JSSM Guidelines for erectile dysfunction


Yasusuke Kimoto md, Department of Urology, Spinal Injuries Center, 550-4 Igisu, Iizuka, Fukuoka 820-8508, Japan. Email: ykimoto@jcom.home.ne.jp


How these Guidelines came to be written and financed

The preparation of these Guidelines was decided upon by the Board of Directors of the Japanese Society for Sexual Medicine (JSSM) in 2006. It was then approved by the Society members, and developed with the support of the Japanese Urological Association. The need for these Guidelines arose in 1999, when oral medication for the treatment of erectile dysfunction (ED) was approved for sale in Japan. Until that time, only a few specialists were involved in the treatment of ED. With the appearance of an effective oral agent, general physicians had more opportunities to treat ED. This led to our decision to produce these Guidelines for the diagnosis and treatment of ED, suitable for general physicians. The funding came entirely from the Japanese Society for Sexual Medicine, with the approval of the Board of Directors.

Target readership

As mentioned above, these Guidelines are aimed at the general physician. In this case, the term general physician includes most urologists who do not specialize in the area of sexual dysfunctions.


We initially based our deliberations on the record of the International Consultation on Erectile and Sexual Dysfunctions,1 and the 2005 version of the European Association of Urology (EAU) ED Guidelines,2 the most recent guidelines available at present. We searched specialist ED journals, The Journal of Sexual Medicine, International Journal of Impotence Research, and journals published by the Japanese Society for Sexual Medicine, for papers published after the above guidelines were issued. We also searched the specialist urology journals, Journal of Urology, Urology, BJU International, and European Urology.

We confined our PubMed search to ‘review OR meta-analysis’ from 2005 or later, using the key words ‘erectile dysfunction OR sexual dysfunction’. Based on the resulting material, preparation committee members Drs Kimoto, Nagao, Sasaki and Marumo (all urologists) gathered and conferred together at branch meetings of the Japanese Society for Sexual Medicine and the general meeting of the Japanese Urological Association, and produced some draft guidelines. These were sent to committee members Drs Ishikura, Takahashi, Nishi and Ms Futamatsu for their opinions in their specialist fields. The draft guidelines, revised according to their opinions, were then presented to the General Council of the Japanese Society for Sexual Medicine. Further changes were made following their deliberations, and the Guidelines were ratified. Levels of evidence for the references quoted in the Guidelines, and the recommendation grades, were based on the ‘Minds 2007 Manual for Production of Clinical Practice guidelines’3 (see Table 1).

Table 1.  Level of evidence and Recommendation grade by Minds3
Levels of evidence
ISystematic reviews or meta-analyzes of randomized controlled trials (RCTs)
IIOne or more RCTs
IVaAnalytic epidemiological studies (cohort studies)
IVbAnalytic epidemiological studies (case-control studies, horizontal studies)
VDescriptive studies (case reports and case series)
VIExpert opinion (group or individual), not based on patient data
Recommendation grade
AGood supporting evidence, strongly recommend this action
BFair supporting evidence, recommend this action
C1A lack of supporting evidence, but recommend this action
C2A lack of supporting evidence, recommend against this action
DEvidence for lack of efficacy or harm, recommend against this action

Plans for revisions

The Third International Consultation on Erectile and Sexual Dysfunctions is scheduled for 2009, and the proceedings should be release in 2010. We plan to prepare a Japanese version based on their deliberations, work progressing through 2010 and 2011, and the revised Japanese Guidelines should be ready for publishing in 2011. The production committee for the revised edition will be selected mainly from participants in the 2009 International Consultation.

Conflicts of interest

All of the members of the Guidelines preparation committee have over the past 3 years received payments in their respective specialty fields, in the form of lecture fees and teaching fees, from Pfizer, Bayer, and Eli-Lilly Japan.


The authors would like to express their gratitude to the representatives of the Japanese Urological Association, and in particular Dr Akihiko Okuyama, the Association President, for their invaluable assistance in the preparation of these Guidelines.

Epidemiology of ED

A recent review article4 estimated that at present 5–20% of all adult males worldwide experience moderate to complete ED.NB This proportion is expected to further increase with the aging population, doubling by the year 2025.5 An epidemiological survey was conducted in Japan with a questionnaire using the same format as the Massachusetts Male Aging Study (MMAS).6 Subject selection was based on the basic resident registry, and the survey was conducted using a postal and visit format. This allowed an accurate statistical survey and an accurate estimation of the prevalence of this condition on a national scale. A total of 2000 subjects randomly selected from 100 locations around Japan were surveyed, with a 55.2% response rate. As with the MMAS, the prevalence of ED rose with age. From the prevalence found in the survey, 11 300 000 Japanese males aged 30–79 were estimated to have moderate to complete ED.7 The rapidly aging population of Japan leads us to anticipate a further rise in ED patients in the future.

NBDegree of severity classified by patients themselves. After explanation of a simple definition of ED, patients were asked to identify which of the following best fit their own symptoms:

No ED: always able to get and keep erection good enough for sexual intercourse

Mild ED: usually able to get and keep erection good enough for sexual intercourse

Moderate ED: sometimes able to get and keep erection good enough for sexual intercourse

Complete ED: never able to get and keep erection good enough for sexual intercourse

Classification of the severity of ED using a single question has been shown to correlate very closely with the International Index of Erectile Function (IIEF),8 an index with confirmed validity9.

