The Effect of Finasteride in Men with Benign Prostatic Hyperplasia

Subjects

This was a multicenter, double-blind, placebo-controlled study conducted at 25 centers in the United States and 5 centers in Canada. The study was approved by the institutional review board at each center, and all the men gave written informed consent. A total of 895 men ranging in age from 40 to 83 years who had benign prostatic hyperplasia were enrolled in the study. All the men had some symptoms of urinary obstruction, an enlarged prostate gland on digital rectal examination, and maximal urinary-flow rates of less than 15 ml per second (with a voided volume of 150 ml or more). Men with very low urinary-flow rates were not excluded, unless they were considered to be at risk for total urinary obstruction. Any man whose rectal examination suggested the presence of prostatic cancer was required to have a biopsy with negative results for entry into the study. Men with a residual urinary volume of more than 350 ml, a serum prostate-specific—antigen concentration of ≥40 μg per liter, or evidence of prostatic cancer, urinary tract infection, chronic prostatitis, or neurogenic bladder were excluded.

Study Protocol

After a two-week period during which the men received a placebo in single-blind fashion, they were randomly assigned in a double-blind fashion to one of three treatment groups: 297 were assigned to receive a 5-mg tablet of finasteride daily, 298 to receive a 1-mg tablet of finasteride daily, and 300 to receive a placebo tablet each day. The tablets were taken once each day for 12 months. Each man was evaluated monthly by the same investigator, at which time urinary symptoms, side effects, and the degree of compliance with the treatment regimen were assessed and the urinary-flow rate and residual urinary volume were measured. Prostatic volume was measured at base line and after 6 and 12 months of treatment. In addition, each man had a complete ophthalmologic examination at base line and after 12 months. Every three months, serum aminotransferases, urea nitrogen, creatinine, sodium, potassium, calcium, and glucose were measured and a complete blood count was performed. After month 12, the men could enter an open study in which they received 5 mg of finasteride daily. Compliance with treatment was excellent as determined by counting the number of tablets remaining at each visit and by periodic measurements of serum dihydrotestosterone concentrations.

Table 1. Table 1. Reasons for Withdrawal from the Finasteride Study.*

During the study, 40 (13 percent) of the men in the group receiving 5 mg of finasteride, 28 (9 percent) of those receiving 1 mg of finasteride, and 37 (12 percent) of those receiving placebo withdrew from the study for the reasons shown in Table 1.

Evaluation Procedures

The patients' symptoms were assessed on the basis of their responses to a questionnaire modified from that of Boyarsky et al.¹⁵ The total symptom score was calculated from the responses to nine questions about urinary hesitancy, terminal dribbling, impairment of the size and force of the urinary stream, interruption of urination, incomplete emptying, urgency, dysuria, clothes wetting, and straining or pushing to start urinary flow. Each symptom was scored according to a five-point scale, with a score of 0 indicating the absence of a symptom and a score of 4 the presence of a severe symptom, so that the worst possible score was 36. The total score was divided into two subgroups: one covering changes due to obstructive uropathy and the other covering changes involving nonobstructive uropathy. The obstructive-uropathy score was calculated from the response to the first five questions, and the nonobstructive-uropathy score was calculated from the response to the last four questions.

The symptom score was validated in studies of normal men and men with benign prostatic hyperplasia before and after transurethral resection of the prostate. The mean (±SD) score in the 20 normal men was 0.9±1.8 (unpublished data). The mean score among 20 men studied in the office of a urologist before and 1 to 3 months after surgery decreased from 10.9±5.9 to 3.7±3.8, and it decreased from a preoperative mean of 8.1±3.5 to 4.4±3.6 among 7 men who responded to a questionnaire mailed 12 months after surgery.

