J Psychopharmacol OnlineFirst, published on September 18, 2008 as doi: /

Transcription

1 J Psychopharmacol OnlineFirst, published on September 18, 2008 as doi: / Original Papers Better sexual acceptability of agomelatine (25 and 50 mg) compared with paroxetine (20 mg) in healthy male volunteers. An 8-week, placebocontrolled study using the PRSEXDQ- SALSEX scale Journal of Psychopharmacology 00(00) (2008) British Association for Psychopharmacology ISSN SAGE Publications Ltd, Los Angeles, London, New Delhi and Singapore / AL Montejo Servicio de Psiquiatría, Hospital Universitario de Salamanca, Salamanca, Spain. N Prieto Servicio de Psiquiatría, Hospital Universitario de Salamanca, Salamanca, Spain. A Terleira Clinical Pharmacology Studies Unit, Clinical Pharmacology Service, Hospital Clinico San Carlos, Madrid, Spain. J Matias Servicio de Psiquiatría, Hospital Universitario de Salamanca, Salamanca, Spain. S Alonso Clinical Pharmacology Studies Unit, Clinical Pharmacology Service, Hospital Clinico San Carlos, Madrid, Spain. G Paniagua Servicio de Psiquiatría, Hospital Universitario de Salamanca, Salamanca, Spain. S Naval Clinical Pharmacology Studies Unit, Clinical Pharmacology Service, Hospital Clinico San Carlos, Madrid, Spain. DG Parra Servicio de Psiquiatría, Hospital Universitario de Salamanca, Salamanca, Spain. C Gabriel Institut de Recherches Internationales Servier, Courbevoie, France. E Mocaër Institut de Recherches Internationales Servier, Courbevoie, France. A Portolés Clinical Pharmacology Studies Unit, Clinical Pharmacology Service, Hospital Clinico San Carlos, Madrid, Spain. Abstract Sexual dysfunction (SD) is a common and underestimated effect of antidepressants. Healthy volunteers are the most adequate group to study this adverse event avoiding influence of depression itself. Sexual acceptability of agomelatine (a melatonergic agonist and 5HT 2C antagonist) paroxetine and placebo by using the Psychotropic-Related Sexual Dysfunction Salamanca Sex Questionnaire (PRSEXDQ-SALSEX) was explored. A total of 92 healthy male volunteers were randomised to agomelatine (25 or 50 mg), paroxetine 20 mg or placebo for 8 weeks. SD, defined as at least one sexual impairment in one of the following PRSEXDQ-SALSEX items (decreased libido, delayed orgasm/ejaculation, anorgasmia/no ejaculation and erectile dysfunction), was evaluated at baseline and after 2, 4 and 8 weeks. At the last post-baseline assessment, SD was significantly lower in each agomelatine group (22.7% on 25 mg and 4.8% on 50 mg) than in the paroxetine group (85.7%; p < ). In the placebo group, 8.7% of volunteers reported a SD. The percentages of volunteers with moderate or severe SD were 4.5% for agomelatine 25 mg, 4.8% for agomelatine 50 mg, 61.9% for paroxetine 20 mg and 0% in the placebo group (p agomelatine versus paroxetine). There is a much lower risk of having SD with agomelatine than paroxetine in healthy male volunteers, which confirms the better sexual acceptability profile of agomelatine compared with the SSRIs. Key words agomelatine; antidepressant; PRSexDQ; salsex; scales; sexual dysfunction; SSRIs Corresponding author: AL Montejo, Hospital Universitario de Salamanca, Servicio de Psiquiatría, Av Comuneros 27, Salamanca 37003, Spain. amontejo@usal.es

2 2 Better sexual acceptability of agomelatine versus paroxetine Introduction Most of the currently available antidepressants produce sexual dysfunction (SD) in men and women (Rosen, et al., 1999; Montejo, et al., 2001; Clayton, et al., 2002; Delgado, et al., 2005; Kennedy, et al., 2006; Clayton, et al., 2007), and affect all phases of sexual activity, by decreasing desire, arousal, orgasm and ejaculation. Such side effects affect the patient s quality of life, can lead to therapeutic non-compliance and often interfere with recovery from a depressive episode (Roose, 1999; Rothschild, 2000; Zajecka, 2000; Montejo, et al., 2001). The incidence of SD is strongly dependent on the antidepressant used (Clayton and Montejo, 2006). A higher incidence (between 50 and 70%) is observed with drugs with a high profile of 5-HT reuptake blockade such as selective serotonin reuptake inhibitors (SSRIs), monoamine oxidase inhibitors (MAOIs), tricyclic agents or SNRIs. Drugs that predominantly increase noradrenaline or dopamine uptake and the 5-HT 2 receptor blockers appear to have few negative effects on sexual functioning (Clayton, et al., 2002; Montejo, et al., 1999). Agomelatine is an antidepressant with melatonergic agonist and 5-HT 2C antagonist properties that has shown an efficacy (Loo, et al., 2002; Kennedy and Emsley, 2006; Olié and Kasper, 2007) at least as similar as available antidepressants for treating patients with major depressive disorder (MDD) (Lemoine, et al., 2007; Kasper, et al., 2008; Kennedy, et al., 2008). No deleterious effect on sexual function has been observed for agomelatine through the whole development plan. Sexual emergent adverse events frequency was low and similar between patients treated with agomelatine and placebo. Sex scales (ASEX, SexFX) carried out during the short-term efficacy studies, confirmed the good sexual tolerability for patients taking agomelatine (Kennedy, 2005). For agomelatine as for other antidepressants, however, most of the studies evaluating antidepressant effects on sexuality have been performed in depressed patients, that is, in conditions where the therapeutic effect of drugs on mood can partially mask concomitant negative effects on sexual functioning related to the pharmacodynamic effect. However, depression itself could deteriorate sexual relationships before the antidepressant intake, and antidepressant-related SD is quite difficult to be measured unless to select only patients without SD from the beginning. Waldinger, et al. (1994) have accurately measured the ejaculation delaying in non-depressed men with lifelong premature ejaculation and receiving SSRIs and other antidepressants evaluating the intravaginal ejaculation latency time (IELT) by the use of a stopwatch showing that SSRIs, mainly paroxetine, provoked strong ejaculation delay (Waldinger, et al., 2004). On the contrary, nefazodone and mirtazapine did not delay ejaculation (Waldinger, et al., 2001a, 2003) because of their antagonism on 5-HT 2C receptors. To date, only few studies have evaluated the impact of antidepressant therapy in a population without depressive symptoms (Kennedy, et al., 1996; Nafziger, et al., 1999). More studies in nondepressed men and women are certainly required to provide the clearest evidence of sexual side effects exclusively related to the drug. Therefore, the present double blind, comparative and placebo-controlled study was designed 1) to assess the sexual acceptability of agomelatine (doses 25 and 50 mg) versus paroxetine (20 mg) treatment for 8 weeks in healthy male volunteers, using placebo to assess assay sensitivity (compared with paroxetine) and 2) to show the absence of hormonal modifications and to provide safety data. Contrary to studies performed by Waldinger, et al. (1997), this study is not aimed to evaluate the effects of agomelatine on premature ejaculation. To this aim, the validated Psychotropic-Related Sexual Dysfunction Questionnaire (PRSEXDQ-SALSEX) (Montejo, et al., 2000, 2008) and the International Index of Erectile Function (IIEF) (Rosen, et al., 1997) were used. The sensitivity of the trial was assessed by using a placebo arm and, as comparator, the SSRI paroxetine (20 mg), an antidepressant with wellknown impairment of sexual function in depressed patients (Montejo, et al., 1996, 2001; Rosen, et al., 1999). Methods This phase I study used a randomised, double blind, 4-group (agomelatine 25 and 50 mg, paroxetine 20 mg and placebo) parallel design, in healthy male volunteers, and was conducted in two clinical centres in Spain in agreement with the principles of Good Clinical Practice and the Declaration of Helsinki. The relevant local Ethics Committees approved the protocol, and all volunteers freely gave their written informed consent before their selection in the study. Volunteers In order to be included, healthy male volunteers aged 18 30, had to be non-smokers or moderate smokers (<10 cigarettes/ day) with a body mass index between 20.0 and 30.0, and had to have a stable relationship for at least 6 months, with a regular (at least 1/week) and satisfactory sexual activity. To be eligible, volunteers had also to have a score null at the PRSEXDQ-SALSEX rating scale (scoring 0 15; 0 = no SD and 15 = maximum SD), a total score of IIEF > 60 and anxiety and depression scores 7 on the Hospital Anxiety Depression Scale. Volunteers also had to have a negative drug screening (amphetamine, benzodiazepines, cocaine, opioids, cannabis) and a negative breath alcohol test. Any use of sedative hypnotics, including benzodiazepines, zolpidem or zopiclone had to be discontinued at least 14 days before entering the study. Any other psychotropic medication, including antidepressants and anti-psychotics had to be discontinued at least 1 month preceding the selection. A foreseeable need for treatment with drugs known to induce or inhibit CYP 1A2, CYP 2C9 and 2C19 (fluvoxamine, cimetidine, fluoxetine, paroxetine, sertraline, phenytoin, barbiturates, rifampicin, lidocaine, phenacetin, mexiletine, quinolones, fluconazole, flavonoids, omeprazole, lansoprazole, pan-

3 Better sexual acceptability of agomelatine versus paroxetine 3 toprazole) was a non-selection criterion. These drugs had also not to be taken within the month before inclusion. Treatment with drugs that could interfere with sexual hormones (hormonal treatment, dopaminergic agonist and antagonists, codeine and opioid analgesics) or treatments with drugs capable of interfering with sexual intercourse (β-blockers, antihypertensive, hypo-cholesterolaemic and psychotropic drugs) were also not authorised. No other medications were allowed concomitantly during the study except paracetamol (1 g/day) when necessary. Treatments Two doses of agomelatine were tested, 25 and 50 mg, versus paroxetine 20 mg (10 mg during the first week of treatment) and placebo was used as a study validator via the comparison paroxetine placebo. During the study, treatments were taken twice daily by oral route: one red capsule (containing either placebo or paroxetine) in the morning during the breakfast and one yellow capsule in the evening (08:00 p.m.), containing either placebo or agomelatine at 25 or 50 mg (i.e., containing one or two tablets of agomelatine 25 mg). To assess drug compliance, agomelatine salivary and paroxetine urinary concentrations were assessed by HPLC MS/MS assay during the study. All analytical results were transferred to the database after database lock. Study treatments were of identical appearance (whatever the treatment arm and the dosage) to protect the blinding. No case of unblinding occurred during the study. The blind was broken after database lock. Study design Volunteers first underwent a 2 4 weeks selection period without treatment and then were randomised to one of the four treatment arms, agomelatine 25 mg, agomelatine 50 mg, paroxetine 20 mg or placebo. The randomisation list was generated in blind and was based on a permuted-block (non-adaptive) randomisation without stratification. Complete permutation blocks were sent to each centre, and at the inclusion visit, the investigator allocated to each volunteer a therapeutic unit number in the chronological order of inclusion in the centre. Visits were performed for selection and inclusion, then weekly from W1 to W8 during the double-blind treatment period. A follow-up visit was planned 7 days after the treatment discontinuation. The primary acceptability criterion was SD assessed by the PRSEXDQ-SALSEX (Montejo, et al., 2000). The PRSEXDQ- SALSEX consists of seven items pertaining to SD. The first item is a screening item to assess whether the patient has any sort of SD. The second item assesses whether the patient has spontaneously reported any SD to his or her physician. The items 3 7 assess five dimensions of SD according to severity or frequency: loss of libido (0 = nil, 1 = mild, 2 = moderate, 3 = severe), delayed orgasm or ejaculation (0 = nil, 1 = mild, 2 = moderate, 3 = severe), anorgasmia or no ejaculation (0 = never, 1 = occasionally, 2 = often, 3 = always), erectile dysfunction in men/vaginal lubrication dysfunction in women (0 = never, 1 = occasionally 2 = often, 3 = always) and patient s tolerance of the SD (0 = no SD, 1 = good, 2 = fair, 3 = poor). Only items 3 7 account for the total score of the PRSEXDQ-SALSEX (total score ranges from 0 to 15). SD was defined as at least one sexual impairment in one of the four following items of PRSEXDQ-SALSEX: decreased libido (item 3), delayed orgasm/ejaculation (item 4), anorgasmia/no ejaculation (item 5) and erectile dysfunction (item 6). A sexual impairment corresponded to a score 1 for items 4, 5, 6 or a score 2 for item 3. Secondary sexual acceptability criteria included moderate or severe SD defined as at least one moderate or severe sexual impairment in one of the PRSEXDQ-SALSEX items 3 6, that is, with a score = 3 for item 3 or a score 2 for items 4, 5 or 6. The following were also considered as secondary sexual acceptability criteria: the sustained SD defined as a dysfunction which appears and persists until the end of study; dysfunction relative to each individual PRSEXDQ-SALSEX item and the PRSEXDQ-SALSEX total score (sum of items 3, 4, 5, 6 and 7). Secondary acceptability criteria also included the 15-item total score of the self-rating questionnaire IIEF (Rosen, et al., 1997), which measures erectile function. The primary and secondary acceptability criteria were assessed at selection, inclusion, W2, W4, W8 and in case of withdrawal. Safety evaluations included collection of adverse events and measurements of blood pressure and heart rate at all visits. To determine any possible association between biological measures and SD, cortisol, prolactin and testosterone (total and free) plasma levels were determined at selection, inclusion, W2, W4, W8 and in case of withdrawal. Biological tests were performed at selection and follow-up visits. For each adverse event, emergence (defined as a new or a worsening event under study treatment), intensity (e.g., mild, moderate, severe), seriousness (serious, non-serious) and causality (related, not related to treatment) were considered. Statistical analysis Statistical analyses were performed on SAS software, version 8.2 (SAS INSTITUTE, SAS Institute Inc. 100 SAS Campus Drive Cary, NC USA). The percentage of volunteers with SD was compared between each agomelatine dose and paroxetine (at the same type one error of 5%) for last post-baseline value using a Cochran-Mantel-Haenszel (CMH) test adjusted for centre (main analysis) and a chi-square test (sensitivity analysis). Comparisons were also performed at each visit using the CMH test. Similar analyses comparing paroxetine and placebo were carried out to assess assay sensitivity. The same adjusted analyses were performed for moderate or severe SD, SD relative to each PRSEXDQ-SALSEX individual item (for last post-baseline value and at each visit), and for

4 4 Better sexual acceptability of agomelatine versus paroxetine sustained SD (until the end of study). Comparisons were performed using the CMH test. Moreover, a comparison between each dose of agomelatine and placebo was also performed on the SD and sustained SD using an adjusted CMH test. The PRSEXDQ-SALSEX mean total score at each visit and last post-baseline value, and IIEF mean change from baseline to last post-baseline assessment were compared between each agomelatine dose and paroxetine and between paroxetine and placebo using a two-sided Student s t-test. Results A total of 94 volunteers were selected by two centres in Spain, of which 92 were included and randomly assigned to one of the four treatment groups (23 volunteers per group). Eighty-eight volunteers completed the study and four discontinued (one volunteer in the agomelatine 25 mg group because of bad compliance, three volunteers in the paroxetine group including one for pharyngotonsillitis not related to the study drug and two for personal reason). In randomised volunteers, the baseline demographic characteristics of the treatment groups were generally similar, with a mean ± standard deviation age of 25.0 ± 3.5, 24.9 ± 2.4, 24.7 ± 2.3 and 23.8 ± 2.7 years old in the agomelatine 25 and 50 mg, paroxetine and placebo groups, respectively. The sexual acceptability and hormonal profile at baseline was similar in all treatment groups. No clinically relevant difference between groups was observed on the mean treatment duration (54.9 ± 9.2 days). The compliance was very satisfactory with 90/92 volunteers reporting a compliance 70%. Sexual acceptability Sexual acceptability was analysed in the Sexual Acceptability Set defined as all included volunteers having a SD assessment after at least 2 weeks and compliant to the treatment (according to saliva or urine samples in active treatment groups). This set was constituted of 87 volunteers (94.6% of included volunteers). Five volunteers were excluded from this set because of no available SD assessment after 2 weeks (two volunteers in the paroxetine group and one in the agomelatine 25 mg group) or poor salivary compliance (two volunteers in the agomelatine 50 mg group). The percentage of volunteers with a SD was significantly lower in each agomelatine group than in the paroxetine group at each visit from W2 to W8 (Figure 1). Similarly, at last postbaseline value, 5 of 22 volunteers in the agomelatine 25 mg group (22.7%) and 1 of 21 volunteers in the agomelatine 50 mg group (4.8%) versus 18 of 21 volunteers in the paroxetine group (85.7%) had a SD according to the PRSEXDQ- SALSEX (p < for each comparison, adjusted CMH test). Compared with paroxetine, the adjusted risk of SD at last assessment was significantly reduced by 74% in the agomelatine 25 mg group (RR = 0.26, 95% CI [0.12; 0.58]), and by 94% in the agomelatine 50 mg group (RR = 0.06, 95%CI [0.01; 0.38]). Moreover, the percentage of volunteers with a SD at last assessment was significantly higher in the paroxetine group than in the placebo group (2 of 23 volunteers, 8.7%) (p < ), which provides the evidence of the assay sensitivity. Compared with placebo, the adjusted risk of SD was significantly increased by 9.4 (RR = 9.4, 95%CI [ ]) at last post-baseline assessment. All these results were confirmed by the unadjusted analysis. The results in favour of agomelatine were also