Reproductive Endocrine Unit and National Center for Infertility Research Massachusetts General Hospital Boston, Massachusetts 02114
Address correspondence and requests for reprints to: Frances J. Hayes, MB, MRCPI, Reproductive Endocrine Unit and National Center for Infertility Research, Massachusetts General Hospital, 55 Fruit Street, Boston, Massachusetts 02114. E-mail: Hayes.Frances{at}MGH.Harvard.edu
![]() |
Introduction |
---|
![]() ![]() ![]() |
---|
The recent resurgence of interest in a therapy that first entered the
clinical arena more than 60 yr ago reflects two major factors, the
development of more patient-friendly testosterone formulations and the
expansion of the clinical indications for androgen replacement. Shortly
after its isolation and synthesis in 1935, it became apparent that
testosterone had negligible oral bioavailability and a very short
duration of action due to extensive hepatic metabolism. Attention was,
therefore, focused on developing alternative modes of testosterone
delivery. Initial progress in this area was slow due to the reluctance
of the pharmaceutical industry to invest heavily in the limited market
of male hypogonadism. However, the expansion of the indications for
testosterone therapy to include wasting disorders such as human
immunodeficiency virus (HIV) infection and, more controversially, the
aging male has now opened up a whole new potential market for androgen
therapy. Currently, there are 34 million Americans over the age of
65 yr, and it is estimated that this number will double in the next 30
yr (1). Given these statistics, it is clear that if testosterone
therapy can be shown to either slow the progression or reverse some of
the features of aging, the market potential is huge.
Testosterone therapy has evolved considerably since the days of
the 19th century French physiologist Charles Brown-Séquard, who
extolled the virtues of a guinea pig testicular extract in restoring
waning potency and virility. However, it has taken a long time to
develop an optimal testosterone formulation combining favorable
pharmacokinetic and safety profiles with ease of administration. To
date, all available androgen formulations have had inherent
limitations. Oral 17-alkylated androgens such as
methyltestosterone are rarely used due to their lack of
potency and potential hepatoxicity. Although not available in the
United States, testosterone undecanoate has been used in Australia,
Canada, and Europe since the 1970s. The undecanoate preparation avoids
first pass metabolism due to its preferential absorption through the
lymphatic system, but its clinical use is limited by its short
half-life, necessitating multiple daily doses. Traditionally,
long-acting esters such as testosterone enanthate or cypionate have
dominated the androgen market and have been used successfully to
alleviate the symptoms of male hypogonadism for more than 50 yr.
Testosterone esters are administered intramuscularly every 2 to 3 weeks
and have the advantage of being the cheapest form of androgen
replacement, provided that the patient or a family member can
administer the injection. In addition to the requirement for an
intramuscular injection, however, these agents have the additional
disadvantage of an unfavorable pharmacokinetic profile, characterized
by large fluctuations in testosterone concentrations accompanied in
some men by undesired changes in mood, libido, and energy levels. In
Australia, a popular method of testosterone administration involves sc
insertion of pellets, which can produce normal levels of testosterone
for up to 6 months, although spontaneous extrusion of the pellets may
occasionally occur.
The development of transdermal testosterone preparations in the late
1980s was an important advance in androgen replacement therapy. The
first patch to come on the market required application to shaved
scrotal skin to permit adequate absorption. In hypogonadal men, the
scrotal transdermal system (Testoderm; Alza Corp., Mountain View, CA) produces testosterone levels that are
in the mid-normal range 4 to 8 h after application (2). However,
many men find this site of application unacceptable. In addition, in
men with hypogonadism of prepubertal onset, the surface area of the
scrotum may not be large enough to accommodate the patch. In recent
years, availability of nongenital patches has led to a significant
decline in the use of scrotal preparations. Permeation-enhanced
transdermal delivery systems with a reservoir containing testosterone
in an alcohol base (Androderm, Watson Laboratory, Corona,
CA) have been shown to maintain serum testosterone levels more
consistently within the normal range than an equivalent intramuscular
dose (3). The main drawback of this patch is skin irritation, which is
encountered in up to one third of patients. Although application of
triamcinolone cream to the skin before the patch can alleviate
erythema and pruritus, adverse skin reactions necessitate
discontinuation of therapy in 10% of subjects. A second transdermal
preparation (Testoderm TTS; Alza Corp.),
which has a larger surface area and does not contain a reservoir, is
associated with less skin irritation. However, this patch does not
always adhere well to the skin, particularly in men who engage in
strenuous exercise. Testosterone patches have the disadvantage of
permitting relatively limited flexibility in dosing and are 10 times
more expensive than testosterone esters.
