SEARCH

SEARCH BY CITATION

Keywords:

  • benign prostatic hyperplasia;
  • progression;
  • systematic review;
  • trials;
  • placebo

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

OBJECTIVE

To systematically review the placebo arms of randomized trials of medical therapy for benign prostatic hyperplasia (BPH) and thus estimate rates of progression.

METHODS

We searched PubMed, the Cochrane Trials Register and the USA Food and Drug Administration website. Retrieved citations were reviewed in two stages: a title and abstract screen, followed by a review of selected full-length articles. The inclusion criterion for studies were trials of BPH with >100 patients in the placebo group.

RESULTS

From 1774 citations screened, 16 eligible studies were found with a total of 12 158 patients. The range of mean baseline age was 62.6–66.5 years, for prostate volume 33.9–61.0 mL and for maximum urinary flow 8.6–11.6 mL/s. With studies spanning 12–48 months, the rates of surgery were 1–10% and for acute urinary retention 0.4–6.6%. Most studies showed progression, based on a worsening in clinical outcomes of the change from baseline in prostate volume and maximum urinary flow. Statistical heterogeneity prevented a quantitative synthesis of the data.

CONCLUSION

This systematic review confirms BPH disease progression in the form of increased prostate volume, reduction in maximum urinary flow rate and an increase in the risk of acute urinary retention and surgery. To provide quantitative estimates of effect, access to data from individual participants would be required.


Abbreviations
AUR

acute urinary retention

Qmax

maximum urinary flow rate

MTOPS

Medical Therapy of Prostatic Symptoms

FDA

Food and Drug Administration.

INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

BPH is the most common urological condition in men aged ≥50 years; depending on the exact definition used, a half to two-thirds of men in their seventh decade will have the condition [1]. As the disease progresses the prostate volume tends to increase, maximum urinary flow rate (Qmax) decreases, symptoms worsen and the likelihood of serious long-term complications of acute urinary retention (AUR) and the need for surgery (as well as bleeding, infection and renal impairment) increase [2,3].

Using ultrasonography, the community-based Olmstead County Study found an annual growth rate of the prostate of 1.6% in men aged 40–79 years, over a 7-year period [2]. Qualitatively similar results were found in the Baltimore Longitudinal Study [4]. Unfortunately, prostate size does not correlate closely with symptoms or Qmax, although the larger the size the greater the likelihood of future clinical deterioration.

Modifiers of disease progression have been identified by analyses of placebo groups of selected trials and longitudinal studies. Analysis of placebo-treated men who had pelvic MRI found that serum PSA level was strongly correlated with future prostate growth [5]. Of 881 placebo-treated men over 4 years, worsening of symptoms and Qmax developed only in those with a prostate volume of >40 mL and a PSA level of ≥1.4 ng/mL [6]. From the analysis of the same trial, the PSA level also predicted the 4-year incidence of AUR and need for surgery [6]. Similar results were reported from the analysis of the Medical Therapy of Prostatic Symptoms (MTOPS) [7]. The predictive value of other variables, e.g. Qmax and symptoms, remains uncertain [8].

Although the notion of disease progression in BPH is generally accepted, there remains uncertainty about both the rate and the determinants of progression. To try to address this uncertainty, we systematically reviewed BPH progression as reported within the placebo arms of randomized trials of medical therapy for BPH.

METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

A literature search was conducted followed by the application of explicit criteria to distil the titles and abstracts found by the search. The distilled list provided the basis for seeking the full articles for a further review for eligibility, and the same inclusion criteria were again applied. The reference list of the full articles was searched for other references. Those full articles that remained eligible were reviewed and the data abstracted and tabulated. Abstracts and full papers in English, Spanish, French and Italian were treated in the same manner as described above. Medline and the Cochrane Controlled Trials Database were searched. The USA Food and Drug Administration (FDA) website was also searched for placebo trial data. The search strategy for Medline, over the years 1951–2005, used the following keywords terms: ‘Benign prostatic hyperplasia’ (Medical Subject Heading) and (Inciden* or cohort or prospective or population-based or prognos* or predict* or trial). The search in the Cochrane Library used ‘benign prostatic hyperplasia’. The FDA website was searched with the following terms: ‘benign prostatic hyperplasia’‘NDA’.

