Taureau
Administrator
Abraham Morgentaler
Harvard Medical School, Beth Israel Deaconess Medical Center, Boston, Massachusetts
Address all correspondence and requests for reprints to: Dr. Abraham Morgentaler, One Brookline Place, Brookline, Massachusetts 02445. E-mail: [email protected].
Given the amount and complexity of new information regarding androgen deficiency and its treatment, The Endocrine Society’s Clinical Practice Guidelines (1) ("the Guidelines") are a welcome addition to the literature, representing the consensus recommendations of some of the most thoughtful "greybeards" in the field. However, the absence of large-scale clinical testosterone trials leaves room for alternative approaches, and I am grateful for the opportunity to share my views as a urologist specializing in the treatment of male sexual dysfunction.
During my training, I had been taught that androgen deficiency was rare and testosterone treatment ineffective. I was therefore surprised, when I began my practice in 1988, to discover that low levels of total and free testosterone were frequently associated with erectile dysfunction (ED), and testosterone injections regularly improved erections. I was further gratified when patients thanked me for making them "feel normal again." When patients complained that their symptoms recurred 10–14 d after their last testosterone injection, coinciding with the anticipated testosterone nadir, I became convinced this was not a placebo effect and that men could determine when their testosterone levels were low.
Today, testosterone therapy is recognized as an effective treatment for diminished libido and ED in many men who have low testosterone levels. Other sexual symptoms may also improve, including difficulty achieving orgasm, reduced orgasmic intensity, reduced sexual sensation in the genitalia, and reduced ejaculate volume (2).
In the androgen-deficient man with ED, my bias is to treat first with testosterone rather than a phosphodiesterase type-5 inhibitor (PDE5i), due to the additional symptomatic benefits that might occur, and also because the good responders will be "cured" without the need to plan for sex, as is necessary with PDE5i treatment. If ED persists after a trial of testosterone, I then add a PDE5i. Testosterone treatment may also "rescue" erections among men who failed PDE5i therapy. In one report, one third of androgen-deficient men who failed sildenafil developed adequate erections with testosterone alone, and another third responded with the combination of testosterone and sildenafil (3).
Testosterone has direct effects on the penis as well as the brain. Testosterone receptors are present in the corpora cavernosa, and castration results in decreased production of cavernosal nitric oxide, the mediator of erection. In rabbits castration abolished nerve-stimulated erections and was associated with loss of cavernosal smooth muscle and, remarkably, the de novo appearance of adipocytes in the subtunical area critical for veno-occlusion (4).
Despite acknowledging that "the testosterone concentration below which testosterone administration improves outcomes is unknown and may vary among individuals and among target organs," the Guidelines still makes the firm recommendation that only men with "unequivocally low" testosterone, specifically less than 300 ng/dl, be considered androgen deficient. In the absence of clinical correlation, this and the various other testosterone thresholds proposed over the years must all be recognized as arbitrary.
From a clinical perspective, there is little justification for denying a trial of treatment to symptomatic men with testosterone greater than 300 ng/dl if they have evidence of low levels of bioavailable or free testosterone. Men with elevated SHBG may be truly androgen deficient yet their total T may appear "normal." Prospective clinical trials are silent on the efficacy of treatment in men with testosterone greater than 300 ng/dl, because these men are generally excluded. Yet in one retrospective, uncontrolled study of men with sexual dysfunction and low free testosterone, 73.6% of men with total testosterone greater than 300 ng/dl reported improved erections (5). Conversely, there is no guarantee that men with "unequivocally low" testosterone will respond to treatment, since less than half of those with testosterone less than 200 ng/dl reported improved erections (5). The diagnosis of androgen deficiency clearly resists easy categorization by total testosterone values.
My approach to diagnosis emphasizes clinical history, with confirmation provided by low levels of either total or free testosterone. I usually begin testosterone treatment with transdermal gel, making sure that treatment levels reach the mid-normal range or higher. Raising testosterone only into the low-normal range will be subtherapeutic for most men. If there is no symptomatic benefit by 3 months, I discontinue treatment. I therefore use testosterone treatment as a therapeutic trial rather than a commitment to lifelong therapy.
In my experience, the analog free testosterone assay corresponds well with clinical presentation and therapeutic response. Indeed, I rely more heavily on the results of this assay than total testosterone, which varies widely with SHBG concentrations. Criticism of the analog assay as inaccurate in the Guidelines stems from the work of Vermeulen et al. (6), who reported that a direct comparison of analog and equilibrium dialysis assays in patient samples yielded a substantially lower result for the analog assay. However, the correlation between the assays was excellent (r = 0.937). The discordance is solved by a simple downward adjustment in reference values for the analog assay, already incorporated by most manufacturers and laboratories. In my practice, values less than 1.5 ng/dl suggest the presence of androgen deficiency. A recent review (7) also dismisses the analog assay, providing three references for this assessment: the report by Vermeulen et al., a report in women, and a letter describing the prevalence of analog assays in the literature. Clearly, none of these addresses the clinical utility of the analog assay in men. Unlike equilibrium dialysis and various calculations for free testosterone, the analog assay is widely available and easy to interpret. I encourage clinicians to determine its value for themselves.
