Testosterone and erectile physiology

Abstract

The role of testosterone deficiency in sexual dysfunction is an important aspect of aging, because it affects such a large proportion of men over 50 years old. A number of age-related factors can cause sexual dysfunction (in particular erectile dysfunction) and testosterone deficiency, such as chronic illness and multiple medications, and the causative link between hypogonadism and erectile dysfunction is still debated. However, studies in castrated animals have proven that addition of testosterone, and its conversion to dihydrotestosterone, can restore erectile function. It appears that testosterone achieves this by peripheral mechanisms (endothelial dependent and independent) and central mechanisms. Testosterone replacement therapy is therefore effective for erectile dysfunction in men with hypogonadism, with success rates of 35–40%. Testosterone supplementation is also important in men who fail on phosphodiesterase type-5 inhibitors, because a minimum plasma concentration of testosterone is required for the successful restoration of erectile function with these agents. Testosterone gels are now the preferred formulation for testosterone supplementation and they can be highly beneficial in a proportion of men with erectile dysfunction.

Keywords:

• Testosterone
• hypogonadism
• erectile dysfunction

Introduction

Testosterone is important for a number of body functions such as mood, cognition and vitality, and is also important for the integrity of the bones and muscle/fat composition. However, sexual dysfunction is recognized as one of the most important symptoms associated with reduced testosterone levels [1]. Sexual dysfunction has many facets including low libido, poor erectile quality, ejaculatory or orgasmic dysfunction, reduced spontaneous erections, and reduced sexual activity and frequency [2].

Because of this relationship between testosterone deficiency and sexual dysfunction, testosterone supplementation can be of benefit in older as well as younger hypogonadal men. Increasing the testosterone levels can restore libido and improve sexual function (such as increased number of spontaneous erections), as well as improving muscle mass, increasing lean body mass and decreasing body fat [2],[3].

In this article the relationship between testosterone levels and sexual dysfunction (and in particular erectile dysfunction) is discussed, and the possible benefits of testosterone replacement are summarized.

Prevalence of testosterone deficiency

Both sexual dysfunction and testosterone deficiency increase in prevalence with age in males. The exact nature of testosterone deficiency has therefore been studied extensively. Several cross-sectional [4-9] and longitudinal studies [10],[11] established the fact that testosterone levels decline with age in men. An estimated 8% of men aged 40–49 years have testosterone levels below 325 ng/dl and this figure increases to 12% of those aged 50–59 years, 19% of those aged 60–69 years, 28% of those aged 70–79 years and 49% of those aged 80 years or more. This decline in testosterone appears to be constant with age and there is no apparent period of accelerated decline. In fact, using data from the Massachusetts Male Aging Study, an annual decrease of 1.6% in total testosterone was estimated and this began from the third decade of life [12].

In addition to this age-related decline in total testosterone, there are even more dramatic decreases in unbound free testosterone [13],[14] and weakly bound bioavailable testosterone [15],[16]. Whereas a lower percentage of men in the 40–49 years age group have low free testosterone (defined as <153 nmol/l) than have low total testosterone, this dramatically increases to 9% in those aged 50–59 years, 34% of those aged 60–69 years, 68% of those aged 70–79 years and 91% of those aged 80 years or more [14]. Similarly, the proportion of men with low bioavailable testosterone (<70 ng/dl) increases from 7% in the 40–49 years age group to 30% in those aged 50–59 years, 44% of those aged 60–69 years and 70% of those aged 70 years or more [15]. This is presumably related to the increased levels of sex hormone-binding globulin (SHBG) in older men [10]. These results suggest that 25% of men with hypogonadism would be missed if total testosterone alone was measured, and measuring bioavailable testosterone may be more relevant [15].

Generally, the decline in testosterone is a result of secondary hypogonadism (but in some cases alterations in testicular function may also cause this decline). However, concomitant diseases contribute significantly to the incidence of testosterone deficiency. As will be discussed in more detail later in this paper, testosterone deficiency is frequently associated with erectile dysfunction – 18.7% of men with erectile dysfunction also had low testosterone (280 ng/dl) in one large-scale study [17]. Approximately 30% of men with HIV have hypogonadism [18], but the effects of this could be estimated incorrectly if only total testosterone is measured because SHBG is increased in HIV infection [19]. The proportion of men with low testosterone could be as high as 50% in men with AIDS. A similar or higher proportion of men with type 2 diabetes have low testosterone. In a recent study, 33% of patients with type 2 diabetes had low free testosterone (the value was 43.7% if total testosterone measurements were used) [20]. These low levels of testosterone in type 2 diabetes were irrespective of glycaemic control, duration of disease or the presence of complications. This link with low testosterone can be seen in many conditions and it is likely that men with any chronic disease are at a high risk of secondary hypogonadism due to the stress of the illness suppressing the hypothalamic-pituitary function.

