madman
Super Moderator
Testosterone replacement therapy and erectile dysfunction (2021)
Ifeanyi C. Onyeji and Raul I. Clavijo
Testosterone (T) deficiency and erectile dysfunction (ED) are independently functionally and socially impairing, and their concurrence in men can be challenging to treat. Successful management requires an understanding of the mechanisms through which T underlies normal erectile function. While the literature elucidating some of these mechanisms is vast (e.g., androgen regulation of the activity of nitric oxygen synthase and phosphodiesterase type 5) for others it is scarce (e.g., catalysts of castration induced corporal fibrosis). The randomized controlled trial data for the efficacy of T replacement as mono- or combination therapy to treat ED has been conflicting. Positive results were frequently not clinically meaningful. Meta-analyses have been helpful in illuminating trends that seem to be promising. Consensus is still lacking in several areas, such as the threshold of low T severity for which replacement therapy is most beneficial; the timing for initiating combination therapy; and the duration of treatment.
INTRODUCTION
A steroid hormone, testosterone (T) is synthesized by Leydig cells and upon reaching its target tissue, it can be metabolized to either dihydrotestosterone (DHT) by 5-alpha reductase or estradiol by aromatase [1]. T or DHT (the latter being more potent) then binds to a high-affinity intracellular receptor protein and this complex enters the nucleus and binds to DNA at recognition sites to regulate gene transcription [2]. T is regulated by the hypothalamic-pituitary-gonadal axis, and longitudinal data has shown that its production decreases with aging [3]. T deficiency or hypogonadism can be accompanied by symptoms across multiple systems, including fatigue, mood disturbance, diminished libido, and erectile dysfunction (ED) [4–6]. The parameters with the most consistent evidence for improvement after testosterone replacement therapy (TRT) are anemia, bone mineral density, lean body mass, depressive symptoms, and ED [7–10]
ED is defined as the “consistent or recurrent inability to attain and/or maintain penile erection sufficient for sexual satisfaction” [11]. Its etiology is multi-factorial in most men with risk factors that include psychiatric disorders, endocrine dysfunction, vascular disease, penile structural anomalies, substance abuse, and medication use [12–17]. Likewise, its management consists of lifestyle modification, psychotherapy, pharmacotherapy, and surgery [18–21]. In select hypogonadal men with ED, TRT is invaluable, and its implementation requires a review of penile erection physiology and interrogation of the clinical trials on which its efficacy is based.
METHODS
The aim of this article is to summarize current knowledge on the mechanisms through which T mediates penile erectile physiology and to discuss the efficacy of its supplementation in treating ED. It is not meant to serve as a systematic review on the topic.
*EFFECTS OF T ON THE MOLECULAR AND CELLULAR PHYSIOLOGY OF PENILE ERECTION
Penile erection is a dynamic neurovascular process involving relaxation of corporal cavernosal smooth muscle, dilation of arterioles, compression of the venous plexus, and expansion of the tunica albuginea [22]. Animal castration studies, which proliferated in the second half of the 20th century, and human studies have unequivocally demonstrated that T regulates each of these phases of penile erection physiology [23–39].
*Regulation of NOS expression
NO is a vital mediator of penile erectile function. It is synthesized by the conversion of L-arginine (which can be synthesized from L-citrulline) to NO by NOS. NOS has three isoforms: inducible nitrogen oxide synthase (iNOS), neuronal nitrogen oxide synthase (nNOS), and endothelial nitrogen oxide synthase (eNOS). In the NO-cyclic guanosine monophosphate pathway, after NO is released from cavernous nerve terminals, it diffuses into smooth muscle cells and binds soluble guanylate cyclase. Guanylate cyclase facilitates the conversion of guanosine triphosphate to cyclic guanosine monophosphate (cGMP), and this pathway decreases intracellular calcium which promotes smooth muscle relaxation [23]. The relationship between T and NO was evaluated by administering T or placebo to surgically castrated rats and measuring changes in erectile response [intracavernosal pressure/ mean arterial pressure] after direct electrostimulation and the transcription of nNOS messenger ribonucleic acid (mRNA) in the CC [24]. Both the magnitude of the erectile response and nNOS gene expression was significantly greater in T-repleted rats. Thus, T may enhance penile erection by upregulating NOS and potentiating NO signaling. Although another study [25] did not find such a T-mediated increase in nNOS activity, several others have supported the findings from this work [26–28].