Risk factors


In both of the abovementioned MMAS6 and the Japanese surveys,7 the prevalence of ED increases with age. In an epidemiological survey comparing 289 residents in a Japanese community and 2115 residents in a community in the USA, the decline in erectile function was more marked in the Japanese responders, with 71% of those in their 70s reporting moderate or complete ED.10 An increased prevalence of ED with age was also seen in a Japanese survey of company employees and their families.11 Similar results were seen in a Japanese survey of 4609 subjects undergoing outpatient treatment for chronic conditions,12 and a survey that analyzed data from 2084 routine health checks.13 The reasons for the increase in ED with age are thought to be the physiological aging process as well as the risk factors mentioned below, that also tend to increase with age.14


Tobacco smoking is known to cause vascular endothelial damage, making it a risk factor for coronary artery disease and cerebrovascular disease. Smoking is also an independent risk factor for ED.15 Although the MMAS did not find smoking to be an independent risk factor for ED, smoking did significantly increase the risk of ED in subjects with coronary artery disease or hypertension.6 In the MMAS follow-up study the odds ratio for ED in smokers versus non-smokers was 1.97.16 A meta-analysis yielded an ED prevalence of 40% in smokers, significantly higher than that of 28% in the general population.17 Subclassification of smokers according to their level of additional passive smoking (no passive smoking, only at home, only at work, both at home and at work) showed that the risk of ED increased in a dose dependent manner as the level of passive smoking increased.16 A dose dependency relationship has also been shown between smoking and the degree of stenosis of the arteries of the corpus cavernosum,18 and between smoking and a reduction in nocturnal erections.19

Unfortunately, there have been no vertical studies with large subject numbers of the effects of smoking cessation. A horizontal study conducted with 4500 retired military personnel showed a lower prevalence of ED in ex-smokers than in those presently smoking (2% vs 3.7%).20

This could be called indirect evidence that smoking cessation lowers the risk of ED.14 Guay et al. reported that when heavy smokers with ED gave up smoking for 1 month, they experienced rapid and significant improvement in the fullness and rigidity of their erections.21 These results indicate that smoking cessation improves erectile function in younger subjects, but there have been a number of reports of no amelioration of ED with smoking cessation.22,23

In a large scale survey of over 110 000 Japanese junior and senior high school students, 29.9% of male students in the first year of junior high school had ever smoked (4.6% smoked every day). Smoking rates rose each year, reaching 55.6% ever smokers in the third year of senior high school (7.1% smoked every day).24 Considering the present situation where many smokers begin smoking before the age of 18, we should warn male high school students that in order to maintain good sexual function they should not smoke, and if they do smoke, they should quit immediately [Recommendation Grade C1].


According to the MMAS, 15% of patients on treatment for hypertension have complete ED.6 A subsequent survey conducted in seven Western countries also demonstrated a prevalence of 26% for ED in hypertensive respondents.25 Conversely, a study that analyzed the data for 28 000 US recipients of managed care found that 41.2% of ED sufferers also had hypertension, whereas only 19.2% of males without ED suffered from hypertension. The odds ratio for hypertension against ED was 1.383, after correction for age and other background factors.26 We must consider adverse reactions of antihypertensive medication (discussed below under Pharmacological ED) as a possible cause of ED in hypertensive patients, although most patients are already aware of some symptoms of ED before commencing antihypertensive medication.26 It appears definite that hypertension is an independent risk factor for ED. The odds ratios for hypertension against ED in two Japanese studies were significant at 2.79 and 1.79, respectively,11,12 although a survey of data from routine health checks yielded an odds ratio of 0.68 that was not significant.13

From a study of spontaneously hypertensive rats (SHR), the etiological mechanism of ED in hypertensive subjects is thought to involve structural changes in the penile vasculature and corpus cavernosum, and hypertensive endothelial damage.27


A review of 23 epidemiological studies published after 1958 estimated the prevalence of ED in diabetic patients at 26–35%.28 According to the MMAS, the risk of complete ED in diabetics was three times that in non-diabetics.6

A cohort study of over 30 000 middle-aged and elderly medical workers calculated an odds ratio of 1.32 for the risk of ED in diabetics vs non-diabetics.29 An Italian survey of 10 000 diabetics demonstrated an increase with age in the prevalence of ED, from 4.6% in the twenties to 45.5% in the sixties, and an overall rate of 35.8%.30 A Japanese survey of 1118 diabetic subjects (age range 40–79) found 64% of respondents with at least moderate ED according to the IIEF5.31 The odds ratios for diabetes vs ED in the three previously mentioned Japanese epidemiological surveys were all significant at 3.84, 2.88 and 2.54, respectively.11–13

A review article in 1998 stated that ED in diabetics is basically organic ED.32 Chronic and progressive in nature, ED can also be an early symptom of diabetes. Libido is generally maintained. A survey of 126 men with ED as the presenting complaint revealed that 17% had already been diagnosed with diabetes, 4.7% were newly diagnosed with diabetes, and 12% were found to have impaired glucose tolerance. Furthermore, the diagnosis had been missed in previous tests for diabetes in 80% of cases.33 Measurement of blood sugar levels is therefore an essential part of the assessment of ED patients [Recommendation Grade B].

In the diabetic patients, the prevalence of silent coronary artery disease is very high. Silent coronary artery disease is a strong predictor of coronary events and cardiac death. In uncomplicated type 2 diabetic patients, sensitivity, specificity and positive predictive values for ED were 33.8%, 95.3% and 88.2%, respectively. ED may be a potential marker with silent coronary artery disease among diabetic patients. An exercise ECG is suggested to be useful before starting a treatment for ED in diabetic patients34[Recommendation Grade B].

