As a library, NLM provides access to scientific literature.

Inclusion in an NLM
database does not imply endorsement of, or agreement with, the contents by
NLM or the National Institutes of Health.
Learn more: PMC Disclaimer | PMC Copyright Notice

Nat Rev Dis Primers. Author manuscript; available in PMC 2017 Feb 4. PMCID: PMC5027992
Published in final edited form as: NIHMSID: NIHMS815695
Nat Rev Dis Primers. 2016; 2: 16003. PMID: 27188339
Published online 2016 Feb 4. doi: 10.1038/nrdp.2016.3

Erectile dysfunction
Faysal A. Yafi,1 Lawrence Jenkins,2 Maarten Albersen,3 Giovanni Corona,4 Andrea M. Isidori,5
Shari Goldfarb,6 Mario Maggi,7 Christian J. Nelson,6 Sharon Parish,8 Andrea Salonia,9 Ronny Tan,10
John P. Mulhall,2 and Wayne J. G. Hellstrom1

Abstract

Erectile dysfunction is a multidimensional but common male sexual dysfunction that in‐
volves an alteration in any of the components of the erectile response, including organic,
relational and psychological. Roles for nonendocrine (neurogenic, vasculogenic and iatro‐
genic) and endocrine pathways have been proposed. Owing to its strong association with
metabolic syndrome and cardiovascular disease, cardiac assessment may be warranted in
men with symptoms of erectile dysfunction. Minimally invasive interventions to relieve the
symptoms of erectile dysfunction include lifestyle modifications, oral drugs, injected va‐
sodilator agents and vacuum erection devices. Surgical therapies are reserved for the subset
of patients who have contraindications to these nonsurgical interventions, those who experi‐
ence adverse effects from (or are refractory to) medical therapy and those who also have
penile fibrosis or penile vascular insufficiency. Erectile dysfunction can have deleterious
effects on a man’s quality of life; most patients have symptoms of depression and anxiety
related to sexual performance. These symptoms, in turn, affect his partner’s sexual experi‐
ence and the couple’s quality of life. This Primer highlights numerous aspects of erectile

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 1 of 51
:
 dysfunction, summarizes new treatment targets and ongoing preclinical studies that evalu‐
ate new pharmacotherapies, and covers the topic of regenerative medicine, which represents
the future of sexual medicine.

The erect penis has always been a symbol of a man’s virility and sexual prowess. Although
it is not a lethal condition, the interest surrounding erectile dysfunction and its remedies has
been constant throughout the ages1–5 (FIG. 1). Erectile dysfunction is the inability to
achieve or maintain an erection that is sufficient for satisfactory sexual performance, and
affects a considerable proportion of men at least occasionally1. Two major aspects of the
male erection, the reflex erection and psychogenic erection, can be involved in the dysfunc‐
tion and are subject to therapeutic intervention: the reflex erection is achieved by directly
touching the penile shaft and is under the control of the peripheral nerves and the lower
parts of the spinal cord; and the psychogenic erection is achieved by erotic or emotional
stimuli, and uses the limbic system of the brain. The severity of erectile dysfunction is often
described as mild, moderate or severe according to the five-item International Index of
Erectile Function (IIEF-5) questionnaire, with a score of 1–7 indicating severe, 8–11 mod‐
erate, 12–16 mild–moderate, 17–21 mild and 22–25 no erectile dysfunction.

Figure 1

Timeline of the understanding and treatment of erectile dysfunction

As it became understood that an erection is a predominantly vasculogenic process, filling the cavernosal
bodies with blood became one of the key features of different modalities of treatment of erectile dys‐
function. For example, the vacuum erection device of today took form when tyre technician Geddings
Osbon invented the youth equivalency device in 1960, which combines the effect of a vacuum that
draws blood into the penis and the penile ring placed at the base of the penis to occlude venous return.
Injectable therapies became prominent following the infamous Brindley lecture ‘Cavernosal α-blockade:
a technique for treating erectile impotence’ at the American Urological Association Meeting in Las Ve‐
gas, Nevada, USA, in 1983 (REF. 222).

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 2 of 51
:
 In the past, erectile dysfunction was considered, in most cases, to be a purely psychogenic
disorder, but current evidence suggests that more than 80% of cases have an organic aetiol‐
ogy. Causes of organic erectile dysfunction can now be broadly divided into nonendocrine
and endocrine. Of the nonendocrine aetiologies, vasculogenic (affecting blood supply) is
the most common and can involve arterial inflow disorders and abnormalities of venous
outflow (corporeal veno-occlusion); there are also neurogenic (affecting innervation and
nervous function) and iatrogenic (relating to a medical or surgical treatment) aetiologies. In
terms of endocrine factors leading to erectile dysfunction, reduced serum testosterone lev‐
els have been implicated, but the exact mechanism has not been fully elucidated. Often, or‐
ganic erectile dysfunction involves a psychological component; that is, regardless of the
precipitating event, erectile dysfunction imposes negative effects on interpersonal relation‐
ships, mood and quality of life.

Importantly, erectile dysfunction is no longer simply confined to sexual activities but acts as
an indicator of systemic endothelial dysfunction1. From a clinical standpoint, erectile dys‐
function often precedes cardiovascular events and can be used as an early marker to identify
men at high risk of major cardiovascular disease6. In this Primer, we describe the different
aetiologies of erectile dysfunction and the currently available treatments.

Epidemiology

Several studies have explored the epidemiology of erectile dysfunction by considering dif‐
ferent settings and populations. Given that erectile dysfunction is regarded as a condition
that is more prevalent in older men, two milestone studies have provided valuable results in
this setting: the Massachusetts Male Ageing Study (MMAS) and the European Male Age‐
ing Study (EMAS)7,8. The MMAS showed a combined prevalence of mild to moderate
erectile dysfunction of 52% in men aged 40–70 years; erectile dysfunction was strongly re‐
lated to age, health status and emotional function7. Conversely, the EMAS, the largest Eu‐
ropean multicentre population-based study of ageing men (40–79 years), reported a preva‐
lence of erectile dysfunction ranging from 6% to 64% depending on different age subgroups
and increasing with age, with an average prevalence of 30% (REF. 8) (FIG. 2). Few studies
have evaluated erectile dysfunction prevalence worldwide9–12. What emerges from these
studies is a systematically higher prevalence of erectile dysfunction in the United States and
eastern and southeastern Asian countries than in Europe or South America. Several factors
can account for these differences, including cultural or socioeconomic variables; however,
further studies are required to identify and discriminate possible genetic influences from
environmental impact. Data on erectile dysfunction incidence are less abundant; new cases
range from 19 to 66 per 1,000 men every year in studies in the United States, Brazil and the
Netherlands13–15. However, these results are not robust owing to short follow-up duration,
as well as heterogeneity of the ages and limited geographical locations of the participants.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 3 of 51
:
 Figure 2

Increasing prevalence of erectile dysfunction with age

Data from the European Male Ageing Study show that erectile dysfunction increases with age. Impor‐
tantly, the prevalence of severe erectile dysfunction (defined as an international index of erectile func‐
tion score of 1–7) increases at a steeper rate than that of moderate erectile dysfunction (score of 8–11)
in men over 60 years of age.

Epidemiological data have indicated a strong association between erectile dysfunction and
lower urinary tract symptoms (LUTS) in men with benign prostatic hyperplasia (BPH);
LUTS are associated with urinary obstruction caused by benign enlargement of the
prostate. This association is maintained even after adjusting for potential confounding fac‐
tors, such as age and comorbid conditions9,15,16. Both erectile dysfunction and LUTS in
those with BPH have a high prevalence in ageing men and have common risk factors, such
as hypertension and cardiovascular disorders, cigarette smoking, obesity, dyslipidaemia,
diabetes mellitus, metabolic syndrome, stress, anxiety and depression17.

Despite being studied thoroughly in men over 40 years of age18, the prevalence of erectile
dysfunction in younger men is rarely regarded as interesting19,20. In this context, a recent
naturalistic study (a study in which the participant is observed without any manipulation by
the researcher) reported that one man out of four seeking medical help for erectile dysfunc‐
tion in the real-life setting is <40 years of age21. Moreover, another study showed that
22.1% of men <40 years of age had low (<21) Sexual Health Inventory for Men (SHIM)
scores22. Studies evaluating erectile dysfunction epidemiology in a younger population will
help to clarify the prevalence and incidence in this age group. Although the data are mainly
focused on real-life experiences of clinicians, and not supported by population-based stud‐
ies, it is likely that most cases of erectile dysfunction in younger men have a psychogenic
basis. In this context, clues to suggest a psychogenic aetiology usually include sudden on‐

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 4 of 51
:
 set, good quality spontaneous or self-stimulated erections, major life events or previous
psychological problems. However, these factors should not disregard the importance of the
components of distress and discomfort that are brought by the onset and persistence of
erectile dysfunction and by problems related to sexuality. Such components include anxiety
and alexithymia (lack of emotional awareness) and even the possibility that, perhaps espe‐
cially in younger patients, erectile dysfunction can be both a symptom and a sentinel mark‐
er of serious organic problems20,23–26.

