Series: The Heart Scare That Almost Killed TRT
Subtitle: How a 2010 study, a flawed meta-analysis, and a media panic nearly derailed testosterone therapy for a generation
Post: 4 of 5
Tags: TRAVERSE trial, NEJM 2023, cardiovascular safety, TRT, MACE, non-inferiority trial, hypogonadism
Word Count: ~1,200
When the medical community finally got the large randomized trial it had been demanding since 2010, it arrived in June 2023 in the pages of the New England Journal of Medicine — arguably the most prestigious medical journal in the world — and it carried enough statistical weight to end the cardiovascular debate that had been running for thirteen years.
The TRAVERSE trial. Five thousand two hundred and forty-six men. Thirty-three months of average follow-up. A pre-specified primary endpoint. And a result that was, depending on how you interpret “the system working,” either a vindication of testosterone therapy, a vindication of scientific patience, or a fairly blistering indictment of how one flawed 2013 observational study shaped a decade of clinical practice.
Possibly all three.
What the Trial Was Designed to Do
The TRAVERSE trial — Testosterone Replacement therapy for Assessment of long-term Vascular Events and efficacy ResponSE in hypogonadal men — was a multicenter, randomized, double-blind, placebo-controlled trial. It was funded by AbbVie, the manufacturer of AndroGel, which created some pre-publication skepticism, though the trial design included independent adjudication of outcomes and a prespecified statistical analysis plan registered before results were unblinded.
The trial enrolled 5,246 men between 2018 and 2021 at sites across the United States. Eligibility criteria were specific:
- Men aged 45 to 80 years
- Confirmed hypogonadism, defined as two morning testosterone readings below 300 ng/dL (10.4 nmol/L)
- Established cardiovascular disease, OR at least two cardiovascular risk factors (such as hypertension, dyslipidemia, obesity, impaired fasting glucose, or current smoking)
This population was not chosen accidentally. It was designed to directly address the critique raised by the Basaria 2010 study — that small samples in extreme populations couldn’t speak to the general population — by enrolling a large sample of men who were actually at meaningful cardiovascular risk. If testosterone therapy was going to harm hearts, this was the population where you’d expect to see it.
The intervention: testosterone 1.62% transdermal gel, titrated to maintain testosterone levels in the 350–750 ng/dL range, versus placebo gel. Both groups received identical-looking packaging and identical application instructions.
The Primary Endpoint: MACE
The trial’s primary endpoint was major adverse cardiovascular events (MACE) — a composite of death from cardiovascular causes, non-fatal myocardial infarction (heart attack), and non-fatal stroke. This is the standard composite endpoint used in cardiovascular outcome trials across medicine, from statin studies to antidiabetic drug trials.
The trial was designed as a non-inferiority trial — which is a specific and important statistical framework worth explaining, because the terminology confused some early media coverage.
A superiority trial asks: does Treatment A produce better outcomes than Treatment B? A non-inferiority trial asks a more modest question: is Treatment A not meaningfully worse than Treatment B, within a pre-defined margin? The margin for TRAVERSE was set at a hazard ratio of 1.5 — meaning the trial was powered to determine whether testosterone gel increased MACE risk by more than 50% relative to placebo.
Choosing non-inferiority rather than superiority as the design was appropriate here, because nobody expected testosterone to protect against heart attacks (that would require a superiority design). The question was whether it caused them.
The Results
After an average follow-up of approximately 33 months, the primary endpoint results were:
- MACE occurred in 7.0% of men in the testosterone group
- MACE occurred in 7.3% of men in the placebo group
- Hazard ratio: 0.96 (95% confidence interval: 0.78 to 1.17)
The upper bound of the confidence interval — 1.17 — was well below the pre-specified non-inferiority margin of 1.50. The trial achieved its primary endpoint.
Testosterone therapy was non-inferior to placebo for major adverse cardiovascular events.
To put it more plainly: in a carefully conducted randomized trial of over 5,000 men with hypogonadism and meaningful cardiovascular risk, followed for nearly three years, testosterone therapy did not increase the rate of heart attacks, strokes, or cardiovascular deaths relative to placebo.
Secondary and Safety Findings
The TRAVERSE trial also evaluated a range of secondary endpoints and safety outcomes. Several findings are worth noting:
Pulmonary embolism and deep vein thrombosis: Testosterone-treated men had a slightly higher rate of pulmonary embolism (0.9% vs. 0.5%). This difference was statistically significant and is consistent with the known effect of testosterone on erythropoiesis (red blood cell production) and hematocrit. Men on testosterone should be monitored for polycythemia.
Atrial fibrillation: Testosterone-treated men had a statistically higher rate of new-onset atrial fibrillation (3.5% vs. 2.4%). The clinical significance of this finding — particularly in men without preexisting AF — remains an area of ongoing discussion.
Prostate safety: The trial found that testosterone therapy was associated with a higher rate of prostate biopsy and a higher rate of prostate cancer diagnosis, though the absolute numbers were small. This finding reinforced the existing recommendation that men on TRT require regular PSA monitoring. (This topic deserves its own dedicated coverage — prostate safety in TRT is a distinct discussion from cardiovascular safety.)
Fractures, diabetes, and sexual function: Secondary endpoints showed improvements in sexual desire and function in the testosterone group, consistent with the established efficacy of TRT. Effects on fracture rates and glycemic control were not statistically significant in this follow-up window.
What “Non-Inferior” Actually Means
The language of “non-inferiority” occasionally gets misread as “barely safe” or “not quite good.” It is worth being precise. In a regulatory and clinical trial context, establishing non-inferiority within a pre-specified margin is considered a rigorous and clinically meaningful result. It is the same standard applied to many commonly used medications where the goal is to show that a new treatment doesn’t cause harm rather than to show that it provides additional benefit.
In the TRAVERSE context, non-inferiority means: when you take 5,246 men with documented hypogonadism and meaningful cardiovascular risk, randomly assign half to testosterone treatment for three years, and count the heart attacks, strokes, and cardiovascular deaths — you don’t get more of them in the testosterone group. The signal the field had feared since 2010 was not present in a trial designed and powered to detect it.
The FDA Response
Following the TRAVERSE publication, the FDA revised its prescribing guidance. While the agency did not eliminate all cardiovascular language from testosterone labels, it removed the most aggressive warning language that had been added in 2014 and updated its guidance to reflect the TRAVERSE findings. The regulatory framing shifted from “potential cardiovascular harm” to a more nuanced acknowledgment of the evidence from the large randomized trial.
Next up: Post 5 — “What TRAVERSE Means for TRT Today: Settled Questions, Open Questions, and What Men on Testosterone Should Actually Know.”
