VOL. 16 NO. 3  1998

These pages are best viewed with Netscape version 3.0 or higher or Internet Explorer version 3.0 or higher. When viewed with other browsers, some characters or attributes may not be rendered correctly.

Erectile Dysfunction in Diabetes:
Pills for Penile Failure

Aaron Vinik, MD, PhD, FCP, FACP, and Donald Richardson, MD

In Brief

An understanding of the physiological mechanism of erection has led to the development of new oral therapies for erectile dysfunction that target different sites in the sexual arousal process. Apomorphine activates the arousal center of the brain. Phentolamine increases penile blood flow. Sildenafil enhances the action of nitric oxide, an endothelial-derived vasodilator and smooth muscle relaxant. These developments constitute a significant advance in a much-neglected area of male medicine.

A November 1997 Newsweek article by John Leland included the following paragraph as an introduction:

"New drugs awaiting approval by the FDA could be a boon to millions of men with erectile dysfunction (ED). They’re painless and discreet (no needles), a potential bonanza for pharmaceutical companies and the booming industry of male medicine. And if they make enough boomer guys feel like virile teenagers again, they could trigger another sexual revolution."

For millions of American men of the baby-boom era who are "marching toward their softening years," this could be the long-awaited miracle: better performance under the covers through chemistry.

The National Institutes of Health estimates that 10–20 million men in the United States have ED. The estimate worldwide is around 140 million. But this may be only the tip of the iceberg. There are 38 million men of the baby-boom era whose anatomies are beginning to break down and who may feel the need to restore their virility. For many, a limp penis equates with a limp ego and loss of gender identity, and even if men are not sexually active, the ability to have an erection and provide satisfaction on demand is essential for self-esteem. "If women can have face-lifts," these men ask, "why are men not entitled to penile upward mobility?"

Even more than the mid-life model red convertible, this is the thing that men will spend their money on. The introduction of MUSE, the pellet-in-the-shaft-of-the-penis treatment, raised the amount of money spent on ED in the United States from $6.4 million in 1995 to $117 million in the year since its introduction in January 1998. It has already been prescribed for 665,000 men.

The past year has seen an explosion in companies seeking to exploit the gigantic market for male medicine. Many are developing new pills, creams, and other medications to replace the old favorites, such as the Spanish beetle cantharides and yohimbine, the extract of male fern, which for years have been used as aphrodisiacs.

For centuries, physicians have tried to pump iron into failing male members. (In fact, iron prostheses are one form of treatment still used.) The mysteries of erection are now being solved, and new insights into the physiology of erectile function are being gained.

We now know that sexual arousal triggers the release of the vasodilator nitric oxide (NO) from the nerve endings and endothelium of the blood vessels of the corpora cavernosae in the shaft of the penis. The smooth muscle is normally in a contracted state, restricting blood flow to a mere trickle. When NO activates the cGMP system, the smooth muscle of the blood vessels, as well as that in the body of the shaft of the penis, is relaxed, allowing blood to fill the spaces. The engorged corpora pinch off the blood vessels draining the penis, trapping the blood in the corpora and producing an erection.

Leaky vessels would clearly compromise this action. Fortunately, this problem is found in only a small portion of men with ED.

In diabetes, a defective release of NO compromises the ability to achieve and sustain an erection. By using potent vasodilators or agents that block the inactivation of cGMP by antagonizing the specific phosphodiesterase (PDIE5) in the penis, the action of NO is enhanced, thus overcoming the defect in most conditions causing ED.

The pharmaceutical industry has raced to capitalize on this penile resuscitation bandwagon. Home injections of alprostadil, either as Caverject or the similar preparation Edex, have received FDA approval and are now in use throughout the country. A California company called Vivus introduced the same drug as MUSE, the soft-pellet treatment mentioned above. But these prostaglandin analogs are associated with pain, hematomas, priapism, fibrosis, or bother, and are not the answer for many patients.

One new drug has reached the U.S. market and two more are in their final phases of approval. Apomorphine (Spontane) is a morphine analog that stimulates the centers in the brain involved in the erectile process. It is useful in up to 70% of people with a mild disorder. Phentolamine (Vasomax) is an adrenergic inhibitor that has been used by injection but is now being introduced in pill form and has the ability to improve erections in 40% of mild cases. Sildenafil (Viagra) seems to be the most promising of the new therapies. Its action is to preserve cGMP, thereby increasing blood flow to the penis. It is ostensibly effective in up to 80% of patients.

Sildenafil, which reached the market in March of this year, was originally developed to treat hypertension and angina, but failed. However, it was noted that research volunteers would not return their medication, and some detective work revealed that the secrecy was due to improved sexual function. Patients insisted on the new hypertension medication that contrasted starkly with the usual impotence-inducing therapies for hypertension. Indeed, in one San Francisco urology clinic, patients not admitted to the study broke into the clinic seeking to steal the medication. Others in the trial sold their medication on the street.

Approximately 4,500 subjects had been treated with sildenafil before its release, with success rates between 48 and 81%, depending on the underlying cause of ED. The drug is now being prescribed to thousands of men per week and is thought to have a billion-dollar potential in its first year, with an anticipated $5 billion in sales within 3 years.

In the sections below, we review the physiology of erection and discuss the problems commonly encountered in diabetes. We also provide a new algorithm for the management of ED in diabetes and summarize the studies that have included sufficient subjects with diabetes for us to make a judgment call.

ED is defined as the consistent inability to attain and maintain an erection adequate for sexual intercourse. The term impotence implies failure of the individual. It should be abandoned in favor of terms such as penile failure, penile unreliability, or ED, which place the responsibility with the offending organ and not with the person.

An estimated 10–20 million, or more than 10%, of American men have ED. One in three men will experience the problem.

The major causes of ED in the United States are illustrated in Figure 1. Diabetes accounts for 30%, and vascular disease, of which diabetes is the dominant contributor, accounts for 40%. Surgery to the prostate and urogenital area, spinal cord injuries, and multiple sclerosis combined make up about 24%, while endocrine disorders cause only about 3–6%.

The etiology of ED in diabetes is multifactorial. Neuropathy, vascular disease, diabetes control, nutrition, endocrine disorders, psychogenic factors, and drugs used in the treatment of diabetes and its complications play a role.

After the age of 70, two-thirds of the general population have ED. The incidence of ED in diabetic men between the ages of 20–29 years is 9%, but this rises to 95% by age 70.

ED may be the presenting symptom of diabetes, and more than 50% of people with diabetes notice the onset of ED within 10 years of onset of the diabetes. Indeed, the later the onset of diabetes in life, the earlier the onset of ED. It may herald the development of diabetes and even precede abnormalities in glucose tolerance.

A large segment of the public, as well as many members of the health professions, remain uninformed or misinformed about ED. This lack of information, added to the pervasive reluctance of physicians to deal candidly with their patients’ sexual problems, has resulted in patients being denied treatment. While we strive to prolong the life of people with diabetes, we should not do so miserably, when there are now a variety of therapeutic options.

Before entering into a discussion of these treatment options, it might be well to consider the reasons for concern for patients with ED. ED is a marker for the development of generalized vascular disease and for premature demise from a myocardial infarct. It has been said that the penis is a divining rod capable of identifying individuals at risk for a vascular catastrophe long before one occurs. As a corollary, the prevalence of diabetes in men with ED is 50%, compared with 7% in an age-matched population without ED. ED cosegregates with elevated total and low-density lipoprotein (LDL) cholesterol, low high-density lipoprotein (HDL) cholesterol, and reduced levels of DHEA-S, all known risk factors for coronary vascular disease. Penile failure may be protective for men wishing to embark on foolish pursuits!

Pathophysiology of ED in Diabetes
An erection is dependent on orchestrated actions of the muscles, nerves, and blood vessels of the penis. It is a hemodynamic event governed by the integrity of smooth muscles in the arteriolar wall and sinusoids of the corpora cavernosa. It requires intact blood flow and sphincteric competence, both of which are regulated by the nervous system.

pg109.gif (30230 bytes)

Figure 1.  Major causes of erectile Dysfunction in the United States.

In the flaccid state, the corporeal smooth muscle is tonically contracted, allowing only a small amount of blood (1–4 ml/100 g tissue) to enter the penis. In the proper hormonal milieu, with either psychological or physical stimulation of the genital organs, autonomic nerves are activated, releasing a variety of neurotransmitters. These cause vasodilatation of the helicene arteries and relaxation of the corporeal smooth muscle, both of which produce large increases in penile blood flow, causing trabecular engorgement and entrapment of blood within the corpora by compression of nerves against a dense nondistensible fibrous sheath, the tunic albuginea.

pg110.gif (87314 bytes)
Figure 2. Schematic illustration of the anatomic structure of the penis.

Recently, it has become apparent that the endothelium plays an important role inasmuch as cholinergic activation is dependent on endothelial release of the potent vasodilators NO and prostacyclin, while neuropeptides, such as VIP, may be released from the nerves per se (Figure 2). The synthesis of NO from its substrate arginine is catalyzed by the enzyme NO synthase (NOS) and is defective in diabetes.

NO achieves its smooth-muscle relaxation effect by increasing the formation of cGMP, which in turn mobilizes Ca++, thereby invoking mechanisms for smooth-muscle relaxation. cGMP is hydrolyzed to 5’GMP by the enzyme phosphodiesterase (PDE). Inhibition of PDE by sildenafil potentiates the activity of NO by preserving higher corporal levels of the cyclic nucleotide. Thus, sildenafil potentiates the action of natural stimuli to arousal. It does not directly increase cGMP.

Prostacyclin, on the other hand, is synthesized from arachidonate and induces smooth-muscle relaxation by activating cAMP. Thus, it is likely that the two could be complementary and that one may work where the other has failed because of their differences in mechanism for regulating smooth-muscle relaxation. Prostacyclin has the ability to induce erections in the absence of erotic stimuli.

Both penile NOS and the spinal motor neurons innervating the striated erectile muscles (bulbocavernosus and ischiocavernosus) are androgen-dependent, and in diabetes NO is depressed in direct correlation with testosterone. The spinal motor neurons innervating the striated erectile muscles are also androgen-dependent. Thus, there appear to be two mechanisms outside the brain that would support an extra-libidinous role for androgens in penile tumescence.

Detumescence is heralded by the return of tone to the corporeal smooth muscle with reduction in size of the vascular sinuses and release of the compression of the subtunical vessels, allowing the corpora to drain and become flaccid. The major regulator of detumescence is norepinephrine acting via postsynaptic alpha-1 adrenergic nerves modulated by presynaptic alpha-2 receptor activity. It is for this reason that detumescence can be achieved with infusion of an alpha- adrenergic agonist, such as phenylephrine, and that erection can be achieved with an alpha-adrenergic-blocking drug, such as phentolamine.

Exceptions to this rule abound, and people with severe autonomic adrenergic insufficiency do not have priapism, suggesting once again that there are alternate modulators of corporeal detumescence.

pg111.gif (68667 bytes)
Figure 3. Schematic illustration of the interaction between endothelial cells and smooth muscle of the corpora cavernosa.

The relationship between erection and detumescence is outlined in Figure 3, and the various candidates for a role in this complex process are given in Table 1.

Table 1. Potential Autonomic and Nonadrenergic/Noncholinergic Modulators of Erectile Function
Modulator Detumescence Tumescence
Acetylcholine Yes
Norepinephrine (alpha-1 adrenergic) Yes
Peptide Histidine Methionine Yes
Neuropeptide Y Yes/No Yes/No
Somatostatin Yes/No Yes/No
Calcitonin Gene-Related Peptide
Substance P Yes
Histamine Yes Yes
Vasoactive Intestinal Peptide Yes

Prostaglandin E1


Prostaglandin F2 alpha


Thromboxane A2

Calcium Yes
Endothelin-1 Yes
Endothelial-Derived Relaxing
            Factor (Nitric Oxide)
Prostanoids Yes Yes
cGMP Yes Yes
Phosphodiesterase 5 inhibitors Yes

The partnership among the sympathetic and parasympathetic nervous system, the endothelium, and smooth-muscle function may be disrupted by diseases of blood vessels or nerves, such as occurs in diabetes. In diabetes, the development of autonomic neuropathy is partly responsible for the loss of cholinergic activation of the erectile process. In the penis, the acetylcholine acts upon the vascular endothelium to release NO and prostacyclin, both of which are defective in diabetes. There is also evidence that nonadrenergic/noncholinergic nerve function is hampered with decreased content of VIP, substance P, and other vasodilator neurotransmitters (Table 1).

The association of diabetes with atherosclerosis and microvascular disease further compounds the problem, and recent data indicate that circulating levels of the potent vasoconstrictor endothelin may be increased in diabetes. Thus, ED in diabetes is no simple matter and derives from a host of abnormalities.

Symptoms of ED
The symptoms of organic ED are gradual in onset and progress with time. The earliest complaints are decreased rigidity with incomplete tumescence before total failure. It occurs with all partners, and there is no loss of libido. Morning erections are lost.

Sudden loss of erections with a particular partner, while maintaining morning erections and nocturnal penile tumescence, suggests a psychogenic cause. Psychogenic factors may, however, be superimposed on organic dysfunction in diabetes.

The neurological manifestations are those of dysfunction of the autonomic nervous system (ANS) and include constipation, diarrhea, orthostasis, gustatory sweating, and postprandial fullness, among others. ANS dysfunction can be diagnosed simply on the basis of loss of beat-to-beat variation in heart rate (<10 bpm) with deep slow (6 breaths/minute) respiration.

Vascular disease is usually manifest by buttock claudication but may be due to stenosis of the internal pudendal artery. A penile/brachial index of <0.7 indicates diminished blood supply.

A venous leak manifests as unresponsiveness to vasodilators and needs to be evaluated by penile Doppler sonography.

Evaluation of ED
Initial assessment of patients with ED should be carried out in the presence of their significant other or sexual partner, if possible. Health care providers should make an attempt to interview the partner to obtain an impression of the overall relationship and the impact that the return of erections will have on the relationship. Extreme care must be exercised in these situations since, in many instances, a patient’s desire is to be able to have erections, but not necessarily to actually have sexual relations (with his wife).

A thorough work-up for impotence will include: medical and sexual history; physical and psychological evaluations; blood tests for diabetes and a check of levels of testosterone, prolactin, and thyroid hormones; tests for nocturnal erections (absence of erections during sleep suggests a physical cause of impotence); tests to assess penile, pelvic, and spinal nerve function; and tests to assess penile blood supply and blood pressure.

pg112.gif (48846 bytes)

Figure 4. Evaluation of diabetic patients with erectile dysfunction.

Diagnosis of the cause of ED is made by a logical stepwise progression (see algorithm in Figure 4).

In all instances, a careful history must be taken to determine the rapidity of onset of ED, presence of morning erections, uniformity of sexual dysfunction with all partners, evidence of autonomic nervous dysfunction, vascular insufficiency, hormonal inadequacy, and drugs used in the treatment of satellite disorders.

Physical examination must include an evaluation of the ANS, vascular supply, and the hypothalamic/pituitary gonadal axis. All patients failing a trial of an orally active agent should receive penile injection of a vasodilator intracavernously for diagnostic purposes and possible choice of further therapeutic options.

Problematic cases should be tested for nocturnal penile tumescence (NPT). Normal NPT defines psychogenic ED, and a negative response to vasodilators implies vascular insufficiency. It is not yet clear at this point in time whether patients with ANS dysfunction, vascular insufficiency, or both, are more or less likely to respond to the oral agents.

Treatment of ED is oriented towards removing the cause. Hormonal abnormalities require correction. Drugs for hypertension, cardiac failure, and depression or anxiety should be changed to those with no, or a lower, potential for ED. Glycemic control should be optimized.

There are three treatment options for diabetic men with ED: 1) no treatment (withdrawal from offending medications with psychosexual counseling), 2) medical treatment, or 3) surgery.

Table 2. Drugs Known to Cause ED and Commonly Used in Diabetic Patients
Antihypertensive agents
Thiazide diuretics
Agents acting on the central nervous system
Tricyclic antidepressants
Drugs acting on the endocrine system
Gonadotropin antagonist
Fibric acid derivatives

No Treatment
Before considering any form of treatment, every effort should be made to have the patient withdraw from drugs likely to promote ED, including alcohol (remember that the Bard pointed out that it enhanced the desire but decreased the performance), and to eliminate smoking even if nicotine patches are necessary. (Table 2).

Table 3. Sexual Dysfunction With Selected Antihypertensive Agents
Drug Percentage of Patients With ED Comments
Thiazide 4–30% Dose-dependent, less frequent with <2.5 ng
Spironolactone 3–20% Dose-dependent gynecomastia common
Sympatholytics (methyldopa, clonidine, reserpine, guanethidine) 8–80% Decreased libido
Beta-adrenergic blockers 0–43% Dose-dependent, lowest with beta-1 specific (pindolol, nadolol, atenolol), may occur with eye drops
Alpha-blockers Minimal May induce priapism
Labetolol Minimal Ejaculatory changes common
ACE inhibitors and antagonists Minimal
Vasodilators (hydralazine, minoxidil) Minimal Some patients report aphrodisiac effects of hydralazine
Calcium-channel antagonists Minimal

Choices of antihypertensive medications should be made on the basis of the least offensive agent (Table 3). Antidepressant medications can be selected for those with minimal tendency to compromise sexual function (Table 4).

Table 4. Sexual Dysfunction With Selected Psychiatric Medications


Percentage of
Patients With ED


Antipsychotics 0–54% More common with alpha-adrenergic and anticholinergic effects
Monoamine oxidase inhibitors 16–31% Also anorgasmia
Tricylics Significant Also decreased libido, anorgasmia, may be less with desipramine
Selective serotonin reuptake inhibitors 8–16% Also decreased libido, ejaculatory changes
Benzodiazepines Reported
Trazodone Priapism

Patients most likely to choose the no-treatment option are those with a long-standing disorder who have given up the thought of sexual relations and have transferred their energies to other avenues of reward. There are also patients with serious underlying vascular or other disease that limits their physical activity who should not pursue the level of exertion required to generate sexual satisfaction.

Whatever the decision, it is important to determine what alternatives to intercourse the couple has established or, if necessary, to educate them about the possible means of achieving joint satisfaction despite ED. In many instances, loss of this area of interpersonal reward is simply due to ignorance of the possibilities.

Psychosexual counseling is mandatory in these situations. Sexual therapy is vital for people with diabetes, since the chronic condition is fraught with situational stresses, performance anxiety, and problems in relationships, especially new ones in which the sexual partners have not yet found a common ground. Depression is rife in people with diabetes and its chronic complications. Attention must be given to psychological counseling and, if necessary, pharmacological management of the depression, which may enhance sexual performance in itself.

The development of ED may severely affect compliance with a regimen, and all men placed on antihypertensive medications should be counseled in this regard. The prevalence of ED may range from 9% in patients on diuretics to 30% in patients on combination drugs (Table 3). Thus, in hypertensive patients, a rise in blood pressure should lead to the suspicion that the patient has stopped taking his medication because of the loss of sexual functioning.

Table 5. Some Therapeutic Options for Erectile Dysfunction
Medication Action Use Pros and Cons
Viagra (sildenafil) Relaxes smooth muscle in corpora Taken 1 hour before sex, requires stimulation Headache, diarrhea, flushing, hypotension, disturbed color vision. Greatest help with partial ED
Spontane (apomorphine)
In final trials
CNS stimulant that triggers erections (dopa-minergic) Works in up to one-third of mild cases Requires stimulation.
Vasomax (phentolamine)
Under FDA review
Alpha-adrenergic antagonist relaxes smooth muscle and increases blood flow Taken 20–40 minutes before sex. Helps 60–80% of those tested. Requires stimulation.
Fewer side effects than sildenafil.
MUSE (alprostadil)
Approved 1997
Applicator inserted into urethra Applied 5–10 minutes before sex. Erection lasts 1 hour Can be used 2 times/day.
Not for use with pregnant partner.
Caverject (alprostadil)
Approved 1997
Injected into corpus cavernosa. Relaxes smooth muscle and dilates blood vessels Injected 20 minutes before sex. Lasts >1 hour. Effective >50%. May be very painful.
Cannot be used every day.
Edex (alprostadil) Different formulation alprostadil Injected with smaller needle. Injected10 minutes to 2 hours before sex. Erections last >1 hour. Cheaper than Caverject.
Invicorp (VIP and phentolamine).
FDA submission 1998
Relaxes smooth muscle Just before sex. More effective than either agent above. Requires stimulation.
No pain.
Vacuum pumps
Approved 1982
Draws blood into corpora by negative pressure Just before sex.
Erections last until elastic ring removed
Clumsy. Interferes with foreplay. Difficult ejaculation.
Penile implants
Since 1966
Rigid or malleable rod or inflatable Bent or pumped into erect state Destroys erectile tissue. Complications high in diabetes.
Rods embarrassing.

Medical Treatment
Several new drugs with therapeutic potential are now or soon will be available (Table 5). Again, though, it is essential for patients first to be removed, if possible, from drugs that are known to cause ED (Tables 3 and 4).

While the exact degree of erectile function, which is androgen dependent, remains unfirm, androgen replacement therapy does provide a return to normal sexual function for many hypogonadal men without diabetes. In hypo- or hypergonadotropic hypogonadal men, testosterone given by any of several routes results in a two- to threefold increase in measures of nocturnal penile tumescence and sexual activity. Clearly, any deficiency of androgens related to diabetes is of concern for those who provide for patients with ED. In contrast, administration of androgens to people with ED and normal gonadal function is usually not beneficial.

Great care must be exercised in diabetic patients before testosterone is given to be sure that there is no prostatic hypertrophy and levels of PSA <2 U/ml. Precautions must also be taken since testosterone can increase RBC mass, occasionally precipitating congestive heart failure in those at risk, or stroke.

Another possible caveat for the use of androgens is in men with an established lipid disorder. The high triglycerides and low HDL cholesterol levels in type 2 diabetes, for example, may be accentuated by the androgen and require modification of drug therapy for the dyslipidemia to avoid increasing the risk of a macrovascular event. However, several studies suggest that androgens may actually improve the lipoprotein profile, or at least not cause worsening. Nonetheless, it is probably wise to at least determine the lipoprotein profile before and 6 months after starting treatment with an androgen.

