(2001) 13, 41±45 ß 2001 Nature Publishing Group All rights reserved 0955-9930/01 $15.00 www.nature.com/ijir Assessment of as needed use of pharmacotherapy and the pause-squeeze technique in premature ejaculation IA Abdel-Hamid 1 *, EA El Naggar 2 and A-H El Gilany 3 1 Department of Andrology, Mansoura Faculty of Medicine, Mansoura, Egypt; 2 Department of Psychiatry, Mansoura Faculty of Medicine, Mansoura, Egypt; and 3 Department of Community Medicine, Mansoura Faculty of Medicine, Mansoura, Egypt The objective was to compare the ef cacy and safety of the as needed use of clomipramine, sertraline, paroxetine, sildena l and the pause-squeeze technique in treatment of primary premature ejaculation. A prospective double blind randomized crossover study involving 31 patients was performed. Treatment phases comprised ve 4-week consecutive treatment periods, each separated by a two-week washout period. Patients were randomly assigned to receive each of the 4 drugs and use pause-squeeze on an as needed basis. Drugs were administered 3 to 5 hours before anticipated coitus. Anxiety score and ejaculation latency time were measured before treatment, after each treatment, and during washout periods. Sexual satisfaction score was measured after each treatment. The median ejaculation latency time was signi cantly increased from the pretreatment median of 1 minute to 4 minutes, 3 minutes, 4 minutes, 15 minutes and 3 minutes during treatment with clomipramine, sertraline, paroxetine, sildena l and pause-squeeze technique, respectively (all P ˆ 0.0001). Sildena l was superior to other modalities in terms of ejaculation latency and satisfaction (P ˆ 0.0001). The three antidepressants were comparable to each other in terms of ef cacy (P > 0.05). Paroxetine was superior to the pause-squeeze technique in terms of ef cacy (P < 0.05). In conclusion, sildena l appears to be superior to other modalities and a valid alternative in treatment of premature ejaculation. The 3 antidepressants were equivalent to each other in terms of ef cacy and safety. Paroxetine was superior to pause-squeeze technique in terms of ef cacy. (2001) 13, 41±45. Keywords: premature ejaculation; behavioral therapy; antidepressants; sildena l Introduction Premature ejaculation has been treated with various modalities. These modalities include behavioral therapy, 1,2 topical applications, 3 oral pharmacotherapy, 4±6 and intracavernosal vasoactive drug injection. 7 Oral pharmacotherapy such as the tricyclic antidepressant, clomipramine, and serotonin reuptake inhibitors, are associated with variable rates of success when taken daily. 5,8 In contrast some reports suggest that these drugs may be effective when received as needed. 4,9 Although behavioral therapy requires partner cooperation, some reports *Correspondence: IA Abdel-Hamid, Department of Andrology, Mansoura Faculty of Medicine, PO Box 35516, Mansoura, Egypt. E-mail: ahamidia@mum.mans.eun.eg Received 3 August 2000; accepted 9 September 2000 documented the success of this modality when used as needed in the treatment of premature ejaculation. 2,10 Recent knowledge has been gained with oral sildena l treatment of psychogenic impotence. Sildena l is a potent and selective inhibitor of cyclic guanosine monophosphate (cgmp)-speci c phosphodiesterase type 5. It thus enhances the relaxant effect of nitric oxide released in response to sexual stimulation by increasing cgmp concentrations in the corporal smooth muscle. 11 Since premature ejaculation, like psychogenic impotence, is a performance-anxiety problem in an otherwise normal male, it could be anticipated that oral sildena l would work in the treatment of premature ejaculation. Since the comparison of these different modalities when used as needed has not been evaluated, this study was undertaken to compare the ef cacy and safety of as needed use of oral clomipramine, sertraline, paroxetine, sildena l and the Masters and Johnson pause-squeeze technique in the treatment of primary premature ejaculation.