Included were randomized trials of patients with BPH, and excluded were articles if they had: no placebo group, no relevant data; review article with no data; BPH with other diseases, e.g. prostatic cancer, where BPH was not analysed separately; duplicate publications; cross-over trials; open follow-up on therapy only. A follow-up of <6 months was chosen as an exclusion criterion after examining the range of duration of the studies, but before any analysis was conducted.

The following characteristics of the studies were abstracted according to a pre-designed format: study title and year of publication; country; study design and type of population; duration of intervention; stage(s) of disease at baseline; mean (and/or range) age at baseline; randomized or non-randomized study; loss to follow-up in the placebo group; background treatment and management; numbers at baseline in the placebo group; comments, definitions of disease and progression measures and other relevant aspects; on-treatment analysis. For binary variables, incidence rate or frequencies and proportions were abstracted. For continuous variables the means, mean differences, sds, pooled sds, sems, pooled sems, P values and 95% CI at baseline and at the end of the study for prostate volume, Qmax and symptom scores were abstracted. Baseline PSA levels and sds, sems, P values and 95% CI were abstracted.

Two independent observers reviewed and abstracted data from the full articles, and reviewed and abstracted data from abstracts where full articles were not obtainable. Any disagreements were resolved by discussion.

Data were analysed for each outcome separately and by duration of study. The change from baseline was the measure of choice for the analysis of continuous variables. The absolute rate or frequency was the measure used for binary outcomes. Fixed and random effects models were used. Statistical tests of homogeneity were used on each outcome, with P = 0.1 as the criterion for rejecting the null hypothesis. There was also visual examination of Forest plots to explore heterogeneity. Binary outcomes are presented as log odds probabilities and 95% CIs. Continuous outcomes are presented as the mean weighted difference and 95% CIs. Mean baseline variables (prostate volume, Qmax, symptom scores, PSA levels and age) were analysed as covariates in meta-regression models; the Stata statistical package (Stata Corp, College Station, TX, USA) was used for the analysis [9].

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

In all, 1774 abstracts and references were reviewed from a search of both Medline and the Cochrane Controlled Trial Register. From the list, 1689 were found to be ineligible, leaving 85 that were selected for review of the full article; 29 studies were found to be eligible, leaving 16 studies after duplicate publications had been excluded. Five references to new drug applications were found on the FDA website, but no abstractable data complying with the inclusion and exclusion criteria were found.

The characteristics of the 16 included trials with abstractable data are detailed in Table 1[10–23]. In summary, they were published between 1992 and 2005, and all were double-blind placebo-controlled trials. They originated in the USA and Europe, except one which included patients from New Zealand and Mexico [13]. One study was a pooled analysis of three similar double-blind placebo-controlled trials of the same drug [21]. The trials varied in the inclusion and exclusion criteria adopted. For example, the use of other medications, e.g. α-blockers and 5α-reductase inhibitors, for BPH varied. Most had a high PSA level as an exclusion criterion. The treatment arms of the trials from which the placebo data arose varied from finasteride, terazosin, alfuzosin, doxazosin, saw palmetto and dutasteride, or a combination. The range of mean age was 62.6–66.5 years, but the age range of men in many individual studies was 40–>80 years. The duration of the studies was 1–4 years, the range of mean prostate volume 33.9–61.0 mL and of Qmax 8.6–11.6 mL/s. The stage of disease was rarely reported in the studies, except indirectly by prostate volume or symptom scores.

Table 1.  Characteristics of trials of BPH included in the analysis
StudyDiagnosis/inclusion criteria (age, years), {Qmax, mL/s}Study duration, monthsTherapy arm(s)N in placebo armMean baseline (for placebo arm)
Age, yearsPSA, ng/mLProstate vol, mLQmax, mL/sSymptom score
  1. NR, not reported; SS, symptom score; SI, symptom index; BI, bother index; PVR, postvoid residual urine volume.

