Escitalopram (S-Enantiomer of Citalopram): Clinical Efficacy and Onset of Action Predicted from a Rat Model

Transcription

1 C Pharmacology & Toxicology 2001, 88, Printed in Denmark. All rights reserved Copyright C ISSN Escitalopram (S-Enantiomer of Citalopram): Clinical Efficacy and Onset of Action Predicted from a Rat Model Stuart A. Montgomery 1, Henrik Loft 2, Connie Sánchez 3, Elin Heldbo Reines 4 and Mariusz Papp 5 1 Imperial College London, London, United Kingdom, 2 International Clinical Research, Department of Biostatistics, H. Lundbeck A/S, 3 Neuropharmacology, H. Lundbeck A/S, 4 International Clinical Research, Department of Mood Disorders, H. Lundbeck A/S, Copenhagen, Denmark, and 5 Institute of Pharmacology, Polish Academy of Sciences, Krakow, Poland (Received March 3, 2001; Accepted March 20, 2001) Abstract: Escitalopram is the active S-enantiomer of citalopram. In a chronic mild stress model of depression in rats, treatments with both escitalopram and citalopram were effective; however, a faster time to onset of efficacy compared to vehicle treatment was observed for escitalopram-treated (5 mg/kg/day) than for citalopram-treated (10 mg/kg/day) rats at Week 1. To study the predictability of this observation in the clinic, we analysed 4-week data from an 8-week, doubleblind, randomised, placebo-controlled, flexible-dose study that compared escitalopram and citalopram to placebo in primary care patients with major depressive disorder (baseline Montgomery and Åsberg Depression Rating Scale (MADRS) scores Ø22 and Æ40). Since the flexible dosing started after Week 4, analysis of 4-week data ensured that the patients received fixed doses of 10 mg/day escitalopram (155 patients), 20 mg/day citalopram (160 patients), or placebo (154 patients). The efficacy analysis showed a significantly superior therapeutic effect for escitalopram versus placebo from Week 1 onwards (observed cases) with an adjusted mean change in MADRS at Week 4 (last observation carried forward) of 2.7 points (PΩ0.002). By comparison, 20 mg/day citalopram did not demonstrate a statistically significant effect compared to placebo. Escitalopram was well tolerated with an adverse event profile similar to that of citalopram. The preclinical observation that escitalopram possesses a faster time to onset of efficacy than citalopram was also seen in primary care patients with major depressive disorder. Thus, escitalopram is efficacious in depression and the effect occurs earlier than for citalopram. Currently, citalopram is one of the most widely prescribed antidepressants. It is a potent, highly selective serotonin reuptake inhibitor with high affinity for serotonin reuptake sites and low affinity for noradrenaline and dopamine reuptake sites. Citalopram exists as a chiral compound that is a racemic mixture of R- and S-enantiomers in a 1:1 ratio. Escitalopram, the active S-enantiomer of citalopram, mediates serotonin reuptake inhibition (Hyttel et al. 1992; Owens et al. 2001) and antidepressant activity (Hogg & Sánchez 1999), and has increased selectivity and potency compared to citalopram. Depression is a very common and widespread disorder with an estimated lifetime prevalence of 16 to 17% in both Europe and the United States (Kessler et al. 1994; Lépine et al. 1997). According to the World Health Organisation (Murray & Lopez 1997), depression is now associated with more years of disability than any other chronic disorder. Some common complicating factors in the treatment of depression are that the therapy itself may be associated with adverse drug reactions or it may fail to alleviate the condition. There is an unmet need for improvements to the currently available drugs with some of the key areas for improvement being: faster onset of action than current ther- Author for correspondence: Elin H. Reines, H. Lundbeck A/S, International Clinical Research, Department of Mood Disorders, Ottiliavej 9, DK-2500 Valby, Denmark (fax π , er/lundbeck.com). apies, increase of the proportion of patients who respond to treatment, better long-term efficacy, effect in resistant depression, and effect in currently untreatable conditions (Montgomery 1999). Pharmacological studies in animals have been conducted to examine the potential antidepressant effect of escitalopram. The chronic mild stress model was used since it is a behavioural procedure, with high predictive validity, for determining the efficacy of antidepressant compounds (Willner 1997). It consists of sequential exposures to a variety of mild stressors for a prolonged period and results in behavioural hedonic deficits in rats. This condition is thought to model a key symptom of the depressive disorder, that is, a loss of interest or pleasure. In this particular adaptation of the chronic mild stress model, a decrease in consumption of palatable sucrose solution is the hedonic measure, and it has been consistently shown that chronic treatment with antidepressant drugs (including tricyclic antidepressants, selective serotonin reuptake inhibitors, and monoamine oxidase inhibitors) over several weeks gradually normalises the sucrose intake in rats (Willner 1997). Using the rat chronic mild stress model, the efficacy and time to onset was explored for treatment with escitalopram and citalopram. To explore whether this behavioural rat model is a suitable predictor of the onset of action in man, we analysed 4-week fixed-dose data from a large 8-week phase III clinical study in patients with major depressive disorder in a primary care setting.

