Over the past 15 years, an increasing number of publications have reported the pharmacological treatment of premature ejaculation with a variety of different medications, which act centrally or locally to retard the psychoneurological control of ejaculation and subsequent orgasm. It is well established that major tranquilizers and SSRIs retard ejaculation significantly and will, in a small percentage of men, result in anejaculation. The efficacy of SSRIs in delaying ejaculation combined with the low side-effect profile make them first-line agents for premature ejaculation, administered either on a daily or an on-demand basis.

Psychosexual counseling

In many relationships, premature ejaculation causes few if any problems. In others, the couple may reach an accommodation of the problem through various strategies – young men with a short refractory period may often experience a second and more controlled ejaculation during a subsequent episode of lovemaking. Frequently however, premature ejaculation eventually leads to significant relationship problems, with the partner regarding the man as selfish and developing a pattern of sexual avoidance. This only worsens the severity of the prematurity on the occasions when intercourse does occur.

The cornerstones of behavioral treatment are the Seman’s ‘stop–start’ maneuver and its modification proposed by Masters and Johnson, the squeeze technique. Both are based on the theory that premature ejaculation occurs because the man fails to pay sufficient attention to pre-orgasmic levels of sexual tension. As most men with premature ejaculation are aware of their anxiety and the sources of that anxiety tend to be relatively superficial, treatment success with these behavioral approaches is relatively good in the short term but convincing long-term treatment outcome data are lacking.

Pharmacological treatment of premature ejaculation

Pharmacological modulation of ejaculatory threshold represents a novel and refreshing approach to the treatment of premature ejaculation and a radical departure from the psychosexual model of treatment, previously regarded as the cornerstone of treatment. The introduction of the SSRIs has revolutionized the approach to and treatment of premature ejaculation. SSRIs encompass five compounds: citalopram, fluoxetine, fluvoxamine, paroxetine, and sertra-line, all with a similar pharmacological mechanism of action. Although the methodology of the initial drug treatment studies was rather poor, later double-blind and placebo-controlled studies replicated the genuine effect of clomipramine and SSRIs to delay ejaculation. In spite of a development towards more evidence-based drug treatment research, the majority of studies still lack adequate design and methodology. A recent meta-analysis of all drug treatment studies demonstrated that only 14.4% had been performed according to the established criteria of evidence-based medicine, and that open-design studies and studies using subjective reporting or questionnaires showed a higher variability in ejaculation delay than double-blind studies in which the ejaculation delay was prospectively assessed with a stopwatch.

Daily treatment with selective serotonin reuptake inhibitors

Daily treatment can be performed with paroxetine 20–40mg, clomipramine 10–50mg, sertraline 50–100mg, or fluoxetine 20–40mg. Paroxetine appears to exert the strongest ejaculation delay, increasing intravaginal ejaculatory latency time approximately 8.8-fold over baseline. Ejaculation delay usually occurs within 5–10 days but may occur earlier. Adverse effects are usually minor, start in the first week of treatment, and gradually disappear within 2–3 weeks; they include fatigue, yawning, mild nausea, loose stools, and perspiration. Diminished libido or mild erectile dysfunction is infrequently reported. Significant agitation is reported by a small number of patients and treatment with SSRIs should be avoided in men with a history of bipolar depression.

Daily treatment with alpha-1-adrenoceptor antagonists

Ejaculation is a sympathetic spinal cord reflex that could theoretically be delayed by alpha-1-adrenoceptor blockers. Several authors have reported their experience with the selective alpha-1-blockers alfuzosin and terazosin in the treatment of premature ejaculation. In a double-blind, placebo-controlled study, Cavallini reported that alfuzosin (6 mg/day) and terazosin (5 mg/day) were effective in delaying ejaculation in approximately 50% of the cases. Similarly, Basar reported that terazosin was effective in 67% of men. However, both studies were limited by the use of subjective study end-points of patient impression of change and sexual satisfaction, and neither evaluated objective end-points such as IELT.

However, in the rat model, systemic injection of tamsulosin impaired bulbospongiosus muscle contractile capacity and bladder neck and seminal vesicle pressures, whereas alfuzosin did not. Consistent with this, Hellstrom et al. reported decreased ejaculate volume in almost 90% of subjects and anejaculation in approximately 35% of participants with tam-sulosin, but he failed to observe anejaculation with alfuzosin. Contrary to these results, a comparison of tamsulosin with alfuzosin in men symptomatic BPH and normal ejaculatory latency demonstrated minimal ejaculatory dysfunction. Additional controlled studies are required to determine the role of alpha-1-blockers in the treatment of premature ejaculation.

