Study Designs

Four study designs have been used to investigate α-adrenergic blockers for BPH: titration to fixed dose, titration to response, titration to maximal dose, and randomized dose withdrawal.

Subjects enrolled in titration to fixed dose studies receive one of several predetermined final doses independent of clinical response unless significant adverse effects are encountered. An advantage of this study design is that dose-dependent efficacy and safety of different doses are determined. A disadvantage is the requirement for a large sample size to identify statistically significant differences between placebo and all of the treatment groups.

Titration to response design allows the investigators to titrate the dose to a threshold response or maximal dose. An advantage of this design is a smaller sample size because all subjects receiving active treatment are analyzed as a composite group independent of final dose. A disadvantage of this design is that the maximal therapeutic effect may be underestimated if the titration is not to maximal response. The data are also misleading if expressed in terms of group mean changes according to final dose because all nonresponders are titrated to the maximal dose in the absence of toxicity.

A randomized dose withdrawal design begins with an open-label dose titration. All responders are randomized to active drug or placebo. An advantage of this design is the enrichment of responders. A disadvantage is that the results are not generalizable to untreated patients.

A titration to maximal dose design, like titration to response, requires a relatively small sample size because there is only one active treatment group. This study design defines maximal clinical response achievable in practice, providing the maximal dose is also the most efficacious tolerable dose.

Dose Response

Multicenter, randomized, placebo-controlled studies have consistently shown that symptom and flow improvement is dependent on the dose of the α₁ blockers. The differences between the effectiveness of different doses were often not statistically significant because these dose-ranging studies were not adequately powered to show significant differences between dose groups. MacDiarmid and coworkers (1999) have provided the most compelling evidence for a positive correlation relationship between dose and effectiveness of α₁ blockers in the treatment of BPH. Responders to 4 mg of doxazosin were randomized in a double-blind manner to receive 4 mg or 8 mg of doxazosin. The improvement observed in the 8-mg group was 3.7 symptom units greater than in the 4-mg group (P = .03). In phase 3 trials, the impact of dose observed in the responders is diluted by the lack of effect in the nonresponders. In clinical practice, nonresponders are withdrawn from treatment.

Terazosin

Randomized, double-blind, multicenter, placebo-controlled studies have consistently demonstrated the efficacy and safety of terazosin for BPH (Di Silverio, 1992; Lepor et al, 1992; Lloyd et al, 1992; Brawer et al, 1993; Elhilali et al, 1996; Lepor et al, 1996; Roehrborn et al, 1996) (Table 92–2). The multicenter, double-blind, parallel-group, randomized, placebo-controlled study of once-a-day administration of terazosin to patients with symptomatic BPH reported by Lepor and associates (1992) is representative of the expectations of terazosin therapy. Two hundred eighty-five patients entered the double-blind treatment receiving either placebo or 2, 5, or 10 mg of terazosin once daily. Statistically significant decreases from baseline obstructive, irritative, and total symptom scores were observed for all terazosin treatment groups. The level of improvements in the symptom scores were dose dependent. The 10-mg terazosin treatment group exhibited significantly greater decreases in mean irritative and total symptom scores relative to the placebo group. The 5- and 10-mg terazosin treatment groups exhibited a significantly greater mean decrease in obstructive scores relative to the placebo group. The percentages of patients experiencing a greater than 30% improvement in the total symptom scores for the placebo, 2-, 5-, and 10-mg treatment groups were 40%, 51%, 57%, and 69%, respectively (Fig. 92–5). The percentage of patients experiencing greater than 30% improvement in total symptom score in the 10-mg treatment groups was significantly greater than that of the placebo group.

Table 92–2 Efficacy of Terazosin in Benign Prostatic Hyperplasia

Figure 92–5 Two hundred eighty-five patients were enrolled in a randomized double-blind study comparing placebo (Plb) and 2 mg, 5 mg, and 10 mg terazosin administered once daily. Percentages of patients experiencing more than 30% improvement in total symptom scores and peak urinary flow rates are shown.

(From Lepor H. Medical therapy for benign prostatic hyperplasia. Urology 1993;42:483–501.)

A statistically significant improvement from baseline was seen in the peak and mean urinary flow rates for all the treatment groups. The effect of terazosin on PFR was also dose dependent. The 10-mg treatment group exhibited a significantly greater increase from baseline in peak and mean urinary flow rates relative to the placebo group. The percentages of patients experiencing a greater than 30% increase in PFR in the placebo, 2-, 5-, and 10-mg treatment groups were 26%, 40%, 35%, and 52%, respectively. A significantly greater proportion of patients in the 10-mg terazosin treatment group exhibited a greater than 30% improvement in PFR compared with the placebo group. Overall, the adverse events in the four treatment groups were minor and reversible. Although a higher incidence of asthenia, flu syndrome, and dizziness were observed in the terazosin treatment groups, the differences from placebo were not statistically significant. There was a significantly greater incidence of postural hypotension in the 5-mg terazosin group than in the placebo group. The incidence of syncope for all terazosin-treated patients was less than 0.5%.

The relationships between percentage change in total symptom score and PFR versus baseline age, prostate size, PFR, PVR, and total symptom score were examined to identify clinical or urodynamic factors that predicted response to terazosin therapy. No significant association was observed between treatment effect and any of these baseline factors.

There is legitimate concern whether the results of multicenter studies conducted primarily at tertiary medical centers is generalizable to community practice. Roehrborn and coworkers (1996) reported the results of the Hytrin Community Assessment Trial (HYCAT), which enrolled 2084 men into a 1-year randomized double-blind study comparing the safety and effectiveness of terazosin versus placebo. The overwhelming majority of the subjects were enrolled by urologists in community practice. The daily dose of terazosin was titrated up to 10 mg based on the discretion of the principal investigators. The symptom scores in the placebo and terazosin group throughout the study are shown in (Fig. 92–6). The treatment-related improvement (terazosin minus placebo) in the AUASI score and in urinary PFR was 1.4 mL/sec and 3.9 symptom units, respectively. The treatment-related incidences of dizziness, asthenia, and peripheral edema were 5.9%, 4.6%, and 3.1%, respectively.

Figure 92–6 Time course of improvements in the American Urological Association (AUA) Symptom Index (A), the AUA Bothersome Score (AUA BS) (B), BPH Impact Index (QOL BII) (C), and Quality of Life question (QOL Q4) (D) for placebo and terazosin treatment groups. A lowering of the score represents clinical improvement. Weeks 1 to 2 and 2 to 5 are the mandatory titration periods for 1 and 2 mg of terazosin (and placebo), respectively.

(A to D, redrawn from Roehrborn CG. Effectiveness and safety of terazosin versus placebo in the treatment of men with symptomatic benign prostatic hyperplasia in the HYCAT study. Urology 1996;47:159–68.)

Lepor (1995) presented an interim report of 494 patients entered into an open-label extension study demonstrating the durable clinical response of terazosin. Durations of follow-up ranged from 3 to 42 months. The percentage of patients on final terazosin doses of 1, 2, 5, 10, and 20 mg were 7%, 12%, 26%, 34%, and 21%, respectively. Of the 494 patients, 213 (43.1%) withdrew prematurely: 55 (11%) because of therapeutic failure, 96 (19%) because of adverse events, and 62 (13%) because of administrative reasons.

At all follow-up visits the group mean PFRs were significantly higher than baseline values (Fig. 92–7). At baseline, PFR was 10.0 mL/sec. From 3 to 42 months, improvement ranged from 2.3 to 4.0 mL/sec. Between months 3 and 42, at least a 30% improvement in PFR from baseline was observed in 40% to 59% of the patients. At all follow-up intervals, the group mean Boyarsky symptom scores were significantly lower than at baseline; this was true of obstructive, irritative, and total scores (Fig. 92–8). From 3 months onward, improvement ranged from 4.0 to 5.4 points. Between months 3 and 42, at least a 30% improvement in total symptom score from baseline was observed in 62.4% to 77.1% of the patients.

