- Email: [email protected]
Effects of Testosterone Replacement Therapy on Lower Urinary Tract Symptoms: A Systematic Review and Meta-analysis
EURURO-6643; No. of Pages 8 EUROPEAN UROLOGY XXX (2016) XXX–XXX
available at www.sciencedirect.com journal homepage: www.europeanurology.com
Review – Benign Prostatic Hyperplasia
Effects of Testosterone Replacement Therapy on Lower Urinary Tract Symptoms: A Systematic Review and Meta-analysis Taylor P. Kohn a,y, Douglas A. Mata b,y, Ranjith Ramasamy c, Larry I. Lipshultz d,* a
Baylor College of Medicine, Houston, TX, USA;
c
Department of Urology, Miller School of Medicine, University of Miami, Miami, FL, USA; d Scott Department of Urology, Baylor College of Medicine, Houston,
b
Department of Pathology, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA;
TX, USA
Article info
Abstract
Article history: Accepted January 22, 2016
Context: There is a potential risk that testosterone replacement therapy (TRT) may exacerbate lower urinary tract symptoms (LUTS) among aging men with late-onset hypogonadism (LOH) because of testosterone’s growth-promoting effects on the prostate. Objective: To compare the change in LUTS severity as assessed using the International Prostate Symptom Score (IPSS) between men receiving TRT versus placebo for the treatment of LOH. Evidence acquisition: Systematic search of MEDLINE, Embase, ClinicalTrials.gov, and The Cochrane Library for randomized controlled trials of TRT for LOH published between January 1992 and September 2015. Studies were eligible for inclusion if they were a randomized control trial, used TRT, and assessed LUTS outcomes using the IPSS. Estimates were pooled using random effects meta-analysis. Differences by study-level characteristics were estimated using meta-regression. Evidence synthesis: Data were extracted from 14 trials involving 2029 participants. The average age was 64.5 yr and the average follow-up was 34.4 mo. Seven studies used topical, five used injectable, and two used oral testosterone. There was no statistically significant difference in pooled changes in IPSS from baseline to follow up in men treated with TRT compared with those receiving placebo (–0.41 points [95% confidence interval: –0.89 to 0.07; I2 = 0%, p = 0.28] vs. 0.12 points [95% confidence interval: –0.32 to 0.55; I2 = 0%, p = 0.81], between-group difference p > 0.05). No between-group differences were noted in subanalyses that controlled for potential confounders such as type of testosterone, change in testosterone, aging male symptom scale, or prostate-specific antigen levels (p > 0.05 for all comparisons). Conclusions: In this meta-analysis of 14 clinical trials of TRT for LOH, the change in IPSS was similar among men receiving TRT versus placebo, suggesting that TRT treatment does not worsen LUTS among men with LOH. Patient summary: In this analysis of 14 clinical trials, testosterone replacement therapy did not worsen lower urinary tract symptoms among men being treated for late-onset hypogonadism. # 2016 European Association of Urology. Published by Elsevier B.V. All rights reserved.
Associate Editor: James Catto Keywords: Benign prostatic hyperplasia Hypogonadism Lower urinary tract symptoms Meta-analysis Prostate Testosterone
y These authors contributed equally to this report. * Corresponding author. Scott Department of Urology, Baylor College of Medicine, 6624 Fannin Street, Suite 1700, Houston, TX 77030, USA. Tel. +1-713-798-6163; Fax: +1-713-798-6007. E-mail address: [email protected] (L.I. Lipshultz).
http://dx.doi.org/10.1016/j.eururo.2016.01.043 0302-2838/# 2016 European Association of Urology. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: Kohn TP, et al. Effects of Testosterone Replacement Therapy on Lower Urinary Tract Symptoms: A Systematic Review and Meta-analysis. Eur Urol (2016), http://dx.doi.org/10.1016/j.eururo.2016.01.043
EURURO-6643; No. of Pages 8 2
EUROPEAN UROLOGY XXX (2016) XXX–XXX
1.
Introduction
Longitudinal studies have demonstrated that men experience a decline in total serum testosterone beginning at 40 yr of age, and that 30% of men meet criteria for late-onset hypogonadism (LOH) by age 70 yr [1,2]. LOH is defined by a decrease in serum testosterone as well as symptoms such as decreased libido, depression, erectile dysfunction, and fatigue [3]. Testosterone replacement therapy (TRT), an accepted treatment for LOH, has been shown to effectively ameliorate many of its symptoms [4–9]. There is a theoretical risk that TRT could exacerbate lower urinary tract symptoms (LUTS). Although some studies have demonstrated that TRT does not exacerbate LUTS among men with LOH, their results have yet to be formally synthesized [8,10–25]. To clarify the available evidence, we performed a systematic review and meta-analysis of randomized controlled trials to determine whether TRT affected LUTS as assessed using the International Prostate Symptom Score (IPSS).
2015 that reported IPSS for hypogonadal men receiving TRT were identified using electronic searches of MEDLINE, Embase, ClinicalTrials.gov, and The Cochrane Library, by scanning the reference lists of articles identified and by correspondence with study investigators using the approach recommended by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines [27]. A flow diagram for study selection can be seen in Figure 1. The computer-based searches combined terms related to TRT, LOH, and study design, without language restriction (full details of the search strategy are provided in Supplementary Methods 1). 2.2.
Inclusion criteria and trial selection
2.
Evidence acquisition
Studies were included if they were a randomized controlled trials of TRT for LOH that reported on LUTS as measured by IPSS, a validated seven-question questionnaire that assesses urinary frequency, nocturia, weak stream, hesitancy, intermittent stream, incomplete emptying, and urgency on a scale of 0 to 5 [26]. The most comprehensive publication was used when there were several involving the same study population.
2.1.
Search strategy
2.3.
Randomized controlled trials published between January 1992—the year the IPSS was developed [26]—and September
Data extraction
The following information was extracted independently by two trained investigators (T.P.K. and D.A.M.) using a
1028 Records idenfied through database searching 402 MEDLINE 600 Embase 25 ClinicalTrials.gov 1 The Cochrane Library
895 Excluded based on review of tle and abstract 673 Wrong populaon or outcome 223 Duplicates
133 Arcles screened
119 Excluded aer review of full text 92 Did not report IPSS data 14 Reported on the same populaon 6 Wrong populaon or treatment 3 Did not report complete IPSS data 2 Review arcles Nonrandomized control trial 2
14 Full-text arcles included Fig. 1 – Flow diagram for study selection. IPSS = International Prostate Symptom Score.
Please cite this article in press as: Kohn TP, et al. Effects of Testosterone Replacement Therapy on Lower Urinary Tract Symptoms: A Systematic Review and Meta-analysis. Eur Urol (2016), http://dx.doi.org/10.1016/j.eururo.2016.01.043
EURURO-6643; No. of Pages 8 EUROPEAN UROLOGY XXX (2016) XXX–XXX
standardized form: authors and publication year, participant inclusion and exclusion criteria, sample size, length of follow up, geographic locale in which the study took place, mean or median participant age, method of TRT administration, IPSS, prostate-specific antigen (PSA) levels, aging male symptoms (AMS) scale, and serum testosterone levels. All discrepancies were resolved by discussion and adjudication of a third reviewer (R.R.). 2.4.
Quality assessment
The risk of bias in the included randomized trials was assessed using the Cochrane Risk of Bias Assessment tool in the domains of randomization, sequence generation, allocation concealment, blinding, completeness of outcome data, selective outcome reporting, and other potential sources of bias [28]. Domains were independently assessed by two trained investigators (T.P.K. and D.A.M.). All discrepancies were resolved by discussion and adjudication by a third reviewer (R.R.). Risk of bias graph and summary were generated via RevMan software version 5.2. (Cochrane, Freiburg, Germany) [29]. 2.5.
Data synthesis and analysis
The mean differences in IPSS measured prior to initiating and then after treatment with either TRT or placebo were calculated for each individual study. Overall differences were then calculated by pooling the study-specific estimates using random effects meta-analysis that included between-study heterogeneity [30]. Between-study heterogeneity was assessed with standard chi-square tests and the I2 statistic (ie, the percentage of variability in prevalence estimates due to heterogeneity rather than sampling error or chance) [31,32] and by comparing results from studies grouped according to prespecified study-level characteristics (type of testosterone, change in PSA, change in testosterone levels, and change in AMS scale) using stratified meta-analysis and meta-regression [33,34]. The influence of individual studies on the overall summary estimates was examined by serially excluding each study in a sensitivity analysis. Bias secondary to small study effects was investigated using funnel plot and Egger’s test [35,36]. All analyses were performed using R Foundation for Statistical Computing 3.2.2 (The R Foundation, Vienna, Austria) [37]. Statistical tests were two-sided and used a significance threshold of p < 0.05.
3
secondary outcome. Eight studies did not publish an IPSS score as an exclusion criterion, five studies excluded men with IPSS scores 20, and one study excluded men with IPSS scores 14. Seven studies used topical, five used injectable, and two used oral testosterone. Three studies measured serum testosterone with radioimmunoassay, three with immunochemical assays, two with chemiluminescence assays, one with gas chromatography, and one with liquid chromatography. Two studies did not specify how testosterone was measured. The majority of studies were found to be at low risk of bias (risk of bias for each individual study can be found in Supplemental Table 1). 3.2.