Mechanisms/pathophysiology

The penis remains in its flaccid state when the smooth muscle is contracted. The smooth
muscle contraction is regulated by a combination of adrenergic (noradrenaline) control, in‐
trinsic myogenic control and endothelium-derived contracting factors (prostaglandin and
endothelins; FIG. 3)27–29. Upon sexual stimulation, erection occurs after nitric oxide (NO)
is released from non-adrenergic noncholinergic (NANC) nerve fibres and acetylcholine is
released from parasympathetic cholinergic nerve fibres (FIG. 4); the result of the ensuing
signalling pathways is increased cyclic GMP (cGMP) concentrations, decreased intracellu‐
lar Ca2+ levels and smooth muscle cell relaxation29,30. As the smooth muscle relaxes, blood
is able to fill the lacunar spaces in the corpora cavernosa, leading to compression of the
subtunical venules, thereby blocking the venous outflow (veno-occlusion). The process is
reversed as cGMP is hydrolysed by phosphodiesterase type 5 (PDE5)29,30. Erectile dysfunc‐
tion can occur when any of these processes is interrupted.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 5 of 51
:
 Figure 3

Penile smooth muscle contraction — the flaccid state

a | Ca2+ influx into cells is regulated by noradrenaline signalling and levels of inositol-1,4,5-trisphos‐
phate (Ins(1,4,5)P3, which is produced from phosphatidylinositol-4,5-bisphosphate (PtdIns(4,5)P2) by
phospholipase C) in the cells; increased intracellular Ca2+ binds to calmodulin, facilitating the forma‐
tion of the calmodulin–myosin light chain kinase (MLCK) complex. This leads to the phosphorylation
of MLC, resulting in smooth muscle contraction and a flaccid penis. Noradrenaline signalling also in‐
hibits adenylyl cyclase and modulates the RHO-associated protein kinase (ROCK) pathway, which in‐
creases the sensitivity of MLC to Ca2+, a process negatively regulated by testosterone. Endothelins and
prostaglandins from the endothelium also trigger an increase in intracellular Ca2+ to promote smooth
muscle contraction. b | When the smooth muscle is contracted, inflow of blood through the cavernous
artery is minimal, and blood outflows freely through the subtunical venular plexus. ER, endoplasmic
reticulum; MLCP, myosin light chain phosphatase.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 6 of 51
:
 Figure 4

Penile smooth muscle relaxation — the erect state

a | Upon sexual stimulation, normal erection occurs after nitric oxide (NO) release from non-adrenergic
non-cholinergic (NANC) nerve fibres causes the activation of guanylyl cyclase, which raises the concen‐
tration of cyclic GMP, and after parasympathetic cholinergic nerve fibres release acetylcholine, which
activates adenylyl cyclase to increase the levels of cyclic AMP. Signalling pathways that are triggered
decrease intracellular Ca2+ levels and lead to smooth muscle cell relaxation. b | As the smooth muscle
relaxes, blood is able to fill the lacunar spaces in the cavernosa, leading to compression of the subtuni‐
cal venules, thereby blocking the venous outflow. The process is reversed as cGMP is hydrolysed by
phosphodiesterase type 5 (PDE5). ER, endoplasmic reticulum; InsP3, inositol trisphosphate; NOS, NO
synthase; PtdIns(4,5)P2, phosphatidylinositol 4,5 bisphosphate.

Psychogenic erectile dysfunction

Non-organic erectile dysfunction is also known as psychogenic or adrenaline-mediated

erectile dysfunction (noradrenaline-mediated or sympathetic-mediated erectile dysfunc‐
tion). It has not been well studied but is an important factor to consider when evaluating
and managing men with this condition. Stress, depression and anxiety are generally defined
as heightened anxiety related to the inability to achieve and maintain an erection before or
during sexual relations, and are commonly associated with psychogenic erectile dysfunc‐
tion (BOX 1). This association is unsurprising, given that noradrenaline is the primary
erectolytic (anti-erectile) neurotransmitter31 (FIG. 3).

Box 1

Symptoms of erectile dysfunction

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 7 of 51
:
 Psychogenic

Sudden onset
Intermittent function (variability, situational)
Loss of sustaining capability
Excellent nocturnal erection
Response to phosphodiesterase type 5 inhibitors is likely to be excellent

Organic

Gradual onset
Often progressive
Consistently poor response
Erection better in standing position than lying down (in the presence of venous
leak)

Nonendocrine causes

Neurogenic Neurogenic erectile dysfunction is caused by a deficit in nerve signalling to the

corpora cavernosa. Such deficits can be secondary to, for example, spinal cord injury, mul‐
tiple sclerosis, Parkinson disease, lumbar disc disease, traumatic brain injury, radical pelvic
surgery (radical prostatectomy, radical cystectomy, abdominoperineal resection) and dia‐
betes. Upper motor neuron lesions (above spinal nerve T10) do not result in local changes
in the penis but can inhibit the central nervous system (CNS)-mediated control of the erec‐
tion. By contrast, sacral lesions (S2–S4 are typically responsible for reflexogenic erections)
cause functional and structural alterations owing to the decreased innervation32. The func‐
tional change resulting from such injuries is the reduction in NO load that is available to the
smooth muscle. The structural changes centre on apoptosis of the smooth muscle and en‐
dothelial cells of the blood vessels, as well as upregulation of fibrogenetic cytokines that
lead to collagenization of the smooth muscle. These changes result in veno-occlusive dys‐
function (venous leak)33–37.

Vasculogenic Vascular disease and endothelial dysfunction lead to erectile dysfunction

through reduced blood inflow, arterial insufficiency or arterial stenosis. Vasculogenic erec‐
tile dysfunction is by far the most common aetiology of organic erectile dysfunction. In‐
deed, erectile dysfunction can be a manifestation of an underlying vascular disorder. The
risk of developing vasculogenic erectile dysfunction is increased in men with hypertension
(odds ratio (OR) of 3.04 for those on anti-hypertensive medication, and 1.35 for those not

38–41

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 8 of 51
:
 on medication), diabetes (OR 2.57) and dyslipidaemia (OR 1.83)38–41. Cigarette smoking
has also been shown to increase the risk of erectile dysfunction (OR 1.4)39,40,42, although
smoking cessation can reduce the risk32. Vasculogenic erectile dysfunction does not devel‐
op from high blood pressure itself but is secondary to the arterial wall changes (decreased
elasticity) in response to the increase in blood pressure. In addition, atherosclerosis related
to diabetes, dyslipidaemia and/or cigarette smoking can lead to arterial stenosis and com‐
pound the vascular injury.

Hypoxia from decreased corpora cavernosal oxygenation can cause a decrease in

prostaglandin E1 levels, which normally inhibit pro-fibrotic cytokines, such as transforming
growth factor β1 (TGFβ1)43. These pro-fibrotic cytokines promote collagen deposition, re‐
placing the smooth muscle and resulting in decreased elasticity of the penis, as has been
shown in several rat models44. As the smooth muscle to collagen ratio decreases and colla‐
gen content increases, the ability of the cavernosa to compress the subtunical veins decreas‐
es, leading to corporal veno-occlusive dysfunction45.

Iatrogenic The most common iatrogenic cause of erectile dysfunction is radical pelvic
surgery. Generally, the damage that occurs during these procedures is primarily neurogenic
in nature (cavernous nerve injury) but accessory pudendal artery injury can also
contribute46. Pelvic fractures can also cause erectile dysfunction in a similar manner, owing
to nerve distraction injury and arterial trauma.

Various medications have also been shown to be associated with the development of erec‐
tile dysfunction40 (BOX 2). Medications that are used to treat hypertension (thiazide diuret‐
ics and β-blockers) are most commonly associated with erectile dysfunction, but others,
including psychotherapeutics, anti-androgens, anti-ulcer drugs, opiates and digoxin, have
also been linked with the condition40. However, whether the erectile dysfunction results
directly from the medication itself or the underlying disease — for example, hypertension
— is difficult to define. The Treatment of Mild Hypertension Study (TOMHS) compared
five anti-hypertensive drugs with a placebo for changes in quality of life (sexual function
was ascertained by physician interviews)47. Chlorthalidone (a diuretic drug used to treat
hypertension) had the greatest effect on sexual function at 2 years after treatment, but the
placebo achieved almost the same level at 4 years. Accordingly, chlorthalidone may potenti‐
ate erectile dysfunction earlier in those who are likely to develop the condition later in life.

Box 2

Medications associated with erectile dysfunction

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 9 of 51
:
 Thiazide diuretics, β-blockers and spironolactone used to treat hypertension
Digoxin used to treat atrial fibrillation
5α-reductase inhibitors used to treat benign prostatic hyperplasia
Anti-androgens used to treat prostate cancer
Luteinizing hormone-releasing agonists and antagonists used to treat prostate
cancer
Tricyclic antidepressants, selective serotonin reuptake inhibitors,
benzodiazepines, antipsychotics and phenytoin used to treat depression and
other psychiatric conditions
H2 blockers used to treat ulcers
Opiates used to treat pain

Endocrine causes Androgens are considered the major hormonal regulator of penile devel‐
opment and physiology48,49; however, the role of testosterone replacement therapy in erec‐
tile dysfunction is controversial because of discrepancies in the findings from clinical trials,
and the fact that both hypogonadism and erectile dysfunction are common in ageing. The
increasing association of erectile dysfunction and the progressive decline of androgen levels
with ageing does not necessarily imply a causal link.

