FDA reviewed five observational studies4-8 and two meta-analyses of placebo-controlled trials9,10 to examine the risk of cardiovascular events associated with testosterone replacement therapy (TRT). The five observational studies were retrospective cohort studies that reported conflicting results. Two of these studies found statistically significant cardiovascular harm with TRT (Vigen and Finkle), 4-5 two studies found a statistically significant mortality benefit with TRT (Shores and Muraleedharan), 6-7 and one study was inconclusive (Baillargeon).8
The Vigen study evaluated male veterans who underwent angiography and had low testosterone concentrations. On average, testosterone-treated men were 64 years old and untreated men were 61 years old. This study found an increased risk with TRT compared to no TRT for the composite cardiovascular outcome of myocardial infarction, stroke, and death (Hazard Ratio [HR]=1.29, 95% Confidence Interval [CI]: 1.04-1.58).4
The Finkle study evaluated TRT users in a large claims database. The men included in this study were on average 54 years old. This study found an increased risk of non-fatal myocardial infarction during the 90 days following an initial prescription for TRT compared to the pre-TRT period (Relative Risk [RR]=1.36, 95% CI: 1.03-1.81).5
The Shores study evaluated a population of male veterans older than 40 years of age with low testosterone and found a decreased risk of all-cause mortality with TRT compared to no TRT (HR=0.61, 95% CI: 0.42-0.88).6
The Muraleedharan study evaluated men with type 2 diabetes in the United Kingdom. The main analysis assessed mortality in men with low serum testosterone concentrations compared to men with normal serum testosterone concentrations. Mortality was also assessed in a subgroup analysis of treated and untreated men with low serum testosterone; an increased risk of all-cause mortality in men with no TRT compared to those on TRT was found (HR=2.30, 95% CI: 1.30-3.90).7
Finally, the Baillargeon study evaluated men older than 65 years of age enrolled in Medicare and found no overall increase in risk of hospitalization for myocardial infarction when comparing those treated with TRT to those receiving no TRT (HR=0.84, 95% CI: 0.69-1.02).8
The Xu meta-analysis involved 27 published, randomized, placebo-controlled trials representing 2,994 mostly middle-aged and older male participants (1,773 treated with testosterone and 1,261 treated with placebo) who reported 180 cardiovascular-related adverse events.9 This study found that testosterone therapy was associated with an increased risk of adverse cardiovascular events (Odds Ratio [OR]=1.5, 95% CI: 1.1-2.1); however, methodological issues limit conclusions. These limitations include inconsistent and incomplete reporting of adverse events; substantial clinical heterogeneity in the design and conduct of the component trials and the types of cardiovascular outcomes included in the analyses; potential bias resulting from selection of component trials; and variable quality of the trials, particularly with regard to ascertainment of cardiovascular safety outcomes and balance in cardiovascular risk factors and discontinuation rates across study arms.
The Corona meta-analysis involved 26 published, randomized, controlled trials, 20 of which were also included in the Xu meta-analysis. The included studies represented 3,236 men (1,895 men treated with testosterone, 1,341 men treated with placebo) who reported 51 major adverse cardiovascular events, defined as cardiovascular death, non-fatal myocardial infarction or stroke, and serious acute coronary syndromes or heart failure.10 This study did not find a statistically significant increased risk of these cardiovascular events associated with testosterone treatment. Similar to the first meta-analysis, this study had methodological issues that limit conclusions. These issues include incomplete adverse event reporting in the published trials, clinical trial heterogeneity, possible treatment arm imbalances in cardiac risk factors, high or unbalanced discontinuation rates in some component trials, and the potential for bias in trial selection and interpretation of reported adverse events.
The five observational studies and the Xu meta-analysis were discussed at a joint meeting of the Bone, Reproductive and Urologic Drugs Advisory Committee and the Drug Safety and Risk Management Advisory Committee on September 17, 2014. Based on these findings, the advisory committee members were in general agreement that the signal of cardiovascular risk is weak and that only a prospective, well-controlled clinical trial could determine whether testosterone causes cardiovascular harm. The Corona study was recently published and could not be reviewed in time to be presented at the Advisory Committee meeting; however, we have reviewed the study and factored its findings into our overall assessment.
For complete reviews, background information, and minutes of the September 17, 2014, Advisory Committee meeting, click here.
Here is the original post:
FDA Drug Safety Communication: FDA cautions about using testosterone ...