Effect of TRT on change in IPSS
There was no statistically significant difference in pooled change in IPSS from baseline to follow up in men treated with TRT compared with those receiving placebo (–0.41 points [95% confidence interval: –0.89 to 0.07; I2 = 0%, p = 0.28] vs. 0.12 points [95% CI: –0.32 to 0.55; I2 = 0%, p = 0.81]; betweengroup difference, Q = 2.57, p = 0.11; Fig. 2). Sensitivity analysis, in which the analysis was serially repeated after the exclusion of each study, demonstrated that in the control group no individual study affected the overall prevalence estimate by more than an absolute difference of 0.08. In the TRT group, one study, Basaria et al [23] was found to affect the overall prevelnce estimate by an absolute difference of 0.21 (p = 0.01; Supplementary Table 2). 3.3.
Effect of TRT on change in IPSS according to study-level
characteristics
Among the 14 studies, no between-group differences were noted in subanalyses that controlled for potential confounders such as type of testosterone (Fig. 3), change in PSA, change in testosterone levels, or change in AMS scale (p > 0.05 for all comparisons; Table 2). 3.4.
Assessment of publication bias
Visual inspection of the funnel plot revealed minimal asymmetry, suggesting that the pooled estimates were unlikely to be importantly biased secondary to small study effects (Supplementary Fig. 1). The Egger et al [36] regression asymmetry test supported this finding (control: z = –0.49, p = 0.62; TRT: z = –1.84, p = 0.07). Discussion
3.
Evidence synthesis
4.
3.1.
Study characteristics
Our systematic review and meta-analysis of 14 randomized controlled trials involving 2029 men demonstrated that there was no statistically significant change in IPSS among hypogonadal men receiving TRT compared with men receiving a placebo. Neither was a clinically significant change observed as only a change in IPSS greater than 3 points as perceived by the men [38]. This negative result suggests that TRT administration does not worsen LUTS in hypogonadal men with no, mild, or moderate LUTS.
Fourteen randomized controlled trials involving 2029 participants were included in this meta-analysis (Table 1). The median number of participants per study was 145 (range, 37–321), the mean age was 64.5 yr, and the mean follow up was 34.4 mo. Five studies took place in North America, four in Asia, four in Europe, and one in Australia. Two of the trials studied IPSS as a primary outcome while 12 studied it as a
Please cite this article in press as: Kohn TP, et al. Effects of Testosterone Replacement Therapy on Lower Urinary Tract Symptoms: A Systematic Review and Meta-analysis. Eur Urol (2016), http://dx.doi.org/10.1016/j.eururo.2016.01.043
EURURO-6643; No. of Pages 8 4
EUROPEAN UROLOGY XXX (2016) XXX–XXX
Table 1 – Selected characteristics of the 14 studies included in this systematic review Author
Length of follow up (wk)
Testosterone inclusion criteria
2015 2015
16 26
2015 Meuleman et al [18] Konaka et al [16] 2015
28
Total T < 300 ng/dl Total T < 400 ng/dl or freeT <50 pg/ml Free T < 0.26nmol/l
52
Total T < 11.8nmol/l
Tan et al [22]
2013
48
Total T < 12nmol/l
Behre et al [11] Shigehara et al [20]
2012 2011
28 52
Total T < 15nmol/l Free T < 8.5 pg/ml
Kalinchenko et al [14]
2010
30
Total T < 350 ng/dl or free T < 65 pg/ml
Srinivas-Shankar et al [21] Kenny et al [8] Emmelot-Vonk et al [13] Chiang et al [12]
2010
28
2010 2008
52 28
Total T <345 ng/dl or free T < 7.2 ng/dl Total T <350 ng/dL Total T < 13.7nmol/l
2007
14
Marks et al [17]
2006
28
Kenny et al [15]
2001
52
Paduch et al [19] Basaria et al [23]
Yr
Testosterone Type
Total T < 300 ng/dl or free T < 8.7 pg/ml Total T < 300 ng/dl
Bioavailable T < 4.44nmol/L
Dose
Topical 60 mg of 2% T Topical 7.5 g of 1% T Oral
240 mg/d T undecanoate IM 250 mg T enanthate every 4 wk IM 1 g of T undecanoate at 0 wk, 6 wk, 18 wk, 30 wk, and 42 wk Topical 5 g of 1% T IM 250 mg T enanthate every 4 wk 1 g of T IM undecanoate at 0 wk, 6 wk, and 18 wk Topical 5 g of 1% T Topical 5 g of 1% T Oral 160 mg/d T undecanoate Topical 5 g of 1% T IM
150 mg T enanthate biweekly Topical Two 2.5 mg transdermal T patches
Sample size
Age (yr)
TRT Control TRT Control
Change in IPSS TRT
Control
PSA (ng/ml) TRT
Control
30 129
35 119
52.7 66.9
48.4 68.3
–1.1 1.02
0.5 0.56
0.12 0.69
–0.04 0.25
69
71
58.6
58.4
–1.33
0.42
0.04
0.03
120
100
65.7
67.6
–0.56
0.88
0.19
0.21
56
58
53.1
53.8
–2.5
–1.6
0.55
0.15
166 23
155 23
61.9 72
62.1 68.9
–0.7 –3.2
0.6 –0.5
0.19 0.322
104
65
51.6
52.8
–0.6
–0.5
0
0.2
132
132
73.7
73.9
–0.2
0.4
0.5
0
53 113
46 110
77.9 67.1
76.3 67.4
0.1 0.3
–0.1 0.1
20
17
47.9
56.1
–1.9
–2
–
–
20
19
64.5
65.75
–0.25
–0.5
0.72
1.145
20
24
76
75
0.3
1.8
0.6
0.3
–0.05 0
–0.08 0.305
–0.02 0
IPSS = International Prostate Symptom Score; PSA = prostate-specific antigen; T = testosterone; TRT = testosterone replacement therapy; IM = intramuscular.
It is important to note that several of the trials excluded participants with severe LUTS, defined by an IPSS > 19. The mean IPSS among all men included in this meta-analysis had a baseline IPSS of 7.15, and there was no difference in baseline IPSS between the treatment and control groups
Table 2 – Meta-regression by change in prostate-specific antigen, testosterone levels, and aging male symptom score Control Meta-regression
Slope
Lower CI
Upper CI
Q
p value
Change in T Change in PSA Change in AMS
–0.02 –0.97 0.02
–0.04 –3.82 –0.35
0.00 1.88 0.39
3.19 0.45 0.01
0.07 0.50 0.91
Meta-regression
Slope
Lower CI
Upper CI
Q
p value
Change in T Change in PSA Change in AMS
–0.01 1.14 –0.06
–0.01 –0.76 –0.21
0.00 3.03 0.10
3.20 1.38 0.52
0.07 0.24 0.47
TRT arm
AMS = Aging Male Symptoms Scale; CI = confidence interval; PSA = prostatespecific antigen; T = testosterone; TRT = testosterone replacement therapy.
(7.15 vs. 7.16). Changes in testosterone delivery method, PSA level, and testosterone level were not significantly associated with change in IPSS. AMS scales were extracted from the studies to determine if an overall improvement in health caused men to notice worsened LUTS. AMS scales measure health-related quality of life and symptoms of aging men; a lower score corresponds to an improved health-related quality of life. AMS scales decreased by an average of 4.7 points in men not receiving testosterone compared with a decrease of 7.6 points in men placed on TRT. These changes, however, were not significantly associated with a change in IPSS. Therefore, men were no more aware of worsened or improved IPSS when placed on TRT when compared with men receiving a placebo. Studies collected data on IPSS and PSA at various time intervals throughout the trial. Only one trial reported blinded results beyond 12 mo—Basaria et al [23] who reported results through 36 mo. In this study by Basaria et al [23], we selected a time point within 12 mo as a paper by Saad et al [39] suggests that the effect of testosterone on PSA and prostate volume was first observed at 3 mo after treatment initiation, and then treatment effect plateaued after 12 mo. Additionally, Saad et al [39] state that the changes in PSA and prostate volume after 12 mo is likely
Please cite this article in press as: Kohn TP, et al. Effects of Testosterone Replacement Therapy on Lower Urinary Tract Symptoms: A Systematic Review and Meta-analysis. Eur Urol (2016), http://dx.doi.org/10.1016/j.eururo.2016.01.043
EURURO-6643; No. of Pages 8 5
EUROPEAN UROLOGY XXX (2016) XXX–XXX
MD
95% CI
Control Kenny et al [15] 2001 Kenny et al [8] 2010 Paduch et al [19] 2015 Meuleman et al [18] 2015 Konaka et al [16] 2015 Behre et al [11] 2012 Chiang et al [12] 2007 Shigehara et al [20] 2011 Marks et al [17] 2006 Emmelot-Vonk et al [13] 2008 Basaria et al [23] 2015 Kalinchenko et al [14] 2010 Srinivas-Shankar et al [21] 2010 Tan et al [22] 2013 Random effects model
−1.80 −0.10 −1.20 0.42 0.88 0.60 −2.00 −0.50 −0.50 0.10 0.56 −0.50 0.40 −1.60 0.12
(−4.64 to 1.04) (−2.67 to 2.47) (−3.71 to 1.31) (−0.54 to 1.38) (−1.14 to 2.90) (−0.81 to 2.01) (−5.68 to 1.68) (−5.46 to 4.46) (−5.07 to 4.07) (−1.15 to 1.35) (−0.67 to 1.79) (−1.77 to 0.77) (−0.72 to 1.52) (−4.62 to 1.42) (−0.32 to 0.55)
TRT Kenny et al [15] 2001 Kenny et al [8] 2010 Paduch et al [19] 2015 Meuleman et al [18] 2015 Konaka et al [16] 2015 Behre et al [11] 2012 Chiang et al [12] 2007 Shigehara et al [20] 2011 Marks et al [17] 2006 Emmelot-Vonk et al [13] 2008 Basaria et al [23] 2015 Kalinchenko et al [14] 2010 Srinivas-Shankar et al [21] 2010 Tan et al [22] 2013 Random effects model
0.30 0.10 0.60 −1.33 −0.56 −0.70 −1.60 −3.20 −0.25 0.30 1.08 −0.60 −0.20 −2.50 −0.41
(−3.37 to 3.97) (−2.13 to 2.33) (−1.89 to 3.09) (−2.31 to −0.35) (−2.72 to 1.60) (−1.84 to 0.44) (−5.34 to 2.14) (−8.47 to 2.07) (−4.56 to 4.06) (−0.99 to 1.59) (−0.19 to 2.35) (−1.52 to 0.32) (−1.47 to 1.07) (−5.05 to 0.05) (−0.89 to 0.07)
Study
I2 = 0%, tau2 = 0, p = 0.81
I2 = 15.6%, tau2 = 0.12, p = 0.28
−10
−5
0
5
10
Mean difference Fig. 2 – Forest plot of change in International Prostate Symptom Score for men on testosterone replacement therapy verses placebo. CI = confidence interval; MD = mean difference.
