expressed at higher levels than other α1AR subtypes in the female urethra (including the bladder neck) (9–12). Therefore, the efficacy of α1ARAs for the treatment of female LUTS may be explained by the targeting of two possible mechanisms. The first is dysfunction of the bladder neck and urethra, causing functional outlet obstruction and secondary detrusor overactivity, which is similar to bladder outlet obstruction in men with BPH. The second possibility is increased α1AR activity in the detrusor, causing frequency and urgency (18); however, from the meta-analysis results, only total I-PSS and I-PSS QOL improved following α1ARA treatment with no alteration of the MUFR and AFR compared with those of the placebo groups. If α1ARAs relax the bladder neck and urethral muscle, it is possible that MUFR and AFR could improve. More studies are required to confirm whether the difference of total I-PSS and I-PSS QOL is associated with MUFR and AFR in the control group. In terms of terazosin and tamsulosin blocking α1A- and α1DARs (Table I), the bladder neck, urethral muscle and detrusor would be relaxed under the effect of these drugs. It is hypothesized that the fact that there was no clear alteration of the MUFR and AFR was as a result of the functional bladder outlet obstruction remitting and the detrusor pressure reducing over time. Highly selective α1AARAs may improve the MUFR and AFR for women with LUTS including functional bladder outlet obstruction along with possible detrusor overactivity.

In the present analysis, α1ARAs were indicated to be more effective than placebo in reducing total I-PSS and improving QOL in females with LUTS; however, MUFR and AFR did not increase significantly and it was not clear whether the alteration of total I-PSS and IPSS QOL was associated with changes of MUFR and AFR for females with LUTS under α1ARA treatment. As the sample size was small and the patients were from a limited geographical area, further clinical trials are required to expand the sample size and select participants of different ethnicities. More high-quality, multi-center, randomized controlled trials are required.

References