Read the complete EBM focus and earn CME
Reference: (JAMA 2013 Nov 6;310(17):1829) (level 2 [mid-level] evidence)
Testosterone replacement therapy has become increasingly common in recent years and has shown efficacy for improving sexual dysfunction, bone mineral density, exercise capacity, and metabolic parameters in men with androgen deficiency or chronic conditions including heart failure and diabetes. However, efficacy trials have generally been of relatively short duration, which may limit the ability to assess potential harms. A recent 6-month randomized trial evaluating testosterone gel in frail elderly men was terminated early due to increased rates of cardiovascular events in the testosterone group (N Engl J Med 2010 Jul 8;363(2):109) raising concerns about whether this was specific to this trial or might have been reflect a larger problem. Now, a new retrospective cohort study has assessed long-term risks of testosterone therapy in men with low testosterone levels and high cardiovascular risk.
A total of 8,709 men who visited a Veteran Administration facility for coronary angiography from 2005 to 2011 and subsequently had serum testosterone < 300 ng/dL (< 10.4 nmol/L), were followed for mean 27.5 months. At the time of angiography, 20% had a previous myocardial infarction (MI), 50% had diabetes. More than 80% had at least 20% stenosis in ≥ 1 epicardial vessel on angiography. Following angiography, 14% began testosterone therapy at some point during the follow-up period (63.3% had testosterone patch, 35.7% had injection, and 1.1% had gel). The median time from angiography to beginning of therapy was 531 days. The primary outcome was a composite of all-cause mortality, MI and ischemic stroke.
There were a number of baseline differences between men who had testosterone therapy and men who did not. The no-testosterone group was older (mean difference 3 years) and had significantly higher rates of coronary artery obstruction, hypertension, hyperlipidemia, heart failure, previous MI, COPD, peripheral vascular disease and cerebrovascular disease. In an analysis weighted by the probability of treatment with testosterone and adjusted for comorbidities, testosterone therapy was associated with an increased risk of primary outcome events (adjusted hazard ratio 1.29, 95% CI 1.04-1.58). At 3 years, the cumulative event rates were 25.7% in testosterone group vs. 19.9% in no-testosterone group. These data are consistent with the results of a recent systematic review assessing the effects of testosterone therapy in 2,994 with low testosterone or chronic diseases (BMC Med 2013 Apr 18;11:108).
While it may be difficult to minimize bias in an observational study of this type, results from studies like this will often prompt longer follow-up in subsequent randomized trials. That would certainly be helpful in this case. However, until such data are available, this information and its limitations should be part of the discussion for men considering testosterone replacement therapy.
For more information, see the Testosterone therapy in men topic in DynaMed.