SEARCH

SEARCH BY CITATION

Keywords:

  • 5-α-reductase inhibitor;
  • prostate cancer;
  • prevention;
  • systematic review

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. MEN CONSIDERED AT INCREASED RISK FOR PROSTATE CANCER: THE REDUCE TRIAL ()
  7. DISCUSSION
  8. ACKNOWLEDGEMENTS
  9. CONFLICT OF INTEREST
  10. REFERENCES

OBJECTIVE

  • • 
    To estimate the benefits and harms of 5-α-reductase inhibitors (5-α-RIs) in preventing prostate cancer.

MATERIALS AND METHODS

  • • 
    We searched MEDLINE and the Cochrane Collaboration Library through June 2010 to identify randomized trials.
  • • 
    We included articles if they examined 5-α-RI vs control, were ≥1 year in duration and provided clinical outcomes.
  • • 
    Our primary outcome was prostate cancer period-prevalence ‘for-cause’.

RESULTS

  • • 
    Eight studies met inclusion criteria but only the Prostate Cancer Prevention Trial and the Reduction by Dutasteride of Prostate Cancer Events were designed to assess the impact of 5-α-RIs on prostate cancer period-prevalence. The mean age of enrolees was 64 years, 92% were White, and mean PSA level was 3.1 ng/mL. For-cause prostate cancers comprised 54% of all cancers detected in placebo-controlled studies.
  • • 
    Compared with placebo, 5-α-RI resulted in a 25% relative risk (RR) reduction in prostate cancers detected for-cause [RR 0.75, 95% confidence interval (CI) 0.67–0.83; 1.4% absolute risk reduction (3.5% vs 4.9%)]. One BPH trial reported that the risk of prostate cancers detected for-cause was significantly reduced with dutasteride and combined dutasteride plus tamsulosin compared with tamsulosin monotherapy.
  • • 
    Six trials vs placebo assessed prostate cancers detected overall. There was a 26% RR reduction favouring 5-α-RI [RR 0.74, 95% CI 0.55–1.00; 2.9% absolute risk reduction (6.3% vs 9.2%)]. There were reductions across categories of age, race and family history of prostate cancer.
  • • 
    One placebo-controlled trial of men that investigators considered at greater risk for prostate cancer (based on age, elevated PSA level and having a previous suspicion of prostate cancer leading to a prostate biopsy) reported that dutasteride did not reduce prostate cancers detected for-cause based on needle-biopsy but did reduce risk of overall incident prostate cancer detected by biopsy by 23%[RR 0.77, 95% CI 0.70–0.85; absolute reduction 16.1% vs 20.8%]. There were reductions across age, family history of prostate cancer, PSA level, and prostate volume subgroups.
  • • 
    Incidences of erectile dysfunction, ejaculate volume, decreased libido, and gynaecomastia were greater with 5-α-RI vs placebo.

CONCLUSIONS

  • • 
    5-α-RIs reduce the risk of being diagnosed with prostate cancer among men who are screened regularly for prostate cancer.
  • • 
    Information is inadequate to assess the effect of 5-α-RIs on prostate cancer or all-cause mortality.
  • • 
    5-α-RIs increase sexual and erectile dysfunction.

Abbreviations
CombAT

the Combination of Avodart and Tamsulosin (study)

PCPT

Prostate Cancer Prevention Trial

PLESS

Proscar Long-term Efficacy and Safety Study

PROSPECT

PROscar Safety Plus Efficacy Canadian Two year Study

REDUCE

the Reduction by Dutasteride of Prostate Cancer Events (trial)

5-α-RI

5α-Reductase inhibitors

RR

relative risk.

INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. MEN CONSIDERED AT INCREASED RISK FOR PROSTATE CANCER: THE REDUCE TRIAL ()
  7. DISCUSSION
  8. ACKNOWLEDGEMENTS
  9. CONFLICT OF INTEREST
  10. REFERENCES

Safe and effective methods to prevent or delay prostate cancer development are needed. Interventions that alter circulating androgen levels or inhibit 5-α-reductase have potential as chemopreventive agents because testosterone, after conversion to 5-α-dihydrotestosterone, controls prostate mitotic activity and potentially cancer development. Two 5-α-reductase inhibitors (5-α-RIs) exist. Finasteride (Proscar) is selective for the type 2 isoenzyme and dutasteride (Avodart) inhibits both type 1 and type 2 isoenzymes. These agents are used to treat LUTS associated with BPH and, in the case of finasteride (in lower doses), male androgenic alopecia. Their role as chemopreventive agents has resulted in randomized trials designed to assess their impact on the period prevalence of biopsy-confirmed prostate cancer and reports on prostate cancer period-prevalence from long-term BPH treatment trials [1,2]. We updated a previous systematic review [3] to address the benefits and harms of 5-α-RIs for prostate cancer chemoprevention.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. MEN CONSIDERED AT INCREASED RISK FOR PROSTATE CANCER: THE REDUCE TRIAL ()
  7. DISCUSSION
  8. ACKNOWLEDGEMENTS
  9. CONFLICT OF INTEREST
  10. REFERENCES

LITERATURE SEARCH STRATEGY AND INCLUSION CRITERIA

We searched MEDLINE and the Cochrane Library through June 2010 using the MeSH (Medical Subject Headings) terms and text words: ‘finasteride’, ‘dutasteride’, ‘prostatic neoplasms’, ‘azasteroids’, ‘reductase inhibitors’ and ‘enzyme inhibitors’. We searched reviews, reference lists, and personal files and received additional unpublished information via communication with study sponsors. We included randomized trials that evaluated a 5-α-RI vs a control, lasted ≥1 year(s) and reported prostate cancer period prevalence.