42 Methods From May 1999 to June 2000, 31 heterosexual men complaining of primary premature ejaculation from the beginning of their sexual life were enrolled in this study. Premature ejaculation was de ned as intravaginal ejaculation latency time (IVELT) of less than 2 minutes and the patient complained of little, if any, control over their ejaculation. The patients were recruited from the outpatient clinic of the andrology and sexology unit, Mansoura University Hospital, Mansoura, Egypt. All patients gave their verbal consent to participate in the study. Inclusion criteria include persons married for at least 1 y and willingness to attempt sexual intercourse at least twice per week with a cooperative female partner. Exclusion criteria include the following: (a) history of a psychiatric disorder; (b) current physical illness (eg, diabetes, liver disease and so forth); (c) previous surgery or drug known to affect sexual function; (d) current substance abuse (alcohol or drug); (e) patients with secondary premature ejaculation combined with erectile dysfunction. All patients were asked not to use condoms or topical penile applications. The study utilized a prospective randomized double blind crossover design. Treatment phases comprised ve 4-week consecutive treatment periods, each separated by a two-week washout period. Each patient was informed that he would be treated with 5 different modalities of therapies of identical action to ascertain which of the 5 was most useful. The patients were randomly assigned to receive clomipramine hydrochloride 25 mg (Anafranil 1, Novartispharma, Cairo, Egypt), sertraline hydrochloride 50 mg (Lustral 1, P zer, Cairo, Egypt), paroxetine hydrochloride 20 mg (Seroxat 1, Smith- KlineBeecham pharmaceuticals, Brentford, England), sildena l citrate 50 mg (Viagra 1, P zer, New York, USA) and to use the pause-squeeze technique according Masters and Johnson. 2 The drugs were administered as needed 3 to 5 hours before planned intercourse and not more than twice a week. The pause-squeeze technique was used during intercourse. Each patient was randomly assigned to receive any of the available treatments as the rst treatment and a sequence of treatment regimens. The assignment was unknown by the patient or the investigator. Subjects were asked to complete a questionnaire, which we designed, that contained items such as intravaginal ejaculation latency time of the last 2 consecutive experiences of intercourse, frequency of intercourse and possible side effects. Intravaginal ejaculation latency time was measured by the patient using a watch. Anxiety measurement was assessed using an Arabic questionnaire (0 ± 30 scores) designed by Shaheen and Elrakhawy 12 with higher scores indicating greater anxiety. We used the rst 9-items (0 ± 30 scores) of the sexual satisfaction questionnaire (patient version) designed by Althof et al 13 for measuring the degree of sexual satisfaction among erectile dysfunction patients. We found that the rst 9 items are applicable to premature ejaculation. Higher scores indicate greater satisfaction. This questionnaire was translated into Arabic by two independent translators. All measurements were obtained before treatment, after each treatment and after each washout period. Twenty healthy men who reported a sexual history free of symptoms of premature ejaculation acted as a control group. The control group was subjected to measurement of anxiety score. After the completion of the study, the unblinded code revealed that 7 patients began with clomipramine, 6 with sertraline, 6 with paroxetine, 6 with sildena l and 6 patients began with the pause-squeeze technique. Statistical analysis The variables in this study were statistically processed using SPSS program for windows, standard version, release 8.0. The data were subjected to Kolmogorov-Smirnov one sample test, to test for normal distribution. This test showed that all the outcome variables were nonparametric. Nonparametric statistical tests were used to assess differences in the measurements. Friedman's two-way analysis of variance was used for comparison between all the treatment periods. Wilcoxon signed rank test was used for evaluation of measures between baseline and after each treatment and also between each two treatment periods. The relationship between parameters was quanti ed by using the Spearman rank correlation coef cient. Chisquare and Fisher exact tests were used for comparison of the incidence of side effects among different treatments. Student's t-test and Chi-square test were used for comparison between the study group and control group. A two-tailed P-value < 0.05 was considered signi cant. Results Table 1 shows the baseline characteristics of the patient and control groups. Of 31 patients, 25 (80.6%) completed the 5 treatment periods. The 6 patients who dropped out of the study used between 1 and 4 modalities (mean 3). Reasons for dropout included lack of ef cacy (clomipramine 2, sertraline 2, paroxetine 2 and the pause-squeeze technique 2), side effects (sildena l 2) and lack of ef cacy plus side effects (clomipramine 1). Before treatment the anxiety score was signi cantly higher among patients in comparison with the control group (Table 1). After 4-week treatments with clomipramine,