related to aging rather than testosterone therapy. Thus, we chose time points at or within that first 12 mo, as literature suggests that this is when TRT is primarily responsible for increases in PSA and prostate volume. In Basairia et al [23], from 6 mo to 36 mo, PSA levels increased in the control arm by 0.2 ng/dl and in the TRT arm by 0.3 ng/dl and IPSS decreased by 1.8 points in the control arm and by 0.7 points in the TRT arm. Thus, even after 36 mo of TRT, no significant or clinical changes in IPSS are observed. The 2010 Endocrinology Clinical Society Practice Guidelines for testosterone therapy in men with androgen deficiency syndromes recommend against the use of testosterone in men with severe LUTS, defined as IPSS > 19 [40]. Only two studies to date has detailed inclusion of men with severe LUTS as many studies use IPSS > 19 as an exclusion criterion. Tan et al [22] included 10 men in the placebo arm and seven men in the treatment arm with severe IPSS (IPSS > 19), yet this study did not report changes in IPSS specifically for this group. In a retrospective study of
120 men, Pearl et al [25] found that patients with severe IPSS once started on TRT had an average decrease of IPSS of 7.42 [25]. This decrease in IPSS was found to be significant compared with patients with baseline mild and medium IPSS. This study, however, did not quantify how many patients had a baseline severe IPSS or if the patients who were prescribed medication for bothersome LUTS during the study were removed. Thus, while this study shows that TRT does not worsen IPSS, it is difficult to determine if this is due to TRT or prescribed LUTS medication. Two reviews have been published on this topic [41,42]. In a review of the 2010 Endocrinology Clinical Society Practice Guidelines, Seftel et al [41] called for this recommendation to be reexamined as studies seem to refute the belief that TRT exacerbates LUTS. In this meta-analysis, only randomized controlled trials were included. One observational study was identified that had both a therapeutic group as well as a control group, but was excluded given that observational studies may be
Please cite this article in press as: Kohn TP, et al. Effects of Testosterone Replacement Therapy on Lower Urinary Tract Symptoms: A Systematic Review and Meta-analysis. Eur Urol (2016), http://dx.doi.org/10.1016/j.eururo.2016.01.043
EURURO-6643; No. of Pages 8 6
EUROPEAN UROLOGY XXX (2016) XXX–XXX
Study
Mean difference
95% CI
–0.29
[–1.27; 0.68]
0.30
[–0.46; 1.06]
0.17
[–0.47; 0.80]
Intramuscular
–0.82
[–1.60; –0.04]
Oral
–0.57
[–2.17; 1.02]
0.01
[–0.62; 0.64]
–0.41
[–0.89; 0.07]
Mean difference
Control Intramuscular I2 = 0%, tau2 = 0, p = 0.71
Oral I2 = 0%, tau2 = 0, p = 0.69
Topical I2 = 0%, tau2 = 0, p = 0.51
TRT Arm I2=0%, tau2=0, p = 0.59 I2 = 74.2%, tau2 = 0.99, p = 0.05
Topical
I2 = 0%, tau2 = 0, p = 0.51
Random effects model I2 = 15.6%, tau2 = 0.12, p = 0.28
–10
–5
0
5
10
Fig. 3 – Subanalyses by study-level characteristics. CI = confidence interval; TRT = testosterone replacement therapy.
confounded by unmeasured variables [25]. No prior metaanalysis has looked primarily at changes in LUTS for men receving TRT. One earlier meta-analysis by Cui and Zhang [10] in 2013 performed a subanalysis of 346 men, looking at changes in LUTS for men on androgen replacement therapy, which found no significant difference in IPSS between men on TRT and men who were not. Our meta-analysis included a greater number of participants and studies, as well as four studies that have been published since Cui and Zhang [10]. In addition, they examined studies that used both testosterone and dihydrotestosterone therapies, while our meta-analysis focused solely on TRT. Our study has important limitations. As mentioned, few studies examine men with severe LUTS. Additional factors such as age, prostatitis, and bladder dysfunction could also contribute to potential changes in IPSS levels. Less than 30% of studies report post-void residual volume and prostate volume, thus limiting relevant information analyzed for lower urinary tract functions. Measurements of testosterone levels were performed using various assay methods, introducing potential heterogeneity. Furthermore, to deterimine hypogonadism some studies included only men with a low serum testosterone and exhibited other hypogonadal symptoms, while other studies only required low serum testosterone levels for inclusion regardless of hypogonadal symptoms. This meta-analysis demonstrates that in hypogonadal men with mild and moderate LUTS, no significant changes occur in IPSS when treated with TRT. Our analysis lacks hypogonadal men with severe LUTS as the current Endocrinology Clinical Society Practice Guidelines recommend against treating this population with TRT. Our meta-analysis encourages the clinical question: would hypogonadal men
with severe LUTS and treated with TRT also see stable IPSS over the course of treatment as is seen in hypogonadal men with mild and moderate LUTS? Thus, we recommend that a randomized controlled trial be conducted specifically exploring the effects of testosterone on IPSS in hypogonadal men with severe LUTS. 5.
Conclusion
In this meta-analysis of randomized controlled trials evaluating the effect of TRT on LUTS, the changes in IPSS were similar among men who received TRT versus those who did not. This finding suggests that hypogonadal men treated with TRT should not experience aggravation of LUTS after therapy. Author contributions: Larry I. Lipshultz had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Kohn, Mata, Ramasamy, Lipshultz. Acquisition of data: Kohn, Mata, Ramasamy. Analysis and interpretation of data: Kohn, Mata, Ramasamy, Lipshultz. Drafting of the manuscript: Kohn, Mata, Ramasamy, Lipshultz. Critical revision of the manuscript for important intellectual content: Kohn, Mata, Ramasamy, Lipshultz. Statistical analysis: Mata, Kohn, Ramasamy. Obtaining funding: None. Administrative, technical, or material support: Kohn, Mata, Ramasamy, Lipshultz. Supervision: Mata, Ramasamy, Lipshultz. Other: None. Financial disclosures: Larry I. Lipshultz certifies that all conflicts of interest, including specific financial interests and relationships and
Please cite this article in press as: Kohn TP, et al. Effects of Testosterone Replacement Therapy on Lower Urinary Tract Symptoms: A Systematic Review and Meta-analysis. Eur Urol (2016), http://dx.doi.org/10.1016/j.eururo.2016.01.043
EURURO-6643; No. of Pages 8 EUROPEAN UROLOGY XXX (2016) XXX–XXX
affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: None.
7
and other parameters in older men: a randomized controlled trial. JAMA 2008;299:39–52. [14] Kalinchenko SY, Tishova YA, Mskhalaya GJ, Gooren LJ, Giltay EJ, Saad F. Effects of testosterone supplementation on markers of the metabolic syndrome and inflammation in hypogonadal men with
Funding/Support and role of the sponsor: None.
the metabolic syndrome: the double-blinded placebo-controlled Moscow study. Clin Endocrinol 2010;73:602–12.
Appendix A. Supplementary data
[15] Kenny AM, Prestwood KM, Gruman CA, Marcello KM, Raisz LG. Effects of transdermal testosterone on bone and muscle in older men with low bioavailable testosterone levels. J Gerontol A Biol Sci
Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j. eururo.2016.01.043.