Most early studies aiming to understand the role of reduced testosterone on erectile func‐
tion focused on androgen ablation — a model that cannot be easily translated to erectile
dysfunction in humans50,51. From these studies and metabolic models, three sites of action
for androgens have been described: the nuclei in the CNS52, the spinal neurons and pelvic
ganglia, and the genital tissues53 (FIG. 5). Part of the erectile response to testosterone is
mediated through sexual desire (the male sex drive depends on testosterone), but mechanis‐
tic studies have documented a direct role of testosterone on cavernous smooth muscle cells,
involving NO, RHO-associated protein kinase (ROCK), PDE5 and the adrenergic response.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 10 of 51
:
 Figure 5

Levels of androgen action in the control of sexual response

Physiological effects of testosterone have been described in the regions of the brain that control sexual
arousal at the spinal cord level (affecting neuronal firing from the pelvic ganglia) and within the penis
(regulating endothelial and smooth muscle cell function). Testosterone has been shown to modulate the
release of nitric oxide from non-adrenergic non-cholinergic fibres, and the functioning of nitric oxide
synthase in endothelial cells. In the smooth muscle, testosterone modulates the activity of phosphodi‐
esterase type 5, the kinase that regulates Ca2+ and K+ levels, and adrenergic receptor sensitivity. Figure
modified with permission from REF. 53, Elsevier.

Effects on smooth muscle cells All animal studies support the idea that castration (reduc‐
tion in testosterone levels) causes a rapid drop in intracavernous pressure, owing to both
reduced arterial inflow and altered veno-occlusion during stimulated erections54 — castra‐
tion is associated with a rapid reduction in neuronal nitric oxide synthase (nNOS)55 and
pelvic ganglion activity56. However, in models of hypogonadism or in castration, the effects
of testosterone replacement on nNOS have been variable, with some studies revealing in‐
creased expression but unaltered activity, and other studies showing no effect55,57. Studies
carried out in animals treated with l-NG-nitroargininemethyl ester (l-NAME, a NOS in‐
hibitor) revealed that androgens trigger additional, NO-independent mechanisms that still
require intact cGMP generation to control veno-occlusion58. That is, androgens require
cGMP to produce an erection, which suggests that androgens modulate the erectile re‐
sponse through redundant mechanisms that involve cGMP generation.

Among these NO-independent targets is the ROCK pathway59, which contributes to tonic
smooth muscle cell contraction via calcium sensitization (FIG. 3a). Hypogonadism has
been shown to induce activation of ROCK1 (REF. 51), which counteracts smooth muscle

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 11 of 51
:
 cell relaxation. However, hypogonadism does not activate ROCK2, which is increased in
response to testosterone in endothelial cells60. Additional studies are necessary to under‐
stand the role of androgens in ROCK-dependent modulation of erection.

NO-independent, pro-erectile mechanisms of androgens also include regulation of expres‐

sion of smooth muscle myosin isoforms61 and sphingosine-1-phosphate (S1P)62,63 binding
to its natural receptors: a family of G protein-coupled receptors that are widely expressed in
the cardiovascular system. S1P receptor activation sustains constriction of smooth muscle
cells via phospholipase C (which cleaves phosphatidylinositol-4,5-bisphosphate (Ptd‐
Ins(4,5)P2) into inositol-1,4,5-trisphosphate (Ins(1,4,5)P3), leading to increased Ca2+) and
the ROCK pathways. In endothelial cells, SIP receptor activation triggers the phosphoinosi‐
tide 3-kinase (PI3K)–AKT pathway, enabling crosstalk between the ROCK and the en‐
dothelial nitric oxide synthase (eNOS) pathways61. The latter findings reinforce a beneficial
role of androgens on several overlapping NO-independent pathways (SIP, PI3K–AKT and
ROCK) favouring erectile response.

Several in vitro studies using animal and human tissues have shown that PDE5 expression
is upregulated by androgens57,64,65. However, recent studies have questioned this evidence,
suggesting that low PDE5 in hypogonadism simply reflects the overall reduction in smooth
muscle cell content66. Indeed, androgen deprivation triggers apoptosis of smooth muscle
cells, extracellular matrix deposition57 and accumulation of lipid droplets in mesenchymeal
cells (especially in the subtunical region) contributing to impaired veno-occlusion67. In
general, cGMP levels, regulated by the activity of PDE5 (the primary enzyme involved in
cGMP degradation), seem crucial for any direct57,64,65 or indirect68 androgenic regulation
within the penis58.

A recognized mechanism of action of testosterone in particular is the regulation of α1-

adrenergic responsiveness of smooth muscle cells57,69. Castration in animals has also been
shown to be associated with a decreased density of NANC innervating fibres and reduced
NANC-mediated relaxation in isolated corpora cavernosa strips57. These data suggest an
effect of testosterone on the postganglionic parasympathetic neurons, or even further up‐
stream within the autonomic nervous system53. In line with this hypothesis, the effects of
castration on penile haemodynamics, including NOS activity, can be transiently reversed in
vivo by short-term electrical stimulation of the cavernosal nerve55. Accordingly, androgens
might be necessary to support adequate neuronal stimulation to the corpora cavernosa,
maintaining tissue structural integrity; denervation, as can occur following prostate surgery,
and castration share some histological similarities53.

Hypogonadism and erectile dysfunction Further evidence for the role of androgens in erec‐
tile dysfunction comes from clinical studies. In the 1980s, Bancroft performed pivotal stud‐
ies to discriminate central effects from peripheral effects of testosterone replacement thera‐

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 12 of 51
:
 py. He showed that in acute settings, erectile capacity in response to visual stimulation is
less sensitive to androgen than sexual interest, fantasies and cognitive sexual activities70.
That is, androgen enhances the sexual response to sexual fantasy more than it enhances the
response to visual stimuli, which has implications for the kind of sexual activity measured
in the research setting. Experimental endogenous hypogonadism induced by gonadotropin-
releasing hormone (GnRH) agonists71 in supraphysiological-dose studies72 generated the
threshold hypothesis, confirmed by epidemiological data, that at least 8 nmol l−1 of testos‐
terone in sera is required for erectile function. However, some hypogonadal men retain
near-normal sexual activity despite very low testosterone levels73. In young adults, the an‐
drogen dependency of erectile function is maintained at threshold values that are far below
those required to maintain the function of other target organs (that is, <8 nmol l−1 or 230
ng dl−1). However, erectile function despite low androgen levels may not apply to elderly
men who have comorbidities, possibly owing to changes in androgen receptor expression
and activity. To match testosterone levels to an individual’s own requirement, the concept
of compensated or subclinical hypogonadism74 has been introduced (FIG. 6). In this set‐
ting, it is suggested that when testosterone declines from a previously higher level, a rise in
the levels of luteinizing hormone might be a biomarker for insufficient androgenization74,75.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 13 of 51
:
 Figure 6

Schematic representation of different types of hypogonadism

a| The normal pituitary–testicular axis. Gonadotropin-releasing hormone (GnRH) stimulates the release
of luteinizing hormone (LH). This triggers the testes to respond by producing adequate levels of testos‐
terone, which, in turn, exerts negative feedback control on the hypothalamus and pituitary gland. Both
circulating LH and testosterone are in the normal range. b | Secondary hypogonadism. The pituitary
release of LH is impaired, the testes are no longer stimulated and testosterone production drops; both
circulating LH and testosterone are reduced. c | Subclinical or compensated hypogonadism. The testicu‐
lar responsiveness to LH is impaired, testosterone production is maintained owing to overstimulation by
LH; circulating testosterone is normal or borderline, whereas LH is increased. In this case, the system is
driven to its maximal capacity and no further adjustment can be achieved. d | Primary hypogonadism. In
testicular failure, increases in LH can no longer sustain testosterone levels: circulating testosterone is
low and LH is high.

Other evidence for a role of testosterone in erectile dysfunction comes from clinical trial
data on testosterone replacement therapy. The few available randomized clinical trials ad‐
dressing the roles of treatment with testosterone in erectile dysfunction have been exten‐
sively reviewed, with the largest and most updated meta-analysis confirming significant
beneficial effects on various domains of erectile function, but only in men with testosterone
levels of less than 12 nmol l−1 (345 ng dl−1) at baseline76. Regression and subgroup analy‐
ses emphasized a role of ageing as a possible moderator of responsiveness to testosterone in
those with erectile dysfunction76. Another relevant emerging aspect is the time course of
testosterone effects (that is, the length of treatment necessary to achieve the maximum re‐
sult). A recent systematic review77 and randomized clinical trials78–80 revealed that al‐
though the effects on libido, ejaculation and sexual activity were apparent within just 2–3
weeks of commencing treatment, the effects on erectile function may take up to 6–12
months to be evident. Recently, the largest and longest trial addressing the effects of testos‐
terone replacement therapy on subclinical atherosclerosis progression in older men showed

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 14 of 51
:
 no significant difference in IIEF score compared with the placebo at 18 months or 36
months after the start of treatment. However, erectile dysfunction was not an inclusion cri‐
terion for the trial, and the relatively high baseline total IIEF score suggests that only some
of the participants had erectile dysfunction at enrolment81.