[10](≤80); {≥5, ≤15}≥2 symptoms indicating moderate BPH; PSA ≤10 ng/mL; PVR ≤150 mL24Finasteride 34665.5NR41.710.513.1
[11](>49); AUA score ≥8; {<15}12Saw palmetto  11363 1.633.9 11.615
[12](>45); PSA ≤10 ng/mL; moderate -to-severe symptoms12Finasteride 59665.1NRNRNRNR
[13](40–89); {<15}; voided volume >150 mL12Finasteride 25566 5.746.3 8.618.2
[14](40–89); {<15}; voided volume >150 mL12Finasteride 30064 4.161.0 9.6 9.8
[15](50–80); IPSS ≥12; {≥5}; voided volume ≥150 mL12Doxazosin; finasteride; combined 25364 2.63610.817.2
[16](45–80); mean SS ≥8; {mean 4–15}; voided volume ≥125 mL; mean PVR <300 mL12Finasteride, terazosin, finasteride + terazosin 30565 2.438.410.415.8
[17](50–75); {5–15}; voided volume 150 mL; PSA < 10 ng/mL; PVR < 150 mL24 48Finasteride Finasteride1591 151663.4 64NR  2.839.2 5510.9 1114.3 15
[18]Moderate to severe symptoms {<15}; voided volume ≥150 mL        
[7](≥50); AUA SS 8–30; {4–15}; voided volume ≥125 mL48Doxazosin combination 73762.6 2.335.210.516.8
[19](≤80); {5–15}; ≥2 moderate symptoms of BPH; voided volume ≥150 mL; PSA = 10 ng/mL; PVR ≥150 mL24Finasteride 30363NR45.810.916.6
[20](≥55); AUA-SS ≥13 and AUA-BS ≥812Terazosin 103165.7 2.4NR 9.620.1
[6]Moderate-to-severe, based on symptoms; {<15}; normal prostate biopsy if PSA 4.0–9.9 ng/mL48 150364 2.855 1115
[21](≥55); 6-month history of LUTS related to BPH; IPSS ≥13; {5–12} voided volume ≥150 mL; PVR ≥350 mL; PSA 1.4–10 ng/mL24Alfuzosin 76366.5 3.650 8.819.2
[22](≥50); prostate (TRUS) ≥30 mL; AUA-SI ≥12; {≤15}; PSA ≥ 1.5 ng/mL24 Dutasteride215866.1 454.010.417.1
[23](≥45); AUA SS 9–35; PSA ≤10 ng/mL12Finasteride 57963.4NRNRNR18.35

Many studies did not report sufficiently adequately the sd or sem of the change analyses for use of the data in this analysis, and hence the number of useable studies for each of the outcomes examined varies. In some studies CIs were provided with no sems or sds, and thus we assumed that the sd was a quarter of the 95% CI to extract more useable data.

On initial examination it was clear that the rates of progression were sufficiently heterogeneous, rendering a formal meta-analysis inappropriate. To explain the heterogeneity, we fitted models with age and duration, singly and together. Linear trends with age and duration were assumed because there were too few subjects to assess nonlinear trends. In the light of these initial findings, a descriptive systematic review of the data was carried out, together with Forest plots of the results, but without quantitative pooling.

Based on 14 studies (Fig. 1), the probability of surgery varied enormously by study, at 1–10%. Due to the heterogeneity the analysis was repeated for a duration of 1, 2 and 4 years (not shown). The risk of surgery depended on duration; the risk after 2 years was about twice that after 1 year, but there was still enormous statistical heterogeneity within the duration groups. There was only one study with a duration of ≥4 years.

image

Figure 1. The risk of surgery in 14 studies.

Download figure to PowerPoint

Based on 11 studies (Fig. 2), the probability of AUR varied by study, at 0.4–6.6%. Due to the heterogeneity the analysis was repeated for a duration of 1, 2 and 4 years (not shown). The risk of AUR depended on duration; the risk after 2 years was about twice that after 1 year, but there was still enormous statistical heterogeneity within the duration groups. There was only one study with a duration of ≥4 years.

image

Figure 2. Risk of acute urinary retention in 11 studies: details as Fig. 1.

Download figure to PowerPoint

Figure 3 shows the absolute change in prostate volume from baseline. All of the studies did not report data that permitted sems to be calculated and these are depicted solely as point estimates. Due to significant heterogeneity, no pooled estimate was calculated; the values were 0.8–11.5 mL. Age, prostate volume, Qmax or PSA level at baseline did not appear to be correlated with the degree of change.

image

Figure 3. Prostate volume change from baseline.

Download figure to PowerPoint

Figure 4 shows the absolute change in Qmax from baseline; due to heterogeneity, no pooled estimate was calculated, and values were – 0.3 to 1.4 mL/s. Age, prostate volume, Qmax or PSA level at baseline did not appear to be correlated with the degree of change.

image

Figure 4. Maximum urinary flow change from baseline: details as Fig. 3.