Urinary-outflow obstruction was measured with a urinary flow-meter (Urodyn 1000, Dantec, Mahwah, N.J.). In normal men, the maximal urinary flow exceeded 15 ml per second and their mean urinary flow was more than 6.9 ml per second.¹⁶ Measurements of maximal and mean urinary-flow rates were used in the analyses only if at least 150 ml of urine was excreted. Approximately 9 percent of the urinary-flow measurements were excluded on this basis. Residual urinary volume was measured by transabdominal ultrasonography.

Prostatic volume was measured by magnetic resonance imaging at base line and after 3, 6, and 12 months of treatment. Measurements were made in the axial, sagittal, and coronal planes, and the volume of the prostate was calculated according to the following formula: volume (in milliliters) = (4/3)π(axial value/2)(coronal value/2)(sagittal value/2). With the use of this approach the mean (±SD) prostatic volume in 23 men ranging in age from 31 to 50 years (mean, 41) with no urinary symptoms was 25±19 ml.

Laboratory Tests

Serum testosterone and dihydrotestosterone were measured by radioimmunoassay after Chromatographic separation at Endocrine Sciences (Tarzana, Calif.).¹² The sensitivity of the assay was 5 ng per deciliter (0.17 nmol per liter) for testosterone and 2 ng per deciliter (0.07 nmol per liter) for dihydrotestosterone, and the interassay variations were 9.6 percent and 11.7 percent, respectively. The range of serum testosterone in normal men was 344 to 1029 ng per deciliter (12 to 36 nmol per liter), and of serum dihydrotestosterone, 30 to 88 ng per deciliter (1.0 to 2.9 nmol per liter). Serum luteinizing hormone was measured by radioimmunoassay with human pituitary standards obtained from the World Health Organization. Serum prostate-specific antigen was measured by MRL Laboratories (Cincinnati) with a solid-phase, two-site immunoradiometric assay (Hybritech, La Jolla, Calif). The normal concentration was less than 4 μg per liter, the sensitivity of the assay was 0.2 μg per liter, and the interassay variation was 3.8 percent. The measurements were made as the samples were collected, and the results were not released from the central laboratory until the end of the study.

Statistical Analysis

Table 2. Table 2. Base-Line Characteristics of Men with Benign Prostatic Hyperplasia, According to Treatment-Group Assignment.*

Table 3. Table 3. Symptom Scores, Maximal Urinary-Flow Rates, and Prostatic Volume in Men with Benign Prostatic Hyperplasia Treated with Finasteride or Placebo.*

An intention-to-treat analysis was used in evaluating the results. The last results in men who did not complete the study were carried forward through month 12. An analysis of the data based only on results in men for whom data were available at each time point (per-protocol analysis) produced results that were similar to those of the intention-to-treat analysis. An analysis of variance was used to compare the effects of treatment on the total, obstructive-uropathy, and nonobstructive-uropathy symptom scores; prostatic volume; maximal urinary-flow rate; and serum hormone and prostate-specific—antigen values at each time point. The analysis of variance model analyzed the effects of interactions between treatment groups, between study centers, and between treatment groups and study centers. No significant qualitative interaction was found. All analyses of variance were performed with both parametric methods (with use of the actual change from base line) and nonparametric methods (a rank transformation of these values across treatment groups and centers) to ensure the consistency of inferences. Because of a general lack of symmetry in the distributions, the median values are given for prostatic volume and serum prostate-specific antigen, except in Tables 2 and 3. The values for all other variables are given as mean changes. Within-group tests were done with a paired t-test or with Puri's extension of the signed-rank test, with data stratified according to center. The Cochran—MantelHaenszel test was used to analyze categorical variables, with data stratified according to center. All tests of significance were two-tailed, and all P values of 0.05 or less were considered to indicate significance. No adjustment was made for the performance of multiple tests.

The base-line characteristics of the men enrolled in the study are shown in Tables 2 and 3. The three treatment groups were similar except that the total voided volume was significantly lower (P<0.05) and the obstructive-symptom scores significantly higher (P<0.05) in the men who received 1 mg of finasteride than in those who received placebo.

