Efficacy and Safety of Long-TermTreatment with the Dual 5a-Reductase Inhibitor Dutasteride in Men with Symptomatic Benign Prostatic Hyperplasia

European Urology European Urology 46 (2004) 488 495 Efficacy and Safety of Long-TermTreatment with the Dual 5a-Reductase Inhibitor Dutasteride in Men with Symptomatic Benign Prostatic Hyperplasia Frans Debruyne a, *, Jack Barkin b, Peter van Erps c, Mario Reis d, Teuvo L.J. Tammela e, Claus Roehrborn f on behalf of the ARIA3001, ARIA3002 and ARIB3003 Study Investigators a 426 Department of Urology, University Hospital Nijmegen, Geert Grooteplein 10, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands b CMX Research Inc. and University of Toronto, Toronto, Ontario, Canada c Department of Urology, Middelheim Hospital, Antwerp, Belgium d Serviço de Urologia, Hospital S. João, Porto, Portugal e Tampere University Hospital and Medical School, University of Tampere, Tampere, Finland f The University of Texas Southwestern Medical Center, Dallas, TX, USA Accepted 3 May 2004 Available online 23 July 2004 Abstract Objectives: Dutasteride, a dual inhibitor of Type 1 and Type 2 5a-reductase, has been shown to improve disease measures in patients with symptomatic benign prostatic hyperplasia (BPH) in three randomised, placebo-controlled, large-scale, 2-year Phase III clinical studies. This paper reports the pooled results of a 2-year open-label extension of the three randomised studies assessing the long-term efficacy and safety of dutasteride. Methods: Patients randomised to dutasteride or placebo in the double-blind portion of the Phase III studies were eligible for a 2-year open-label extension, where all patients received dutasteride 0.5 mg daily (dutasteride/ dutasteride [D/D] group and placebo/dutasteride [P/D group]). Results: Significant improvements in AUA-SI score and Q max were observed from Month 24 to 48 in both study groups. At Month 48, patients in the D/D group had significantly greater improvements in AUA-SI score and Q max, and significantly greater reductions in prostate volume, than those in the P/D group. Acute urinary retention and BPH-related surgery occurred in a small percentage of patients during the open-label phase. No new safety issues were noted with long-term therapy. Onset of new drug-related adverse events were reported most frequently at the start of therapy and declined over time in patients receiving dutasteride. Conclusions: Long-term treatment with dutasteride results in continuing improvements in urinary symptoms and flow rate, and further reductions in TPV, in men with symptomatic BPH. The reduction in risk of AUR and BPHrelated surgery, seen in the double-blind phase, was durable over 4-year treatment. Dutasteride was also well tolerated in long-term use. # 2004 Elsevier B.V. All rights reserved. Keywords: BPH; Dutasteride; Lower urinary tract symptoms; 5a-reductase inhibitor; Long term 1. Introduction * Corresponding author. Tel. +31 243 613735; Fax: +31 243 541031. E-mail address: f.debruyne@uro.umcn.nl (F. Debruyne). Of the two commonly used pharmacotherapeutic interventions for men with symptomatic benign prostatic hyperplasia (BPH), the 5a-reductase inhibitors (5ARIs) and a 1 blockers, only the 5ARIs have been 0302-2838/$ see front matter # 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.eururo.2004.05.008

F. Debruyne et al. / European Urology 46 (2004) 488 495 489 shown to modify the underlying pathology. Data from three large-scale, randomised clinical studies demonstrate that treatment with the 5ARIs dutasteride or finasteride results in significant reductions in prostate volume associated with improvements in urinary symptoms and flow, and significant reductions in the risks of acute urinary retention (AUR) and BPH-related surgery [1 3]. In contrast, a 1 blockers, which reduce the dynamic elements of BPH but do not reduce prostate volume, are associated with rapid improvements in symptoms and flow, but have not been shown to reduce the long-term risk of AUR or BPH-related surgery in blinded, randomised and placebo-controlled studies [3,4]. BPH is a progressive disease in many men, and underlying this