Figure 92–7 Mean change in peak flow rate between baseline and 42 months in terazosin-treated patients. The numbers across the top of the graph indicate the number of patients available at each time interval. All data points were significantly different from baseline at the P ≤ .05 level.

(From Lepor H and the Multicenter Study Group. Long-term efficacy and safety of terazosin in patients with benign prostatic hyperplasia. Urology 1995;45:406–13.)

Figure 92–8 Mean change in Boyarsky symptom scores from baseline to 42 months. Baseline scores were 10.5 for total score, 6.2 for obstructive scores, and 4.3 for irritative scores. The numbers across the top of the graph indicate number of patients available at each time interval. All changes were significant at the P ≤ .05 level.

(From Lepor H and the Multicenter Study Group. Long-term efficacy and safety of terazosin in patients with benign prostatic hyperplasia. Urology 1995;45:406–13.)

Kirby (1998b) has examined the mean changes in blood pressure according to whether subjects were normotensive or hypertensive at baseline (Table 92–3). In normotensive patients, small, clinically insignificant decreases in blood pressure were generally noted. Untreated hypertensive men had larger and clinically significant decreases in blood pressure. In men with medically controlled hypertension, terazosin had no clinically significant effect on blood pressure, whereas in men with poorly controlled medically treated hypertension, terazosin significantly lowered blood pressure. In all clinical circumstances, terazosin’s effect on blood pressure was consistently physiologically desirable. The ability to treat two common coexisting conditions (BPH and hypertension) is a potentially desirable feature of the drug.

Table 92–3 Effect of Terazosin on Blood Pressure in Normotensive and Hypertensive Men

Doxazosin

The half-life of doxazosin is longer than that of terazosin (22 vs. 12 hours). The efficacy, safety, and durability of clinical response of doxazosin has been demonstrated in multicenter, randomized, double-blind, placebo-controlled studies (Chapple et al, 1994; Fawzy et al, 1995; Gillenwater et al, 1995) (Table 92–4) and a long-term open-label extension study (Lepor, 1995). Fawzy and associates (1995) reported a 16-week multicenter, randomized, double-blind, placebo-controlled titration to response study in 100 normotensive subjects with BPH. Of the 41 evaluable subjects receiving doxazosin, 88% underwent titration to the maximal dose (8 mg). The group mean changes in PFR and symptom score were significantly greater in the doxazosin group compared with the placebo group (see Table 92–4). The magnitudes of these treatment-related effects are similar to those of terazosin. The systolic blood pressure changes in normotensive subjects are greater than those with terazosin. The treatment-related incidences of dizziness, fatigue, headache, somnolence, hypotension, and nausea were 20%, 8%, 8%, 6%, 8%, and 8%, respectively. The percentages of subjects withdrawing because of an adverse event were 14% and 2.1% in the doxazosin and placebo groups, respectively. The treatment-related incidence of adverse clinical events in this doxazosin study appears slightly higher than that of terazosin and may be a result of its greater effect on blood pressure.

Table 92–4 Efficacy of Doxazosin in Benign Prostatic Hyperplasia

Gillenwater and coworkers (1995) reported a multicenter, randomized, double-blind, placebo-controlled titration to fixed dose study comparing placebo versus 2, 4, 8, and 12 mg of doxazosin in 248 men with mild to moderate essential hypertension. The group mean changes in PFR and Boyarsky symptom score are summarized in Table 92–4 according to treatment groups. Because relatively small numbers of subjects were randomized into the individual treatment groups the failure to demonstrate statistical significance between placebo and some of the active treatment groups reflects the small sample size. The group mean improvement in PFR was dose dependent and statistically significant relative to placebo for all active treatment groups. The mean improvements in symptom scores relative to placebo for the group were statistically significant for the 4- and 8-mg doxazosin groups. Statistically and clinically significant changes in systolic blood pressure were observed between the placebo and the 4-, 8-, and 12-mg doxazosin groups. Lowering of blood pressure was a desirable outcome in these hypertensive patients. The overall treatment-related incidence of dizziness and fatigue was 15% and 10%, respectively. The percentages of subjects withdrawing because of an adverse event in the doxazosin versus placebo groups were 11.1% and 4.1%, respectively. Statistically significant changes in symptom scores and PFR relative to baseline have been reported in a long-term open-label doxazosin extension study (Lepor, 1995). The initial improvements in symptom scores and PFR in 450 subjects were maintained for up to 42 months. Kirby (1995) summarized the effects of doxazosin on blood pressure in normotensive and hypertensive men enrolled into two double-blind, placebo-controlled trials (Fawzy et al, 1995; Gillenwater et al, 1995). The treatment-related group mean reductions in sitting systolic blood pressure in the normotensive and hypertensive subjects were 3 and 17 mm Hg, respectively. The treatment-related group mean reductions in sitting diastolic blood pressure in the normotensive and hypertensive subjects were 4 and 3 mm Hg, respectively.

With the standard preparation of doxazosin, multiple titration steps are used to obtain optimal therapeutic response. Often doxazosin standard is started at 1 mg/day and titrated through 2 and 4 mg/day to 8 mg/day to obtain the optimal response. The controlled release gastrointestinal therapeutic system (GITS) formulation of doxazosin reduces the plasma peak-to-trough ratio to minimize the need for titration. To compare the two formulations in 795 men with BPH, doxazosin standard was initiated at 1 mg/day, titrated to 2 mg/day after 1 week to 4 mg/day at 3 weeks and 8 mg/day at 7 weeks if indicated. This regimen was compared with doxazosin GITS initiated at 4 mg once daily and titrated to 8 mg once daily after 7 weeks if indicated and to a placebo group over 13 weeks. The symptoms of BPH were measured with the IPSS, which has seven questions (covering frequency, nocturia, weak urinary stream, hesitancy, intermittence, incomplete emptying, and urgency) scored 0 (absent) to 5 (severe). On the IPSS there was an improvement of −8.4 and −8.0 with doxazosin standard and GITS, respectively, compared with −6.0 in the placebo group. Doxazosin standard and GITS produced clinically comparable increases in mean PFR, compared with placebo, with a greater improvement observed earlier following treatment with doxazosin GITS than with doxazosin standard. A similar number of patients in both doxazosin groups was titrated to the maximum dose of 8 mg for both formulations. The incidence of adverse effects was slightly higher with doxazosin standard than doxazosin GITS and placebo, which caused a similar incidence.

The previous study and another with 680 men were combined to further analyze the comparison between doxazosin standard and GITS (Kirby et al, 2003). In addition to confirming the results given earlier, in a subgroup reporting sexual dysfunction at baseline there was a modest clinical improvement in sexual function with both preparations of doxazosin.. Treatment-related adverse events occurred in 16.1% of patients on doxazosin GITS, 25.3% of patients on doxazosin standard, and 7.7% of placebo patients. Headache and dizziness occurred in 6.0% and 5.3% of doxazosin GITS patients, compared with 5.1% and 9.1% of doxazosin standard patients, respectively (placebo, 4.5% and 1.9%, respectively). Fewer patients on doxazosin GITS (5.7%) or placebo (2.6%) discontinued treatment because of adverse events than on doxazosin standard (7.2%). The reduction in blood pressure was not clinically significant in the normotensive patients, but there were clinically significant reductions in blood pressure with both preparations of doxazosin (placebo, 3.9/5.0; doxazosin standard, 7.4/6.1; doxazosin GITS, 9.4/6.8). A comparison of the nonconcurrent multicenter, randomized, double-blind, placebo-controlled studies of terazosin (see Table 92–2) and doxazosin (see Table 92–4) shows similar efficacy. Studies of doxazosin versus tamsulosin (Kirby et al, 2003) and doxazosin versus alfuzosin (de Reijke and Klarskov, 2004) revealed only minor differences in safety and efficacy.