2 ESCITALOPRAM: CLINICAL EFFICACY AND ONSET OF ACTION 283 Materials and Methods Rat chronic mild stress study. Animals. Male Wistar rats (Gorzkowska, Warsaw, Poland) were brought into the laboratory 2 months before the start of the experiment. Except as described below, the animals were individually housed at 22 2 æc with food and water freely available, and maintained on a 12-hr light/dark cycle (lights on at 8:00 a.m.). The study was conducted in compliance with the Animal Protection Bill of August 21, 1997, and was approved by the Bioethical Committee at the Institute of Pharmacology, Polish Academy of Sciences, Krakow, Poland. Stress procedure. Initially, the animals were trained to consume a 1% sucrose solution. Training consisted of 10 1-hr baseline tests (twice weekly) in which sucrose was presented to the rats in the home cage, after 14 hr without food and water. Sucrose intake was measured at weekly intervals by weighing preweighed bottles containing the sucrose solution. On the basis of their sucrose intakes in the final baseline test, the rats were divided into two matched groups. One group was subjected to the chronic mild stress procedure for a period of 9 consecutive weeks. Each week of the stress regimen consisted of two periods of food or water deprivation, 45 æ cage tilt, intermittent illumination (lights on and off every 2 hr), soiled cage (250 ml water in sawdust bedding), paired housing, low intensity stroboscopic illumination (150 flashes/min.), and no stress. All stressors were 10 to 14 hr in duration and were applied individually and continuously, day and night. Control rats were housed in separate rooms and had no contact with the stressed rats. All the rats were deprived of food and water for the 14 hr preceding each sucrose test, but otherwise food and water were freely available in the home cage. Drug administration. On the basis of sucrose intake scores after 3 weeks of stress, both stressed and control rats were further divided into matched subgroups (nω8), and for the following 5 weeks they received daily intraperitoneal injections of vehicle (1.0 ml/kg saline), 10 mg/kg citalopram, or 5.0 or 10 mg/kg escitalopram. Previously, a dose of 10 mg/kg citalopram was shown to cause stable and comparable effects after chronic treatment in studies using the same experimental protocol (Przegalinski et al. 1995; Sánchez & Papp 2000). The drugs were administered at approximately 10 a.m. and the sucrose tests were carried out 24 hr after the last drug injection. Stress was continued throughout the entire treatment period. Statistics. Sucrose intake data were analysed by multiple one-way analyses of variance comparing drug effects to corresponding vehicle-treatment groups at the 5% significance level. Fisher s Least Significant Difference (LSD) procedure was used for post hoc comparisons of means when relevant. The principal statistical software used was Statistica for Windows A (version 5), StatSoft Inc., Tulsa, OK 74104, USA. Clinical study. Clinical study design and patient selection. This multi-national, parallel-group study conducted in 69 primary care centres in 8 countries (Belgium, Canada, Finland, France, Norway, Sweden, Switzerland, and the United Kingdom) compared the efficacy and safety of escitalopram and the active reference, citalopram, to placebo. Patients fulfilled the Diagnostic and Statistical Manual of Mental Disorders, 4th Edition (DSM-IV), criteria for major depressive disorder with a baseline MADRS Ø22 and Æ40. After a 1-week, single-blind placebo phase, patients were randomised into an 8- week double-blind phase where they received 10 mg/day escitalopram, 20 mg/day citalopram, or placebo, with the option of having their dosage doubled after 4 or 6 weeks of treatment. The study was approved by the relevant ethics committees and conducted in compliance with the principles of Good Clinical Practice, the Declaration of Helsinki and its amendments in force at the initiation of the study, and the Loi Huriet (Law ) in France. Prior to enrolment in the study, all patients gave their written informed consent. Analysis. The objective of this analysis was to evaluate the effect of fixed doses, 10 mg/day escitalopram and 20 mg/day citalopram, versus placebo on depressive symptoms during the first 4 weeks of the study. Assessments. The Montgomery and Åsberg Depression Rating Scale (MADRS) is a validated scale for measuring a patient s depressive symptoms in clinical studies (Montgomery & Åsberg 1979). The MADRS consists of 10 items, each rated on a scale from 0 (no symptoms) to 6 (severe symptoms). In this analysis, the MADRS scores of interest were those at Weeks 0, 1, 2, 3, and 4 (or an early termination visit). The ratings were carried out by the same person at