The latest development in androgen replacement is an open testosterone delivery system using a 1% hydroalcoholic gel (AndroGel; Unimed Pharmaceuticals, Inc., Buffalo Grove, IL). Pharmacokinetic studies of this gel indicate that application to hypogonadal men increases testosterone levels to the normal range within 30 min with steady-state levels achieved by 24 h (4). In the August 2000 issue of JCEM (85:28392853), the same investigators report that 6 months treatment with this testosterone gel formulation improves sexual function, increases lean body mass and muscle strength, and decreases fat mass in a large cohort of hypogonadal men (5). Whereas similar beneficial effects were observed with the testosterone patch, the gel was better tolerated and associated with less skin irritation and a lower discontinuation rate. The main adverse effect observed was a significant, dose-related increase in hematocrit levels, which exceeded the normal range in 18% of subjects who received the higher testosterone gel dose. Prostate-specific antigen (PSA) levels also increased in a dose-dependent fashion but remained within the normal range. No significant changes were observed in the lipid profile following testosterone therapy.
It is important to put the benefits of this new mode of testosterone delivery into perspective. The testosterone gel is clearly an important advance in androgen replacement in that it is more user-friendly than existing formulations and, therefore, likely to improve patient compliance. However, the lay press is already attributing far-reaching clinical implications to this formulation in the absence of strong scientific data. The concept that a relative hypogonadism (i.e. testosterone levels in the low-normal range) may account for many of the symptoms of aging has recently gained popularity. The normal aging process is accompanied by physiological changes in target organs that are sites of androgen action. Given the decline in androgen levels with aging that has been documented in many epidemiological studies, it is tempting to attribute some of the features of aging to androgen deficiency (6). However, symptoms of aging, such as decrease in muscle strength, energy levels, libido, and potency, are nonspecific and are more likely to be multifactorial in origin as opposed to reliable indices of androgen status. In addition, there is considerable interindividual variation in the decline in testosterone with aging, and in many instances mean levels remain well within the normal adult male range.
If an absolute or relative androgen deficiency does, indeed, contribute to age-related physiologic changes, one could argue that a trial of testosterone therapy should be considered for all older men with testosterone levels in the low-normal range. There are three main problems with this approach. First, assessment of androgen status by a single testosterone level can be misleading. Second, the potential benefit of testosterone in the aging male is a question that is more appropriately addressed by large, prospective, randomized trials than by anecdotal experience. Third, testosterone is a Schedule III controlled substance with the potential to cause significant adverse effects if prescribed for inappropriate indications and without proper medical supervision.
In the absence of good biologic markers of androgen action, androgen status has traditionally been assessed by measurement of plasma testosterone levels. Whereas hypogonadism is easy to diagnose when testosterone levels are markedly diminished (i.e. <100 ng/dL), it is more difficult to interpret the biological significance of testosterone levels just below or in the low-normal range. When assigning a diagnosis of androgen deficiency based on a single testosterone measurement, it is important to be cognizant of the fact that as many as 15% of healthy young men with normal sexual function have a testosterone level below the normal range at some point during a 24-h period of blood sampling (7). In younger men in whom testosterone levels may fall by as much as 30% in the course of a day, the time at which the blood sample is drawn is critical for its interpretation. In many instances, samples that fall below the normal range when measured in the late afternoon will be normal when repeated in a morning sample. Obtaining a morning testosterone level may be less critical in older men given that the diurnal rhythm of testosterone secretion is significantly attenuated with aging (8). Given this normal pulsatile and diurnal rhythm of testosterone secretion, a more careful characterization of the hypothalamic-pituitary-testicular axis may be necessary, in some instances, to prevent misdiagnosis of hypogonadism and the institution of inappropriate therapy.