α-Adrenergic blockers such as doxazosin may influence smooth muscle growth in the prostate. In BPH patients treated with α₁-adrenoceptor antagonists there is a decreased expression of myosin heavy chain messenger RNA, a functional marker for the smooth muscle phenotype (Lin et al, 2001).

Biopsy and prostatectomy specimens from untreated and doxazosin-treated BPH patients suggest that doxazosin may induce apoptosis in both the epithelial and stromal cells with little effect on cell proliferation. The apoptosis was associated with a decrease in smooth muscle α-actin expression and stromal regression. Another study showed that the mean pretreatment baseline apoptosis was 1.9% and 1.0% for the epithelial and stromal prostate components. The mean apoptotic indexes increased after 3 months of doxazosin treatment for BPH to 6% in the glandular epithelial and 12% in the smooth muscle cells. By 12 months after treatment, epithelial apoptosis had decreased to constitutive levels while the apoptotic index of prostatic stromal cells remained high (Kyprianou et al, 1998).

In primary cultures of human prostate stroma cells, doxazosin increased apoptosis and decreased cell numbers. Transforming growth factor (TGF)-β₁ also decreased cell numbers, and because doxazosin increased the levels of TGF-β₁ in the cells it was suggested that the effect of doxazosin may be mediated through TGF-β₁ (Ilio et al, 2001).

The ability of doxazosin to induce apoptosis may be shared with the other quinazoline-based α₁-adrenergic receptor antagonists terazosin and prazosin, although it seems unlikely that this effect is α₁-adrenergic receptor mediated (Gonzalez-Juanatey et al, 2003). The apoptotic effects of the quinazoline-based α₁ antagonists may be linked to their ability to inhibit HERG potassium channels, which has been demonstrated using cloned channels expressed in Xenopus oocytes (Thomas et al, 2004).

Recent data also suggest that doxazosin may have a mildly beneficial impact on sexual function in men suffering from BPH (Kirby et al, 2005). The mechanism for this effect is uncertain but may be the result of a vasodilatory action within the corpora cavernosa.

Tamsulosin

Tamsulosin is currently the most widely employed α₁ antagonist investigated for BPH (Foglar et al, 1995). One of the features of tamsulosin is that it exhibits some degree of specificity for the α_1A-adrenergic receptor (Foglar et al, 1995). The efficacy and safety of tamsulosin has been investigated in four multicenter, randomized, double-blind, placebo-controlled studies (Kawabe et al, 1990; Abrams et al, 1995, Lepor et al, 1998; Narayan and Tewari, 1998) (Table 92–5).

Table 92–5 Efficacy of Tamsulosin in Benign Prostatic Hyperplasia

Lepor and coworkers (1998) reported a multicenter, randomized, double-blind, placebo-controlled study of 756 American men with clinical BPH randomized to receive placebo or 0.4 or 0.8 mg of tamsulosin for 13 weeks. The mean changes in AUASI score, PFR, and adverse events are summarized in Table 92–5. The symptom score improvements were significantly greater in the 0.8-mg tamsulosin group compared with the 0.4-mg group. The treatment-related incidences of dizziness, asthenia, rhinitis, and abnormal ejaculation in the 4-mg group were 5%, 3%, 3%, and 6%, respectively, and in the 0.8-mg group they were 6%, 3%, 9%, and 18%, respectively. The mean changes in systolic and diastolic blood pressure in the placebo and tamsulosin groups were not significantly different for both hypertensive and normotensive subjects. In the subjects who were hypertensive and those whose blood pressure was uncontrolled, the systolic blood pressure changes in the placebo, 0.4-mg, and 0.8-mg groups were −8.4, −7.2, and −10.2 mm Hg, respectively. The advantage of not lowering blood pressure in men who are hypertensive at baseline is controversial.

Of the 618 subjects who completed the 13-week randomized study reported by Lepor and coworkers (1998), 418 (68%) continued into the 40-week extension study on the same double-blind medication and dose. The symptom and flow rate improvements observed at the end of the 13-week study were maintained throughout the 40-week extension study.

Narayan and associates (1998) reported the results of a randomized, double-blind, placebo-controlled trial comparing the safety and effectiveness of 0.4 and 0.8 mg of tamsulosin versus placebo. Seven hundred thirty-five men were randomized in the study. The active treatment was 13 weeks. The treatment-related improvements in the AUASI score and PFR were comparable with those reported by Lepor and coworkers (1998). The differences between 0.4 mg and 0.8 mg were not statistically significant; however, the study lacked statistical power to show clinically significant differences between the active treatment groups. The treatment-related incidences of asthenia, dizziness, rhinitis, and abnormal ejaculation observed for the 0.4-mg tamsulosin group were 2%, 5%, 3%, and 11%, respectively; and for the 0.8-mg tamsulosin group they were 3%, 8%, 9%, and 18%, respectively. The incidences of retrograde ejaculation and rhinitis were significantly greater in the 0.8-mg group compared with the 0.4-mg group. No statistically or clinically significant differences were observed for systolic blood pressure between any of the treatment groups.

A systematic review of tamsulosin therapy for BPH has been published (Wilt et al, 2002b). This included 14 studies with a total of 4122 patients. The mean change in symptom score was 12% for the 0.4-mg dosage and 16% with the 0.8-mg dosage. Improvements in flow rate were 1.1 mL/sec for both dosages. Adverse events were generally mild and included dizziness, rhinitis, and abnormal ejaculation. These increased in a dose-dependent manner with discontinuations due to such effects similar to placebo at 0.2 mg but increasing to 16% with the 0.8-mg/day dosage.

Alfuzosin

Jardin and colleagues (1991) reported the first large-scale, multicenter, randomized, placebo-controlled trial demonstrating that alfuzosin was safe and effective for the treatment of BPH (Table 92–6). A long-term open-label extension study showed that the effectiveness of alfuzosin was durable up to 30 months (Jardin et al, 1994). The primary limitation of alfuzosin was a requirement for multiple daily doses (2.5 mg three times a day or 5 mg twice a day). In the absence of any demonstrable advantage over the once-a-day drugs like terazosin, doxazosin, and tamsulosin there was no compelling reason to prescribe alfuzosin.

Table 92–6 Efficacy of Alfuzosin in Benign Prostatic Hyperplasia

Extended-release or slow-release (SR) alfuzosin is a new formulation that allows for a once-daily dosing regimen without dose titration. Buzelin and coworkers (1997b) reported the first randomized, multicenter, placebo-controlled trial evaluating the safety and effectiveness of SR alfuzosin for the treatment of BPH. Three hundred and ninety subjects were randomized to once-daily 5 mg alfuzosin versus placebo for 12 weeks. The treatment-related improvements in the IPSS and PFR were −1.6 symptom unit and 1.3 mL/sec, respectively. The incidence of dropouts because of adverse events was 4.6% and 7.1% in the SR alfuzosin and placebo groups, respectively. The 2-mm Hg change in systolic and diastolic blood pressure was not significantly different from that in the placebo group. The incidences of dizziness and asthenia were similar in the SR alfuzosin and placebo groups.