*Regulation of penile arterial inflow and venous outflow
Increased blood flows into the cavernous sinuses and venous occlusion of these sinuses are requisite to achieving and sustaining an erection. These two phases were measured in the CC of castrated rats using Doppler sonography during an electrostimulated erection [29]. Castrated rats had significantly decreased blood flow into their CC and higher leakage of venous blood compared to their T-supplemented counterparts (Table 1). In a subsequent study, this group showed that the loss of venoocclusion in the CC of castrated rats was reversed by administering a NO donor medication [30]. Hence, T seems to regulate both the arteriogenic and venogenic phases of penile erection, and its impact on the veno-occlusive mechanism may be NO-mediated.
*Regulation of corporal cavernosal contractility
Sphingolipids are known to modulate vascular tone and the activity of Shingosine-1-phosphate (SIP) has been shown to promote contraction of the corporal cavernosal smooth muscle (CCSM) in rats [31]. SIP has three G protein-coupled protein receptors, two of which exert their effect via the RhoA/Rho-kinase signaling pathway. In brief, the binding of RhoA to Rho-kinase activates the latter and triggers signaling that increases calcium sensitization in penile smooth muscle cells and enhances contraction [32]. The tonic contraction of the CCSM is believed to be mediated by this calcium sensitizing pathway [32]. The interplay among T, SIP and CCSM contractility was explored in a rat model [33]. T deprivation significantly increased serum SIP levels, SIP-2, and SIP-3 receptor mRNA transcription; led to a twofold increase in CCSM contractility; and diminished erection response scores. T supplementation normalized these effects (Table 1). Thus, T deficiency may contribute to ED by upregulating SIP activity and impending the relaxation of the CCSM.
*Regulation of dorsal nerve innervation
T has been shown to be integral to the structural maintenance of peripheral autonomic neurons that innervate the male reproductive tract so its influence on the penile dorsal nerve—the primary source of afferent input—warranted investigation [34]. A study using light microscopy to examine dorsal nerve sections from control and castrated rats showed that castration increased the axonal cross-sectional area (i.e. swelling) of the dorsal nerve neurons and decreased the density of the dorsal nerve fibers [35]. T partially reversed the latter finding but this positive trend did not reach statistical significance. In another study, however, T supplementation significantly normalized the number of nicotinamide adenine dinucleotide phosphate-stained nerve fibers in the CC and dorsal nerve that had been diminished by castration [22]. These findings suggest both a quantitative and qualitative relationship between T and dorsal nerve proliferation and structural integrity, respectively
*Regulation of phosphodiesterase type 5
Given the impact of phosphodiesterase type 5 (PDE5) inhibitors on the management of ED, there has been a considerable inquiry into the interplay between PDE5 and T. An immunohistochemical analysis of human CC tissue found extensive PDE5 labeling in the smooth muscle cells [36]. The rabbit model showed that compared to controls, hypogonadotrophic hypogonadal rabbits had decreased transcription of PDE5 mRNA as well as decreased PDE5 labeling in the smooth muscle cells [36]. Interestingly, T supplementation restored PDE5 activity in the CC. The direct effect of T on PDE5, which seems counter-productive, has been replicated in other etiologies of T deficiency. Sickle cell disease (SCD) is associated with primary hypogonadism due to testicular failure and SCD-induced priapism has been attributed in part to the decreased activity of PDE5, resulting in unchecked accumulation of cGMP [37, 38]. In a mouse model of human sickle cell disease with T deficiency, T was found to correct the excessive erectile response to electrostimulation seen in sickle cell mice and also upregulated phosphorylation and activation of PDE5 in the CC to wild-type levels [39]. The finding that T enhances both PDE5 and NOS activity seems to undermine its potential positive effect on erectile function, but one hypothesis is that this mechanism leads to a recycling of factors that are necessary to respond to the next sexual stimuli [39].
*Regulation of penile structural integrity and fibroelasticity
Rat CC is filled with smooth muscle trabeculae surrounding endothelium-lined sinusoids and connective tissue. DNA fragmentation is an established marker of cellular apoptosis. The presence of DNA fragments and apoptotic nuclei were analyzed with fluorescence microscopy in the CC of castrated rats, castrated rats post-T supplementation, and controls [40]. The authors found that T deprivation significantly increased DNA fragmentation and the apoptotic rate of CC cells. Unfortunately, the study could not localize which cell types underwent apoptosis but others have confirmed that castration induces smooth muscle cell apoptosis [41, 42].
