Eur Urol. Author manuscript; available in PMC 2008 October 7.
Eur Urol. 2007 July; 52(1): 54–70.
Abdulmaged M. Traish,ab* Irwin Goldstein,c and Noel N. Kimb
Androgens are essential for the development and growth of the penis, and they regulate erectile physiology by multiple mechanisms. Our goal is to provide a concise overview of the basic research and how this knowledge can be translated into a new clinical paradigm for patient management. In addition, this new paradigm may serve as a basis for stimulating constructive debate regarding the use of testosterone in men, and to promote new, innovative basic and clinical research to further understand the underlying mechanisms of androgen action in restoring erectile physiology.
A literature review was performed utilizing the US National Library of Medicine’s PubMed database.
On the basis of evidence derived from laboratory animal studies and clinical data, we postulate that androgen insufficiency disrupts cellular-signaling pathways and produces pathologic alterations in penile tissues, leading to erectile dysfunction. In this review, we discuss androgen-dependent cellular, molecular, and physiologic mechanisms modulating erectile function in the animal model, and the implication of this knowledge in testosterone use in the clinical setting to treat erectile dysfunction. The new clinical paradigm incorporates many of the consensed points of view discussed in traditional consensed algorithms exclusively designed for men with androgen insufficiency. There are, however, novel and innovative differences with this new clinical paradigm. This paradigm represents a fresh effort to provide mandatory and optional management strategies for men with both androgen insufficiency and erectile dysfunction.
The new clinical paradigm is evidence-based and represents one of the first attempts to address a logical management plan for men with concomitant hormonal and sexual health concerns.
The health of the penile vascular tissues and the perineal and ischiocavernosus muscles that support the proximal penis is essential for normal erectile function [1–4]. The role of androgens in regulating erectile physiology in humans is of considerable importance and merits continued investigation. The literature is replete with articles and anecdotes suggesting that androgens have little or a passive role in erectile function. In contrast, a significant and accumulating body of knowledge suggests that androgens play an important role in erectile physiology in humans. These inconsistencies may be due to the fact that much of the literature is based on clinical studies with varying methodologies and patient populations. In addition, genetic, health, and cultural factors are usually not considered. Nevertheless, animal studies have provided some basic foundation for our understanding of erectile physiology and the role androgens play in this process. In this review, we discuss knowledge gained from animal studies to provide a succinct analysis of the cellular, molecular, and physiologic mechanisms of androgens in erectile physiology, and how such knowledge may be translated into a new clinical paradigm for the management of patients with androgen deficiency and erectile dysfunction (ED). Our objective is to engage readers in a constructive and stimulating debate regarding the use of testosterone in men, and to promote new, innovative basic and clinical research to further understand the underlying cellular and molecular mechanisms of androgen action in restoring erectile physiology.
2. Modulation of erectile physiology by androgens: cellular, molecular, and physiologic mechanisms
2.1. Testosterone regulates nerve structure and function
The studies of Meusburger and Keast  and Keast et al  have provided elegant demonstrations on the potential role of androgens in maintaining the structure and function of many pelvic ganglion neurons. They suggest that testosterone is critical for the maturation and maintenance of terminal axon density and neuropeptide expression in the vas deferens. Giuliano et al  suggested that testosterone acting peripherally to the spinal cord enhances the erectile response of the cavernous nerve. Rogers et al  demonstrated that castration altered the dorsal nerve ultrastructure in the rat concomitant with loss of erectile function. The authors further showed that testosterone treatment of castrated animals restored the nerve fibers and myelin sheath structure, similar to that observed in the sham (control) group. Baba et al [9,10] reported that the integrity of NADPH diaphorase-stained nerve fibers in the rat corpus cavernosum and dorsal nerve is dependent on androgens. Recently, we examined the effects of castration on the structural integrity and function of the cavernosal nerve (Traish et al, unpublished observations). We noted that there were marked structural changes in the cavernosal nerve from castrated animals compared with control (sham-operated animals) or castrated animals treated with androgens (Fig. 1). These structural alterations may be responsible in part for the marked reduction in the intracavernosal pressure (attenuated blood flow) observed in the experimental animals . In addition, recent studies have demonstrated that penile erection in rats, elicited by stimulation of the medial preoptic area, is testosterone-dependent . Thus, testosterone may regulate central mechanisms of penile erection, as well as peripheral neural mechanisms. Clearly, more in-depth investigations are warranted to define the exact role of androgens on the penile nerve network and to determine how androgens modulate penile response to sexual stimulation.