progression is continuing prostate growth [2,5 12]. Therapy aimed at reducing prostate volume, and the accompanying benefits in improving symptoms and lessening the risk of BPH morbidity, has a logical place in the medical management of men at risk of progressive BPH. However, the long-term nature of BPH requires a therapy that reduces and maintains reductions in prostate volume, has lasting effects on symptoms and minimizes the risks of AUR and BPH-related surgery. A number of studies with the 5ARI finasteride have examined its long-term effect over 4 to 8 years [1,3,13 15]. These studies demonstrate that finasteride treatment is associated with a maximal reduction in prostate volume after 2 years of therapy that is maintained without further reductions over the remaining course of the studies. 5ARIs achieve reductions in prostate volume through the inhibition of the nuclear-bound steroid 5a-reductase (5AR) isoenzymes, which chemically reduce testosterone to dihydrotestosterone (DHT), the principal androgen stimulating prostatic growth. Dutasteride is the first selective dual inhibitor of both Type 1 and Type 2 5AR, whereas finasteride is a selective mono-inhibitor of 5AR Type 2 [16,17]. Dual inhibition of 5AR with dutasteride results in nearcomplete and consistent suppression of serum DHT, with >85% of men achieving a 90% reduction within 4 weeks. Finasteride suppresses serum DHT by approximately 70%, with only 49% of treated men achieving this reduction in one data series [18]. The efficacy and safety of dutasteride in men with symptomatic BPH has been demonstrated in an analysis of three, large-scale, randomised, placebo-controlled Phase III studies (ARIA3001, ARIA3002 and ARIB3003). This paper examines the results from 2 years of double-blind treatment from these studies and an additional 2 years of open-label treatment, providing a total of 4 years of efficacy and safety data for dutasteride. The pattern of effect of long-term dual 5AR inhibition on prostate volume, urinary symptoms and flow, AUR and need for BPH-related surgery are reviewed. 2. Materials and methods ARIA3001, ARIA3002 and ARIB3003 were randomised, double-blind, placebo-controlled studies of the efficacy and safety of dutasteride 0.5 mg once daily in the treatment of men with symptomatic BPH. The design of these studies has been previously reported [2]. Principle inclusion and exclusion criteria are shown in Table 1. All three studies had an optional 2-year open-label extension period in which patients initially receiving dutasteride in the double-blind phase were maintained on dutasteride (D/D group), while those initially receiving placebo were switched to open-label dutasteride (P/D group). At the start of the double-blind phase, patients underwent a 1-month, single-blind placebo run-in period, and were subsequently randomised to treatment with dutasteride 0.5 mg or placebo once daily for 2 years. Patients who completed the 2 years of double-blind treatment were eligible to participate in an additional 2 years of open-label treatment with daily dutasteride 0.5 mg. Consenting patients were scheduled to return for assessments on an outpatient basis at 27, 30, 33, 36, 39, 42, 45 and 48 months postrandomisation. The objectives of the 2-year open-label phase were to assess the safety and tolerability of long-term dutasteride therapy. Efficacy endpoints reported from the studies were change in American Urological Association Symptom Index (AUA-SI), total prostate volume (TPV) and maximal urinary flow rate (Q max ). AUA-SI and Q max were evaluated at baseline and months 1, 3, 6, 12, 18, 24, 30, 36, 42 and 48. Transrectal ultrasound, including measurement of TPV, was carried out at baseline and 6, 12, 24 and 48 months in all three studies. TPV was also assessed at 1 month in ARIA3001 and at 3 months in ARIA3002. Samples for serum DHT and testosterone were obtained at baseline and at yearly intervals, and for serum prostate specific antigen (PSA) at baseline and months 1, 3, 6, 12, 18, 24, 36, and 48. Safety was assessed through adverse event (AE) reporting, clinical laboratory assessments and yearly physical examinations, which