SR alfuzosin (10 mg once a day) has been compared with IR alfuzosin (2.5 mg three times daily) and placebo (van Kerrebroeck et al, 2000). After a 1-month placebo lead-in, 447 patients were randomly assigned in equal proportions to the three treatment groups for 3 months. The improvement in the IPSS was 6.9, 6.4, and 4.9 in the alfuzosin 10-mg/day, alfuzosin 2.5-mg three times a day, and placebo groups, respectively. The symptom improvement observed in both active treatment groups was significantly greater than that in the placebo group. The improvements in the filling and voiding subscores and quality of life index were also significantly greater in the active treatment group relative to the placebo group. The improvement in the PFR was 2.3 mL/sec, 3.2 mL/sec, and 1.4 mL/sec in the SR alfuzosin, IR alfuzosin, and placebo groups, respectively. The modest improvements in the PFR were significantly greater in both active treatment groups compared with placebo. The incidences of dizziness were 2.1%, 4.7%, and 1.3%; and those for asthenia were 3.5%, 0.7%, and 2.6% in the SR alfuzosin, IR alfuzosin, and placebo groups, respectively. No sexual dysfunction was reported in the 10-mg/day alfuzosin group. There were no statistically or clinically significant treatment-related effects on blood pressure in normotensive or hypertensive subjects. Of those men who were hypertensive at baseline, the mean reductions in the standing blood pressure were 8.1, 8.6, and 5.8 mm Hg, respectively, in the SR alfuzosin, IR alfuzosin, and placebo groups.

Alfuzosin has been shown to have a beneficial effect on the quality of life of men suffering from BPH. In the ALFUS study (Roehrborn et al, 2001) the quality of life index improved by 18% in both active groups compared with 8% in the placebo study arm (P = .002). In an open extension of the ALFORTI trial there was a 35% improvement from baseline of quality of life (van Kerrebroeck et al, 2002). Some part of this effect may be the result of a mildly beneficial impact on sexual function (van Moorselaar et al, 2005).

Because of the lack of adverse effects and blood pressure changes, alfuzosin has been described as a uroselective drug (Kirby, 1998a). SR alfuzosin exhibits no pharmacologic uroselectivity for any of the α₁ subtypes (Andersson et al, 1997). In-vivo studies in the conscious rat have shown that alfuzosin reduces urethral pressure without significantly altering blood pressure (Martin et al, 1995). This experimental observation does not prove clinical uroselectivity because terazosin and doxazosin do not alter blood pressure in normotensive subjects. Another explanation for the lack of adverse events has been the low penetration of alfuzosin into the brain (Rouquier et al, 1994). It is also important to consider that the better tolerance may simply be related to a lower level of α₁ blockade because the treatment-related improvement of the 10 mg of alfuzosin appears to be less than that achieved with 10 mg of terazosin and 8 mg of doxazosin.

The long-term effectiveness of IR alfuzosin 2.5 mg three times a day is supported by an open-label prospective 3-year trial involving 3228 men with clinical BPH (Lukacs et al, 2000). The improvements in symptom score in BPH-specific health-related quality of life index observed at the 3-month visit were maintained throughout the 36 months of follow-up. A total of 20.1% of the men withdrew from the study. Only 4.2% of the men discontinued therapy because of an adverse event. The other reasons for withdrawal were death, 7.6%; loss of follow-up, 1.7%; lack of efficacy, 1.8%; study withdrawal owing to personal reasons, 0.8%; concomitant disease, 0.7%; and other reasons, 3.3%. Only 0.3% of men experienced AUR.

Silodosin

In Japan, silodosin, a selective α₁-adrenergic receptor antagonist, is the BPH market leader and received U.S. Food and Drug Administration (FDA) approval in October 2008 (Watson Pharmaceuticals, Inc). In two phase 3 studies, 8 mg once-daily silodosin taken for 12 weeks resulted in significant and rapid improvement of the IPSS compared with placebo. Silodosin was shown to increase urine flow in 2 to 6 hours after the initial dose. Improvement of symptoms was realized in 3 to 4 days, with the majority of patients, including men on concomitant cardiovascular medications, achieving at least a 3-point improvement in IPSS score, regardless of age or severity of symptoms. It was associated with a low incidence of orthostatic hypotension and syncope, fainting, and dizziness. As the newest agent available silodosin has had to satisfy more rigorous cardiovascular requirements by the FDA than other, older agents. Silodosin demonstrated minimal effects on the cardiovascular system, without any meaningful prolongation of the QT interval. The most common drug-related side effect was retrograde ejaculation (better described as anejaculation). Rates of discontinuing therapy due to retrograde ejaculation were low. The second most commonly reported adverse event was dizziness, the rate of which being only slightly higher than in placebo-treated patients. There were no reported events of symptomatic orthostatic hypotension or dizziness when administered with a single dose of medications for erectile dysfunction (ED) in healthy male subjects.

A randomized, double-blind, placebo-controlled study comparing silodosin to tamsulosin or placebo study was conducted at 88 centers in Japan (Kawabe et al, 2006). Men aged 50 years or older with an IPSS of 8 or higher, a quality of life score of 3 or more, a PFR of less than 15 mL/sec, a prostate volume of 20 mL or more, and a PVR urine volume of less than 100 mL were eligible for enrollment. Patients were randomized to receive silodosin 4 mg twice daily, tamsulosin 0.2 mg once daily (i.e., a lower dose than used in the United States or Europe), or placebo, for 12 weeks: 457 men were randomized (silodosin, 176; tamsulosin, 192; and placebo, 89). The change in the total IPSS from baseline in the silodosin, tamsulosin, and placebo groups was −8.3, −6.8, and −5.3, respectively. There was a significant decrease in the IPSS versus placebo in the silodosin group from 1 week. In the early-stage comparison, silodosin showed a significant decrease in IPSS versus tamsulosin at 2 weeks. The change in quality of life score from baseline was −1.7, −1.4, and −1.1 in the silodosin, tamsulosin, and placebo groups, respectively. Silodosin showed a significant improvement in quality of life score versus placebo. In the subgroup of patients with severe symptoms (IPSS ≥ 20) silodosin also gave a significantly better improvement than placebo (−12.4 vs. −8.7). The incidence rates of adverse events and drug-related adverse events were, respectively, 88.6%, 82.3%, and 71.6% and 69.7%, 47.4%, and 36.4%, respectively. The most common adverse event in the silodosin group was abnormal ejaculation, which occurred more often in the silodosin than in the tamsulosin group (22.3% vs. 1.6%). However, only 5 men (2.9%) discontinued treatment for abnormal ejaculation. It is unclear how men in other populations will respond to this effect. Measurement by Noguchi and colleagues (2008) of intraurethral pressure in the prostatic urethra and intraluminal pressure in the vas deferens in anesthetized male dogs given the α₁-adrenergic receptor agonist phenylephrine and then a series of α₁-adrenergic receptor antagonists showed that silodosin had the highest selectivity for the vas deferens (7.5-fold), followed by naftopidil (4.3-fold), alfuzosin (3.8-fold), tamsulosin (2.6-fold), and prazosin (2.5-fold). These results suggest that high tissue selectivity for the vas deferens over the urethra may contribute to the incidence of abnormal ejaculation. These results (and U.S. data presented at the AUA meeting in 2009) do, however, suggest that silodosin is clinically useful for treating LUTS associated with BPH.

Naftopidil

Naftopidil, a relative selective α_1D-adrenergic receptor antagonist, seems more effective for patients showing a dominant expression of the α_1D-adrenergic receptor subtype. The α₁ adrenergic receptor antagonists tamsulosin and naftopidil were administered to 96 patients with BPH for 8 weeks in a crossover study. With the administration of both drugs, the IPSS significantly decreased and the maximum urinary flow significantly increased. Whereas naftopidil monotherapy decreased the IPSS for storage symptoms, tamsulosin monotherapy decreased the IPSS for voiding symptoms (Ikemoto et al, 2003). It remains to be seen how clinically important this new agent will turn out to be.

Finasteride

Finasteride is a competitive inhibitor of the enzyme 5α-reductase (Vermeulen et al, 1989). Finasteride lowers serum and intraprostatic DHT levels. At least two isozymes (types 1 and 2) of 5α-reductase exist (Jenkins et al, 1992). Finasteride is a selective inhibitor of the type 2 isozyme. Finasteride does not reduce DHT levels to castrate levels because circulating testosterone is converted to DHT by type 1 isozymes that exist in skin and liver (Thigpen et al, 1993). Gormley and coworkers (1992) reported the first multicenter, randomized, double-blind, placebo-controlled trial investigating the safety and efficacy of finasteride in 895 men with BPH. The subjects were randomized to receive placebo or 1 or 5 mg of finasteride for 1 year. This study is often referred to as the North American Finasteride Trial.