This apoptotic effect of T deficiency on CCSM is pertinent to Peyronie’s disease (PD). Although the definitive mechanism of PD has yet to be elucidated, it is reputed to derive from the healing process after microtrauma in the tunica which promotes the deposition of collagen matrix—while suppressing its degradation —resulting in fibrotic inelastic plaques in the CC [43, 44]. A decrease in the CCSM relative to the connective tissue matrix (i.e., smooth muscle to connective tissue ratio), is thought to engender CC fibrosis and possibly PD [45]. Two studies using similar staining techniques showed that T deficiency decreased the smooth muscle to connective tissue ratio and that T repletion greatly improved but did not quite normalize it [42, 46]. As seen in Table 1, the latter authors also found that extracellular collagen proteins were more abundant in castrated rats. These histology studies suggest that by maintaining the survival of individual CCSM cells, T may deter connective tissue fibrosis and maximize the aggregate sinusoidal elasticity, which limits excessive venous outflow that would result in poorly sustained erections. It remains unclear if T supplementation would have a significant effect on the fibrotic plaques seen in clinical PD.
*EFFICACY OF T-REPLACEMENT THERAPY ON CLINICAL ERECTIONS
*T-replacement monotherapy
Several clinical trials have investigated the efficacy of TRT in ameliorating ED, and it is commendable that different formulations of T supplementation and the gold standard questionnaire for measuring erectile function (i.e., the International Index of Erectile Function, IIEF) were utilized in these studies [47]. An increase in the IIEF erectile function domain (EFD) score of 4 points after treatment with oral pharmacotherapy has been shown to indicate a meaningful clinical improvement [47]. This threshold varies depending on the severity of baseline ED with mild, moderate, and severe ED requiring changes of 2, 5, and 7 points, respectively [47].
As T decreases with aging, it was imperative to evaluate the benefit of TRT on ED in hypogonadal older men. In the Sexual Function Trial, 470 men aged 65 or older with symptomatic hypogonadism (serum T <275 ng/dL) were randomized to either T gel 1% or placebo for 12 months and their EFD scores were assessed quarterly [48]. The TRT group had a statistically significant greater increase in EFD score (3.1 ± 6.9 vs 1.0 ± 6.0), but this change of fewer than 4 points does would likely not enhance clinical performance. In one of the largest multi-institutional, double-blinded, randomized control trials (RCT), 715 men with hypogonadism (mean of two total serum T levels less than 300 ng/dL and at least one symptom of low T) received either daily 60 mg of topical T solution 2% or placebo [10]. Approximately 81% of the patients had an EFD score <26, indicating some degree of ED. Two noteworthy shortcomings are that 25% of patients in the T group did not achieve normal T levels when the final IIEF questionnaire was completed and that the authors did not report a sub-group analysis focusing on the severity of baseline ED or T deficiency. Stratification by baseline ED severity could reveal why the mean increase in EFD score in the T group compared with placebo (3.4 ± 0.47 vs 0.7 ± 0.47) was statistically significant but did not meet the criteria for a meaningful clinical difference (Table 2).
Three meta-analyses have sought to answer the question of the T deficiency threshold at which TRT is effective for men with ED, and the results are conflicting. The first stratified 615 subjects from 17 RCTs by baseline serum T (<201 ng/dL, 201–346, and >346) and found that the positive effect on T replacement on EFD score was stronger in those with the lowest baseline T compared to those with the highest levels (1.26 vs 0.86) [49]. This direct relationship between T deficiency severity and improvement in erectile function was statistically significant (see Table 2), but because the authors did not report the mean change in the EFD score, the clinical importance of this finding is unclear. The results of the second study, which included a similar number of RCTs, contradict those of the first. It found a statistically significant increase in erectile function after T replacement in men with low normal and normal baseline T (>300 ng/dL) and not in those with severe T deficiency (<300 ng/dL) (see Table 2) [50]. Again, clinical significance cannot be ascertained for similar reasons. The third study, however, provided such data. The greater improvement in EFD score in the group with the lowest T serum compared to those with a milder form of T deficiency was statistically significant (mean change 2.95 vs 1.97) but would not impact clinical performance [51]. Of note, the inverse relationship between the efficacy of TRT and a eugonadal state may explain why a contemporary study did not find a benefit of T supplementation on erectile function, as nearly half of the study population had normal baseline levels of T [52].
*None of these meta-analyses assessed the impact of baseline severity of ED on the efficacy of TRT. One study [47] hinted that the effect TRT diminishes in men with severe ED, which supports the speculation that whereas men with severe T deficiency and mild-moderate ED may benefit from TRT alone, those with severe T deficiency and severe ED may warrant dual therapy.
*Combination therapy with T replacement and phosphodiesterase type 5 inhibitors Phosphodiesterase type 5 inhibitors
(PDE5i) revolutionized the treatment of ED but despite their success as first-line agents, ~30–35% do not respond adequately [53–55]. As such, TRT has been combined with PDE5i to improve outcomes and the evidence for the efficacy of combination therapy is inconsistent.
