Although the behavioral techniques described above are generally successful in the treatment of premature ejaculation, there are some limitations associated with and concerns about this mode of treatment. While most patients who are able to comply with the full treatment course do well, about 25% of patients do not respond to behavioral treatments (Hawton et al. 1986). In addition, the treatment requires a cooperative partner for optimal outcome, which is not always feasible. There is also some evidence that the benefits from behavioral techniques are not always maintained in the long term. Hawton et al. (1986) found that in a 3-year follow-up, 75% of patients showed no lasting improvement over baseline. De Amicis et al. (1985) also found that immediate posttherapy improvement was not sustained by 3 years posttreatment. Some patients prefer pharmacotherapy because it is more easily hidden from a partner than scheduled psychotherapy sessions and requires less of a time and behavioral investment. The issue of payment for intensive behavioral therapy must be considered in the age of managed care. The cost-effectiveness argument is made less compelling by the availability of patient self-help books and films (e.g., Kaplan 1989), which can greatly reduce the need for frequent psychotherapy sessions.

Pharmacological options for the treatment of premature ejaculation have become more varied and better studied in recent years. Pharmacological agents that block sympathetic arousal, raise serotonin or prolactin levels, or have sedative effects may have the effect of delaying ejaculation (Buffum 1992; Segraves 1988, 1989, 1993). Multiple medications have been reported in individual case reports to be useful in the treatment of premature ejaculation. These include Korean red ginseng (Choi et al. 1995), prilocaine-lidocaine cream (Berkovitch et al. 1995), intracavernous injection of vasoactive drugs (Fein 1990), lorazepam (Segraves 1987), a-blockers (Beretta et al. 1986), and b-blockers (Cooper and Magnus 1984). While these options may have some efficacy, they are all associated with unwanted side effects that make them of limited value for most patients.

Interest in the role of antidepressants in the treatment of premature ejaculation arises from the many reports of delayed ejaculation as a side effect of these medications. Several reviews have described the potential utility of antidepressants in the management of premature ejaculation (Balon 1996; Rosen et al. 1999; St. Lawrence and Madakasira 1992). Multiple reports exist documenting the effect on ejaculation of various psychotropic agents, including the monoamine oxidase inhibitors (MAOIs) tranylcypromine (Simpson et al. 1965), phenelzine (Harrison et al. 1985; Rapp 1979), and isocarboxazid (Gross 1982); lithium (Blay et al. 1982); the tricyclic antidepressants imipramine (Glass 1981), clomipramine (Klug 1984; Quirk and Einarson 1983; Yassa 1982), and amitriptyline (Ninninger 1978); the atypicals trazodone, amoxapine, and maprotiline (Segraves 1992); and the serotonin reuptake-inhibiting antidepressants (SRIs) (Ashton et al. 1997; Forster and King 1994; McGilp 1993; Montejo-Gonzalez et al. 1997; Wise 1994). Delayed ejaculation is thus a potential side effect of nearly all antidepressants and appears to be dose-related and reversible with discontinuation of the medication. For the majority of patients taking antidepressants, this side effect is undesirable and can lead to noncompliance. However, for patients suffering from premature ejaculation, the side effect can be used to advantage.

The earliest mention of the use of antidepressants to treat premature ejaculation was a report by Bennett, who in 1961 described the efficacy of the MAOIs iproniazid and isocarboxazid in the treatment of three patients with premature ejaculation (Bennett 1961). Improvement was noted within 6 weeks of initiation of treatment, and relapse occurred within 2 weeks of discontinuation of the drug. None of the patients developed failure of erection, and the dosage was low, with minimal side effects other than the desired delay in ejaculation.

The tricyclic clomipramine was the first agent to show consistent promise in the pharmacological treatment of premature ejaculation. The earliest study in this indication was that of Eaton (1973), who found that clomipramine improved ejaculation latency in 12 of 13 patients treated. Goodman’s study in 1980 employed a placebo-controlled design and found no difference in ejaculation latency between patients receiving clomipramine 40 mg/day and those receiving placebo. However, when the study was made open label and the dosage gradually increased, 9 of the 16 patients responded (Goodman 1980). Girgis et al. (1992), in a placebo-controlled crossover study of men with premature ejaculation, found significant improvement in patients taking clomipramine. Interestingly, the effects were generally maintained when these patients were switched to the placebo group. Fewer of the patients who started on placebo and were switched to clomipramine showed improvement. Other studies of clomipramine (Althof 1995; Althof et al. 1995; Assalian 1988; Haensel et al. 1996; Kim and Seo 1998; Segraves et al. 1993; Strassberg et al. 1999) have generally shown similar results, indicating that this agent is very useful in the treatment of premature ejaculation. However, in many of these studies, side effects, especially anticholinergic effects, were significant and limited compliance to some degree.