Download figure to PowerPoint

Figure 5 shows the change in symptom score from baseline. Due to heterogeneity, no pooled estimate was calculated. Age, prostate volume, Qmax or PSA level at baseline did not appear to be correlated with the degree of change.

image

Figure 5. Change in symptoms from baseline: details as Fig. 3.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

The principal finding relates to the unexpected degree of heterogeneity seen in all the pre-specified outcomes. The extent of this heterogeneity rendered inappropriate any formal pooling to provide a single summary of the studies.

The rates of surgery were 1–10% and the rates of AUR 0.4–6.6%; these rates tended to be worse with a longer follow-up. When mean study age and mean study duration were accounted for, there remained considerable heterogeneity, the implication being that the studies differed in important and significant ways. The differences in overall design, the types of patients that were randomized and the way in which patients allocated to the placebo arms were managed means that it is invalid to aggregate the data with a view to generating a single estimate of the rate of BPH-related progression.

Before considering the results further it is important to address three methodological issues that might have resulted in an underestimate of the rates of disease progression. The first relates to the presence of forced unilateral regression to the mean occurring in some of the variables that constitute disease progression. This phenomenon occurs because studies of BPH tend to have as their inclusion/exclusion criteria requirements for certain variables (Qmax, urinary symptom score, etc.) to be above or below a certain value. The values of these variables tend to fluctuate over time. It is understandable therefore that if men are only entered into a study if their Qmax is e.g. <12 mL/s, that it is a statistical likelihood that the next time it is measured it will be greater than this. This regression to the mean will have the effect of masking some early progression events in men allocated to the placebo group.

The second relates to censoring; most of the trials censored patients when a progression event occurred. Most commonly this is a worsening in urinary symptoms, as expressed by an increase in a symptom score of a given magnitude. Therefore progression events that might have occurred subsequent to the first event will not be recorded. Although this method of analysing progression events provides sound estimates of time to progression, it almost certainly underestimates the burden of progression. At least one non-randomized study of progression in a treated cohort identified symptom deterioration as a major risk factor for both AUR and surgery [24].

The third relates to the adequacy of reporting standards in the studies that were included. In several studies the data were incomplete, in others they were presented in a way that did not permit quantitative synthesis.

Whilst these limitations impede our ability to conduct a quantitative analysis, the results as obtained can be presented descriptively. All but two studies [10,11] showed progression, as evidenced by a change from baseline to increased prostate size and a reduction in Qmax. By contrast, symptom scores improved in all but two studies [10,13], probably as a result of the ‘regression to the mean’ effects described above, in addition to possible ‘white coat’ effects.

In the present analysis, none of mean age, study duration, the presence of a placebo run-in, prostate volume, Qmax, symptom score or PSA level correlated with the change from baseline. This is in contrast to other studies that identified PSA level to be positively associated with both AUR and future surgery [5,6]. However, one of these studies only reported an association between PSA level and worsening symptoms and Qmax in men with a PSA level of ≥1.4 ng/mL and a prostate volume of >40 mL [6]. We think that this difference arose from a lack of homogeneity both within patients in individual studies and between studies in the aggregate data available for this review.

It is difficult to estimate the extent to which these results can be used to inform an individual on his future likelihood of progression. Given the differences that we observed among studies, it is likely that any generalization to a ‘normal patient’ would be of limited value [24]. It is clearly difficult to generalize results of disease progression to patients who are very different from those in the trials. This applies to making comparisons with community-based cohort studies. Table 2 shows the differences for several variables of disease severity between the range of mean values at baseline in the trials of this review and the mean values found in a community-based cohort study [25]. As expected, the trial patients were older, had more symptoms, a lower Qmax, a larger prostate volume and higher PSA levels. A valid comparison requires rates of progression between subjects in trials with similar subjects in a community cohort. Although the community-based cohort study from Olmsted County found higher rates of surgery, AUR and symptom progression than in the MTOPS trial, another community-based study found a 12% rate of surgery over a median period of 2.4 years [26]. This compares with 2.6–10.2% in trials of 2 years’ duration in the present review. Therefore, it would appear that the rate of progression in the trials underestimates the rate of progression in the community, although community- based cohort studies also appear to be heterogeneous in their design, especially the duration of follow-up [2,4,25,26], and generalizing from them might also not provide the entire answer. To advance the development of a valid comparison requires pooling individual participant data from the trials, and comparing this with individual participant data from cohort studies stratified by important variables such as age and duration of follow-up.

