Erectile dysfunction (ED) is a highly prevalent disease, as well as a major men ’ s sexual concern. As the proportion of older people in the population increases, it has been estimated that the worldwide prevalence of erectile dysfunction will double from 152 million men in 1995 to 322 million men in 2025. Today, phosphodiesterase (PDE) type 5 inhibitors are considered the first-choice treatment for erectile dysfunction by both physicians and patients. The safety of these drugs is becoming more important since more than 30 million men are treated worldwide with a PDE-5 inhibitor, especially in the primary care setting. This chapter summarizes current knowledge on safety and adverse events of PDE-5 inhibitors, providing clinical guidance on essential topics related to drug interactions and contraindications.

Adverse events and discontinuation rates

The currently available PDE-5 inhibitors (sildenafil, tadalafil, and vardenafil) have been shown to be efficacious and safe treatment options for patients with ED. Being of the same class of drugs, they share common side-effects, although differences have been noticed in the reported frequency and severity of these symptoms; therefore an analysis follows for each of the three.


Sildenafil was the first commercially available PDE-5 inhibitor, and its safety profile has been assessed in numerous studies. The most commonly reported treatment-related adverse events are headache, facial flushing, dyspepsia, dizziness, rhinitis, and abnormal vision. Adverse events are typically transient and mild to moderate, and they are due to the vasodilatory effects of the medication; these include headache, flushing, and nasal congestion. Dyspepsia is probably secondary to relaxation of the smooth muscle of the gastro-esophageal sphincter. Visual disturbances (blurred vision, flashing lights, blue haze, and change in color perception) occur as a result of weak inhibition of PDE-6 in the retina. They are coincident with peak plasma concentrations of sildenafil and are transient and fully reversible. None persisted 6 hours after taking sildenafil and are rarely a reason for discontinuing treatment. In this review including 5 fixed- and 6 flexible-dose trials, placebo patients were more likely to drop out than treatment patients from treatment-related adverse events (2 % for sildenafil vs 2.3 % for placebo, differences not statistically significant).

Long-term safety data from open-label extension studies revealed very low rates of treatment-emergent adverse events. However, this population differed from those of the pooled double-blind studies in that it was self-selected. Therefore, it is likely that the population included mostly men who had previously experienced a good response to sildenafil. Post-marketing case series reported a higher incidence of adverse events, especially for headache (9–39 %), flushing (7–33 %) and abnormal vision (5–11 %). Patients may tolerate sildenafil differently based on existing comorbidities. Ischemic heart disease and hypertension are associated with higher incidence of discontinuation rates caused by adverse events than diabetes (3.6 %, 2.3% , and 1.9 % respectively).

Multiple studies have demonstrated that sildenafil is well tolerated by elderly patients with ED, irrespective of comorbid conditions. Rates of serious adverse events and discontinuation of treatment owing to adverse events are similar between placebo and sildenafil groups. The safety profile of sildena-fil remains unchanged in patients with spinal cord injury or depression treated with serotonin reuptake inhibitors.


The most frequent adverse events reported by patients on tadalafil were headache, flushing, nasopharyngitis, and back pain or myalgia. These adverse events were primarily mild or moderate and generally self-limited over continuous use of the drug. Only few patients discontinued treatment in each of the treatment groups (1.3 % for placebo, 1.6 % for tadalafil 10mg, 3.2 % for tadalafil 20mg), although the difference was significant between tadalafil and placebo (p= 0.026). The long-term safety and tolerability of tadalafil assessed in a 24-month extension trial (all patients started at a dose of tadalafil 10mg but dose could be increased to 20mg or decreased to 5mg). No unexpected adverse event was recorded. Serious adverse events occurred in 8.6 % of patients but no consistent pattern of serious adverse events was assessed as causally associated with tadalafil administration. The discontinuation rate because of adverse events was 6.3 %, while 3.1 % of the patients discontinued as a result of adverse events that were assessed by the investigator as being possibly related to tadalafil.