Changes in Hormone Secretion and Prostate-Specific—Antigen Levels

Figure 1. Figure 1. Effect of Treatment with Placebo (Circles), 1 mg of Finasteride (Triangles), or 5 mg of Finasteride (Squares) on Mean (±SE) Serum Dihydrotestosterone (Panel A) and Median (±95 Percent Confidence Interval) Serum Prostate-Specific—Antigen (Panel B) Concentrations in Men with Benign Prostatic Hyperplasia.

The stippled area in Panel A indicates the normal range in men. To convert values for dihydrotestosterone to nanomoles per liter, multiply by 0.033. Month 0 represents the base line. Values before month 0 were obtained during the two-week placebo run-in period.

The mean base-line concentrations of serum dihydrotestosterone, testosterone, luteinizing hormone, and prostate-specific antigen were similar in the three groups. The mean serum dihydrotestosterone concentrations decreased significantly in the two finasteride-treated groups (P<0.001) during the first two weeks of treatment and did not change thereafter (Fig. 1A). The decrease in serum dihydrotestosterone concentrations among the men who received 5 mg of finasteride daily was significantly greater than that among the men who received 1 mg of finasteride daily (P≤0.01) at all times from 2 weeks to 12 months.

In both finasteride-treated groups, serum testosterone concentrations increased approximately 8 to 10 percent after two weeks of treatment, and they remained increased thereafter. The mean (±SD) values in the men who received 5 mg of finasteride were 484±173 ng per deciliter (17±6 nmol per liter) after 2 weeks and 488±173 ng per deciliter ( 17±6 nmol per liter) after 12 months, as compared with 492±161 ng per deciliter (17±6 nmol per liter) and 489±161 ng per deciliter (17±6 nmol per liter), respectively, in the men who received 1 mg of finasteride. Despite the increases, all values were within the normal range at all times. The values in the two finasteride-treated groups were significantly higher (P<0.001) than those in the placebo group at all times, but there was no significant difference between the two finasteride-treated groups.

Serum luteinizing hormone concentrations increased in all three groups during the first two months. However, the increases in both finasteride-treated groups were significantly higher than those in the placebo group (P<0.05). Among the men who received 5 mg of finasteride, the mean (±SD) serum luteinizing hormone concentration increased from 6.6±2.6 U per liter to 7.4±2.7 U per liter after 2 months and to 7.6±2.9 U per liter after 12 months. Among the men who received 1 mg of finasteride, the serum luteinizing hormone concentration increased from 6.8±2.7 U per liter to 7.7±3.3 U per liter after 2 months and to 7.8±3.4 U per liter after 12 months, and in the men who received placebo, the concentration increased from 6.8±2.8 U per liter to 7.2±3.1 U per liter and 7.3±3.4 U per liter, respectively.

The median serum prostate-specific—antigen concentrations in the placebo group did not change during the 12-month treatment period (Fig. 1B). In contrast the men in both finasteride-treated groups had significant reductions in serum prostate-specific antigen (P<0.001 for the comparison with the placebo group) at all times from month 3 through month 12 of treatment (Fig. 1B). The median decrease was 50 percent among the men who received 5 mg of finasteride and 48 percent among those who received 1 mg of finasteride.

Changes in Symptom Scores

Figure 2. Figure 2. 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<0.05), the pound symbols (P<0.01), and the dagger (P<0.001) indicate significant differences between the finasteride-treated groups and the placebo group. Month 0 represents the base line.

The symptom scores at base line and at 12 months are shown in Table 3, and the results of treatment at each time point are shown in Figure 2. In the placebo group the total symptom scores decreased during the first four months, indicating that these patients had fewer or less severe symptoms, but then increased gradually toward base-line values. In contrast, the men treated with 5 mg of finasteride had a significant decrease in total symptom scores at several points from week 2 through month 12 (P<0.02 at months 2 and 7, P<0.01 at months 10 and 11, and P<0.001 at month 12), as compared with the placebo group. The maximal decrease in the scores at 12 months was 2.7 points in the group given 5 mg of finasteride (Table 3). The men treated with 1 mg of finasteride had no significant changes in the symptom scores.