2.2. Testosterone regulates nitric oxide synthase expression and activity
The nitric oxide synthase/cyclic guanosine monophosphate (NOS/cGMP) pathway has been deemed critical for erectile function . Nitric oxide (NO) mediates relaxation of the vascular smooth muscle of the resistance arteries of the corpus cavernosum and the trabeculae to facilitate penile erection. A preponderance of evidence supports a role for androgens in regulating the expression and activity of NOS isoforms in the corpus cavernosum in animal models [14–25]. In castrated animals, testosterone or 5α-dihydrotestosterone (DHT) administration restored the erectile response and NOS expression in the penis [9–11,16,18,19,21,23,24]. Interestingly, very few studies ventured beyond these initial observations demonstrating effects of testosterone on NOS expression. Further studies are needed to delineate the molecular basis of androgen receptor activation of the NOS genes and the battery of factors that modulate androgen receptor activity in penile tissue. While the focus on the NOS/cGMP pathway provided a stimulus for understanding the physiology of erection, many other pathways have received little attention. For instance, the role of prostanoids/eicosanoids and growth factors in regulating erectile physiology has yet to be fully investigated. Reassessment of the multiple pathways involved in this very critical physiologic function is warranted.
2.3. Testosterone regulates phosphodiesterase (type) 5
Phosphodiesterase (type) 5 (PDE5) hydrolyzes cGMP in vascular and trabecular smooth muscle into GMP. Activation of PDE5 terminates NO-induced, cGMP-mediated smooth muscle relaxation, resulting in restoration of basal smooth muscle contractility and penile flaccidity. In penile tissue, the balance between the intracellular levels of cGMP and GMP is primarily regulated by the activities of NOS and PDE5. Thus, it is likely that any disruption in the expression or activity of these enzymes will lead to pathophysiology. Castration has been shown to reduce the expression and activity of PDE5 in rabbits and rats [11,27,28], and androgen supplementation has been shown to upregulate the expression and activity of PDE5 [11,26–28]. Further, administration of PDE inhibitor alone to medically or surgically castrated animals has little effect on the intracavernosal pressure in response to pelvic nerve stimulation [27,29], suggesting that androgens are critical not only for regulating NOS activity, but also in modulating PDE5 activity. While these actions may be seen as a paradox, in that androgens are upregulating both signal initiators (NOS) and signal terminators (PDE5), we interpret this to be a homeostatic mechanism that maintains a relatively constant ratio of critical enzymes for this pathway (Fig. 2). We postulate that PDE5 expression may be controlled by NO. Upregulation of NOS by androgens may lead to increased NO synthesis, which then upregulates PDE5 expression and activity. Conversely, NOS downregulation by androgen deprivation results in downregulation of PDE5 expression and activity. Studies are underway to define this delicate and crucial mechanism in androgen action.
2.4. Testosterone regulates cellular growth and differentiation
Androgen deprivation by surgical or medical castration results in a significant reduction in trabecular smooth muscle content and marked increase in connective tissue deposition [26,29]. These structural alterations are also associated with loss of erectile function. Using transmission electron microscopy, the cavernosal smooth muscle in castrated animals appears disorganized with a large number of cytoplasmic vacuoles, whereas, in the intact animals, the smooth muscle cells exhibit normal morphology and are arranged in clusters [1,8]. Shen et al  have demonstrated that the structure of the tunica albuginea in rats is also influenced by androgens. Four weeks after castration, the tunica was thinner with fewer elastic fibers, and the collagen appeared more disorganized. Depletion of elastic fibers and replacement fibrosis was also noted in intact rats treated with finasteride, although the thickness of the tunica did not differ from intact controls. Taken together, these results suggest that androgens have a profound effect on cellular structure and organization of the corpus cavernosum, and that these alterations may contribute to the loss of erectile function. Such studies have not been performed in human penile tissue.
In addition to the alterations in smooth muscle and connective tissue, fat-containing cells have been observed in the subtunical region of penile tissue sections from orchiectomized animals . The alterations in cavernosal tissue composition and structure were accompanied by a reduced erectile response to pelvic nerve stimulation [11,31]. It is interesting to speculate that the presence of fat cells in the subtunical region of the corpus cavernosum may contribute to venous leakage in the orchiectomized or androgen-deficient animal. Abnormal deposition of fat-containing cells and reduced relaxation response to nitroprusside and acetylcholine has also been observed in penile corpus cavernosum of intact rabbits that were administered the endocrine disrupters bisphenol A and tetrachlorodibenzodioxin (TCDD) [32,33]. Specifically, bisphenol A has been shown to accelerate terminal differentiation of 3T3L1 fibroblasts into adipocytes through the PI3 kinse pathway . Interestingly, in developmental studies, Goyal et al [35–38] have shown that administration of estradiol valerate or the estrogen receptor agonist diethylstilbestrol into 2-day-old rats resulted in infertile mature animals (120 d) and accumulation of fat-containing cells in the penile corpus cavernosum. In contrast, animals treated with vehicle exhibited no fat-containing cells and remained fertile. The authors demonstrated that estrogen treatment was associated with low plasma testosterone levels, which may have contributed to alterations in penile anatomy and morphology, infertility, and ED. Since estrogens are known to act as antiandrogens in some tissues [39–41], these studies point to the potential importance of androgens in maintaining penile corpus cavernosum structure.