Visual disturbances are very rare since tadalafil clinically lacks selectivity for PDE-6. Myalgia or back pain is a common adverse event related to tadalafil, in contrast to the other two PDE-5 inhibitors but its pathophysiology is poorly understood. Owing to the fact that tadalafil is also an inhibitor of PDE-11, an enzyme encountered also in the testis, safety concerns about sperm effects were raised. A study conducted in men older than 45 years revealed no effect on spermatogenesis or reproductive hormones after chronic administration of tadalafil (10mg and 20mg) for 6 months. Preliminary results indicate that tadalafil may acutely (1–2 hours after administration of a single 20mg dose) impair sperm motility in young, infertile patients. However, there are no data showing that this fact may be of any clinical significance.

Studies across different ethnic groups and various risk factors revealed the same pattern of adverse events and discontinuation rates, without statistically significant differences between groups. This was also the case in patients treated in tertiary care academic centers where more severe erectile dysfunction and ED-associated comorbidities are expected compared with patients treated in primary care centers. Finally, the safety pattern of tadalafil did not change when administered continuously (20mg, 3 times per week or 5–10mg daily) compared with on-demand use in general erectile dysfunction populations o r in erectile dysfunction patients with diabetes mellitus.


The most frequent adverse events reported by patients on vardenafil were headache, flushing, and rhinitis, consistent with the vasodilatory properties of PDE-5 inhibitors. Discontinuation rates because of treatment-emergent adverse events were similar to those for placebo (2 % for placebo vs 3 % for vardenafil). Treatment-emergent adverse events were generally of mild-to-moderate intensity and rapidly decreased during long-term treatment. During one 104-week study, the incidence of treatment-emergent adverse events was greatest during the first 4 weeks of the study, and rapidly decreased during long-term treatment, similar to the situation seen with other PDE-5 inhibitors. Overall, 1.8 % of patients in the vardenafil 10mg group and 2 % of patients in the vardenafil 20mg group discontinued treatment because of adverse events.

Visual disturbances are reported more rarely than with sildenafil. Serious adverse events were infrequent, reported by 1–5 % of patients receiving vardenafil 5–20mg and 3–5 % of placebo recipients in short-term studies and in 11% and 13 % of patients receiving vardenafil 10mg and 20mg, respectively, in the 104-week study, but only one event (reduced visual acuity) was judged as possibly related to vardenafil treatment.

In the post-marketing surveillance study involving almost 30,000 patients in Germany, adverse events were reported by only 1.4 % of men. The incidence of adverse events with vardenafil under ‘real-life’ conditions appears to be much lower than that seen in clinical trials. Similar results have been reported in other general erectile dysfunction populations, in aging men, in patients with diabetes mellitus and in patients with spinal cord injury.

Phosphodiesterase type 5 inhibitors: Cardiovascular safety

Non-arteritic anterior ischemic optic neuropathy

The non-arteritic type of anterior ischemic neuropathy is not due to inflammation of the arteries; rather, it results from transient poor circulation or loss of circulation in the capillaries of the optic nerve head, causing infarction of the anterior optic nerve. It is a rare disorder (the annual incidence is 2.3 to 10.3 per 100,000). The pathophysiology of non-arteritic anterior ischemic neuropathy (NAION) is controversial. It is postulated that decreased perfusion in the pial vasculature (the capillaries feeding the optic nerve head) results in hypo-perfusion and ischemia of the optic nerve head. A vascular nature is suggested by the abrupt onset of symptoms typical of ischemia, the increased incidence in older age, and the strong association with vascular risk factors. NAION onset is typically characterized by sudden painless monocular visual loss that may progress over hours to weeks without any proven effective treatment.

A total of 196 post-marketing reports of NAION in patients using PDE-5 inhibitors (168 reports with sildenafil, 18 reports with tadalafil, and 10 reports with vardenafil) had been received by the US Food and Drug Administration (FDA) as of June, 2007. Thirty-six of these cases appear to be of the NAION subtype. In 26 cases, the loss of vision was described as continuing or permanent. Furthermore, five cases have been reported in Canada. There are 17 NAION case reports published in association with sildenafil and three in association with tadalafil. A few of the cases associated with sildena-fil use and one case associated with tadalafil experienced temporary partial visual loss that became a fixed visual loss upon rechallenging. No case reports have been published in association with vardenafil usage.

Based on the similar risk profiles for both NAION and ED, it is not unexpected to observe NAION cases in men with ED. Establishing a causal relationship between NAION and a specific risk factor (including the use of a particular medication) is problematic since most patients with NAION have one or more concurrent systemic or local risk factors that place them at risk for optic head ischemia. On 8 July, 2005, the FDA approved updated labeling for all three PDE-5 inhibitors, reflecting current knowledge of the possible association with NAION. At this time, it is not possible to determine whether PDE-5 inhibitors were the cause of the loss of eyesight or whether the problem is related to other factors such as high blood pressure or diabetes, or to a combination of these problems. The FDA advises patients to stop taking these medicines and call a doctor or healthcare provider right away if they experience a sudden or decreased vision loss in one or both eyes. Patients taking or considering taking these products should inform their healthcare professionals if they have ever had severe loss of vision, which might reflect a prior episode of NAION. Such patients are at an increased risk of developing NAION again and PDE-5 inhibitors should not be considered until possible risk factors are optimized. Aspirin is the only medication that shows some potential benefit in reducing the frequency of second eye involvement in patients with NAION. Therefore, a reasonable and informed consent should be provided by the clinician to patients regarding the possible but low risk of NAION in association with PDE-5 inhibitors.


Consistent with the effects of PDE-5 inhibition on the nitric oxide (NO)–cGMP pathway, all PDE-5 inhibitors may potentiate the hypotensive effects of nitrates. Therefore, the administration of all PDE-5 inhibitors is contraindicated in patients who are using any form of organic nitrates, either regularly or intermittently. Patients with known hereditary degenerative retinal disorders, including retinitis pigmentosa, were not included in clinical trials, and PDE-5 inhibitor use in these patients is not recommended.

Other precautions

Several precautions exist in patient subpopulations and as a result of drug interactions. Renal insufficiency alters the pharmacokinetics of PDE-5 inhibitors, increasing both the area under the curve (AUC) and the maximum plasma concentration. A starting oral dose of 25mg should be considered for sildenafil in patients with severe renal insufficiency. No modifications are necessary in patients with mild or moderate renal insufficiency. The dose of tadalafil should be limited to 5mg not more than once daily in patients with severe renal insufficiency or end-stage renal disease. A starting dose of 5mg not more than once daily is recommended for patients with moderate renal insufficiency; the maximum recommended dose is 10mg not more than once in every 48 hours. No dose adjustment is required in patients with mild renal insufficiency. In patients with moderate or severe renal insufficiency, the AUC of vardenafil was 20–30 % higher than in normal subjects. Vardenafil pharmacokinetics have not been evaluated in patients requiring renal dialysis.

In patients with hepatic cirrhosis (Child–Pugh score A and B), a starting oral dose of 25mg should be considered for sildenafil. Tadalafil dose should not exceed 10mg in these patients, and use in patients with severe hepatic impairment (Child–Pugh score C), is not recommended. Finally, a starting dose of vardenafil 5mg is recommended for patients with moderate hepatic impairment, and the maximum dose should not exceed 10mg. Vardenafil has not been evaluated in patients with severe hepatic impairment.

Drug interactions may play a significant role in the safety profile of PDE-5 inhibitors. Besides precautions already reported, Table Effects of other drugs on phosphodiesterase (PDE) type 5 inhibitors summarizes effects of PDE-5 inhibitors on other drugs. Physicians must be aware of these effects, since dose and treatment modifications may be necessary. The safety of PDE-5 inhibitors in patients with bleeding disorders and patients with active peptic ulceration is not documented. Therefore, they should be administered to these patients only after careful benefit–risk assessment and with caution. Co-administration with aspirin does not prolong bleeding time, relative to aspirin alone.

Table Effects of other drugs on phosphodiesterase (PDE) type 5 inhibitors

Study drug (dose studied for each PDE-5 inhibitor)


Mechanism of interaction Sildenafil Tadalafil Vardenafil
Cimetidine (800mg, N/A, 400mg bd) Non-specific CYP inhibitor 56 % increase in plasma concentrations (50mg) N/A (nizatidine – another H 2 antagonist – has no effect) No effect
Erythromycin (500mg tid, N/A, 500mg bd) CYP3A4 inhibitor 182 % increase in AUC (100mg) N/A (would probably increase tadalafil exposure) Three-fold increase in C max , four-fold increase in AUC
Ritonavir (500mg bd, 200mg bd, 800mg tid) P450 inhibitor 300 % increase in C max (100mg), 1000 % increase in AUC (100mg) No change in C max , 124 % increase in AUC 13-fold increase in C max , 49-fold increase in AUC
Saquinavir (1200mg tid, N/A, N/A) CYP3A4 inhibitor 140 % increase in C max (100mg), 210 % increase in AUC (100mg) N/A (expected to increase tadalafil exposure) N/A (expected to increase vardenafil exposure)
Indinavir (N/A, N/A, 800mg tid) N/A (expected to increase sildenafil exposure) N/A (expected to increase tadalafil exposure) Seven-fold increase in C max , 16-fold increase in AUC
Ketoconazole (N/A, 400mg daily, 200mg daily) CYP3A4 inhibitor (strong) N/A (expected to increase sildenafil exposure) 22 % increase in C max (20mg), 15 % increase in C max (10mg), 312 % increase in AUC (20mg), 107 % increase in AUC (20mg) 4-fold increase in C max , 10-fold increase in AUC
Itraconazole CYP3A4 inhibitor (strong) N/A (expected to increase sildenafil exposure) N/A (would probably increase tadalafil exposure) N/A (would probably increase vardenafil exposure)
Grapefruit CYP3A4 inhibitor (weak) N/A (may give rise to modest increases in plasma levels) N/A (would probably increase tadalafil exposure) N/A (would probably increase vardenafil exposure)
Rifampin (N/A, 600mg daily, N/A) CYP3A4 inducer N/A (expected to decrease sildenafil exposure) 46 % decrease in C max (10mg), 88 % decrease in AUC (10mg) N/A (expected to decrease vardenafil exposure)

Both alcohol and PDE-5 inhibitors act as mild vasodilators. When mild vasodilators are taken in combination, blood-pressure-lowering effects of each individual compound may be increased. Regular alcohol consumption does not alter pharmacokinetics of PDE-5 inhibitors. However, substantial consumption of alcohol (e.g. 5 units or greater) in combination with a PDE-5 inhibitor can increase the potential for orthostatic signs and symptoms, including increase in heart rate, decrease in standing blood pressure, dizziness, and headache.

PDE-5 inhibitors should be used with caution in patients with anatomical deformation of the penis (such as angulation, cavernosal fibrosis, or Peyronie’s disease), or in patients who have conditions that may predispose them to priapism (such as sickle cell anemia, multiple myeloma, or leukemia). There have been rare reports of prolonged erections for more than 4 hours and priapism (painful erections lasting longer than 6 hours) for this class of compounds. In the event that an erection persists for longer than 4 hours, the patient should seek immediate medical assistance.

The safety and efficacy of combinations of PDE-5 inhibitors with other treatments for erectile dysfunction have not been studied in pre-marketing trials. However, there are data in the literature supporting combination of PDE-5 inhibitors with intraurethral alprostadil and intracavernosal injections in non-responders to PDE-5 inhibitors alone. Such combinations may increase side-effects and they are not recommended in everyday clinical practice.

Phosphodiesterase type 5 inhibitors: Summary

Numerous clinical trials and post-marketing studies have established that PDE-5 inhibitors have an excellent overall safety profile. Adverse events are typically transient, mild-to-moderate in severity, and self-limiting with continuous use. Despite early concerns, the cardiovascular safety of PDE-5 inhibitors is also excellent with the incidence rates of cardiovascular adverse events being similar to those with placebo. Furthermore, recent data show that PDE-5 inhibitors may actually have a cardioprotective role. However, certain limitations and precautions exist. Physicians must be aware of these in order to treat erectile dysfunction patients safely and effectively.

Konstantinos Hatzimouratidis and Dimitrios Hatzichristou

Selections from the book: “Textbook of Erectile Dysfunction”, 2009

Share →