The mechanism of the onset of ED in diabetic patients is thought to include autonomic neuropathy, disorders of the penile vasculature and endothelial damage.28


Overseas data indicate that total cholesterol levels are raised in 26% of men with ED.35 A follow-up study of 3250 males aged 25–83 with dyslipidemia (no ED at study commencement) found that the odds ratio for the risk of ED in subjects with a total cholesterol level >240 (mg/dL) versus those with a level <180 was 1.83. The odds ratio for ED in subjects with a high density lipoprotein (HDL) level >60 (mg/dL) versus those with a level <30 was 0.3.36 The MMAS demonstrated a negative correlation between HDL levels and the rate of ED.6

The abovementioned three Japanese epidemiological studies did not show any significant differences in the prevalence of ED with any lipid parameters,11–13 so it may be that dyslipidemia does not necessarily increase the risk of ED in Japanese men.

The effect of lipid-lowering agents on erectile function is a controversial area. In a study of ED patients in whom an elevated total cholesterol level was the only risk factor, administration of atorvastatin significantly reduced total cholesterol levels (P < 0.001), and the IIEF erectile function domain also improved significantly from 14.2 to 20.7 (P < 0.001).37 On the other hand, both statin and fibrate classes of lipid-lowering agents have been implicated as possibly causing pharmacological ED.38

The athelosclerosis in the pelvic vascular beds is presumed to be the cause of dyslipidemia-induced ED.39

Overweight and exercise

In comparison with males in the healthy weight range, the odds ratio for ED is 60% greater in overweight males, after correction for other factors.40 The MMAS follow-up study also found a positive correlation between body mass index (BMI) and exacerbation of ED, and a negative correlation between BMI and amelioration of ED.41 A Boston study that followed over 1000 men aged 40–70 for an average of 8.8 years found a positive correlation between overweight and ED (P < 0.006), and a negative correlation between exercise and ED (P < 0.01).23 The Japanese analysis of data from routine health checks defined overweight as a BMI ≥25 kg/m2, and yielded an odds ratio of 0.94 for overweight vs ED, that was not significant.13

In a cohort study of over 30 000 medical workers aged 53–90, the risk of ED in the group doing at least 32.6 metabolic equivalents (METS) weekly versus the group doing only 2.7 METS exercises a week was 0.7 (95% confidence interval 0.6–0.7). The risk of ED in the group with a BMI >28.7 versus the group with a BMI <23.2 was 1.3 (1.2–1.4).42

An intervention trial was conducted with 110 overweight males aged 35–55, with an IIEF5 score ≤21 points, but no diabetes, hypertension or dyslipidemia. The intervention group was given detailed individual advice regarding calorie intake and exercise, bringing about weight reduction ≥10%. The control group was given verbal and written information about diet and exercise. After 2 years, the intervention group had reduced their average BMI from 36.9 to 31.2, whereas the control group had only dropped from 36.4 to 35.7 (P < 0.01). The IIEF5 score rose from 13.9 to 17 (P < 0.01) in the intervention group, but was unchanged in the control group, going from 13.5 to 13.6.43 An appropriate level of exercise should therefore be recommended from both preventive and therapeutic viewpoints [Recommendation Grade B].


Using the MMAS data, a study was conducted into the relationship between depression and ED, defining depression as a score of at least 16 points on the Center for Epidemiological Studies Depression (CES-D) scale. Multiple regression analysis revealed a strong correlation, with an odds ratio for depression and ED of 1.82 (1.21–2.73).44 A vertical study conducted in Finland that followed up 1380 men in their sixth to eighth decades found a strong correlation between ED and depression. The adjusted odds ratio for ED in subjects with depression at the commencement of the study vs those not on treatment for depression was 2.6, and the same odds ratio versus those on treatment for depression was 3.3. Conversely, the adjusted odds ratio for depression in subjects with ED at the commencement of the study versus those with no ED was 1.9.45 In a placebo-controlled randomized clinical trial of sildenafil in the treatment of 152 ED patients with depression, symptoms of depression improved significantly in subjects whose erectile function improved with administration of the test drug (responders) in comparison with non-responders.46 This suggests a strong correlation between the two conditions.

Lower urinary tract symptoms/benign prostatic hyperplasia

A survey conducted with 10 000 residents in seven Western countries yielded an odds ratio for lower urinary tract symptoms (LUTS) versus ED of 8.9 (6.85–11.55), greater than that of 1.96 for diabetes versus ED.47 In a Japanese survey of subjects with chronic conditions, the odds ratio for prostatic disease was 4.8,12 and a significant correlation was also seen in the routine health check data.11 The mechanism of ED in men with LUTS is thought to involve common factors such as pelvic ischemia.48 Treatments for LUTS and benign prostatic hyperplasia (BPH), both pharmacological using alpha-blockers,49 and surgical,50 have been shown to improve erectile function, so aggressive treatment of these conditions is recommended [Recommendation Grade B].