The changes in the obstructive-symptom scores were similar to those in the total scores. The mean obstructive-symptom score at base line was approximately 7 (Table 3). At 12 months, the scores were lower by 1.9 points in the group given 5 mg of finasteride, by 1.4 in the group given 1 mg of finasteride, and by 0.8 in the group given placebo. The scores decreased in all three groups during the first nine months of treatment. The scores continued to decrease in the group given 5 mg of finasteride but increased in the group given placebo, such that the scores in these two groups were significantly different (P<0.001) at 12 months. The scores in the group given 1 mg of finasteride did not change after 9 months, but they were significantly different from the scores in the group given placebo at 12 months (P<0.05). The nonobstructive-symptom scores decreased in the group given 5 mg of finasteride but not in the other two groups. The scores in the group given 5 mg of finasteride were significantly lower than those in the placebo group at 10 months (P = 0.05) and 12 months (P<0.01) (Table 3).

Changes in Urinary Flow Rate

Figure 3. Figure 3. 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 stippled area indicates the range in which urinary flow was considered to be obstructed. Month 0 represents the base line. Values before month 0 were obtained during the two-week placebo run-in period. The asterisks (P<0.05), pound symbols (P<0.01 ), and daggers (P<0.001 ) indicate significant differences between the finasteride-treated groups and the placebo group.

The maximal urinary-flow rates increased slightly during the two-week pretreatment period in all three groups. During the 12 months of treatment, the maximal urinary-flow rates increased progressively in both finasteride-treated groups, but not in the group given placebo (Table 3, Fig. 3). At 6 and 12 months the maximal flow rates in both finasteride-treated groups were significantly greater than both the base-line values and the values in the placebo group (P<0.05 to P<0.001). After 12 months of treatment the maximal urinary-flow rate had increased by at least 3 ml per second in 31 percent of the men who received 5 mg of finasteride, 30 percent of those who received 1 mg of finasteride, and 17 percent of those who received placebo. The differences between the placebo group and the two finasteride-treated groups were significant (P≤0.001). The maximal urinary-flow rates at base line were not significantly different between the men who had increases in maximal urinary flow of at least 3 ml per second and the men who did not have such increases.

At each measurement from 5 to 12 months, the mean urinary-flow rates increased slightly but significantly more in the men treated with 5 mg of finasteride than in the men given placebo. The mean values at base line and after 12 months were 5.6±2.1 and 6.4±2.7 ml per second in the group given 5 mg of finasteride, 5.3±2.2 and 6.1±2.4 ml per second in the group given 1 mg of finasteride, and 5.6±2.1 and 5.8±2.3 ml per second in the group given placebo. There were no significant differences in residual urinary volume between the groups at any time.

Changes in the Size of the Prostate

Figure 4. Figure 4. Median Change (±95 Percent Confidence Interval) in Prostatic Volume in Men with Benign Prostatic Hyperplasia during Treatment with Placebo (Circles), 1 mg of Finasteride (Triangles), or 5 mg of Finasteride (Squares).

Month 0 represents the base line.

During the first six months, the median size of the prostate decreased progressively in both finasteride-treated groups (Fig. 4), after which it did not change significantly, and it was significantly smaller in both finasteride-treated groups than in the placebo group at all times (P<0.001). After 12 months of treatment, the prostate had shrunk by 19 percent from base line in the group given 5 mg of finasteride, by 18 percent in the group given 1 mg of finasteride, and by 3 percent in the group given placebo (Table 3). The prostatic volume had decreased by at least 20 percent in 43 percent of the men given 5 mg of finasteride after 6 months of treatment and in 49 percent after 12 months. In the group given 1 mg of finasteride, 48 percent and 47 percent of the men had a reduction of at least 20 percent after 6 and 12 months, respectively. In the placebo group, this degree of reduction occurred in 18 percent and 20 percent of the men at 6 and 12 months, respectively. At all times the difference in the percentage of men with a reduction in prostatic size of at least 20 percent was significantly greater in the finasteride-treated groups than in the placebo group (P<0.001).