In a study of 140 men with ED and low serum T, the patients were first optimized to a clinically effective dose of sildenafil and then randomized to either topical T gel 1% or placebo gel [56]. T normalization in the experimental group was verified. Although the EFD scores increased from baseline to PDE5i optimization by a mean of 7.7 points in the entire cohort, after receiving T or placebo for 14 weeks, erectile function did not differ statistically between the two groups (mean change of ~1.0 in both groups). Stratification by age and severity of T deficiency did not change the results.
These findings contrast with those from earlier work in men who had failed sildenafil monotherapy and went on to receive either T gel or placebo [57]. Here, the improvement in erectile function was both statistically significant and clinically meaningful (mean change in EFD of 4.4 vs 2.1), the first of the clinical trials discussed thus far. The TADTEST study added to these results. In their analysis of non-responders to oral PDE5i therapy, the authors found that a baseline T <300 ng/dL was the threshold for which T replacement had a statistically and clinically significant impact on ED (mean change EFD 6.2 vs 2.3) [58]. Thus, there seems to be a clinical role for TRT in ED and in particular, in patients with severe T deficiency and refractory ED. Their results also refute the notion that sequential combination therapy is not effective because PDE5i exhausts the pathways on which T would act to enhance erectile function [56].
When using tadalafil as the PDE5i in combination therapy, the daily dosage might be more efficacious than on-demand dosing. In a study with 9 months of follow-up, 60 hypogonadal men received 1000 mg of T undecanoate and then half of the subjects took daily tadalafil 5 mg and the other took tadalafil 10–20 mg for on-demand use [59]. Improvement in total IIEF score peaked at week 30 and significantly favored the daily-dosed PDE5i group compared to the on-demand group (see Table 2). Notably, both groups experienced a decline in their erectile function gains and serum T levels 6 weeks after stopping combination therapy, suggesting that this therapy may require long-term use for sustained effect.
The most recent meta-analysis on the efficacy of combination therapy included eight RCTs with 913 hypogonadal men with ED. TRT plus PDE5i was compared to PDE5i plus placebo [60]. Areas of heterogeneity include criteria for hypogonadism (total serum T ≤ 330 to 400 ng/dL); baseline ED criteria (IIEF-EFD score <25 or poor response to the previous PDE5i); formulation of TRT (topical gel 1% or oral undecanoate of varying dosage titrations); PDE5i therapy (sildenafil 50–100 mg on-demand or tadalafil 5–10 mg daily); assessment of change in erectile function (IIEF-15 EFD score or the abbreviated IIEF-5 score); and follow-up duration (range 8–14 weeks). The pooled change in erectile function favored the combination group (Table 2). However, stratification analysis revealed that the beneficial effect of combination therapy did not reach statistical significance in men with baseline T levels greater than 288 ng/dL (vs ≤ 288 ng/dL). This may indicate that in men with ED and serum T at the lower limits of normal, PDE5i alone could be offered upfront and after the failure of monotherapy, TRT could be used adjunctively.
CONCLUSIONS AND FUTURE RESEARCH
The literature presented herein illustrates that T mediates erectile function, but it is imperative to note that T is not necessary for satisfactory erections. Two small prospective studies found that erectile responses to visual erotic stimuli were equivalent between hypogonadal men and healthy controls [61, 62]. However, nocturnal penile tumescence and spontaneous daytime erections were diminished in hypogonadal men. This points to the observation that motivation and interest through sexual stimuli can overcome deficiencies in the physiology of the erectile response brought upon by hypogonadism.
Several questions remain despite the robust literature that exists. In the physiology realm, certain pathways may present an opportunity for novel drug targets. For example, the study that examined castration-induced upregulation of sphingosine-1- phosphate and consequent contraction of rat CCSM showed that administering an S1P receptor antagonist reinstated CCSM relaxation [33]. From the structural aspect, corporal cavernosal fibrosis is associated with T deprivation but few studies have assessed the restorative effects of T on animal models of PD.
Clinically, the optimal population of men with ED who will benefit most from TRT is unclear. There were inconsistent results on the impact of age on treatment efficacy and while multiple RCTs have focused on just older men, a parallel large study on younger men is still lacking. The threshold of ED and T deficiency severity for which TRT is efficacious, the dosing of TRT, the timing for initiating combination therapy, and the duration of the treatment have yet to be outlined in clinical guidelines. Presumably, TRT monotherapy could be first-line for patients with severe T deficiency but milder forms of ED (with PDE5i as salvage therapy if the response is poor), whereas combination TRT and PDE5i would be first-line for those with severe T deficiency and severe ED.