Several medications are known to trigger ED. Table 2 shows some typical agents that can cause ED. If patients are told that medication is the likely cause of their ED, this will lead to reduced compliance with their treatment. Agents used to treat hypertension and psychiatric disorders make up the majority of medications known to cause ED, and control of these conditions is far more important than treatment of ED, although for some patients the priorities will be reversed.51

Table 2.  Medications that may cause erectile dysfunction (ED)
Drug typeDrug classGeneric nameED causality degree
  1. +, case reports; ++, low frequency; +++, moderate frequency; ++++, high frequency; ±, reports of ED and improved erectile function. ACE inhibitor, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blocker; SSRI, selective serotonin reuptake inhibitor.

AntihypertensivesThiazide diureticHydrochlorothiazide+++
Loop diureticFrusemide+
K sparing diureticSpironolactone+++
Central sympathetic antagonistMethyl dopa+++
Peripheral sympathetic antagonistReserpine+++
Alpha-beta blockerLabetalol++++
Ca antagonistNifedipine++
ACE inhibitorEnalapril++
AntidepressantsTricyclic antidepressantImipramine+++
Antiulcer drugsH2 receptor antagonistCimetidine+++
Dopamine antagonistSulpiride+++
Male hormone antagonistsAntiandrogenChlormadinone acetate+++
LHRH analogLeuprorelin+++
Lipid lowering agentsStatinAtorvastatin±

For treatment of hypertension, angiotensin converting enzyme (ACE) inhibitors and angiotensin receptor blockers (ARBs) are recommended because they are less likely to cause ED51[Recommendation Grade B].

It is a problem that even if the offending medication is removed, erectile function does not always recover.51 General physicians are therefore encouraged to discuss this specific potential adverse reaction with patients before prescribing any medication that may cause ED [Recommendation Grade B].

ED as a marker or early presentation of cardiovascular disease

Overseas data indicate that many patients with cardiovascular disease have symptoms of ED, and that cardiovascular disease and ED have a common pathological basis (endothelial damage) and a number of common risk factors.52 In diabetics, ED is the most effective predictive factor for coronary artery disease.34 Analysis of the placebo group in the Prostate Cancer Prevention Trial (PCPT) showed that the presence or onset of ED was a significant predictor of subsequent cardiovascular events (hazard ratio 1.25).53

Even in Japan, with considerably less ischemic heart disease than Western countries, a study of 58 subjects presenting with ED, including stress electrocardiography, found 18 subjects (31%) with ischemic heart disease or in the high risk group.54

Accordingly, one should always keep the possibility of ischemic heart disease in mind when seeing a patient with ED of uncertain etiology [Recommendation Grade B].

Management of patients with cardiovascular risk factors

A group of cardiologists and urologists met at Princeton University in 1999 and 2004, and issued recommendations based on the abovementioned correlation between ED and cardiovascular disease. Patients should be managed in accordance with the recommendations of the Princeton Consensus Panels (Fig. 1) [Recommendation Grade B].55,56 Assessment of cardiac function should also be repeated on a regular basis, at a recommended interval of 6 months.

Figure 1.

Princeton Consensus Panel treatment algorithm.

Princeton Consensus Panel Risk Categories

Low risk group

  • • asymptomatic, less than three risk factors (other than maleness)
  • • controlled hypertension
  • • mild, stable angina pectoris
  • • postsuccessful coronary revascularization
  • • asymptomatic, no ischemia on stress test post myocardial infarction (6–8 weeks)
  • • mild valvular disease
  • • left ventricular dysfunction (New York Heart Association [NYHA] class I).

Intermediate or interminate risk group

  • • asymptomatic, three or more risk factors (other than maleness)
  • • moderate, stable angina pectoris
  • • history of myocardial infarction (2–6 weeks)
  • • left ventricular dysfunction or congestive heart failure (NYHA class II)
  • • non-cardiac sequelae of atherosclerotic disease (e.g. cerebrovascular disease).

High risk group

  • • Unstable or refractory angina pectoris
  • • uncontrolled hypertension
  • • congestive heart failure (NYHA class III/IV)
  • • recent myocardial infarction (<2 weeks)
  • • high risk arrhythmia
  • • obstructive hypertrophic cardiomyopathy
  • • moderate to severe valvular disease.

Addendum main risk factors for cardiovascular disease

  • • age
  • • maleness
  • • hypertension
  • • diabetes mellitus
  • • smoking
  • • dyslipidemia
  • • sedentary lifestyle
  • • family history of premature coronary artery disease.


Basic assessment

Consulting room environment

There are particular problems in Japan with privacy and allocation of sufficient time. Consultations should be conducted in a relaxed atmosphere, with no time pressures, in a room from which conversations cannot be overheard.

It is desirable to involve the partner in the assessment and any decisions about treatment.


The history should concentrate on the time of onset of ED (temporal relationship to the onset of any underlying conditions), time course of the onset (sudden or gradual), other sexual dysfunctions (problems with ejaculation, libido, or orgasm), relationship with their partner, any previous treatment and their effects. When attempting to differentiate between physiological and psychogenic ED, the diagnosis of psychogenic ED on the basis of no obvious organic cause. Psychogenic ED is a positive diagnosis made when a definite psychological cause has been identified.57

Other conditions

In addition to the previously mentioned risk factors (p. 2), check for other conditions such as kidney disease, neurological diseases, trauma, surgery or radiation to the pelvic region, voiding difficulties, and sleep disorders.

Medications and drug use

The previously mentioned prescription medications are not the only ones to cause ED, so always check for use of over the counter (OTC) medications. Also ask about smoking (record as package-years) and alcohol intake.