Several studies have recently reported growing concerns regarding the safety of sex steroid
replacement therapy (in men and women). These studies questioned the physiological roles
of the various hormones, and purposefully sought to amplify some sort of ‘hormonophobia’
by exaggerating or misrepresenting safety concerns82. However, leaving aside the contro‐
versies surrounding both the alarming83 and reassuring studies84 — contrasts that impose a
risk–benefit evaluation before any treatment — it seems clear that these considerations nei‐
ther apply to young adults with hypogonadism80 nor question the physiological role of
testosterone in erectile function.

Finally, little data have addressed the roles of other hormones in erectile dysfunction85. In‐
deed, possible roles have been documented for thyroid hormones, prolactin, growth hor‐
mone and insulin-like growth factor 1, dehydroepiandrosterone and oxytocin. Although
these hormones play a part in the pathophysiology of erection, their epidemiological impact
is likely to be small and is awaiting confirmation. After testosterone, prolactin is the most
commonly altered hormone in men with sexual dysfunction; its main effect is to inhibit go‐
nadotropin secretion to induce hypogonadism. Thus, prolactin should be considered for
screening, together with testosterone and luteinizing hormone (FIG. 6), in men with erectile
dysfunction.

Diagnosis, screening and prevention

Risk factors

The identification of the pathogenetic factors involved in erectile dysfunction is the corner‐
stone of an accurate diagnosis and successful treatment.

Lifestyle factors and diabetes Alcohol and smoking habits have consistently been shown to
affect erectile function. Evidence from observational studies suggests a positive dose–re‐
sponse association between quantity and duration of smoking and the risk of erectile dys‐
function86,87. Similar results have been documented for alcohol abuse88. In addition, diets
that are low in whole-grain foods, legumes, vegetables and fruits, and high in red meat,
full-fat dairy products, and sugary foods and beverages are all associated with an increased
risk of erectile dysfunction89. Finally, meta-analysis of available evidence demonstrates that
moderate and more frequent physical activity are associated with reduced risk of erectile
dysfunction90. Accordingly, both cross-sectional and prospective epidemiological studies

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 15 of 51
:
 suggest that obesity and metabolic syndrome are associated with an increased risk of erec‐
tile dysfunction91. It is conceivable that obesity-associated hypogonadism and increased
cardiovascular risk are correlated with the higher prevalence of erectile dysfunction in over‐
weight and obese men (see below). However, recent clinical and experimental studies sug‐
gest that the association between erectile dysfunction and central obesity is independent
from obesity-associated comorbidities and hypogonadism91. Although increased levels of
tumour necrosis factor (TNF; also known as TNFα) — a cytokine involved in systemic in‐
flammation — could be a mediator of these conditions92, further studies are needed to con‐
firm this. Sexual health is impaired by type 1 and type 2 diabetes and even by a pre-diabetic
status93–95. Peripheral neuropathy, atherosclerosis of large blood vessels, endothelial dys‐
function of arterioles and the associated hypogonadism all contribute to diabetes-related
sexual dysfunction93–95.

Cardiovascular disease Arteriogenic erectile dysfunction and cardiovascular disease are

considered different manifestations of a common underlying vascular disorder. Three inde‐
pendent meta-analyses have documented that erectile dysfunction needs to be considered
the harbinger of coronary heart disease96–98 and as a predictor of future silent cardiac
events99. This association is particularly important in younger men (<55 years of age) and
those with erectile dysfunction and no other comorbidites98, emphasizing the importance of
early diagnosis and correct management of erectile dysfunction-associated morbidities. Ac‐
cordingly, the Princeton III Consensus Recommendations for the management of erectile
dysfunction and cardiovascular disease indicate that incident erectile dysfunction has a sim‐
ilar, or even greater, predictive value for cardiovascular disease than traditional risk factors
such as diabetes, hypertension, dyslipidaemia or smoking100.

BPH and LUTS The presence of LUTS alongside BPH represents another important issue
in men with erectile dysfunction. The Multinational Survey of The Ageing Male (MSAM-
7) study — a multinational survey conducted in the United States and six European coun‐
tries — demonstrated that the presence of LUTS is an independent risk factor for erectile
dysfunction, although the pathological reason for this association is unclear9. Common al‐
terations in the NO–cGMP pathway, enhancement of RHOA–ROCK signalling and pelvic
atherosclerosis are often considered the most important mechanisms involved in determin‐
ing the two conditions.

Psychogenic and relationship factors Aside from organic factors, psychogenic and relation‐
ship domains need to be evaluated in men with symptoms of erectile dysfunction. All sexu‐
al dysfunctions, even the most documented organic types (such as diabetes-associated erec‐
tile dysfunction), are stressful and can lead to psychological disturbances101. Performance
anxiety is a common issue in men with sexual dysfunction, often leading to avoidance of

101

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 16 of 51
:
 sex, loss of self-esteem and depression101. Psychiatric symptoms are often comorbid in pa‐
tients with erectile dysfunction101. In addition, many psychotropic drugs can induce erectile
and other sexual problems101.

Although considered less often, the quality of a relationship represents an essential determi‐
nant of successful sexual activity. In fact, any sexual dysfunction in one member of the cou‐
ple will affect the couple as a whole, causing distress, partner issues and further exacerba‐
tion of the original sexual problem (see below)102. Interestingly, a patient’s perception of
reduced partner interest represents an independent predictor of incident cardiovascular
events103. Hence, the physical relationship between partners should be considered not only
as enjoyable, but also as a strategy for improving overall health and life expectancy.

Diagnostic work-up

The basic work-up of patients seeking medical care for erectile dysfunction needs to in‐
clude an evaluation of all the aforementioned factors, including: establishing an accurate
medical and sexual history; a careful general and focused genitourinary examination; and a
minimum number of hormonal and routine biochemical tests (FIG. 7). Other optional tests
can be considered in specific situations (see below).

Figure 7

Suggested diagnostic work-up for patients with erectile dysfunction

The basic work-up of patients seeking medical care for erectile dysfunction should include an accurate
medical and sexual history, a careful general and andrological physical examination, hormonal evalua‐
tion (total testosterone and sex hormone-binding globulin in all men, prolactin and thyroid hormone
evaluation in some men) and routine biochemical exams (total and high-density lipoprotein cholesterol,
triglycerides, fasting glucose and glycosylated haemoglobin). Second-line evaluation should be limited
to those with abnormal first-line results.

Given the personal, interpersonal, social and occupational implications of sexual dysfunc‐
tion, assessing sexual history is not an easy task. Finding the correct way to ask questions
and ‘decode’ answers on sexual health and illnesses is necessary to avoid embarrassing the
patient. Hence, expert-guided, validated and standardized sexual inventories — structured
interviews and self-reported questionnaires (SRQs) — can help both inexperienced and sea‐
soned clinicians to address sexual health and related conditions104. Both structured inter‐

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 17 of 51
:
 views and SRQs are composed of a set of standardized questions requiring a finite re‐
sponse. Several SRQs have been published, mainly focusing on evaluating erectile dysfunc‐
tion severity (for example, the IIEF) or erectile dysfunction treatment outcomes (for exam‐
ple, the Erectile Dysfunction Inventory for Treatment Satisfaction (EDITS))104. Structured
interviews are generally considered a more reliable instrument than SRQs in evaluating
sexual history and causes of erectile dysfunction, as they tend to achieve a closer patient–
physician relationship and reduce the risk of misunderstanding. So far, the only validated
structured interview on erectile dysfunction that has demonstrated sufficient utility in sever‐
al clinical studies is the Structured Interview on Erectile Dysfunction (SIEDY)104. SIEDY
is a 13-item interview composed of three scales that identify and quantify important do‐
mains in men with erectile dysfunction (organic, scale 1; marital, scale 2; and intrapsychic,
scale 3).

The physical examination of patients includes evaluation of the chest (including heart
rhythm, breathing and signs of gynaecomastia (enlargement of the breasts)), penis, prostate
and testes, and of the distribution of body hair105. Small testes and/or small prostate vol‐
ume, according to the patient’s age, might imply underlying hypogonadism. Similarly, other
possible signs of hypogonadism include gynaecomastia as well as a decrease in beard and
body hair growth. Assessment of the peripheral vascular system is also important to deter‐
mine the characteristics of the pulse, to ascertain the presence of an arterial bruit (a vascu‐
lar sound that is associated with turbulent blood flow). Increased pulse rate (tachycardia)
might suggest hyperthyroidism, whereas reduced pulse rate (bradycardia) might be evident
in men with heart block (arrhythmia), hypothyroidism or in those who use certain drugs
(for example, β-blockers). Diminished or absent pulses in the various arteries examined
could be indicative of impaired blood flow caused by atherosclerosis. The evaluation of the
penis in the flaccid condition might show the presence of Peyronie disease (involving pal‐
pable fibrous plaques), phimosis (congenital narrowing of the opening of the foreskin) or
frenulum breve (whereby the tissue under the glans penis that connects to the foreskin is
too short and restricts the movement of the foreskin), which can all contribute to erectile
dysfunction. Measurement of blood pressure, waist circumference and body mass index is
also performed105.

