SCIENCE@home

Oral Phosphodiesterase-5 Inhibitors and Hormonal Treatments for Erectile Dysfunction: A Systematic Review and Meta-analysis

A. Tsertsvadze,H. Fink,F. Yazdi,R. MacDonald,A. Bella,M. Ansari,C. Garritty,K. Soares-Weiser,R. Daniel,M. Sampson,Steven Fox,D. Moher,T. Wilt

Published 2009 in Annals of Internal Medicine

ABSTRACT

Erectile dysfunction (ED) is a common male sexual disorder and is defined as the persistent inability to achieve or maintain penile erection sufficient for satisfactory sexual performance (1). Advanced age, diabetes, vascular diseases, psychiatric disorders, and possibly hypogonadism are associated with increased prevalence of ED (14). According to data from the Massachusetts Male Aging Study (2), the prevalence of ED in men aged 40 to 70 years is about 50%. Unless contraindicated, oral phosphodiesterase-5 (PDE-5) inhibitors, such as sildenafil, tadalafil, or vardenafil, are currently first-line treatments of ED (5). Alternative treatments include hormones, vacuum constriction devices, intraurethral suppositories, intracavernosal injections, or surgery (for example, penile prosthesis implants) (57). Direct comparisons between treatments have been limited, and the relative long-term efficacy and safety profiles of different therapies have not been adequately explored. Estimates from the National Health and Nutrition Examination Survey suggested that annual U.S. treatment costs of ED could reach $15 billion if all affected men sought care (8). In the past decade, the use of diagnostic tests for underlying causes of ED markedly decreased and use of pharmacologic therapy, especially with oral PDE-5 inhibitors, increased (8). In 2005, sales of sildenafil, tadalafil, and vardenafil were $1.6 billion, $747 million, and $327 million, respectively (911). The value of hormonal blood tests (such as testosterone) in routine evaluation of men with ED is uncertain (1215). The European Association of Urology and the British Society for Sexual Medicine guidelines recommend endocrinologic screening in the initial evaluation of all men with ED (1520). In contrast, the American Urological Association recommends that hormone testing in men with ED be based on initial clinical assessment (for example, decreased libido, small testes, and reduced body hair) or failure of initial PDE-5 therapy management (5). Whether this targeted approach for identifying and treating hormonal disorders as underlying causes of ED is appropriate has not been rigorously evaluated (13, 15, 20, 21). The aims of our review were to systematically identify and synthesize the published evidence to determine 1) the relative benefits and harms of oral PDE-5 inhibitors and hormonal treatments of ED, including the effect of patient characteristics and comorbid conditions on the likelihood of treatment success and 2) the clinical value of hormonal blood testing for identifying treatable causes of ED (for example, hypogonadism) and improving its outcomes. We summarized and updated evidence from a technical review prepared for the Agency for Healthcare Research and Quality (22) and adapted the article in collaboration with the American College of Physicians' Clinical Efficacy Assessment Subcommittee to inform the development of its clinical practice guideline on this topic. The article focuses more specifically than the Agency for Healthcare Research and Quality technical review on treatments likely to be prescribed by primary care physicians. The topic of diagnosis and treatment of ED, as a subject for systematic review, was originally nominated by the American College of Physicians. Methods Data Sources and Searches We searched for English-language articles in MEDLINE (1966 to May 2007), EMBASE (1980 to week 22 of 2007), Cochrane Central Register of Controlled Trials (second quarter of 2007), PsycINFO (1985 to June 2007), AMED (1985 to June 2007), and SCOPUS (2006). Search terms were impotence; erectile dysfunction; randomized, controlled trial; and controlled clinical trial. We also scanned reference lists of retrieved publications. We updated the review by searching MEDLINE and EMBASE (May 2007 to April 2009). Study Selection To assess the relative benefits and harms of pharmacologic treatments for ED, we selected randomized, controlled trials (RCTs) of pharmacologic treatments in men aged 18 years or older with ED. Treatments not generally prescribed by primary care physicians, such as vacuum constriction devices, intraurethral suppositories, intracavernosal injections, or psychotherapy, were considered beyond the scope of this review and are addressed in the technical report. To assess the risks for nonarteritic anterior ischemic optic neuropathy (NAION) in PDE-5 inhibitor users, we selected RCTs; nonrandomized, controlled trials; and observational studies. To assess the clinical value of routine hormonal blood tests in men with ED, we selected studies that reported prevalence of hypogonadism, hyperprolactinemia, or both in men with ED and all RCTs comparing hormone treatment alone or in combination versus control in men with ED. We excluded reviews, pooled analyses, editorials, commentaries, and letters. Two independent reviewers screened all identified titles and abstracts and, for articles considered potentially