On-demand treatment with selective serotonin reuptake inhibitors

Administration of clomipramine, paroxetine, sertraline, or fluoxetine 4–6 hours before intercourse is efficacious and well tolerated but is associated with less ejaculatory delay than daily treatment. Daily administration of an SSRI is associated with superior increases in intravaginal ejaculatory latency time compared with on-demand administration, owing to the greatly enhanced 5-HT neuro-transmission resulting from several adaptive processes, which may include presynaptic 5-HT-1A and 5-HT-1B or 5-HT-1D receptor desensitization. On-demand treatment may be combined with either an initial trial of daily treatment or concomitant low-dose daily treatment.

The assertion that on-demand drug treatment of premature ejaculation is preferable to daily dosing parallels the rationale for the treatment of erectile dysfunction but is contrary to personal experience and the results of the only premature ejaculation drug preference study. Whilst many men suffering from premature ejaculation who infrequently engage in sexual intercourse may prefer on-demand treatment, the majority of men in established relationships prefer the convenience of daily medication.

On-demand treatment with dapoxetine

A number of rapid-acting, short half-life SSRIs are under investigation as on-demand treatments for premature ejaculation. Dapoxetine is an SSRI and was originally developed as a short half-life drug for the treatment of depression. In December 2004, a new drug application was submitted to the Food and Drug Administration in the USA for dapoxetine hydrochloride as a drug to treat premature ejaculation. Dapoxetine is a potent SSRI (binding affinity, pKi = 8 nM), structurally similar to fluoxetine. Equilibrium radioligand binding studies using human cells demonstrate that dapoxetine binds to 5-HT, norepinephrine and dopamine reuptake transporters and inhibits uptake in the following rank order of potency: 5-HT> norepinephrine >> dopamine. Brain PET studies have demonstrated significant displaceable binding of radiolabeled dapoxetine in the cerebral cortex and subcortical gray matter.

Dapoxetine undergoes rapid absorption and elimination, resulting in minimal accumulation, and it has dose-proportional pharmacokinetics, which are unaffected by multiple dosing. The pharmacokinetic profile of dapoxetine suggests that it is a candidate for on-demand treatment of premature ejaculation. The pharmacokinetics of both single doses and multiple doses over 6–9 days (30mg, 60mg, 100mg, 140mg, or 160mg) have been evaluated. Dapoxetine has a time to maximum concentration of 1.4–2.0 hours, and it rapidly achieves peak plasma concentration following oral administration. Both plasma concentration and area under the curve (AUC) are dose-dependent up to 100mg. The mean half-life of dapox-etine after a single dose is 0.5–0.8 hours and plasma concentrations rapidly decline to about 5% of peak plasma concentration at 24 hours. The pharmacokinetics of dapoxetine and its metabolites were not affected by repeated daily dosing, and state plasma concentrations were reached within 4 days, with only modest accumulation of dapoxetine (approximately 1.5-fold). Food does not have a clinically significant effect on dapox-etine pharmacokinetics.

No drug–drug interactions associated with dapoxetine have been reported. Co-administration of dapoxetine with ethanol did not produce significant changes in the pharmacokinetics of dapoxetine and its metabolites. Drug interaction studies demonstrate that tadalafil, a phosphodiesterase (PDE) type 5 inhibitor used in the treatment of ED, did not affect the phar-macokinetics of dapoxetine, whereas sildenafil increased the dapoxetine AUC by 22%. However, this was not regarded as clinically important. Dapoxetine did not appear to affect the pharmacokinetics of tadalafil or sildenafil.

Preliminary data suggest that dapoxetine administered 1–2 hours prior to planned intercourse is modestly effective and well tolerated, superior to placebo, and increases IELT two- to three-fold over baseline in a dose-dependent fashion. In randomized, double-blind, placebo-controlled, multicenter, phase 3 12-week clinical trials involving 2614 men with a mean baseline intravaginal ejaculatory latency time ≤2 minutes, dapoxetine 30mg or 60mg was more effective than placebo for all study end-points. Mean intravaginal ejaculatory latency time increased from 0.91 minutes at baseline to 2.78 minutes (3.0-fold) and 3.32 minutes (3.6-fold) at study end with dapoxetine 30mg and 60mg, respectively, compared with a 1.9-fold increase with placebo. Mean patient rating of control over ejaculation as fair, good, or very good increased from 2.8% at baseline to 51.8% and 58.4% at study end with dapoxetine 30mg and 60mg, respectively. Treatment-related side effects were uncommon and dose-dependent; they included nausea, diarrhea, headache, and dizziness, and they were responsible for study discontinuation in 4% (30mg) and 10% (60mg) of subjects.