SUMMARY OF OUTCOMES ASSESSED

Our primary outcome was period-prevalence, the proportion of randomized men found to have prostate cancer detected ‘for-cause’ over the trial period. We defined for-cause cancers as those that:

  • 1
    were suspected clinically because of symptoms, abnormal DRE, or abnormal PSA level, and were confirmed on biopsy
  • 2
    a recommendation was made for biopsy per the study protocol (e.g. due to increasing PSA levels) that was never done, and end-of-study biopsy showed prostate cancer
  • 3
    end-of-study biopsy in the setting of a PSA level of >4 ng/mL and/or suspicious DRE showed prostate cancer.

We assessed prostate cancer stage distribution, Gleason scores, and prostate cancer period-prevalence by age, race, baseline PSA level, and family history categories. Secondary outcomes included overall prostate cancer detected, all-cause mortality, prostate cancer-specific mortality and adverse events.

STATISTICAL ANALYSIS

We categorized trials as long (>2 years) or mid-term (1–2 years) duration. Analyses were performed using RevMan 4.2 software [4]. Relative risks (RRs) and absolute risk differences with 95% CIs were calculated for categorical outcomes using a random effects model. Heterogeneity between studies was assessed using the chi-square test for heterogeneity (present if P < 0.1) and the I2 test (present if >50%) [5]. Individual trial outcomes were assessed, with emphasis on the Prostate Cancer Prevention Trial (PCPT) and the Reduction by Dutasteride of Prostate Cancer Events trail (REDUCE) because they were designed to assess whether 5-α-RIs prevented or delayed prostate cancer.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. MEN CONSIDERED AT INCREASED RISK FOR PROSTATE CANCER: THE REDUCE TRIAL ()
  7. DISCUSSION
  8. ACKNOWLEDGEMENTS
  9. CONFLICT OF INTEREST
  10. REFERENCES

We identified 728 MEDLINE and 315 Cochrane Library abstracts. Abstract screening excluded 919 abstracts that failed to meet criteria. The remaining 124 articles were reviewed and 12 manuscripts representing eight studies met final inclusion criteria [1,2,6–15].

STUDY QUALITY

Four trials had adequate treatment allocation concealment [2,6,9,14]. Seven studies indicated subjects and investigators were ‘blinded’ to treatment assignment [2,6,7,9,11,13,14]. Five studies recorded using an intention-to-treat analysis [6,7,9,11,14]. The primary intention-to-treat analysis for the PCPT included men receiving a diagnosis of prostate cancer during the study or men who underwent an end-of-study biopsy [1]. For the REDUCE trial, the efficacy population included men with a negative baseline prostate biopsy who received at least one dose of the study medication [2]. There was adequate description of study withdrawals and discontinuations by treatment group in all studies [1,2,6,7,9,11].

TRIAL CHARACTERISTICS (TABLE 1)

Table 1.  Characteristics for 5-α-RI studies
Study (reference)/locationNumber randomized/Interventions dose/dayBaseline demographic characteristicsStudyduration, yearsProstate cancerdiagnosis for cause
  1. PIN, prostatic intraepithelial neoplasia.