Table 2.  A comparison of the reviewed trials and the community-based Olmsted County cohort study [25]
 [25] means*Trial-like cohortSystematic review (range of means)
  • *

    Based on MTOP trial eligibility.

Age, years61.56463.0–66.5
Symptom score (AUA-SI) 6.612.815.0–20.1
Qmax, mL/s18.3 11 8.6–11.6
Prostate volume, mL3236.733.9–61.0
PSA, ng/mL 1.7 2.4 1.6–5.7

In this systematic review we detected disease progression in the studies included in the analysis. Progression was observed in terms of increasing prostate volume, decrease in Qmax and an increase in the future risk of AUR and surgery. The only way that any precision could be applied in relation to the rates of progression would be to obtain and undertake a similar analysis of the individual-participant data. Without such data, a single (pooled) estimate is unlikely to be representative of the population from which the patients arose.

ACKNOWLEDGEMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

The study was funded by GlaxoSmithKline SA, Spain, the manufacturer of dutasteride, a drug used to treat benign prostatic hyperplasia.

CONFLICT OF INTEREST

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES

M.G.L. is an employee of GSK; M.E. is an Investigator in the GSK-sponsored Reduce trial in the UK. He has received lecture fees and has acted as a paid consultant for GSK. He is supported by the Comprehensive Biomedical Centre located at University College London and University College London Hospitals NHS Trust. J.M.F. and B.D. have acted as paid consultants for GSK. N.Q. is Director of OXON Epidemiology Limited, which carries out consultancy work for GSK. Contributors: N.Q. and M.G.L. contributed to the design, analysis, interpretation, and drafted and revised the manuscript. M.E., J.M.F., R.D. contributed to the interpretation and revision of the manuscript. N.Q. and M.G.L. are the guarantors. Ethical approval was not applicable as all data were derived from published articles.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. ACKNOWLEDGEMENTS
  8. CONFLICT OF INTEREST
  9. REFERENCES
  • 1
    Garraway WM, Collins GN, Lee RJ. High prevalence of benign prostatic hypertrophy in the community. Lancet 1991; 338: 46971
  • 2
    Rhodes T, Girman CJ, Jaconsen SJ, Roberts RO, Guess HA, Lieber MM. Longitudinal prostate growth rates during 5 years in randomly selected community men 40–79 years old. J Urol 1999; 161: 1174
  • 3
    Homma Y, Imago C, Takahahi S, Kawabe K, Aso Y. Urinary symptoms and urodynamics in a normal elderly population. Scand J Urol Nephrol (Suppl) 1994; 157: 27
  • 4
    Williams AM, Simon I, Landis PK et al. Prostates growth rate determined from MRI data: age-related longitudinal changes. J Androl 1999; 20: 474
  • 5
    Bosch JL, Bohnen AM, Groeneveld FP. Validity of digital rectal examination and serum prostate specific antigen in the estimation of prostate volume in community-based men aged 50–78 years: the Krimpen Study. Eur Urol 2004; 46: 7539
  • 6
    Roehrborn CG, McConnell JD, Lieber MM et al., for the PLESS Study Group. Serum prostate specific antigen concentration is a powerful predictor of acute urinary retention and need for surgery in men with clinical benign prostatic hyperplasia. Urology 1999; 53: 47380
  • 7
    McConnell JD, Roehrborn CG, Bautista OM et al., Medical Therapy of Prostatic Symptoms (MTOPS) Research Group. The long-term effect of doxazosin, finasteride, and combination therapy on the clinical progression of benign prostatic hyperplasia. N Engl J Med 2003; 349: 238798
  • 8
    Marks LS, Roehrborn CG, Andriole GL. Prevention of Benign Prostatic Hyperplasia Disease. J Urol 2006; 176: 1299306
  • 9
    StataCorp. Stata Statistical Software: Release 8. College Station, TX: StataCorp LP, 2003
  • 10
    Andersen JT, Ekman P, Wolf H et al. Can finasteride reverse the progress of benign prostatic hyperplasia? A two-year placebo-controlled study. The Scandinavian BPH Study Group. Urology 1995; 46: 6317