However, Waldinger et al. have suggested that dapoxetine-induced ejaculation delay is marginal and has been systematically overestimated by the misleading use of arithmetic mean intravaginal ejaculatory latency times as opposed to geometric mean intravaginal ejaculatory latency times or median IELTs at study end. They assert that the use of median intravaginal ejaculatory latency times at study end would result in a substantially lower study end intravaginal ejaculatory latency time fold increase of 1.2, 1.9 and 2.3 for placebo, dapoxetine 30mg, and dapoxetine 60mg, respectively. This is of marginal clinical significance, is only slightly higher than the 1.4-fold increase of placebo and is substantially less than the 8.8-fold increase seen with daily dosing of paroxetine. Waldinger et al. also assert that the dapoxetine phase 3 study lacks an adequate methodology of assessment of dapoxetine-related adverse effects, since adverse effects were retrospectively assessed with a non-validated questionnaire at each monthly visit to the clinic. They cite the disparity between the incidence of dapoxetine adverse effects in volunteer control pharmacoki-netic studies and the dapoxetine phase 3 study as consistent with this aspect of inadequate study design and the risk of underestimating the incidence of adverse effects like dizziness, headache, and diarrhea.

It is likely that dapoxetine, despite its modest effect upon ejaculatory latency, has a place in the management of premature ejaculation, which will eventually be determined by market forces once the challenge of regulatory approval has been met.

On-demand treatment with tramadol

The efficacy of on-demand tramadol in the treatment of premature ejaculation was recently reported. Tramadol is a centrally acting synthetic opioid analgesic with an unclear mode of action that is thought to include binding of parent and M1 metabolite to micro-opioid receptors and weak inhibition of reuptake of norepinephrine and 5-HT. Serotonin syndrome has been reported as an adverse effect of tramadol alone or in combination with SSRIs. In a double-blind, placebo-controlled study, the on-demand use of tramadol 50mg, taken 2 hours prior to intercourse, exerted a clinically relevant ejaculation delay in men with premature ejaculation with a 12.7-fold increase in IELT. In a single-blind, placebo-controlled study, the on-demand use of tramadol 25mg, taken 1–2 hours prior to intercourse, was associated with a 6.3-fold increase in IELT. Additional flexible dose studies and long-term follow-up studies to evaluate the risk of opioid addiction are required.

Anesthetic topical ointments

The use of topical local anesthetics such as lidocaine and prilocaine as a cream, gel, or spray is well established and is moderately effective in retarding ejaculation. A recent study reported that a metered-dose aerosol spray containing a eutec-tic mixture of lidocaine and prilocaine produced a 2.4-fold increase in baseline intravaginal ejaculatory latency time and significant improvements in ejaculatory control and in both patient and partner sexual quality of life. They may be associated with significant penile hypoesthesia and possible transvaginal absorption, resulting in vaginal numbness and resultant female anorgasmia unless a condom is used.

Intracavernous injection of vasoactive drugs

Intracavernous self-injection treatment of premature ejaculation has been reported but is without any evidence-based support for the efficacy of this strategy. Fein reported an open study of eight men treated with a combination of papaverine and phen-tolamine administered by intracavernous auto-injection in which the treatment success was defined as prolongation of erection after ejaculation and not by any measure of ejaculatory latency. Three of eight men stated that they were cured and suspended treatment while the other five men continued using the medication. In the absence of well-controlled studies, treatment of premature ejaculation by intracavernous auto-injection cannot be recommended.

Phosphodiesterase type 5 inhibitors

Medications that inhibit PDE-5 isoenzyme – sildenafil, tadalafil, and vardenafil – are effective treatments for ED. Several authors have reported their experience with PDE-5 inhibitors alone or in combination with SSRIs as a treatment for premature ejaculation. The putative role of PDE-5 inhibitors as a treatment for premature ejaculation is based upon the role of the nitric oxide (NO)–cGMP transduction system as a central and peripheral mediator of inhibitory non-adrenergic, non-cholinergic nitrergic neu-rotransmission in the urogenital system. Several studies suggest that elevation of extracellular NO in the MPOA accelerates dopamine release and facilitates male copulatory behavior of rats, whereas a decrease of NO reduces their cop-ulatory behavior. Hull et al. demonstrated that microin-jection of the NO synthase (NOS) inhibitor, N-nitro- L-arginine methyl ester (NAME) decreased the number of erections, but also increased the number of seminal emissions and decreased the latency to the first seminal emission. The results indicate that not only does NO promote erection in intact male rats, but it may also inhibit seminal emission.