A few biochemical and hormonal parameters are of value in patients with erectile dysfunc‐
tion. However, levels of cholesterol, triglycerides, fasting glucose and glycosylated haemo‐
globin (HbA1c) are important determinants of cardiovascular and metabolic risk stratifica‐
tion105,106. Total testosterone and sex hormone-binding globulin for the evaluation of calcu‐
lated free testosterone105,106 are sufficient parameters to rule out hypogonadism. Prolactin
and thyroid hormone evaluation are limited to a subset of patients105,106.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 18 of 51
:
 The vast majority of men with erectile dysfunction are managed within the primary care
setting. However, in the presence of abnormal biochemical or hormonal values, further di‐
agnostic tests are advisable (second-line evaluation). If the fasting plasma glucose level is
100–126 mg dl−1, or HbA1c is >5.7%, an oral glucose tolerance test can be used to exclude
overt type 1 and type 2 diabetes mellitus. The necessity of performing further cardiovascu‐
lar evaluation should be based on the criteria of the Princeton III Consensus Panel100 (
TABLE 1). In the presence of reduced total testosterone and/or calculated free testosterone,
obtaining prolactin and gonadotropin levels will determine the source (central or peripher‐
al) of the problem.

Table 1

Princeton III Consensus recommendations

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 19 of 51
:
 Profile Description Sexual activity and PDE5
inhibitor use

Low Fewer than three risk factors for coronary artery Cleared to resume sexual
disease* (excluding sex) activity
Controlled hypertension Cleared to take PDE5
Class I or II stable angina‡ inhibitors
Successful coronary revascularization
History of uncomplicated myocardial infarction
Mild valvular disease, congestive heart failure
without left ventricular dysfunction and/or New
York Heart Association class I heart failure

Intermediate At least three risk factors for coronary artery Cardiac evaluation necessary
disease* (excluding sex) prior to resuming sexual
Class I or II stable angina‡ activity
Recent myocardial infarction (within 2–6 No contraindication to PDE5
weeks) inhibitor use
Left ventricular dysfunction and/or New York
Heart Association class II congestive heart
failure
Noncardiac sequela from atherosclerotic disease
(stroke and/or peripheral vascular disease)

High Unstable or refractory angina Sexual activity delayed until

Uncontrolled hypertension cardiac condition stabilized
New York Heart Association class III–IV
congestive heart failure
Recent myocardial infarction (within 2 weeks)
High-risk arrhythmias
Severe cardiomyopathy
Moderate to severe vascular disease
Princeton III Consensus recommendations for risk stratification and cardiovascular evaluation for sexual
activity100. PDE5, phosphodiesterase type 5.
*
Major cardiovascular risk factors include age, male gender, hypertension, type 1 and type 2 diabetes
mellitus, smoking, dyslipidaemia, a sedentary lifestyle and a family history of premature cardiovascular
disease.
‡
Defined by the Canadian Cardiovascular Society (see REF. 223).

Recent data have documented that penile duplex Doppler ultrasound (PDDU) can be per‐
formed in both flaccid (before vasodilator stimulation) and dynamic states (after vasodilator
stimulation) to further improve the stratification of cardiovascular risk in men with erectile
103,107,108

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 20 of 51
:
 dysfunction103,107,108. Nocturnal penile tumescence and rigidity testing using the RigiScan
device (GOTOP Medical, St Paul, Minnesota, USA) is currently rarely carried out109; its
use is limited to testing the presence of nocturnal spontaneous erectile activity for medico-
legal purposes when the presence of naturally occurring erections needs to be demonstrat‐
ed. Arteriography and dynamic infusion cavernosometry (measuring cavernosal blood pres‐
sure) and cavernosography (to assess venous leak) are carried out only in young men who
are potential candidates for vascular reconstructive surgery109.

Management

In the absence of a specific correctable aetiology, the management of erectile dysfunction is

largely empirical and performed in a step-wise manner. That is, initial treatment is based on
lifestyle modification followed by first-line therapies using PDE5 inhibitors and vacuum
erection devices (VEDs). Second-line therapies consist of an intraurethral suppository
(IUS) of prostaglandin E1 (alprostadil) and intracavernosal injection (ICI) with vasoactive
substances. Surgical intervention is reserved as the final option after conservative options
have been discussed or attempted.

Lifestyle modification

Lifestyle modifications can have a major role in managing erectile dysfunction, especially
in the younger patient. The physician can identify reversible risk factors that contribute to
the patient’s erectile dysfunction, such as medications, poor diet, low exercise, en‐
docrinopathies and anxiety. Although epidemiological evidence seems to support a role for
lifestyle factors in erectile dysfunction, limited data are available, suggesting that the treat‐
ment of underlying risk factors and coexisting illnesses will ultimately improve erectile
dysfunction110. The major limitation remains the paucity of interventional studies assessing
the effect of lifestyle changes on erectile function.

The available data support the recommendation that adults should do 30 minutes of moder‐
ate-intensity aerobic activity most days of the week110. Weight loss in obese men, and
switching from a Western diet to a Mediterranean diet, plus exercise, has been shown to
improve erectile dysfunction outcomes111–113. It has also been shown that, in a group of
patients with congestive heart failure, short-term moderate exercise training can improve
sexual function114. A review of the patient’s medications might reveal some drugs that are
well known to have erectile dysfunction as an adverse effect (BOX 2). However, much of
the data on iatrogenic erectile dysfunction are based on observational studies, with severe
limitations.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 21 of 51
:
 Smoking has been shown to have a direct relationship with erectile dysfunction, and a dose
response relationship has been suggested because men have increased erectile difficulties
with greater numbers of packets of cigarettes smoked or more years of smoking38,115,116.
One study suggested that cessation of cigarette smoking can improve erectile function117.
Mild alcohol consumption might improve erectile function by reducing anxiety; however,
chronic use or alcohol abuse can have lasting effects on the liver, leading to low levels of
testosterone and increased levels of oestrogen, both of which can contribute to erectile dys‐
function118,119. Patients with psychological stressors (such as performance anxiety, relation‐
ship issues and current life stress)120 may benefit from confidence restoration with erecto‐
genic medications and/or counselling with a psychologist or other health care professional
specializing in sexual dysfunction121. Accordingly, the European Association of Urology
states that “lifestyle changes and risk factor modification must precede or accompany any
[erectile dysfunction] treatment”, and classifies the level of evidence for lifestyle modifica‐
tion as 1b with a grade A recommendation122.

Nonsurgical interventions

PDE5 inhibitors Oral phosphodiesterase inhibitors (initially sildenafil, and later, vardenafil,
tadalafil, avanafil, and others available outside of the United States (TABLE 2)) have
changed the management of erectile dysfunction and created a sexual revolution123,124.

Table 2

Properties of available phosphodiesterase type 5 inhibitors

Drug Trade name Peak absorption post Serum half-life Take on empty
name (company) ingestion (hours) (hours) stomach?

Sildenafil* Viagra (Pfizer) 1–2 3–5 Yes

Vardenafil* Levitra 1–2 3–5 Yes

(GlaxoSmithKline)

Tadalafil Cialis (Lilly) 2–4 18 No

Avanafil Stendra (Mitsubishi 0.5 6 No

Tanabe)

*
Consider taking 1–2 hours prior to a meal.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 22 of 51
:
 Mechanistically, these drugs competitively inhibit PDE5, leading to a build-up of cGMP
upon NO release, initiating a cascade of events that lead to smooth muscle relaxation and
promotion of an erection29,125 (FIG. 4). Sildenafil and vardenafil reach their median Cmax
(maximum observed plasma concentration) values 1 hour after administration and have a
half-life of 3–5 hours126–128 (TABLE 2). The medications should be administered with ade‐
quate time before sexual intercourse to allow peak absorption of the drug. Patients must be
instructed on optimal conditions for the medications to work effectively. Sildenafil and var‐
denafil should be taken on an empty stomach because lipids in foods can decrease and delay
absorption126; tadalafil and avanafil are not as strongly affected by food127,128. Patients
should be reminded that PDE5 inhibitors still require sexual stimulation, both physical and
mental, to create arousal and initially raise the available levels of NO in an effort to gener‐
ate cGMP production129.

PDE5 inhibitors have been beneficial in correcting erectile dysfunction in a wide range of
patients with varying aetiologies of sexual dysfunction. Sildenafil has been shown to im‐
prove erections, leading to successful intercourse in 63% of men with general erectile dys‐
function compared with 29% of men using a placebo130. A study in 2001 showed that 59%
of patients with type 2 diabetes mellitus were able to have successful intercourse while tak‐
ing sildenafil compared with only 14% of those using a placebo131. In hypogonadal patients
who have not responded to treatment with PDE5 inhibitors alone, recent studies have sug‐
gested that combination of testosterone supplementation and a PDE5 inhibitor can improve
erectogenic outcomes132.

In men with prostate cancer who have undergone nerve-sparing radical prostatectomy, erec‐
tile function declines while the cavernous nerves recover from the surgical trauma. Al‐
though data regarding the efficacy of penile rehabilitation in radical prostatectomy patients
are mixed, the design of studies outside of the Pfizer-sponsored sildenafil study133 are
fraught with significant methodological limitations. Thus, so far no study has defined the
exact role of PDE5 inhibitors in penile rehabilitation in this patient population. Of note, one
randomized placebo-controlled trial in men who have undergone radiotherapy for prostate
cancer has demonstrated greater preservation of sexual function in those treated with PDE5
inhibitors versus a placebo134.