Ifeanyi C. Onyeji and Raul I. Clavijo
Testosterone (T) deficiency and erectile dysfunction (ED) are independently functionally and socially impairing, and their concurrence in men can be challenging to treat. Successful management requires an understanding of the mechanisms through which T underlies normal erectile function. While the literature elucidating some of these mechanisms is vast (e.g., androgen regulation of the activity of nitric oxygen synthase and phosphodiesterase type 5) for others it is scarce (e.g., catalysts of castration induced corporal fibrosis). The randomized controlled trial data for the efficacy of T replacement as mono- or combination therapy to treat ED has been conflicting. Positive results were frequently not clinically meaningful. Meta-analyses have been helpful in illuminating trends that seem to be promising. Consensus is still lacking in several areas, such as the threshold of low T severity for which replacement therapy is most beneficial; the timing for initiating combination therapy; and the duration of treatment.
INTRODUCTION
A steroid hormone, testosterone (T) is synthesized by Leydig cells and upon reaching its target tissue, it can be metabolized to either dihydrotestosterone (DHT) by 5-alpha reductase or estradiol by aromatase [1]. T or DHT (the latter being more potent) then binds to a high-affinity intracellular receptor protein and this complex enters the nucleus and binds to DNA at recognition sites to regulate gene transcription [2]. T is regulated by the hypothalamic-pituitary-gonadal axis, and longitudinal data has shown that its production decreases with aging [3]. T deficiency or hypogonadism can be accompanied by symptoms across multiple systems, including fatigue, mood disturbance, diminished libido, and erectile dysfunction (ED) [4–6]. The parameters with the most consistent evidence for improvement after testosterone replacement therapy (TRT) are anemia, bone mineral density, lean body mass, depressive symptoms, and ED [7–10]
ED is defined as the “consistent or recurrent inability to attain and/or maintain penile erection sufficient for sexual satisfaction” [11]. Its etiology is multi-factorial in most men with risk factors that include psychiatric disorders, endocrine dysfunction, vascular disease, penile structural anomalies, substance abuse, and medication use [12–17]. Likewise, its management consists of lifestyle modification, psychotherapy, pharmacotherapy, and surgery [18–21]. In select hypogonadal men with ED, TRT is invaluable, and its implementation requires a review of penile erection physiology and interrogation of the clinical trials on which its efficacy is based.
METHODS
The aim of this article is to summarize current knowledge on the mechanisms through which T mediates penile erectile physiology and to discuss the efficacy of its supplementation in treating ED. It is not meant to serve as a systematic review on the topic.
*EFFECTS OF T ON THE MOLECULAR AND CELLULAR PHYSIOLOGY OF PENILE ERECTION
Penile erection is a dynamic neurovascular process involving relaxation of corporal cavernosal smooth muscle, dilation of arterioles, compression of the venous plexus, and expansion of the tunica albuginea [22]. Animal castration studies, which proliferated in the second half of the 20th century, and human studies have unequivocally demonstrated that T regulates each of these phases of penile erection physiology [23–39].
*Regulation of NOS expression
NO is a vital mediator of penile erectile function. It is synthesized by the conversion of L-arginine (which can be synthesized from L-citrulline) to NO by NOS. NOS has three isoforms: inducible nitrogen oxide synthase (iNOS), neuronal nitrogen oxide synthase (nNOS), and endothelial nitrogen oxide synthase (eNOS). In the NO-cyclic guanosine monophosphate pathway, after NO is released from cavernous nerve terminals, it diffuses into smooth muscle cells and binds soluble guanylate cyclase. Guanylate cyclase facilitates the conversion of guanosine triphosphate to cyclic guanosine monophosphate (cGMP), and this pathway decreases intracellular calcium which promotes smooth muscle relaxation [23]. The relationship between T and NO was evaluated by administering T or placebo to surgically castrated rats and measuring changes in erectile response [intracavernosal pressure/ mean arterial pressure] after direct electrostimulation and the transcription of nNOS messenger ribonucleic acid (mRNA) in the CC [24]. Both the magnitude of the erectile response and nNOS gene expression was significantly greater in T-repleted rats. Thus, T may enhance penile erection by upregulating NOS and potentiating NO signaling. Although another study [25] did not find such a T-mediated increase in nNOS activity, several others have supported the findings from this work [26–28].