NOS isoenzymes are present in human seminal vesicle smooth muscle. Several authors have reported the effects of NO donor drugs on electrically induced contractions and on tissue levels of cGMP and cAMP in isolated human seminal vesicle smooth muscle preparations and have concluded that NO might be involved in the control of secretory activity and smooth muscle function of human seminal vesicles. Consistent with this, Kriegsfeld reported that mice homozygous for endothelial NOS (eNOS) gene deletion have striking ejaculatory anomalies. A significantly higher percentage of eNOS gene deletion mice than normal controls ejaculated during the testing period, requiring less stimulation and fewer mounts and intromissions.

A recent systematic review of 14 studies published in peer-reviewed journals or the proceedings of major international and regional scientific meetings on the PDE-5 inhibitor treatment for premature ejaculation examined the role of NO as a neurotransmitter involved in the central and peripheral control of ejaculation, the methodology of PDE-5 inhibitor treatment studies, the adherence of methodology to the contemporary consensus of ideal premature ejaculation drug trial design, the impact of methodology on treatment outcomes, and the role of PDE-5 inhibitors in the treatment of premature ejaculation. These studies cover a total of 1102 subjects suffering from premature ejaculation and treated with sildenafil, tadalafil, or vardenafil, either as monotherapy or in combination with SSRIs, clomipramine, or topical anesthetics.

Most of these studies support a role for PDE-5 inhibitors in the treatment of premature ejaculation and speculate multiple mechanisms, including a central effect involving increased NO and reduced sympathetic tone, smooth muscle dilatation of the vas deferens and seminal vesicles, which may oppose sympathetic vasoconstriction and delay ejaculation, reduced performance anxiety as a result of better erections, and down-regulation of the erectile threshold to a lower level of arousal so that increased levels of arousal are required to achieve the ejaculation threshold.

The small number of publications and the lack of sufficient data preclude any meta-analysis of results. However, examination of the methodology of these studies, the adherence of methodology to the contemporary consensus of ideal clinical trial design, and the impact of study methodology on treatment outcomes fails to provide any robust empirical evidence to support a role of PDE-5 inhibitors in the treatment of premature ejaculation, with the exception of men with premature ejaculation and comorbid ED. Of the 14 studies reviewed, only one fulfilled these criteria and this study failed to confirm any significant treatment effect on IELT.

Caution should be exercised in interpreting data from inadequately designed studies of PDE-5 inhibitors and on-demand SSRI treatment, and their results must be regarded as unreliable. The extremely broad range of intravaginal ejaculatory latency time fold increases reported with sildenafil (2.7–15.0, mean 6.6), combined sildenafil and on-demand sertraline (3.3–10.0, mean 6.9), and combined sildenafil and on-demand paroxetine (6.6–14.9, mean 10.7) in this systematic review is testament to the unreliability of inadequate study design. In contrast to these findings, the range of placebo intravaginal ejaculatory latency time fold-increases was relatively narrow (IELT-range 1.2–1.6, mean 1.4) and was identical to the mean 1.4 intravaginal ejaculatory latency time fold-increase reported in a meta-analysis of other premature ejaculation drug studies.

Treatment of premature ejaculation and comorbid erectile dysfunction

There is evidence to suggest that PDE-5 inhibitors alone or in combination with an SSRI may have a role in the management of acquired premature ejaculation in men with comorbid ED. This systematic review includes three studies of patients with premature ejaculation with comorbid erectile dysfunction treated with a PDE-5 inhibitor alone or in combination with sertraline. In 45 men with premature ejaculation and comorbid erectile dysfunction treated with flexible doses of sildenafil (50–100mg) for periods of 1–3 months, Li et al. reported improved erectile function in 40 (89%) and reduced severity of premature ejaculation in 27 (60%). Improved erectile function was reported by all of the 27 men with reduced severity of premature ejaculation, of whom 81.5% described themselves as satisfied or very satisfied. Contrary to these findings, only 1 of the 18 men (5.6%) who did not obtain improvement of premature ejaculation reported treatment satisfaction.

In a group of 37 men with primary or acquired premature ejaculation and a baseline score from the erectile function domain of the International Index of Erectile Function (IIEF) of 20.9 (consistent with mild ED), Sommer et al. reported a 9.7-fold intravaginal ejaculatory latency time increase and normalization of erectile function (score of 26.9) with vardenafil treatment as opposed to a 4.4-fold intravaginal ejaculatory latency time increase with on-demand sertraline.