included focused Table 1 Principal inclusion and exclusion criteria for entry into the double-blind phase Inclusion criteria Men eligible aged 50 years Diagnosis of BPH by history and physical examination Prostate volume measured by transrectal ultrasonography of 30 cc American Urological Association-Symptom Index (AUA-SI) score 12 (moderate-to-severe symptoms) Maximal urinary flow rate (Q max ) 15 ml/s Serum prostate specific antigen level of 1.5 ng/ml Exclusion criteria Post-void residual volumes >250 ml History of prostate cancer Previous prostate surgery Prior history of AUR within 3 months of screening Use of an a 1 blocker within 2 weeks or any previous use of a 5ARI Serum prostate specific antigen level 10 ng/ml

490 F. Debruyne et al. / European Urology 46 (2004) 488 495 gynaecomastia evaluations. Data on AUR and BPH-related surgical events were collected throughout the 4-year period. The data from the three trials were pooled for analysis. For effects on DHT, testosterone, AUA-SI, Q max and TPV, results from the open-label intent-to-treat (ITT) population, who received at least one dose of study medication during the open-label phase, were reported (D/D vs. P/D). In addition, results from the completer population for change in AUA-SI score are reported. Completer analyses for changes in TPV and Q max, which in common with the AUA-SI data were similar to those for the ITT population, are not reported. Changes in reported variables were calculated from baseline values established at the start of the double-blind phase. Mean change from baseline or mean percent change from baseline ( standard deviation of the mean) were calculated for continuous parameters (AUA-SI, Q max, TPV, PSA) at each scheduled time point in the double-blind and open-label phases. Statistical comparisons between the two treatment groups (P/D and D/D) at the scheduled open-label time-points in terms of the change (or percentage change) from baseline for the continuous parameters were carried out using a general linear model with effects for baseline, treatment, protocol and investigator cluster. Statistical comparisons within each of the treatment groups comparing the difference between the Month 48 and Month 24 values were carried out using a t-test. Statistical significance was evaluated at the 0.05 level. As AUR and BPH-related surgery were events that resulted in the withdrawal of patients from the study, risk reductions calculated for AUR and BPH-related surgery were based on the double-blind ITT population (men who received at least one dose of study medication during the double-blind phase) for baseline to 4 years, to ensure that men who had either event in the double-blind phase, and did not therefore enter the open-label phase, were included. Adverse events reported in 1% of patients were reported by year for the double-blind ITT population (baseline to 2 years) and the open-label ITT population (2 to 4 years). 3. Results 3.1. Patient demographics A total of 4325 men were randomised into the double-blind phase of the studies: 2158 to placebo and 2167 to dutasteride (Fig. 1). For the open-label phase, 2340 patients were enrolled, 1188 of whom previously received dutasteride (D/D group) and 1152 Fig. 1. Subject accountability for double-blind and open-label study phases. of whom previously received placebo (P/D group). A proportion of subjects not entering the open-label phase were as a result of more than 25% of study centres electing not to participate further following the double-blind phase. However, 16 of the 19 countries participating in the double-blind phase also participated in the open-label phase. Table 2 includes data for patients from the open-label ITT population. For patients from the open-label ITT population, no significant differences at the start of the double-blind phase in baseline parameters were observed between patients in the D/D group and P/D group, except for higher mean Q max values in the P/D group. Men in the D/D group who entered the open-label phase had characteristics that were similar to those that elected not to continue with the study (Table 3). 