Long-term trials (>2 years)    
REDUCE 2010 [2]N= 8 231Mean age 63 years4Yes
MultinationalDutasteride 0.5 mgRace: White 91%, Black 2%, Hispanic 4%, Asian 2%, Other 1%.  
PlaceboMean PSA level 5.9 ng/mL  
 Mean prostate volume 45.7 mL  
 Family history of prostate cancer (first-degree relative) 13%  
 Mean baseline AUA/IPSS score 8.7  
CombAT 2010 [14,15]N= 4 844Mean age 66 years4Yes
MultinationalDutasteride 0.5 mgRace: White 88%  
Combined dutasteride 0.5 mg and tamsulosin 0.4 mgMean PSA level 4 ng/mL  
Mean prostate volume 55 mL  
Mean baseline AUA/IPSS score 16.5  
Tamsulosin 0.4 mg   
PCPT 2003 [1]N= 18 882Mean age 63 years7Yes
USAFinasteride 5 mgAge, 45–64 years 62%; ≥65 years 38%  
PlaceboRace: White 92%, Black 3.8%, Hispanic 2.6%, Other 1.5%  
 PSA level 1.2 ng/mL  
 Family history of prostate cancer (first degree relative) 15.4%  
 Mean (Sd)baseline AUA/IPSS score 6.7 (4.8)  
PLESS 1998 [6]N= 3 040Race: White 96%.4Yes
USAFinasteride 5 mgMean PSA level 2.8 ng/mL  
PlaceboMean prostate volume 54.5 mL  
 Mean baseline AUA/IPSS score 15  
Mid-term trials (1–2 years)    
ARIA 3001–3003N= 4 325Mean age 66 years2Yes
2002 [7,8]Dutasteride 0.5 mgRace: White 92%, Black 4%, Hispanic 7.3%,  
Multinational Asian 1%, Other 1.5%.  
PlaceboMean PSA level 4.0 ng/mL  
 Mean prostate volume 54.5 mL  
 Mean baseline AUA/IPSS score 17.1  
PROSPECT [9]N= 613Mean age 63 years2Yes
CanadaFinasteride 5 mgRace not reported.  
PlaceboMean PSA level not reported.  
 Mean prostate volume 44.9 mL  
 Mean baseline (quasi) AUA/IPSS score 16.2  
Cote 1998 [10]N= 58Mean age 68 years1No
USAFinasteride 5 mgRace not reported.  
Observation (watchful waiting)Mean PSA level 9.8 ng/mL (all >4.0 ng/mL)  
Mean baseline AUA/IPSS score not reported.  
Study objective was to examine effect of finasteride on prostate cellular proliferation and high-grade PIN.  
Pre-existing high-grade PIN: observation 5 men; finasteride 8 men.  
Finasteride Study Group [11–13]N= 1 645Mean age 65 years1Yes
USA and multinationalFinasteride 1 mg or 5 mgRace: White 95.8%, Black 1.6%, Other 2.6%.  
PlaceboMean PSA level 4.7 ng/mL  
 Mean prostate volume 55.0 mL  
 Mean baseline AUA/IPSS score not reported  

Eight studies, randomizing 41 638 men, provided information related to prostate cancer period-prevalence [1,2,6,8–10,12,14]. Two reports that assessed the effectiveness and safety of dutasteride included a pooled analysis of three trials [7,8] and the Finasteride Study Group report consisted of two trials [11,13]. Study duration ranged from 1 to 7 years, with four trials enrolling a total of 34 997 (84%) men lasting ≥4 years [1,2,6,14]. The PCPT and REDUCE assessed the impact of a 5-α-RI on prostate cancer period-prevalence as a primary endpoint [1,2]. Five studies assessed the effect of 5-α-RIs on BPH [6,7,9,11,13,14].

Overall, the subject mean age was 64 years (range of means 63–68). Enrollees were mostly White (92%; five studies) [1,2,6,7,11]. The mean (range) baseline PSA level was 3.1 (1.2–9.8) ng/mL (seven studies) [1,2,6,7,10,11,13,14] and mean (range) prostate volume was 51.2 (45.7–55.0) mL (six studies) [2,6,7,9,11,13,14]. The mean baseline PSA level was 1.2 ng/mL for men in the PCPT and 5.9 ng/mL for men in the REDUCE trial [1,2]. About 15% of subjects in the PCPT and REDUCE trials reported a family history of prostate cancer [1,2]. The mean IPSS or AUA Symptom score in the BPH trials was 16, indicating moderate LUTS [6,7,14].

PROSTATE CANCER PERIOD PREVALENCE

Five studies provided data for ‘prostate cancer detected for-cause among all subjects randomized.’ 5-α-RI therapy reduced for-cause cancers by 25% in relative terms (RR 0.75, 95% CI 0.67–0.83) and 1.4% in absolute terms [1,6,8,9,12] (Table 2). All but the Finasteride Study Group trial favoured 5-α-RI. For-cause prostate cancers comprised 54% (1188/2211) of all cancers with the PCPT contributing 85% of ‘for cause’ cancers. In the the Combination of Avodart and Tamsulosin (CombAT) BPH trial, prostate cancer was reported in 42 men (2.6%) receiving dutasteride vs 37 (2.3%) in the combination dutasteride group and 63 (3.9%) in the tamsulosin group [14]. Both dutasteride monotherapy and combined dutasteride plus tamsulosin reduced ‘for-cause cancers’ compared with tamsulosin, with RRs of 0.66 (95% CI 0.45–0.97) and 0.59 (95% CI 0.39–0.88), respectively.