Record what type of exercise (walking, jogging, golf, cycling, etc.) is done for how long and how often. This is done to determine whether there is a lack of exercise, and also as a rough measure of cardiac function. See Table 3 for oxygen consumption estimates.58

Table 3.  Predicted oxygen consumption
  1. METS, metabolic equivalents.

Walking on flat surface at 3.2 km/h2 METS
Walking on flat surface at 4.8 km/h3 METS
Sexual intercourse (preorgasm)2–3 METS
Sexual intercourse (at orgasm)3–4 METS
Cycling on flat surface at 16 km/h6–7 METS
Stress test (Bruce stage 4)13 METS

Erectile function questionnaire

The IIEF559 Japanese version (Table 4)60 is generally used. The linguistic validity and retest reliability of the Japanese translation have been confirmed.61

Table 4.  International Index of Erectile Function (IIEF5)
In the past 6 months 
  1. Grading using the IIEF5 questionnaire

  2. Total score

  3. Severe

  4. 1–7

  5. Moderate

  6. 8–11

  7. Mild to Moderate

  8. 12–16

  9. Mild

  10. 17–21

  11. No ED

  12. 22–25

(1) How do you rate your confidence that you could get and keep an erection?
 Very low1
 Very high5
(2) When you had erections with sexual stimulation, how often were your erections hard enough for penetration?
 Almost never/never1
 A few times2
 Sometimes (about half the time)3
 Most times4
 Almost always/always5
(3) During sexual intercourse, how often were you able to maintain your erection after you had penetrated your partner?
 Almost never/never1
 A few times2
 Sometimes (about half the time)3
 Most times4
 Almost always/always5
(4) During sexual intercourse, how difficult was it to maintain your erection to completion of intercourse?
 Extremely difficult1
 Very difficult2
 Slightly difficult4
 Not difficult5
(5) When you attempted sexual intercourse, how often was it satisfactory for you?
 Almost never/never1
 A few times2
 Sometimes (about half the time)3
 Most times4
 Almost always/always5

Physical examination

This should include measurement of height and weight, and calculation of the body mass index (BMI), cardiovascular (blood pressure and pulse) and neurological examinations, checking secondary sex characteristics and the external genitalia (e.g. penile deformity or sclerosis, testicular atrophy), and a digital prostate examination in patients over 50 (check for hypertrophy/cancer).

Laboratory investigations

If no results are available from within the previous year, test the random blood sugar, total cholesterol, low density cholesterol (LDL), high density cholesterol (HDL), and fasting triglyceride levels.

If there are any findings suggestive of hypogonadism, morning free testosterone levels should be measured.NB

NBOverseas guidelines recommend measurement of total testosterone levels,1,2 but this is thought to be due to cost considerations. Comparison with total testosterone levels measured overseas should be made with caution because the kits are different, giving different reference values and minimum levels. The Japanese ‘Late Onset Hypogonadism Treatment Guide’ recommends measurement of free testosterone levels.62

ED diagnostic flow chart

A diagnostic flow chart of ED is shown in Figure 2.

Figure 2.

Erectile dysfunction (ED) diagnostic flow chart.

Optional and specific investigations (specialist initiated)

These may include psychiatric evaluation, nocturnal penile tumescence test, intracavernous injection test,NB Doppler ultrasonography, cavernosometry, cavernosography, penile angiography, computed tomography (CT) angiography, and specific endocrine tests.

NBEligibility for workers' compensation benefits for ED in Japan requires rigiscan evaluation of nocturnal penile tumescence and vascular assessment using an intracavernous injection of prostaglandin E1 (PGE1). 63



The goal of therapy should be viewed as restoration of a satisfactory sexual life, not only a rigid erection.1,2

Management of risk factors and concurrent disease [Recommendation Grade B]

  • • If the patient is a smoker, recommend smoking cessation.
  • • Advise lifestyle changes and increased exercise if the patient is clearly overweight or lacks exercise.
  • • If hypertension, diabetes or depression has been newly diagnosed, refer the patient to a specialist for the appropriate treatment. If LUTS/BPH is newly diagnosed, consider treatment with an alpha-blocker or urological referral.
  • • If pharmacological ED is suspected, the patient's treating physician should be contacted, and cessation/dosage reduction/change of the causative medication considered, as long as treatment of the underlying condition is not adversely affected.

Identification and treatment of curable ED

Testosterone deficiency

Due to the diurnal variation in testosterone production, samples should be taken in the morning. Based on data from 1143 healthy Japanese men (aged 20–77 years), the Japanese reference values for free testosterone are as follows: low <8.5 pg/mL; borderline ≥8.5, <11.8 pg/mL; normal ≥11.8 pg/mL. The Japanese reference range for total testosterone is 2.01–7.50 ng/mL.64

A meta-analysis of the effects of testosterone on male sexual function concluded that testosterone replacement is an effective treatment for patients with vasculogenic ED and low testosterone levels65[Recommendation Grade B].

There is, however, a tendency for the benefits of this treatment to decline over time. Testosterone supplementation is contraindicated in patients with prostate cancer. Rectal examination and determination of the prostate specific antigen (PSA) level must be carried out before commencement of therapy.