*Regulation of penile arterial inflow and venous outflow
Increased blood flows into the cavernous sinuses and venous occlusion of these sinuses are requisite to achieving and sustaining an erection. These two phases were measured in the CC of castrated rats using Doppler sonography during an electrostimulated erection [29]. Castrated rats had significantly decreased blood flow into their CC and higher leakage of venous blood compared to their T-supplemented counterparts (Table 1). In a subsequent study, this group showed that the loss of venoocclusion in the CC of castrated rats was reversed by administering a NO donor medication [30]. Hence, T seems to regulate both the arteriogenic and venogenic phases of penile erection, and its impact on the veno-occlusive mechanism may be NO-mediated.
*Regulation of corporal cavernosal contractility
Sphingolipids are known to modulate vascular tone and the activity of Shingosine-1-phosphate (SIP) has been shown to promote contraction of the corporal cavernosal smooth muscle (CCSM) in rats [31]. SIP has three G protein-coupled protein receptors, two of which exert their effect via the RhoA/Rho-kinase signaling pathway. In brief, the binding of RhoA to Rho-kinase activates the latter and triggers signaling that increases calcium sensitization in penile smooth muscle cells and enhances contraction [32]. The tonic contraction of the CCSM is believed to be mediated by this calcium sensitizing pathway [32]. The interplay among T, SIP and CCSM contractility was explored in a rat model [33]. T deprivation significantly increased serum SIP levels, SIP-2, and SIP-3 receptor mRNA transcription; led to a twofold increase in CCSM contractility; and diminished erection response scores. T supplementation normalized these effects (Table 1). Thus, T deficiency may contribute to ED by upregulating SIP activity and impending the relaxation of the CCSM.
*Regulation of dorsal nerve innervation
T has been shown to be integral to the structural maintenance of peripheral autonomic neurons that innervate the male reproductive tract so its influence on the penile dorsal nerve—the primary source of afferent input—warranted investigation [34]. A study using light microscopy to examine dorsal nerve sections from control and castrated rats showed that castration increased the axonal cross-sectional area (i.e. swelling) of the dorsal nerve neurons and decreased the density of the dorsal nerve fibers [35]. T partially reversed the latter finding but this positive trend did not reach statistical significance. In another study, however, T supplementation significantly normalized the number of nicotinamide adenine dinucleotide phosphate-stained nerve fibers in the CC and dorsal nerve that had been diminished by castration [22]. These findings suggest both a quantitative and qualitative relationship between T and dorsal nerve proliferation and structural integrity, respectively
*Regulation of phosphodiesterase type 5
Given the impact of phosphodiesterase type 5 (PDE5) inhibitors on the management of ED, there has been a considerable inquiry into the interplay between PDE5 and T. An immunohistochemical analysis of human CC tissue found extensive PDE5 labeling in the smooth muscle cells [36]. The rabbit model showed that compared to controls, hypogonadotrophic hypogonadal rabbits had decreased transcription of PDE5 mRNA as well as decreased PDE5 labeling in the smooth muscle cells [36]. Interestingly, T supplementation restored PDE5 activity in the CC. The direct effect of T on PDE5, which seems counter-productive, has been replicated in other etiologies of T deficiency. Sickle cell disease (SCD) is associated with primary hypogonadism due to testicular failure and SCD-induced priapism has been attributed in part to the decreased activity of PDE5, resulting in unchecked accumulation of cGMP [37, 38]. In a mouse model of human sickle cell disease with T deficiency, T was found to correct the excessive erectile response to electrostimulation seen in sickle cell mice and also upregulated phosphorylation and activation of PDE5 in the CC to wild-type levels [39]. The finding that T enhances both PDE5 and NOS activity seems to undermine its potential positive effect on erectile function, but one hypothesis is that this mechanism leads to a recycling of factors that are necessary to respond to the next sexual stimuli [39].
*Regulation of penile structural integrity and fibroelasticity
Rat CC is filled with smooth muscle trabeculae surrounding endothelium-lined sinusoids and connective tissue. DNA fragmentation is an established marker of cellular apoptosis. The presence of DNA fragments and apoptotic nuclei were analyzed with fluorescence microscopy in the CC of castrated rats, castrated rats post-T supplementation, and controls [40]. The authors found that T deprivation significantly increased DNA fragmentation and the apoptotic rate of CC cells. Unfortunately, the study could not localize which cell types underwent apoptosis but others have confirmed that castration induces smooth muscle cell apoptosis [41, 42].