There is renewed interest in understanding the mechanisms by which androgens regulate growth and differentiation of vascular smooth muscle cells. Bhasin et al  and Singh et al [43,44] hypothesized that androgens promote the commitment of pluripotent stem cells into a muscle lineage and inhibit their differentiation into an adipocyte lineage. The total number of circulating vascular progenitor cells may also be dependent on testosterone levels . Regulation of progenitor cell differentiation is a complex process, dependent on numerous hormones, growth factors, and specific activation of a cascade of gene expression [42,46–51].Critical regulators of adipocyte differentiation include C/EBPα (CCAAT/enhancer binding protein), PPARγ2 (peroxisome proliferator-activated receptor), and LPL (lipoprotein lipase) [47,48,52–57]. Alternatively, transdifferentiation of smooth muscle cells into other phenotypes may occur [58–61]. Inhibition of 5α-reductase activity induces stromal remodeling and smooth muscle dedifferentiation in the prostate, suggesting that 5α-DHT deficiency promotes smooth muscle dedifferentiation . While these mechanisms of precursor cell differentiation or smooth muscle transdifferentiation have yet to be investigated in penile tissue, future studies using expression of biochemical markers as well as changes in ultrastructure are needed to test these hypotheses in the corpus cavernosum under varying states of androgen deprivation and supplementation. Thus, the penis is a unique model system that contains multiple tissue types with differing androgen responses. We postulate that, in the penile corpus cavernosum, androgens are critical for promoting and maintaining the myogenic lineage (Fig. 3).
2.5. Testosterone restores erectile function in diabetic animals
Zhang et al  have demonstrated that alloxan-induced diabetes in rabbit and streptozotocin-induced diabetes in rats resulted in reduced plasma testosterone and atrophy of androgen-dependent accessory glands. Supplementation with testosterone in diabetic rats increased the erectile response, and the expression of PDE5 and endothelial and neural NO synthases. In organ baths, relaxation to acetylcholine was enhanced in corpus cavernosum tissue strips from diabetic animals treated with testosterone. The authors concluded that normalizing plasma testosterone in diabetic animals restores NOS and PDE5, and reinstates sensitivity to relaxant stimuli and responsiveness to sildenafil in vivo.
2.6. Erectile function depends on a threshold dose of testosterone
Armagan et al  have demonstrated that erectile function in rats is maintained by a wide range of systemic testosterone levels, as low as 10–12% of normal physiologic plasma concentrations. However, below these concentrations, erectile function is significantly attenuated, and this attenuation is positively correlated with the plasma concentration of testosterone. Testosterone levels in the range of 10% of the normal physiologic plasma concentration may represent a threshold value, below which erectile function declines in a dose-dependent fashion. This concept of threshold value is supported by recent clinical studies . Interestingly, in rats, prostate tissue mass was positively correlated to plasma testosterone levels across the entire range of testosterone concentrations examined. In addition, the significance of the correlation between plasma testosterone and androgen-dependent tissue growth (weights) was variable, with the seminal vesicles exhibiting the most significant correlation. These data suggest that different androgen-dependent tissues have varying sensitivities to circulating testosterone levels that can be manifested through both trophic and functional responses.
3. A clinical paradigm for the combined management of androgen insufficiency and ED
Androgens have been used to treat sexual problems  as well as to augment vasodilation [66–69] in patients with angina and claudication for more than six decades. In light of the established historic link of androgens to both facilitating sexual function and vasodilatory function, it is not surprising that the contemporary management of aging men and their sexual health concerns involves frequent use of PDE5 inhibitors and off-label use of androgens [70–76]. These clinical uses are based in part on the recent explosion of basic science and clinical data concerning androgens and erectile physiology [1,3,77–83]. Such basic science and clinical trial data support an evidence-based diagnostic and treatment paradigm for men with both androgen insufficiency and ED. Prevalence rates of androgen insufficiency and ED in aging men have been reported from 1.7%  to 35% , which translates to millions of men being afflicted with both disorders. In the following section, we outline an integrated approach for the management of men with both androgen insufficiency and ED, including step care strategies that engage identification of the hormonal and sexual problems, patient and partner education, modification of reversible causes, hormonal and nonhormonal therapies, and other treatments (Fig. 4). The remaining sections of this review encompass information from guidelines provided by the European Association of Urology, International Society of Andrology (ISA), and International Society for the Study of the Aging Male (ISSAM) . In addition, we propose a new clinical paradigm for patient management based on knowledge gained from basic science research. The new clinical paradigm incorporates many of the consensed points of view discussed in traditional consensed algorithms exclusively designed for men with androgen insufficiency. There are, however, novel and innovative differences that represent a fresh effort to provide mandatory and optional management strategies for aging men with not just androgen insufficiency alone, but for men with both androgen insufficiency and ED.