Table 1 Baseline characteristics of patients and control group Patients n ˆ 31 Control n ˆ 20 Signi cance test Age (y) Range 27 ± 42 25 ± 46 Mean s.d. 34.09 4.29 34.8 6.44 t ˆ 0.47 0.54 Median 34.0 33.5 Duration of marriage (y) Range 1.5 ± 10.0 1 ± 10 t ˆ 0.39 0.39 Mean s.d. 3.6 2.1 3.4 2.1 Median 3.0 3.0 Education level High 18 (58.1%) 11 (55.0%) w 2 ˆ 0.047 0.83 Middle 13 (41.9%) 9 (45%) Anxiety score (Baseline) Range 5 ± 25 1 ± 9 t ˆ 6.54 0.0001 Mean ˆ s.d. 12.7 5.8 3.7 2.7 Median 12 3.0 sertraline, paroxetine, sildena l and the pausesqueeze technique, the median intravaginal ejaculation latency time was signi cantly increased from the pretreatment median of 1 minute to 4 minutes, 3 minutes, 4 minutes, 15 minutes and 3 minutes respectively (Wilcoxon z ˆ 7 4.54, 74.63, 74.71, 74.63 and 74.55, respectively, all P < 0.0001). According to Friedman's test, treatment with sildena l caused a signi cant increase in the median IVELT, median sexual satisfaction score and median IVELT during the washout period (Table 2). The most effective treatment in prolongation of IVELT was sildena l in 28 patients (90.3%), followed by paroxetine (80.6%), sertraline (71.2%), clomipramine (71%) and the pause-squeeze technique (54.8%). Clomipramine, sertraline and paroxetine were more or less equivalent to each other in terms of ejaculation latency time and sexual satisfaction score (all P > 0.05). Paroxetine was found to be superior to the pause-squeeze technique in terms of ejaculatory latency and sexual satisfaction score (Wilcoxon zˆ 72.05, P ˆ 0.04, zˆ 72.24, P ˆ 0.025, respectively). Sexual satisfaction scores showed statistically signi cant positive correlation with IVELT during all treatment periods (all P ˆ 0.01). We noted P signi cant positive correlation (r ˆ 0.666, P ˆ 0.01) between anxiety score and IVELT during treatment with the pause-squeeze technique. Moreover, we found signi cant negative correlation between anxiety score and sexual satisfaction score during treatment with the pause-squeeze technique (r ˆ 70.547, P ˆ 0.01), clomipramine (r ˆ 70.381, P < 0.05), sildena l (r ˆ 70.573, P ˆ 0.01). Table 3 displays overall incidence and types of the reported side effects for each treatment. No adverse effects on sexual function were noted. No statistically signi cant differences were found with respect to side effects of treatment between different drugs. Most of the side effects were mild to moderate in severity. Discussion In the present study, treatment with clomipramine, sertraline, paroxetine, sildena l and pause-squeeze technique as needed resulted in a statistically signi cant and clinically relevant delay of intravaginal ejaculation latency time in patients with primary premature ejaculation. It is well established that the improvement of ejaculatory delay after as Table 3 Side effects of different drugs Sertraline Paroxetine Sildena l Clomipramine No. patients with side effects (%) 3 (10.3%) 5 (17.2%) 5 (17.9%) 7 (25%) P-value* 0.27 0.69 0.75 No. of side effects Dry mouth 2 3 Anorexia 1 Nausea 1 1 1 Headache 2 Flushing 2 Drowsiness 1 1 Sleepiness 2 Nasal congestion 1 Yawning 2 *Compared to incidence of side effect with clomipramine. 43 Table 2 Comparison between different modalities Baseline Clomipramine Sertraline Paroxetine Sildena l Squeeze technique Friedman w 2 df P Anxiety score Median 12 11 11 9 8 12 (Range) (5 ± 25) (4 ± 22) (5 ± 22) (5 ± 23) (4 ± 15) (5 ± 21) 55.15 5 0.0001 IVELT (min) Median 1 4 3 4 15 3 (Range) (0.5 ± 1.5) (1 ± 8) (1 ± 10) (2 ± 10) (5 ± 30) (1 ± 7) 92.53 5 0.0001 Sexual satisfaction score Median 11 10 12 30 6 (Range) (0 ± 25) (0 ± 31) (0 ± 29) (17 ± 34) (0 ± 22) 57.87 4 0.0001 IVELT during washout periods (min) Median 1 1 1 1.75 1 (Range) (0.5 ± 1.5) (0.5 ± 2) (0.5 ± 2) (0.5 ± 8) (0.5 ± 1.5) 32.52 5 0.0001
44 needed treatment with clomipramine, sertraline and paroxetine is attributed to central inhibition of serotonin reuptake, 6 as it has been suggested that the serotoninergic system has inhibitory in uence on ejaculation. 5 This improvement associated with antidepressant use cannot be ascribed to a decrease in anxiety score, as this did not occur. Similar ndings, dosages and methods of administration were reported. 