Although measurement of total testosterone by RIA has traditionally
been considered the most reliable method of diagnosing androgen
deficiency, there are situations where it may be misleading. More than
98% of testosterone is protein bound: 40% to sex hormone-binding
globulin (SHBG) and 60% to albumin. Aging is associated with an
increase in SHBG levels and consequently a slower decline in total
testosterone (
0.4% per yr) than in the free fraction (
1.2% per
yr) (9). For this reason, some investigators have advocated use of free
testosterone to diagnose hypogonadism on the basis that it is the most
sensitive marker of androgen status. However, accurate measurement of
free testosterone involves a dialysis procedure, a requirement that
makes the assay both time-consuming and inconvenient. The alternative,
which is to measure free testosterone by one of the widely used
commercial kits, is notoriously unreliable. In some studies, use of
free testosterone kits has been shown to underestimate plasma free
testosterone levels by as much as 100% (10). Therefore, reliance on
these kits to diagnose male hypogonadism has the potential to markedly
overestimate the true prevalence of androgen deficiency. A third
measure of androgen status is provided by bioavailable or
non-SHBG-bound testosterone. This assay, which involves ammonium
sulfate precipitation, does not suffer from the methodological
limitations of the free testosterone kits and may be a better marker of
androgen status than total testosterone in states of altered SHBG
secretion, such as obesity. Whereas androgen status is typically
defined solely on the basis of plasma testosterone levels, the androgen
receptor (AR) is clearly another important determinant of androgen
action. Indeed, whereas the majority of studies demonstrate no
correlation between endogenous testosterone levels and the subsequent
development of prostate cancer, an inverse correlation has been
established between the length of the AR polyglutamine tract and both
an increased risk and earlier onset of prostate cancer (11). Therefore,
polymorphisms in the AR may be as important as absolute testosterone
levels in mediating androgen effects.
When discussing the physiologic role of testosterone, it is important to make a distinction between overt hypogonadism and the more subtle decline in testosterone levels seen with aging. It is clear that overt hypogonadism has a significant negative impact on a variety of parameters, including body composition, bone mineral density (BMD), and sexual function, and that these effects can be reversed with androgen replacement. However, it is not clear what impact the age-related decline in testosterone has on these same parameters. In the last decade, a number of studies have tried to address this question by examining the impact of testosterone supplementation in the aging male. To date, these studies have yielded inconsistent results. The discrepancy in results may reflect the relatively small number of patients studied, variable selection criteria and definitions of "hypogonadism," differences in dose and duration of androgen therapy, and the different methodologies used to assess study end points.
Given the prevalence of osteoporosis in elderly men and its associated
morbidity and mortality, determining whether testosterone therapy may
have a beneficial impact on bone in men with testosterone levels in the
low normal range is an important public health issue. Based on current
data, the proportion of the variance in BMD in healthy elderly men that
is due to testosterone is controversial (12). Some, but not all,
studies have demonstrated a positive correlation between BMD and either
total or free testosterone. Overall however, hormone levels were
estimated to account for only 5% of the age-and weight-adjusted
variance in BMD in a large cohort of healthy, elderly men (13). The
largest and longest study of testosterone administration to a cohort of
healthy, elderly men with a baseline testosterone level of less than
475 ng/dL failed to show any significant improvement in BMD after 36
months of therapy (14). However, subset analysis of this study
demonstrated that subjects with baseline testosterone levels less than
200 ng/dL had a significant improvement in BMD in the lumbar spine,
whereas the latter did not change in those with a baseline testosterone
level more than 400 ng/dL. Regardless of the baseline testosterone
level, however, no change was observed in BMD at the hip. The absence
of a beneficial effect of testosterone on BMD at the hip is
particularly important given that most of the morbidity and mortality
of osteoporosis relates to hip fracture. Therefore, a convincing role
for testosterone has yet to be definitively established in reversing
the decline in BMD with aging. Some argue that the testosterone levels
achieved in this study were not high enough to see an effect, although
mean levels almost doubled. An alternative possibility is that it is
estradiol rather than testosterone that is the major steroid regulator
of bone in the male. This hypothesis is supported by the demonstration
of osteoporosis in men with congenital estrogen deficiency despite high
levels of testosterone. In addition, some epidemiological studies
demonstrate a better correlation between BMD and estradiol than
testosterone (15). Estradiol does not demonstrate the same age-related
decline as testosterone (9). The failure of estradiol produced by
aromatization of exogenously administered testosterone to have a
significant beneficial effect on BMD in men with normal baseline
hormone levels suggests that there may be a threshold for estrogens
beneficial effect on bone.