Med Sci 2001;56:M266–72. [16] Konaka H, Sugimoto K, Orikasa H, et al. Effects of long-term androgen replacement therapy on the physical and mental statuses of aging males with late-onset hypogonadism: a multicenter ran-
References
domized controlled trial in Japan (EARTH Study). Asian J Androl 2016;18:25–34.
[1] Feldman HA, Longcope C, Derby CA, et al. Age trends in the level of
[17] Marks LS, Mazer NA, Mostaghel E, et al. Effect of testosterone
serum testosterone and other hormones in middle-aged men:
replacement therapy on prostate tissue in men with late-onset
longitudinal results from the Massachusetts male aging study.
hypogonadism: a randomized controlled trial. JAMA 2006;296:
J Clin Endocrinol Metab 2002;87:589–98. [2] Harman SM, Metter EJ, Tobin JD, Pearson J, Blackman MR. Longitu-
2351–61. [18] Meuleman EJ, Legros JJ, Bouloux PM, et al. Effects of long-term oral
dinal effects of aging on serum total and free testosterone levels in
testosterone undecanoate therapy on urinary symptoms: data from
healthy men. Baltimore Longitudinal Study of Aging. J Clin Endocrinol
a 1-year, placebo-controlled, dose-ranging trial in aging men with
Metab 2001;86:724–31. [3] Schubert M, Jockenhovel F. Late-onset hypogonadism in the aging
symptomatic hypogonadism. Aging Male 2015;18:157–63. [19] Paduch DA, Polzer PK, Ni X, Basaria S. Testosterone replacement in
male (LOH): definition, diagnostic and clinical aspects. J Endocrinol
androgen-deficient men with ejaculatory dysfunction: A random-
Invest 2005;28:23–7.
ized controlled trial. J Clin Endocrinol Metab 2015;100:2956–62.
[4] Wang C, Alexander G, Berman N, et al. Testosterone replacement
[20] Shigehara K, Sugimoto K, Konaka H, et al. Androgen replacement
therapy improves mood in hypogonadal men–a clinical research
therapy contributes to improving lower urinary tract symptoms in
center study. J Clin Endocrinol Metab 1996;81:3578–83.
patients with hypogonadism and benign prostate hypertrophy: a
[5] Wang C, Swerdloff RS, Iranmanesh A, et al. Transdermal testosterone gel improves sexual function, mood, muscle strength, and body
randomised controlled study. Aging Male 2011;14:53–8. [21] Srinivas-Shankar U, Roberts SA, Connolly MJ, et al. Effects of tes-
composition parameters in hypogonadal men. J Clin Endocrinol
tosterone on muscle strength, physical function, body composition,
Metab 2000;85:2839–53.
and quality of life in intermediate-frail and frail elderly men: a
[6] Seidman SN, Rabkin JG. Testosterone replacement therapy for hypogonadal men with SSRI-refractory depression. J Affect Disord 1998;48:157–61. [7] Snyder PJ, Peachey H, Berlin JA, et al. Effects of testosterone replacement in hypogonadal men. J Clin Endocrinol Metab 2000; 85:2670–7. [8] Kenny AM, Kleppinger A, Annis K, et al. Effects of transdermal
randomized, double-blind, placebo-controlled study. J Clin Endocrinol Metab 2010;95:639–50. [22] Tan WS, Low WY, Ng CJ, et al. Efficacy and safety of long-acting intramuscular testosterone undecanoate in aging men: a randomised controlled study. BJU Int 2013;111:1130–40. [23] Basaria S, Harman SM, Travison TG, et al. Effects of testosterone administration for 3 years on subclinical atherosclerosis progres-
testosterone on bone and muscle in older men with low bioavail-
sion in older men with low or low-normal testosterone levels:
able testosterone levels, low bone mass, and physical frailty. J Am
A randomized clinical trial. JAMA 2015;314:570–81.
Geriatr Soc 2010;58:1134–43.
[24] Pechersky AV, Mazurov VI, Semiglazov VF, Karpischenko AI,
[9] Dobs AS, Meikle AW, Arver S, Sanders SW, Caramelli KE, Mazer NA.
Mikhailichenko VV, Udintsev AV. Androgen administration in
Pharmacokinetics, efficacy, and safety of a permeation-enhanced
middle-aged and ageing men: effects of oral testosterone undecano-
testosterone transdermal system in comparison with bi-weekly
ate on dihydrotestosterone, oestradiol and prostate volume. Int J
injections of testosterone enanthate for the treatment of hypogonadal men. J Clin Endocrinol Metab 1999;84:3469–78. [10] Cui Y, Zhang Y. The effect of androgen-replacement therapy on prostate growth: a systematic review and meta-analysis. Eur Urol 2013;64:811–22.
Androl 2002; 25:119–25. [25] Pearl JA, Berhanu D, Franc¸ois N, et al. Testosterone supplementation does not worsen lower urinary tract symptoms. J Urol 2013;190: 1828–33. [26] Barry MJ, Fowler Jr FJ, O’Leary MP, et al. The American Urological
[11] Behre HM, Tammela TL, Arver S, et al. A randomized, double-blind,
Association symptom index for benign prostatic hyperplasia. The
placebo-controlled trial of testosterone gel on body composition
Measurement Committee of the American Urological Association.
and health-related quality-of-life in men with hypogonadal to low-
J Urol 1992;148:1549–57.
normal levels of serum testosterone and symptoms of androgen
[27] Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items
deficiency over 6 months with 12 months open-label follow-up.
for systematic reviews and meta-analyses: the PRISMA statement.
Aging Male 2012;15:198–207.
Ann Intern Med 2009;151:264–9.
[12] Chiang HS, Hwang TI, Hsui YS, et al. Transdermal testosterone gel
[28] Higgins JP, Green S. Cochrane handbook for systematic reviews of
increases serum testosterone levels in hypogonadal men in Taiwan
interventions v.5.1.0. The Cochrane Collaboration; 2011. www.
with improvements in sexual function. Int J Impot Res 2007; 19:411–7. [13] Emmelot-Vonk MH, Verhaar HJ, Nakhai Pour HR, et al. Effect of testosterone supplementation on functional mobility, cognition,
cochrane-handbook.org. [29] Review Manager (RevMan) Computer Program. Version 5.2. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration; 2014.
Please cite this article in press as: Kohn TP, et al. Effects of Testosterone Replacement Therapy on Lower Urinary Tract Symptoms: A Systematic Review and Meta-analysis. Eur Urol (2016), http://dx.doi.org/10.1016/j.eururo.2016.01.043
EURURO-6643; No. of Pages 8 8
EUROPEAN UROLOGY XXX (2016) XXX–XXX
[30] Borenstein M, Hedges LV, Higgins JP, Rothstein HR. A basic intro-
[38] Barry MJ, Williford WO, Chang Y, et al. Benign prostatic hyperplasia
duction to fixed-effect and random-effects models for meta-
specific health status measures in clinical research: how much
analysis. Res Synth Methods 2010;1:97–111. [31] Higgins JP, Thompson SG. Quantifying heterogeneity in a metaanalysis. Stat Med 2002;21:1539–58. [32] Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003;327:557–60. [33] Sterne JA, Juni P, Schulz KF, Altman DG, Bartlett C, Egger M.
change in the American Urological Association symptom index and the benign prostatic hyperplasia impact index is perceptible to patients? J Urol 1995;154:1770–4. [39] Saad F, Aversa A, Isidori AM, Zafalon L, Zitzmann M, Gooren L. Onset of effects of testosterone treatment and time span until maximum effects are achieved. Eur J Endocrinol 2011;165:675–85.
Statistical methods for assessing the influence of study character-
[40] Bhasin S, Cunningham GR, Hayes FJ, et al. Testosterone therapy in
istics on treatment effects in ‘meta-epidemiological’ research. Stat
men with androgen deficiency syndromes: an Endocrine Society
Med 2002;21:1513–24.
clinical practice guideline. J Clin Endocrinol Metab 2010;95:
[34] van Houwelingen HC, Arends LR, Stijnen T. Advanced methods in meta-analysis: multivariate approach and meta-regression. Stat Med 2002;21:589–624. [35] Sterne JA, Egger M. Funnel plots for detecting bias in meta-analysis: guidelines on choice of axis. J Clin Epidemiol 2001;54:1046–55.
2536–59. [41] Seftel AD, Kathrins M, Niederberger C. Critical update of the 2010 Endocrine Society Clinical Practice Guidelines for Male Hypogonadism: A systematic analysis. Mayo Clin Proc 2015;90: 1104–15.
[36] Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-
[42] Kathrins M, Doersch K, Nimeh T, Canto A, Niederberger C, Seftel A.
analysis detected by a simple, graphical test. BMJ 1997;315:629–34.
The Relationship between Testosterone Replacement Therapy and
[37] R-Core-Team. R: A language and environment for statistical com-
Lower Urinary Tract Symptoms: A systematic review. Urology. In
puting. Vienna, Austria: R Foundation for Statistical Computing;
press. http://dx.doi.org/10.1016/j.urology.2015.11.006.
2012.