3.1. Step care 1: Identification of androgen insufficiency and ED
Androgen insufficiency [82,83] is considered to exist as a syndrome in which there are (1) nonspecific signs and symptoms, such as low sexual interest, muscle weakness, feeling sad and melancholy, or having an inadequate erectile response to PDE5 inhibitors in a man with ED and (2) biochemical blood test values suspicious for low levels of physiologically relevant androgens. ED is the persistent or consistent inability to obtain and/or maintain a sufficient erection for satisfactory sexual activity . Presenting symptoms exist in several other syndromes and vary widely among individuals. Thus, a detailed medical workup is required .
3.1.1. Sexual, psychosocial, and medication history
The sexual symptoms of androgen insufficiency are varied and include decreased sexual interest; diminished erectile quality, particularly of nocturnal erections; muted, delayed or absent orgasms; decreased genital sensation; and reduced sexual pleasure [82,83,86,88–90]. In addition, sexual dysfunction may affect the patient’s self-esteem, coping ability, and occupational and social roles . Androgen insufficiency is associated with changes in mood, diminished well-being, blunted motivation, changes in spatial orientation, reduced intellectual ability, fatigue, depression, and anger/irritability [82,83,86,88–90].
Failure to respond to a maximum dose of oral PDE5 with maximum erection hardness may be the first sign of androgen insufficiency [70–76]. This perspective is based on the observation that androgens may directly control the expression and activity of NOS in human corpus cavernosum [91–94]. The clinical responsiveness of PDE5 inhibitors appears to be strongly linked to NOS activity in vascular tissues [70–76].
3.1.2. Screening questionnaires
Screening questionnaires may be used to aid in the clinical diagnosis of androgen insufficiency. Androgen Deficiency of the Aging Male (ADAM) is useful for identifying the presence or absence of androgen insufficiency symptoms [95,96], but has poor specificity in aging men. The Aging Male Scale (AMS) is a more extensive, validated instrument . The low testosterone screener of Smith and colleagues  is also useful to reliably detect androgen insufficiency. Similarly, ANDROTEST is a structured interview for the screening of androgen insufficiency in men with sexual dysfunction . However, it should be noted that validated questionnaires cannot replace a detailed history and physical examination [4,82,83]. Questionnaires are separate and unique aspects of the “identification” aspect of the new clinical paradigm. Some of the questionnaires are psychometrically validated and distinct, and used for outcomes assessment.
3.1.3. Physical examination
A focused physical examination including endocrinologic examination should be performed on every patient, especially if the response to a PDE5 inhibitor is not robust. Androgen insufficiency is associated with small, less firm testes; decreased beard and body hair growth; skin thinning; a decrease in lean body mass; an increase in body fat and decrease in muscle mass and strength; and the development of breast tissue [82,83]. Small, less firm testes are consistent with hypergonadotrophic hypogonadism (primary testicular failure). However, this characteristic may not be the case in hypogonadothrophic hypogonadism.
3.1.4. Mandatory laboratory testing
In this new clinical paradigm, laboratory tests are subdivided into mandatory and optional in men with both androgen insufficiency and ED. In this section, we describe the mandatory laboratory tests (Fig. 4).
The diagnosis of androgen insufficiency in men is based on a suggestive clinical picture and the biochemical demonstration of androgen deficiency. Abnormal total testosterone values [82,83] alone are not sufficient reason to institute therapy. In men with minimal symptoms and markedly reduced testosterone values (eg, <200 ng/dl), a discussion should ensue with the patient concerning risks and benefits of therapy. It should be noted that the various laboratory ranges currently considered normal for androgens in men are not always reliable [82,83] and are, at best, an approximation of androgen status. They do not take into account the localized, tissue-specific metabolism of androgens into bioactive metabolites (intracrine mechanisms) or differences in androgen sensitivity, in which the response of target organs to a given androgen concentration will vary in different individuals [82,83].
There is no universally accepted cut-off value of total testosterone that unambiguously defines the state of androgen insufficiency [64,86,100,101]. Since total testosterone values fall with age and change with circadian rhythm, the ideal time to clinically measure total testosterone is in the early morning. Because of loss of pulsatile hypothalamic functions with aging, starting as early as age 40 , blood tests can be measured at any time in aging men [82,83,102].