Table 2.  Prostate cancer period prevalence
StudyProstate cancer detected ‘for cause’Overall prostate cancer period prevalence
5-α-RI, n/NPlacebo, n/NRR (95% CI)5-α-RI, n/NPlacebo, n/NRR (95% CI)
Finasteride, mid-term treatment duration (1–2 years)
 Cote 1998 [10]   8/27   1/257.41 (1.00–55.09)
 PROSPECT 1996 [9]  3/310  6/3030.49 (0.12–1.94)  3/310   6/3030.49 (0.12–1.94)
 FSG 1992–93 [12]  7/1 090  3/5551.19 (0.31–4.58)  8/1 090   4/5551.02 (0.31–3.37)
 Subtotal 10/1 400  9/8580.74 (0.30–1.82) 19/1 427  11/8831.29 (0.33–5.02)
Finasteride, long-term treatment duration (>2 years)
 PCPT 2003 [1]435/9 423571/9 4590.76 (0.68–0.86)803/9 4231147/9 4590.70 (0.65–0.77)
 PLESS 1998 [6] 36/1 524 45/1 5160.80 (0.52–1.23) 72/1 524  77/1 5160.93 (0.68–1.27)
 Subtotal471/10 947616/10 9750.77 (0.68–0.86)875/10 9471224/10 9750.78 (0.60–1.01)
Dustasteride, mid-term treatment duration (1–2 years)
 ARIA 2002 [8] 27/2 167 55/2 1580.49 (0.31–0.77) 27/2 167  55/2 1580.49 (0.31–0.77)
 Total508/14 541680/14 0160.75 (0.67–0.83)921/14 5411290/14 0160.74 (0.55–1.00)
Dutasteride for men at increased risk for prostate cancer, long-term treatment duration (>2 years)
 Reduce 2010 [2] 63/4 105 86/4 1260.74 (0.53–1.02)659/4 105 858/4 1260.77 (0.70–0.85)
Dutasteride vs combined dutasteride and tamsulosin, long-term treatment duration (>2 years)
 CombAT 2010 [14] 42/1 623 37/1 6101.13 (0.73–1.74) 42/1 623  37/1 6101.13 (0.73–1.74)
Dutasteride vs doxazosin, long-term treatment duration (>2 years)
 CombAT 2010 [14] 42/1 623 63/1 6110.66 (0.45–0.97) 42/1 623  63/1 6110.66 (0.45–0.97)
Combined dutasteride and tamsulosin vs. tamsulosin, long-term treatment duration (>2 years)
 CombAT 2010 [14] 37/1 610 63/1 6110.59 (0.39–0.88) 37/1 610  63/1 6110.59 (0.39–0.88)

Overall prostate cancer period prevalence was obtained from three long and mid-term studies [1,2,6,9,10,12] (Table 2). There was a 26% relative reduction of borderline statistical significance in overall prostate cancers and a 2.9% absolute reduction (6.3% vs 9.2%). There was evidence of moderate statistical heterogeneity (P= 0.04; I2= 56.0%). Heterogeneity was primarily based on one very small mid-term trial of men with PSA levels of >4 ng/mL that included some men with baseline prostate biopsies positive for prostatic intraepithelial neoplasia [10]. When this trial was excluded, the overall results became statistically significant (RR 0.72, 95% CI 0.58–0.88). ARIA (ARIAD Pharmaceuticals Inc.), the only report evaluating dutasteride, showed a 51% relative reduction in prostate cancers detected with an absolute reduction of 1.3% (1.2% vs 2.5%) [8]. The PCPT provided 88% of overall prostate cancers [1].

SUBGROUP ANALYSIS (TABLE 3)

Table 3.  Period prevalence of prostate cancer according to subgroup
Subgroup/reference5-α-RI, n/N (%)Placebo, n/N (%)RR (95% CI)
  1. PCa, prostate cancer.

