Testosterone therapy and prostate cancer: what the trials now show

The fear that testosterone therapy causes prostate cancer, which made doctors wary of the treatment for most of a century, is not supported by the trials that have since tested it [s1][s2]. In screened men with low testosterone, randomised evidence shows no increase in prostate cancer, and the largest safety analysis to date found the rates of prostate cancer and other prostate events no higher on testosterone than on placebo [s1]. The caveats are real and worth stating, but the headline claim — treat the deficiency and you seed a tumour — has not held up.

Where the fear came from

The worry has a respectable origin: androgen-deprivation therapy shrinks advanced prostate cancer, so it seemed to follow that adding testosterone would do the reverse and accelerate it [s3]. In 2009 a review in European Urology argued that the logic breaks down at normal physiology [s3]. Its "saturation model" held that the androgen receptors driving prostate tissue are fully occupied at fairly low testosterone concentrations; above that point, extra testosterone has little further to grip, so raising a deficient man's level back into the normal range does not keep stimulating growth in step [s3]. That reframing — the prostate responds to androgen deprivation, but not symmetrically to androgen addition above a low threshold — is what the trials went on to test.

What the largest trial found

TRAVERSE, reported in 2023, is the reason the question can now be answered with more than registries. It randomised 5,246 men aged 45 to 80 who had low testosterone and either established cardiovascular disease or a high risk of it, assigning them to daily testosterone gel or placebo, with a mean follow-up of 33 months [s4]. It was built to test cardiovascular safety, which it met — a primary cardiac event occurred in 7.0% on testosterone versus 7.3% on placebo [s4]. But its size, length and formal adjudication also made it the best prostate-safety dataset available.

That prostate analysis, published in 2024, found that among these carefully screened men the incidences of high-grade or any prostate cancer, of acute urinary retention, of surgery for benign enlargement, of prostate biopsy and of new drug treatment for urinary symptoms were low and did not differ between the testosterone and placebo groups [s1]. Testosterone did not worsen lower-urinary-tract symptoms [s1]. The one consistent signal was that testosterone raised prostate-specific antigen more than placebo during the first year of treatment — a change that, left unmonitored, can trigger biopsies rather than reflect cancer [s1]. The authors also noted the counter-intuitive backdrop: men with untreated low testosterone have a reduced risk of dying from prostate cancer, not a raised one [s1].

Pooling the trials

A 2026 systematic review drew together 41 randomised trials covering 11,161 men and reached the same place [s2]. Testosterone therapy was not associated with a significant increase in prostate cancer events (odds ratio 0.88, 95% confidence interval 0.52 to 1.51), in clinically significant prostate cancer (odds ratio 1.13, 95% confidence interval 0.39 to 3.26), or in major cardiovascular events (odds ratio 0.83, 95% confidence interval 0.52 to 1.32) [s2]. Every one of those intervals crosses 1.0 — the point of no effect — so the honest reading is "no detectable increase," with confidence limits wide enough to rule out a large harm but not every small one [s2].

What this does not settle

The reassurance is bounded by who was studied. Trial participants were screened to exclude men already at high risk of prostate cancer, and the follow-up is measured in a few years, not decades [s1][s2]. None of this speaks to giving testosterone to a man with active, untreated prostate cancer, and the PSA rise means men on treatment need a monitoring plan so that a benign hormonal bump is not mistaken for disease — or, worse, a real cancer missed [s1]. The evidence supports the short-to-medium-term safety of treating genuine deficiency; it is not a licence to push testosterone at men who do not need it [s2].

That distinction runs through the wider debate over what testosterone therapy actually delivers and the marketing of testosterone "optimisation" to men whose levels are normal. It bears on how a low reading is worked up in the first place — see low testosterone and libido and male hypogonadism — and on why a raised PSA is now read through MRI before biopsy rather than the needle first. For men who do have prostate cancer, the trade-offs sit in the evidence on active surveillance.