The mean baseline prostate volumes in the placebo and 1- and 5-mg finasteride groups were 61, 61, and 59 cm³, respectively. The primary outcome measures were group mean changes in a modified Boyarsky symptom score (maximum score, 36) and PFR (Figs. 92-9 and 92-10). The group mean changes in symptom score, PFR, and prostate volume are shown in Table 92–8. The group mean percentage changes in symptom score at 12 months in the placebo and 1- and 5-mg finasteride groups were −2%, 9%, and 21%, respectively. The mean percentage changes in PFR were 8%, 23%, and 22% in the placebo and 1- and 5-mg finasteride groups, respectively. The mean percentage changes in prostate volume were −3%, −18%, and −19% in the placebo and 1- and 5-mg finasteride groups, respectively. The difference between the mean changes in PFR and prostate volume was statistically significant for both the 1- and the 5-mg finasteride groups versus the placebo group. The difference between the mean changes in symptom scores was statistically significant only for placebo versus 5 mg of finasteride. The dose-dependent symptom response was not associated with a dose-dependent prostate volume or PFR response. The changes in prostate volume were not directly related to the magnitude of the clinical response to finasteride. These observations suggest that the efficacy of finasteride may not be mediated exclusively by reduction of prostate volume. The incidences of decreased libido, ejaculatory disorder, and impotence were significantly greater in the finasteride groups compared with the placebo group. The treatment-related incidences of decreased libido, ejaculatory disorder, and impotence in the 1- and 5-mg treatment groups were 4.7%, 2.7%, and 3.7%, respectively, and 3.4%, 2.7%, and 1.7%, respectively. The percentage of subjects withdrawing because of an adverse clinical event was equivalent in the three treatment groups. Prostate volume regression was maximal at 6 months. The greatest change in symptom scores and PFR occurred within the first 2 months of initiating active treatment.

Figure 92–9 Mean (±SE) change in the total symptom score in men with benign prostatic hyperplasia during treatment with placebo (circles), 1 mg of finasteride (triangles), or 5 mg of finasteride (squares). The asterisks (P ≤ .05), the pound symbols (P < .01), and the dagger (P < .001) indicate significant differences between the finasteride-treated groups and the placebo group. Month 0 represents the baseline.

(From Gormley GJ, Stoner E, Brusekewitz RC, et al. The effect of finasteride in men with benign prostatic hyperplasia. N Engl J Med 1992;327:1185–91.

Figure 92–10 Mean (±SE) maximal urinary flow rates in men with benign prostatic hyperplasia during treatment with placebo (circles), 1 mg of finasteride (triangles), or 5 mg of finasteride (squares). The blue area indicates the range in which urinary flow was considered to be obstructed. Month 0 represents the baseline. Values before month 0 were obtained during the 2-week placebo run-in period. The asterisks (P < .05), pound symbols (P < .001), and daggers (P < .01) indicate significant differences between the finasteride-treated groups and the placebo group.

(From Gormley GJ, Stoner E, Brusekewitz RC, et al. The effect of finasteride in men with benign prostatic hyperplasia. N Engl J Med 1992;327:1185–91.

Table 92–8 Efficacy of Finasteride for Benign Prostatic Hyperplasia

The Finasteride Study Group (1993) reported another multicenter, randomized, double-blind, placebo-controlled clinical trial that is referred to as the International Finasteride Study. The effect of finasteride on symptom scores, PFR, and prostate volume is in agreement with the findings of the North American Finasteride Trial. Andersen and associates (1995) reported the results of a multicenter, randomized, double-blind, placebo-controlled study investigating the safety and efficiency of finasteride in 707 Scandinavian subjects maintained on randomized treatment for 2 years. The mean baseline prostate volumes in the placebo and finasteride groups were 41.7 and 40.6 cm³, respectively. A selection bias for enrolling large prostates did not exist in this study. A modified Boyarsky symptom score was used to capture changes in symptom score. The differences in the group mean changes in the symptom score and PFR between finasteride and placebo after 12 and 24 months on active treatment are summarized in Table 92–8. The difference between the group mean changes in symptom scores and PFR after 1 year of randomized treatment is slightly less than that in the North American Finasteride trial? This may be attributed to the smaller baseline prostate volumes.

Although the proportion of subjects experiencing any adverse clinical event and the number of withdrawals from adverse clinical events were similar to those in the finasteride and placebo groups there were more patients with sexual dysfunction in the finasteride versus placebo groups (19% vs. 10%). The time-dependent symptom score changes demonstrate that the placebo response returns to baseline between 1 to 2 years whereas the finasteride response remains stable. The authors interpret this to show that finasteride halts or alters the natural history of the disease. The mean differences between the symptom scores at 12 and 24 months in the finasteride and placebo groups were only −0.3 and 0.6 symptom unit, respectively.

Marberger and coworkers (1998) reported the results of a 2-year randomized, placebo-controlled trial of 3270 men receiving finasteride versus placebo that are comparable with those from the report of Andersen and associates (1995). The mean baseline prostate volumes in the placebo and finasteride groups were 39.2 and 38.7 cm³, respectively. The baseline prostate volume is the lowest of all the finasteride trials. The treatment-related improvements in the quasi-AUASI at 1 and 2 years were 1.0 and 1.7 symptom units, respectively. The incidences of AUR were 1.0% and 2.5% in the finasteride and placebo groups, respectively.

Stoner and associates (1994) reported on the safety and efficacy of 3 years of therapy with finasteride. Subjects participating in the North American Finasteride Study and International Finasteride Study were offered the opportunity to participate in an open-label extension study after completing 1 year of randomized therapy. The long-term (3 years) efficacy and safety analysis was limited to the 543 subjects randomized to 5 mg. Of the 543 subjects, 297 (55%) completed 3 years of treatment and were evaluable. Of the 246 unevaluable patients, 178 were dropouts and 68 were placed in a category indicating insufficient data. Inspection of the group mean changes in outcome measures reveals that the most precipitous changes in symptom scores, PFR, and prostate volume occurred between the 12- and the 18-month assessment, which coincides with transferring from blinded to unblinded treatment. After 18 months the time-dependent changes were stable, suggesting durability of response. A subsequent report of the open-label extension study demonstrated the durability of finasteride effective up to 5 years (Hudson et al, 1999).

Boyle and coworkers (1996) reported a meta-analysis of six randomized, placebo-controlled clinical trials with finasteride. The mean group changes in symptom scores and PFR correlated with the mean baseline prostate volume. This observation accounts for the variability of treatment effect observed in the different studies.

The Proscar Long-Term Efficacy and Safety Study (PLESS) represented at the time the longest-duration multicenter, randomized, double-blind, placebo-controlled study reported in the medical therapy of BPH literature (McConnell et al, 1998). The 3040 men with moderate to severe urinary symptoms were randomized to receive daily finasteride, 5 mg, versus placebo for 4 years. A quasi-AUASI was used. The baseline prostate volume in the study population was approximately 55 cm³, indicating a bias for enrolling men with markedly enlarged prostates. The mean group changes in symptom score, PFR, and prostate volume throughout the study are shown for the finasteride and placebo in Figure 92–11. The treatment-related effects of finasteride on symptom score, PFR, and prostate volume were 2.0 symptom units, 1.7 mL/sec, and 32% size reduction for those subjects on active treatment at the end of the study. The symptom and PFR improvement was modest and consistent with findings of prior finasteride trials. The PLESS demonstrated the durability of symptom and flow improvements by finasteride and very modest progression of these end points in the placebo group.