Patients should be counselled on the possible adverse events of these drugs, which may in‐
clude headache, heartburn, facial flushing, nasal congestion and visual disturbances (owing
to cross-reactivity with PDE6)124. Myalgia (muscle pain) is more common with tadalafil
than the other PDE5 inhibitors. The use of a PDE5 inhibitor with nitrate-containing med‐
ications (for example, those used to treat angina) can result in a dangerously low blood
pressure135. Also, PDE5 inhibitors and α-adrenergic receptor blockers, often used for treat‐
ment of BPH, need to be taken at least 4 hours apart. Priapism (prolonged erection of >4
hours) is a concern, but is a rare occurrence with PDE5 inhibitor therapy (approximately

136

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 23 of 51
:
 <0.1% of patients)136. Some vision-related conditions are cause for increased precautions,
including macular degeneration, retinitis pigmentosa and nonarteritic anterior ischaemic
optic neuropathy (NAION). Although no definitive data show causality in NAION, PDE5
inhibitors are contraindicated in patients with vision loss owing to NAION136 (the drugs
exert mild inhibitory action against PDE6, which is present exclusively in rod and cone re‐
ceptors). Concerns have also been raised regarding PDE5 inhibitor use and auditory
changes (hearing loss and tinnitus); however, limited data support this link134. Patients who
are concerned or who have hearing loss need to be warned about this risk, and may consid‐
er other options for erectile dysfunction treatment.

Vacuum erection devices A VED is a cylindrical mechanical device placed over the penis
and then pumped, which creates a negative pressure vacuum to draw blood into the
penis137. The blood fills the lacunar spaces within the corpora cavernosa, causing tumes‐
cence. Devices are often used in conjunction with a constriction band that is placed at the
base of the penis after tumescence is achieved, to prevent the backflow of blood. A study in
1991showed that 75% of diabetic men were able to have sexual intercourse when using a
VED to achieve tumescence138. However, discontinuation rates of up to 30% were also re‐
ported owing to bruising on the penis, pivoting at the base of the penis, coldness and/or
numbness of the penis, pain related to the constriction band and/or decreased ability to
achieve orgasm139. Patients benefit from a training session to help optimize their under‐
standing of proper use of VED. Obtaining a tight seal of the cylinder against the body using
lubricant and/or trimming the pubic hair is important for success with the VED. Men with
large pubic fat pads and/or a buried penis may have difficulty placing the device because
they have a less usable penile shaft140.

Several adverse reactions have been noted with VEDs use that should be pointed out to pa‐
tients. These adverse effects include petechiae (capillary bleeding) and haematoma (a
swelling of clotted blood)141. Patients need to be warned that constriction band use to main‐
tain tumescence can give a grey-blue colour to the penis, and that the penis can become
cool to the touch, owing to obstructed venous outflow. Few data exist on the use of VEDs
for rehabilitation after prostatectomy; however, one randomized prospective study showed
that daily VED use can preserve penile length after radical prostatectomy142.

Intraurethral suppository The use of IUS involves the placement of a prostaglandin E1-
loaded pellet within the urethra before sexual intercourse. After insertion of the pellet, the
patient should massage that area of the penis to help disperse the medication. The drug is
absorbed through the urethra into the corpora cavernosa and increases the intracellular lev‐
els of cyclic AMP (cAMP), leading to decreased intracellular Ca2+ levels, increased smooth
muscle relaxation and tumescence143.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 24 of 51
:
 As a second-line treatment to PDE5 inhibitors, one study showed IUS efficacy in 56% of
patients; those with organic erectile dysfunction were more likely to respond144. Another
study showed that, across all patients, 65% of men were able to have sexual intercourse,
including patients with diabetes144, but a second study found that only 46% of patients with
diabetes could have intercourse145.

After radical prostatectomy, one study showed that IUS use for penile rehabilitation was not
significantly different to sildenafil. However, the IUS group had a significant dropout rate
and lower compliance rate compared with the sildenafil group146.

Patients need to be trained on the technique of the IUS before use and should be advised
that pain or burning may occur with use of this medication. The painful sensation can be
intolerable for some patients: one study showed that 39% of patients discontinued use ow‐
ing to pain147.

Priapism is a risk in patients using IUS, but the actual rate of occurrence is low. Other ad‐
verse effects include penile pain, urethral pain and dizziness148. Patients need to be cau‐
tioned before using IUS if they have any diseases of the urethra, such as urethral stricture.
Furthermore, there can also be transference of prostaglandin E1 to the partner, and condoms
should be used if the partner is pregnant to avoid premature labour.

Intracavernosal injection ICI involves the use of vasoactive substances injected directly into
the corpora cavernosa via a small needle. These vasoactive agents include prostaglandin E1,
papaverine and phentolamine (and sometimes atropine), which work alone or in combina‐
tion to elicit an erection. Prostaglandin E1 has been approved by the FDA as a single-agent
ICI for erectile dysfunction and increases cAMP levels. Papaverine is a nonspecific phos‐
phodiesterase inhibitor that leads to increased levels of cAMP and cGMP. Phentolamine is
an α1-adrenergic receptor inhibitor that helps to prevent vasoconstriction to maintain
tumescence149.

These vasoactive drugs are available in varying strengths and combinations made by phar‐
macies: monotherapy with prostaglandin E1; bi-mixture of papaverine and phentolamine;
and tri-mixture of prostaglandin E1, papaverine and phentolamine. Patients require training
on how to prepare the medications with syringes for home use; an in-office test dose can be
useful to characterize the patient’s response. The dose can then be titrated up or down under
supervision to one that is satisfactory to the patient. Many patients have an understandable
fear of injecting the penis, and overcoming this is the first step to successful treatment150.

Initial satisfaction rates are high, with up to 94% of patients achieving a satisfactory erec‐
tion with in-office titration151. However, dropout rates with ICI are also high, with 46–80%
abandoning treatment in the first year152,153. Reasons for dropouts included lack of partner,

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 25 of 51
:
 high cost, problems with the concept of penile injection and a desire for a permanent solu‐
tion154.

A major concern with ICI use is priapism. If priapism occurs, the patient will need to seek
urgent medical attention, which might require aspiration of blood, surgical shunt formation
or ICI of phenylephrine to cause cavernosal vasoconstriction155. Pain has also been fre‐
quently described in up to 50% of patients and is usually attributed to prostaglandin E1, but
often attenuates over time. Bruising can also be observed, as might be expected with any
injected medication.

Injections are useful for patients after radical prostatectomy, if they fail to respond to PDE5
inhibitor therapy. Penile injections have been used frequently in the penile rehabilitation
setting based on survey data156, but its acceptance is not widespread.

Surgical interventions

Penile implants Although oral and vacuum erection therapies are effective first-line and
second-line management options for most men with erectile dysfunction, surgical interven‐
tions remain an important modality in certain situations. Indeed, surgery might be suited for
patients who have contraindications, experience adverse effects or are refractory to medical
therapy; patients with erectile dysfunction and penile fibrosis secondary to Peyronie dis‐
ease, prolonged priapism or severe infections; and patients with structural and/or vascular
penile defects from genital or pelvic trauma. The current surgical armamentarium includes
insertion of a penile prosthesis and vascular reconstructive surgery.

Penile implants include malleable and inflatable devices157 (FIG. 8). Malleable prostheses
consist of semi rigid cylinders that can be bent upwards for sexual intercourse and down‐
wards to conceal when not in use. These devices are easy to handle, have low mechanical
failure rates and are optimal for patients who have diminished manual dexterity. Unfortu‐
nately, owing to their constant rigidity, they can be uncomfortable, can cause social embar‐
rassment and are at a higher risk of erosion158. Two-piece inflatable penile prostheses
(IPPs) consist of two cylinders with a scrotal pump, which enables transfer of fluid to the
cylinder chambers when an erection is desired. Although it provides similar rigidity to a
malleable device, it also enables better flaccidity when deflated159. Its main indication is in
patients in whom placement of an abdominal fluid reservoir is contraindicated or not feasi‐
ble. For all remaining patients, implantation of a three-piece IPP is considered the gold
standard in North America. This system consists of paired corporal cylinders, a scrotal
pump and an abdominal reservoir filled with saline. This device provides the greatest de‐
gree of girth expansion and penile rigidity (owing to its larger fluid reservoir) when an erec‐
tion is desired, and the most flaccidity when deflated157. Prosthesis implantation can be per‐
formed using infrapubic, penoscrotal or subcoronal approaches.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 26 of 51
:
 Figure 8

Penile prostheses

a | The malleable penile prosthesis involves two semi-rigid rods that are placed in the corpora caver‐
nosa. The implant does change in size; it is bent upwards before intercourse. b | The two-piece inflatable
penile prosthesis (IPP) involves placement of two inflatable cylinders in the corpora cavernosa and a
pump in the scrotum. As the pump is used, fluid is transferred between the pump and the cylinders. c |
The three-piece IPP involves the placement of two inflatable cylinders (in the corpora cavernosa), a
pump in the scrotum and a fluid reservoir in the lower abdomen alongside the bladder. Pressure applied
to the pump causes a transfer of fluid from the reservoir to the cylinders, leading to penile rigidity. The
pump has a release valve or button to transfer the fluid back from the cylinders to the reservoir at the
end of intercourse.