Patients on supplementation should be periodically checked for efficacy, as well as hepatic function and prostate status. Another potential adverse reaction of concern is polycythemia. The abovementioned meta-analysis examined data from 17 RCTs, involving 656 subjects (testosterone group 284 subjects, placebo group 284, crossover group 88). Doubts remain concerning safety with long-term administration, due to low subject numbers and a median follow-up period of only 3 months (range 1–36 months). A possible publishing bias has also been pointed out. The results of this meta-analysis indicate that administration of testosterone to males whose testosterone levels are not reduced does not improve erectile function. Considering the increase in breast cancer in women on hormone replacement therapy,66 some have warned not to make a similar mistake with men.67

Testosterone preparations available in Japan are oral methyltestosterone, injectable testosterone proprionate and testosterone enanthate, and testosterone cream (OTC). Hepatic dysfunction is a problem with the oral preparation,68 and there is a lack of data concerning blood levels with use of creams,69 so intramuscular injections are generally used.

Post-traumatic arteriogenic ED in young men

In younger patients with traumatic arterial damage, ED may be cured or improved by increasing arterial inflow through arterial bypass surgery (long-term success rates 60–70%).70 In a 5-year follow-up study, self-assessment of symptoms by 51 Japanese patients yielded a 67.5% efficacy rate.71 Younger patients like these should be assessed and treated by a surgeon experienced in this area [Recommendation Grade B].

Psychogenic ED

The patient and their partner should be given education and counseling, along with pharmacotherapy for ED if appropriate.

Education involves explanation of the erectile mechanism in an easily understood way with the use of educational materials (pamphlets and materials from the makers of the phosphodiesterase 5 (PDE5) inhibitors can be used). It should be emphasized that ED is a disease, and a treatable one.

In general, psychotherapy takes time, and is variable in its efficacy.57 Some will undoubtedly require specialist treatment, however, and in the following cases the patient should be referred to a mental health specialist:

  • 1ED from the first attempt at sexual intercourse, with no apparent organic cause.
  • 2PDE5 inhibitors ineffective, despite no history of neurological or vascular disease.
  • 3Patient or partner has experienced sexual abuse or sexual trauma.
  • 4Depression (particularly untreated).72


Phosphodiesterase 5 inhibitors

Phosphodiesterase 5 (PDE5) inhibitors, found in abundance in the corpus cavernosum, is an enzyme that breaks down cyclic GMP, an intracellular second messenger for nitric oxide (NO). PDE5 inhibitors competitively antagonize the action of PDE5, raising intracellular levels of cyclic GMP in corpus cavernosum smooth muscle, leading to smooth muscle relaxation in the corpus cavernosum and an erection in response to sexual stimulation.

In Japan, the three agents in this class presently available on prescription are sildenafil, vardenafil and tadalafil.

  • (a) Sildenafil (Viagra) [Recommendation Grade A]

This was the first PDE5 inhibitor clinically used worldwide. It has been prescribed to over 1 000 000 patients in Japan (more than 20 000 000 worldwide). Onset of action occurs 30–60 min after oral administration.73 A meta-analysis of 11 placebo-controlled RCTs collated data from over 3000 subjects, assessing efficacy using the global assessment question (GAQ), ‘Did the treatment that you used for your erections improve your erections during the time of therapy?’ The efficacy rate, or proportion of the sildenafil group who answered ‘yes’ for the 12-week study period, was 76% (placebo group 22%). Of these, improvement in erections was reported by 65% (placebo 12%) of subjects with severe ED according to the IIEF.74

A safety study collated the results of a placebo-controlled double blind trial and an open label study (sildenafil group 6986 subjects, placebo group 5054), examining the incidence of cardiovascular events, and area of concern when sildenafil was first released. No difference was seen in the incidence of myocardial infarction between the sildenafil group (0.58/100 person-years), placebo group (0.95/100 person-years), nor was any difference seen in the overall death rate of 0.37/100 versus 0.53/100.75 A Japanese clinical trial yielded similar efficacy and safety rates to overseas studies.76 Only the 25 mg and 50 mg dosages are allowed in Japan, so we are unable to use the 100 mg tablets available elsewhere. If sildenafil is taken after a meal, absorption is delayed, lowering blood concentration levels and reducing its effectiveness. A study of 12 non-Japanese subjects examining the interaction between alcohol and sildenafil, a commonly seen combination, found that alcohol (0.5 g/kg body weight) had no effect on the pharmacokinetics of sildenafil.

Diastolic blood pressure in supine and reclined positions were significantly lower in the alcohol/sildenafil group than in the alcohol/placebo group.77

  • (b) Vardenafil (Levitra) [Recommendation Grade A]

Onset of action occurs 30 min after oral administration.78 A meta-analysis of nine placebo-controlled RCTs collated data from over 4286 subjects participating in 12 week studies. Improved erections were reported by 69% of the vardenafil group, and 26% of the placebo group (P < 0.00001).79 In a 2-year long-term study, 90–92% of subjects who completed the trial answered ‘yes’ to the GAQ. There were no cardiovascular events.80 A Japanese clinical trial yielded similar efficacy and safety rates to overseas studies.81

Vardenafil is contraindicated in patients taking Class Ia (e.g. quinidine, procainamide) or Class III antiarrhythmic agents (e.g. amiodarone), and in patients with long QT syndrome.56 The dosages available in Japan are 5, 10 and 20 mg (although 20 mg tablets cannot be used in elderly patients aged over 65). The efficacy of vardenafil is reduced if it is taken after a high fat meal. The interaction between alcohol and vardenafil was examined in a study of 12 healthy non-Japanese subjects. It found that alcohol (0.5 g/kg body weight) had no effect on the pharmacokinetics of vardenafil.82 Comparisons between an alcohol group, vardenafil group, and alcohol/vardenafil group, found only that alcohol raised the heart rate, with no effects of vardenafil.