Please cite this article in press as: Kohn TP, et al. Effects of Testosterone Replacement Therapy on Lower Urinary Tract Symptoms: A Systematic Review and Meta-analysis. Eur Urol (2016), http://dx.doi.org/10.1016/j.eururo.2016.01.043
available at www.sciencedirect.com journal homepage: www.europeanurology.com
Review – Benign Prostatic Hyperplasia
Effects of Testosterone Replacement Therapy on Lower Urinary Tract Symptoms: A Systematic Review and Meta-analysis Taylor P. Kohn a,y, Douglas A. Mata b,y, Ranjith Ramasamy c, Larry I. Lipshultz d,* a
Baylor College of Medicine, Houston, TX, USA;
c
Department of Urology, Miller School of Medicine, University of Miami, Miami, FL, USA; d Scott Department of Urology, Baylor College of Medicine, Houston,
b
Department of Pathology, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, USA;
TX, USA
Article info
Abstract
Article history: Accepted January 22, 2016
Context: There is a potential risk that testosterone replacement therapy (TRT) may exacerbate lower urinary tract symptoms (LUTS) among aging men with late-onset hypogonadism (LOH) because of testosterone’s growth-promoting effects on the prostate. Objective: To compare the change in LUTS severity as assessed using the International Prostate Symptom Score (IPSS) between men receiving TRT versus placebo for the treatment of LOH. Evidence acquisition: Systematic search of MEDLINE, Embase, ClinicalTrials.gov, and The Cochrane Library for randomized controlled trials of TRT for LOH published between January 1992 and September 2015. Studies were eligible for inclusion if they were a randomized control trial, used TRT, and assessed LUTS outcomes using the IPSS. Estimates were pooled using random effects meta-analysis. Differences by study-level characteristics were estimated using meta-regression. Evidence synthesis: Data were extracted from 14 trials involving 2029 participants. The average age was 64.5 yr and the average follow-up was 34.4 mo. Seven studies used topical, five used injectable, and two used oral testosterone. There was no statistically significant difference in pooled changes in IPSS from baseline to follow up in men treated with TRT compared with those receiving placebo (–0.41 points [95% confidence interval: –0.89 to 0.07; I2 = 0%, p = 0.28] vs. 0.12 points [95% confidence interval: –0.32 to 0.55; I2 = 0%, p = 0.81], between-group difference p > 0.05). No between-group differences were noted in subanalyses that controlled for potential confounders such as type of testosterone, change in testosterone, aging male symptom scale, or prostate-specific antigen levels (p > 0.05 for all comparisons). Conclusions: In this meta-analysis of 14 clinical trials of TRT for LOH, the change in IPSS was similar among men receiving TRT versus placebo, suggesting that TRT treatment does not worsen LUTS among men with LOH. Patient summary: In this analysis of 14 clinical trials, testosterone replacement therapy did not worsen lower urinary tract symptoms among men being treated for late-onset hypogonadism. # 2016 European Association of Urology. Published by Elsevier B.V. All rights reserved.
Associate Editor: James Catto Keywords: Benign prostatic hyperplasia Hypogonadism Lower urinary tract symptoms Meta-analysis Prostate Testosterone
y These authors contributed equally to this report. * Corresponding author. Scott Department of Urology, Baylor College of Medicine, 6624 Fannin Street, Suite 1700, Houston, TX 77030, USA. Tel. +1-713-798-6163; Fax: +1-713-798-6007. E-mail address: [email protected] (L.I. Lipshultz).
http://dx.doi.org/10.1016/j.eururo.2016.01.043 0302-2838/# 2016 European Association of Urology. Published by Elsevier B.V. All rights reserved.
Please cite this article in press as: Kohn TP, et al. Effects of Testosterone Replacement Therapy on Lower Urinary Tract Symptoms: A Systematic Review and Meta-analysis. Eur Urol (2016), http://dx.doi.org/10.1016/j.eururo.2016.01.043
EURURO-6643; No. of Pages 8 2
EUROPEAN UROLOGY XXX (2016) XXX–XXX
1.
Introduction
Longitudinal studies have demonstrated that men experience a decline in total serum testosterone beginning at 40 yr of age, and that 30% of men meet criteria for late-onset hypogonadism (LOH) by age 70 yr [1,2]. LOH is defined by a decrease in serum testosterone as well as symptoms such as decreased libido, depression, erectile dysfunction, and fatigue [3]. Testosterone replacement therapy (TRT), an accepted treatment for LOH, has been shown to effectively ameliorate many of its symptoms [4–9]. There is a theoretical risk that TRT could exacerbate lower urinary tract symptoms (LUTS). Although some studies have demonstrated that TRT does not exacerbate LUTS among men with LOH, their results have yet to be formally synthesized [8,10–25]. To clarify the available evidence, we performed a systematic review and meta-analysis of randomized controlled trials to determine whether TRT affected LUTS as assessed using the International Prostate Symptom Score (IPSS).
2015 that reported IPSS for hypogonadal men receiving TRT were identified using electronic searches of MEDLINE, Embase, ClinicalTrials.gov, and The Cochrane Library, by scanning the reference lists of articles identified and by correspondence with study investigators using the approach recommended by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines [27]. A flow diagram for study selection can be seen in Figure 1. The computer-based searches combined terms related to TRT, LOH, and study design, without language restriction (full details of the search strategy are provided in Supplementary Methods 1). 2.2.
Inclusion criteria and trial selection
2.
Evidence acquisition
Studies were included if they were a randomized controlled trials of TRT for LOH that reported on LUTS as measured by IPSS, a validated seven-question questionnaire that assesses urinary frequency, nocturia, weak stream, hesitancy, intermittent stream, incomplete emptying, and urgency on a scale of 0 to 5 [26]. The most comprehensive publication was used when there were several involving the same study population.
2.1.
Search strategy
2.3.
Randomized controlled trials published between January 1992—the year the IPSS was developed [26]—and September
Data extraction
The following information was extracted independently by two trained investigators (T.P.K. and D.A.M.) using a
1028 Records idenfied through database searching 402 MEDLINE 600 Embase 25 ClinicalTrials.gov 1 The Cochrane Library
895 Excluded based on review of tle and abstract 673 Wrong populaon or outcome 223 Duplicates
133 Arcles screened
119 Excluded aer review of full text 92 Did not report IPSS data 14 Reported on the same populaon 6 Wrong populaon or treatment 3 Did not report complete IPSS data 2 Review arcles Nonrandomized control trial 2
14 Full-text arcles included Fig. 1 – Flow diagram for study selection. IPSS = International Prostate Symptom Score.
Please cite this article in press as: Kohn TP, et al. Effects of Testosterone Replacement Therapy on Lower Urinary Tract Symptoms: A Systematic Review and Meta-analysis. Eur Urol (2016), http://dx.doi.org/10.1016/j.eururo.2016.01.043
EURURO-6643; No. of Pages 8 EUROPEAN UROLOGY XXX (2016) XXX–XXX
standardized form: authors and publication year, participant inclusion and exclusion criteria, sample size, length of follow up, geographic locale in which the study took place, mean or median participant age, method of TRT administration, IPSS, prostate-specific antigen (PSA) levels, aging male symptoms (AMS) scale, and serum testosterone levels. All discrepancies were resolved by discussion and adjudication of a third reviewer (R.R.). 2.4.
Quality assessment
The risk of bias in the included randomized trials was assessed using the Cochrane Risk of Bias Assessment tool in the domains of randomization, sequence generation, allocation concealment, blinding, completeness of outcome data, selective outcome reporting, and other potential sources of bias [28]. Domains were independently assessed by two trained investigators (T.P.K. and D.A.M.). All discrepancies were resolved by discussion and adjudication by a third reviewer (R.R.). Risk of bias graph and summary were generated via RevMan software version 5.2. (Cochrane, Freiburg, Germany) [29]. 2.5.
Data synthesis and analysis
The mean differences in IPSS measured prior to initiating and then after treatment with either TRT or placebo were calculated for each individual study. Overall differences were then calculated by pooling the study-specific estimates using random effects meta-analysis that included between-study heterogeneity [30]. Between-study heterogeneity was assessed with standard chi-square tests and the I2 statistic (ie, the percentage of variability in prevalence estimates due to heterogeneity rather than sampling error or chance) [31,32] and by comparing results from studies grouped according to prespecified study-level characteristics (type of testosterone, change in PSA, change in testosterone levels, and change in AMS scale) using stratified meta-analysis and meta-regression [33,34]. The influence of individual studies on the overall summary estimates was examined by serially excluding each study in a sensitivity analysis. Bias secondary to small study effects was investigated using funnel plot and Egger’s test [35,36]. All analyses were performed using R Foundation for Statistical Computing 3.2.2 (The R Foundation, Vienna, Austria) [37]. Statistical tests were two-sided and used a significance threshold of p < 0.05.
3
secondary outcome. Eight studies did not publish an IPSS score as an exclusion criterion, five studies excluded men with IPSS scores 20, and one study excluded men with IPSS scores 14. Seven studies used topical, five used injectable, and two used oral testosterone. Three studies measured serum testosterone with radioimmunoassay, three with immunochemical assays, two with chemiluminescence assays, one with gas chromatography, and one with liquid chromatography. Two studies did not specify how testosterone was measured. The majority of studies were found to be at low risk of bias (risk of bias for each individual study can be found in Supplemental Table 1). 3.2.