In clinical practice, most men are referred for testosterone replacement therapy because of declining sexual function. Contrary to what is being suggested by the lay press, current data do not support a beneficial role of testosterone therapy for middle-aged men with sexual dysfunction and testosterone levels in the low-normal range. The study by Wang et al. (5) as well as previous studies using other modes of testosterone delivery demonstrate that libido and potency are diminished in hypogonadal men and are restored with testosterone therapy (6). However, the majority of studies have demonstrated no correlation between libido or potency and testosterone levels within the normal range. Similarly, increasing testosterone levels within the normal range rarely has been reported to have a beneficial impact on sexual function. In the study by Wang et al. (5), the improvement in sexual function was independent of the dose or mode of testosterone delivery despite the fact that serum testosterone levels were almost 2-fold higher in the group that received the high-dose testosterone gel compared with those who received the patch. These data, therefore, support the concept of a threshold level of testosterone in the low normal range, below which libido and sexual function are impaired and above which there is no further enhancement of response.
Several studies have demonstrated that testosterone administration has a beneficial impact on body composition. This anabolic effect of testosterone is seen in both hypogonadal men treated with physiologic doses of testosterone as well as eugonadal men receiving pharmacologic levels of testosterone (16). Studies of androgen administration to older men have demonstrated a consistent increase in lean body mass, but a variable improvement in muscle strength (17). None of these studies have attempted to correlate the changes in muscle strength to functional performance. However, to date, all of the studies of androgen replacement have been done exclusively in healthy older men. Therefore, it remains to be seen what impact testosterone administration will have on frail, elderly men, who, in general, tend to have testosterone levels 1015% lower than healthy, age-matched controls (9). This may well be the population that will derive most benefit from testosterone replacement. It will be particularly interesting to see if changes in muscle strength in this group can translate into an improvement in functional activity and potentially a reduction in falls and fractures.
Testosterone therapy has the potential to cause a number of adverse events, which is why it is classified as a Schedule III controlled substance. Almost all studies of androgen therapy have demonstrated a significant increase in hematocrit levels due to stimulation of erythropoiesis, which has the potential to cause symptoms related to hyperviscosity. Administration of testosterone to hypogonadal men also causes a modest increase in prostate volume comparable with that seen in age-matched eugonadal men. To date, testosterone administration has not been reported to produce clinical symptoms of benign prostatic hypertrophy, although it is important to bear in mind that men with any suggestion of prostatic outflow obstruction have been excluded from these studies. Despite an increase in PSA levels within the normal range, there are no data to suggest a link between testosterone administration and prostate cancer, although longer follow-up is needed. The impact of androgens on lipids seems to vary depending on the dose, route of administration, and whether aromatizable or nonaromatizable androgens are used. To date, most studies of testosterone administration to healthy, older men have demonstrated a decrease in total cholesterol, with either no change or a slight decrease in high-density lipoprotein cholesterol. Abnormalities in liver function tests are generally limited to the orally administered androgens. Therefore, all patients on androgen replacement require monitoring with hematocrit levels, liver function tests, a lipid profile and PSA, and should be under the supervision of a physician with expertise in this area.
In summary, therefore, androgen replacement therapy in men is a new and exciting field of study. As "designer androgens" become available it should be possible to develop formulations with a desired profile of activity such that they are stimulatory to the desired target organ (e.g. bone) but have a neutral effect on erythropoiesis or the prostate. Available data are still insufficient to permit any major conclusions about the role of androgen replacement in the treatment of age-related physiological changes. In designing intervention studies on androgen replacement in men, it is imperative to learn from the studies of the impact of HRT in women. Whereas several observational studies had suggested that HRT was associated with a significant reduction in cardiovascular mortality, the one prospective, randomized, placebo-controlled trial of estrogen replacement demonstrated no cardiovascular benefit in terms of secondary prevention (18). Therefore, a large, randomized, placebo-controlled, multicenter study would be the optimal way to address the potential benefit of androgen replacement in the male.
Received July 17, 2000.
Accepted July 17, 2000.
![]() |
References |
---|
![]() ![]() ![]() |
---|