Although many diseases are associated with low testosterone levels, one of the most comprehensive studies of the longitudinal changes in testosterone suggested that decreases in testosterone with age were independent of chronic illness [14]. Only cancer was shown to have a significant effect on testosterone levels, while other chronic disease such as type 2 diabetes and coronary heart disease did not [14].

Like testosterone deficiency, erectile dysfunction increases with age (from 5% in men aged 40 years to 15% in men aged 70 years) [21], and a similar range of diseases are associated with, or are risk factors for, sexual dysfunction. In one medical endocrine-based centre, a large proportion of men presenting with sexual dysfunction (mainly erectile dysfunction but many men had several manifestations of sexual dysfunction) had hypogonadism (36.3%) – although this highlights the link between testosterone levels and erectile function, the actual proportion of men with hypogonadism in this study reflected an endocrine referral bias [22]. Other frequently occurring comorbidities in this population included hypertension (35.8%) and diabetes (23.1%) [22].

Testosterone and erectile function

Despite an apparent association between testosterone deficiency and erectile dysfunction, we must appreciate that the two conditions are overlapping but independent disorders. One condition is not always the inevitable consequence of the other. Different studies put the proportion of men with erectile dysfunction who also have low testosterone at between 2.1% and 21%[16],[23-25]. Despite this, there is a definite link between testosterone levels and erectile dysfunction in a proportion of patients.

The basic model that has been used to demonstrate the effects of testosterone on erectile physiology was described by Traish et al. [26]. Erectile function was tested in castrated New Zealand white rabbits and non-castrated controls, by stimulating the cavernosal nerve in the presence or absence of testosterone and then measuring the intracavernosal pressure (ICP). In control animals, the pelvic stimulation causes increases in the ICP from around 20 mmHg in the flaccid state to approximately 90 mmHg (Figure 1). Castration however reduces this increase in ICP (to approximately 30 mmHg after pelvic nerve stimulation) while having no effect on systemic blood pressure. However, when these castrated animals were injected subcutaneously with testosterone (14 mg/day for 7 days) the ICP increase in response to pelvic stimulation was restored (Figure 1). The possibility that the effect of testosterone is due to its aromatization to estradiol was discounted by doing the same experiment with estradiol (14 mg/day for 7 days), which did not restore the ICP increase [26].

Figure 1. The effects of castration and testosterone replacement on nerve-stimulated erectile function in a rabbit mode. (a) Mean systemic arterial pressure in different treatment groups. (b) Intracavernosal pressure with or without nerve stimulation in different treatment groups. C, intact control animals; V, castrated animals receiving vehicle only; T, castrated animals receiving testosterone; and E, castrated animals receiving estradiol [26]. Copyright © 1999, The Endocrine Society. Reprinted with permission from Traish AM, Park K, Dhir V, Kim NN, Moreland RB, Goldstein I. Effects of castration and androgen replacement on erectile function in a rabbit model. Endocrinology 1999;140:1861–1868. All rights reserved.

Using the same rabbit model but both surgically and chemically castrating the animals by injecting the luteinizing hormone-receptor hormone (LH-RH) agonist leuprolide acetate showed that the phosphodiesterase type 5 (PDE-5) inhibitor vardenafil did not restore ICP in castrated or chemically castrated animals [27]. This suggests that in hypogonadism, PDE-5 inhibitors cannot help restore erectile function and that a certain level of testosterone must be present for these agents to have their effect – this has obvious clinical significance that will be discussed later.