3.2. Effects on serum dihydrotestosterone Patients in the D/D group experienced a median decrease in DHT concentration of 93.7% from baseline to Month 24, which was maintained to Month 48 (95.3%). Patients in the P/D group had a median 5.9% increase at Month 24 and a 95.4% decrease at Month 48. At this time point, 87% of patients in the D/D group and 89% of patients in the P/D group had a reduction in DHT 90%. 3.3. Effects on prostate volume, symptoms and urinary flow From baseline to Month 24, patients treated with dutasteride had a mean reduction in TPVof 26.0% vs. a mean increase of 1.4% for placebo-treated patients (p < 0.001 between treatment groups, Fig. 2). At Month 48, patients switched from placebo to dutasteride at Month 24 (P/D group) had a reduction from baseline in TPVof 21.7% that was significant vs. Month 24 (p < 0.001). TPV continued to decrease from Months 24 48 for D/D-treated patients, although this decrease was not statistically significant (p = 0.07). The overall reduction in TPV from baseline to Month 48 in D/D-treated patients was significantly greater than that observed in P/D patients ( 27.3%, p < 0.001). From baseline to Month 24, patients treated with dutasteride had a mean reduction in AUA-SI score of 4.4 points vs. 2.5 points for placebo-treated patients (p < 0.001 between treatment groups, Fig. 3a). At Month 48, patients switched from placebo to dutasteride at Month 24 (P/D group) had a reduction from baseline in AUA-SI score of 5.6 points that was significant vs. Month 24. (p < 0.001). AUA-SI score decreased significantly from Months 24 to 48 for D/D-treated patients (p < 0.001), with the overall reduction from baseline significantly greater than that

F. Debruyne et al. / European Urology 46 (2004) 488 495 491 Table 2 Baseline, 24- and 48-month parameters for the open-label ITT population Parameter Baseline Month 24 Month 48 P/D D/D P/D D/D P/D D/D Serum DHT (pcg/ml) 407.0 201.9 427.9 232.1 424.9 194.2 39.1 75.7 29.7 63.0 31.1 60.7 Serum testosterone (pcg/ml) 3959 1595 4042 1707 3990 1472 4797 1792 4718 1688 4904 2186 Total prostate volume (cc) 53.9 20.9 56.1 24.2 54.4 25.3 41.3 20.2 42.2 20.2 41.3 23.1 Transition zone volume (cc) 26.9 16.0 28.6 18.0 28.6 19.2 21.5 14.2 22.8 18.8 22.4 16.0 AUA-SI 16.9 5.9 16.6 5.7 14.5 7.2 12.3 6.7 11.3 6.4 10.2 6.1 Q max (ml/s) 10.7 3.7 10.2 3.6 11.3 4.6 12.5 5.6 12.6 5.1 12.8 5.5 Serum PSA (ng/ml) 3.9 2.1 4.1 2.2 4.3 2.8 1.9 1.9 2.0 1.8 1.7 1.8 P/D = placebo/dutasteride-treated subjects; D/D = dutasteride/dutasteride-treated subjects. DHT = dihydrotestosterone; AUA-SI = American Urological Association Symptom Index; Q max = maximal flow rate; PSA = prostate specific antigen. Data presented as means standard deviations. observed in patients from the P/D group (6.5 points, p < 0.001). Changes in symptom scores were greater at Month 24 in the completer population vs. the ITT population (Fig. 3b), with a reduction of 5.0 points in the D/D group and 2.7 points in the P/D group (p < 0.001 between treatment groups). The significant difference between the D/D and P/D groups was maintained from Month 24 to 48. Over the first 24 months of treatment with dutasteride, the proportion of patients who had severe symptoms (AUA-SI score of 20 to 35) decreased in both the P/D and D/D groups, but to a greater extent in the D/D (Fig. 4). In the D/D group, the proportion of patients who had severe symptoms decreased from 30.0% at baseline to 8.0% at Month 48. From baseline to Month 24, patients treated with dutasteride had a mean improvement in Q max of 2.2 ml/s vs. 0.6 ml/s for placebo-treated patients (p < 0.001 between treatment groups, Fig. 5). At Month 48, patients switched from placebo to dutasteride at Month 24 (P/D group) had an improvement in Q max of 1.9 ml/s that was significant vs. Month 24 (p < 0.001). Q max increased significantly from Months 24 to 48 for D/D-treated patients (p = 0.007), with the overall increase from baseline significantly greater than that observed in patients from the P/D group (2.7 ml/s, p = 0.042). 