observed at each visit from W2 with a significant (adjusted) risk reduction by 85% at W2, 78% at W4, and 74% at W8 for agomelatine 25 mg and by 92% at W2, and 94% at W4 and W8 for agomelatine 50 mg. Statistically significant differences between each dose of agomelatine and placebo group were underscored neither at last post-baseline assessment nor at each visit from W2. The adjusted relative risk of agomelatine versus placebo was 2.59 for agomelatine 25 mg (95% CI [0.52; 12.96]) and 0.54 for agomelatine 50 mg (95% CI [0.05; 5.44]) at last assessment. Moderate or severe SD The results of the comparison of the percentage of volunteers with a moderate or severe SD between each dose of agomelatine and paroxetine at each time and for the last post-baseline assessment are presented in Figure 2. At last assessment, only one volunteer (4.5%) in the agomelatine 25 mg group and one volunteer (4.8%) in the agomelatine 50 mg group reported a moderate or severe SD. In the paroxetine group, 13 volunteers (61.9%) had at least one moderate or severe dysfunction. By comparison with paroxetine, the adjusted risk of having a moderate or severe SD on agomelatine was significantly reduced by 93% in the agomelatine 25 mg group at last assessment (RR = 0.07, 95% CI [0.01; 0.51], Figure 1 Percentage of volunteers with a sexual dysfunction (SD). Evolution by visit. At each visit from W2 to W8, volunteers receiving paroxetine reported significantly more SD than those of the placebo group. On the contrary, volunteers receiving either dose of agomelatine experienced significantly less SD than volunteers in the paroxetine group. CMH test adjusted for centre: ***p < (agomelatine 25 or 50 mg versus paroxetine), p < 0.01; p < (paroxetine versus placebo).

5 Better sexual acceptability of agomelatine versus paroxetine 5 p < ) and by 92% in the agomelatine 50 mg group (RR = 0.08, 95% CI [0.01; 0.54], p = ). No moderate or severe SD was reported in the placebo group (0/23). The frequency of moderate or severe SD was significantly higher on paroxetine than on placebo (p < ). The percentage of volunteers with moderate or severe SD showed no relevant differences between each agomelatine dose and placebo. Similar results were observed at each visit from W2. Sustained SD Throughout the study period, SD were sustained in five volunteers (22.7%) in the agomelatine 25 mg group, one volunteer (4.8%) in the agomelatine 50 mg group, whereas 17 volunteers (81.0%) in the paroxetine group reported SD at each subsequent visit after first onset. Compared with paroxetine, the adjusted risk of having a sustained SD with agomelatine was significantly reduced by 72% at dose 25 mg (RR = 0.28, 95% CI [0.12; 0.63], p = ) and by 94% at dose 50 mg (RR = 0.06, 95% CI [0.01; 0.40], p < ). Throughout the study period, the percentage of volunteers with a sustained SD in the placebo group was 4.3% (1/23). Compared with placebo, the adjusted risk of having a sustained SD with paroxetine was significantly increased by 17.7 (p < ). The adjusted relative risk of sustained SD of agomelatine versus placebo was 5.17 for agomelatine 25 mg (95% CI [0.63; 42.66]) and 1.07 for agomelatine 50 mg (95% CI [0.07; 15.54]). Adjusted CMH test showed no statistically significant difference between both doses of agomelatine and placebo. 5-item total score of PRSEXDQ-SALSEX The mean ± standard deviation total score (ranging 0 = no SD to 15 = maximum SD) was significantly lower in both agomelatine groups compared with paroxetine at each visit and at last assessment (0.9 ± 2.0 for agomelatine 25 mg, 0.2 ± 0.9 for agomelatine 50 mg, compared with 5.2 ± 3.6 for paroxetine, p < , two-sided Student s t-test). The PRSEXDQ-SALSEX 5-item total score observed in the placebo group was 0.5 ± 1.2, close to those of the agomelatine groups. In the paroxetine group, the mean total score was significantly higher than in the placebo group (p < two-sided Student s t-test) (Figure 3). PRSEXDQ-SALSEX: dysfunction relative to each individual item For all PRSEXDQ-SALSEX items at last evaluation, results were statistically significantly in favour of both agomelatine doses compared with paroxetine group, except for item decreased libido for agomelatine 25 mg dose (Figure 4). The greatest difference in favour of agomelatine was noted for the dysfunction delayed orgasm/ejaculation, which was reported in 2 (9.1%) and 1 (4.8%) volunteers in the agomelatine 25 and 50 mg groups, respectively, versus 17 (81.0%) in the paroxetine group (p < for each comparison, adjusted CMH test). One dysfunction (4.3%) was reported in the placebo group; this finding was close to those of the agomelatine groups. On paroxetine, at last evaluation, the frequency of volunteers with a SD relative to each item was statistically significantly higher than on placebo, except for erectile dysfunction. IIEF questionnaire At last post-baseline assessment, the mean changes from baseline in the IIEF total score were close to null in volunteers given agomelatine (both doses) or the placebo, whereas the mean IIEF total score decreased by 7.8 ± 8.8 in the paroxetine group. The difference between each dose of agomelatine and paroxetine and between paroxetine and placebo was statistically significant (p = and p < on Figure 2 Percentage of volunteers with a moderate or severe sexual dysfunction (SD). Evolution by visit. At each visit, the frequency of moderate or severe SD reported by volunteers in the paroxetine group was significantly higher than in volunteers receiving placebo. On the contrary, rates of volunteers reporting one moderate to severe dysfunction were similar between both agomelatine groups and placebo group. Adjusted CMH test: *p < 0.05; **p < 0.01; ***p < (agomelatine 25 or 50 mg versus paroxetine) p < 0.05; p < (paroxetine versus placebo). Figure 3 The 5-item PRSEXDQ-SALSEX total score. Evolution by visit. At each visit, the mean PRSEXDQ-SALSEX total score was significantly higher in paroxetine group compared with placebo. On the contrary, total score was lower in both agomelatine groups compared with paroxetine. Twosided Student s t-test: **p < 0.01; ***p < (agomelatine 25 or 50 mg versus paroxetine), p < 0.05; p < (paroxetine versus placebo).