Figure 92–11 The effect of finasteride or placebo on symptom scores (on the quasi-AUASI), prostate volume, and maximal urinary flow rate over time. Values are mean (±SE) changes from baseline. The numbers below the panels show the numbers of patients with valid data who remained in the study.

(From 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. N Engl J Med 1998;338:557–63.)

The unique findings of the PLESS were related to incidences of both AUR and surgical intervention for BPH (Fig. 92–12). The cumulative incidences of AUR at 4 years in the finasteride and placebo groups were 7% and 3%, respectively (57% risk reduction). The cumulative risk of undergoing BPH-related surgery was 10% and 5% in the placebo and finasteride groups, respectively (55% risk reduction). In those men with prostate volumes greater than 55 cm³, the risk reduction of AUR and/or surgical intervention for finasteride was 70%. The risk reduction of AUR and BPH-related surgery was clinically relevant, especially in men with very large prostates. Men with significantly enlarged prostates and LUTS should therefore be advised of their significant risk of urinary retention and the beneficial effects of finasteride.

Figure 92–12 Probability of undergoing surgery for benign prostatic hyperplasia or the development of acute urinary retention during the 4-year study period in the placebo and finasteride groups. The figure shows life-table analyses of the proportion of men with each outcome. The numbers of events shown below the graphs are those that occurred during each 1-year interval. The numbers of men at risk are those at the beginning of the respective 1-year intervals. Data on men who died, were given a diagnosis of prostate cancer, or were lost to follow-up were censored at the time of death, diagnosis of prostate cancer, or discontinuation of therapy.

(From 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. N Engl J Med 1998;338:557–63.

The PLESS study also provided insights into the impact of finasteride on the detection of prostate cancer. The decision to pursue the diagnosis of prostate cancer in the study was left at the discretion of the principal investigator and therefore represents standard practice. The detection of prostate cancer was not significantly different in the placebo and finasteride group, suggesting that finasteride does not mask the diagnosis of prostate cancer (Andriole et al, 1998).

Tammela and Kontturi (1993) reported a randomized, double-blind, placebo-controlled study examining the effects of finasteride on BOO in 36 men on a routine waiting list for prostatectomy. The mean prostate volumes in the finasteride and placebo study groups were 50 and 48 cm³, respectively. The mean baseline Pdet at maximum PFR in the placebo and finasteride groups were 115 and 126 cm H₂O, respectively, indicating that the subjects were severely obstructed. The group mean changes in Pdet at PFR was +3 and −39 cm H₂O in the placebo and finasteride groups, respectively. Although the difference between placebo and finasteride was highly statistically significant, the overwhelming majority of the finasteride-treated subjects remained obstructed after treatment. The marked treatment-related changes in Pdet were not associated with statistically significant changes in symptom scores. The authors did not comment on whether the changes in Pdet correlated with changes in prostate volume. Of the subjects participating in the randomized, double-blind study, 27 completed a 4-year open-label extension study (Tammela and Kontturi, 1995). The Pdet at PFR showed further improvements over time.

PSA has become widely accepted as a screening instrument for prostate cancer (Tchetgen and Oesterling, 1995). An elevated or significantly rising PSA value is often an indication for prostatic biopsy. Finasteride reduces group mean serum PSA levels approximately 50% (Guess et al, 1993). The effect of finasteride on individual serum PSA levels is highly variable. Because of the variable effect on PSA, men who are candidates for early detection of prostate cancer should have their PSA level determined before beginning finasteride therapy. A biopsy should be performed if the PSA level is elevated. A repeat biopsy should be performed for progressively rising PSA levels after initiation of therapy.

The Prostate Cancer Prevention Trial (PCPT) (Thompson et al, 2003) showed that finasteride reduced the incidence of diagnosed prostate cancer (vs. placebo). There was an increased incidence of high-grade disease, probably owing to technical pathologic reasons.

Gross hematuria is a relatively rare, yet troublesome, manifestation of BPH. Puchner and Miller (1995) reported an uncontrolled personal experience of 18 BPH patients treated with finasteride for refractory gross hematuria secondary to BPH. In the 18 reported cases, 12 of the patients had undergone a prior prostatectomy. Finasteride was very effective in relieving the postprostatectomy-associated gross hematuria. Miller and Puchner (1998) reported a follow-up series that demonstrated the long-term effectiveness of finasteride for the treatment of hematuria from BPH. Carlin and coworkers (1997) reported resolution of gross hematuria in 12 men treated with finasteride. These preliminary observations have been confirmed by a randomized, double-blind, placebo-controlled study by Foley and associates (2000) demonstrating that finasteride prevents recurrent gross hematuria secondary to BPH. Gross refractory hematuria recurred within 1 year for 63% and 14% of men randomized to placebo and finasteride, respectively.

Dutasteride

Dutasteride is a dual inhibitor of 5α-reductase types 1 and 2 and therefore has a greater impact on suppressing serum DHT levels (Clark et al, 2004). In a randomized controlled trial of 4325 men (2951 completed) Roehrborn and colleagues (2002) reported that serum DHT was reduced by 90.2%. The symptom score was improved by 4.5 points (21.4%) (P < .001), and the maximal flow rate improved significantly by 2.2 mL/sec (P < .001) at 24 months. The risk reduction of AUR was 57% and the risk reduction of BPH-related surgery was 48% compared with placebo. Debruyne and coworkers (2004) reported the pooled results of a 2-year open-label extension study in which both dutasteride- and placebo-treated groups received dutasteride 0.5 mg/day. Significant improvements in symptom scores and PFRs were observed in both study groups. It was concluded that long-term treatment with dutasteride results in continuing improvements in both symptoms and PFR and that the risk reduction of AUR and BPH-related surgery was durable over 4 years. Similar to finasteride, the principal side effects were loss of libido and ED, but these were most frequently seen at the start of therapy and declined over time with treatment. Roehrborn and associates (2004) reported similar results and confirmed 93% suppression of DHT at 4 years. Recently, the results of the REDUCE study of dutasteride versus placebo in terms of their ability to prevent prostate cancer have been reported by Andriole and associates (2009). A 23% reduction in prostate cancer was seen in the dutasteride study arm, with no increase in less well-differentiated cancers in the dutasteride arm. Beneficial effects on symptoms, PFR, and risk of progression were also seen in this group of men at high risk for BPH progression as well as for a prostate cancer diagnosis.

Further re-analysis has suggested some risk of high-grade cancers, and in December 2010 the FDA Oncologic Drugs Advisory Committee voted against use of dutasteride for reduction in the risk of prostate cancer. The manufacturer has withdrawn this application.

Zanoterone

Zanoterone is a steroidal competitive androgen receptor antagonist (Juniewicz et al, 1993). Berger and coworkers (1995) reported a multicenter, randomized, double-blind, placebo-controlled study in 463 subjects receiving placebo and 100, 200, 400, or 800 mg of zanoterone for 6 months. The group mean changes in AUASI score were not reported; however, the differences between placebo and all of the zanoterone groups were not statistically significant. The mean baseline prostate volumes ranged between 37.7 and 42.2 cm³ in the five treatment groups. The differences between the percent group mean changes in prostate volume in the placebo (−6%) and active treatment groups (−4% to −8%) were not significantly different. Interestingly, the differences between the percent group mean changes in serum PSA in all of the active treatment groups were significantly greater than those in the placebo group, despite the lack of an apparent drug effect on prostate volume. The difference between the group mean changes in PFR for placebo and the 200-mg dose was 0.7 mL/sec. Fifty-six percent and 22% of all zanoterone subjects reported breast pain and gynecomastia, respectively. The incidence and severity of adverse clinical events and the equivocal efficacy precluded further development of this drug for BPH.

