Wierman ME. Risks of Different Testosterone Preparations: Too Much, Too Little, Just Right. JAMA Intern Med. Published online May 11, 2015. http://archinte.jamanetwork.com/article.aspx?articleID=2293076
There is a developing story that the increased use of testosterone therapy without a diagnosis of hypogonadism is associated with an elevated cardiovascular risk for some patients. In this issue of JAMA Internal Medicine, Layton and coauthors1 deepen our knowledge with their analysis of 3 large databases (Truven MarketScan and Medicare cohorts from the United States and the Clinical Practice Research Datalink cohort from the United Kingdom) to ask whether they could detect a difference in cardiovascular or cerebrovascular events and mortality in patients newly prescribed testosterone therapy in the form of injectable testosterone compared with testosterone gels or patches.
The strengths of the study include the large number of men in the data sets given testosterone therapy (>430 000), patient diversity, and difference in provider prescribing habits with stricter use of testosterone therapy in the United Kingdom compared with the United States. Layton et al1 examined cardiovascular event risk within the first year after testosterone prescription because a prior clinical trial2 and several large epidemiologic studies3,4 suggest the association is observed within this time frame. This new large data analysis demonstrated that prescription of injectable depo-testosterone was associated with a higher overall cardiovascular risk, including myocardial infarction, unstable angina, and stroke (hazard ratio
, 1.26; 95% CI, 1.18-1.35), hospitalization (HR, 1.16; 95% CI, 1.13-1.19), and death (HR, 1.34; 95% CI, 1.15-1.56) compared with testosterone gel or patch formulations. The outcomes of the analysis support the hypothesis that intermittent pharmacologic levels of testosterone, which are inevitable with an injectable depo-testosterone administration every 2 to 3 weeks, result in an association with an increase in the combined outcome of cardiovascular and cerebrovascular morbidity and death. The absolute risk between testosterone therapy groups was small in this epidemiologic analysis, but the risk does raise concern considering the marked increase in testosterone prescriptions both in the United States and internationally5 and prior associations in large data sets of increased risks in men receiving vs those not receiving treatment.3,4 In addition, Layton and colleagues1 did not include a comparison group of men not receiving testosterone, so overall risks and benefits of the therapy cannot be compared between the groups.
Layton et al1 acknowledge the limitations of the study, including lack of information concerning the indication for treatment and only partial data concerning testosterone levels before initiating or during therapy. However, their findings are consistent with those of other large epidemiologic studies3,4 and the recent alarming US Food and Drug Administration report that more than 25% of testosterone prescriptions in the United States are written without determination of a level and that more than 30% of patients receiving testosterone therapy do not have follow-up laboratory testing.5 Taken together, these reports suggest that testosterone therapy is often being prescribed off label as an anabolic steroid and not for a diagnosis of hypogonadism, which requires signs and symptoms together with a testosterone level below the reference range.6 The results from the analysis by Layton and coworkers1 are consistent with most, but not all, recent epidemiologic data sets6; together, these studies can inform both patients and health care professionals. It is somewhat reassuring that, more recently, there was an increase in testosterone gel and patch prescriptions over injectable prescriptions since those transdermal formulations have been noted to carry a smaller risk of pharmacologic dosing unless at dosages prescribed above the label indications.
The other important outcome in the data set was the lack of evidence for a signal of an increased association of venous thromboembolism with testosterone therapy (hazard ratio, 0.92; 95% CI, 0.76-1.11). The prior studies showing an increased risk of venous thromboembolism with testosterone use, which influenced the US Food and Drug Administration to add a comment to the prescribing label, most likely reflect either patients given pharmacologic testosterone, which is converted by aromatase into high levels of estrogens, or those rare patients with underlying genetic hypercoagulable risk.
As with all epidemiologic studies, the results reported by Layton and colleagues1 show associations, not necessarily causal relationships. There were limited data on whether testosterone therapy was prescribed for hypogonadism, whether the testosterone levels were low on initiation of therapy or at follow-up, or whether men who received injectable formulations were somehow at higher cardiovascular risk than those given gels or patches.
The new data provided by Layton et al1 support the Endocrine Society guidelines,6 which recommend testosterone therapy for the treatment of hypogonadism rather than just for specific symptoms alone or for a testosterone laboratory-determined cutoff level. Careful evaluation of each patient for an underlying cause that may be reversible is needed, as well as treatment with physiologic testosterone replacement strategies with close monitoring and follow-up to evaluate risk and benefit. The results of the Layton et al1 study can reassure physicians who rationally provide treatment for men with true hypogonadism with approaches that result in physiologic levels of testosterone, which is a safe and effective therapy. Finally, the study raises the issue of whether injectable depo-testosterone or other formulations that consistently result in levels outside the physiologic range should be restricted or at least more carefully monitored for cardiovascular risk.
Abundant epidemiologic data suggest that both very low and high testosterone levels are associated with increased cardiovascular risk. Thus, as with Goldilocks and her porridge, too much is bad, too little is bad, and just right is needed when it comes to testosterone treatment strategies in men with hypogonadism.