The relatively benign profile of the SRIs makes them excellent candidates for treatment of premature ejaculation. Over the last few years, a number of case reports of SRI treatment of premature ejaculation have been published. These (Forster and King 1994; Kaplan 1994; McHale and Phanjoo 1994; Mendels 1994; Power-Smith 1994; Swartz 1994) indicate good response with minimal side effects.

In general, results of placebo-controlled trials have also been promising. Lee et al. (1996) used fluoxetine dosages of 20-60 mg/day to treat 19 men with premature ejaculation. They found subjective improvement in ejaculation latency, sexual satisfaction, and desire, as well as decreased anxiety associated with intercourse. In a placebo-controlled comparison of fluoxetine at two dosages (20 and 40 mg/day), Kara et al. (1996) reported that both dosages led to increased ejaculation latency. Mendels et al. (1995) conducted a placebo comparison of the effects on premature ejaculation of sertraline dosages ranging from 50 to 200 mg/day. They found sertraline-dose-related increases in ejaculation latency, number of successful intercourse attempts, and subjective satisfaction. Ludovico et al. (1996) used sertraline to treat 32 men with premature ejaculation and reported that all improved within 2 weeks. Side effects were mild, but within 3 weeks of discontinuing the medication at the end of the study, 90% of the men had experienced recurrence of their symptoms. Kindler et al. (1997) studied men with premature ejaculation and comorbid anxiety in an open-label trial with fluoxetine at 20 mg/day. Improvement of the premature ejaculation was seen within 2 weeks, whereas improvement in anxiety symptoms took 4 weeks. Waldinger et al. (1997) reported the results of a double-blind, placebo-controlled trial of paroxetine at 20 or 40 mg/day. Although both 20 and 40 mg/day were successful in retarding ejaculation, no differences in response rate were noted between the two dosages. Kim and Seo (1998) compared fluoxetine, sertraline, clomipramine, and placebo in 36 men with premature ejaculation. Although the best results, in terms of patient and partner satisfaction as well as ejaculation latency, were found with clomipramine, clomipramine produced significantly more side effects than the other drugs. Nearly as satisfactory results were seen with sertraline. Waldinger et al. (1998) compared fluoxetine 20 mg/day, fluvoxamine 100 mg/day, paroxetine 20 mg/day, and sertraline 50 mg/day in a placebo-controlled, double-blind 6-week study. With the exception of fluvoxamine, all of the agents were superior to placebo in delaying intravaginal ejaculation latency. The absence of response to fluvoxamine was most likely a consequence of the relatively low dosage used; Montejo-Gonzalez et al. (1997) demonstrated that fluvoxamine does commonly lead to sexual dysfunction, including orgasmic delay. Paroxetine produced the greatest delay, followed by sertraline and fluoxetine. Other placebo-controlled studies were conducted by Haensel et al. (1998) with clomipramine and by McMahon (1998) with sertraline. Paick and Kilm (1998) found utility for 50 mg of sertraline taken on an “as needed” basis the evening of anticipated sexual activity. This strategy was helpful in 15 (83%) of the 18 men tested, thus raising the hope that some patients may not need to take medication on a scheduled basis to achieve desired results.

The use of medication to treat premature ejaculation is not without its detractors. Whether patients can utilize these agents on an “as needed” versus a scheduled basis is unclear and perhaps subject to individual variability. Dose and duration are not known with certainty. Gains may be lost after medication discontinuation. Medication can be expensive and has the potential to cause further undesirable side effects. The most problematic adverse reactions involve an “overshooting” of the desired orgasmic delay, which could potentially cause a change in orgasmic intensity, a decrease in sexual drive, or even complete anorgasmia (Nitenson and Cole 1993; Segraves 1993). Occasionally, these results can be successfully managed with dosage reduction or with alternative medications described in the following section.

In summary, use of psychotherapeutic and/or pharmacological strategies in treating premature ejaculation produces some of the most successful results in all of medicine. These strategies allow the practitioner to ably manage a very common sexual dysfunction with little negative impact on the patient.