This apoptotic effect of T deficiency on CCSM is pertinent to Peyronie’s disease (PD). Although the definitive mechanism of PD has yet to be elucidated, it is reputed to derive from the healing process after microtrauma in the tunica which promotes the deposition of collagen matrix—while suppressing its degradation —resulting in fibrotic inelastic plaques in the CC [43, 44]. A decrease in the CCSM relative to the connective tissue matrix (i.e., smooth muscle to connective tissue ratio), is thought to engender CC fibrosis and possibly PD [45]. Two studies using similar staining techniques showed that T deficiency decreased the smooth muscle to connective tissue ratio and that T repletion greatly improved but did not quite normalize it [42, 46]. As seen in Table 1, the latter authors also found that extracellular collagen proteins were more abundant in castrated rats. These histology studies suggest that by maintaining the survival of individual CCSM cells, T may deter connective tissue fibrosis and maximize the aggregate sinusoidal elasticity, which limits excessive venous outflow that would result in poorly sustained erections. It remains unclear if T supplementation would have a significant effect on the fibrotic plaques seen in clinical PD.
*EFFICACY OF T-REPLACEMENT THERAPY ON CLINICAL ERECTIONS
*T-replacement monotherapy
Several clinical trials have investigated the efficacy of TRT in ameliorating ED, and it is commendable that different formulations of T supplementation and the gold standard questionnaire for measuring erectile function (i.e., the International Index of Erectile Function, IIEF) were utilized in these studies [47]. An increase in the IIEF erectile function domain (EFD) score of 4 points after treatment with oral pharmacotherapy has been shown to indicate a meaningful clinical improvement [47]. This threshold varies depending on the severity of baseline ED with mild, moderate, and severe ED requiring changes of 2, 5, and 7 points, respectively [47].
As T decreases with aging, it was imperative to evaluate the benefit of TRT on ED in hypogonadal older men. In the Sexual Function Trial, 470 men aged 65 or older with symptomatic hypogonadism (serum T <275 ng/dL) were randomized to either T gel 1% or placebo for 12 months and their EFD scores were assessed quarterly [48]. The TRT group had a statistically significant greater increase in EFD score (3.1 ± 6.9 vs 1.0 ± 6.0), but this change of fewer than 4 points does would likely not enhance clinical performance. In one of the largest multi-institutional, double-blinded, randomized control trials (RCT), 715 men with hypogonadism (mean of two total serum T levels less than 300 ng/dL and at least one symptom of low T) received either daily 60 mg of topical T solution 2% or placebo [10]. Approximately 81% of the patients had an EFD score <26, indicating some degree of ED. Two noteworthy shortcomings are that 25% of patients in the T group did not achieve normal T levels when the final IIEF questionnaire was completed and that the authors did not report a sub-group analysis focusing on the severity of baseline ED or T deficiency. Stratification by baseline ED severity could reveal why the mean increase in EFD score in the T group compared with placebo (3.4 ± 0.47 vs 0.7 ± 0.47) was statistically significant but did not meet the criteria for a meaningful clinical difference (Table 2).
Three meta-analyses have sought to answer the question of the T deficiency threshold at which TRT is effective for men with ED, and the results are conflicting. The first stratified 615 subjects from 17 RCTs by baseline serum T (<201 ng/dL, 201–346, and >346) and found that the positive effect on T replacement on EFD score was stronger in those with the lowest baseline T compared to those with the highest levels (1.26 vs 0.86) [49]. This direct relationship between T deficiency severity and improvement in erectile function was statistically significant (see Table 2), but because the authors did not report the mean change in the EFD score, the clinical importance of this finding is unclear. The results of the second study, which included a similar number of RCTs, contradict those of the first. It found a statistically significant increase in erectile function after T replacement in men with low normal and normal baseline T (>300 ng/dL) and not in those with severe T deficiency (<300 ng/dL) (see Table 2) [50]. Again, clinical significance cannot be ascertained for similar reasons. The third study, however, provided such data. The greater improvement in EFD score in the group with the lowest T serum compared to those with a milder form of T deficiency was statistically significant (mean change 2.95 vs 1.97) but would not impact clinical performance [51]. Of note, the inverse relationship between the efficacy of TRT and a eugonadal state may explain why a contemporary study did not find a benefit of T supplementation on erectile function, as nearly half of the study population had normal baseline levels of T [52].
*None of these meta-analyses assessed the impact of baseline severity of ED on the efficacy of TRT. One study [47] hinted that the effect TRT diminishes in men with severe ED, which supports the speculation that whereas men with severe T deficiency and mild-moderate ED may benefit from TRT alone, those with severe T deficiency and severe ED may warrant dual therapy.
*Combination therapy with T replacement and phosphodiesterase type 5 inhibitors Phosphodiesterase type 5 inhibitors
(PDE5i) revolutionized the treatment of ED but despite their success as first-line agents, ~30–35% do not respond adequately [53–55]. As such, TRT has been combined with PDE5i to improve outcomes and the evidence for the efficacy of combination therapy is inconsistent.
