|Adverse events associated with testosterone replacement in middle-aged and older men: a meta-analysis of randomized, placebo-controlled trials
|Calof OM, Singh AB, Lee ML, Kenny AM, Urban RJ, Tenover JL, Bhasin S
This review evaluated the risk of adverse events associated with testosterone replacement in older men. The authors concluded that testosterone replacement is associated with a significantly higher risk of prostate events and of a haematocrit of more than 50% relative to placebo. A limited search, unclear review methodology, and the unknown quality of the primary studies limit the reliability of the conclusions.
To determine the risk of adverse events associated with testosterone replacement in older men.
The authors searched MEDLINE from 1966 to April 2004 for both English and non-English articles; the search terms were stated. In addition, experts searched publications for further relevant papers and letters were sent to authors to clarify information provided in the publications.
Study designs of evaluations included in the review
Double-blind, randomised controlled trials (RCTs) lasting at least 90 days were eligible for inclusion. The mean duration of the included studies was 10 months (range: 90 days to 3 years).
Specific interventions included in the review
Studies that investigated the use of testosterone or its esters in replacement doses were included in the review. Studies that used androgens other than testosterone, or that used supraphysiologic doses of testosterone, were excluded. Studies that included medications other than testosterone were also excluded, unless they had a clearly defined testosterone-only arm. Testosterone dose and formulation varied across studies: formulations included testosterone patch, injectable testosterone esters, testosterone gel and oral testosterone undecanoate.
Participants included in the review
Men aged 45 years or older with low or low-normal testosterone levels, and without an acute illness, were included in the review. Studies of participants who had unstable disease conditions, or who were infected with the human immunodeficiency virus, were excluded. Where stated, the mean age of the participants in the included studies ranged from 51.9 to 81.0 years. One study that included both younger and older men, but where the majority of the population was over 45 years, was also included in the review. Most studies were in asymptomatic men selected by level of testosterone.
Outcomes assessed in the review
The primary review outcome measure was the all-cause prostate event rate. Prostate events included prostate biopsies, prostate cancers, increase in International Prostate Symptom Score (IPSS) of greater than 4, prostate-specific antigen (PSA) of greater than 4 ng/mL or a PSA increment of at least 1.5 ng/mL during treatment, and acute urinary retention. Cardiovascular outcomes included all cardiovascular events, myocardial infarction, chest pain or ischaemia, coronary procedure or coronary artery bypass graft, and vascular events or cerebrovascular accidents. Other categories of adverse events that were assessed were haematological, lipids, respiratory and/or sleep apnoea, other adverse events and death.
How were decisions on the relevance of primary studies made?
The authors did not state how the papers were selected for the review, or how many reviewers performed the selection.
Assessment of study quality
Study quality was assessed according to randomisation, concealment, blinding, and whether the control group was placebo-treated or not. The authors did not state how the papers were assessed for quality, or how many reviewers performed the quality assessment.
Three reviewers extracted the data. The rates of each of the seven categories of adverse event were extracted from each study.
Methods of synthesis
How were the studies combined?
A pooled event rate per 1,000 patient-years was calculated for each outcome for each treatment group, based on an average study duration of 10 months. Odds ratios (ORs) were pooled using a random-effects model, weighted for sample size. The Clopper-Pearson method was used to calculate 95% confidence intervals (CIs).
How were differences between studies investigated?
Differences between the studies were discussed in the text and forest plots were presented for some meta-analyses. A sensitivity analysis was carried out after excluding 6 studies. Studies were excluded because two did not use a placebo-control, three included participants receiving glucocorticoids, and one included participants with chronic angina pectoris.
Results of the review
Nineteen RCTs (1,084 men) were included in the review.
There were significantly more prostate events in men receiving testosterone than in men receiving placebo (OR 1.78, 95% CI: 1.07, 2.95). There was no statistically significant difference between treatment groups in terms of rates of prostate cancer, increases in PSA levels, prostate biopsies, or increase in IPSS scores.
There was no statistically significant difference between treatment groups in the rate for all cardiovascular events.
An increase in haematocrit over 50% was the most common testosterone-related adverse event. There was a significantly higher number of participants with haematocrit over 50% in the testosterone group than in the placebo group (OR 3.69, 95% CI: 1.82, 7.51).
Of the 11 studies that reported plasma lipid values, four reported a significant decrease and seven reported no significant change in high-density lipoprotein cholesterol levels, four found a decrease in total cholesterol levels, and two found a decrease in low-density lipoprotein cholesterol level in men receiving testosterone.
The 4 studies that used a testosterone patch reported a high frequency of skin irritation at the application site (range: 17 to 40%). The frequency of men with a new diagnosis of sleep apnoea during treatment was not statistically significantly different between treatment groups. No deaths were reported in the testosterone-treated group, while two were reported in the placebo-treated group. Other adverse events were reported.
The exclusion of six studies from the sensitivity analysis did not change the findings.
Testosterone replacement in older men is associated with a significantly higher risk of identifying prostate events and a haematocrit of more than 50% relative to placebo. The increase in haematocrit was the most common adverse event associated with testosterone replacement. The data confirm the need to monitor haematocrit and PSA, and for digital examination of the prostate during testosterone replacement in older men.
The authors set out a clear question at the beginning of the review, with inclusion criteria clearly defined in terms of the participants, interventions, outcomes and study design. Only one database was searched, which increased the possibility that relevant studies might have been missed. In addition, publication bias was not assessed. However, the search was not restricted to English language papers, which helps to reduce the risk of language bias. It was unclear whether measures were taken to reduce bias in the study selection and quality assessment processes. Although three reviewers extracted data, it was not clear whether the extraction was carried out in duplicate or independently. The quality of the studies was assessed on appropriate criteria; however, the results of the assessment were not reported. Although only double-blind studies were eligible for inclusion in the review, it was unclear whether all included studies met this criterion.
Statistical heterogeneity was not assessed and, although forest plots were presented for some meta-analyses, no comments about statistical heterogeneity were made. In addition, the authors pointed out several sources of clinical heterogeneity and bias, which suggested that statistical pooling of the studies might not have been appropriate.
The limited search, incomplete reporting of review methods, and lack of reporting of a quality assessment mean that the reliability of the authors' conclusions is unclear.
Implications of the review for practice and research
Practice: The authors stated that older men who decide to receive testosterone treatment should be warned about the increased risk of prostate events and increased haematocrit, and should be monitored by periodic evaluation of haematocrit and PSA and by digital rectal examination of the prostate.
Research: The authors stated that sources of bias that could contribute to the observed frequency of prostate events in testosterone-treated men should be taken into account, and strategies for minimising bias should be incorporated in the design of testosterone trials. The authors stated that studies of longer duration would be needed to determine the effects of testosterone on prostate and cardiovascular risk.
Calof OM, Singh AB, Lee ML, Kenny AM, Urban RJ, Tenover JL, Bhasin S. Adverse events associated with testosterone replacement in middle-aged and older men: a meta-analysis of randomized, placebo-controlled trials. Journals of Gerontology Series A - Biological Sciences and Medical Sciences 2005; 60(11): 1451-1457
Subject indexing assigned by NLM
Aged; Cardiovascular System /drug effects; Hematocrit; Hormone Replacement Therapy /adverse effects; Lipids /blood; Male; Middle Aged; Placebos; Prostate /drug effects; Sensitivity and Specificity; Testosterone /administration & dosage /adverse effects
Date bibliographic record published
Date abstract record published
This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.