  • (c) Tadalafil (Cialis) [Recommendation Grade A]

Onset of action occurs 30 min after oral administration, and is maintained for 36 h.83 This prolonged duration of action is the big difference between this agent and the abovementioned two. Another characteristic of tadalafil is that it also inhibits PDE11. The significance of inhibiting PDE11, mainly found in the prostate, testis and skeletal muscle, is unknown. In a meta-analysis of five placebo-controlled RCTs, 41–81% of 1112 subjects with mild to moderate ED reported improved erections (placebo 35%). They further reported successful intercourse in 73–80% of sexual opportunities from 30 min to 36 h after taking tadalafil.84

Back pain is a typical adverse reaction of tadalafil, occurring at a frequency of 5%. A safety study collated the results of a placebo-controlled double blind trial (tadalafil group 12 487 subjects, placebo group 2047) and an open label study, found no difference in the incidence of cardiovascular events between the tadalafil group (0.40/100 person-years) and the placebo group (0.43/100 person-years).85 A Japanese clinical trial yielded similar efficacy and safety rates to overseas studies.86 The dosages available in Japan are 5, 10 and 20 mg. Absorption of tadalafil is affected very little by meals.

The interaction between alcohol and tadalafil was examined in a study of around 100 healthy non-Japanese subjects. It found that alcohol (0.6 g/kg body weight) had no effect on the pharmacokinetics of tadalafil. Comparisons between a tadalafil/alcohol group and a tadalafil/placebo group found only that tadalafil enhances the hypotensive effect of alcohol.87

  • (d) Comparisons of the three PDE5 inhibitors

The efficacy and tolerability of these three agents have been confirmed in comparative trials and postmarketing surveillance. A number of comparative studies have examined the question of which PDE5 inhibitor is superior.88,89 Such studies are inevitably imperfect due to the fact that the primary endpoint relies on subjects' impressions, and the fact that blind trials are impossible because of the characteristics of on demand dosing, differences in onset and duration of action, and reduced absorption with meals. The best course is therefore to explain to each patient the characteristics of each agent (see Table 5), and have them make the choice [Recommendation Grade C1].

Table 5.  Characteristics of the phosphodiesterase 5 (PDE5) inhibitors
  1. NA, not available.

Tmax (h)0.80.7–0.92
T1/2 (h)3–54–517.5
Protein binding rate (%)969197
Bioavailability (%)4015NA
Use with alpha-blockersCaution requiredCaution requiredCaution required
Effect of mealsDelayed absorption and onset of actionNone (reduced effect with high fat meal)None
  • (e) Evaluation of efficacy

Clinical efficacy is often evident with the first dose. Due to the need to determine the appropriate dosage and sexual stimulus, patients should be carefully advised on the optimum use of their medication. Therapeutic effect therefore improves with repeated use.

Even after a lack of response to several doses of a PDE5 inhibitor, it is still premature to call it ineffective. A study of 236 subjects referred by their general practitioner due to apparent non-response to sildenafil found that 98 subjects (41.5%) responded after re-education using videotapes and written material.

The reason for the initial treatment failure was inappropriate use (e.g. after a heavy meal, no sexual stimulation, wrong timing, only tried a few times) in 81% of cases.90

Reported salvage rates for initial treatment failures are of the order of 42–59%,91–95 indicating that half of these patients can be salvaged with re-education. The European Association of Urology (EAU) Guidelines accordingly state that a minimum of four doses is required for non-response to be diagnosed2[Recommendation Grade B].

There are also undoubtedly many cases where treatment is stopped despite initial success. One study reported a dropout rate of 57% over 3 years.96 A Japanese study of 667 subjects with initial treatment success found that 48% discontinued treatment over the 3-year follow-up period.97

  • (f) Adverse reactions and contraindicated concomitant medication

Concomitant use of PDE5 inhibitors and nitrates lowers blood pressure to dangerous levels, and is therefore contraindicated. Caution is advised in administering a PDE5 inhibitor in conjunction with an alpha-blocker (either for treatment of hypertension, e.g. terazosin, or for BPH, e.g. tamsulosin), due to augmentation of the hypotensive effect of the latter, so the former should be introduced at a low dosage.56

The PDE5 inhibitors have known class-specific adverse reactions. Some typical adverse reactions are headache, dyspepsia, facial flushing, nasal congestion, and visual disturbance (caused by PDE6 inhibition – glare, seeing everything tinted blue, unrelated to nonarteritic anterior ischemic optic neuropathy [NAION]: see below). These adverse reactions are usually mild to moderate in degree, and transitory.

There have been a few reports of priapism, an early cause of concern, associated with sildenafil,98 and one case with tadalafil.99 Although this is an extremely rare adverse reaction, an erection lasting more than 4 h requires immediate treatment.