PSA level at study entry 0.0 to <4.0 ng/mL   
 PCPT [1]803/9423 (8.5) 1147/9456 (12.1)0.70 (0.64–0.77)
 Finasteride Long-term Efficacy and 28/1149 (2.4)  32/1154 (2.8)0.88 (0.53–1.45)
 Safety Study Group (PLESS) [5]   
  Total831/10 572 (7.9) 1179/10 610 (11.1)0.7 (0.65–0.77)
PSA level at study entry ≥4.0 ng/mL   
 Cote [10]  8/27 (29.6)  1/25 (4.0)7.41 (1.00–55.09)
 PLESS [6] 44/374 (11.8) 45/357 (12.6)0.93 (0.63–1.38)
  Total 52/401 (13.0) 46/382 (12.0)2.08 (0.28–15.43)
Gleason score 7   
 PCPT [1]190/9423 (2.0)184/9459 (1.9)1.04 (0.85–1.27)
 PLESS [6] (estimated from graph)  11/1524 (0.72) 12/1516 (0.79)0.91 (0.40–2.06)
  Total201/10 947 (1.8)196/10 975 (1.9)1.03 (0.85–1.25)
PCPT: Gleason score 7 [1]   
 PCa overall/number randomized190/9423 (2.0)184/9459 (1.9)1.04 (0.85–1.27)
 PCa overall/included in analysis190/4368 (4.3)184/4692 (3.9)1.11 (0.91–1.35)
 PCa detected for cause/number randomized 118/9423 (12.5)103/9459 (10.9)1.15 (0.88–1.50)
 PCa detected for cause/biopsy performed for cause 118/1639 (7.2)103/1934 (5.3)1.35 (1.05–1.75)
Gleason scores 8–10   
 PCPT [1] 90/9423 (0.96) 53/9459 (0.56)1.70 (1.22–2.39)
 PLESS [6] (estimated from graph)  1/1524 (0.07)  7/1516 (0.46)0.14 (0.02–1.15)
PCPT: Gleason scores 8–10 [1]   
 PCa overall/number randomized 90/9423 (0.96)53/9459 (0.56)1.70 (1.22–2.39)
 PCa overall/included in analysis 90/4368 (2.1)53/4692 (1.1)1.82 (1.30–2.55)
 PCa detected for cause/number randomized 70/9423 (0.74)45/9459 (0.48)1.56 (1.07–2.27)
 PCa detected for cause/biopsy performed for cause 70/1639 (4.3)45/1934 (2.3)1.84 (1.27–2.65)
PCPT, additional subgroups [1]   
 Age, years:   
  55–59205/2954 (6.9)309/2954 (10.5)066 (0.56–0.79)
  60–64254/2970 (8.6)357/2825 (12.6)0.68 (0.58–0.79)
  ≥65344/3498 (9.8)481/3677 (13.1)0.75 (0.66–0.86)
 Race:   
  Non-Hispanic White739/8667 (8.5)1067/8713 (12.2)0.70 (0.64–0.76)
  African-American 41/356 (11.5)  50/353 (14.2)0.81 (0.55–1.20)
  Hispanic 19/262 (7.3)  23/237 (9.7)0.75 (0.42–1.34)
 Family history of PCa in first-degree relative:   
  Yes176/1458 (12.1)241/1455 (16.6)0.73 (0.61–0.87)
  No627/7965 (7.9)906/8002 (11.3)0.70 (0.63–0.77)
REDUCE subgroups [2]   
 Age, years:   
  <65342/1953 (17.5)462/2053 (22.5)0.76 (0.67–0.87)
  ≥65317/1352 (23.4)396/1371 (28.9)0.78 (0.68–0.89)
 Baseline PSA levels, ng/mL:   
  <4.9194/1095 (17.7)259/1173 (22.1)0.77 (0.65–0.92)
  4.9 to <6.8239/1121 (21.3)308/1138 (27.1)0.77 (0.65–0.89)
  ≥6.8225/1086 (20.7)290/1105 (26.2)0.77 (0.65–0.90)
 Prostate volume, mL:   
  <36.6268/1065 (25.2)350/1121 (31.2)0.79 (0.69–0.92)
  36.6 to <51.8214/1093 (19.6)250/1133 (22.1)0.84 (0.71–0.96)
  ≥51.8169/1097 (15.4)244/1112 (21.9)0.68 (0.57–0.82)
 Family history of PCa:   
  Yes105/448 (23.4)141/437 (32.3)0.69 (0.54–0.88)
  No554/2853 (19.4)717/2987 (24.0)0.78 (0.71–0.87)

Three trials [1,6,10] provided data according to baseline PSA levels and two trials [1,6] reported histological grade. Only PCPT provided information related to race, age at entry and family prostate cancer history [1]. Some subgroup results are available only on prostate cancer detected overall rather than prostate cancer detected for-cause. Risk reduction occurred among men with baseline PSA levels of <4.0 ng/mL (n= 21 982; RR 0.70, 95% CI 0.64–0.77). Among men who had baseline PSA levels of ≥4 ng/mL (n= 783), there was no reduction although fewer than 100 prostate cancers were diagnosed and CIs were very wide (RR 2.08, 95% CI 0.28–15.43) [6,10]. All men with baseline PSA levels of ≥4 ng/mL had a prostate biopsy at entry that was negative for prostate cancer.

In the PCPT there were slightly, and statistically significantly, more high Gleason score tumours (7–10 or 8–10) in men on finasteride but not in the smaller and shorter-term Proscar Long-term Efficacy and Safety Study (PLESS) in which scores of 8–10 were slightly, but not statistically significantly greater in the placebo arm [1,6]. Although Gleason score distribution in PLESS was reported ‘to be similar’ between finasteride and placebo, ≈5% of all cancers were Gleason 8–10 with seven occurring in the placebo-treated group and one in the finasteride group (P= 0.07) [6].

In the PCPT, Gleason 7–10 cancers comprised 22% of all cancers detected in the placebo group compared with 37% in the finasteride group [1]. The cumulative percentages for Gleason 7–10 cancers diagnosed in biopsies performed for-cause were 29.4% and 47.8% for placebo and finasteride groups, respectively. Gleason 8–10 tumours were identified in 90 men in the finasteride group compared with 53 men in the placebo arm (RR 1.70, 95% CI 1.22–2.39) [1]. In the CombAT trial, relative reductions vs placebo in both low and high Gleason cancers were similar to the overall reduction (40%) [15]. Among the 142 patients developing prostate cancer there were 12, 22, and 27 Gleason 7–10 tumours over the 4 years study period in the combination, dutasteride, and tamsulosin groups, respectively.

Reduction in overall prostate cancers detected due to finasteride was consistent regardless of age, race or family history in the PCPT [1]. Fewer than 10% of men were of non-White race; CIs were wide and risk reductions were not statistically significant. Both the relative and absolute reductions due to finasteride in detected prostate cancers were similar among men with or without a first-degree family history of prostate cancer.