each visit, whenever possible. To increase the inter-rater reliability, only persons experienced with MADRS rating and trained as raters during a co-rating session were allowed to rate patients. The Clinical Global Impressions (CGI) Scale (Guy 1976) consists of two sub-scales: O CGI improvement scale. This single-item rating scale evaluates a patient s total improvement from baseline on a defined 7-point scale with regard to whether the improvement is related to the study product. The investigator (rater) rated the patient from 1 (very much improved) to 7 (very much worse). CGI improvement was rated at Weeks 1, 2, 3, and 4 (or an early termination visit). O GCI severity scale. This single-item rating scale evaluates a patient s severity of depression on a defined 7-point scale based on the investigator s total clinical experience with depressed patients. The investigator (rater) rated the patient from 1 (normal, not at all ill) to 7 (among the most extremely ill patients). CGI severity was rated at Weeks 0, 1, 2, 3, and 4 (or an early termination visit). Safety assessments included recording of adverse events. Adverse events were summarised using descriptive statistics for the all-patients-treated set for the 4-week analysis. There were other efficacy and safety assessments performed which are outside the scope of this analysis, and which will be presented in a comprehensive article on the study. Statistical analysis. A minimum of 360 patients (120 patients in each treatment group) for the full-analysis set was expected to provide a power of at least 85% to detect a significant difference in mean change from baseline to final assessment (Week 8) in the MADRS total score between the escitalopram and placebo treatment groups. This assumed a signal-to-noise ratio of 0.40 at a significance level of 5%. The full-analysis set, also known as the intention-to-treat data set, comprised all randomised patients who took at least one dose of double-blind study product and who had at least one postbaseline assessment of the MADRS total score. Efficacy data for the fullanalysis set were analysed using visit windows and the principle of observed cases for Weeks 1, 2, 3, and 4, and last observation carried forward for the last visit. Efficacy analysis of the change in MADRS total score from baseline to Week 4 and the change in MADRS total score from baseline to each visit were based on a general linear model (ANCOVA) with factors for treatment group and centre, and with baseline score as a covariate. The principal statistical software used was SAS A (SAS Campus Drive, Cary, North Carolina 27513, USA), version 8.1. Results Animal pharmacology. Chronic mild stress caused a gradual decrease in the consumption of 1% sucrose solution. In the final baseline test,

3 284 STUART A. MONTGOMERY ET AL. Fig. 1. Graph of the effect of chronic treatment with 1 ml/kg vehicle, 5 mg/kg escitalopram, or 10 mg/kg citalopram on the consumption of a 1% sucrose solution in controls (closed symbols) and in rats exposed to chronic mild stress (open symbols). Treatment commenced following 3 weeks of stress. In the stressed groups (nω8/ group), escitalopram increased the consumption of sucrose solution to unstressed control levels by Week 1, while a similar effect was observed in the citalopram stressed group by Week 2. Values shown are the means (S.E.M. has been omitted for clarity). Statistical significance relative to vehicle- or drug-treated controls is indicated by: *P 0.05, **P 0.01, ***P Statistical significance relative to drug-treated stressed rats at Week 0 is indicated by: π P 0.05, ππ P 0.01, πππ P all animals consumed approximately 15 g of sucrose solution. In the following 3 weeks of stress, sucrose intake remained at similar levels in control animals, but decreased significantly to 8 g in stressed animals (PΩ0.001). The difference between control and stressed animals treated with vehicle persisted at similar levels for the remainder of the experiment (fig. 1). Fig. 1 shows the effect of 5 mg/kg/day escitalopram and 10 mg/kg/day citalopram in the chronic mild stress model of depression in rats, with citalopram acting as the active reference. Neither escitalopram nor citalopram affected sucrose consumption in unstressed controls. In the stressed groups, escitalopram increased the consumption of sucrose solution to unstressed control levels by Week 1 (PΩ0.029). A similar effect was observed in the citalopram-stressed group by Week 2 (PΩ0.050). In the same study, a parallel group treated with 10 mg/kg/day escitalopram also produced an increase of sucrose consumption to unstressed control levels by Week 1 (PΩ0.049) (data not shown). The effects of escitalopram and citalopram were still observed in stressed groups at the end of