4,9,14,15 On the other hand, the excellent success associated with sildena l use could be attributed to three possible mechanisms. The rst is reduction in performance anxiety, because of signi cant decrease in anxiety score during treatment and this could explain the signi cant negative correlation between anxiety score and sexual satisfaction score. The second possible explanation is that sildena l may maintain erection and increase the erection time, and ejaculation latency time was reported to be dependent on erection time. 6 So, it is anticipated that sildena l could increase the ejaculation latency time. The last mechanism is a possible central effect. Little is known about the intensity of stimulus required to induce ejaculation or the in uence of the cerebral cortex on the ejaculation re ex, although the peripheral neural pathway involved in ejaculation is fairly well understood. Phosphodiesterase enzymes were shown to be present in the rat brain, 17 and nitric oxide, the intracellular messenger of sildena l, was reported to perform numerous physiological functions, one of which is neural communication in the brain. 18 In addition, nitric oxide-stimulated guanylyl cyclase activity within a cellular environment is more complex than previously assumed and associated with diversity of cellular responses. 19 Hence it is possible that sildena l could have a central effect. Furthermore, it has been suggested that nitric oxide activity in the medial preoptic area tonically inhibits ejaculation by decreasing sympathetic tone. 20 Behavioral therapy, such as the Semans pause maneuver 1 which was modi ed to the pausesqueeze technique by Masters and Johnson, 2 is considered the gold standard for treatment of premature ejaculation. 21 While these techniques are harmless and usually a painless treatment modality and have been fairly successful at rates of 60 to 95%, 22 they require partner cooperation and usually the patients fail to maintain gains for the longterm. 10 In our experience, the pause-squeeze technique was associated with the lowest success rate (54.8%) compared with other modalities. In addition, the technique was associated with lower sexual satisfaction score and lower ejaculation latency in comparison with sildena l and paroxetine, but comparable with clomipramine and sertraline. Contributing factors for this lower success may be that some patients are embarrassed to involve their partners in the technique and others nd that it is time consuming and does not work effectively. Other authors have found that treatment with sex therapy led to improvement for a minority of patients. 10 Our data documented signi cant positive correlation between ejaculation latency and sexual satisfaction score during different treatment periods. Since the higher sexual satisfaction could re ect an increase in the control over ejaculation, this nding is especially interesting, in view of the fact that ejaculation latency and control tend to be related. 23 In view of the tendency of premature ejaculation to relapse following discontinuation of the three antidepressants and the pause-squeeze technique, because of the mean half-lives of clomipramine, sertraline and paroxetine are about 24, 26 and 21 hours, respectively. 24,25 In addition, behavioral techniques had not always been successful in the longterm. 10 Although the half life of sildena l is three to ve hours 11 a signi cant delay of ejaculation latency time during the washout period in 29% of patients was surprising. This improvement could be due to reduction in performance anxiety induced by initial successful sildena l treatment. Three limitations of the present study must be addressed. First, the sexual satisfaction instrument used 13 can not be applied before treatment to give the baseline satisfaction because it is composed of 9 items analysing all aspects of sexual satisfaction related to treatment. However, the questionnaire is valuable in comparative and controlled studies. 13 Moreover, it must be remembered that men who participate in clinical trials are extremely motivated and as such have a very low satisfaction level. For example, the range of sexual satisfaction scores during treatment with the 3 antidepressants and the pause-squeeze technique started from zero. The second limitation is that the patients gave a retrospective assessment of the mean ejaculation latency before treatment, after each treatment and after the washout period. It is dif cult for study participants to recall events in the past. It would be better to record ejaculation latency time at each coitus. However by asking the patients to report speci cally on the last 2 consecutive experiences of intercourse, the validity of the assessment of latency became more reliable. The last limitation is the lack of a placebo control period. This is because our work included two different modalities (drugs vs behavioral) and it is unacceptable to include a placebo control period of treatment. Furthermore, in crossover studies the patients may serve as their own control. 26 Moreover a placebo or untreated period may not be appropriate for diseases for which there is proven effective treatment or a well-established standard treatment. 27 It is known that selective serotonin reuptake inhibitors have a more favorable side effect pro le than clomipramine either when given on a daily basis 5,8 or on demand. 4,9,15 This is because of selective inhibition of serotonin reuptake and mini-