6 6 Better sexual acceptability of agomelatine versus paroxetine Figure 4 Sexual dysfunction as per each single PRSEXDQ-SALSEX s item: delayed orgasm/ejaculation (A), anorgasmia/no ejaculation (B), decreased libido (C) and erectile dysfunction (D). At last evaluation, all PRSEXDQ-SALSEX items had statistically significantly lower frequency in both agomelatine groups than in paroxetine group except for item decreased libido for agomelatine 25 mg dose. The greatest difference in favour of agomelatine was noted for the dysfunction delayed orgasms/ejaculation. On paroxetine, all items at last evaluation had a statistically significantly higher frequency than on placebo, except for erectile dysfunction. Adjusted CMH test: *p < 0.05; **p < 0.01; ***p < (agomelatine 25 or 50 mg versus paroxetine) p < 0.05; p < (paroxetine versus placebo). agomelatine 25 and 50 mg, respectively, and p = for the placebo paroxetine comparison). Safety In the safety set (N = 92), defined as all included volunteers who took at least one dose of study treatment, 91.3% of the volunteers in the agomelatine 25 mg group (21 of 23), 65.2% of the volunteers in the agomelatine 50 mg group (15 of 23), 91.3% of the volunteers in the paroxetine group (21 of 23) and 73.9% in the placebo group (17 of 23) reported at least one emergent adverse event (Table 1). Medication was stopped in one volunteer on paroxetine because of the occurrence of an emergent adverse event (pharyngotonsillitis) not related to the drug. The most common treatment-emergent adverse events reported in agomelatine groups were especially somnolence, which was experienced by 10 volunteers (43.5%) receiving agomelatine 25 mg, by 11 volunteers (47.8%) receiving agomelatine 50 mg, compared with 5 volunteers (21.7%) receiving paroxetine and 6 volunteers (26.1%) in the placebo group. Volunteers of the paroxetine group mainly experienced sexual disorders with a markedly higher incidence than on placebo and on agomelatine. No sexual-related adverse events were reported at follow-up visit, after 7 days treatment discontinuation showing that all volunteers recovered their previous normal sexual functioning. All emergent adverse events in the agomelatine or placebo groups were of mild or moderate intensity. In the paroxetine group, five volunteers (21.7%) had seven severe emergent adverse events, all but one were related to SD. In all treatment groups, somnolence and SD were considered as treatment related. There were no death and no serious adverse events reported during the study.

7 Better sexual acceptability of agomelatine versus paroxetine 7 Table 1 Summary of main adverse events results Primary SOC Preferred term a Agomelatine 25 mg (N = 23) Agomelatine 50 mg (N = 23) Paroxetine (N = 23) Placebo (N = 23) n % n % n % n % Volunteers with at least one emergent adverse event Volunteers withdrawn from the study because of adverse event Volunteers with serious adverse event Nervous system disorders Somnolence Headache Dizziness Gastrointestinal disorders Diarrhoea Nausea Sexual function disturbances have not been included as AE to avoid confusion, as they are evaluated by the other variables. a Most commonly affected system organ classes (>20% of volunteers), and most common emergent adverse events (at least three volunteers in any groups) more frequently reported in the active treatment groups than in the placebo group. As regards to hormone profiles, there were no clinically relevant changes in the mean plasma levels of prolactin, total cortisol, total and free testosterone during the study in the active treatment groups compared with the placebo group. Vital signs mean values remained stable from baseline to the last observation in all treatment groups. No clinically relevant change in mean values of biochemical and haematological parameters were also found. Discussion The present study is one of the few randomised clinical trials evaluating the sexual acceptability of an antidepressant (here agomelatine and paroxetine) in a population without depressed symptoms (Kennedy, et al., 1996; Nafziger, et al., 1999), to provide evidence of sexual side effects exclusively related to the pharmacodynamic effects of the above-mentioned antidepressants (without masking effects of mood improvement related to improvement of depression). The sexual acceptability of agomelatine is clear-cut and particularly superior to that of paroxetine. Healthy male subjects had a significantly better sexual acceptability (lower risk of having SD) on agomelatine 25 or 50 mg than on paroxetine 20 mg. The frequencies of SD with agomelatine 25 or 50 mg were low and, although the study was not specifically designed for a direct placebo-agomelatine comparison, frequency of SD with agomelatine was analogous to that of placebo. A robust difference in the sexual acceptability of both antidepressants, in favour of agomelatine, is observed from the second week of treatment, and is maintained subsequently. Although few SD events were observed on agomelatine at the dose of 25 mg, it is of note that they were of mild intensity, and not dosedependent as they were not found at the 50 mg dose. Accordingly, compared with paroxetine, volunteers from both agomelatine groups presented significantly less of moderate or severe SD, and of sustained SD. Consistently, the mean PRSEXDQ- SALSEX 5-item total score was significantly lower in both agomelatine