Effect of TRT on change in IPSS
There was no statistically significant difference in pooled change in IPSS from baseline to follow up in men treated with TRT compared with those receiving placebo (–0.41 points [95% confidence interval: –0.89 to 0.07; I2 = 0%, p = 0.28] vs. 0.12 points [95% CI: –0.32 to 0.55; I2 = 0%, p = 0.81]; betweengroup difference, Q = 2.57, p = 0.11; Fig. 2). Sensitivity analysis, in which the analysis was serially repeated after the exclusion of each study, demonstrated that in the control group no individual study affected the overall prevalence estimate by more than an absolute difference of 0.08. In the TRT group, one study, Basaria et al [23] was found to affect the overall prevelnce estimate by an absolute difference of 0.21 (p = 0.01; Supplementary Table 2). 3.3.
Effect of TRT on change in IPSS according to study-level
characteristics
Among the 14 studies, no between-group differences were noted in subanalyses that controlled for potential confounders such as type of testosterone (Fig. 3), change in PSA, change in testosterone levels, or change in AMS scale (p > 0.05 for all comparisons; Table 2). 3.4.
Assessment of publication bias
Visual inspection of the funnel plot revealed minimal asymmetry, suggesting that the pooled estimates were unlikely to be importantly biased secondary to small study effects (Supplementary Fig. 1). The Egger et al [36] regression asymmetry test supported this finding (control: z = –0.49, p = 0.62; TRT: z = –1.84, p = 0.07). Discussion
3.
Evidence synthesis
4.
3.1.
Study characteristics
Our systematic review and meta-analysis of 14 randomized controlled trials involving 2029 men demonstrated that there was no statistically significant change in IPSS among hypogonadal men receiving TRT compared with men receiving a placebo. Neither was a clinically significant change observed as only a change in IPSS greater than 3 points as perceived by the men [38]. This negative result suggests that TRT administration does not worsen LUTS in hypogonadal men with no, mild, or moderate LUTS.
Fourteen randomized controlled trials involving 2029 participants were included in this meta-analysis (Table 1). The median number of participants per study was 145 (range, 37–321), the mean age was 64.5 yr, and the mean follow up was 34.4 mo. Five studies took place in North America, four in Asia, four in Europe, and one in Australia. Two of the trials studied IPSS as a primary outcome while 12 studied it as a
Please cite this article in press as: Kohn TP, et al. Effects of Testosterone Replacement Therapy on Lower Urinary Tract Symptoms: A Systematic Review and Meta-analysis. Eur Urol (2016), http://dx.doi.org/10.1016/j.eururo.2016.01.043
EURURO-6643; No. of Pages 8 4
EUROPEAN UROLOGY XXX (2016) XXX–XXX
Table 1 – Selected characteristics of the 14 studies included in this systematic review Author
Length of follow up (wk)
Testosterone inclusion criteria
2015 2015
16 26
2015 Meuleman et al [18] Konaka et al [16] 2015
28
Total T < 300 ng/dl Total T < 400 ng/dl or freeT <50 pg/ml Free T < 0.26nmol/l
52
Total T < 11.8nmol/l
Tan et al [22]
2013
48
Total T < 12nmol/l
Behre et al [11] Shigehara et al [20]
2012 2011
28 52
Total T < 15nmol/l Free T < 8.5 pg/ml
Kalinchenko et al [14]
2010
30
Total T < 350 ng/dl or free T < 65 pg/ml
Srinivas-Shankar et al [21] Kenny et al [8] Emmelot-Vonk et al [13] Chiang et al [12]
2010
28
2010 2008
52 28
Total T <345 ng/dl or free T < 7.2 ng/dl Total T <350 ng/dL Total T < 13.7nmol/l
2007
14
Marks et al [17]
2006
28
Kenny et al [15]
2001
52
Paduch et al [19] Basaria et al [23]
Yr
Testosterone Type
Total T < 300 ng/dl or free T < 8.7 pg/ml Total T < 300 ng/dl
Bioavailable T < 4.44nmol/L
Dose
Topical 60 mg of 2% T Topical 7.5 g of 1% T Oral
240 mg/d T undecanoate IM 250 mg T enanthate every 4 wk IM 1 g of T undecanoate at 0 wk, 6 wk, 18 wk, 30 wk, and 42 wk Topical 5 g of 1% T IM 250 mg T enanthate every 4 wk 1 g of T IM undecanoate at 0 wk, 6 wk, and 18 wk Topical 5 g of 1% T Topical 5 g of 1% T Oral 160 mg/d T undecanoate Topical 5 g of 1% T IM
150 mg T enanthate biweekly Topical Two 2.5 mg transdermal T patches
Sample size
Age (yr)
TRT Control TRT Control
Change in IPSS TRT
Control
PSA (ng/ml) TRT
Control
30 129
35 119
52.7 66.9
48.4 68.3
–1.1 1.02
0.5 0.56
0.12 0.69
–0.04 0.25
69
71
58.6
58.4
–1.33
0.42
0.04
0.03
120
100
65.7
67.6
–0.56
0.88
0.19
0.21
56
58
53.1
53.8
–2.5
–1.6
0.55
0.15
166 23
155 23
61.9 72
62.1 68.9
–0.7 –3.2
0.6 –0.5
0.19 0.322
104
65
51.6
52.8
–0.6
–0.5
0
0.2
132
132
73.7
73.9
–0.2
0.4
0.5
0
53 113
46 110
77.9 67.1
76.3 67.4
0.1 0.3
–0.1 0.1
20
17
47.9
56.1
–1.9
–2
–
–
20
19
64.5
65.75
–0.25
–0.5
0.72
1.145
20
24
76
75
0.3
1.8
0.6
0.3
–0.05 0
–0.08 0.305
–0.02 0
IPSS = International Prostate Symptom Score; PSA = prostate-specific antigen; T = testosterone; TRT = testosterone replacement therapy; IM = intramuscular.
It is important to note that several of the trials excluded participants with severe LUTS, defined by an IPSS > 19. The mean IPSS among all men included in this meta-analysis had a baseline IPSS of 7.15, and there was no difference in baseline IPSS between the treatment and control groups
Table 2 – Meta-regression by change in prostate-specific antigen, testosterone levels, and aging male symptom score Control Meta-regression
Slope
Lower CI
Upper CI
Q
p value
Change in T Change in PSA Change in AMS
–0.02 –0.97 0.02
–0.04 –3.82 –0.35
0.00 1.88 0.39
3.19 0.45 0.01
0.07 0.50 0.91
Meta-regression
Slope
Lower CI
Upper CI
Q
p value
Change in T Change in PSA Change in AMS
–0.01 1.14 –0.06
–0.01 –0.76 –0.21
0.00 3.03 0.10
3.20 1.38 0.52
0.07 0.24 0.47
TRT arm
AMS = Aging Male Symptoms Scale; CI = confidence interval; PSA = prostatespecific antigen; T = testosterone; TRT = testosterone replacement therapy.
(7.15 vs. 7.16). Changes in testosterone delivery method, PSA level, and testosterone level were not significantly associated with change in IPSS. AMS scales were extracted from the studies to determine if an overall improvement in health caused men to notice worsened LUTS. AMS scales measure health-related quality of life and symptoms of aging men; a lower score corresponds to an improved health-related quality of life. AMS scales decreased by an average of 4.7 points in men not receiving testosterone compared with a decrease of 7.6 points in men placed on TRT. These changes, however, were not significantly associated with a change in IPSS. Therefore, men were no more aware of worsened or improved IPSS when placed on TRT when compared with men receiving a placebo. Studies collected data on IPSS and PSA at various time intervals throughout the trial. Only one trial reported blinded results beyond 12 mo—Basaria et al [23] who reported results through 36 mo. In this study by Basaria et al [23], we selected a time point within 12 mo as a paper by Saad et al [39] suggests that the effect of testosterone on PSA and prostate volume was first observed at 3 mo after treatment initiation, and then treatment effect plateaued after 12 mo. Additionally, Saad et al [39] state that the changes in PSA and prostate volume after 12 mo is likely
Please cite this article in press as: Kohn TP, et al. Effects of Testosterone Replacement Therapy on Lower Urinary Tract Symptoms: A Systematic Review and Meta-analysis. Eur Urol (2016), http://dx.doi.org/10.1016/j.eururo.2016.01.043
EURURO-6643; No. of Pages 8 5
EUROPEAN UROLOGY XXX (2016) XXX–XXX
MD
95% CI
Control Kenny et al [15] 2001 Kenny et al [8] 2010 Paduch et al [19] 2015 Meuleman et al [18] 2015 Konaka et al [16] 2015 Behre et al [11] 2012 Chiang et al [12] 2007 Shigehara et al [20] 2011 Marks et al [17] 2006 Emmelot-Vonk et al [13] 2008 Basaria et al [23] 2015 Kalinchenko et al [14] 2010 Srinivas-Shankar et al [21] 2010 Tan et al [22] 2013 Random effects model
−1.80 −0.10 −1.20 0.42 0.88 0.60 −2.00 −0.50 −0.50 0.10 0.56 −0.50 0.40 −1.60 0.12
(−4.64 to 1.04) (−2.67 to 2.47) (−3.71 to 1.31) (−0.54 to 1.38) (−1.14 to 2.90) (−0.81 to 2.01) (−5.68 to 1.68) (−5.46 to 4.46) (−5.07 to 4.07) (−1.15 to 1.35) (−0.67 to 1.79) (−1.77 to 0.77) (−0.72 to 1.52) (−4.62 to 1.42) (−0.32 to 0.55)
TRT Kenny et al [15] 2001 Kenny et al [8] 2010 Paduch et al [19] 2015 Meuleman et al [18] 2015 Konaka et al [16] 2015 Behre et al [11] 2012 Chiang et al [12] 2007 Shigehara et al [20] 2011 Marks et al [17] 2006 Emmelot-Vonk et al [13] 2008 Basaria et al [23] 2015 Kalinchenko et al [14] 2010 Srinivas-Shankar et al [21] 2010 Tan et al [22] 2013 Random effects model
0.30 0.10 0.60 −1.33 −0.56 −0.70 −1.60 −3.20 −0.25 0.30 1.08 −0.60 −0.20 −2.50 −0.41
(−3.37 to 3.97) (−2.13 to 2.33) (−1.89 to 3.09) (−2.31 to −0.35) (−2.72 to 1.60) (−1.84 to 0.44) (−5.34 to 2.14) (−8.47 to 2.07) (−4.56 to 4.06) (−0.99 to 1.59) (−0.19 to 2.35) (−1.52 to 0.32) (−1.47 to 1.07) (−5.05 to 0.05) (−0.89 to 0.07)
Study
I2 = 0%, tau2 = 0, p = 0.81
I2 = 15.6%, tau2 = 0.12, p = 0.28
−10
−5
0
5
10
Mean difference Fig. 2 – Forest plot of change in International Prostate Symptom Score for men on testosterone replacement therapy verses placebo. CI = confidence interval; MD = mean difference.