3.4. Effects on AUR and BPH-related surgery AUR occurred in a small percentage of men in the P/D group during the open-label phase at rates consistent with those seen in dutasteride-treated patients between Months 0 24 and 24 48 (Table 4). BPHrelated surgery occurred in a small percentage of men in both the P/D and D/D groups between Months 24 and 48 at slightly lower rates than seen in D/Dtreated men during Months 0 24. 3.5. Safety and tolerability The most common drug-related AEs were sexual events (impotence, decreased libido, and ejaculation disorders) and gynaecomastia (Table 5). The onset of most new drug-related sexual AEs occurred within the first 6 months of therapy. Among patients who received dutasteride throughout the 48-month study period, the incidence of most drug-related sexual AEs decreased with duration of treatment. The incidence of drugrelated gynaecomastia was low and remained constant over the treatment period. Among patients who Table 3 Comparison of AUA-SI, prostate volume and Q max for patients who entered the open-label phase, versus those who did not Placebo Dutasteride Entered open-label (n = 1152) Did not enter open-label (n = 1006) Entered open-label (n = 1188) Did not enter open-label (n = 979) AUA-SI score 14.5 7.16 * 15.5 7.56 * 12.3 6.68 12.1 6.49 Change in AUA-SI score from baseline 2.5 6.67 * 1.6 7.30 * 4.4 6.52 4.7 6.96 Prostate volume (cc) 54.4 25.31 53.1 24.65 41.3 20.19 40.8 22.24 Change in prostate volume from baseline (%) 1.4 26.16 2.8 24.74 26.0 19.38 24.7 21.03 Q max (ml/s) 11.3 4.60 * 10.9 5.48 * 12.5 5.57 12.7 5.75 Change in Q max (ml/s) 0.6 4.57 0.9 5.12 2.2 5.15 2.3 5.42 Data are derived from the Month 24 visit (double-blind ITT population) and are means standard deviations. * p < 0.05.

492 F. Debruyne et al. / European Urology 46 (2004) 488 495 Fig. 2. Mean change (%) in TPV from baseline over 48 months. Red bars = placebo/dutasteride-treated subjects (P/D); blue bars = dutasteride/dutasteride-treated subjects (D/D). Month 1 results are from ARIA3001 only; Month 3 results are from ARIA3002 only; Month 6 to 48 results are from ARIA3001, ARIA3002 and ARIB3003. *p < 0.001 for differences between treatment groups. y p < 0.001 for change from Month 24 to Month 48. z p = 0.07 for change from Month 24 to Month 48. received dutasteride in the open-label phase only, the incidence of events was similar to those experienced by D/D-treated patients at the start of therapy. The incidence of events in the P/D group also declined between Fig. 4. Proportions of subjects in the placebo/dutasteride group (P/D) and dutasteride/dutasteride group (D/D) with severe (AUA-SI 20 35) symptoms at baseline, Month 24 and Month 48. Months 36 and 48 following 1 year of dutasteride therapy. The incidence of drug-related sexual function AEs that led to withdrawal was <1% in the open-label phase. 3.6. Effects on testosterone and serum PSA Patients in the D/D group experienced a median increase in testosterone concentration of 19.7% from baseline to Month 24, which was maintained to Month 48 (21.9%). Patients in the P/D group had a median 2.2% increase at Month 24 and a 20.7% increase at Month 48. The mean and median testosterone levels for P/D- and D/D-treated patients remained within the normal range at Month 48. Serum PSA decreased in the D/D group from a mean of 4.1 ng/ml at baseline to 1.9 ng/ml at Month 24 (52.9% mean decrease) and 1.7 ng/ml at Month 48 (57.2% mean decrease). Patients in the P/D group experienced a mean increase in PSA of 15.0% from Fig. 3. Mean change in AUA-SI scores from baseline over 48 months: a, open-label ITT population; b, open-label completer population. Red line with diamonds = placebo/dutasteride-treated subjects (P/D); blue line with squares = dutasteride/dutasteride-treated subjects (D/D). *p < 0.001 between treatment groups; yp < 0.001 for differences for both treatment groups from Month 24. Fig. 5. Mean change in Q max from baseline over 48 months. Red line with diamonds = placebo/dutasteride-treated subjects (P/D); blue line with squares = dutasteride/dutasteride-treated subjects (D/D). *p < 0.001 between treatment groups; zp = 0.032 between treatment groups; yp = 0.042 between treatment groups; p < 0.001 vs. Month 24; p = 0.039 vs. Month 24; p = 0.007 vs. Month 24.