Side Effects and Unrelated Complications

Table 4. Table 4. Side Effects in Men with Benign Prostatic Hyperplasia Treated with Finasteride or Placebo for 12 Months.*

The symptoms considered by the investigators to be possibly, probably, or definitely drug-related in at least 0.5 percent of the men are listed in Table 4. The overall frequency of these symptoms was similar in the three groups. The proportion of men reporting decreased libido and decreased ejaculate volume was significantly higher (P<0.05) in both finasteride-treated groups than in the placebo group, and the proportion of men who reported impotence was significantly higher (P<0.05) in the group given 1 mg of finasteride than in the placebo group. Only eight men withdrew from the study because of sexual dysfunction (one in the placebo group, three in the group given 1 mg of finasteride, and four in the group given 5 mg of finasteride). Two men, each taking 1 mg of finasteride daily, died during the study: one died of cardiac arrest, and one committed suicide. Prostatic cancer was diagnosed in four men (one in the placebo group, two in the group given 1 mg of finasteride, and one in the group given 5 mg of finasteride). Three of these four men had serum prostate-specific—antigen concentrations of more than 10 μg per liter at base line. The three men taking finasteride who were found to have prostatic cancer had reductions in serum prostate-specific—antigen levels during treatment, whereas the man in the placebo group had an increase. Eleven men had prostatic surgery during the study. One man had a radical prostatectomy for prostatic cancer, and 10 men had transurethral resections (3 in the group given 5 mg of finasteride, 4 in the group given 1 mg of finasteride, and 3 in the placebo group). There was no significant effect of treatment on the pulse rate, blood pressure, or visual acuity. Only eight men had serious abnormalities in laboratory values during the study. One man in the group given 5 mg of finasteride had an increase in serum creatinine, calcium, potassium, and urea nitrogen concentrations. One man in the group given 1 mg of finasteride had a decrease in red-cell and white-cell counts, and one man had an increase in serum prostate-specific—antigen concentrations. In the placebo group, one man became positive for the human immunodeficiency virus and three men had hematuria.

This study was undertaken to evaluate the safety and efficacy of finasteride in a large number of men with benign prostatic hyperplasia. Serum dihydrotestosterone concentrations decreased to levels present after castration soon after the initiation of finasteride therapy; the greatest degree of suppression occurred in the men who received 5 mg of finasteride daily. Serum testosterone and luteinizing hormone concentrations increased only slightly and remained well within the normal range throughout the study.

Prostatic volume decreased significantly within three months in both finasteride-treated groups. The median reductions in prostatic volume after 12 months were 19 percent in the group given 5 mg of finasteride daily, 18 percent in the group given 1 mg of finasteride daily, and 3 percent in the group given placebo. Approximately 50 percent of the men who received finasteride had a decrease of 20 percent or more in the size of the prostate, a response similar to that reported in men treated with an analogue of gonadotropin-releasing hormone (24 percent reduction in size)⁶ or surgical castration (31 percent reduction in size).⁵ The reduction in prostatic volume¹⁷ in the finasteride-treated men was consistent with the significant decreases in serum prostate-specific—antigen concentrations in these men, and thus were indicative of a decrease in prostatic epithelial-cell function. The three men receiving finasteride who were found to have prostatic cancer during the study also had decreases in serum prostate-specific—antigen concentrations. Therefore, a decrease in the concentration of serum prostate-specific antigen during finasteride therapy does not mean that a man does not have prostatic cancer, nor should it be interpreted to mean that he has been adequately treated for prostatic cancer.