mal effect on other monamines. 14 Our results revealed that clomipramine showed the highest incidence of side effects, although the differences were not statistically signi cant. In general, the current study showed that the on demand use of clomipramine, sertraline and paroxetine is associated with mild and low incidence of side effects when compared with continuous administration as reported by other investigators. 5,6,8 In conclusion, this study highlighted some differences between the 5 modalities. Clomipramine, sertraline and paroxetine were equivalent to each other in terms of ef cacy and safety when used as needed; but sildena l was decidedly the most effective, demonstrating its tolerability and proving that it is a valid alternative for treatment of premature ejaculation. This study demonstrates also that paroxetine is superior to the pause-squeeze technique in terms of ejaculation latency and sexual satisfaction. Our data justify further extended studies to con rm the ef cacy of sildena l in the treatment of premature ejaculation and to explore its possible mechanisms of action. References 1 Semans JH. Premature ejaculation: a new approach. South Med J 1956; 49: 353 ± 357. 2 Masters W, Johnson V. Human Sexual Inadequacy. Little Brown & Co: Boston, 1970, pp 92 ± 115. 3 Choi HK et al. Clinical study of ss-cream in patients with lifelong premature ejaculation. Urology 2000; 55: 257 ± 261. 4 Haensel SM, Rowland DL, Kallan KT. Clomipramine and sexual function in men with premature ejaculation and controls. J Urol 1996; 156: 1310 ± 1315. 5 Kim SC, Seo KK. Ef cacy and safety of uoxetine, sertraline and clomipramine in patients with premature ejaculation: a double blind placebo controlled study. J Urol 1998; 159: 425 ± 427. 6 Balon R. Antidepressants in the treatment of premature ejaculation. J Sex Marital Ther 1996; 22: 85 ± 96. 7 Fein RL. Intracavernous medication for treatment of premature ejaculation. Urology 1990; 35: 301 ± 303. 8 Girgis SM, EL Haggar S, EL Hermouzy S. A double blind trial of clomipramine in premature ejaculation. Andrologia 1982; 14: 364 ± 368. 9 Kim SW, Paick JS. Short-term analysis of the effects of as needed use of sertraline at 5 pm for the treatment of premature ejaculation. Urology 1999; 54: 544 ± 547. 10 Metz ME et al. Premature ejaculation: a psychophysiological review. J Sex Marital Ther 1997; 23: 3 ± 23. 11 Padma-Nathan H, Shabsigh R. Sildena l citrate (Viagra 1 ): a review. AUA update series 1999; 18: 274 ± 280. 12 Shaheen O, Elrakhawy Y. ABC of Psychiatry, part I. EL Ahram press; Cairo, 1971, pp 55 ± 65. 13 Althof SE et al. EDITS: development of questionnaires for evaluating satisfaction with treatment for erectile dysfunction. Urology 1999; 53: 793 ± 799. 14 Strassberg DS et al. Clomipramine in the treatment of rapid [premature] ejaculation. J Sex Marital Ther 1999; 25: 89 ± 101. 15 MacMahon CG, Touma K. Treatment of premature ejaculation with paroxetine hydrochloride as needed: 2 single-blind placebo controlled crossover studies. J Urol 1999; 161: 1826 ± 1830. 16 Kaeya Y, Deguchi A, Yokota Y. Analysis of measured values of ejaculation time in healthy males. J Sex Marital Ther 1997; 23: 25 ± 28. 17 Kotera J, Fujishige K, Omori K. Immunohistochemical localisation of cgmp-binding cgm-speci c phosphodiesterase (PDE5) in rat tissues. J Histochem Cytochem 2000; 48: 685 ± 693. 18 Garthwaite J, Boulton CL. Nitric-oxide signaling in the central nervous system. Ann Rev physiol 1995; 57: 683 ± 706. 19 Bellamy TC, Wood J, Goodwin DA, Garthwaite J. Rapid desensitization of the nitric oxide receptor, soluble guanylyl cyclase, underlies diversity of cellular cgmp responses. Proc Natl Acad Sci USA 2000; 97: 2928 ± 2933. 20 Pfaus JG. Neurobiology of sexual behavior. Curr Opin Neurobiol 1999; 9: 751 ± 758. 21 Seftel AD, Althof SE. Premature ejaculation. In: Mulcahy JJ (ed). Diagnosis and Management of Male Sexual Dysfunction. Igaku-Shoin: New York, 1997, pp 196 ± 203. 22 Hawton K, Catalan J. Prognostic factors in sex therapy. Behav Res Ther 1986; 24: 377 ± 385. 23 Rowland DL, Strassberg DS, degouveia Brazao CA, Slob AK. Ejaculatory latency and control in men with premature ejaculation: an analysis across sexual activities using multiple sources of information. J Psychosom Res 2000; 48: 69 ± 77. 24 Potter WZ, Manji Hk, Rudorfer MV: Tricyclics and tetracyclics. In: Schatzberg AF and Nemeroff CB, (eds). Psychopharmacology. American Psychiatric Press, Inc: Washington, DC, 1994, pp 141 ± 160. 25 Majeroni BA, Hess A. The pharmacologic treatment of depression. J Am Board Fam Pract 1998; 11: 127 ± 139. 26 Bigby M, Gadenne A. Understanding and evaluating clinical trials. J Am Acad Dermatol 1996; 34: 555 ± 590. 27 Rothman KJ, Michels KB. The continuing unethical use of placebo controls (comments). New Engl J Med 1994; 331: 394 ± 398. 45