Total testosterone measurements can be misleading, because only unbound testosterone can act within cells to regulate gene expression. In normal males, 2% of testosterone is free (unbound), 30–60% is bound to sex hormone–binding globulin (SHBG) with high affinity, and the remainder is bound with much lower avidity to albumin and other proteins . SHBG has a higher affinity for testosterone than for estradiol, and changes in SHBG reduce or amplify the hormonal milieu. Thus SHBG, in part, regulates androgen function, and it is clinically relevant in each patient suspected of having androgen insufficiency to be aware of the SHBG value. High values of SHBG will lower the unbound physiologically available testosterone [82,83].
The health care provider should assess free testosterone in all patients. However, it should be stressed that different assay techniques can yield different measurements. Antibody-based, free testosterone assays using a testosterone analogue are considered unreliable. Equilibrium dialysis, the gold standard, is usually difficult and time-consuming, and therefore not widely used clinically . Bioavailable testosterone measures free and albumin-bound fractions of testosterone, and is reliable and accessible. Bioavailable testosterone values fall with increasing age, especially as total testosterone falls and SHBG values increase [82,83].
A contemporary management strategy for the health care provider is to determine total testosterone (ng/dl), SHBG (nmol/l), and albumin (g/dl) concentrations. These values can then be used to determine free testosterone with a calculator [82,83,86] that is available on the Web page of the ISSAM (www.issam.ch/freetesto.htm). The use of this calculator is free of charge and results in values that correlate well to free testosterone determined by equilibrium dialysis. In most cases of “healthy” men, the albumin value can be assumed to be 4.3 g/dl. However, when doing clinical research or in aging men with a chronic disorder, it is advisable to determine the individual’s actual albumin value. A calculated free testosterone value less than 5 ng/dl is considered abnormal. When the total testosterone value is borderline, calculated free testosterone values are useful to help confirm androgen insufficiency . In studies using this approach, 17.6% of men with ED had criteria for androgen insufficiency . Further, hypertension, aging, absence of nocturnal erections, and low erectile function scores were associated with low calculated free testosterone levels .
126.96.36.199. Prostate-specific antigen
Androgen administration is absolutely contraindicated in men with or suspected of having carcinoma of the prostate [82,83,86]. Determination of serum prostate-specific antigen ((PSA) [107,108] and digital rectal examination (DRE) are mandatory as baseline measurements of prostate health prior to therapy with androgens. Many health care providers now consider a PSA of 0–2.5 ng/ml as low and values greater than 2.6 to 10 ng/ml as elevated. Both PSA and DRE examinations should be repeated every 3–6 mo for the first 12 mo, and annually thereafter. Transrectal biopsy of the prostate is indicated if the DRE or the PSA are abnormal or if the PSA increases 0.75 ng/ml in one calendar year [107–110]. If the PSA increases during androgen therapy and the biopsy is negative for prostate cancer, androgen therapy can continue with repeated PSA testing and DRE every 3–6 mo. While there is no evidence that androgen therapy causes prostate cancer, it may accelerate an existing underlying prostate cancer [107–110].
3.1.5. Optional laboratory testing
Serum dihydrotestosterone (DHT) determination may be valuable, because for some androgen-dependent functions, testosterone is a prohormone peripherally converted to DHT via the enzyme 5-alpha reductase. Supraphysiologic levels of DHT may be observed following topical testosterone gel administration, which is associated with acne and scalp hair loss . The presumed mechanism is related to the presence of high concentrations of 5-alpha reductase enzyme in skin and the much greater skin surface area of testosterone application using the gels compared with the patch. Successful management of side effects may be achieved with low doses of 5-alpha reductase enzyme inhibitors.
Subphysiologic levels of DHT may occur with the medical treatment for lower urinary tract symptoms (LUTS) that engages clinical use of 5-alpha reductase inhibitors and lowers the circulating level of DHT by as much as 80% . The 5-alpha reductase inhibitors finasteride and dutasteride are reportedly associated with a greater risk of ED, ejaculatory dysfunction, and decreased libido compared with placebo [113,114]. In animal studies, finasteride treatment resulted in significantly lower DHT levels, and multiple ultrastructural alterations of the tunica albuginea and penile erectile tissues . DHT has also been shown to be an independent hormonal predictor of increased frequency of orgasms in men .
Hyperprolactinemia is an uncommon cause of androgen insufficiency and ED [82,83]. However, if a patient presents with signs and symptoms of diminished sexual interest and gynecomastia, and has biochemical evidence of androgen insufficiency, determination of serum prolactin is recommended . Direct role of prolactin in male libido has been proposed . Although rare, elevated serum prolactin levels are associated with potentially high morbidity disease and pituitary tumors.