MORTALITY

No study was designed to assess the impact of 5-α-RI on mortality. All-cause as well as prostate cancer-specific mortality was low, at 5.6% and 0.05%, respectively. There were no differences in all-cause or prostate cancer mortality between finasteride and placebo in any trial (RR all-cause mortality 1.05, 95% CI 0.94–1.18).

MEN CONSIDERED AT INCREASED RISK FOR PROSTATE CANCER: THE REDUCE TRIAL (TABLE 2)

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. MEN CONSIDERED AT INCREASED RISK FOR PROSTATE CANCER: THE REDUCE TRIAL ()
  7. DISCUSSION
  8. ACKNOWLEDGEMENTS
  9. CONFLICT OF INTEREST
  10. REFERENCES

One trial (N= 6729) assessed the effect of a 5-α-RI among men considered at increased risk for prostate cancer based primarily on age and PSA values [2]. Specifically men were eligible if they were aged 50–75 years, had PSA level of 2.5–10 ng/mL (for men aged 50–60 years) or 3.0–10 ng/mL (for men aged >60 years), and had a previous suspicion for prostate cancer leading to a prostate biopsy within 6 months of study enrolment.

Dustasteride did not statistically significantly reduce the incidence of ‘prostate cancer detected for-cause’ based only on needle-biopsies over the 4-year trial period. Incidence in the dutasteride group was 1.5% vs 2.1% in the placebo arm (RR 0.74, 95% CI 0.53–1.02; personal communication, Roger Rittmaster, August 4, 2010). Dustasteride did reduce the risk of overall incident prostate cancer (RR 0.77, 95% CI 0.70–0.85) among men who underwent a biopsy or prostate surgery. About 16% of men in the dutasteride group were diagnosed with prostate cancer vs 21% of men in the placebo group, absolute risk reduction of 5% (95% CI 3% to 6%). There was risk reduction with dutasteride across all pre-specified baseline subgroups including age (<65 or ≥65 years), family history of prostate cancer, prostate volume, and PSA level (Table 3). In contrast to the PCPT, there was no difference in the overall number of cancers with Gleason scores between 7 and 10, 220 (6.7%) in the dutasteride arm vs 233 (6.8%) in the placebo arm (P= 0.81). There were 29 cancers classified Gleason 8–10 in the dutasteride group vs 19 in the placebo group (P= 0.15). However, during the third and fourth years there were more tumours with Gleason scores between 8 and 10 in the dutasteride group (n= 12) compared to only one in the placebo group (P= 0.003). There was no difference in overall mortality though few men died over the course of the trial (1.7% of men randomized to dutasteride vs 1.9% of men randomized to placebo, P= 0.65). None of the deaths were attributed to prostate cancer.

STUDY DISCONTINUATIONS AND ADVERSE EFFECTS (TABLE 3)

Across trials, study discontinuations and/or lost to follow-up did not differ among men assigned to placebo or 5-α-RI. However, both the PCPT and the REDUCE trial excluded individuals who did not comply during a placebo-based ‘run-in’ phase and those randomized who did not take at least one dose of study medication or had a positive baseline prostate biopsy. Side-effects of finasteride represented the primary reason for the difference in the proportion of men who temporarily discontinued treatment in the PCPT: 18% in finasteride vs 10% in placebo [1]. In the REDUCE trial, more subjects in the dutasteride group permanently discontinued treatment compared with placebo, 4% vs 2% (P < 0.001) [2]. Sexual or erectile dysfunction or endocrine effects were more common with 5-α-RI than placebo. Incidence of erectile dysfunction was greater for 5-α-RI compared with placebo in the PCPT and the REDUCE trial, with absolute risk differences of 6% and 3%, respectively [1,2]. Absolute risk differences for four mid-term trials ranged from 0% to 10%[7,9,11,13]. Decreased libido was reported with greater frequency by men in the 5-α-RI groups in four of five studies (absolute risk differences ranging from 2% to 6%) [1,2,7,9,11]. In three large studies, incidence of gynaecomastia was greater for men allocated to 5-α-RI vs placebo, with absolute differences ranging from 1% to 2%[1,2,6]. Decreased ejaculate volume was greater in the 5-α-RI groups compared with placebo in all studies reporting this adverse event [1,2,6,11]. Absolute risk differences generally ranged from 1% to 3% but the PCPT reported 13%. The degree to which a particular side-effect led to withdrawal is not known. In the REDUCE trial, incidence of a composite category of ‘cardiac failure’, defined as congestive heart failure, cardiac failure, acute cardiac failure, ventricular failure, cardiopulmonary failure, or congestive cardiomyopathy was greater in the dutasteride group (0.7%, n= 30) compared with the placebo group (0.4%, n= 16; P= 0.03). There was no difference between groups in deaths from cardiovascular events.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. MEN CONSIDERED AT INCREASED RISK FOR PROSTATE CANCER: THE REDUCE TRIAL ()
  7. DISCUSSION
  8. ACKNOWLEDGEMENTS
  9. CONFLICT OF INTEREST
  10. REFERENCES

5-α-RIs reduced the absolute period-prevalence of ‘for-cause’ detected prostate cancer in all men randomized by ≈1%. This reduction occurred regardless of race, family history or age. 5-α-RIs were generally well tolerated among trial participants for up to 7 years with overall discontinuations and discontinuations due to adverse events similar to placebo. However, adverse events with 5-α-RI were increased, particularly those related to sexual and erectile dysfunction.

The findings and clinical interpretations have limitations:

  • 1
    Only the PCPT and the REDUCE trial were designed to assess the impact of 5-α-RIs on prostate cancer period-prevalence [1,2]. However, data from long-term BPH treatment trials that provided prostate cancer detection information are consistent with these findings.
  • 2
    Trial authors’ definitions of ‘for-cause’ included some cancers detected due to study protocol directed biopsies. Our primary outcome in this review emphasised prostate cancers detected for-cause among all men randomized. We think that this approach represents outcomes occurring in clinical settings, reduces bias by preserving the randomization process and focuses on the clinically most relevant prostate cancers. Additionally, it decreases selection bias that could occur in the situation in which only enrolees receiving end-of-study evaluation were analysed. Our ‘for cause cancers among patients randomized’ outcome, applied to the PCPT data, resulted in a RR reduction of 26% and an absolute reduction of 1.4%. This contrasts to the 10% RR reduction and 3% absolute reduction if limiting the definition to patients undergoing a biopsy for cause or a 25% relative and 6% absolute reduction in period prevalence if one includes patients with a PSA elevation even though some of them did not have the biopsy.
  • 3
    Men in all of the trials were screened regularly for prostate cancer with PSA level measurement and DRE. We were unable to assess the impact of 5-α-RIs on prostate cancer among men not actively screened. Regular screening for PSA levels approximately doubles the incidence of a diagnosis of prostate cancer. The RR reduction of being diagnosed with prostate cancer of ≈26% must be interpreted in this context. There was a net increase in prostate cancer of nearly 50% for the combined strategy of screening plus chemoprevention. Results from randomized trials of prostate cancer screening have resulted in uncertainty about the net benefit of prostate cancer screening with PSA [16].
  • 4
    Findings probably overestimate the number of cancers detected compared with clinical settings even where frequent PSA testing and follow-up prostate biopsies occur.
  • 5
    Studies do not answer whether reported risk reductions resulted from shrinkage or growth inhibition of existing tumours rather than cancer prevention.
  • 6
    There is still uncertainty regarding the impact of 5-α-RIs on tumours of potentially greatest lethal potential including those with Gleason score 8–10. While PLESS reported a reduction in high-grade tumours with finasteride, the PCPT and the REDUCE trail (between the third and fourth years) reported the opposite effect [1,2,6]. In the REDUCE trial, there were more tumours with Gleason scores between 8 and 10 in the dutasteride group compared with the placebo group during the third and fourth years. However, a large number of cancers had been removed from the placebo group at the 2-year biopsy and therefore were not available for subsequent biopsy and potential upgrading at the 4-year biopsy. Controversy continues as to the cause of the increase in high-grade tumours seen in the PCPT. A panel of pathologists evaluating high-grade tumour biopsies as well as radical prostatectomy specimens from the PCPT concluded that finasteride effects on tumour morphology did not explain the increase in high-grade cancer [17]. Reports have discussed the impact of finasteride on prostate volume and sampling and the potential role this may play in a changed sensitivity for prostate cancer detection and detection of high-grade disease [17–19].
  • 7
    Finally, evidence is insufficient regarding age to initiate treatment or the optimal duration of chemoprevention. The mean age of enrolees ranged from 63 to 68 years with none aged <55 years. The longest duration of treatment and follow-up was 7 years. A >80% reduction in PSA levels due to 5-α-RIs occurs within 12 weeks of initiation and PSA values return to baseline within a similar period after discontinuation. Whether this is directly associated with initiation and/or loss of prostate cancer prevention activity is not known. Long-term cancer prevention might be of greatest benefit in younger men, although this must be balanced with adverse effects on sexual function. Because the presence and severity of BPH symptoms are less common in younger men, the potential advantage for relieving bothersome symptoms or preventing BPH progression may be of less importance. Potential candidates for chemoprevention include men who express interest in chemoprevention and regular prostate cancer screening.

Future research is needed to determine whether 5-α-RIs reduce all-cause and disease-specific morbidity and mortality; appropriate ‘PSA adjustments’ and indications for prostate biopsy in men on 5-α-RIs and whether the effect of 1 mg finasteride for androgenic alopecia on serum PSA levels is associated with a reduction in prostate cancer period-prevalence [20].

In conclusion, 5-α-RIs reduce prostate cancer risk in men who undergo regular prostate cancer screening using PSA level measurements and DRE with follow-up prostate biopsies. The impact of 5-α-RIs on absolute or relative rates of prostate cancer in men who are not being regularly screened is not clear. Effects are consistent across race, family history and age but were limited to men with baseline PSA values of <4.0 ng/mL. Information is inadequate to assess the impact on mortality. Long-term compliance in randomized trials is high. Decisions to initiate chemoprevention with 5-α-RIs should incorporate this information.

ACKNOWLEDGEMENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. MEN CONSIDERED AT INCREASED RISK FOR PROSTATE CANCER: THE REDUCE TRIAL ()
  7. DISCUSSION
  8. ACKNOWLEDGEMENTS
  9. CONFLICT OF INTEREST
  10. REFERENCES

Supported in part by a contract through the American Urological Association and the American Society for Clinical Oncology.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. MEN CONSIDERED AT INCREASED RISK FOR PROSTATE CANCER: THE REDUCE TRIAL ()
  7. DISCUSSION
  8. ACKNOWLEDGEMENTS
  9. CONFLICT OF INTEREST
  10. REFERENCES
  • 1
    Thompson IM, Goodman PJ, Tangen CM et al. The influence of finasteride on the development of prostate cancer. N Engl J Med 2003; 349: 21524
  • 2
    Andriole GL, Bostwick DG, Brawley OW et al. REDUCE Study Group Effect of dutasteride on the risk of prostate cancer. N Engl J Med 2010; 362: 1192202
  • 3
    Wilt TJ, MacDonald R, Hagerty K, Schellhammer P, Kramer BS. Five-alpha-reductase inhibitors for prostate cancer prevention. Cochrane Database Syst Rev 2008; 2: CD007091
  • 4
    Anonymous. Review Manager [Computer Program], Version 4.1. For Windows. Oxford: The Cochrane Collaboration, 2001
  • 5
    Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003; 327: 55760
  • 6
    Andriole GL, Guess HA, Epstein JI et al. Treatment with finasteride preserves usefulness of prostate-specific antigen in the detection of prostate cancer: results of a randomized, double-blind, placebo-controlled clinical trial. PLESS Study Group. Proscar Long-term Efficacy and Safety Study. Urology 1998; 52: 19502
  • 7
    Roehrborn CG, Boyle P, Nickel JC, Hoefner K, Andriole G. ARIA3001 ARIA3002 and ARIA3003 Study Investigators. Efficacy and safety of a dual inhibitor of 5-alpha-reductase types 1 and 2 (dutasteride) in men with benign prostatic hyperplasia. Urology 2002; 60: 43441
  • 8
    Andriole GL, Roehrborn C, Schulman KM, Slawin KM, Somerville M, Rittmaster RS. Effect of dutasteride on the detection of prostate cancer in men with benign prostatic hyperplasia. Urology 2004; 64: 53743
  • 9
    Nickel JC, Fradet Y, Boake RC et al. Efficacy and safety of finasteride therapy for benign prostatic hyperplasia: results of a 2-year randomized controlled trial (the PROSPECT study). PROscar Safety Plus Efficacy Canadian Two year Study. CMAJ 1996; 155: 12519
  • 10
    Cote RJ, Skinner EC, Salem CE et al. The effect of finasteride on the prostate gland in men with elevated serum prostate-specific antigen levels. Br J Cancer 1998; 78: 4138
  • 11
    Gormley GJ, Stoner E, Bruskewitz RC et al. The effect of finasteride in men with benign prostatic hyperplasia. The Finasteride Study Group. N Engl J Med 1992; 327: 118591
  • 12
    Stoner E, Round E, Ferguson D, Gormley GJ. Clinical experience of the detection of prostate cancer in patients with benign prostate hyperplasia treated with finasteride. The Finasteride Study Group. J Urol 1994; 151: 1296300
  • 13
    Finasteride Study Group. Finasteride (MK-906) in the treatment of benign prostatic hyperplasia. The Finasteride Study Group. Prostate 1993; 22: 2919
  • 14
    Roehrborn CG, Siami P, Barkin J et al.; CombAT Study Group. The effects of combination therapy with dutasteride and tamsulosin on clinical outcomes in men with symptomatic benign prostatic hyperplasia: 4-year results from the CombAT study. Eur Urol 2010; 57: 12331
  • 15
    Roehrborn CG, Andriole GL, Wilson TH, Castro R, Rittmaster RS. Effect of dutasteride on prostate biopsy rates and the diagnosis of prostate cancer in men with lower urinary tract symptoms and enlarged prostates in the Combination of Avodart and Tamsulosin (CombAT) trial. Eur Urol 2010 (in press)
  • 16
    Ilic D, O’Connor D, Green S, Wilt T. Screening for prostate cancer. Cochrane Database Syst Rev 2010 (in press)
  • 17
    Lucia MS, Epstein JI, Goodman PJ et al. Finasteride and high-grade prostate cancer in the Prostate Cancer Prevention Trial. J Natl Cancer Inst 2007; 99: 137583
  • 18
    Thompson IM, Chi C, Ankerst DP et al. Effect of finasteride on the sensitivity of PSA for detecting prostate cancer. J Natl Cancer Inst 2006; 98: 112833
  • 19
    D’Amico AV, Barry MJ. Prostate cancer prevention and finasteride. J Urol 2006; 176: 20103
  • 20
    D’Amico AV, Roehrborn CG. Effect of 1 mg/day finasteride on concentrations of serum prostate-specific antigen in men with androgenic alopecia: a randomised controlled trial. Lancet Oncol 2007; 8: 215