the 5-week study period. Body weight remained unaffected by drug treatment throughout the study in both unstressed and stressed groups (data not shown). Clinical study. A total of 471 patients were randomised into the study. The all-patient-treated set comprised 469 patients and the fullanalysis set 468 patients. In the full-analysis set there were 155 patients in the escitalopram group, 159 patients in the citalopram group, and 154 patients in the placebo group. There was an approximately 3 to 1 ratio of women to men in each treatment group, and almost all patients were Caucasian. The mean age was 43 years (S.D. 11). At baseline, the mean MADRS total score was approximately 29 for each treatment group, which signifies moderate to severely ill patients. The efficacy analysis of the adjusted mean change in MADRS total score showed a significantly superior therapeutic effect for escitalopram versus placebo from Week 1 (PΩ0.023) to Week 4 (PΩ0.002) (observed cases). At Week 4, the adjusted mean change in MADRS total score (last observation carried forward) for escitalopram versus placebo was 2.7 points (PΩ0.002) compared to a statistically insignificant change of 1.5 points for citalopram versus placebo (table 1 and fig. 2). Escitalopram was statistically significantly superior to placebo on both the CGI improvement and severity subscales from Week 1 (observed cases) (PÆ0.05) onwards, while there was no statistically significant difference between citalopram and placebo during the 4-week period (data not shown). At Week 4 (last observation carried forward), escitalopram was statistically significantly superior to placebo while there was no statistically significant difference between citalopram versus placebo (data not shown). The withdrawal rate from the study was low with 9.7% in the placebo group, 5.8% in the escitalopram group, and 5.0% in the citalopram group. Approximately 3% of patients were withdrawn due to adverse events in each treatment group during the entire study. During the 4-week period, the adverse events occurring in ±10% of patients in any treatment group were nausea and headache. Headache was more frequent in the placebo group, and the incidences Fig. 2. Graph of the adjusted mean change in MADRS total score from baseline (table 1) showing the significantly superior therapeutic effect for escitalopram versus placebo from Week 1 onwards and the statistically insignificant difference between citalopram and placebo during the first 4 weeks of treatment. Efficacy analyses were performed using visit windows with observed case (OC) analysis used for Weeks 1, 2, 3, and 4, and last observation carried forward (LOCF) shown for the Week 4 last visit. All P-values relate to differences between escitalopram and placebo: *P 0.05, **P 0.01.

4 ESCITALOPRAM: CLINICAL EFFICACY AND ONSET OF ACTION 285 Table 1. Adjusted mean change from baseline in MADRS total scores. Least squares Difference to Treatment group Visit a Analysis set N mean b placebo S.E. P-value Placebo Week 1 (OC) 149 ª3.13 Week 2 (OC) 140 ª6.76 Week 3 (OC) 141 ª8.84 Week 4 (OC) 144 ª9.06 Last (LOCF) 154 ª8.68 Escitalopram Week 1 (OC) 154 ª4.40 ª Week 2 (OC) 146 ª8.20 ª Week 3 (OC) 143 ª10.62 ª Week 4 (OC) 145 ª11.82 ª Last (LOCF) 155 ª11.34 ª Citalopram Week 1 (OC) 154 ª3.92 ª Week 2 (OC) 150 ª7.46 ª Week 3 (OC) 151 ª9.74 ª Week 4 (OC) 154 ª10.50 ª Last (LOCF) 159 ª10.21 ª a Using visit windows. b Analysis was based on a general linear model (ANCOVA) with factors for treatment group and centre, and with baseline score as a covariate. NΩnumber of patients, S.E.Ωstandard error, OCΩobserved cases, LOCFΩlast observation carried forward. of both these adverse events were comparable in the escitalopram and citalopram groups. Discussion This analysis explored whether the observations of faster onset for escitalopram in the behavioural pharmacological chronic mild stress model in rats could be repeated in a clinical setting with fixed dosages of escitalopram, citalopram, and placebo administered for a short time (4 weeks). All traditional antidepressants (tricyclic antidepressants and monoamine oxidase inhibitors) tested in the chronic mild stress model required treatrnent periods of at least 3 to 5 weeks before they could normalise the stress-induced behavioural deficit, and the only clinically effective antidepressant treatment that has been shown to restore normal responsiveness to reward in stressed animals after a single week of administration is electroconvulsive shock (Willner 1997). The observations from the pharmacological chronic mild stress model showed a 1-week time to onset of effect for escitalopram compared to placebo versus the 2-week time to onset of effect for citalopram compared to placebo. The improved time to efficacy