Several of the Guidelines’ recommendations fall into the realm of testosterone "lore" and should be discarded. One is the impractical requirement to obtain blood in the early morning. There is no evidence that morning levels more accurately diagnose androgen deficiency than afternoon levels, especially since diurnal variation is substantially blunted in older men (8). Another is the requirement to confirm low testosterone by repeat testing. Given the variation in testosterone levels and the arbitrary nature of diagnostic thresholds, what would be the justification in denying treatment to a symptomatic man whose initial testosterone level was 290 ng/dl and whose repeat level was 315 ng/dl?
Finally, age-adjusted testosterone reference values should be eliminated. Since testosterone values decline with age, it makes no sense to define "normal" by comparing individuals to populations of similarly aged men who have also experienced a decline in values. Imagine if we denied eyeglasses to all but those with visual acuity in the lowest 2.5% (2 SD from the mean) of their contemporaries!
Review of the available evidence suggests that the long-standing fear of stimulating prostate cancer with testosterone is without a scientific basis (9). One explanation for this lack of increased risk is that testosterone treatment fails to increase concentrations of testosterone or dihydrotestosterone within the prostate itself (10).
A new concern is the association of low levels of testosterone with prostate cancer. Approximately 15% of hypogonadal men with a prostate-specific antigen (PSA) 4.0 ng/ml or less have biopsy-detectable cancer, and the risk of cancer was increased for men with more severe reductions in testosterone (11). Low levels of testosterone have also been associated with high-grade prostate cancer, higher stage at presentation, and worse prognosis (9).
With regard to PSA monitoring, I fear the recommendations in the Guidelines are too lax. Urologists view yearly PSA increases of 0.7–0.9 ng/ml as suspicious and generally perform prostate biopsy for an increase of 1.0 ng/ml or greater (2).
The diagnosis of androgen deficiency requires only an ear attuned to the characteristic symptoms and a blood test providing evidence of reduced levels of total or free testosterone. Treatment provides an opportunity for gratifying results, for patients and clinicians alike.
Harvard Medical School, Beth Israel Deaconess Medical Center, Boston, Massachusetts
Address all correspondence and requests for reprints to: Dr. Abraham Morgentaler, One Brookline Place, Brookline, Massachusetts 02445. E-mail: [email protected].
Given the amount and complexity of new information regarding androgen deficiency and its treatment, The Endocrine Society’s Clinical Practice Guidelines (1) ("the Guidelines") are a welcome addition to the literature, representing the consensus recommendations of some of the most thoughtful "greybeards" in the field. However, the absence of large-scale clinical testosterone trials leaves room for alternative approaches, and I am grateful for the opportunity to share my views as a urologist specializing in the treatment of male sexual dysfunction.
During my training, I had been taught that androgen deficiency was rare and testosterone treatment ineffective. I was therefore surprised, when I began my practice in 1988, to discover that low levels of total and free testosterone were frequently associated with erectile dysfunction (ED), and testosterone injections regularly improved erections. I was further gratified when patients thanked me for making them "feel normal again." When patients complained that their symptoms recurred 10–14 d after their last testosterone injection, coinciding with the anticipated testosterone nadir, I became convinced this was not a placebo effect and that men could determine when their testosterone levels were low.
Today, testosterone therapy is recognized as an effective treatment for diminished libido and ED in many men who have low testosterone levels. Other sexual symptoms may also improve, including difficulty achieving orgasm, reduced orgasmic intensity, reduced sexual sensation in the genitalia, and reduced ejaculate volume (2).
In the androgen-deficient man with ED, my bias is to treat first with testosterone rather than a phosphodiesterase type-5 inhibitor (PDE5i), due to the additional symptomatic benefits that might occur, and also because the good responders will be "cured" without the need to plan for sex, as is necessary with PDE5i treatment. If ED persists after a trial of testosterone, I then add a PDE5i. Testosterone treatment may also "rescue" erections among men who failed PDE5i therapy. In one report, one third of androgen-deficient men who failed sildenafil developed adequate erections with testosterone alone, and another third responded with the combination of testosterone and sildenafil (3).
Testosterone has direct effects on the penis as well as the brain. Testosterone receptors are present in the corpora cavernosa, and castration results in decreased production of cavernosal nitric oxide, the mediator of erection. In rabbits castration abolished nerve-stimulated erections and was associated with loss of cavernosal smooth muscle and, remarkably, the de novo appearance of adipocytes in the subtunical area critical for veno-occlusion (4).