eligible, abstracted their full-text reports. Discrepancies were discussed and resolved by consensus. Appendix Figure 1 shows the literature selection process. Appendix Figure 1. Literature search and selection. NAION = nonarteritic anterior ischemic optic neuropathy. Data Extraction and Quality Assessment Two reviewers independently abstracted data on study, population, and treatment characteristics. Treatment efficacy outcomes were the proportion of successful sexual intercourse attempts based on either participants' diaries or event logs (erection sufficiently hard and long-lasting for satisfactory intercourse) or participants' responses to question 3 of the Sexual Encounter Profile (SEP) (erection lasted long enough for successful intercourse) (23); the improvement in erectile function based on either participants' self-reports of improved erections (global assessment or efficacy question 1) or the mean Erectile Function domain score of the International Index of Erectile Function (IIEF) (24); and participants' responses to IIEF questions 3 (successful penile penetration) and 4 (maintenance of erection after penetration) (24). Abstracted adverse events data were the number of patients with any adverse event, specific adverse events, withdrawals due to adverse events, serious adverse events, and serious cardiovascular adverse events. We assessed the prevalence of hypogonadism or hyperprolactinemia by using the definitions provided by study authors, even though these may have differed between studies. We evaluated the overall strength of evidence by using a method developed by the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) group (25). We used the Jadad scale to assess the methodological and reporting quality of RCTs (score range, 0 to 5; higher score indicates better quality) (26). We judged the adequacy of allocation concealment to be adequate, inadequate, or unclear by using the approach proposed by Schulz and Grimes (27). We assessed the quality of studies reporting serum hormonal levels by using a subset of the Quality Assessment Tool of Diagnostic Accuracy Studies (questions 1, 2, 8, 12, and 14) (28). To explore the overall risk for bias, we generated risk-for-bias graphs (29). Appendix A includes all evidence and quality assessment tables and risk-for-bias graphs for the included studies. Supplement 1. Appendix A Data Synthesis and Analysis We qualitatively summarized data on study (design, reporting quality, and sample size), population (age, severity of ED, and comorbid conditions), and treatment characteristics (dose, frequency, and duration). We considered studies suitable for pooling if they used the same design (RCT), enrolled similar populations (trials restricted to participants with a specific comorbid condition vs. those enrolling participants with a heterogeneous profile of comorbid conditions), evaluated the same type of treatment (for example, sildenafil), and reported the same efficacy or safety outcomes. We used DerSimonian and Laird random-effects models to generate pooled estimates of relative risks (RRs) and weighted mean differences (WMDs) with 95% CIs (30). To avoid double-counting during pooling of a trial with several PDE-5 dose groups versus placebo, we used a generic inverse variance method to combine data from these groups for a single estimate of mean response rate versus placebo. Statistical heterogeneity was evaluated using a chi-square test and the I 2 statistic (low= 25.0%; moderate= 50.0%; high= 75.0%) and was explored through subgroup and sensitivity analyses (for example, study quality and random- or fixed-effects model) (31). We defined the subgroups a priori with respect to severity (mild, severe, or moderate) and cause (psychogenic, mixed, or organic) of ED, treatment (for example, dose [50 mg or 100 mg], dosing [fixed or flexible], type [sildenafil, tadalafil, or vardenafil], and duration [12 weeks vs. >12 weeks]), and underlying or concurrent condition (for example, diabetes, cardiovascular disease, depression, or prostatectomy). When studies did not adequately report summary statistics (such as treatment group mean score and SD), we calculated the needed variables if data for individual patients were reported. If a study reported only an SE of the mean response, we converted it to an SD. Trials were not incorporated into meta-analyses if the needed data (means [SDs]) could not be derived, and crossover trials did not report precrossover data. We included trials with no events for harms in the meta-analyses. We examined the extent of publication bias through visual inspection of funnel plot asymmetry (32) and linear regressionbased tests proposed by Egger and colleagues (33). We performed analyses with R software, version 2.4.0 (www.r-project.org), and STATA software, version 11 (StataCorp, College Station, Texas). Role of the Funding Source The Agency for Healthcare Research and Quality provided funding. The funding source suggested the initial research questions and p

PUBLICATION RECORD

CITATION MAP

EXTRACTION MAP

CLAIMS

  • No claims are published for this paper.

CONCEPTS

  • No concepts are published for this paper.

REFERENCES

Showing 1-100 of 236 references · Page 1 of 3

123
Next

CITED BY

Showing 1-100 of 190 citing papers · Page 1 of 2

12
Next