The results of the present study are concerning. Previous studies that were more pharma-epidemiological received a lot of press about 2 years ago for concluding that people who received testosterone had a greater risk of developing some cardiac event. All those studies were vastly criticized, rightly so, since large databases suffer from incomplete information on patients. One doesn’t know the circumstances behind why a patient was or was not placed on testosterone. Were they too sick to be treated, or too healthy? Were they truly taking the hormone or did they discontinue? They were nonrandomized in these clinical databases and in fact, about a third of those patients didn’t even have testosterone levels drawn before they were treated. Those studies using large pharmacy-oriented databases were interesting, but did not produce firm conclusions. The controversy has raged on since other non-randomized studies showed metabolic benefits in treated men.
The difference here is that this is a randomized study, in which elderly men were treated with testosterone vs. placebo. Since it was a relatively short 12-month study, clinical outcomes did not change, rather they used a surrogate marker of CAD, ie, coronary artery noncalcified plaque volume on coronary CTA, as an indicator of coronary atherosclerosis. This measure has been associated with myocardial ischemia and subsequent CV adverse events. The investigators reported an increase in their primary outcome, as well as total plaque volume on coronary CTA in the testosterone-treated group.
There are several issues to consider before making any changes in clinical practice based on these findings. Firstly, can these findings be generalized to all men? The patients in this study, by design, were older (more than 65 years of age), many of whom were found to have significant heart disease already with elevated coronary artery calcium scores. This population varies from the group of men who are most likely to be treated in clinical practice, ie, the middle-aged, 45-to-65-year-olds. The results may thus not apply to the younger, larger populations who are taking the medication.
Secondly, how well do we trust the change in coronary CTA, as a surrogate marker of CAD progression? There was a statistical change over the period of observation, but it remains questionable whether we trust this information enough to make clinical decisions based on it. This is not a `real’ angiogram and these are people who didn’t have any kind of real events. How sure are we that if someone has a change in plaque, it’s going to be associated with a clinical event, and in what amount of time? It may suggest that a patient may have a clinical cardiac event 10 years from now, but patients could stay relatively healthy with a high calcium score for a long time.
Thirdly, what is the total risk-benefit ratio? The JAMA journals also published two other investigations from this study, although the data may not have been for the exact same patients. The authors found that testosterone treatment was associated with a statistically significant increase in bone density as measured by DEXA scan (Snyder PJ, et al. JAMA Intern Med. 2017;doi:10.1001/jamainternmed.2016.9539), and an improvement in anemia (Roy CN, et al. JAMA Intern Med. 2017;doi:10.1001/jamainternmed.2016.9540). Bone density is also an intermediate marker, in this case a surrogate for fracture risk. The increased bone density thus is an important finding for men who are at increased risk of osteoporosis and fractures. Similarly, improvements in hematocrit could also be beneficial. Previously in this same study, testosterone treatment was linked to slight improvements in sexual function (Snyder PJ, et al. N Engl J Med. 2016;doi:10.1056/NEJMoa1506119). Thus, clinicians need to balance the risks and benefits.
There are clear indications for testosterone treatment, eg, when there is no question that the patient must be treated because of extremely low levels. Examples include pituitary disease, or congenital or acquired testicular damage. In those situations, we commonly understand the mechanism to explain the hypogonadism, and the men are symptomatic. Treatment in these situations would be recommended, even if the man has known heart disease, assuming there is careful monitoring afterwards.
A study to evaluate the effect of testosterone on clinical CV outcomes is being planned, but it will take a large sample size and many years of observation. Unfortunately, we will need to wait a long time for this. Until then, treating men with testosterone should, like all medications, come only after careful thought and discussions with the patient. Clinicians should offer testosterone therapy to men with low serum testosterone levels, who are symptomatic and have an underlying mechanism to explain their hypogonadism. Physicians should always emphasize lifestyle intervention and direct treatment for CV risk factors. Also, depending on the circumstances, testosterone therapy need not be continued forever if there is no symptomatic improvement or physiological benefit.
Adrian Dobs, MD, MHS
Professor of Medicine and Oncology
Johns Hopkins Medicine
Disclosures: Dobs reports no relevant financial disclosures.