related to aging rather than testosterone therapy. Thus, we chose time points at or within that first 12 mo, as literature suggests that this is when TRT is primarily responsible for increases in PSA and prostate volume. In Basairia et al [23], from 6 mo to 36 mo, PSA levels increased in the control arm by 0.2 ng/dl and in the TRT arm by 0.3 ng/dl and IPSS decreased by 1.8 points in the control arm and by 0.7 points in the TRT arm. Thus, even after 36 mo of TRT, no significant or clinical changes in IPSS are observed. The 2010 Endocrinology Clinical Society Practice Guidelines for testosterone therapy in men with androgen deficiency syndromes recommend against the use of testosterone in men with severe LUTS, defined as IPSS > 19 [40]. Only two studies to date has detailed inclusion of men with severe LUTS as many studies use IPSS > 19 as an exclusion criterion. Tan et al [22] included 10 men in the placebo arm and seven men in the treatment arm with severe IPSS (IPSS > 19), yet this study did not report changes in IPSS specifically for this group. In a retrospective study of
120 men, Pearl et al [25] found that patients with severe IPSS once started on TRT had an average decrease of IPSS of 7.42 [25]. This decrease in IPSS was found to be significant compared with patients with baseline mild and medium IPSS. This study, however, did not quantify how many patients had a baseline severe IPSS or if the patients who were prescribed medication for bothersome LUTS during the study were removed. Thus, while this study shows that TRT does not worsen IPSS, it is difficult to determine if this is due to TRT or prescribed LUTS medication. Two reviews have been published on this topic [41,42]. In a review of the 2010 Endocrinology Clinical Society Practice Guidelines, Seftel et al [41] called for this recommendation to be reexamined as studies seem to refute the belief that TRT exacerbates LUTS. In this meta-analysis, only randomized controlled trials were included. One observational study was identified that had both a therapeutic group as well as a control group, but was excluded given that observational studies may be
Please cite this article in press as: Kohn TP, et al. Effects of Testosterone Replacement Therapy on Lower Urinary Tract Symptoms: A Systematic Review and Meta-analysis. Eur Urol (2016), http://dx.doi.org/10.1016/j.eururo.2016.01.043
EURURO-6643; No. of Pages 8 6
EUROPEAN UROLOGY XXX (2016) XXX–XXX
Study
Mean difference
95% CI
–0.29
[–1.27; 0.68]
0.30
[–0.46; 1.06]
0.17
[–0.47; 0.80]
Intramuscular
–0.82
[–1.60; –0.04]
Oral
–0.57
[–2.17; 1.02]
0.01
[–0.62; 0.64]
–0.41
[–0.89; 0.07]
Mean difference
Control Intramuscular I2 = 0%, tau2 = 0, p = 0.71
Oral I2 = 0%, tau2 = 0, p = 0.69
Topical I2 = 0%, tau2 = 0, p = 0.51
TRT Arm I2=0%, tau2=0, p = 0.59 I2 = 74.2%, tau2 = 0.99, p = 0.05
Topical
I2 = 0%, tau2 = 0, p = 0.51
Random effects model I2 = 15.6%, tau2 = 0.12, p = 0.28
–10
–5
0
5
10
Fig. 3 – Subanalyses by study-level characteristics. CI = confidence interval; TRT = testosterone replacement therapy.
confounded by unmeasured variables [25]. No prior metaanalysis has looked primarily at changes in LUTS for men receving TRT. One earlier meta-analysis by Cui and Zhang [10] in 2013 performed a subanalysis of 346 men, looking at changes in LUTS for men on androgen replacement therapy, which found no significant difference in IPSS between men on TRT and men who were not. Our meta-analysis included a greater number of participants and studies, as well as four studies that have been published since Cui and Zhang [10]. In addition, they examined studies that used both testosterone and dihydrotestosterone therapies, while our meta-analysis focused solely on TRT. Our study has important limitations. As mentioned, few studies examine men with severe LUTS. Additional factors such as age, prostatitis, and bladder dysfunction could also contribute to potential changes in IPSS levels. Less than 30% of studies report post-void residual volume and prostate volume, thus limiting relevant information analyzed for lower urinary tract functions. Measurements of testosterone levels were performed using various assay methods, introducing potential heterogeneity. Furthermore, to deterimine hypogonadism some studies included only men with a low serum testosterone and exhibited other hypogonadal symptoms, while other studies only required low serum testosterone levels for inclusion regardless of hypogonadal symptoms. This meta-analysis demonstrates that in hypogonadal men with mild and moderate LUTS, no significant changes occur in IPSS when treated with TRT. Our analysis lacks hypogonadal men with severe LUTS as the current Endocrinology Clinical Society Practice Guidelines recommend against treating this population with TRT. Our meta-analysis encourages the clinical question: would hypogonadal men
with severe LUTS and treated with TRT also see stable IPSS over the course of treatment as is seen in hypogonadal men with mild and moderate LUTS? Thus, we recommend that a randomized controlled trial be conducted specifically exploring the effects of testosterone on IPSS in hypogonadal men with severe LUTS. 5.
Conclusion
In this meta-analysis of randomized controlled trials evaluating the effect of TRT on LUTS, the changes in IPSS were similar among men who received TRT versus those who did not. This finding suggests that hypogonadal men treated with TRT should not experience aggravation of LUTS after therapy. Author contributions: Larry I. Lipshultz had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Kohn, Mata, Ramasamy, Lipshultz. Acquisition of data: Kohn, Mata, Ramasamy. Analysis and interpretation of data: Kohn, Mata, Ramasamy, Lipshultz. Drafting of the manuscript: Kohn, Mata, Ramasamy, Lipshultz. Critical revision of the manuscript for important intellectual content: Kohn, Mata, Ramasamy, Lipshultz. Statistical analysis: Mata, Kohn, Ramasamy. Obtaining funding: None. Administrative, technical, or material support: Kohn, Mata, Ramasamy, Lipshultz. Supervision: Mata, Ramasamy, Lipshultz. Other: None. Financial disclosures: Larry I. Lipshultz certifies that all conflicts of interest, including specific financial interests and relationships and
Please cite this article in press as: Kohn TP, et al. Effects of Testosterone Replacement Therapy on Lower Urinary Tract Symptoms: A Systematic Review and Meta-analysis. Eur Urol (2016), http://dx.doi.org/10.1016/j.eururo.2016.01.043
EURURO-6643; No. of Pages 8 EUROPEAN UROLOGY XXX (2016) XXX–XXX
affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: None.
7
and other parameters in older men: a randomized controlled trial. JAMA 2008;299:39–52. [14] Kalinchenko SY, Tishova YA, Mskhalaya GJ, Gooren LJ, Giltay EJ, Saad F. Effects of testosterone supplementation on markers of the metabolic syndrome and inflammation in hypogonadal men with
Funding/Support and role of the sponsor: None.
the metabolic syndrome: the double-blinded placebo-controlled Moscow study. Clin Endocrinol 2010;73:602–12.
Appendix A. Supplementary data
[15] Kenny AM, Prestwood KM, Gruman CA, Marcello KM, Raisz LG. Effects of transdermal testosterone on bone and muscle in older men with low bioavailable testosterone levels. J Gerontol A Biol Sci
Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j. eururo.2016.01.043.