F. Debruyne et al. / European Urology 46 (2004) 488 495 493 Table 4 Crude incidence rates of subjects with AUR or BPH-related surgical intervention in the double-blind and open-label phases (double-blind ITT population) Study phase AUR BPH-related surgery P/D (n = 2158) D/D (n = 2167) P/D (n = 2158) D/D (n = 2167) Double-blind phase Month 0 6 1.3% 0.9% 0.6% 0.6% Month 6 12 1.0% 0.4% 1.3% 0.6% Month 12 18 1.1% 0.2% 1.0% 0.7% Month 18 24 0.9% 0.4% 1.4% 0.4% Open-label phase Month 24 30 0.6% 0.1% 0.3% 0.2% Month 30 36 0.6% 0.4% 0.5% 0.2% Month 36 42 0.3% 0.4% 0.1% 0.4% Month 42 48 0.6% 0.3% 0.2% 0.1% Incidence for 48 months 5.1% 2.4% 4.5% 2.6% P/D = placebo double-blind/dutasteride open-label; D/D = dutasteride double-blind/dutasteride open-label. a baseline of 3.9 ng/ml to 4.3 ng/ml at Month 24. Following initiation of dutasteride treatment at Month 24, PSA reduced to 2.0 ng/ml at Month 48: a mean decrease of 48.4% vs. baseline. 4. Discussion Treatment with the dual 5ARI dutasteride for 4 years results in sustained and consistent DHT suppression, a continued reduction in prostate volume and continued improvements in symptoms and urinary flow. The median reduction in DHT from baseline was 95.3%, with 87% of patients having a reduction 90%. The extent and consistency of these DHT reductions are in accordance with previously published data [18]. Sustained reductions in TPV occurred in men in the D/D group over 48 months. Although a reduction in TPV between Months 24 and 48 was observed in these subjects, this did not reach statistical significance. However, subjects in the D/D group had significantly greater reductions in TPV vs. those who received dutasteride for only the latter 24 months, suggesting that for 4 vs. 2 years of treatment, a longer duration of therapy increases the likelihood of achieving a maximal reduction in TPV. Long-term treatment with dutasteride resulted in continued improvements in AUA-SI score and urinary flow. In D/D-treated patients, the change in AUA-SI score from Months 24 to 48 was statistically significant. Patients previously receiving placebo during the double-blind phase had a significant improvement in symptom score after receiving 24 months of open-label dutasteride, but this improvement did not achieve the same magnitude as that of the D/D group. The magnitudes of improvement in symptom score from baseline of 4.4 points at Month 24 and 6.5 points at Month 48 in D/D-treated subjects were in excess of that defined as a clinically relevant decrease in symptoms (3.1 points) [19]. Over the 4-year period of this study therefore, a longer duration of dutasteride therapy (4 years) resulted in significantly greater reductions in TPV and improvements in symptoms and urinary flow Table 5 Onset of double-blind and open-label drug-related AEs occurring in 1% of subjects in either treatment group Events Double-blind phase Double-blind ITT population Open-label phase Open-label ITT population Month 0 12 Month 12 24 Month 24 36 Month 36 48 P/D (n = 2158) D/D (n = 2167) P/D (n = 1736) D/D (n = 1744) P/D (n = 1152) D/D (n = 1188) P/D (n = 968) D/D (n = 1041) Any drug-related event 11.7% 15.5% 3.7% 5.7% 10.5% 6.4% 2.8% 2.6% Impotence 3.0% 6.0% 1.2% 1.7% 2.8% 1.4% 0.4% 0.4% Decreased libido 1.9% 3.7% 0.3% 0.6% 2.4% 0.4% 0.2% 0.1% Ejaculation disorders 0.7% 1.8% 0.1% 0.5% 1.2% 0.3% 0.3% 0.1% Gynaecomastia * 0.5% 1.3% 0.3% 1.3% 1.3% 1.8% 0.9% 0.7% P/D = placebo double-blind/dutasteride open-label; D/D = dutasteride double-blind/dutasteride open-label. * Includes breast/nipple tenderness and breast enlargement.