In castrated rats, the relative importance of testosterone and dihydrotestosterone (DHT; produced by the 5-α reduction of testosterone) for erectile function has been tested. By implanting castrated rats with SILASTIC implants containing testosterone or DHT, with or without daily injections of the 5-α-reductase inhibitor finasteride (which prevents the conversion of testosterone to DHT), it was shown that erectile response was evident with testosterone without finasteride and with DHT with or without finasteride but not with testosterone plus finasteride [28]. DHT is therefore the active androgen in the restoration of erectile function in castrated animals. In this study, as well as using ICP measurements to assess erectile function, nitric oxide synthase (NOS) activity was measured in the penile cytosol. NOS levels correlated with the ICP measurements and this suggested that the effects of testosterone and DHT were at least partially mediated by NOS levels, and therefore endothelial activity, in the penis [28]. Other studies have subsequently confirmed these results and suggest that testosterone acts at the genomic level to regulate expression of the gene encoding the NOS enzyme [29].

However, testosterone appears to have other non-nitric-oxide-dependent activity. Castrated rats were implanted with a testosterone pellet with or without L-nitro-L-arginine methyl ester (L-NAME), an inhibitor of NOS, in their drinking water for 1 week before undergoing pelvic stimulation and having their ICP measured [30]. L-NAME completely reduced ICP to almost zero in castrated animals, but the addition of testosterone at least partially restored the ICP response to pelvic stimulation even in the presence of L-NAME. It was suggested that this non-nitric-oxide-dependent pathway may involve the activation of guanylate cyclase, but that other mechanisms may also be involved [30].

In summary, the effects of testosterone on erectile function depend initially upon the conversion of testosterone to DHT. The activity of DHT is then mediated by an endothelial-dependent mechanism. However, other non-endothelial-dependent mechanisms are also responsible for the restoration of erectile function by testosterone. Importantly for the clinic, there appears to be a minimum level of testosterone needed for PDE-5 inhibitors to restore erectile function. The use of testosterone replacement therapy will therefore have clinical benefit in some men with erectile dysfunction.

Clinical benefits of testosterone therapy

From the data on the prevalence of hypogonadism and erectile dysfunction, and the apparent links between the two conditions, it seems likely that testosterone replacement therapy would have a positive effect on erectile function. Indeed, in both primary and secondary hypogonadism accompanied by erectile dysfunction, testosterone replacement therapy produces improvements in erectile dysfunction in approximately 35–40%[21]. The lack of response in 60–65% of men stems from the fact that there are likely to be other underlying diseases or medications contributing to the erectile dysfunction.

An additional but important role for testosterone replacement therapy in erectile dysfunction is to restore the response to PDE-5 inhibitors. A recent study recruited 20 men with arteriogenic erectile dysfunction (evaluated by dynamic colour duplex ultrasound; D-CDU) who had normal sexual desire but low-normal testosterone (mean total testosterone 368 ng/dl and mean free testosterone 75 pg/ml) and who did not respond to the PDE-5 inhibitor sildenafil on six consecutive attempts [31]. These men were randomized to receive placebo or transdermal testosterone (5 mg/day) for 1 month. After taking sildenafil 100 mg, the arterial inflow to cavernous arteries (measured by D-CDU) was significantly higher in patients receiving testosterone than in those receiving placebo (p < 0.05) (Figure 2) [31]. This corresponded to a significant improvement in the erectile function domain score of the International Index for Erectile Function (IIEF) in the testosterone-treated patients but not in the placebo group (p < 0.05), which in turn was associated with changes in the Global Assessment Question (GAQ) (80% and 10%, respectively, p < 0.01) [31]. In this study, the effects of testosterone appear to be direct vascular effects on NO-mediated vasodilation and not an effect on libido.

Figure 2. Arterial inflow to cavernous arteries (measured by dynamic colour duplex ultrasound) after taking sildenafil 100 mg at: B baseline; S + T after 1 month of testosterone (5 mg/day by transdermal patch); or S + P after 1 month of placebo [31].

The increased response of men with hypogonadism to sildenafil when given testosterone was confirmed by measuring nocturnal penile tumescence (NPT) frequency [32]. As well as being non-responders to sildenafil, the men in this study did not respond to apomorphine 3 mg, which acts by stimulating the brain to send signals down the spinal cord to the penis. However, after being given transdermal testosterone (5 mg/day) for 6 months, patients were responsive to apomorphine (increased number of NPT and increased maximum rigidity) suggesting that testosterone acts both centrally and peripherally to improve erectile function [32].