One subset of patients who benefit from implantation of an IPP is men with Peyronie dis‐
ease and concomitant erectile dysfunction. Although penile plication or incision and graft‐
ing procedures are options to correct the penile curvature in men with good erections, inser‐
tion of a prosthesis is the recommended approach for those with poor erections, often due
to vascular insufficiency160. For these men, owing to higher patient and partner satisfaction
rates, a three-piece IPP is preferred over a malleable device161. IPP placement alone can
often correct the penile deformity, but it is sometimes necessary to use adjunctive straight‐
ening manoeuvres — such as penile modelling (manual straightening), plication, incision
of plaque and grafting — to optimize results162.

Corporal fibrosis is another circumstance in which penile prosthesis implantation is indicat‐

ed. Such fibrosis usually occurs secondary to priapism, previous prosthesis infection or
trauma. Surgical techniques include excision or incision of the scar, corporal excavation,
corporotomies with or without grafting, use of cavernotomes (to scrape the fibrotic areas),
implant downsizing and transcorporal resection163.

Contemporary data using newer devices (for example, devices made with sturdier materials
or antibiotic-coated devices) demonstrate improved results with rates of overall freedom
from reoperation and mechanical failure at 10 years of 74.9% and 81.3%, respectively164.
Reported complications following IPP implantation include infection, distal cylinder ero‐
sion, auto-inflation, pump migration and reservoir displacement165. Overall, IPP placement
is associated with excellent patient (92–100%) and partner (91–95%) satisfaction rates, sig‐
nificantly higher than satisfaction rates reported for oral, intraurethral and intracavernosal

166,167

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 27 of 51
:
 medical therapies166,167. However, because of its associated high cost, potential complica‐
tions and invasiveness of the procedure, IPP implantation should only be offered to patients
who fail more conservative measures (such as those described above).

Penile revascularization Building on the same principles as bypass surgery for coronary
artery disease, penile revascularization surgery techniques were developed to anastomose
the inferior epigastric artery to either the dorsal artery or deep dorsal vein (arterialization),
with or without venous ligation165, to improve penile vascular inflow while reducing ve‐
nous outflow. However, multicentre studies have revealed that arteriogenic erectile dysfunc‐
tion is not common, and is more often associated with systemic multifactorial disease, pos‐
sibly caused by a combination of endothelial and corporal smooth muscle dysfunction,
leading to a drop in penile revascularization procedures performed clinically165. A review
of the published literature shows that the overall sexual satisfaction rate following penile
revascularization (12%) is lower than that following placement of an IPP (93%)165. Current‐
ly, penile revascularization is recommended for younger men (<55 years) who are not dia‐
betic, non-smokers and have a documented isolated stenotic segment of the internal puden‐
dal artery without concomitant venous leak. The success rates are reported to be as high as
80% in some populations168,169. Potential complications of penile revascularization include
glans hyperaemia, shunt thrombosis and inguinal hernias170.

Currently, no standardized method is available to accurately and systematically diagnose

corporal veno-occlusive dysfunction independently of general arterial insufficiency171.
Most medical experts and organizations think that venous leak is a consequence rather than
an isolated aetiology of erectile dysfunction. As such, both the American Urological Asso‐
ciation and the International Society for Sexual Medicine concur that “surgeries performed
with the intent to limit the venous outflow of the penis are not recommended” (REFS
171,172).

Quality of life

Impact on the patient

Erectile dysfunction can have a substantial negative effect on a man’s quality of life (BOX
3). Many men with erectile dysfunction experience depressive symptoms and anxiety relat‐
ed to sexual performance and avoid engaging in sexual relations. Many men also avoid
seeking treatment for their sexual dysfunction173–176. The association between the condition
and depression is considered to be bidirectional, with the two conditions reinforcing each
other in a downward spiral173. Importantly, men with depression and erectile dysfunction
have lower libidos than men with erectile dysfunction alone, and are less likely to discuss
their erectile dysfunction with their partners175,177.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 28 of 51
:
 Box 3

Impact of erectile dysfunction on the patient, partner and couple

Intimate relationships and health have a dynamic nature over time

Involvement of both partners’ health in joint overall ‘health’ means that the
illnesses of one partner affect the health of the other
Rates of depression in men with erectile dysfunction are reported to be as high
as 56%
Anxiety in a sexual situation is prevalent and can trigger a loss in sexual
confidence
As a result of this psychological burden, many men with erectile dysfunction
avoid sexual situations and use erectogenic medications
Erectile dysfunction negatively affects a man’s quality of life and satisfaction in
his relationship
Erectile dysfunction is associated with female partner sexual dysfunction
Female sexual function improves with treatment of erectile dysfunction in her
partner with phosophodiesterase type 5 inhibitors and nerve-sparing surgery (in
men with prostate cancer)

The proportion of men with erectile dysfunction has been reported to be as high as 50%,
and data from many well designed population-based studies have demonstrated a significant
relationship between erectile dysfunction and depressive symptoms (OR 1.7, P <
0.05)174,175. These studies were conducted in men aged 40–70 years in the United States,
Finland, Brazil, Japan and Malaysia.

Anxiety is also a concern, and it can lead to the inability to maintain an erection (psy‐
chogenic erectile dysfunction)178. This anxiety increases a man’s focus on the firmness of
his erection, leading to self-consciousness and cognitive distractions that interfere with
arousal and contribute to poor performance179,180. Men who cannot achieve or maintain an
erection as a result of anxiety lose sexual confidence and generally develop greater fears
leading up to the next sexual encounter, which increases the likelihood of failure and rein‐
forces the pressure to perform during future encounters178.

Given the psychological consequences of erectile dysfunction, it is not surprising that many
men have difficulty accepting their condition and avoid treatment. Research indicates that
as many as 69% of men deny the existence of their erectile dysfunction, and the median
time to pursue treatment for erectile dysfunction is >2 years181. Men report that they expe‐

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 29 of 51
:
 rience significant disappointment and shame related to erectile dysfunction; and even when
using erectile dysfunction treatments, men report a fear of failure and a tendency to avoid
sexual situations as a result of this fear182. In fact, data indicate that 50–80% of men discon‐
tinue use of medical interventions (oral medications, ICIs or VEDs) for erectile dysfunction
within 1 year183,184, and dropout rates related to PDE5 inhibitor use are estimated to be 30–
50%185–187. Although men discontinue erectile dysfunction treatment for a number of rea‐
sons, avoidance of sexual activity might be a primary cause182. This avoidance leads to
continued distress, anxiety and potential relationship difficulties.

Considering the potential psychological effect of erectile dysfunction, appropriate psycho‐

logical evaluation, in addition to appropriate medical treatment, is important. This biopsy‐
chosocial perspective implies that physical treatments, as well as psychological interven‐
tions, are needed to ensure that men pursue and continue to use appropriate erectile dys‐
function treatments.

Impact on the partner

Health and sexuality have a bidirectional relationship, and both partners of a couple con‐
tribute assets and/or liabilities to the health endowment. In this context, health is defined as
jointly produced by a man and his intimate partner188. Joint production of health means that
the illnesses of one partner affect the health of the other; in addition, there is a dynamic as‐
pect in both intimate relationships and health over time. For example, if a man has erectile
dysfunction, his partner might report decreased libido because he or she is anticipating a
negative sexual experience rather than a positive one. Similarly, an endometrial cancer sur‐
vivor with stenosis and dyspareunia can have a partner who experiences sexual aversion
because he or she fears intimacy will cause her pain189. Likewise, if a man has premature
ejaculation, his partner may be unable to experience orgasm, which causes distress about
sexual satisfaction190. However, this distress may be reversed when the partner’s sexual
dysfunction is treated.

The Female Experience of Men’s Attitudes to Life Events and Sexuality (FEMALES) study
characterized the sexual experience of female partners of men with erectile dysfunction
through surveys performed before and after the development of their partner’s erectile dys‐
function and in relation to his use of PDE5 inhibitors191. The study showed that after men
develop erectile dysfunction, their partners engage in significantly less sexual activity than
before (P < 0.001). Women with partners taking PDE5 inhibitors reported that they experi‐
enced more frequent sexual desire, arousal and orgasm (P < 0.05) than women with part‐
ners not receiving treatment. A study carried out in Taiwan evaluated the association of
erectile dysfunction with female sexual function and corroborated the results from the FE‐
MALES study192. It demonstrated that female partners of men with erectile dysfunction
had significantly (P < 0.001) lower overall and individual domain scores on the Female

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 30 of 51
:
 Sexual Function Index (FSFI) than women whose partners did not have erectile dysfunc‐
tion. After adjusting for other risk factors, the women’s sexual function was worse in terms
of arousal, orgasm, sexual satisfaction and sexual pain.

Impact on the couple

Sexual dysfunction is not only an issue for the individual but also for the couple. In one tri‐
al, on-demand oral vardenafil improved erectile function as measured by males’ responses
to the Sexual Encounter Profile question 3 (P < 0.0001), and the sexual quality of life of the
couple as measured by the female partners’ responses to the quality of life domain of the
modified Sexual Life Quality Questionnaire at last observation carried forward (P <
0.0001)193. Another vardenafil treatment study showed similar efficacy for erectile dysfunc‐
tion and also improved the sexual satisfaction of the female partners who had not them‐
selves received specific interventions for sexual function194. The female partners’ total FSFI
score and scores for sexual desire, subjective arousal, lubrication, orgasm and satisfaction
increased. This study also showed that the female partner’s sexual function improvements
significantly related to treatment responses in her partner.