Another adverse reaction, NAION, was first reported in 2000, and attracted a great deal of attention in 2005. The package inserts for each PDE5 inhibitor was also changed at that time. NAION is a condition affecting the optic nerve of unknown etiology that presents as a sudden, pain free, visual field defect (usually unilateral). It is often first noticed upon awakening. Fundal examination reveals findings of optic nerve edema. An American epidemiological survey estimated the yearly incidence in Caucasians aged 50 years and over as 2.3–10.2 per 100 000. This yields an estimated 1500–5000 new cases each year in the USA.100,101 The Ischemic Optic Neuropathy Decompression Trial (IONDT) cohort study (418 subjects, 5 years) identified vision impairment at the commencement of the observation period and diabetes as statistically significant risk factors for NAION.102 There are no effective treatments or preventive measures for this condition. The aims of treatment are to prevent enlargement of the lesion in the affected eye, and protect the unaffected eye. NAION and ED share common risk factors, so the relationship between NAION and ED is the same as that between ED and cardiovascular diseases. Although some are of the opinion that concurrent ED and NAION are the product of chance,103 there are also some reproducible case studies, and at present all that can be said is that the relationship between NAION and PDE5 inhibitors is unclear.

There have been no epidemiological studies conducted in Japan, and the incidence of NAION in this country is unknown. The search engine on the Japanese Ministry of Health, Labour and Welfare yielded no reports by the use of any PDE5 inhibitors.104 Patients should be informed of this adverse reaction, and advised to stop their medication and see an ophthalmologist if they experience any visual disturbance [Recommendation Grade C1].

Local treatments

  • 1Intracavernous injections [Recommendation Grade B]

This involves injecting prostaglandin E1 (PGE1) into the corpus cavernosum. The usual dosage is 5–20 μg of PGE1 dissolved in 1 mL of physiological saline. The mechanism of action is that PGE1 binds to EP receptors within smooth muscle cells in the corpus cavernosum, activating adenylate cyclase. This results in increased levels of cyclic AMP, smooth muscle relaxation, and an erection. PGE1 is metabolized by 15-hydroxy-PGE1 dehydrogenase present in the corpus cavernosum.105 This treatment shows high efficacy and moderate tolerability in most ED patients. The main adverse events are local in nature, acutely penile pain (50%) and rarely priapism (ischemic, 1%), and chronically penile fibrosis (2%) and hematoma/subcutaneous hemorrhage (8%).106 In particular, priapismNB is a serious condition requiring urgent treatment, corresponding to the ‘acute penis’. The cavernosous injections should therefore only be prescribed by institutions able to deal with this adverse reaction.

Although this treatment is effective in patients in whom PDE5 inhibitors are ineffective or contraindicated, it has not been approved for clinical use in Japan. Some trials are underway, led by Okayama University.107


Priapism is defined as ‘an erection that is maintained for greater than 4 h despite the absence of both physical and psychological stimulation’. The choice of 4 h for this definition was made on the basis of experience of how long patients take to get to hospital, and how long it takes to organize investigations and treatment. Edema of the corpus cavernosum smooth muscle begins 12 h after the onset of priapism, and after 24 h endothelial cells (a source of NO, the trigger for erections) begin to die off, with platelets adhering to the site. Necrosis of corpus cavernosum smooth muscle is seen after 48 h.108 The time limit of 4 h was set on the basis of this report, to allow a reasonable safety margin.

Here we show you the recommendations by AUA how to diagnose and treat ischemic priapism, a urological emergency.109

  • • Diagnosis

This is readily made on the basis of the history and observation of the erect penis. It can also be readily determined whether it is ischemic or not by aspirating the corpus cavernosum, and carrying out arterial blood gas analysis.

  • • Treatment

First insert a 19 or 21 gauge butterfly needle into the corpus cavernosum, aspirate blood and flush with physiological saline. Then inject 1 mL of 100–500 μg/mL phenylephrine, the α-agonist. Repeat the above procedure every 3–5 min until 1 h has passed. This procedure relieves 43–81% of cases of priapism. Recurrent cases require surgical treatment, and should be referred to a urologist.

  • 2Vacuum constriction devices [Recommendation Grade B]

These devices apply negative pressure, drawing blood into the shaft of the penis, and then apply a rubber band to the base of the penis to keep this blood in the penis. This is therefore a pseudo-erection rather than a physiological erection. Although an erection capable of intercourse is achieved in 90% of cases, reported satisfaction rates vary widely from 27% to 94%.110 The only such devices presently approved in Japan are the Vet-Co (internet sales, http://www.jinseido.co.jp/jin/etc/impo.html)111 and the vacuum constriction device ‘Kanki’ (Takei Ikakouki Co. Ltd: Tel. 03-3255-0711).112 They each cost around 50 000 yen, and directions for use are given in an accompanying videotape, although instruction from a medical practitioner is preferable. Adverse reactions include penile pain and numbness, subcutaneous hemorrhage, and ejaculatory difficulties. Anticoagulant therapy is a relative contraindication for the use of a vacuum erection device.

Surgical treatment

Vascular surgery

We have already discussed arterial bypass surgery (see page 7). Long-term results with venous surgery for venous leakage are poor, and so it cannot be recommended113[Recommendation Grade D].

Penile prosthesis [Recommendation Grade B]

The treatment of last resort for ED is surgical implantation of a prosthesis into the corpus cavernosum. Because it is invasive and irreversible, this procedure should only be carried out after all other treatment modalities have failed.

Infections occur rarely (1–5%), necessitating removal of the prosthesis and possibly resulting in severe scarification and deformity.114 The non-inflatable type AMS600 and Duraphase II, and inflatable type AMS700CXM were available for use in Japan until August 2007, after which time their import has been banned.115 At present, they can only be imported for private use.

ED treatment flow chart

A treatment flow chart for ED is shown in Figure 3.

Figure 3.

Erectile dysfunction (ED) treatment flow chart.