494 F. Debruyne et al. / European Urology 46 (2004) 488 495 vs. a shorter duration of therapy (2 years). In the longer term, it is possible that the 2 treatment groups may eventually converge with regard to the percentage effect on these parameters. Over this treatment period, however, a longer duration of treatment had a greater effect, with no apparent plateau after 4 years of therapy. Dutasteride therapy was well tolerated over 4 years, with a general trend towards a reduction in incidence of the most common sexual AEs over time. The exception is gynaecomastia, the incidence of which remains low and relatively constant throughout the 4-year period (1.3% in Year 1, 1.3% in Year 2, 1.8% in Year 3 and 0.7% in Year 4 for dutasteride-treated patients). Patients initiated on dutasteride therapy after 2 years have a similar pattern of AEs to those starting therapy at baseline, with diminishing incidences over the 24 to 48 month period. Overall, the near maximal suppression of DHT observed with dutasteride does not result in decreased tolerability over a 4-year period. 5. Conclusions In conclusion, long-term treatment with the dual 5ARI dutasteride results in continuing improvements in urinary symptoms and flow rate, and further reductions in TPV, in men with symptomatic BPH. The reduction in risk of AUR and BPH-related surgery, seen in the double-blind phase, was durable over the 4- year term of the studies. Taken together, these findings establish that the disease-modifying benefits of dutasteride are durable in long-term treatment. Dutasteride was also well tolerated in long-term use, with no new safety issues emerging over 4 years of treatment. References [1] McConnell JD, Bruskewitz R, Walsh PC, Andriole G, Lieber MM, Holtgrewe HL, 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. N Engl J Med 1998;338:557 63. [2] Roehrborn CG, Boyle P, Nickel JC, Hoefner K, Andriole G. Efficacy and safety of a dual inhibitor of 5-alpha-reductase types 1 and 2 (dutasteride) in men with benign prostatic hyperplasia. Urology 2002;60:434 41. [3] McConnell JD, Roehrborn CG, Bautista OM, Andriole Jr GL, Dixon CM, Kusek JW, et al. The long-term effect of doxazosin, finasteride, and combination therapy on the clinical progression of benign prostatic hyperplasia. N Engl J Med 2003;349:2387 98. [4] Djavan B, Marberger M. A meta-analysis on the efficacy and tolerability of alpha1-adrenoceptor antagonists in patients with lower urinary tract symptoms suggestive of benign prostatic obstruction. Eur Urol 1999;36:1 13. [5] Jacobsen SJ, Girman CJ, Guess HA, Rhodes T, Oesterling JE, Lieber MM. Natural history of prostatism: longitudinal changes in voiding symptoms in community dwelling men. J Urol 1996;155:595 600. [6] Jacobsen SJ, Jacobson DJ, Girman CJ, Roberts RO, Rhodes T, Guess HA, et al. Natural history of prostatism: risk factors for acute urinary retention. J Urol 1997;158:481 7. [7] Jacobsen SJ, Girman CJ, Jacobson DJ, Liebert MM. Long-term (92- month) natural history of changes in lower urinary tract symptom severity. BJU Int 2000;86:248 9. [8] Jacobsen SJ, Girman CJ, Lieber M. Natural history of benign prostatic hyperplasia. Urology 2001;58:5 16. [9] Lee AJ, Russell AW, Garraway WM, Prescott RJ. Three-year follow up of a community based cohort of men with untreated benign prostatic hyperplasia. Eur Urol 1996;30:11 7. [10] Bruskewitz R, Girman CJ, Fowler J, Rigby OF, Sullivan M, Bracken RB, et al. Effect of finasteride on bother and other health related quality of life aspects associated with benign prostatic hyperplasia. Urology 1999;54:670 8. [11] Flanigan RC, Reda DJ, Wasson JH, Anderson RJ, Abdellatif M, Bruskewitz RC. 5-year outcome of surgical resection and watchful waiting for men with moderately symptomatic benign prostatic hyperplasia: a Department of Veterans Affairs cooperative study. J Urol 1998;160:12 7. [12] Meigs JB, Barry MJ, Giovannucci E, Kawachi I. Incidence rates and risk factors for acute urinary retention: the health professionals follow up study. J Urol 1999;162:376 82. [13] Hudson PB, Boake R, Trachtenberg J, Romas NA, Rosenblatt S, Narayan P, et al. Efficacy of