Med Sci 2001;56:M266–72. [16] Konaka H, Sugimoto K, Orikasa H, et al. Effects of long-term androgen replacement therapy on the physical and mental statuses of aging males with late-onset hypogonadism: a multicenter ran-
References
domized controlled trial in Japan (EARTH Study). Asian J Androl 2016;18:25–34.
[1] Feldman HA, Longcope C, Derby CA, et al. Age trends in the level of
[17] Marks LS, Mazer NA, Mostaghel E, et al. Effect of testosterone
serum testosterone and other hormones in middle-aged men:
replacement therapy on prostate tissue in men with late-onset
longitudinal results from the Massachusetts male aging study.
hypogonadism: a randomized controlled trial. JAMA 2006;296:
J Clin Endocrinol Metab 2002;87:589–98. [2] Harman SM, Metter EJ, Tobin JD, Pearson J, Blackman MR. Longitu-
2351–61. [18] Meuleman EJ, Legros JJ, Bouloux PM, et al. Effects of long-term oral
dinal effects of aging on serum total and free testosterone levels in
testosterone undecanoate therapy on urinary symptoms: data from
healthy men. Baltimore Longitudinal Study of Aging. J Clin Endocrinol
a 1-year, placebo-controlled, dose-ranging trial in aging men with
Metab 2001;86:724–31. [3] Schubert M, Jockenhovel F. Late-onset hypogonadism in the aging
symptomatic hypogonadism. Aging Male 2015;18:157–63. [19] Paduch DA, Polzer PK, Ni X, Basaria S. Testosterone replacement in
male (LOH): definition, diagnostic and clinical aspects. J Endocrinol
androgen-deficient men with ejaculatory dysfunction: A random-
Invest 2005;28:23–7.
ized controlled trial. J Clin Endocrinol Metab 2015;100:2956–62.
[4] Wang C, Alexander G, Berman N, et al. Testosterone replacement
[20] Shigehara K, Sugimoto K, Konaka H, et al. Androgen replacement
therapy improves mood in hypogonadal men–a clinical research
therapy contributes to improving lower urinary tract symptoms in
center study. J Clin Endocrinol Metab 1996;81:3578–83.
patients with hypogonadism and benign prostate hypertrophy: a
[5] Wang C, Swerdloff RS, Iranmanesh A, et al. Transdermal testosterone gel improves sexual function, mood, muscle strength, and body
randomised controlled study. Aging Male 2011;14:53–8. [21] Srinivas-Shankar U, Roberts SA, Connolly MJ, et al. Effects of tes-
composition parameters in hypogonadal men. J Clin Endocrinol
tosterone on muscle strength, physical function, body composition,
Metab 2000;85:2839–53.
and quality of life in intermediate-frail and frail elderly men: a
[6] Seidman SN, Rabkin JG. Testosterone replacement therapy for hypogonadal men with SSRI-refractory depression. J Affect Disord 1998;48:157–61. [7] Snyder PJ, Peachey H, Berlin JA, et al. Effects of testosterone replacement in hypogonadal men. J Clin Endocrinol Metab 2000; 85:2670–7. [8] Kenny AM, Kleppinger A, Annis K, et al. Effects of transdermal
randomized, double-blind, placebo-controlled study. J Clin Endocrinol Metab 2010;95:639–50. [22] Tan WS, Low WY, Ng CJ, et al. Efficacy and safety of long-acting intramuscular testosterone undecanoate in aging men: a randomised controlled study. BJU Int 2013;111:1130–40. [23] Basaria S, Harman SM, Travison TG, et al. Effects of testosterone administration for 3 years on subclinical atherosclerosis progres-
testosterone on bone and muscle in older men with low bioavail-
sion in older men with low or low-normal testosterone levels:
able testosterone levels, low bone mass, and physical frailty. J Am
A randomized clinical trial. JAMA 2015;314:570–81.
Geriatr Soc 2010;58:1134–43.
[24] Pechersky AV, Mazurov VI, Semiglazov VF, Karpischenko AI,
[9] Dobs AS, Meikle AW, Arver S, Sanders SW, Caramelli KE, Mazer NA.
Mikhailichenko VV, Udintsev AV. Androgen administration in
Pharmacokinetics, efficacy, and safety of a permeation-enhanced
middle-aged and ageing men: effects of oral testosterone undecano-
testosterone transdermal system in comparison with bi-weekly
ate on dihydrotestosterone, oestradiol and prostate volume. Int J
injections of testosterone enanthate for the treatment of hypogonadal men. J Clin Endocrinol Metab 1999;84:3469–78. [10] Cui Y, Zhang Y. The effect of androgen-replacement therapy on prostate growth: a systematic review and meta-analysis. Eur Urol 2013;64:811–22.
Androl 2002; 25:119–25. [25] Pearl JA, Berhanu D, Franc¸ois N, et al. Testosterone supplementation does not worsen lower urinary tract symptoms. J Urol 2013;190: 1828–33. [26] Barry MJ, Fowler Jr FJ, O’Leary MP, et al. The American Urological
[11] Behre HM, Tammela TL, Arver S, et al. A randomized, double-blind,
Association symptom index for benign prostatic hyperplasia. The
placebo-controlled trial of testosterone gel on body composition
Measurement Committee of the American Urological Association.
and health-related quality-of-life in men with hypogonadal to low-
J Urol 1992;148:1549–57.
normal levels of serum testosterone and symptoms of androgen
[27] Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items
deficiency over 6 months with 12 months open-label follow-up.
for systematic reviews and meta-analyses: the PRISMA statement.
Aging Male 2012;15:198–207.
Ann Intern Med 2009;151:264–9.
[12] Chiang HS, Hwang TI, Hsui YS, et al. Transdermal testosterone gel
[28] Higgins JP, Green S. Cochrane handbook for systematic reviews of
increases serum testosterone levels in hypogonadal men in Taiwan
interventions v.5.1.0. The Cochrane Collaboration; 2011. www.
with improvements in sexual function. Int J Impot Res 2007; 19:411–7. [13] Emmelot-Vonk MH, Verhaar HJ, Nakhai Pour HR, et al. Effect of testosterone supplementation on functional mobility, cognition,
cochrane-handbook.org. [29] Review Manager (RevMan) Computer Program. Version 5.2. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration; 2014.
Please cite this article in press as: Kohn TP, et al. Effects of Testosterone Replacement Therapy on Lower Urinary Tract Symptoms: A Systematic Review and Meta-analysis. Eur Urol (2016), http://dx.doi.org/10.1016/j.eururo.2016.01.043
EURURO-6643; No. of Pages 8 8
EUROPEAN UROLOGY XXX (2016) XXX–XXX
[30] Borenstein M, Hedges LV, Higgins JP, Rothstein HR. A basic intro-
[38] Barry MJ, Williford WO, Chang Y, et al. Benign prostatic hyperplasia
duction to fixed-effect and random-effects models for meta-
specific health status measures in clinical research: how much
analysis. Res Synth Methods 2010;1:97–111. [31] Higgins JP, Thompson SG. Quantifying heterogeneity in a metaanalysis. Stat Med 2002;21:1539–58. [32] Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ 2003;327:557–60. [33] Sterne JA, Juni P, Schulz KF, Altman DG, Bartlett C, Egger M.
change in the American Urological Association symptom index and the benign prostatic hyperplasia impact index is perceptible to patients? J Urol 1995;154:1770–4. [39] Saad F, Aversa A, Isidori AM, Zafalon L, Zitzmann M, Gooren L. Onset of effects of testosterone treatment and time span until maximum effects are achieved. Eur J Endocrinol 2011;165:675–85.
Statistical methods for assessing the influence of study character-
[40] Bhasin S, Cunningham GR, Hayes FJ, et al. Testosterone therapy in
istics on treatment effects in ‘meta-epidemiological’ research. Stat
men with androgen deficiency syndromes: an Endocrine Society
Med 2002;21:1513–24.
clinical practice guideline. J Clin Endocrinol Metab 2010;95:
[34] van Houwelingen HC, Arends LR, Stijnen T. Advanced methods in meta-analysis: multivariate approach and meta-regression. Stat Med 2002;21:589–624. [35] Sterne JA, Egger M. Funnel plots for detecting bias in meta-analysis: guidelines on choice of axis. J Clin Epidemiol 2001;54:1046–55.
2536–59. [41] Seftel AD, Kathrins M, Niederberger C. Critical update of the 2010 Endocrine Society Clinical Practice Guidelines for Male Hypogonadism: A systematic analysis. Mayo Clin Proc 2015;90: 1104–15.
[36] Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-
[42] Kathrins M, Doersch K, Nimeh T, Canto A, Niederberger C, Seftel A.
analysis detected by a simple, graphical test. BMJ 1997;315:629–34.
The Relationship between Testosterone Replacement Therapy and
[37] R-Core-Team. R: A language and environment for statistical com-
Lower Urinary Tract Symptoms: A systematic review. Urology. In
puting. Vienna, Austria: R Foundation for Statistical Computing;
press. http://dx.doi.org/10.1016/j.urology.2015.11.006.
2012.
Please cite this article in press as: Kohn TP, et al. Effects of Testosterone Replacement Therapy on Lower Urinary Tract Symptoms: A Systematic Review and Meta-analysis. Eur Urol (2016), http://dx.doi.org/10.1016/j.eururo.2016.01.043