  • 11
    Bent S, Kane C, Shinohara K et al. Saw palmetto for benign prostatic hyperplasia. N Engl J Med 2006; 354: 55766
  • 12
    Byrnes CA, Morton AS, Liss CL, Lippert MC, Gillenwater JY, on behalf of the CUSP Investigators. Efficacy, tolerability, and effect on health-related quality of life of finasteride versus placebo in men with symptomatic benign prostatic hyperplasia: a community-based study. Clin Ther 1995; 17: 95669
  • 13
    Finasteride Study Group. Finasteride (MK-906) in the treatment of benign prostatic hyperplasia. Prostate 1993; 22: 2919
  • 14
    Gormley GJ, Stoner E, Bruskewitz RC et al. The effect of finasteride in men with benign prostatic hyperplasia. The Finasteride Study Group. N Eng J Med 1992; 327: 118591
  • 15
    Kirby RS, Roehrborn C, Boyle P et al., Prospective European Doxazosin and Combination Therapy Study Investigators. Efficacy and tolerability of doxazosin and finasteride, alone or in combination, in treatment of symptomatic benign prostatic hyperplasia: the Prospective European Doxazosin and Combination Therapy (PREDICT) trial. Urology 2003; 61: 11926
  • 16
    Lepor H, Williford WO, Barry MJ et al. The efficacy of terazosin, finasteride, or both in benign prostatic hyperplasia. Veterans Affairs Cooperative Studies Benign Prostatic Hyperplasia Study Group. N Engl J Med 1996; 335: 5339
  • 17
    Marberger MJ. Long-term effects of finasteride in patients with benign prostatic hyperplasia: a double-blind, placebo-controlled, multicenter study. The PROWESS Study Group. Urology 1998; 51: 67786
  • 18
    McConnell JD, Bruskewitz R, Walsh P et al. The effect of finasteride on the risk of acute urinary retention and the need for surgical treatment among men with benign prostatic hyperplasia. The Finasteride Long-Term Efficacy and Safety Study Group. N Engl J Med 1998; 338: 6123
  • 19
    Nickel JC, Fradet Y, Boake RC et al., the PROSPECT Study Group. Efficacy and safety of finasteride therapy for benign prostatic hyperplasia: results of a 2-year randomized controlled trial (the PROSPECT Study). Can Med Assoc J 1996; 155: 12519
  • 20
    Roehrborn CG, Oesterling JE, Auerbach S et al. The Hytrin Community Assessment Trial study: a one-year study of terazosin versus placebo in the treatment of men with symptomatic benign prostatic hyperplasia. HYCAT Investigator Group. Urology 1996; 47: 15968
  • 21
    Roehrborn CG, Boyle P, Nickel JC, Hoefner K, Andriole G, on behalf of the ARIA 3001, ARIA 3002 and ARIA 3003 Study Investigators. Efficacy and safety of a dual inhibitor of 5-alpha-reductase types 1 and 2 (dutasteride) in men with benign prostatic hyperplasia. Urology 2002; 60: 43441
  • 22
    Roehrborn CG. Alfuzosin 10 mg once daily prevents overall clinical progression of benign prostatic hyperplasia but not acute urinary retention: results of a 2-year placebo-controlled study. BJU Int 2006; 97: 73441
  • 23
    Tenover JL, Pagano GA, Morton AS, Liss CL, Byrnes CA. Efficacy and tolerability of finasteride in symptomatic benign prostatic hyperplasia: a primary care study. Primary Care Investigator Study Group. Clin Ther 1997; 19: 24358
  • 24
    Emberton M, Hartung R, Matzkin H et al. Symptomatic progression under treatment and history of AUR are the strongest predictors for AUR in men with LUTS treated with alfuzosin 10 mg once daily. Urology 2005; 66: 31622
  • 25
    Roberts RO, Lieber MM, Jacobson DJ, Girman CJ, Jacobsen SJ. Limitations of using outcomes in the placebo arm of a clinical trial of benign prostatic hyperplasia to quantify those in the community. Mayo Clin Proc 2005; 80: 75964
  • 26
    Souverein PC, Van Riemsdijk MM, De La Rosette JJ, Opdam PC, Leufkens HG. Treatment of benign prostatic hyperplasia and occurrence of prostatic surgery and acute urinary retention: a population-based cohort study in the Netherlands. Eur Urol 2004; 47: 50510