AB19. Testosterone replacement therapy: how safe is it?

Testosterone has a ubiquitous role in the male body and the importance of a decline in testosterone levels has wide ranging impact on: regulation of gonadal function, prostate development and growth, libido, cerebral function, behavior, mood, muscle mass, liver function, lipid regulation, bone formation, atherogenesis, erythropoiesis, hair growth and immune function. What the minimum required level of serum testosterone for the optimal health of each of these areas, nor whether each organ system’s biological response to increasing or decreasing testosterone levels follows a ‘dose-response’, ‘threshold’ or other behaviour is unclear. Late-onset hypogonadism (also known as age-associated testosterone deficiency syndrome) is a syndrome associated with advancing age and characterized by a spectrum of symptoms and a biochemical deficiency in serum testosterone levels below the young healthy adult male reference range (280-300 ng/dL; 9.8-10.4 nmol/L- Note: this level may vary in different laboratories). The decrease in serum testosterone levels seems to be a gradual, age-related process resulting in an approximate 1-2% annual decline after age 30 years, with a steep decline in bioavailable and free testosterone levels. The findings Baltimore Longitudinal Study of Aging demonstrated that 30% of men in their eighth have total testosterone values in the hypogonadal range (that is between 200 and 400 ng/dL), and 50% have low free testosterone values (5-9 ng/dL; 0.17-0.31 nmol/L). An estimated 500,000 new cases of late-onset hypogonadism occur annually in the USA, with similar levels reported worldwide. Testosterone deficiency has marked physiological and clinical effects on men in middle age and beyond. With subnormal testosterone levels, the potential positive benefits of TRT on factors such as muscle mass, libido or erectile function are likely a dose-response phenomena, and should be considered differently than the threshold impact on the prostate. The controversies surrounding testosterone replacement therapy (TRT) have been addressed in the past few years. Although the androgenic effects of TRT on normal and malignant prostate cells have been studied for over 70 years, little clinical prospective research exists on the physiological responses of prostate tissues to a wide range of serum testosterone levels. The early, well-designed in vivo studies formed the basis of the concept that testosterone has a threshold or saturation level in all types of androgen-dependent prostate cells. That is, the stimulatory effects of androgens on the prostate reach a point within physiological serum levels above which they no longer have any proliferative effect and serum levels of testosterone and dihydrotestosterone can decrease substantially in both the eugonadal and hypogonadal states without affecting the amount of androgen within the nucleus of the cell. At a certain threshold level (possibly ‘castrate’ level), the intranuclear level of androgen will begin to decrease and the appropriate physiological changes will be triggered. Questions remain as to whether results from experimental studies in the rat can be extrapolated to the situation in humans. Is the human prostate subject to the same homeostatic constraints as has been so well defined in animal experiments, and if so, what is the threshold or saturation level for maximal intracellular androgens and physiological responses in man? The sensitivity of an individual to varying levels of testosterone is also influenced by his genetic makeup, particularly polymorphisms in the androgen receptor, and other upstream signaling and downstream metabolic events, including diabetes mellitus and obesity. Despite decades of research, no compelling evidence exists that increasing testosterone beyond this threshold level has a causative role in prostate cancer, or indeed changes the biology of the disease. Notwithstanding this, the reluctance to utilize testosterone replacement has been incorporated into urological dogma and is largely responsible for the US FDA’s continuing caution about the relationship between therapy and initiation or progression of prostate cancer. Recent international concern has arisen on the potential negative impact of TRT on the cardiovascular system, specifically MI and CVA. This is based on the results of a clinical trial and two observational studies. The first study to identify an association was the Testosterone in Older Men (TOM) trial designed to evaluate the effect of a T gel on muscle strength and functionality in an elderly population. The study did show an increase in upper and lower limb strength but was terminated prematurely due to an increased cardiovascular event rate in men receiving T. But because of the low event rate, the fact that many men reached supraphysiologic levels of serum T and the fact that the study was not designed with any specific cardiovascular endpoints in mind, it is difficult to conclude from this study that T therapy places men at increased risk of CVS disease. The next publication by Vigen et al. was a retrospective, non-randomized, observational analysis of 8,709 men, 61 to 64 years, who had undergone coronary angiography in the VA system with a low serum total testosterone (<300 ng/dL), looking at the CV events after having filled a prescription for TRT, comparing to untreated patients. The actual CVS event rate was twice as high for the untreated men (21.2% vs. 10.1%) but after complex statistical modeling the number of events in the treated men tripled. This study has been widely discredited because of serious flaws. For example, many of the treated men did not achieve normal levels, most men did not fill their scripts for the full 4-year duration of the study, and almost 3,000 men who received TRT prior to angiography were excluded so all men began in the ‘no TRT’ group. Most importantly, the authors inexplicably excluded 1,132 men who suffered stroke or heart attack prior to receiving a testosterone prescription. These men all had events during the study period and all should have been included in the no-testosterone group. This would have increased the rate of events in the no-testosterone group by 71%, likely reversing the results. It is impossible to conclude from this study that testosterone prescriptions increase rates of cardiovascular events. In a third population-based retrospective, non-randomized, observational study (insurance claims) of a cohort of 55,593 men, Finkle and colleagues evaluated the rate of non fatal MI in the three months after either having filled a first prescription for TRT or a PDE5i and compared this rate to the rate of non fatal MI in the preceding year. The authors concluded that the risk of MI is increased in older men and in younger men with pre-existing known heart disease who received testosterone prescriptions. Unfortunately, this study also has flaws in that it is impossible from the design to distinguish whether any observed difference was due to the underlying condition (T deficiency) or to its treatment (T prescription). Also the shorter the exposure time for a drug, the less likely it is responsible for an observed difference and though the authors had long term data (12 months) they did not report, raising a concern that the observed difference no longer persisted over time. Contrary to these three studies, new retrospective studies reveal no CVS dangers of TRT, and in fact suggest a possible protective mechanism. One trial suggests a 7-fold decrease in risk of MI. These studies would support multiple previous publications that suggest that men with normal T levels actually have longer life expectancies than hypogonadal men and that both low and high T levels have negative physiologic impact on male health. In summary, “expert views” from all walks of life (endocrinologists, epidemiologists, gerontologists, public health experts and urologists; lay press and regulatory agencies; pharmaceutical and film industry) is resulting in a cacophony of opinions creating much confusion and detriment to patients and physicians alike. A properly funded and implemented longitudinal study, similar to the Women’s Health Initiative, is required before we can address the true prostate and cardiovascular safety of TRT in the hypogonadal man. Until then, the application of this therapy should be personalized to the needs of the individual.