Despite acknowledging that "the testosterone concentration below which testosterone administration improves outcomes is unknown and may vary among individuals and among target organs," the Guidelines still makes the firm recommendation that only men with "unequivocally low" testosterone, specifically less than 300 ng/dl, be considered androgen deficient. In the absence of clinical correlation, this and the various other testosterone thresholds proposed over the years must all be recognized as arbitrary.
From a clinical perspective, there is little justification for denying a trial of treatment to symptomatic men with testosterone greater than 300 ng/dl if they have evidence of low levels of bioavailable or free testosterone. Men with elevated SHBG may be truly androgen deficient yet their total T may appear "normal." Prospective clinical trials are silent on the efficacy of treatment in men with testosterone greater than 300 ng/dl, because these men are generally excluded. Yet in one retrospective, uncontrolled study of men with sexual dysfunction and low free testosterone, 73.6% of men with total testosterone greater than 300 ng/dl reported improved erections (5). Conversely, there is no guarantee that men with "unequivocally low" testosterone will respond to treatment, since less than half of those with testosterone less than 200 ng/dl reported improved erections (5). The diagnosis of androgen deficiency clearly resists easy categorization by total testosterone values.
My approach to diagnosis emphasizes clinical history, with confirmation provided by low levels of either total or free testosterone. I usually begin testosterone treatment with transdermal gel, making sure that treatment levels reach the mid-normal range or higher. Raising testosterone only into the low-normal range will be subtherapeutic for most men. If there is no symptomatic benefit by 3 months, I discontinue treatment. I therefore use testosterone treatment as a therapeutic trial rather than a commitment to lifelong therapy.
In my experience, the analog free testosterone assay corresponds well with clinical presentation and therapeutic response. Indeed, I rely more heavily on the results of this assay than total testosterone, which varies widely with SHBG concentrations. Criticism of the analog assay as inaccurate in the Guidelines stems from the work of Vermeulen et al. (6), who reported that a direct comparison of analog and equilibrium dialysis assays in patient samples yielded a substantially lower result for the analog assay. However, the correlation between the assays was excellent (r = 0.937). The discordance is solved by a simple downward adjustment in reference values for the analog assay, already incorporated by most manufacturers and laboratories. In my practice, values less than 1.5 ng/dl suggest the presence of androgen deficiency. A recent review (7) also dismisses the analog assay, providing three references for this assessment: the report by Vermeulen et al., a report in women, and a letter describing the prevalence of analog assays in the literature. Clearly, none of these addresses the clinical utility of the analog assay in men. Unlike equilibrium dialysis and various calculations for free testosterone, the analog assay is widely available and easy to interpret. I encourage clinicians to determine its value for themselves.
Several of the Guidelines’ recommendations fall into the realm of testosterone "lore" and should be discarded. One is the impractical requirement to obtain blood in the early morning. There is no evidence that morning levels more accurately diagnose androgen deficiency than afternoon levels, especially since diurnal variation is substantially blunted in older men (8). Another is the requirement to confirm low testosterone by repeat testing. Given the variation in testosterone levels and the arbitrary nature of diagnostic thresholds, what would be the justification in denying treatment to a symptomatic man whose initial testosterone level was 290 ng/dl and whose repeat level was 315 ng/dl?
Finally, age-adjusted testosterone reference values should be eliminated. Since testosterone values decline with age, it makes no sense to define "normal" by comparing individuals to populations of similarly aged men who have also experienced a decline in values. Imagine if we denied eyeglasses to all but those with visual acuity in the lowest 2.5% (2 SD from the mean) of their contemporaries!
Review of the available evidence suggests that the long-standing fear of stimulating prostate cancer with testosterone is without a scientific basis (9). One explanation for this lack of increased risk is that testosterone treatment fails to increase concentrations of testosterone or dihydrotestosterone within the prostate itself (10).
A new concern is the association of low levels of testosterone with prostate cancer. Approximately 15% of hypogonadal men with a prostate-specific antigen (PSA) 4.0 ng/ml or less have biopsy-detectable cancer, and the risk of cancer was increased for men with more severe reductions in testosterone (11). Low levels of testosterone have also been associated with high-grade prostate cancer, higher stage at presentation, and worse prognosis (9).
With regard to PSA monitoring, I fear the recommendations in the Guidelines are too lax. Urologists view yearly PSA increases of 0.7–0.9 ng/ml as suspicious and generally perform prostate biopsy for an increase of 1.0 ng/ml or greater (2).
The diagnosis of androgen deficiency requires only an ear attuned to the characteristic symptoms and a blood test providing evidence of reduced levels of total or free testosterone. Treatment provides an opportunity for gratifying results, for patients and clinicians alike.