In a study of 140 men with ED and low serum T, the patients were first optimized to a clinically effective dose of sildenafil and then randomized to either topical T gel 1% or placebo gel [56]. T normalization in the experimental group was verified. Although the EFD scores increased from baseline to PDE5i optimization by a mean of 7.7 points in the entire cohort, after receiving T or placebo for 14 weeks, erectile function did not differ statistically between the two groups (mean change of ~1.0 in both groups). Stratification by age and severity of T deficiency did not change the results.
These findings contrast with those from earlier work in men who had failed sildenafil monotherapy and went on to receive either T gel or placebo [57]. Here, the improvement in erectile function was both statistically significant and clinically meaningful (mean change in EFD of 4.4 vs 2.1), the first of the clinical trials discussed thus far. The TADTEST study added to these results. In their analysis of non-responders to oral PDE5i therapy, the authors found that a baseline T <300 ng/dL was the threshold for which T replacement had a statistically and clinically significant impact on ED (mean change EFD 6.2 vs 2.3) [58]. Thus, there seems to be a clinical role for TRT in ED and in particular, in patients with severe T deficiency and refractory ED. Their results also refute the notion that sequential combination therapy is not effective because PDE5i exhausts the pathways on which T would act to enhance erectile function [56].
When using tadalafil as the PDE5i in combination therapy, the daily dosage might be more efficacious than on-demand dosing. In a study with 9 months of follow-up, 60 hypogonadal men received 1000 mg of T undecanoate and then half of the subjects took daily tadalafil 5 mg and the other took tadalafil 10–20 mg for on-demand use [59]. Improvement in total IIEF score peaked at week 30 and significantly favored the daily-dosed PDE5i group compared to the on-demand group (see Table 2). Notably, both groups experienced a decline in their erectile function gains and serum T levels 6 weeks after stopping combination therapy, suggesting that this therapy may require long-term use for sustained effect.
The most recent meta-analysis on the efficacy of combination therapy included eight RCTs with 913 hypogonadal men with ED. TRT plus PDE5i was compared to PDE5i plus placebo [60]. Areas of heterogeneity include criteria for hypogonadism (total serum T ≤ 330 to 400 ng/dL); baseline ED criteria (IIEF-EFD score <25 or poor response to the previous PDE5i); formulation of TRT (topical gel 1% or oral undecanoate of varying dosage titrations); PDE5i therapy (sildenafil 50–100 mg on-demand or tadalafil 5–10 mg daily); assessment of change in erectile function (IIEF-15 EFD score or the abbreviated IIEF-5 score); and follow-up duration (range 8–14 weeks). The pooled change in erectile function favored the combination group (Table 2). However, stratification analysis revealed that the beneficial effect of combination therapy did not reach statistical significance in men with baseline T levels greater than 288 ng/dL (vs ≤ 288 ng/dL). This may indicate that in men with ED and serum T at the lower limits of normal, PDE5i alone could be offered upfront and after the failure of monotherapy, TRT could be used adjunctively.
CONCLUSIONS AND FUTURE RESEARCH
The literature presented herein illustrates that T mediates erectile function, but it is imperative to note that T is not necessary for satisfactory erections. Two small prospective studies found that erectile responses to visual erotic stimuli were equivalent between hypogonadal men and healthy controls [61, 62]. However, nocturnal penile tumescence and spontaneous daytime erections were diminished in hypogonadal men. This points to the observation that motivation and interest through sexual stimuli can overcome deficiencies in the physiology of the erectile response brought upon by hypogonadism.
Several questions remain despite the robust literature that exists. In the physiology realm, certain pathways may present an opportunity for novel drug targets. For example, the study that examined castration-induced upregulation of sphingosine-1- phosphate and consequent contraction of rat CCSM showed that administering an S1P receptor antagonist reinstated CCSM relaxation [33]. From the structural aspect, corporal cavernosal fibrosis is associated with T deprivation but few studies have assessed the restorative effects of T on animal models of PD.
Clinically, the optimal population of men with ED who will benefit most from TRT is unclear. There were inconsistent results on the impact of age on treatment efficacy and while multiple RCTs have focused on just older men, a parallel large study on younger men is still lacking. The threshold of ED and T deficiency severity for which TRT is efficacious, the dosing of TRT, the timing for initiating combination therapy, and the duration of the treatment have yet to be outlined in clinical guidelines. Presumably, TRT monotherapy could be first-line for patients with severe T deficiency but milder forms of ED (with PDE5i as salvage therapy if the response is poor), whereas combination TRT and PDE5i would be first-line for those with severe T deficiency and severe ED.