Flutamide

Flutamide is an orally administered nonsteroidal antiandrogen that inhibits the binding of androgen to its receptor (Sufrin and Coffey, 1973). The first reported randomized, double-blind, placebo-controlled trial in BPH examined the safety and efficacy of flutamide in 31 men with symptomatic BPH (Caine et al, 1975). Statistically significant differences were not observed between placebo and 300 mg of flutamide for symptoms, prostate size, PVR, and PFR. Stone (1989) reported a multicenter, randomized, double-blind study comparing flutamide and placebo in men with BPH. Eighty-four patients were randomized to receive 24 weeks of either flutamide, 250 mg three times a day, or placebo. Of the 84 patients, 58 (69%) and 12 (14%) were evaluable 12 and 24 weeks on double-blind treatment. The small sample size at 24 weeks precludes any meaningful conclusions. The between-group comparisons of group mean changes from baseline for placebo versus flutamide were not statistically significant at any time point. The incidences of breast tenderness and diarrhea in the flutamide group were 53% and 11%, respectively. Although the interim analysis was reported in 1989, a subsequent report of the multicenter study has not been published.

Cetrorelix

Cetrorelix is a gonadotropin-releasing hormone antagonist that has been investigated for BPH. A potential advantage of a gonadotropin-releasing hormone antagonist over the luteinizing hormone–releasing hormone agonists in the treatment of BPH is the ability to titrate the level of androgen suppression. This would be clinically relevant if different levels of androgen suppression mediate prostate size reduction and adverse events (hot flashes, decreased libido, ED). An open-label study of 11 men demonstrated that cetrorelix reduced prostate volume and improved LUTS without significant adverse events. Lepor and coworkers (1997) reported a proof of concept randomized, double-blind, placebo-controlled study of cetrorelix in men with BPH. After an 8-day placebo lead in, men received daily subcutaneous injections of placebo or 1 mg of cetrorelix (group C01) for 27 days. One group received loading doses of 10 mg of cetrorelix on the first 4 days of active treatment (group C10). Maximal lowering of testosterone was observed within 24 hours. The testosterone level was reduced to castrate levels in the C10 group and to intermediate testosterone suppression level in the non–loading-dose group (C01) (approximately 20 ng/dL). Men exhibiting a clinical effect were observed for 1 year. In the C10 and C01 groups, the treatment-related improvement in AUASI score was 3.0 and 2.0 symptom units, respectively. In both the C10 and the C01 groups the treatment-related improvement in PFR was 2.0 mL/sec. The treatment-related reductions of prostate volume were 5.5 and 3.0 cm³, respectively. The treatment-related incidence of hot flashes and sexual dysfunction in the C01 group was negligible. During the open-label extension, prostate volume did not return to baseline, suggesting a prolonged effect on the disease. Because of problems with the drug formulation, phase 3 studies with cetrorelix were not pursued until recently (Debruyne et al, 2008).

In this dose-finding study three dosing regimens were explored: one or two injections weekly for 4 weeks. In all groups a rapid improvement in mean IPSS was obtained, with a peak effect of −5.4 to −5.9 (placebo: −2.8). Changes from baseline and differences to placebo were statistically significant up to week 20. Placebo response was less sustained than usually seen with oral preparations.

The primary disadvantage of cetrorelix and other gonadotropin-releasing hormone antagonists will be the requirement for an injection and the cost. If single-injection therapy provides desirable clinical response (over, for example, 6 months) with minimal adverse events there may exist a role for these antagonists in the treatment of BPH.

Aromatase Inhibitors

The rationale for aromatase inhibition is that estrogens may be involved in the pathogenesis of BPH. The estrogenic effect most likely mediates stromal-epithelial interactions that regulate the proliferative activity of the prostate. Several observations support the role of stroma in the development of BPH and the influence of estrogens on prostatic stroma. The inductive potential of prostatic mesenchyme (stroma) is supported by the observation of Cunha and colleagues (1980) in a mouse embryonic animal model. Coffey and Walsh (1990) reported that estrogen treatment of castrated beagles produced a threefold to fourfold increase in the total amount of prostatic stroma. Estrogens also greatly enhanced the ability of androgens to induce BPH in a canine model (Walsh and Wilson, 1976; DeKlerk et al, 1979). This synergistic effect may be mediated by the ability of estrogens to upregulate prostatic androgen receptor content. Stromal hyperplasia can be induced in the prostates of dogs and monkeys treated with aromatizable androgens and prevented by aromatase inhibitors such as atamestane (Habenicht et al, 1987; Habenicht and El Etreby, 1989).

Atamestane is a highly selective aromatase inhibitor that lowers both serum and intraprostatic levels of estradiol and estrone (El Etreby et al, 1991). Gingell and coworkers (1995) reported a multicenter, randomized, double-blind, placebo-controlled study comparing placebo and 400 mg of atamestane in 160 subjects with clinical BPH. Atamestane resulted in a statistically significant decrease in serum estradiol and estrone levels and a statistically significant increase in serum testosterone. No statistically significant group mean differences were observed for changes in Boyarsky symptom score, PFR, or prostate volume between the atamestane and the placebo groups. One of the explanations contributing to the failure of atamestane to achieve clinical efficacy was the increase in testosterone. The development of atamestane for BPH was suspended because of these negative clinical findings. The failure to demonstrate that atamestane causes regression of established BPH or clinical improvement does not negate the influence of estrogens in the pathogenesis of BPH.

Finasteride and dutasteride are the only drugs available that achieve androgen suppression with acceptable tolerability. Both symptoms and flow rates are improved, especially in men with larger glands. Impotence and decreased ejaculatory volume are the primary treatment-related adverse experiences. The literature also suggests finasteride and dutasteride may be offered to men with hematuria secondary to friable prostatic tissue and in those men with LUTS and enlarged prostates who elect to reduce their risk of developing urinary retention.

Serenoa Repens (Saw Palmetto Berry)

The extract of the berry of the American saw palmetto, or dwarf palm plant, Serenoa repens (also known by its botanical name of Sabal serrulatum), is the most popular phytotherapeutic agent available for the treatment of BPH. Although numerous clinical trials with saw palmetto berry extracts have been published, many were uncontrolled open-label studies, thus providing little useful information in determining the efficacy of these phytotherapies. Even though placebo-controlled studies have been published (Table 92–17), most of them are flawed and few of them would meet the generally accepted criteria developed by the International Consultation on BPH for assessing treatment results in men with LUTS (McConnell, 1998). The studies are of limited value because of their small numbers of patients, short duration (only 1 to 3 months), and lack of use of standardized symptoms scores.

Table 92–17 Placebo-Controlled Trials of Serenoa repens (Permixon)

For example, nocturia was the only symptom available for analysis in all the studies reviewed in two meta-analyses (Wilt et al, 1998; Boyle et al, 2000). In the Wilt and colleagues (1998) meta-analysis of 18 trials involving 2939 patients using various S. repens single and combination preparations, the mean weighted difference for nocturia between patients and subjects taking a placebo was −0.76 times per evening (−1.22 to −0.32) for 10 trials. In the Boyle and coworkers (2000) meta-analysis of 13 trials involving 2859 patients using only the Permixon brand of S. repens, the attributable reduction of nocturnal urinations was 0.50 (±0.01) times per night. Wilt and coworkers conducted a systematic review, first published in 1998 and updated in 2002 and again in 2009, to evaluate the efficacy and adverse events of S. repens (Tacklind et al, 2009). Initially their analysis was mildly supportive of the efficacy of S. repens. In this update 9 new trials involving 2053 additional men (a 64.8% increase) were included. For the main comparison—S. repens versus placebo—3 trials were added with 419 subjects and three end points (IPSS, peak urine flow, prostate size). Overall, 5222 subjects from 30 randomized trials lasting from 4 to 60 weeks were assessed. Twenty-six trials were double blinded, and treatment allocation concealment was adequate in 18 studies. With these extra data and especially the double-blind, placebo-controlled trial reported by Bent and associates (2006), they concluded that S. repens was not superior to placebo in improving IPSS or PFR or for reducing prostate size. For nocturia, S. repens was significantly better than placebo (weighted mean difference −0.78 nocturnal visits, 95% CI −1.34 to −0.22, P < 0.05; 9 trials), but with the caveat of significant heterogeneity within studies. A sensitivity analysis, utilizing higher-quality, larger trials demonstrated no significant difference. S. repens was also not superior to finasteride or to tamsulosin. Overall, it was concluded that S. repens was not more effective than placebo for treatment of urinary symptoms consistent with BPH.