groups than in the paroxetine group from the second week, and the difference was maintained until the end of the study period. At last assessment, the dysfunctions relative to each PRSEXDQ-SALSEX individual item (except decreased libido for agomelatine 25 mg) were significantly less frequent on agomelatine than on paroxetine. Moreover, at last evaluation, all PRSEXDQ-SALSEX items but erectile dysfunction had a significantly higher frequency on paroxetine than on placebo. It is generally assumed that after the antidepressant treatment onset, the chronological sequence of sexual alterations in men are: 1) delayed orgasm/ejaculation, 2) anorgasmia/no ejaculation (appearing in the first month after treatment) 3) decrease in libido and 4) erectile dysfunction (appearing and increasing after 1 3 months of the onset) (Montejo, et al., 1997, 2001; Clayton, et al., 2006). Therefore, it may be that an 8-week treatment with paroxetine is too short for the emergence of a decreased libido and an erectile dysfunction and a longer treatment should be needed. Moreover, in agreement with all these results, no changes from baseline in the mean IIEF total score were noted in volunteers given both doses of agomelatine (or the placebo), whereas the mean IIEF total score significantly decreased in the group of volunteers receiving paroxetine. In fact, no deleterious effect on sexual function has been observed throughout the entire development of agomelatine. The frequency of sexual emergent adverse events was always low and similar for patients treated with agomelatine or placebo. The analysis of ASEX scale (McGahuey, et al., 2000) applied during the short-term efficacy studies even showed that agomelatine induced less SD than placebo (3.0% vs 8.6%). Furthermore, a specific study using the SEXFX showed

8 8 Better sexual acceptability of agomelatine versus paroxetine that agomelatine 50 mg administered for 12 weeks to sexually active depressed patients offered a better sexual profile than venlafaxine 150 mg in remitters (Kennedy, et al., 2008). The present study confirms in healthy male volunteers that paroxetine itself is capable of causing high rates of SD, as already reported in depressed patients (Montejo, et al., 1996; Rosen, et al., 1999; Montejo, et al., 2001; Clayton and Montejo, 2006). The known deterioration of the sexual acceptability by paroxetine was observed on both PRSEXDQ-SALSEX and IIEF scales. The deleterious effects of paroxetine on sexual functions could be used in clinical practice for treatment of premature ejaculation (Waldinger, et al., 1994, 1997, 1998, 2001a, b, 2003; Ludovico, et al., 1996; Salonia, et al., 2002; Waldinger and Olivier, 2004; Waldinger, 2007), and, as expected, the greatest deleterious effect of paroxetine detected by the PRSEXDQ-SALSEX is a delayed orgams/ejaculation dysfunction (reported in 17 out of 21 volunteers (81%)). By comparison, agomelatine rarely delays orgasm/ejaculation in healthy male volunteers, which may be related to its 5-HT 2C antagonist action, as two other antidepressants sharing this property, mirtazapine and nefazodone, have been shown to not delay ejaculation as well in non-depressed healthy men with lifelong premature ejaculation (Waldinger, et al., 2001a, 2003). Using the PRSEXDQ-SALSEX scale, similar data were previously reported in patients with MDD (Montejo, et al., 1997, 2001). Moreover, an increase of the risk of drop-outs for patients with antidepressant-related SD, near to 40%, has been shown in patients under chronic treatment (Montejo, 2001), which underlies the importance of having well sexually tolerated treatments. Clinical safety in healthy volunteers given agomelatine 25 and 50 mg were consistent with that observed in depressive patients (Kennedy, 2007) and no unexpected adverse event was observed. Such data give additional support to the good tolerability of both doses of agomelatine. On both doses of agomelatine, volunteers mainly experienced somnolence (mild or moderate in intensity). Somnolence is a known effect with agomelatine, likely related to its agonist properties at MT1 and MT2 melatonergic receptors. However, the incidence of somnolence reported by MDD patients on agomelatine 25/50 mg is generally by far lower (approximately 3%) than in the present study, which may suggest a higher susceptibility for somnolence in healthy volunteers compared with depressed patients. In the paroxetine group, patients mainly reported sexual disorders. After treatment discontinuation, no sexual-related adverse events were reported anymore, which further supports the relationship with the drug. As concerns the hormone profile, no clinically relevant mean changes were found regardless of the treatment group. This is particularly interesting when considering the absence of significant changes of prolactin plasma levels in healthy volunteers. It has been published that chronic paroxetine can induce plasma prolactin increase in healthy volunteers (Cowen and Sargent, 1997), but this effect has been also reported to be only marginal in this population (Schlosser, et al., 2000). It has been hypothesised that hyperprolactinemia is one of the multiple mechanisms that may underlie the SD (Bhasin, et al., 2007). Hyperprolactinemia seems to be related mainly with decreased libido and erectile dysfunction as a consequence of subsequent hypogonadism, but there is no enough information about