A study assessing the impact of prostate cancer treatment on couples’ sexual function dur‐
ing the initiation phase of treatment found that the couples’ ‘complicity’ remained intact
despite decline in the male’s sexual function195. Global marital adjustment was measured
by the marital adjustment test (MAT), a self-reported 15-item questionnaire that successful‐
ly differentiates between distressed and non-distressed couples; couples’ MAT scores were
unchanged before and after robot-assisted laparoscopic prostatectomy. However, postopera‐
tive decreases in IIEF scores and the female partners’ FSFI scores were significantly associ‐
ated within couples. Interestingly, when strategies were used to prevent or limit erectile
dysfunction, sexual function of women improved; bilateral nerve-sparing surgery preserved
not only male but also female sexual function195.

These various interventions demonstrate the interrelationship of partners’ sexuality and the
concept of a sexual unit in long-term relationships. Thus, couple sexuality is a dynamic
process, which adapts and responds to the sexual states of each partner and can be en‐
hanced by interventions directed at improving one partner’s sexual functioning.

Outlook

Although the clinical management of erectile dysfunction has been rather static since the
advent of PDE5 inhibitors, basic and translational research has continued to progress and
deliver new therapeutic targets that are in various stages of development. A crude division

196

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 31 of 51
:
 can be made between pharmacotherapeutic compounds196 on the one hand, and regenera‐
tive medicine197 on the other (FIG. 9).

Figure 9

Potential future treatment options for erectile dysfunction

Current treatments provide temporary symptomatic relief, but do not interfere with the progress of the
disease itself. The first-line treatment of phosphodiesterase type 5 (PDE5) inhibition depends on
bioavailable nitric oxide (NO) to exert any effect. Accordingly, future pharmacological therapies will
need to be efficacious in patients who do not respond to PDE5 inhibition (thereby providing treatment to
men with neurogenic erectile dysfunction) and provide temporary relief or ‘erections on demand’. The
site of action may be the central nervous system (CNS), or a variety of peripheral pathways that control
the balance between vasoconstriction and vasorelaxation. By contrast, the aim of regenerative medicine
is definitive symptomatic relief (cure) by reversing or halting the progression of degeneration in erectile
dysfunction. Regenerative medicine intends to change the course of the disease and in many instances
will regenerate failing cells, tissues or whole organ systems. Depending on the severity of tissue damage
or the severity of the clinical presentation, various tools, such as growth factors, gene transfer, (stem)
cells and tissue engineering could be used to achieve this goal. eNOS, endothelial nitric oxide synthase;
Maxi-K channel, calcium-activated potassium channels; ROCK, RHO-associated protein kinase.

Pharmacotherapeutics

A considerable proportion of men do not respond to oral PDE5 inhibitors and have to resort
to ICI therapy or penile prosthesis implantation to be able to have intercourse198. Refrac‐
toriness to PDE5 inhibitors stems from the need for endogenous NO release from the
NANC fibres and for the endothelium to be able to accumulate cGMP to cause smooth
muscle relaxation during sexual stimulation196. Thus, there is a clear medical need for com‐
pounds (preferentially orally administered ones) that cause smooth muscle relaxation in the
absence of NO. Although both centrally and peripherally acting compounds are being pre‐
clinically investigated, ROCK inhibitors and soluble guanylyl cyclase activators are of par‐
ticular interest because they can be beneficial to patients who do not respond to PDE5 in‐
hibitors. Detailed reviews of the pharmacotherapeutic pipeline in erectile dysfunction are
available elsewhere196–199.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 32 of 51
:
 ROCK inhibitors The ROCK pathway plays an important part in maintaining the flaccid
state of the penis (FIG. 3a). ROCK phosphorylates and inactivates myosin light chain phos‐
phatase, allowing the myosin light chain to stay phosphorylated and, therefore, bound to
smooth muscle actin200. That is, inhibition of ROCK provides a mechanism of smooth mus‐
cle relaxation that is independent of NO. In various animal models of erectile dysfunction
(diabetes, ageing and cavernous nerve injury), ROCK was found to be upregulated, which
makes it an attractive target for the future treatment of erectile dysfunction201–203. In animal
studies, acute administration of the ROCK inhibitor Y-27632 was shown to induce relax‐
ation of rat corpus cavernosum strips in vitro and to cause an increase in intracavernous
pressure in rat models of erectile dysfunction. The inhibition of ROCK elicited an erectile
response by a process that was not mediated by NO or cGMP, and resulted in better erectile
responses than in sham-treated animals201–203. Recently, it was shown that ROCK2 is up‐
regulated in the corpora cavernosa of men with erectile dysfunction, corroborating earlier
results in animal studies204. This indicates that ROCK2 may be an important target, espe‐
cially as targeting an upregulated isoform of ROCK specifically addresses the altered physi‐
ology in the diseased smooth muscle. Furthermore, animal studies have shown that chronic
ROCK inhibition can lead to sustained improvement of erectile function201–203.

Although Y-27632 is still in a preclinical phase, SAR-407899 (another ROCK inhibitor un‐
der development) has advanced to clinical trial. However, although a randomized double-
blind Phase II trial in 20 patients with erectile dysfunction (NCT00914277)196 was con‐
ducted, the results, for unknown reasons, were never made public. Other ROCK inhibitors,
such as cethrin, are currently under clinical investigation for related conditions such as
spinal cord injury, and successful outcomes of these studies might bring promise for erec‐
tile dysfunction following cavernous nerve injury.

Soluble guanylyl cyclase activators Another class of molecules under consideration is the
activators of soluble guanylyl cyclase. In the healthy state, this enzyme is activated by NO
to generate cGMP from GTP196. However, when endogenous NO production is limited,
direct activation of soluble guanylyl cyclase can result in smooth muscle relaxation. Activa‐
tors of soluble guanylyl cyclase have been shown to induce smooth muscle relaxation in
vitro and penile erection in vivo in animal models. One activator, BAY60-4552, was inves‐
tigated in a rat model of erectile dysfunction following cavernous nerve injury in combina‐
tion with vardenafil; the combination therapy resulted in improved erectile function com‐
pared with non-injured sham-treated animals205. A subsequent clinical study showed that
BAY60-4552 and vardenafil had synergistic effects on smooth muscle relaxation in corpus
cavernosum strips from patients who did not respond to PDE5 inhibitors206.

The safety of BAY60-4552 has been tested in a Phase I trial (NCT01110590). When given
in combination with vardenafil, no major adverse effects were reported in men with erectile
dysfunction according to company data207. However, a Phase II study (NCT01168817)208

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 33 of 51
:
 study failed to demonstrate superior efficacy of the combination treatment of BAY60-4552
plus vardenafil versus vardenafil alone, although both active treatments showed superior
efficacy versus a placebo. The drug is apparently not being investigated further for erectile
dysfunction because it was not shown to be superior to vardenafil monotherapy. Aspects of
the study design merit further examination of this compound as a treatment for erectile dys‐
function by either pharmaceutical companies or academia. Further research is also warrant‐
ed in the specific population of patients who do not respond to PDE5 inhibitors.

Regenerative medicine

Currently, the treatment of erectile dysfunction focuses on symptomatic relief and, there‐
fore, tends to provide a temporary reprieve rather than providing a cure or reversing the
cause. As stated above, the existence of a large population of difficult-to-treat patients has
motivated researchers to identify new treatment approaches that focus on cure and restora‐
tion of the underlying cause, rather than providing ad hoc symptomatic treatment. Regener‐
ative medicine has developed extensively over the past few decades, and preclinical trials
have emphasized the benefit of growth factor therapy, gene transfer, stem cells and tissue
engineering for the estoration of erectile function197.

One approach investigating the gene transfer-assisted upregulation of the endogenously

present ‘Maxi-K’ potassium channel into penile smooth muscle cells to increase their relax‐
ant capacities has been examined in a Phase I clinical trial. The treatment seems safe and
preliminary indicators of effectiveness were present in a small number of treated
patients209. Furthermore, a plethora of studies have investigated the preclinical administra‐
tion of stem cells in animal models of erectile dysfunction related to ageing, type 1 and type
2 diabetes, cavernous nerve injury, Peyronie disease and even penile trauma210. Researchers
have generally reached a consensus that the results of these studies, mainly using mes‐
enchymal stem cells from the bone marrow and adipose tissue, show this to be a promising
approach211.

Recently, human umbilical cord blood stem cells have shown beneficial effects on erectile
function when administered into the penises of men with severe type 2 diabetes212. Howev‐
er, this effect was short-lived and not durable. Another Phase I trial investigated intracav‐
ernous administration of bone marrow cells after radical prostatectomy in an open dose-
escalation study and reported no serious adverse effects213. Conclusions on efficacy in these
trials should be drawn with necessary caution, as these trials were designed to study safety
(no control group); however, preliminary efficacy data are promising, with improved erec‐
tile function and penile vascularization measurements in a small set of patients. Although
these initial safety data are encouraging, larger Phase I–III studies and functional results are
eagerly awaited. To date, at least seven stem cells trials have been registered214–220.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5027992/ 16/10/24, 2 44 PM
Page 34 of 51
: