Clinical Monograph for Drug Formulary Review: Erectile Dysfunction Agents

BACKGROUND: Significant advances in the pharmacologic treatment of erectile dysfunction (ERD) have occurred in recent years, most notably the introduction of sildenafil, the first oral selective phosphodiesterase type 5 (PDE5) inhibitor, in 1998. Sildenafil quickly gained acceptance by the medical community and the public because of its broad efficacy for different types of ERD and its ease of use. Two PDE5 inhibitors, vardenafil and tadalafil, have since joined sildenafil to compete in the ERD market. A review was conducted by the Drug Information Service of a pharmacy benefits manager (PBM) to determine the relative merits and place in therapy of commonly used ERD drugs as part of drug formulary management process and decision making by the Pharmacy and Therapeutics (P&T) committee. OBJECTIVES: To provide readers with a comprehensive clinical monograph on ERD drugs written from a managed care perspective. METHODS: The PBM clinical monograph is designed to provide health plans with an evidence-based review of drugs, therapeutic classes, and disease states with a managed care focus. For each therapeutic class or disease review, an extensive and thorough literature search of MEDLINE is conducted for efficacy, safety, effectiveness, and humanistic and economic data. Drug/disease-state databases (Up-to-Date online, MICROMEDEX), U.S. Food and Drug Administration clinical reviews, key Internet sites, medical/pharmacy-related news sites, clinical guidelines, and AMCP dossiers are also reviewed. Formulary drug monographs prepared by the Drug Information Service of the PBM include a critical analysis and summary of disease-oriented and patient-oriented clinical outcomes, effectiveness, and humanistic data. Additional data considered and included in the formulary review process are clinical attributes, patent expirations/generic competition, off-label or pending indications, and pharmacoeconomic data. RESULTS: Despite the lack of head-to-head comparative studies, all 3 PDE5 inhibitors appear to have equivalent efficacy in the treatment of general ERD and ERD associated with diabetes and post-prostatectomy. Sildenafil has additional efficacy data in the management of ERD associated with spinal cord injury and antidepressant medications. Tadalafil has the longest duration of action (up to 36 hours); this feature can be both beneficial (greater sexual spontaneity) or possibly detrimental (greater exposure to drug, delayed adverse events). All 3 PDE5 inhibitors appear to be generally well tolerated and have similar contraindications and warnings. However, vardenafil is the only PDE5 inhibitor with a cardiac conduction precaution. Alprostadil products are recommended in current ERD guidelines as second-line therapy for those who have not responded or cannot take the oral PDE5 inhibitors. Overall, higher clinical efficacy rates are achieved with intracavernous than with transurethral administration. CONCLUSIONS: A large amount of clinical efficacy and safety data has been published since the market launch of sildenafil in 1998. Sildenafil has the greatest body of efficacy and safety evidence. No comparative studies have been conducted with any of the PDE5 inhibitors. Differences in study populations, primary end points, and measurement tools make comparisons difficult. However, all PDE5 inhibitors appear to be roughly equivalent in efficacy, with minor differences in adverse event profiles. Until more comparative data are available, economic considerations will be a significant factor in choosing ERD products for formulary inclusion.


ss I. Introduction
Erectile dysfunction (ERD) has been defined as the persistent (lasting at least 6 months) inability to attain and maintain erection sufficient to permit satisfactory sexual performance. Although ERD is not a life-threatening disorder, it has a profound impact on the quality of life of those who suffer from it. ERD increases progressively with age, but it is not an inevitable consequence of aging. Other age-related conditions may increase the risk of developing ERD.
Based on the Massachusetts Male Aging Study, the probability of ERD of any degree is 40% among 40-year-old men and 70% among 70-year-old men. 1,2 Many diseases-and many medications-may lead to erectile dysfunction. Therefore, an individual evaluation and identification of the underlying causes as well as a reduction in polypharmacy and a substitution of medications should be some of the first approaches in the management of ERD. 3 Significant advances in the pharmacologic treatment of ERD have occurred in recent years, most notably the introduction in 1998 of sildenafil, the first oral selective phosphodiesterase type 5 (PDE5) inhibitor. Sildenafil quickly gained acceptance by the medical community and the public because of its broad efficacy for different types of ERD and its ease of use. Recent guidelines published by the European Association of Urology and the American Association of Clinical Endocrinologists include sildenafil as first-line pharmacologic therapy in the treatment of ERD when nonspecific therapy is appropriate. 3,4 ABSTRACT BACKGROUND: Significant advances in the pharmacologic treatment of erectile dysfunction (ERD) have occurred in recent years, most notably the introduction of sildenafil, the first oral selective phosphodiesterase type 5 (PDE5) inhibitor, in 1998. Sildenafil quickly gained acceptance by the medical community and the public because of its broad efficacy for different types of ERD and its ease of use. Two PDE5 inhibitors, vardenafil and tadalafil, have since joined sildenafil to compete in the ERD market. A review was conducted by the Drug Information Service of a pharmacy benefits manager (PBM) to determine the relative merits and place in therapy of commonly used ERD drugs as part of drug formulary management process and decision making by the Pharmacy & Therapeutics (P&T) committee. OBJECTIVE: To provide readers with a comprehensive clinical monograph on ERD drugs written from a managed care perspective.
METHODS: The PBM clinical monograph is designed to provide health plans with an evidence-based review of drugs, therapeutic classes, and disease states with a managed care focus. For each therapeutic class or disease review, an extensive and thorough literature search of MEDLINE is conducted for efficacy, safety, effectiveness, and humanistic and economic data. Drug/disease-state databases (UptoDate online, MICROMEDEX), U.S. Food and Drug Administration clinical reviews, key Internet sites, medical/pharmacy-related news sites, clinical guidelines, and AMCP dossiers are also reviewed. Formulary drug monographs prepared by the Drug Information Service of the PBM include a critical analysis and summary of disease-oriented and patient-oriented clinical outcomes, effectiveness, and humanistic data. Additional data considered and included in the formulary review process are clinical attributes, patent expirations/generic competition, off-label or pending indications, and pharmacoeconomic data.
RESULTS: Despite the lack of head-to-head comparative studies, all 3 PDE5 inhibitors appear to have equivalent efficacy in the treatment of general ERD and ERD associated with diabetes and postprostatectomy. Sildenafil has additional efficacy data in the management of ERD associated with spinal cord injury and antidepressant medications. Tadalafil has the longest duration of action (up to 36 hours); this feature can be both beneficial (greater sexual spontaneity) or possibly detrimental (greater exposure to drug, delayed adverse events). All 3 PDE5 inhibitors appear to be generally well tolerated and have similar contraindications and warnings. However, vardenafil is the only PDE5 inhibitor with a cardiac conduction precaution. Alprostadil products are recommended in current ERD guidelines as second-line therapy for those who have not responded or cannot take the oral PDE5 inhibitors. Overall, higher clinical efficacy rates are achieved with intracavernous than with transurethral administration.
CONCLUSION: A large amount of clinical efficacy and safety data has been published since the market launch of sildenafil in 1998. Sildenafil has the greatest body of efficacy and safety evidence. No comparative studies have been conducted with any of the PDE5 inhibitors. Differences in study populations, primary end points, and measurement tools make comparisons difficult. However, all PDE5 inhibitors appear to be roughly equivalent in efficacy, with minor differences in adverse event profiles. Until more comparative data are available, economic considerations will be a significant factor in choosing ERD products for formulary inclusion.
Two PDE5 inhibitors, vardenafil and tadalafil, have joined sildenafil to compete in the ERD market. However, PDE5 inhibitors do not work for all patients, and some individuals may have contraindications that preclude their use. Other firstline options include the use of vacuum devices or investigational oral drugs such as oral yohimbine, trazodone, phentolamine, and, in Europe, sublingual apomorphine. Efficacy data is sparse and conflicting for the off-label use of trazodone, yohimbine, and phentolamine in the treatment of ERD. 4 U.S. Food and Drug Administration (FDA)-approved agents recommended as second-line alternatives in ERD guidelines include intracavernosal alprostadil therapy (direct delivery of the drug to the erectile chambers) and transurethral alprostadil delivery (direct delivery to the urethra) ( Table 1).
This monograph will present a short overview of the etiology, risk factors, pathophysiology, and diagnosis of ERD. The focus of this monograph will be an evaluation of pharmacology, pharmacodynamics, pharmacokinetics, clinical efficacy, and the safety of the pharmacologic treatments that are approved by the FDA for the management of ERD.
Testosterone injection, oral tablets, gels, and transdermal systems are indicated for the treatment of ERD associated with hypogonadism. The review of testosterone preparations for the treatment of hypogonadism will be the subject of a separate monograph.

ss II. Overview of ERD Pathophysiology
Penile erection depends on one or two main mechanisms: reflex erection or psychogenic erection. It is a hemodynamic event regulated by the relaxation of the arterial and corporal smooth muscle. The penis consists of paired erection chambers (corpora cavernosa) that are filled with erectile tissue (corporal sinusoids) composed of smooth muscles. Relaxation of the smooth muscle of the corpora cavernosa is mediated by the release of acetylcholine by the parasympathetic nerves. Acetylcholine causes the endothelial cells to release a nonandrogenic, noncholinergic carrier of relaxation signal-nitric oxide. Nitric oxide may stimulate guanylate cyclase to produce cyclic guanosine monophosphate (cGMP), therefore causing a relaxation of the trabecular smooth muscle. Penile erection is a result of neurally mediated increased arterial inflow into the corporal bodies and an increased amount of oxygen that stimulate nitric oxide synthesis by cavernosal nerves and endothelium, along with a decrease or cessation of venous outflow. [5][6][7] The corporal smooth muscle is contracted when the penis is flaccid. The contraction is due to the presence of a normally present adrenergic tone. Smooth muscle relaxation occurs with erection. There are a number of other receptors in penile smooth muscle, including those responsive to vasoactive intestinal polypeptide, dopamine, histamine, prostaglandin, and various others. [5][6][7] Etiology and Risk Factors of ERD Vascular disease is the most common etiology of ERD in elderly men. The risk of vascular ERD increases with smoking, hypercholesterolemia, and diabetes. In addition, many diseases, such as diabetes, stroke, and Parkinson's disease, can cause autonomic dysfunction. This can impair the penile arterial vasodilatation, maintaining the vascular constriction, and therefore preventing erection. Furthermore, a number of medications have been associated with ERD. Medications with anticholinergic properties, such as antidepressants, antipsychotics, and antihistamines, block parasympathetic-mediated penile artery vasodilatation and trabecular smooth muscle relaxation. 8 Causes contributing to ERD may be related to a number of disorders, which are listed in Table 2. ERD is clearly a symptom of many conditions, and certain risk factors have been identified, some of which may be preventable. Diabetes mellitus, hypogonadism, hypertension, vascular disease, high cholesterol or low-density lipoprotein cholesterol, alcohol ingestion, depression, lack of sexual knowledge, poor sexual techniques, and many chronic diseases have all been identified as risk factors. In addition, age is a strong indirect risk factor because it may be associated with increased likelihood of direct risk factors. Smoking is another indirect risk factor that may increase the effects of other risk factors, such as hypertension or vascular disease. Knowledge of the risk factors can guide patients to prevention strategies. 5,7,9,10

Diagnosis of ERD
ERD may be associated with several abnormalities of the endocrine, neurological, and vascular system. Thus, an appropriate evaluation of all men with ERD should include a medical and sexual history, physical exam, psychosocial evaluation, and appropriate laboratory studies. 3 Endocrine evaluation includes hemoglobin A1C, a morning serum testosterone, prolactin, luteinizing hormone, and folliclestimulating hormone (FSH) levels. Other tests, such as complete blood count, urinalysis, creatinine, lipid profile, fasting blood sugar, and thyroid function may be indicated to exclude an unrecognized underlying systemic disease. Neurologic causes may be associated with a history of diabetes, spinal injury, or cerebrovascular accident; a detailed medical history will be essential to identify them. In addition, nocturnal penile tumescence testing may be useful when a primary psychogenic ERD is suspected. An erectile response to an intracavernosal injection of pharmacological test dose of a vasodilatory agent, such as papaverine or PGE1, indicates adequate arterial and veno-occlusive function. For patients who favor noninvasive treatments, such as the oral PDE5 inhibitors, pharmacological injection, intraurethral suppository, or vacuum constrictor devices, no further diagnostic tests are necessary. On the other hand, for patients with unsatisfactory response, penile implant surgery or further diagnostic tests may be appropriate. 3 ss III. Pharmacology/Pharmacodynamics FDA-Approved Therapy Alprostadil (Caverject, Edex, and MUSE) Prostaglandin E1 (alprostadil) is one of the prostaglandins, naturally occurring acidic lipids with a variety of pharmacological effects, including vasodilatation, inhibition of platelet aggregation, and stimulation of intestinal and uterine smooth muscle. It acts by relaxing the trabecular smooth muscles of the corpus cavernosum and increasing the diameter of cavernous arteries, and this leads to erection. In animal studies, the degree and duration of cavernous smooth muscle relaxation appears to be dose dependent. [11][12][13]

PDE5 Inhibitors (Sildenafil, Vardenafil, and Tadalafil)
The mechanism of penile erection involves relaxation of the corpus cavernosal smooth muscle. This occurs through release of nitric oxide during sexual stimulation, which results in increased concentrations of cGMP. Sildenafil, vardenafil, and tadalafil are all competitive inhibitors of the type 5 cGMPspecific PDE5 enzyme. [14][15][16] The result is an enhancement of the effect of nitric oxide secondary to a decrease in degradation of cGMP. PDE5 inhibitors have no effect in the absence of sexual stimulation.
There are 11 families of phosphodiesterase isoenzymes that have been identified in mammalian tissue. While PDE1 through 6 have been extensively studied, PDE7 through 11 have been recently discovered, and thus less is known regarding their distribution and function in the human body.
Sildenafil, vardenafil, and tadalafil are all more selective for the PDE5 isoenzyme than for all other PDE isoenzymes. However, degrees of selectivity vary among the agents, depending on the isoenzyme in question. As illustrated in Table 3, sildenafil is 80 times more selective for PDE5 than for PDE1, but greater than 80 times more selective for PDE6, an isoenzyme heavily concentrated in the retina of the eye. 17,18 In contrast, tadalafil is greater than 700 times more selective for PDE5 than for the PDE6 isoenzyme. This selectivity ratio pattern may explain why the side effect of blue-tinged vision or changes in blue-green color discrimination is reported with sildenafil but is  not expected to occur with tadalafil use. On the other hand, tadalafil is only 14 times more selective for the PDE5 than the PDE11 isoenzyme than sildenafil and vardenafil, which have much higher selectivity ratios. The low selectivity ratio of tadalafil for PDE11, an isoenzyme heavily concentrated in the testes and skeletal muscle, led investigators to conduct safety studies to ascertain what effect tadalafil would have on spermatogenesis. However, 6-month, daily-dosing, placebocontrolled studies with 10 and 20 mg/day of tadalafil produced no clinically relevant effect on spermatogenesis as measured by sperm count and sperm morphology and motility. Additionally, no effect was observed on hormones related to spermatogenesis (luteinizing hormone, FSH, testosterone) with chronic tadalafil use. 19

Hemodynamic Effect
The PDE5 inhibitors all work as vasodilators. Because PDE5 is found in the smooth muscle of the systemic arteries and veins, these agents all have potential to interact with the cardiovascular system. Since many men with ERD also have coexisting hypertension, diabetes, and cardiovascular disease, significant hemodynamic effects from PDE5 inhibitor use could be clinically important. Table 4 summarizes the hemodynamic changes seen with the PDE5 inhibitors in normal healthy volunteers and patients with coronary artery disease. All 3 agents produce minor changes in systolic and diastolic blood pressure, but these changes do not alter response to exercise testing. Careful analysis of population data and vardenafil, sildenafil, and tadalafil clinical data do not show an increase in serious cardiac events associated with PDE5 inhibitor use. 15,[20][21][22][23][24] All PDE5 inhibitors are contraindicated with concomitant administration of nitrates because significant hypotension can result. Sildenafil, vardenafil, and tadalafil are also contraindicated for use with alpha-blockers for the same reason. One exception to this rule is that tadalafil can be safely administered with tamsulosin 0.4 mg daily. 14-16

Effect on Cardiac Conduction
Vardenafil in therapeutic (10 mg) and supratherapeutic (80 mg) doses produced increases in the QT interval similar to that of 400 mg of moxafloxicin. While the clinical impact of these changes is unknown, the coadministration of vardenafil with Class IA and Class III antiarrhythmic medications should be avoided. Patients with congenital QT prolongation should also avoid vardenafil use. 15

ss IV. Pharmacokinetics
The pharmacokinetics of the ERD agents are summarized in Table 5. Sildenafil and vardenafil reach peak plasma concentrations at about 1 hour after administration; tadalafil reaches peak concentrations at 2 hours. Although not well studied, efficacy data for all 3 PDE5 inhibitors indicate that onset of action is earlier (17 to 40 minutes) than when peak serum concentrations are reached. [25][26][27][28] Although all 3 PDE5 inhibitors vie for the claim of earliest onset, only well-designed comparative studies will help answer the question of which agent is the fastest acting. There are no studies that directly compare the onset, duration, or overall efficacy of the PDE5 inhibitors. Unlike sildenafil and vardenafil, peak serum concentrations of tadalafil are not affected by a high-fat meal. 14-16 All 3 PDE5 inhibitors undergo extensive hepatic metabolism and require some dosage adjustment with hepatic dysfunction. The most striking difference between tadalafil, vardenafil, and sildenafil is the long half-life of tadalafil (17.5 hours). This long halflife translates into a prolonged duration of action for tadalafil (up to 36 hours), earning it the name of "le weekend" drug in France.  The bioavailability of intracavernous administration of alprostadil has not been studied. The absorption with transurethral administration of alprostadil appears to be biphasic, with 80% of the dose being absorbed within 10 minutes. [11][12][13] The onset of action after intracavernous injection is within 2 to 5 minutes of administration. The onset of action after transurethral administration is slower, at about 5 to 10 minutes. Following intracavernous and transurethral administration of alprostadil, the drug is either metabolized locally or cleared from the penis into the systemic circulation and then metabolized by the lungs. The mean peripheral plasma concentrations are not significantly greater than baseline levels of endogenous alprostadil. The metabolites are excreted primarily by the kidney. Within 24 hours following administration, about 90% of the dose was excreted in urine, and the remaining 10% was excreted in feces. The effect of age, gender, and renal or hepatic failure on the pharmacokinetics of alprostadil has not been evaluated. However, patients with pulmonary disease may have reduced ability to clear the drug because of pulmonary first-pass metabolism of prostaglandin E1. [11][12][13]29 ss V. Clinical Trials Table Organization The clinical efficacy of the PDE5 inhibitors and intracavernous and transurethral alprostadil are summarized in Tables 6 through 9. The tables are organized in the following manner: • Pivotal placebo-controlled trials in the general ERD population (

Comments:
• Additional analyses were conducted to assess efficacy in the following subgroups: age 65 years, Asians, African Americans, severity of ERD, HTN, vascular disease, diabetes, depression, or psychogenic ERD, history of radical prostatectomy, and spinal cord disorders. While degree of efficacy varied among subgroups, all sildenafil participants had significantly higher efficacy measures than the respective placebo groups. • 48% of the men on sildenafil had at least 1 adverse event compared with 36% of men on placebo (RRI 1.4, CI 1.3-1.6). • Most common events were headache (11%), flushing (12%), dyspepsia (5%), and visual disturbances (3%). • Differences in angina or cardiac chest pain did not reach statistical significance nor did rates of myocardial infarction or death. Erectile Dysfunction Placebo-Controlled Studies: General Population (continued)

TABLE 6
Comments: • Compared with placebo, tadalafil significantly improved all efficacy outcomes.
• High placebo response rate reflects higher proportion of subjects with mild ERD at study entry.

Results:
No response to placebo by either clinical or RigiScan evaluation. Men responding with full erection ranged from roughly 20% (2.5 mcg dose) to 50% (20 mcg dose) by either clinical or RigiScan assessment.

Comments:
• Prolonged erections occurred in 5 men; in 2 men, the erections lasted 4 hours or more. • Mean duration of erection was related to dose. • Penile pain was reported by 23% of the men on intracavernous alprostadil.

Drug regimen:
In-office dose titration phase: Study 1 (N = 85) Edex (intracavernous alprostadil) 1 mcg-20 mcg or placebo Home phase: Study 2 (N = 158) Intracavernous alprostadil 1 mcg-40 mcg or placebo Study 1 responders continued with the home phase; patients continued on optimal dose for 1 week then crossed over to alternate treatment.    in IIEF score from baseline, normalization of IIEF erectile function domain, mean improvement in erectile function score, and percentage improvement over baseline, to name a few.

Sexual Health Inventory for Men (SHIM)
The SHIM is an abbreviated version of the IIEF questionnaire and was designed to allow a more rapid diagnosis of ERD and assignment of ERD severity. The instrument has 6 questions, with a maximum score of 30. ERD is present if the SHIM score is 21 or less. The SHIM primarily measures erectile function, and it does not address measures of orgasmic function, libido, and satisfaction.

Sexual Encounter Profile (SEP) Diary
Assessments of individual sexual encounters are provided by SEP diaries. The SEP diary is intended to be an immediate-recall diary of encounters. The diaries contain 6 questions for the patient and 4 questions for the partner. SEP questions 2 and 3 are common outcomes measures in efficacy studies, and they are very similar to questions 3 and 4 of the IIEF erectile dysfunction domain. However, the SEP questions are answered yes or no while the IIEF questions are assigned a numerical score.

Global Assessment or Global Efficacy Questions
Global assessment or efficacy questions are often used as secondary outcomes measures. The 2 most common questions are: "Did this treatment improve your erections?" and "Did treatment improve your ability to have sexual intercourse?"

Clinical Efficacy Summary General ERD Population: PDE5 Inhibitors
Sildenafil, vardenafil, and tadalafil significantly improve IIEF erectile function domain scores and improve erection quality as compared with placebo in large, double-blind, randomized, controlled trials in the general ERD population. [31][32][33][34][35] There are several outcomes measurements reported in ERD clinical studies,

Dosage regimen and duration:
In-office dose titration phase (1-14 days) to find optimal intracavernous alprostadil or transurethral alprostadil dose (N = 95); transurethral alprostadil maximum 1,000 mcg (transurethral alprostadil group was also offered option of ACTIS penile ring); intracavernous alprostadil maximum 40 mcg At-home, 3-week treatment phase at optimal transurethral alprostadil or intracavernous alprostadil dose ( N = 68); patients were then crossed over to repeat in office and at home phase with the alternate treatment.
Outcomes measure: IIEF erectile domain scores, physician and patient assessments of erection quality, at least 1 erection sufficient for intercourse, at least 75% successful (75% of all attempts are successful)

Comments:
• Penile pain common to both intracavernous alprostadil and transurethral alprostadil (20%-34%) • Application site reaction more common with transurethral alprostadil 10%-15% than with intracavernous alprostadil (2%-4%) • Intracavernous alprostadil had a 2%-3% rate of prolonged erections as compared with none in the transurethral alprostadil group. but perhaps the most meaningful improvement to the patient is the rate of successful intercourse. In a meta-analysis of 14 randomized, placebo-controlled, flexible-dose studies, the subjects on sildenafil 25 mg to 100 mg had a successful intercourse rate of 57% as compared with a rate of 21% with placebo. 31 Combined data from 2 fixed-dose sildenafil studies showed a successful intercourse rate of 43%, 50%, and 51% for the 25 mg, 50 mg, and 100 mg dose, respectively, as compared with the placebo group, which had rates of 14% to 17%. 31 In general, higher sildenafil doses were associated with higher efficacy rates. Also included in the meta-analysis were additional analyses examining efficacy in subgroups stratified by age, race, ERD baseline severity, and ERD etiology. Sildenafil was as efficacious in the Asian and African American subjects as in  whites, who comprise the majority of subjects in ERD studies. While the rate of successful intercourse varied depending on age, ERD severity, and ERD etiology, sildenafil use resulted in significantly greater rates for each subgroup as compared with placebo. 31 In one large, randomized, fixed-dose study, vardenafil, at doses ranging from 5 mg to 20 mg, was able to produce a significantly greater rate of erections adequate for intercourse-50% to 65%-compared with a placebo rate of 32%, which is higher than the reported placebo average of 20%. The high placebo rate seen in this study is intriguing because 30% to 45% of subjects were classified by investigators as having severe ERD. An increase in efficacy was seen with increasing vardenafil dose. 19 Vardenafil significantly improved IIEF erectile function domain scores as compared with placebo regardless of patient age, ERD etiology, or baseline ERD severity. 33 In an integrated analysis of 5 multicenter, double-blind, randomized, fixed-dose studies, tadalafil 2.5 mg, 5 mg, 10 mg, and 20 mg resulted in significantly higher rates of successful intercourse, 36%, 42%, 61%, and 75%, respectively, compared with a placebo rate of 32%. 34 Another multicenter, doubleblind, randomized, fixed-dose study compared the duration of efficacy of 20 mg tadalafil with placebo. 35 At 24 hours postadministration, a 53% rate of successful intercourse attempts was reported in the tadalafil group compared with a 29% rate in the placebo group. Tadalafil remained significantly more efficacious than placebo at 36 hours postdose, with a rate of 59% compared with 28%. 35 The results of this study confirm the long duration of tadalafil, which would be anticipated from its prolonged half-life of 17 hours.
No comparative studies have been done to assess relative efficacy of any one PDE5 inhibitor to another. Until large comparative studies prove otherwise, the efficacy of these products seems roughly equivalent; however, direct comparisons of efficacy and safety should not be made, given the many variables present in populations studied and outcomes measures used.

General ERD Population: Alprostadil
Intracavernous and transurethral administration of alprostadil, while not usually considered first-line therapy, is also effective in the management of ERD in the general population. In alprostadil studies, efficacy is most often measured by physician and patient assessment of erection quality. In one large, multicenter, randomized, fixed-dose study, intracavernous administration of alprostadil at doses of 2.5 mcg, 5 mcg, 10 mcg, and 20 mcg resulted in 20%, 30%, 35%, and 50%, respectively, of men achieving full erections. 36 The mean duration of erection was 37 minutes, and the duration was related to dose. Five men had prolonged erections; in 2 men, the erections lasted 4 hours or more. Penile pain was reported by 23% of intracavernous alprostadil subjects. In a 6-month self-injection extension of the study, the intracavernous alprostadil responders reported being able to have intercourse after the injections 94% of the time. 36 In another placebocontrolled crossover study, intracavernous alprostadil 1 mcg to 40 mcg resulted in 73% to 74% of erections deemed adequate for intercourse (patient assessment) as compared with rates of 7% to 13% for placebo. 37 The median duration of erection was 59 minutes, and prolonged erections lasting 4 to 6 hours were noted in 3% of subjects taking intracavernous alprostadil. The average intracavernous alprostadil dose was not reported in this study. Penile pain and bleeding were other common adverse events. 37 Transurethral alprostadil was significantly more effective than placebo in 2 double-blind, placebo-controlled studies. 38,39 In these studies, using transuretharal alprostadil doses ranging from 125 mcg to 1,000 mcg, the rates of erections deemed adequate for intercourse were 49% to 66%. In the at-home phase of one of the studies, transuretharal alprostadil resulted in a successful intercourse rate of 65% compared with a placebo rate of 19%. Incidence of penile pain ranged from 9% to 19%, hypotension was 3%, and there were no reports of priapism or prolonged erections. 38,39 Two open-label studies compared the efficacy of transurethral alprostadil versus intracavernous alprostadil. In one study, the intracavernous alprostadil product was an extemporaneous preparation 53 ; in the other, the Edex preparation was used. 54 In both studies, intracavernous injections of alprostadil resulted in significantly higher erectile assessment scores or IIEF erectile function domain scores as compared with transurethral alprostadil. In one study, transurethral alprostadil was better tolerated with a lower discontinuation rate due to penile pain 53 ; however, the other study reported similar rates of penile pain and a marked patient preference for injection over transurethral therapy. 54

General ERD Population: Failures on Previous ERD Therapy
One open-label, multicenter study reported that intracavernous alprostadil, in doses up to 40 mcg, was effective in failures with sildenafil therapy. In this study, sildenafil failures had a score of 1.2 or less on the IIEF erectile domain questions 3 and 4. A score of 1 means that sildenafil was almost never or never effective. Use of intracavernous alprostadil resulted in the IIEF scores improving by 2.75 to 2.63 points for 85% to 90% of patients. Penile pain was present in 30% of all intracavernous alprostadil subjects. 56 One open-label, multicenter study examined the efficacy of vardenafil 5 mg to 20 mg in the treatment of ERD in 134 patients determined to be unresponsive to sildenafil. Unresponsiveness was defined as failure with sildenafil on at least 4 out of 6 attempts, with at least one of those attempts at the 100 mg dosage level. Sildenafil failures were randomized to receive either vardenafil (N = 231) or placebo (N = 226) for a treatment period of 12 weeks. Vardenafil use resulted in significantly higher IIEF erectile domain scores than placebo Clinical Monograph for Drug Formulary Review: Erectile Dysfunction Agents and higher rates of maintenance of erection sufficient for intercourse (46% vardenafil versus 16% placebo; P < .001). Overall, 62% of vardenafil subjects stated that their erections were improved compared with 15% of those in the placebo group. 57

Special Populations
Diabetes Several double-blind, placebo-controlled studies have been performed to evaluate the efficacy of sildenafil, vardenafil, and tadalafil in the management of ERD associated with type 1 and type 2 diabetes. [40][41][42][43][44] No direct comparative studies have been performed to assess relative efficacy of one PDE5 inhibitor to another. However, well-designed studies have reported the following rates of successful intercourse: sildenafil 48% versus placebo 12%; vardenafil 49% to 54% versus 23%; tadalafil 28% to 29% versus 1.9%. 44 The lower success rate seen with tadalafil may be due to the high percentage (72%) of patients with severe ERD enrolled in the study.

Postprostatectomy
As with diabetes, several clinical studies have assessed the efficacy of all of the currently available PDE5 inhibitors in the management of ERD postprostatectomy. In this patient population, response to treatment is dependent on subject age, baseline ERD severity, and the type of prostatectomy surgery. In general, bilateral nerve-sparing surgery is associated with the best chance for response with non-nerve-sparing procedures having the lowest response to therapy. However all PDE5 inhibitors are potentially effective in the management of postprostatectomy ERD. [45][46][47][48][49]

Post-Spinal-Cord Injury
Of the PDE5 inhibitors, only sildenafil has been studied in the management of ERD resulting from spinal cord injury. This patient population differs not only in the etiology of ERD but also in age since the average spinal cord injury patient in clinical studies is much younger (38 years) as compared with the ERD patient in the general population (56 years). In one randomized, placebo-controlled crossover study in 178 spinal cord injury patients, doses of sildenafil 50 mg to 100 mg resulted in an intercourse success rate of 55% versus 0% for placebo. Thus, success rates for sildenafil in ERD secondary to spinal cord injury approach rates seen in subjects with other comorbid conditions. 50

Depression
One double-blind, placebo-controlled study has evaluated the efficacy of sildenafil in the management of ERD in patients with depression. Most patients in this study had a diagnosis of mild or moderate major depression and were not treated with antidepressants. Sildenafil 25 mg to 100 mg or placebo was given for 12 weeks. At the end of the study, significantly more patients on sildenafil than placebo (73% versus 14%) had a treatment response as defined by IIEF erectile function treatment scores and positive responses to 2 global efficacy questions. Successful treatment was also associated with an improvement in Hamilton Depression scores and quality-of-life measures. 51 Sildenafil was more effective than placebo (55% versus 4.4%; P < .001) in improving Clinical Global Impression-Sexual Function scores in a study with 90 patients with ERD secondary to treatment with selective serotonin reuptake inhibitor antidepressants. All patients were in remission from major depression and remained on antidepressants during treatment with sildenafil for 6 weeks. 52

Effectiveness Studies
Overall, in controlled clinical studies, sildenafil has an efficacy rate of roughly 60% in the broad ERD population. 31 However, in the real-world setting, refill rates for sildenafil are not as high as would be expected. Of patients tracked for 1 year, only 52% filled a second prescription during that 12-month period and 31% filled greater than 7 prescriptions. 58 In another study, patients in a clinic were followed for 2 years to evaluate their response to sildenafil. 59 Two surveys were conducted. The first survey went to 200 men who had recently been given a prescription for sildenafil. Of these 200 men, only 151 (75%) actually tried the drug. Of those who tried the drug, an overall success rate of 74% was reported. The most common doses used were 50 mg (n = 88) and 100 mg (n = 61). While 38% of patients reported side effects, none discontinued therapy from drug intolerance. Two years later, a second survey was sent out; only 82 patients participated. Of those patients, 17% discontinued because of loss of efficacy and 20% needed to increase their dose by 50 mg. There was no correlation between frequency of use and the need to increase the dose. While the authors concluded that tachyphylaxis to sildenafil was responsible for study results, it is not clear if this is the case. 59 Other reasons for reduced effect over time could have included psychological factors as well as worsening of underlying comorbid conditions, especially progressive vascular disease or poorly controlled diabetes.
Efficacy results in controlled clinical studies are rarely, if ever, duplicated in the real-world setting, and the experience with ERD is no different. However, McCullough et al. did report on several studies designed to identify and improve success rates with sildenafil therapy. 60 The intensive disease management approach utilized in one of the studies yielded impressive results. Overall, 55% of men not previously successful with sildenafil became successful after intensive reeducation and counseling, which included regular follow-up visits with information as to how to take the drug, titration to maximum dose, and a minimum trial of 8 attempts for efficacy assessment. Controlling risk factors for ERD as recommended in current treatment guidelines also was a successful strategy, although men with only 1 risk factor were more likely to respond to intervention than men with multiple risk factors. 60 ss VI. Adverse Events PDE5 Inhibitors Tadalafil, sildenafil, and vardenafil were well tolerated in clinical studies with headache, flushing, and dyspepsia occurring as the most common adverse events. There are no comparative safety data to compare rates of common adverse events, but based on the rates seen in placebo-controlled studies, there appears to be little difference in safety profiles for these most commonly reported events. Discontinuations secondary to adverse events were low for all 3 PDE5 inhibitors, ranging from 1% to 5%. Changes in color vision, which has been reported with sildenafil use, are less frequent with vardenafil and rarely reported with tadalafil. However, tadalafil does seem to be associated with more reports of myalgia and back pain than vardenafil or sildenafil. The muscle aches and back pain usually occur within 12 to 24 hours after tadalafil administration and resolve within 48 hours. Approximately 0.5% of patients discontinued tadalafil because of back pain or myalgia. [14][15][16] Serious Cardiac Events Cardiac mortality rates in the tadalafil clinical study database (N > 4,000 subjects) are consistent with the expected rate in a male population. Across all studies, the incidence rate of myocardial infarction was 0.43 per 100 patient years in the tadalafil-treated patients compared with 0.6 per 100 patient years in the placebo-treated population, which was also consistent with the incidence rate observed with an age-standardized male population. 61 The cardiac safety of sildenafil has been extensively studied.   Pooled results from 53 clinical studies indicated no difference between the incidence of death or myocardial infarction in men with ERD receiving sildenafil or placebo. 62 In a United Kingdom study, 5,601 patients with ERD showed no evidence of increased risk of myocardial or ischemic heart disease during the first 4.9 months of sildenafil therapy. 63 This low risk is supported by open-label safety data from subjects who have been taking sildenafil for up to 4.5 years. 64 Vardenafil has been shown to prolong the cardiac conduction as evidenced by a prolonged QT interval at therapeutic and supratherapeutic doses (Section VII, Contraindications/ Precautions). 15

Intracavernosal and Transurethral Alprostadil
The type and degree of side effects reported in the 2 intracavernous alprostadil formulations are very similar. [11][12][13] No controlled comparative studies are available that directly compared the adverse event rates of these 2 products. As might be expected from a penile injection, local side effects (ecchymosis, hematoma, edema, pain) are prominent with both. Transurethral alprostadil is also associated with a significant occurrence of penile pain, urethral burning, and bleeding. The 2 comparison studies that compared transurethral alprostadil with intracavernous alprostadil injections had conflicting results regarding penile pain and discontinuations due to adverse events. 53,54 Prolonged erection or, in some cases, priapism, can occur with intracavernous alprostadil and transurethral alprostadil. [11][12][13] Tables 10 and 11 display selected common adverse events as reported in respective product labeling for the currently marketed PDE5 inhibitors [14][15][16] and the alprostadil intracavernosal and transurethral products. [11][12][13] Direct comparisons between adverse events rates cannot be made as the event rates displayed are not derived from comparative studies.

ss VII. Contraindications/Precautions
The contraindications, warnings, and precautions for sildenafil, tadalafil, and vardenafil are extremely similar (Table 12). Of note, vardenafil in therapeutic (10 mg) and supratherapeutic (80 mg) doses produced increases in the QT interval similar to that of 400 mg of moxifloxicin. While the clinical impact of these changes is unknown, the coadministration of vardenafil with Class IA and Class III antiarrhythmic medications should be avoided. Patients with congenital QT prolongation should also avoid vardenafil use. 15 The contraindications, warnings, and precautions for intracavernosal and transurethral products are exactly the same with one exception. Transurethral alprostadil should not be used for sexual intercourse with a woman who is Drug and Food Interactions With PDE5 Inhibitors [14][15][16]   All PDE 5 inhibitors are indicated for the treatment of erectile dysfunction pregnant or could become pregnant, unless the couple uses a condom barrier. This precaution is based on animal data that showed embryotoxic effects when alprostadil was administered as a subcutaneous bolus to pregnant female rats (transurethral alprostadil product information). Table 13 lists the contraindications, warnings, and precautions as stated in intracavernous and transurethral alprostadil product information. [11][12][13] ss VIII. Drug/Food Interactions Drug and food interactions with PDE5 inhibitors are presented in Table 14.

ss IX. Use in Pregnancy/Nursing
Transurethral alprostadil should not be used for sexual intercourse with a woman who is pregnant or could become pregnant, unless the couple uses a condom barrier. 13 Vardenafil, sildenafil, and tadalafil are listed as Pregnancy Category B drugs. While no evidence of fetal or embryonic toxicity was found in animal studies, there are no adequate and well-controlled trials of vardenafil, sildenafil, or tadalafil in pregnant women. [14][15][16] In animal studies, tadalafil and vardenafil were secreted into the milk of lactating rats at concentrations 2.4-fold (tadalafil) and 10-fold (vardenafil) greater than found in the plasma. It is not known if these agents are excreted in human breast milk. There is no information on sildenafil and lactation. [14][15][16] ss X. Indications/Dosing The indications, usual adult dose, and dose for special populations for all FDA-approved ERD drugs are listed in Tables 15  and Table 16. [11][12][13][14][15][16] ss XI. Conclusion All 3 PDE5 inhibitors have significant efficacy in the treatment of general ERD and ERD associated with diabetes and postprostatectomy. Placebo-controlled trials have also shown sildenafil to have efficacy for patients with ERD associated with depression and spinal cord injury.
There are no head-to-head clinical studies comparing the efficacy and safety of sildenafil with vardenafil or tadalafil. Sildenafil has by far the highest number of controlled studies confirming its safety and efficacy and is recommended as first-line ERD therapy when a nonspecific therapy is appropriate. The PDE5 inhibitors differ in their duration of action. Sildenafil and vardenafil seem to have similar duration of action of about The recommended frequency is 3 times weekly with 24-hour periods between doses.

Edex (Intracavernous Alprostadil)
Intracavernous alprostadil is indicated for the treatment of erectile dysfunction due to neurogenic, vasculogenic, psychogenic, or mixed etiology.
The dose of intracavernous alprostadil should be individualized for each patient by careful titration under supervision by the physician. Dosage should be initiated at 2.5 mcg and gradually titrated upward according to response and erection duration. For spinal cord injury patients, lower initial doses of 1.25 mg are recommended. No more than 2 doses during an initial titration should be given in 24 hours.
Dosage range 1 to 40 mcg; mean dose in clinical studies was 21.9 mcg.
The recommended frequency is 3 times weekly with 24 hour periods between doses.

MUSE (Transurethral Alprostadil)
Transurethral alprostadil is indicated for the treatment of erectile dysfunction.
The dose of transurethral alprostadil should be individualized for each patient by careful titration under supervision by the physician. Dosage should be initiated at 125 to 250 mcg and gradually titrated upward in a stepwise manner until the patient achieves an erection adequate for intercourse.
Most men in clinical studies required the 500 or the 1,000 mcg dose to achieve an adequate erection.
The maximum frequency is 2 administrations within a 24-hour period. 4 hours, while tadalafil has a duration of action of up to 36 hours. This prolonged duration of action may be a significant advantage for tadalafil since it could allow for increased sexual spontaneity. However, from a side-effect standpoint, it may not be an advantage to have prolonged levels of tadalafil in the systemic circulation.
Tadalafil, sildenafil, and vardenafil have similar common and nonserious adverse events. Yet, tadalafil does have a higher rate of myalgias and back pain that can take several hours to resolve. Vardenafil and especially tadalafil seem to have less propensity for visual changes. However, vardenafil does produce changes in cardiac conduction at therapeutic doses. This could be especially significant if vardenafil is coadministered with CYP3A4 inhibitors because these drugs interfere with vardenafil metabolism. Injectable or transurethral alprostadil remains recommended second-line therapy if first-line therapy is ineffective or contraindicated. Injectable alprostadil results in a quicker onset and a higher success rate than transurethral alprostadil, but it may also have a higher rate of prolonged erections or priapism. Table 17 contains the clinical summary grid that compares and contrasts effectiveness, efficacy, safety, and clinical attributes of the 6 products currently used for the treatment of ERD. Table 18 lists definitions of some of the outcomes terms used in the clinical summary grid. Table 19 contains the comparative costs for a single dose of the ERD agents discussed in this study.

DISCLOSURES
No outside funding supported this study. The author discloses no potential bias or conflict of interest relating to this article.