achieved with escitalopram was a consistent finding in that both doses (5 and 10 mg/kg/day) produced a similar effect. The results for citalopram are consistent with two previously published studies where sucrose consumption was normalised in citalopram-treated rats by Week 2 (Przegalinski et al. 1995; Sánchez & Papp 2000). To date, the predictive validity of the chronic mild stress rat model to man is very high and no false positive or false negative results for any compound have been detected. In the clinical setting described here, escitalopram showed a significantly superior therapeutic effect versus placebo at Week 1 (observed cases). In contrast, there was no statistically significant difference between citalopram and placebo at Week 1. The efficacy of escitalopram was maintained throughout the 4-week period in the absence of a significant difference from placebo for citalopram for the same period. At the Week 4 (last observation carried forward) endpoint, escitalopram was statistically superior to placebo on the adjusted mean change in MADRS total score from baseline; these results are more remarkable given the lack of statistical significance between citalopram (the reference antidepressant) and placebo at this time point. The low withdrawal rate in the study provides additional confidence in the similar observed cases and last observation carried forward results. The CGI severity and improvement analyses provide further support for the MADRS efficacy analyses with escitalopram showing statistically significant superiority to placebo as early as Week 1 while there was no statistical difference for citalopram compared to placebo. The efficacy of escitalopram was sustained through Weeks 2, 3, and 4 (observed cases) and the last visit (last observation carried forward) while citalopram only demonstrated numerical superiority compared to placebo treatment over this same period. The study was not specifically designed to address early onset of efficacy; nevertheless, the early onset finding based on the MADRS results is supported by the CGI severity and improvement results. Future studies should specifically address this hypothesis of early onset of action. Overall, escitalopram was well tolerated with an adverse event profile similar to that of citalopram. In conclusion, the clinical results support the behavioural pharmacological observations and highlight the potential for an improved time to onset of efficacy of escitalopram compared to citalopram in the treatment of depression.

5 286 STUART A. MONTGOMERY ET AL. References Diagnostic and Statistical Manual of Mental Disorders. Fourth Edition (DSM-IV). American Psychiatric Association, Washington DC, USA, Guy, W.: ECDEU Assessment manual for psychopharmacology (rev., 1976). Biometric Laboratory, The George Washington University Kensington, Maryland, USA. US Dept. of Health, Education, and Welfare DHEW Publication No. (ADM) , pp Hogg, S. & C. Sánchez: The antidepressant effects of citalopram are mediated by the S-(π)- and not the R-(ª)-enantiomer. Eur. Neuropsychopharmacology 1999, 9, S213. Hyttel, J., K. P. Bøgesø, J. Perregaard & C. Sánchez: The pharmacological effect of citalopram resides in the (S-(π)-enantiomer. J. Neural. Transm. [Gen Sect] 1992, 88, Kessler, R. C., K. A. McGonagle, S. Zhao, C. B. Nelson, M. Hughes, S. Eshleman, H.-U. Wittchen & K. S. Kendler: Lifetime and 12 month prevalence of DSM III-R psychiatric disorders in the United States. Results from the National Comorbidity Survey. Arch. Gen. Psychiatry 1994, 51, Lépine, J.-P., M. Gastpar, J. Mendlewicz & A. Tylee: Depression in the community: the first pan-european study DEPRES (Depression Research in European Society). Int. Clin. Psychopharmacol. 1997, 12, Montgomery, S. A. & M. Åsberg: A new depression scale designed to be sensitive to change. Brit. J. Psychiatry 1979, 134, Montgomery, S. A.: New developments in the treatment of depression. J. Clin. Psychiatry 1999, 60 [suppl. 14], Murray, C. J. L. & A. D. Lopez: Global mortality, disability, and the contribution of risk factors: Global burden of disease study. Lancet 1997, 349, Owens, M. J., D. L. Knight & C. B. Nemeroff: Second generation SSRls: Human monoamine transporter binding profile of escitalopram and R-fluoxetine. Biol. Psychiatr. 2001, in press. Przegalinski, E., E. Moryl & M. Papp: The effect of 5-HT 1A receptor ligands in chronic mild stress model of depression. Neuropharm. 1995, 35, Sánchez, C. & M. Papp: The selective s 2 ligand Lu has an antidepressant-like profile in the rat chronic mild stress model of depression. Behav. Pharmacol. 2000, 11, Willner, P.: Validity, reliability and utility of the chronic mild stress model of depression: a 10-year review and evaluation. Psychopharmacology 1997, 134,