In an attempt to determine the level of testosterone needed to support normal erectile function when PDE-5 inhibitors are administered, we analysed data from a clinical trial of sildenafil [33]. In 521 men, those who had levels of analogue free testosterone in the normal range (11–35 pg/ml) generally responded well to sildenafil treatment (measured by the mean intercourse success rate). Similarly, if testosterone levels were just below normal levels, most men responded to sildenafil (75–95% response), but when the testosterone level was below 8.1 pg/ml the response was lower (approximately 40–50%) and when it was below 7.4 pg/ml no response to sildenafil was observed (Table I) [33].

Table I.  Response to sildenafil treatment in men with low testosterone levels [33].

Free testosterone level	Mean intercourse success rate
Slightly low (≤10.6 pg/ml), n = 33	75–95%
Moderately low (≤8.1 pg/ml), n = 7	40–50%
Very low (≤7.4 pg/ml), n = 4	0%

Testosterone therapy therefore seems necessary to restore erectile function in men with very low testosterone levels, but testosterone alone may not always be sufficient to achieve this (i.e., both testosterone and PDE-5 inhibition are probably needed in these patients).

Testosterone therapies

A number of methods of delivering testosterone therapy have been employed over the past five decades and more. Long-acting testosterone esters, especially enanthate and cypionate, have been the most commonly utilized. Deep intramuscular injections are given every week or two. Unfortunately blood levels fluctuate with wide peaks and troughs, often with supraphysiological levels being achieved. The result is varying effectiveness and side effects, including polycythemia, fluid retention, mood shifts and perhaps aggravation of sleep apnea. A newer long acting injectable testosterone undecanoate (Nebido®) has been developed. This new testosterone formulation (1000 mg testosterone undecanoate in 4 ml castor oil) possesses good long-term kinetics, i.e. sustained close mimicking of eugonadal testosterone serum levels without supra- or sub-physiological serum concentrations. The generally accepted administration scheme for hypogonadal patients is one injection every 10 weeks to 14 weeks. It is hoped that this preparation will ameliorate the shortcoming of the older preparations; effect in clinical practice will take several years to evaluate.

An oral tablet, methyl testosterone, was popular several decades ago, but was soon abandoned in men when cases of liver abnormalities, including several cases of liver cancer, were reported. Subcutaneous testosterone pellets have been available for a number of decades, and have limited use in some countries and rare usage in others. Three to six pellets are implanted subcutaneously with a trochar every six months. Extrusion of pellets and infection have been noted as possible adverse effects.

Testosterone delivery by transdermal patch offered stable levels similar to the subcutaneous pellets without an invasive procedure. A limiting factor was the skin pruritus or rash that occurs in a third or more of men at the site of application with some of these patches. Some men also noted that the patch would fall off from perspiration or from bathing. These problems were obviated with the development of transdermal gels. Skin irritation is uncommon and the levels remain quite stable in the middle of the normal range.

Are all testosterone topical gels the same?

Testosterone topical gels are now the most frequently used formulation in North America and Europe but these testosterone gels are not identical. After single 50 mg doses of Testim 1% gel (Testim 50 mg gel) and Androgel® (also marketed as Testogel® outside of the USA), the C_max for testosterone DHT and free testosterone were 30%, 19% and 38% higher, respectively, with Testim 1% gel (Testim 50 mg gel) than with Androgel®[34]. Similarly, Testim 1% gel (Testim 50 mg gel) produced 30%, 11% and 47% greater increases in the AUC_0–24 of testosterone, DHT and free testosterone, respectively [34].

Summary

Testosterone is directly involved in erectile physiology. It stimulates central mechanisms including libido but also neurological signals that travel down the spinal cord to exert their effect on the penis. Testosterone also has effects on the peripheral mechanisms of erectile function, and these include endothelial production of nitric oxide and endothelial-independent mechanisms that are still to be elucidated. Aromatization to estradiol is not involved in erectile function but it seems likely that DHT is the active metabolite responsible for the effects on erectile function.

Clinically, the need for a certain level of testosterone for proper functioning of PDE-5 inhibitors is important. However, the exact plasma level threshold of testosterone that is required to influence erectile processes is not yet established (but appears to be lower than the lower threshold for normal testosterone levels). Furthermore, in patients with erectile dysfunction and low testosterone, approximately 35% will respond to testosterone replacement therapy alone. These data highlight a dual role for testosterone therapy in erectile dysfunction (restoring function in a proportion of men with low testosterone and restoring the efficacy of PDE-5 inhibitors), but this message is not always reaching primary care practice and needs to be emphasized so that more men can benefit from the simple and well-tolerated testosterone treatments available.