Despite inherent weaknesses in meta-analyses (Table 92–18) these analyses attempt to maximize the information available from clinical trials using S. repens in the treatment of symptomatic BPH/LUTS and to provide the best information available until the appropriate placebo-controlled, randomized clinical studies are conducted. However, approximately 35% of the prior meta-analyses performed are not accurately predicted by the subsequent randomized controlled trials (LeLorier et al, 1997). Therefore, whereas the efficacy of S. repens has not been completely determined, current meta-analysis suggests that S. repens is not more effective than placebo for treatment of urinary symptoms consistent with BPH.

Table 92–18 Inherent Weaknesses of Meta-Analyses

Pygeum africanum (African Plum)

In addition to the proposed mechanisms of actions previously discussed, P. africanum (Tadenan) has been postulated to have additional beneficial effects on LUTS by having a protective effect on the obstructed bladder. Using a rabbit model with partial BOO, Levin and associates (1996) demonstrated that changes in bladder mass, decrease in compliance, and alterations in contractile response to various forms of stimulation could be tempered by pretreatment with P. africanum.

A review of the published experience with P. africanum identified 2262 patients treated with this extract: 1810 in open-label studies and 452 in comparative trials (Andro and Riffaud, 1995). Twelve double-blind, placebo-controlled studies were completed between 1972 and 1990. Only one study enrolled more than 100 subjects, and none was longer than 12 weeks or used standardized symptom scores. In a 2-month open-label study (Breza et al, 1998) and a 2-month comparison trial of 50 mg twice daily versus 100 mg daily (Chatelain et al, 1999) both trials demonstrated a reduction in IPSS score of approximately 40% (5.7 and 6.4 units) and an improvement in PFRs of approximately 18% (2.1 and 1.7 mL/sec) over baseline. However, without a placebo study arm, the actual magnitude of drug effect cannot be determined. Thus, without at least the data from the Tadenan-IPSS study, the efficacy of P. africanum cannot be determined. Additionally, there are few other significant data available with regard to any of the other P. africanum extracts.

Thus none of those trials meets the guidelines recommended by the International Consultation Conferences on BPH. Another meta-analysis for the Cochrane collaboration (Wilt et al, 2002a) concluded that an effect was possible but not proven. Therefore the data concerning the efficacy of P. africanum are not conclusive.

Hypoxis rooperi (South African Star Grass)

Hypoxis rooperi (Harzol) has been studied in both a 6-month double-blind, placebo-controlled trial of 200 patients (Berges et al, 1995) and, subsequently, an open-label follow-up (Berges et al, 2000). In the initial study, statistically significant improvements were documented for symptom scores (IPSS), quality of life, PFRs, and PVRs (Berges et al, 1995). The placebo group showed appropriate mild improvement in these parameters. The IPSS improved by 7.4 units for β-sitosterol patients and 2.3 units for placebo. Similarly, peak urinary flow rates improved by 5.2 mL/sec for treated patients compared with 1.1 mL/sec for placebo. This magnitude of improvement has not been observed with any other medical therapy previously evaluated for BPH.

During the follow-up study, patients were able to remain on or switch over to Harzol therapy. For those 38 patients who continued Harzol therapy, IPSS improvements were maintained. The 27 patients who received placebo initially and were then treated with Harzol demonstrated similar levels of improvement in IPSS scores and PFRs. Surprisingly, the 14 patients who stopped therapy still maintained similar levels of improvement over the next 12 months. This suggests that intermittent therapy could be an option.

Another product (Azuprostat) contains primarily β-sitosterols from H. rooperi as well as from Pinus (pine) and Picea (spruce). Although this is a combination preparation of extracts, its proposed active ingredient β-sitosterol is common to all three. This product (Azuprostat) was evaluated in a 6-month randomized, placebo-controlled trial with 177 patients (Klippel et al, 1997). Significant improvements in IPSS scores, PFRs, and PVRs were found. IPSS scores improved by 8.2 units for treated patients and 2.8 units for placebo. PFRs improved by 8.9 mL/sec for treated patients and 4.4 mL/sec for placebo. This magnitude of improvement has not been reported for any type of medical therapy for BPH. It is hard to believe these results in which the mean post-treatment PFRs are 19.4 mL/sec, which is a normal value for younger men and not for the typical age of men in the study. If these results are reproducible, they would rival surgical intervention.

Wilt and associates (1999) also produced a meta-analysis for β-sitosterol products. It included four trials (two mentioned earlier and two earlier suboptimal ones) encompassing three different products: Harzol, Azuprostat, and WA184, all of which have different amounts of β-sitosterol. WA184 did not improve PFR (Kadow et al, 1986). Again, this meta-analysis is tempered by the same factors mentioned previously. Their conclusion was that β-sitosterol does improve urologic symptoms and urinary flow rates, but its long-term effectiveness, safety, and ability to prevent the complications of BPH are unknown (Wilt et al, 1999).

Other Extracts

The other extracts listed in Table 92–14—Urtica dioica, Cucurbita pepo, Secale cereale, and Opuntia—have even fewer relevant clinical studies published than the aforementioned ones. Of these extracts, Secale cereale (Cernilton) had two placebo-controlled trials published over a decade ago that did not have standardized scores, used different dosages of the preparation, lasted 12 and 24 weeks, and enrolled only 103 and 60 subjects (Buck et al, 1990). Wilt and associates (1999) in a systematic review and meta-analysis of Cernilton concluded that it modestly improves overall urologic symptoms including nocturia, but additional trials are needed to evaluate clinical effectiveness. More recently Zhang and coworkers (2008) have reported a randomized, double-blind, placebo-controlled clinical trial of a flaxseed lignan extract containing 33% secoisolariciresinol diglucoside (SDG). SDG was evaluated for its ability to alleviate LUTS in 87 subjects with BPH with repeated measurements conducted over a 4-month period using treatment dosages of 0 (placebo), 300, or 600 mg/day SDG. After 4 months of treatment, 78 of the 87 subjects completed the study. For the 0, 300, and 600 mg/day SDG groups, respectively, the IPSS decreased −3.67 ± 1.56, −7.33 ± 1.18, and −6.88 ± 1.43 (mean ± SE, P = .100, < .001, and < .001 compared with baseline), the quality of life score improved by −0.71 ± 0.23, −1.48 ± 0.24, and −1.75 ± 0.25 (mean ± SE, P = .163 and .012 compared with placebo and P = .103, < .001, and < .001 compared with baseline), and the number of subjects whose LUTS grade changed from “moderate to severe” to “mild” increased by 3, 6, and 10 (P = .188, .032, and .012 compared with baseline). PFR insignificantly increased 0.43 ± 1.57, 1.86 ± 1.08, and 2.7 ± 1.93 mL/sec (mean ± SE, no statistical significance reached), and postvoid urine volume decreased insignificantly by −29.4 ± 20.46, −19.2 ± 16.91, and −55.62 ± 36.45 mL (mean ± SE, no statistical significance reached). The observed decreases in IPSS and quality of life score were correlated with the concentrations of plasma total lignans. It was concluded that dietary flaxseed lignan extract appreciably improved LUTS in BPH subjects and that the therapeutic efficacy appeared comparable to that of α₁-adrenergic receptor blockers and 5α-reductase inhibitors.