The high level of correlation between improved erectile function with sildenafil and reduced severity of premature ejaculation reported by Li et al. and the superior intravaginal ejaculatory latency time fold-increase observed with vardenafil reported by Sommer et al. indicates that reduced premature ejaculation severity related to PDE-5 inhibitor use is due to improved erectile function. The intravaginal ejaculatory latency time fold increase observed by Sommer et al. with on-demand sertraline (4.4) is less than that reported in reviewed studies on men with normal erectile function (mean 5.57, range 3.0–8.5), suggesting that men with premature ejaculation and comorbid erectile dysfunction are less responsive to on-demand SSRIs and are best managed with a PDE-5 inhibitor alone or in combination with an SSRI. Furthermore, the report of Chia that the addition of sertraline to sildenafil in the treatment of men with erectile dysfunction with comorbid premature ejaculation was associated with a lesser intravaginal ejaculatory latency time fold increase (3.3) and lower levels of treatment satisfaction than that seen in men with lifelong premature ejaculation and normal erectile function treated with on-demand sertraline, suggests that this group of men are less responsive to pharmacotherapy.

The proposed mechanism of action of PDE-5 inhibitors as monotherapy or in combination with a SSRI in the treatment of acquired premature ejaculation in men with comorbid erectile dysfunction includes the ability to maintain an erection following ejaculation, reduction of the erectile refractory period, and reliance upon a second and more controlled ejaculation during a subsequent episode of intercourse, and a reduction in performance anxiety due to better erections or down-regulation of the erectile threshold to a lower level of arousal so that increased levels of arousal are required to achieve the ejaculation threshold.

The future of drug development

Several in vitro and animal studies have demonstrated that the desensitization of 5-HT-1A receptors, increased activation of postsynaptic 5-HT-2C receptors, and the resultant higher increase in synaptic 5-HT neurotransmission seen with daily dosing of SSRIs can be acutely achieved by blockade of these receptors by administration of an on-demand SSRI and a 5-HT-1A receptor antagonist.

An increasing number of studies report the involvement of central oxytocinergic neurotransmission in the ejaculatory process. In human males, plasma oxytocin levels are elevated during penile erection and at the time of orgasm. Electrical stimulation of the dorsal penile nerve produced excitation in about half of the oxytocin cells in the paraventricular nucleus and supraoptic nucleus of rats. In a rat model, systematic administration of oxytocin facilitated ejaculation by reducing the number of intromissions required for ejaculation, ejaculation latencies, and post-ejaculation intervals. The use of oxytocin receptor antagonists may also have a role but there have been no reports of their efficacy in the treatment of premature ejaculation.

Drug combinations of on-demand rapid-acting SSRIs and 5-HT-1A receptor antagonists or oxytocin receptor antagonists, or single agents that target multiple receptors, may form the foundation of more effective future on-demand medication.


Several authors have reported the use of surgically induced penile hypoesthesia via selective dorsal nerve neurotomy or augmentation of the glans penis with hyaluronic acid gel in the treatment of lifelong premature ejaculation that is refractory to behavioral and pharmacological treatment. The role of surgery in the management of premature ejaculation remains unclear until the results of further studies have been reported.

Men with natural variable premature ejaculation and premature ejaculation-like ejaculatory dysfunction should be managed with psychoeducation, reassurance, or psychotherapy and should not, in general, be treated with pharmacotherapy.

Office management of premature ejaculation

Men with premature ejaculation should be evaluated with a detailed medical and sexual history, a physical examination, and appropriate investigations to establish the true presenting complaint and to identify obvious biological causes such as genital or lower urinary tract infection.

Men with premature ejaculation secondary to ED, other sexual dysfunction, or genitourinary infection should receive appropriate etiology-specific treatment. Men with lifelong premature ejaculation should be initially managed with pharmacotherapy. Men with significant contributing psychogenic or relationship factors may benefit from concomitant behavioral therapy. Men with premature ejaculation secondary to erectile dysfunction can be treated with either ED-specific pharmacotherapy (e.g. PDE-5 inhibitors as monotherapy) or combination therapy with premature ejaculation-specific pharmacotherapy (e.g. daily or on-demand SSRIs). Recurrence of premature ejaculation is highly likely to occur following withdrawal of treatment. Men with acquired premature ejaculation can be treated with pharmacotherapy or behavioral therapy (or both) according to patient and partner preference. Restoration of ejaculatory control in men with acquired premature ejaculation is likely to occur following completion of treatment but is the exception in men with lifelong premature ejaculation. Behavioral therapy may augment pharmacotherapy to enhance relapse prevention.


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

Share →