orgasm or ejaculation failure. However, the present results show that in striking contrast to antipsychotics (Smith, et al., 2002; Compton and Miller, 2002; Ahl, et al., 2004), SD induced by paroxetine, mainly orgasm and ejaculation problems, cannot be simply related to prolactin. Conclusions By using the PRSEXDQ-SALSEX and the IIEF, the present study reports a statistically significantly lower risk of having SD under agomelatine than under paroxetine treatment in healthy male volunteers, and confirms the pharmacodynamic impairment of sexual function induced by paroxetine compared with placebo. We can also conclude that the drug is a very important factor strongly associated to sexual impairment capabilities of some antidepressants, without the bias of the patient s mood state. Although it shall be recognised that a limitation of the present study is its focus on healthy male volunteers, altogether, our findings and those observed previously in depressed patients confirm the better sexual acceptability profile of agomelatine compared with the SSRIs. Acknowledgements We thank all the medical, nursing and administrative personnel of the Clinical Pharmacology Study Unit, Hospital Clínico San Carlos, and the Psychiatry Service, Hospital Universitario Salamanca, with special mention to the nursing team: Vegas C, Paniagua S, Rojas A, Marcos ML, Martin C and Ballesteros C, for their enthusiastic cooperation and their dedication to this project. This study was supported by Servier. Declared disclosures Dr AL Montejo has been a speaker for Lilly, Astra Zeneca, Bristol- Myers Squibb, Servier, Glaxo-SmithKline, Lundbeck, Sanofi Synthelabo. Whyeth. He is an advisor for Lilly and Astra Zeneca and he had received grants from Astra Zeneca, Bristol-Myers Squibb, Lilly, Servier, Glaxo-SmithKline, Lundbeck, Sanofi Synthelabo and Pfizer. Dr N Prieto had received grants from Astra Zeneca, Bristol-Myers Squibb, Lilly, Servier, Lundbeck, Sanofi Synthelabo, Janssen and Pfizer. Dr D Gonzalez Parra had received grants from Servier and Pfizer. Dr J Matías had received grants from Astra Zeneca, Bristol-Myers Squibb, Lilly, Servier, Lundbeck, Sanofi Synthelabo, Janssen and Pfizer. Dr Matías has been a speaker for Glaxo-SmithKline. Dr A Portolés declares neither to have direct or indirect financial interest on any pharmaceutical company nor being member of an advisory board for any pharmaceutical company. In the last 5 years, Dr A Portolés has been investigator of phase-i clinical trials financed by several laboratories (GSK, Pfizer, Belmac, Alter, Combinopharm, Tecnimede, Rubió, cinfa, Servier, TEVA) none of them in competition with this publication.

9 Better sexual acceptability of agomelatine versus paroxetine 9 References Ahl, J, Kinon, BJ, Liu-Seifert, H (2004) Sexual dysfunction associated with neuroleptic-induced hyperprolactinemia improves with reduction in prolactin levels. Ann N Y Acad Sci 1032: Bhasin, S, Enzlin, P, Coviello, A, Basson, R (2007) Sexual dysfunction in men and women with endocrine disorders. Lancet 369: Clayton, A, Kornstein, S, Prakash, A, Mallinckrodt, C, Wohlreich, M (2007) Changes in sexual functioning associated with duloxetine, escitalopram, and placebo in the treatment of patients with major depressive disorder. J Sex Med 4: Clayton, AH, Montejo, AL (2006) Major depressive disorder, antidepressants, and sexual dysfunction. J Clin Psychiatry 67(Suppl. 6): Clayton, AH, Pradko, JF, Croft, HA, Montano, CB, Leadbetter, RA, Bolden-Watson, C, et al. (2002) Prevalence of sexual dysfunction among newer antidepressants. J Clin Psychiatry 63: Compton, MT, Miller, AH (2002) Antipsychotic-induced hyperprolactinemia and sexual dysfunction. Psychopharmacol Bull 36: Cowen, PJ, Sargent, PA (1997) Changes in plasma prolactin during SSRI treatment: evidence for a delayed increase in 5-HT neurotransmission. J Psychopharmacol 11: Delgado, PL, Brannan, SK, Mallinckrodt, CH, Tran, PV, McNamara, RK, Wang, F, et al. (2005) Sexual functioning assessed in 4 doubleblind placebo- and paroxetine-controlled trials of duloxetine for major depressive disorder. J Clin Psychiatry 66: Kasper, S, Laigle, L, Baylé, F (2008) Superior antidepressant efficacy of agomelatine vs sertraline: a randomised, double-blind study. J Clin Psychopharmacol 28: Kennedy, SH (2005) Sexual function in remitted depressed patients following agomelatine and venlafaxine XR treatment. Eur Neuropsychopharmacol 15: S440. Kennedy, SH (2007) Agomelatine: an antidepressant with a novel mechanism of action. Future Neurol 2: Kennedy, SH, Emsley, R (2006) Placebo-controlled trial of agomelatine in the treatment of major depressive disorder. Eur Neuropsychopharmacol 16: Kennedy, SH, Fulton, KA, Bagby, RM, Greene, AL, Cohen, NL, Rafi, TS (2006) Sexual function during bupropion or paroxetine treatment of major depressive disorder. Can J Psychiatry 51: 42. Kennedy, SH, Ralevski, E, Davis, C, Neitzert, C (1996) The effects of moclobemide on sexual desire and function in healthy volunteers. Eur Neuropsychopharmacol 6: Kennedy, SH, Rizvi, S, Fulton, K, Rasmussen, J (2008) A doubleblind comparison of sexual functioning, antidepressant efficacy and tolerability between agomelatine and venlafaxine-xr. J Clin Psychopharmacol (in press). Lemoine, P, Guilleminault, C, Alvarez, E (2007) Improvement in subjective sleep in major depressive disorder with a novel antidepressant, agomelatine: randomized, double-blind comparison with venlafaxine. J Clin Psychiatry 68: Loo, H, Hale, A D haenen, H (2002) Determination of the dose of agomelatine, a melatoninergic agonist and selective 5-HT2C antagonist, in the treatment of major depressive disorder: a placebocontrolled dose range study. Int Clin Psychopharmacol 17: Ludovico, GM, Corvasce, A, Pagliarulo, G, Cirillo-Marucco, E, Marano, A, Pagliarulo, A (1996) Paroxetine in the treatment of premature ejaculation. Br J Urol 77: McGahuey, CA, Gelenberg, AJ, Laukes, CA, Moreno, FA, Delgado, PL, McKnight, KM, et al. (2000) The Arizona Sexual Experience Scale (ASEX): reliability and validity. J Sex Marital Ther 26: Montejo, AI, Llorca, G, Izquierdo, JA, Ledesma, A, Bousono, M, Calcedo, A, et al. (1996) Sexual dysfunction secondary to SSRIs. A comparative analysis in 308 patients. Actas Luso Esp Neurol Psiquiatr Cienc Afines 24: Montejo, AL, Garcia, M, Espada, M, Rico-Villademoros, F, Llorca, G, Izquierdo, JA (2000) Psychometric characteristics of the Psychotropic-Related Sexual Dysfunction Questionnaire. Spanish work group for the study of psychotropic-related sexual dysfunctions. Actas Esp Psiquiatr 28: Montejo, AL, Rico-Villademoros, F (2008) Psychometric properties of the Psychotropic-Related Sexual Dysfunction Questionnaire (PRSexDQ-SALSEX) in patients with schizophrenia and other psychotic disorders. J SEX & Marital Therapy 34: Montejo, AL, Llorca, G, Izquierdo, JA (1997) SSRI-induced sexual dysfunction. Fluoxetine, paroxetine, sertraline and fluvoxamine in a prospective, multi-center and descriptive clinical study of 344 patients. J Sex Marital Ther 23: Montejo, AL, Llorca, G, Izquierdo, JA (1999) Sexual Dysfunction with antidepressive agents. Effect of the change to Amineptine in patients with sexual dysfunction secondary to SSRI. Actas Españolas de Psiquiatría 27: Montejo, AL, Llorca, G, Izquierdo, JA, Rico-Villademoros, F (2001) Incidence of sexual dysfunction associated with antidepressant agents: a prospective multicenter study of 1022 outpatients. Spanish working group for the study of psychotropic-related sexual dysfunction. J Clin Psychiatry 62(Suppl. 3): Nafziger, AN, Bertino, JS, Goss-Bley, AI, Kashuba, AD (1999) Incidence of sexual dysfunction in healthy volunteers on fluvoxamine therapy. J Clin Psychiatry 60: Olié, JP, Kasper, S (2007) Efficacy of agomelatine, a MT1/MT2 receptor agonist with 5-HT2C antagonistic properties, in major depressive disorder. Int J Neuropsychopharmacol 10: Roose, SP (1999) Tolerability and patient compliance. J Clin Psychiatry 60(Suppl. 17): Rosen, RC, Lane, RM, Menza, M (1999) Effects of SSRIs on sexual function: a critical review. J Clin Psychopharmacol 19: Rosen, RC, Riley, A, Wagner, G, Osterloh, IH, Kirkpatrick, J, Mishra, A (1997) The international index of erectile function (IIEF): a multidimensional scale for assessment of erectile dysfunction. Urology 49: Rothschild, AJ (2000) Sexual side effects of antidepressants. J Clin Psychiatry 61(Suppl. 11): Salonia, A, Maga, T, Colombo, R, Scattoni, V, Briganti, A, Cestari, A, et al. (2002) A prospective study comparing paroxetine alone versus paroxetine plus sildenafil in patients with premature ejaculation. J Urol 168: Schlosser, R, Wetzel, H, Dorr, H, Rossbach, W, Hiemke, C, Benkert, O (2000) Effects of subchronic paroxetine administration on nighttime endocrinological profiles in healthy male volunteers. Psychoneuroendocrinology 25: Smith, SM, O Keane, V, Murray, R (2002) Sexual dysfunction in patients taking conventional antipsychotic medication. Br J Psychiatry 181: Waldinger, MD (2007) Premature ejaculation: definition and drug treatment. Drugs 67: Waldinger, MD, Hengeveld, MW, Zwinderman, AH (1994) Paroxetine treatment of premature ejaculation: a double-blind, randomized, placebo-controlled study. Am J Psychiatry 151:

10 10 Better sexual acceptability of agomelatine versus paroxetine Waldinger, MD, Hengeveld, MW, Zwinderman, AH (1997) Ejaculation-retarding properties of paroxetine in patients with primary premature ejaculation: a double-blind, randomized, doseresponse study. Br J Urol 79: Waldinger, MD, Hengeveld, MW, Zwinderman, AH, Olivier, B (1998) Effect of SSRI antidepressants on ejaculation: a double-blind, randomized, placebo-controlled study with fluoxetine, fluvoxamine, paroxetine, and sertraline. J Clin Psychopharmacol 18: Waldinger, MD, Olivier, B (2004) Utility of selective serotonin reuptake inhibitors in premature ejaculation. Curr Opin Investig Drugs 5: Waldinger, MD, Zwinderman, AH, Olivier, B (2001a) Antidepressants and ejaculation: a double-blind, randomized, placebo-controlled, fixed-dose study with paroxetine, sertraline, and nefazodone. J Clin Psychopharmacol 21: Waldinger, MD, Zwinderman, AH, Olivier, B (2001b) SSRIs and ejaculation: a double-blind, randomized, fixed-dose study with paroxetine and citalopram. J Clin Psychopharmacol 21: Waldinger, MD, Zwinderman, AH, Olivier, B (2003) Antidepressants and ejaculation: a double-blind, randomized, fixed-dose study with mirtazapine and paroxetine. J Clin Psychopharmacol 23: Waldinger, MD, Zwinderman, AH, Schweitzer, DH, Olivier, B (2004) Relevance of methodological design for the interpretation of efficacy of drug treatment of premature ejaculation: a systematic review and meta-analysis. Int J Impot Res 16: Zajecka, JM (2000) Clinical issues in long-term treatment with antidepressants. J Clin Psychiatry 61(Suppl. 2):