Androgens can be either long- or short-acting and can be given by the transcutaneous route as Testoderm or Androderm. However, these are expensive. Androgens without a methyl group are favored and include testosterone cypionate and enanthate intramuscularly or subcutaneously.

Doses vary with the individual and the degree of insufficiency, and 100–200 mg every 3 weeks initially can be tailored to the patients needs. Patients can usually sense when the drug is wearing off in anything from 2 weeks to a month, and the frequency of dosing can be adjusted according to libido, sex drive, and the general feeling of well-being, strength, and vitality. Elderly men or those with some enlargement of the prostate should be treated initially with a lower dose, for example, 50 mg every 14 days. Prostatic hypertrophy rarely progresses with this approach.

The patient’s clinical status is the best indication of effectiveness of androgen therapy. Increased libido, energy, and strength occur within days, and physical features may take several months. Prepubertal men will have profound physical and psychological changes that require counseling of both patient and his significant other before starting treatment.

In older men, examination of the prostate is mandatory before starting therapy and at 6-month intervals with treatment. PSA should be measured at 6-month intervals, but a doubling of the value can be due to testosterone itself and is not cause for panic. There is no need to measure testosterone or gonadotropin levels serially.

An added benefit often occurs in those people with pain and discomfort due to the presence of neuropathy. The pain improves with correction of the hypogonadism. Serum levels of testosterone do decline with age and may contribute to the loss of libido, muscle strength, and muscle mass. Preliminary studies suggest that androgen therapy may restore lean body mass, muscle strength, and libido and decrease bone turnover. There cannot, however, be justification for treating older people with androgens simply on the basis of these possible benefits without clear documentation of hypogonadism.

When the hypogonadism is due to elevated levels of prolactin, the treatment of choice is bromergocriptine mesylate (Parlodel). It is started initially with a dose of 2.5 mg taken at bedtime to avoid orthostatic hypotension, with a gradual increase in dose to 2.5 mg three times a day. Carbergoline (Dostinex) has recently been approved for hyperprolactinemia also. The impact on orthostatic blood pressures should be examined periodically, and patients should be given guidelines for prevention of syncope when arising too rapidly from the horizontal position or after taking hot baths.

Self-injection therapy
Over the past decade, there have been a number of reports on the use of intracavernosal treatment of ED with a wide variety of drugs. These include the smooth muscle relaxants papaverine and glyceryl trinitrate, alpha-adrenergic blockers phenoxybenzamine and phentolamine, the calcium-channel antagonist verapamil, the polypeptide VIP, the beta-adrenergic agonist isoprenaline, the antipsychotics/antidepressants trazodone and chlorpromazine, and prostaglandins. Of these, the vasoactive agents papaverine, phentolamine, and prostaglandin E1 (PGE1) are now in use.

These vasoactive substances stimulate the natural erectile process by inhibiting sympathetic tone and relaxing corporeal smooth muscle. Alprostadil and papaverine relax the smooth muscle of the corpora cavernosa, and phentolamine is a competitive inhibitor of alpha-adrenergic receptors. Because of the vascular cross-communication between the corporal bodies, injection into one causes bilateral tumescence.

Papaverine. Papaverine causes arterial dilatation, sinusoidal relaxation, and venous constriction, thereby increasing the blood retained in the corpora. The dose required to initiate and maintain an erection varies with age and cause of ED. In younger patients and those with neurogenic ED, 7.5 mg might be sufficient, while in those with severe vascular disease or in the elderly, doses as high as 60 mg are necessary.

The major problem with papaverine is corporeal fibrosis, which may resemble a Peyronie’s type deformity and is usually the result of repeated injections. Systemic adverse events include vasovagal reactions with bradycardia, hypotension, dizziness, and flushing. These are uncommon and can be treated with atropine. Complications include local infections, hematomas, and pain. Occasionally, systemic effects with abnormalities of liver functions are encountered, which revert with discontinuation of the drug.

Phentolamine. Phentolamine is an alpha-1 and -2 adrenergic antagonist that increases blood flow to the corpora primarily by decreasing arterial resistance and increasing arterial blood flow. It is usually given with papaverine to decrease the likelihood of fibrosis and prolonged erection.

The dose should be 1 mg for every 30 mg of papaverine. These drugs should not be used more than once a week if fibrosis is to be avoided. One of the benefits of alprostadil (see below) may be the fact that it has actions that may actually minimize the development of fibrosis.

Prostaglandin. Alprostadil is a prostaglandin derived from polyunsaturated fats and is naturally occurring in the body. It has been shown to have alpha-adrenergic-blocking properties, dilating blood vessels. It also relaxes smooth muscle directly via a prostacyclin receptor.

It is active in the dose range of 5–20 g with a maximum dose of 60 g, and there is a linear relationship between dose and duration of erection.

Prolonged erections can occur, but fibrosis is rare. The major side effect appears to be pain at the site of injection, but this precludes its use for ED in only 1% of people. It does cause local reactions and hematomas if abused or injected directly into a vessel.

A number of studies have reported on the effects of alprostadil in diabetic patients. In one study, 72 patients received 20 g of PGE1, and erections were achieved in 60%. The success was lower in patients over the age of 60 and in those with duration of diabetes >10 years. Subsequent therapy for 3–15 months was free of side effects save for hematomas at the site of injection in 12.5%.

In another report comparing different vasoactive agents in diabetic and nondiabetic men, 244 men were treated with three different agents for up to 45 months. No difference was found with satisfaction and freedom from side effects, but men with diabetes had a shorter duration of erection.

A new formulation, alprostadil sterile powder, was specifically formulated for treatment of ED, and its effects were examined prospectively in 683 patients, 104 of whom were diabetic (49 type 1, 55 type 2). People with sickle cell disease, endocrine ED, penile deformities, and uncontrolled diabetes and hypertension were excluded from the study. The starting dose of alprostadil was 5–10 g with escalating doses until an erection was obtained. That dose was then used an average of 1.5 times a week for 6 months. The patient and the sexual partner recorded primary efficacy variables.

Men with diabetes required higher mean doses (21 vs. 17 g), fewer of them self-injected in the home-use phase, and fewer of them completed the study. However, in those who actually administered the drug, the overall response rate, duration of erection, and erectile response per injection was similar to nondiabetic subjects. Ninety-one percent had satisfactory intercourse at least once during the home phase, as did 88% of the partners. There were no differences in the rate in type 1 versus type 2 diabetes.

A significant number of diabetic subjects dropped out of the study because of intercurrent medical events, in addition to a higher frequency of penile pain (65 vs. 48%), hematomas (8 vs. 4%), and fibrosis (5 vs. 2%). Intensity of pain was higher in diabetic subjects. Priapism and prolonged erections lasting >4–6 hours were more frequent among those with diabetes. The researchers concluded that diabetic men were less responsive to alprostadil and experienced a higher frequency of adverse events compared with nondiabetic subjects.

Of the 579 nondiabetic men, 69% completed the 6-month study, and 87% reported that their sexual activity was satisfactory. However, for the 31% who did not complete the study, pain or lack of efficacy were prominent complaints. Of the 683 men tested, 50% had some pain.

The commonly used mixture is alprostadil, papaverine, and phentolamine, referred to as trimix. Although urologists have used these drugs for a decade or more, the FDA only approved alprostadil for this use in 1995. While pain may be more frequent with alprostadil than with the mixture, the incidence of prolonged erection (5%) and priapism (1%) are rare and make it the drug of choice.

Recent attempts to solve this problem with the introduction of a medicated urethral system whereby alprostadil is administered in a topical gel via the urethra are discussed below. As successful as intracavernous therapy has become, up to 50% of men discontinue eventually because of pain, loss of effect, or lack of interest.

Caveats to self-injection therapy. Ideal candidates for self-injection therapy are lean individuals with recent onset ED who seek help, in whom the cause is predominantly neuropathic, who are not on anticoagulants, and in whom the relationship with the significant other is such that no extra burden is placed on the relationship by the need to introduce the new form of foreplay. Both parties must be agreeable before embarking on this pursuit. It is also useful during the physical exam to pinch the side of the penis to let the patient know what to expect when the injection is given.

Of the three main agents (papaverine, phentolamine, and prostaglandin), priapism is most frequently seen with papaverine. Treatment of priapism with adrenergic agents can be hazardous in those receiving monoamine oxidase inhibitors. Use of penile vasodilators can also be problematic in those patients who cannot tolerate transient hypotension, have severe psychiatric disease, have poor manual dexterity, have poor vision, or are on anticoagulants. Liver functions should also be checked in those on papaverine.

Prostaglandin can be used to decrease side effects such as pain, corporeal fibrosis, fibrotic nodules, hypotension, and priapism. These drugs are contraindicated in people with gross obesity and in those taking anticoagulants.

The skin should be cleansed properly and the index and middle fingers should occlude the venous drainage by squeezing in an inverted V at the base of the penis. The injection is then given through a 28–30 gauge insulin-type needle laterally to avoid injection into the dorsal vein or the urethra. Massage of the penis enhances action of the erection.

The first injection is always given in the consulting room, which allows for dose titration if necessary and for treatment of any complications, such as prolonged erection or priapism. Patients can then be instructed about dose escalation and the use of enhancing procedures. Application of a constriction band to the base of the penis can enhance the quality of the erection, and some people require provocative visual images to elicit an appropriate response. The quality of the erection is almost always better in the home environment than in the inhibiting circumstances found in most ambulatory care settings.

Patients must be warned to call immediately if an erection persists for longer than 4 hours. Patients should also receive written instructions on what to do under those circumstances, so that another doctor can administer the treatment if necessary. Simple measures, such as immersion in a cold bath or exercise, may be effective. If these fail, decompression can be achieved by injecting an alpha-adrenergic agonist, such as metaraminol or phenylephrine. If this fails, withdrawal of 20–70 ml of blood from the corpus cavernosum is carried out. Oral treatment with phenylpropanolamine or pseudoephedrine can be used.

The Medicated Urethral System for Erection (MUSE) Study Group recently reported the results of a randomized, prospective, double-blind, placebo-controlled trial of a novel delivery system designed to administer alprostadil via the urethral orifice to the corpora cavernosa.

Alprostadil was delivered transurethrally to 1,511 men, 27–88 years of age, who had chronic ED of various causes. The men were first tested in the clinic with up to four doses of the drug (125, 250, 500, and 1,000 g). Those who had sufficient responses were assigned to treatment with either the effective dose of the drug or placebo for 3 months at home.

During the clinic testing, 996 (65.9%) had erections sufficient for intercourse. Of these, 961 reported the results of at least one home treatment. Of the 461 treated with alprostadil, 299 (64.9%) had intercourse successfully at least once, as compared with 93 of 500 who received placebo (18.6%, P < 0.001). On average, 7 of 10 alprostadil treatments were followed by intercourse in the responsive men.

Apparently the results were not dependent on cause or age, and people with diabetes fared no worse than others in the trial. No patients had priapism, unlike with injection therapy, although 8% had pain, and 3.3% had hypotension, neither of which was sufficient to discontinue treatment.

While this was a highly effective study, only 20% of the patients had diabetes, and the cause of the ED in these patients was not defined.

In practice, many patients have found the treatment to be messy and inconvenient. Patients have also had considerable difficulty in introducing the pellet into the urethra if diabetic cheiroarthropathy or neuropathy of the small muscles of the hand was present. Many of our patients have elected to use injections rather than MUSE, which they consider to be bad NEWS. The introduction of sildenafil might obviate the need for this cumbersome form of intervention; however, alprostadil acts through a different pathway. The proportion of patients responding to sildenafil remains to be established.

Sildenafil is a new medicine recently released by Pfizer, Inc., as a treatment for erectile dysfunction. It works by boosting the levels of cGMP in the corpora cavernosa of the penis by inactivating the PDE that hydrolyzes cGMP to 5’GMP.

The formation of cGMP is an essential ingredient in the response to NO, in turn dependent upon the stimulation of endothelial production by neuronal release of acetylcholine (Figure 3). Specificity for cGMP, which is ubiquitous, is achieved by a narrow targeting of the drug to the type 5 isoenzyme of PDE unique to the corpora cavernosa.

Although type 2 and 3 isoenzymes are also found there, the concentrations are relatively low and of little consequence. More importantly, the type 5 isoenzyme is not found in significant concentrations elsewhere. The drug does, however, interact with the type 6 isoenzyme responsible for color vision in the retina.

Increased levels of cGMP mobilize cellular Ca++, facilitating relaxation of the smooth muscle of the helicene vessels and within the corpora, and thereby promoting an erectile response to erotic stimulation. Thus, sildenafil use can involve the sexual partners.

The agent takes about 1 hour to reach maximum effect and so must be taken about 1 hour before anticipated sexual activity. The half-life appears to be about 4 hours, but the duration of erection is limited, unlike that occurring with injection therapy. This militates against the likelihood of developing priapism. Another possible advantage is the fact that it is not a primary stimulator of erection, but rather only a facilitator or enhancer of the erection normally produced by erotic stimulation. Thus, in contrast with the rigid and semirigid penile prostheses, this form of therapy is unlikely to cause embarrassment and, contrasting with alprostadil, requires "normal" foreplay.

In a 24-week, dose-escalation, double-blind, placebo-controlled trial of 532 men with ED of undefined organic cause, 65, 79, and 88% of patients who received sildenafil in respective doses of 25, 50, and 100 mg, reported improved erections, compared with 39% of patients on placebo. Dose-response relationships were found for frequency of penetration, maintenance of hardness, duration of erection, and the number of times satisfactory intercourse was achieved. However, there was no effect on the number of attempts at sexual intercourse. Patients on the drug also reported increased enjoyment of sexual intercourse and increased satisfaction with sex life.

In a separate study, 329 men were treated with sildenafil or placebo with dose escalation to 100 mg based on efficacy and tolerance. Sixty-nine percent of all attempts at intercourse were successful on sildenafil compared with 22% on placebo. The mean number of successful attempts per month was 5.9 for sildenafil and 1.5 for placebo. The most commonly encountered side effects were headache, flushing, dyspepsia, and visual disturbances.

In a small, double-blind, placebo-controlled study involving 21 diabetic patients with a mean age of 50 years (29–66 years) and median duration of ED of 3 years, sildenafil improved the quality of erection in 48% on 25 mg and 52% on 50 mg, compared with the placebo response of 10%.

The major side effects reported were headache, diarrhea, and muscular ache. The authors from Pfizer Central in the United Kingdom concluded that sildenafil is well-tolerated and efficacious in diabetic patients.

Sildenafil was launched on March 27, 1998. In the first 2 months, 1 million prescriptions have been filled.

Caveats in people with diabetes include the coexistence of hypertension in two-thirds of patients, dyslipidemia in half, and coronary artery disease (CAD), and the potential for significant drug interactions. The upside of an orally administered, safe, and effective drug for ED is that many more men with ED will seek help and enjoy the nearly miraculous restoration of sexual function reported by those in the trials.

The downside is the wanton prescription of the drug without proper evaluation. The drug has serious potential for interaction with nitrites and calcium channel blockers and could cause severe hypotension in people taking these drugs. ED in diabetes cosegregates with hypertension, dyslipidemia, and CAD. There is a real risk of coital coronaries (diabetic patients may die smiling!) There has been a tremendous hype in the media since its release and the possibility of abuse, price wars, and street trading is real. Insurers are hesitant to pay for the drug in unlimited amounts, and the long-term safety and efficacy need to be established.

We are still in need of a study on patients with diabetes who have a variety of causes of ED. In these patients, there might be reason to believe that any attempt to enhance the formation or reduce the breakdown of corporal cGMP by PDIE5 might be thwarted by the impaired endothelial and neural function. This is known to occur in the diabetic penis and may not be remediable by preserving the formation of cGMP that nonetheless may be severely compromised.

Vacuum devices
When pharmacological treatment fails, vacuum tumescence devices will work regardless of the cause of ED.

Erection is induced by negative pressure drawing blood into the corpora and retaining the blood in the penis by application of a constriction band to the base of the penis, which obstructs venous drainage. The apparatus consists of a plastic cylinder to which a vacuum pump is attached. The constriction band is made of soft transparent silicone rubber and is rolled over the cylinder and left in place at the base of the penis during intercourse.

Care must be taken to place the band at the base of the penis so as to avoid a constriction band and a rigid penis pivoting around a lax base. The band should not be left in place for more than 30 minutes.

Some complain that the penis feels cold or that ejaculation causes pain. Most find the technique highly acceptable, especially those who have tried injection therapy without success. Partners can often be enticed into treating the application of the device as a form of sexual play before intercourse, and successful couples invariably have learned to add this element to their sexual relations.

Surgery for ED
Penile prostheses
A variety of prosthetic devices have been used for those who fail medical therapy. There are three forms: semirigid, malleable, and inflatable.

The semirigid or malleable rods are implanted into the corpora cavernosa. Unfortunately, this causes the penis to be constantly erect, which often leads to embarrassment when wearing shorts or bathing suits. A possible solution is the use of an inflatable prosthesis, where the pump is housed in the scrotal sac, and the inflatable portion is placed in the corpora.

These are surgical procedures requiring general anesthesia. The main problems are mechanical failure, infection, and erosions. Men with diabetes, spinal cord injuries, or urinary tract infections have an increased risk of prosthesis-associated infection, which in most in-stances requires removal of the device.

Before the advent of injection therapy, we were enthusiastic about the use of penile prostheses. But with vacuum therapy and penile injection now freely available, we almost never insert a prosthesis.

While the great majority of patients with diabetes have ED secondary to autonomic neuropathy, some have severely compromised blood flow. These patients can usually be identified because of their generalized vascular disease and the claudication in their thighs and buttocks, which testifies to vascular insufficiency. The tests necessary to establish the diagnosis have been incompletely validated. Therefore, it is difficult to select patients who can be predicted to have a good outcome.

Dynamic infusion pharmaco-cavernosometry and cavernosonography (DICC), duplex ultrasonography, and possibly arteriography may be needed and are best left to experts in those procedures. If vascular insufficiency is documented, the inferior artery can be anastomosed to the dorsal artery of the penis or the deep cavernosal artery. When venous incompetence prevails, venous leakage can be treated by ligation of the deep dorsal vein of the penis and circumflex arteries.

These procedures are not all that successful, and the severity and extent of vascular disease in people with diabetes renders revascularizing procedures even less successful. The only type of person we would now consider for revascularization would be an individual who is young, with recent onset diabetes, in which it is likely that fairly localized insufficiency is at the root of the ED. In general, older patients with diabetes have diffuse extensive vascular disease in the pelvis with marked calcification of the pudendal vessels, making them poor candidates for an arterial anastomosis.

What is the role of nonadrenergic, noncholinergic modulators of erection, and what will be the role of the endothelial-derived modulators of corporal blood flow? Table 5 lists the various candidates, but at present, their functions are not clear. It is apparent, however, that new agents that can be taken orally and that can affect the production of these modulators by the endothelium are being tested in phase 2 studies, raising interesting possibilities for new therapeutic approaches in the future.

What is the potential for reversing neuropathy, the major cause of ED in patients with diabetes? There are exciting developments on the horizon for the use of aldose reductase inhibitors, ascorbyl dihomogamma-linolenic acid, aminoguanidine, and various growth factors. The agents may overcome the impact of metabolic abnormalities that have not been prevented by glycemic control and show promise for reversing established neuropathy.

Possibly most exciting is the development of various growth factors, for example, nerve growth factor, that are capable of causing regeneration of autonomic nerves, thereby potentially restoring the normal physiological regulation of erection.

Finally, it seems that a component of diabetic neuropathy is related to autoimmune destruction of nerve fibers. New therapies aimed at treating the autoimmunity may prove helpful in this regard.


Vinik AI, Richardson DW: Erectile dysfunction in diabetes. Clinical Diabetes 14:111-20, 1996.

Vinik AI, Richardson DW: Erectile dysfunction in diabetes. Diabetes Reviews, 1998. In press.

Vinik AI, Suwanwalaikorn S: Autonomic neuropathy. In Current Therapy of Diabetes Mellitus. DeFronzo R, Ed. St. Louis, Mo., Mosby-Year Book, 1997, p. 165-76.

Feldman HA, Goldstein I, Hatzichristou DG, Krane RJ, McKinlay JB: Impotence and its medical and psychosocial correlates: results of the Massachusetts Male Aging Study. J Urol 151:54-61, 1994.

Goldstein I, Lue TF, Padma-Nathan H, Rosen RC, Steers WD, Wicker PA for the Sildenafil Study Group: Oral sildenafil in the treatment of erectile dysfunction. N Engl J Med 338:1397-1404, 1998.

Linet OI, Ogrinc FG: Efficacy and safety of intracavernosal alprostadil in men with erectile dysfunction. N Engl J Med 334:873-77, 1996.

Padma-Nathan H, Hellstrom WIG, Kaiser FE, et al: Treatment of men with erectile dysfunction with transurethral alprostadil. N Engl J Med 336:1-7, 1997.

Boolell M, Gepi-Attee S, Gingell JC, Allen MJ: Sildenafil, a novel effective oral therapy for male erectile dysfunction. Br J Urol 78:257-61, 1996.

Burnett AL: Role of nitric oxide in the physiology of erection. Biol Reprod 52:485-89, 1995.

Aaron Vinik, MD, PhD, FCP, FACP, is a professor of internal medicine and anatomy/neurobiology and director of the Diabetes Research Institute at the Eastern Virginia Medical School in Norfolk, Va. Donald Richardson, MD, is an associate professor of internal medicine at the Diabetes Institutes and Center for Endocrinology and Metabolism at the same institution.

Note of Disclosure: Aaron Vinik has received grant support from Alteon and been a consultant for and has received grant support from Genetech, Amgen, Merck, Parke-Davis, and Pfizer. Also, Dr. Vinik has a licensing agreement with Eli Lilly.

Copyright 1998 American Diabetes Association
Updated 8/98
For ADA Related Issues contact

For Technical Issues contact