The maximal urinary-flow rate increased during finasteride treatment. The proportion of men who had an increase of at least 3 ml per second after 12 months was significantly higher in both finasteride-treated groups than in the placebo group. Although the increases in the maximal urinary-flow rate were small as compared with those that occur after transurethral resection of the prostate (8 to 9 ml per second^{18 , 19}), studies of the natural history of benign prostatic hyperplasia²⁰ indicate that the mean decrease in maximal flow is about 0.2 ml per second per year. Thus, an improvement in maximal urinary flow of 3 ml per second represents a shift of about 15 years in the natural course of the disease. Despite the improvement in the maximal urinary flow, the residual volume did not decrease in these patients, possibly as a result of bladder decompensation due to chronic obstruction.

The men who received 5 mg of finasteride daily had a significantly greater decrease in total symptom scores than the men given placebo. The decrease in the scores of the men who received 1 mg of finasteride daily was intermediate. After 12 months, the changes in the total symptom scores in all groups were 30 to 70 percent of the changes that occur in men treated with transurethral resection of the prostate.

Finasteride was well tolerated. The overall frequency of side effects was similar in the three groups, but the men who received finasteride were more likely to have symptoms of sexual dysfunction. The long-term risks, if any, are not known.

In summary, finasteride therapy in men with benign prostatic hyperplasia led to sustained decreases in serum dihydrotestosterone concentrations, followed by decreases in serum prostate-specific—antigen concentrations and prostatic volume. These changes were accompanied by decreases in the number or severity of symptoms of urinary tract obstruction and increases in the urinary-flow rate. The effect of a daily dose of 5 mg of finasteride was greater than that of a 1-mg dose. We conclude that treatment with a 5-mg dose of finasteride is beneficial in men with benign prostatic hyperplasia.

The members of the Finasteride Study Group are as follows: G.L. Andriole, Washington University, St. Louis; R. Boake, University of Alberta, Edmonton; B.R. Bracken, University of Cincinnati Medical Center, Cincinnati; W. Brannan, Ochsner Clinic, New Orleans; R.C. Bruskewitz, University of Wisconsin, Madison; C.E. Cox II, University of Tennessee, Memphis; G.R. Cunningham, Veterans Affairs Medical Center and Baylor College of Medicine, Houston; P.C. Devine and P.F. Schellhammer, Eastern Virginia Medical School, Norfolk; M. Elhilali, Royal Victoria Hospital, Montreal; J. Geller, Mercy Hospital and Medical Center, San Diego, Calif; J.T. Grayhack, Northwestern University Medical School, Chicago; L.H. Harrison, Bowman Gray School of Medicine, Winston-Salem, N.C.; P.B. Hudson, Veteran Affairs Medical Center, Bay Pines, Fla.; J. Imperato and E.D. Vaughan, Jr., New York Hospital, New York; M.M. Lieber, Mayo Clinic, Rochester, Minn.; M.C. Lippert, University of Virginia, Charlottesville; G.H. Malek, Jackson Clinic and Foundation, Madison, Wis.; J.D. McConnell, University of Texas, Dallas; P. Narayan, University of California, San Francisco; J.P. Perreault, Hospital St.-Luc, Montreal; M.I. Resnick, Case Western Reserve University School of Medicine, Cleveland; R. Norman and R. Rittmaster, Dalhousie University, Halifax, N.S.; N.A. Romas and W. Rosner, St. Luke's–Roosevelt Hospital Center, New York; S. Rosenblatt, University of California, Newport Beach; J.B. Roy, University of Oklahoma Health Sciences Center, Oklahoma City; E.J. Seidmon, Temple University Hospital, Philadelphia; J.S. Tenover, Harborview Medical Center, Seattle; J. Trachtenberg, Toronto General Hospital, Toronto; P.C. Walsh, Johns Hopkins Hospital, Baltimore; R.D. Williams, University of Iowa Hospital and Clinic, Iowa City; and G.J. Gormley, E. Stoner, F. Pappas, A.M. Taylor, B. Binkowitz, and J. Ng, Merck Research Laboratories, Rahway, N.J.