finasteride is maintained in patients with benign prostatic hyperplasia treated for 5 years. The North American Finasteride Study Group. Urology 1999;53:690 5. [14] Vaughan D, Imperato-McGinley J, McConnell J, Matsumoto AM, Bracken B, Roy J, et al. Long-term (7 to 8-year) experience with finasteride in men with benign prostatic hyperplasia. Urology 2002;60:1040 4. [15] Lowe FC, McConnell JD, Hudson PB, Romas NA, Boake R, Lieber M, et al. Long-term 6-year experience with finasteride in patients with benign prostatic hyperplasia. Urology 2003;61:791 6. [16] Span PN, Voller MC, Smals AG, Sweep FG, Schalken JA, Feneley MR, et al. Selectivity of finasteride as an in vivo inhibitor of 5alphareductase isozyme enzymatic activity in the human prostate. J Urol 1999;161:332 7. [17] Bramson HN, Hermann D, Batchelor KW, Lee FW, James MK, Frye SV. Unique preclinical characteristics of GG745, a potent dual inhibitor of 5AR. J Pharmacol Exp Ther 1997;282:1496 502. [18] Roehrborn CG, Andriole G, Schalken JA, Wilson T, Clark RV. Dutasteride, a novel dual 5alpha-reductase inhibitor, reduces serum DHT to a greater extent than finasteride and achieves finasteride maximal reduction in a larger proportion of patients. Eur Urol Suppl 2003;2(1):161 [Abstract 635]. [19] Barry MJ, Williford WO, Chang Y, Machi M, Jones KM, Walker- Corkery E, et al. Benign prostatic hyperplasia specific health status measures in clinical research: how much change in the American Urological Association symptom index and the benign prostatic hyperplasia impact index is perceptible to patients? J Urol 1995;154:1770 4.

F. Debruyne et al. / European Urology 46 (2004) 488 495 495 Editorial Comment P. Perrin, Lyon, France paul.perrin@chu-lyon.fr The evaluation of treatment efficacy should be based on a randomised comparison in order to control for selection bias. Indeed, observational studies are not considered to be reliable enough to look for efficacy and many of them have been invalidated by subsequent randomised trials. The open-labelled extension to a randomised study is no more than an observational study for at least three reasons: 1 Only a subgroup of patients initially randomised is included in the extension, breaking the comparability of groups. 2 This subgroup analysis is not able to appraise the safety of the treatment as the patients opting to continue the drug are likely to be those who experienced no, or minimal, treatment-related side-effects. 3 Such a design is inappropriate to compare efficacy according to the length of the treatment because the initiation of the active drug is not the same in the two groups. For these reasons, the data presented in this paper cannot be considered to substantiate the conclusions that treatment over an extended time frame is superior to a delayed treatment over a shorter period. Editorial Comment B. Djavan, Wien, Austria bdjavan@hotmail.com This is the first 4-year multicenter study on dutasteride, but the open-labelled extension of the trial must be interpreted with caution as it lacks randomisation and is faulted with selection bias. However, it is important to recognise that participants for the more important treatment arm, D/D, were comparable to those who did not enter the open-label study in terms of symptom score, prostate volume and Q max after the first 24 months of active drugs. Thus, this study can be viewed as an observational study with 2 important messages: Firstly, continued usage of dutasteride after 24 months provided continuing improvements in symptom scores, Q max and reductions in TPV up to 48 months. Similar results were obtained for the intention-to-treat population as compared to the completer population. Due to the lack of placebo and the rather high drop out rate of 30%, no valid conclusions can be made on the safety of the prolonged usage of dutasteride. Finally, the results achieved in the P/D arm were similar to that obtained in Phase I of the study for men who received dutasteride. This further reinforces the efficacy of this unique 5a-reductase inhibitor in improving urinary flow and symptoms, as well as reducing of total prostate volume and PSA.