For men undergoing exogenous testosterone therapy, serum estradiol determination may be of value. Estradiol is synthesized in men in peripheral organs by metabolism of testosterone via the enzyme aromatase. In aging and obese men, estradiol values increase over time . Basar and colleagues  studied the relationship between scores of the Aging Male Symptoms and serum sex steroid levels and found that estradiol levels were greater in men with aging male symptoms. Since androgens are the precursors of estrogens, administration of exogenous testosterone will result in a potential increase in estradiol values. Recording periodic follow-up estradiol values in men on exogenous testosterone therapy is good medical practice. Estradiol has been shown to inhibit luteinizing hormone (LH) secretion in men (decreasing endogenous testosterone synthesis) and to increase the liver synthesis of SHBG (decreasing free unbound physiologically available testosterone) [82,83]. High estradiol values are considered detrimental to male sexual function. Estradiol values have been noted to be significantly higher in ED patients with venous leak than in controls, supporting the hypothesis that estradiol level can adversely influence penile smooth muscle function .
The physiologic role of dehydroepiandrosterone (DHEA) and DHEA sulphate (DHEA-S) are not well investigated. DHEA may be involved in cognitive, memory, metabolic, vascular, immune, and sexual functions . DHEA is an androgen precursor produced by the adrenal glands that exerts its effects via downstream conversion to testosterone and estradiol . DHEA deficiencies in men have been reportedly associated with various drugs, and endocrine, nonhormonal, and age-related disorders (DHEA steadily decreases from age 40). DHEA-S levels were significantly lower in men with sexual dysfunction, as determined by the International Index of Erectile Function (IIEF) score . Patients with ED and type 1 diabetes had lower levels of DHEA and DHEA-S compared with diabetics without ED . Also, low levels of DHEA and DHEA-S, but not free or total testosterone, were strongly associated with ED. No well-designed clinical trials have definitively substantiated the role of DHEA in these functions in humans, or even the safety and efficacy of DHEA therapy . In a small study, Reiter and colleagues  evaluated the efficacy of DHEA replacement in the treatment of ED and found it was associated with higher mean scores for all five domains of the IIEF with no impact on mean serum levels of PSA or testosterone.
188.8.131.52. Thyroid-stimulating hormone
Both hyperthyroidism and hypothyroidism have been shown to adversely affect sexual function [126,127]. It is likely that androgen therapy would not be successful until thyroid function has been normalized. In examining the clinical and hormonal profiles of patients, screening is performed by obtaining a thyroid-stimulating hormone (TSH) blood value for primary hypothyroidism. In suspected cases of central hypothyroidism, serum-free thyroxine (T4) is considered the best indicator . In men presenting for initial evaluation and therapy of ED, 4.0% had increased TSH .
184.108.40.206. Follicle-stimulating hormone and LH
Serum LH and follicle-stimulating hormone (FSH) determination may also be of value in men with ED and androgen insufficiency. Knowledge of these gonadotropin values will define whether the androgen insufficiency is due to hypogonadotropic hypogonadism versus hypergonadotropic hypogonadism [82,83].
3.2. Step care 2: Patient and partner education
A partner’s sexual health can be affected by the patient’s sexual dysfunction [130–134]. Thus, an essential component in the management of androgen insufficiency and ED is patient and partner education that is uniquely matched to individual needs . Educational subjects include an overview of pertinent anatomy and physiology, relevant pathophysiology, full disclosure of risks and benefits, and appropriate discussion of expectations with treatment. Efforts are made to translate the results of the history taking, physical examination, and laboratory testing into understandable management strategies in the presence of the patient and partner, if possible, with patient and partner’s preferences for management respected and taken into consideration .
3.3. Step care 3: Modifying reversible causes
Both androgen insufficiency and ED are potentially reversible if specific potentially reversible etiologic factors can be addressed. For example, weight loss has been shown to improve testosterone levels, reducing fat mass and estrogen levels [135,136]. Modification may apply to changing prescription or nonprescription drug use and/or altering psychosocial factors .
3.4. Step care 4: Hormonal and nonhormonal pharmacologic treatment
Safe and effective government-approved pharmacologic agents are available to treat androgen insufficiency and ED separately. Pharmacologic treatments are prescribed considering cost and ease of administration. Should optional hormonal blood testing be performed and suspicions for abnormal hormonal blood tests be identified, considerations for hormonal treatment should be discussed with the patient. Hormonal agents [82,83] include testosterone, DHEA, clomiphene citrate, aromatase inhibitors, 5-alpha reductase inhibitors, dopamine agonists, and thyroid therapies [82,83]. For androgen insufficiency, androgen delivery systems, listed chronologically, include oral testosterone , intramuscular depot injections , scrotal transdermal patch systems , nongenital skin transdermal patch systems , hydroalcoholic testosterone gels, [141,142], adhesive buccal tablets , and, recently, long-acting intramuscular depot injections . Nonhormonal treatments include vasodilators such as PDE5 inhibitors and intracavernosal/intraurethral agents . Before considering treatment for androgen insufficiency, a patient should show signs and symptoms and biochemical confirmation of androgen insufficiency, a PSA and DRE not consistent with prostate cancer or a negative prostate biopsy, and an absent history of breast cancer [82,83]. The patient should also meet the definition of ED .
Isidori and colleagues  determined that exogenous testosterone improved the number of nocturnal erections and successful intercourses, sexual thoughts, scores of erectile function, and overall sexual satisfaction in men with low testosterone, but had no effect on eugonadal men. They concluded that the effect of testosterone tended to decline over time and was progressively smaller with increasing baseline T levels, and that long-term safety data were not available . Relative contraindications to be considered include elevated hematocrit, abnormal liver function studies, LUTS, and sleep apnea. A rare adverse event is the enlargement of prostate size, which can be prevented by the administration of finasteride [147,148]. It has also been suggested that large numbers of men with low to low-normal testosterone levels would benefit from testosterone screening when they are evaluated for ED and that testosterone therapy may improve the response of PDE5 inhibitors [70,71,73].
Many men take DHEA without physician supervision, because it is sold over the counter. Saad and colleagues  noted DHEA supplementation had positive effects on the cardiovascular system, body composition, bone mineral density, skin, central nervous system, the immune system, and sexual function. DHEA use may be justified in aging men with periodic evaluations to maintain serum concentrations in the physiologic range . Recent evidence suggests that DHEA has a physiologic role through interaction with specific membrane receptors on the endothelium .
3.4.3. Clomiphene citrate
Exogenous testosterone may be deleterious in men with relative infertility, because it suppresses gonadotropins [82,83]. Alternatively, clomiphene citrate increases gonadotropins  and may be beneficial when the androgen insufficiency is due to hypogonadotropic hypogonadism. Guay et al  and Shabsigh et al  administered off-label clomiphene citrate to men with hypogonadotropic hypogonadism and found significant increases in LH and free testosterone, and improved sexual function. Erectile improvement was lower in men with aging, diabetes, hypertension, coronary artery disease, and multiple medication use. In another study, sexual function improved in ED patients using clomiphene in limited parameters in younger and healthier ED men .
3.4.4. Aromatase inhibitors
Administration of exogenous testosterone will result in increased estradiol values by aromatization. Anastrozole is a potent, highly selective aromatase inhibitor with no intrinsic steroid hormone agonist or antagonist activities . In a recent study, anastrozole administration increased serum bioavailable and total testosterone levels in older men with mild hypogonadism, while estradiol levels remained normal . The sexual benefits of aromatase inhibitor therapy were reported in a case report in which use of an aromatase inhibitor normalized the testosterone level and improved sexual functioning, possibly by a central alteration in the testosterone/estrogen ratio .
3.4.5. 5-Alpha reductase inhibitors
Common, distressing side effects of androgen therapy are hirsutism and acne . The most efficacious pharmacologic therapy to reduce DHT is via 5-alpha reductase inhibition. Mechanical therapies for hirsutism and topical and systemic acne therapies are also available.
3.4.6. Dopamine agonists
Dopamine agonists have been reported to improve sexual function  on the basis of research showing that sexual motivation is modulated by a number of central nervous system neurotransmitter and receptor changes induced, in part, by the action of sex steroids and by the central neurotransmitter dopamine. Dopamine neurotransmitter systems may play a critical intermediary role in the central regulation of sexual arousal and excitation, mood, and incentive-related sexual behavior, particularly in the motivational responses to conditioned external stimuli [160–164]. Although their use is controversial, more research is needed with dopamine agonists for men with androgen insufficiency and ED.
3.4.7. Thyroid hormones
If a patient with androgen insufficiency and ED has a concomitant thyroid abnormality, it is likely that androgen therapy would not be successful until the thyroid state was normalized. In men diagnosed with abnormalities in both thyroid function and sexual function (decreased sexual desire, ED, premature or delayed ejaculation), treatment with methimazole (for hyperthyroidism) or thyroxine (for hypothyroidism) for 8 wk without concomitant PDE5 inhibitor therapy resulted in an improvement in sexual function . In animal studies, hypothyroidism resulted in autonomic neuropathy and endothelial dysfunction, adversely influencing the release or synthesis of NO from nitrergic nerves and endothelium .
3.4.8. Phosphodiesterase inhibitors
Oral PDE5 inhibitors are approved for on-demand administration and are effective in facilitating and enhancing erections following sexual stimulation . In a recent review , a synergistic effect for testosterone therapy and efficacy of PDE5 inhibitor therapy in men with androgen insufficiency and ED was shown. In patients with androgen insufficiency, in whom treatment with testosterone supplementation alone failed, combined treatment with a PDE5 inhibitor and testosterone gel improved erectile function . Likewise, aging men with androgen insufficiency who failed first-line oral PDE5 inhibitor treatment and in whom androgens were not contraindicated had improved erectile function and quality of life when treated with a combination of testosterone and PDE5 inhibitors [74–76]. These findings provide clinical support to the experimental knowledge of the importance of androgens in regulating smooth muscle function. Of interest, sustained improvement in sexual function after 12 mo of PDE5 inhibitor administration has been associated with increased testosterone to estradiol ratio, mainly related to reduction of estradiol levels .
3.4.9. Follow-up strategies
Patients undergoing hormonal treatment for androgen insufficiency and ED should undergo reassessment at regular intervals to ensure optimum patient-physician communication to assess the progress of therapy and the sexual, general medical, and psychosocial status of the patient and partner . Total testosterone, SHBG, albumin (if appropriate), PSA, and DRE should be performed every 3–6 mo until the values are stable and in the appropriate range. Hematocrit and hemoglobulin, liver function tests, and bone density and lipid profile evaluations should be monitored annually. Follow-up also provides the occasion for critical continued education, addressing any relevant patient concerns regarding the treatments, including dose titration or change in medication. Adverse drug reactions or drug interaction effects should be carefully monitored [82,83].
3.5. Step care 5: Other treatments
Men with androgen insufficiency and ED may not respond to the previously discussed interventions and may need to consider such options as a vacuum erection device, intraurethral or intracavernosal administration of alprostadil or other vasoactive agents, or surgical intervention with penile prostheses or reconstructive surgery such as penile revascularization .
4. Summary, conclusions, and future directions
Androgen-dependent mechanisms that regulate genital tissue remodeling in the adult have been poorly defined. Characterization of the molecular and cellular mechanisms by which androgens regulate genital tissue structure and function would provide significant gains in knowledge and understanding of important pathogenic processes. These mechanisms need to be investigated using well-established experimental approaches to assess changes in penile hemodynamics, tissue structure, and cell-specific biomarkers. Such studies in animal models would initiate a novel line of investigation in genital physiology and may provide further scientific rationale for the judicious use of androgens in the management of male ED in men with androgen insufficiency. In light of the similarities in systemic and penile vascular disease, and the role of adipogenesis in the metabolic syndrome, this line of investigation may also stimulate future work on the role of androgens in systemic metabolic and vascular disease. While the NO/cGMP pathway plays a key role in erectile physiology, our knowledge of the downstream events that regulate gene expression in the penis is rudimentary at best. New approaches are needed to develop better understanding of the interplay between PDE5 expression and activation of the NO/cGMP pathway. The effects of androgens on the cavernosal and dorsal nerves also merit further investigation, and defining the effect of androgens on neurotransmitter synthesis and release would be of scientific and clinical value. Finally, tissue remodeling at the vascular, trabecular, and tunica albuginea levels is of paramount importance if we are to understand the relationship between androgen deficiency and venous leakage, and its restoration by androgen treatment.
Both the conditions of androgen insufficiency and ED are highly prevalent medical disorders in aging men with associated multiple risk factors. Good clinical practice requires the use of appropriate step care strategies for patient and goal-directed management. The future will likely see new basic science investigations that will lead to novel treatment strategies. In this fashion, management can be delivered in a more safe and effective manner for the majority of afflicted patients (and partners). It is appreciated that there are some who argue that there is little or no role for androgens in the management of ED. Indeed, healthy skepticism is warranted, but one must keep an open mind and weigh the evidence in making such an important scientific judgment. The emergence of clinical data from well-designed studies should provide the foundation for evidence-based medicine. We must recognize that humans have multiple pathways of generating androgens, not only in the endocrine glands but also in the periphery. It should be noted that a “back door” biosynthetic pathway for the production of 5α-DHTfrom progesterone was reported only recently . Ultimately, the common and binding goal of both clinicians and scientists is to develop better understanding of the role of androgens and erectile function in human health, and to be able to provide the best possible treatment strategies for patients afflicted with androgen deficiency and ED.
Both conditions of androgen insufficiency and ED are highly prevalent medical disorders in aging men with associated multiple risk factors. Good clinical practice requires the use of appropriate step care strategies for patient and goal-directed management. The future will likely see new basic science investigations that will lead to new, safe, and effective treatment strategies. The emergence of clinical data from well-designed studies should provide the foundation for evidence-based medicine. We must recognize that humans have multiple pathways of generating androgens, not only in the endocrine glands but also in the periphery. Ultimately, the common and binding goals of both clinicians and scientists are to develop better understanding of the role of androgens in human health and to be able to provide the best possible treatment strategies for patients afflicted with androgen deficiencies.
This work was supported by National Institutes of Health grants. The authors have nothing to disclose.
Publisher’s Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.