Vortioxetine for Treatment of Generalised Anxiety Disorder
Vortioxetine for Treatment of Generalised Anxiety Disorder
This multicentre, randomised, double-blind, placebo-controlled, active-referenced, parallel-group, phase 3 study was conducted at 72 centres in the United States. The trial was initiated in June, 2008, and completed in February, 2009. The study protocol was approved at each study site either through a central institutional review board, or by individual study site institutional review board according to the requirements of the participating location. The study was conducted in compliance with the US Food and Drug Administration (FDA) code of Federal Regulations Part 21, the International Conference on Harmonisation Tripartite Guidelines for Good Clinical Practice and the World Medical Association Declaration of Helsinki. Eligible patients provided written informed consent prior to screening procedures.
Adult men and women (aged 18–65 years inclusive) were included if they had a primary diagnosis of GAD as defined by the Diagnostic and Statistical Manual of Mental Disorders, 4th edition Text Revision classification code 300.02, a HAM-A total score ≥ 20, a HAM-A score ≥ 2 on both item 1 (anxious mood) and item 2 (tension), and a Montgomery-Åsberg Depression Rating Scale (MADRS) total score ≤ 16 at screening and baseline. Patients were ineligible for study participation if they had received any investigational compound < 30 days or five half-lives prior to screening; had received vortioxetine in a previous clinical study; had a history of lack of response to duloxetine for the treatment of GAD; had any concurrent psychiatric disorder other than GAD or prior history of psychiatric disorders such as manic or hypomanic episode, schizophrenia or substance abuse disorder; had a significant suicide risk in the opinion of the investigator or a score of ≥ 5 on item 10 of the MADRS; or had previously failed to respond to adequate treatment with an SSRI or an SNRI.
Eligible patients were randomised (1:1:1:1:1) to receive placebo, vortioxetine 2.5 mg, vortioxetine 5 mg, vortioxetine 10 mg or duloxetine 60 mg once daily during the 8-week double-blind treatment period using an interactive voice response system (IVRS) (Figure 1). Vortioxetine tablets (2.5, 5 or 10 mg) and duloxetine 60 mg (for treatment) or 30 mg (for tapering dose) capsules were enclosed in brownish-orange capsules; identical capsules containing lactose monohydrate/magnesium stearate filler were used for placebo. Patients were dispensed their double-blind study drug supplies (using IVRS) and were instructed to start taking the medication on the following day and daily thereafter, preferably in the morning, with or without food. No dose titration or escalation period for either vortioxetine or duloxetine was utilised.
(Enlarge Image)
Figure 1.
IVRS, interactive voice response system; QD, once daily
Baseline measurements of HAM-A, Clinical Global Impression-Severity (CGI-S), patient-reported Hospital Anxiety and Depression (HAD) scale and the SDS were obtained prior to randomisation. The HAM-A and CGI-S were assessed at every evaluation visit. Other measures were evaluated on the following schedule: Clinical Global Impression-Improvement of illness (CGI-I) scale at weeks 1, 2, 4, 6 and 8; HAD at weeks 1, 4 and 8; and SDS at weeks 1, 2, 4 and 8.
Adverse events were documented at weeks 0, 1, 2, 4, 6 and 8 and follow up, and were coded by system organ class and preferred term using the Medical Dictionary for Regulatory Activities (MedDRA; International Federation of Pharmaceutical Manufacturers and Associations, Geneva, Switzerland) version 11.0. Vital signs were measured at all visits. Clinical laboratory values, weight and electrocardiograms were evaluated at weeks 0, 4 and 8, and physical examinations were conducted at screening and week 8. Suicidal ideation and behaviour were assessed as an exploratory outcome using the Columbia-Suicide Severity Rating Scale (C-SSRS) at all visits. Sexual functioning was evaluated using the Arizona Sexual Experiences Scale (ASEX).
Analysis Sets. The safety set included all patients who received at least one dose of study medication. The full analysis set (FAS) comprised all randomised patients who received at least one dose of the study drug and had at least one valid postbaseline value for primary efficacy assessment (HAM-A).
Sample Size and Study Power. A standard deviation of 7.0 was assumed for the change from baseline in HAM-A total score (primary end-point), with a total of 750 evaluable subjects (150 per treatment group) estimated to be needed to achieve at least 85% power to detect a difference of 2.5 between the vortioxetine doses vs. placebo by a two-sample t-test with a 0.05 two-sided significance level.
Statistical Methods. Data analysis and descriptive statistics and inferential statistics tabulations were performed using Sas version 9.1.3 (SAS Institute, Inc, Cary, NC).
The primary efficacy end-point was the change from baseline in HAM-A total score at week 8. Comparisons between vortioxetine 5- and 10-mg treatment groups and placebo were performed using the mixed model for repeated measures (MMRM) on all available data.
Secondary Efficacy Analysis. Sensitivity analyses of the primary efficacy end-point – change from baseline in HAM-A total score – were made with an ANCOVA model, adjusting for baseline score, centre and treatment, using both LOCF and observed-case (OC) methods. CGI-S and CGI-I were analysed as continuous variables (based on FAS and both LOCF and OC) using an ANCOVA model, with the CGI-S baseline used as the covariate adjustment in the ANCOVA analysis of CGI-I. CGI-S and CGI-I were also analysed at the last visit using the Mantel–Haenszel test and stratifying by centre. Change from baseline in the HAD anxiety and depression subscales was analysed as continuous variables, similarly to the primary variable. HAM-A response (defined as a ≥ 50% decrease from baseline in HAM-A total score) and remission (defined as a HAM-A total score of ≤ 7) were analysed at all time points by logistic regression, adjusting for baseline score and treatment using both LOCF and OC methods. The CGI-S remission (CGI-S ≤ 2) and the CGI-I response (CGI-I ≤ 2) were analysed similarly.
To control for two-sided type I errors, a prespecified sequential testing procedure was applied to compare the vortioxetine 5- and 10-mg doses with placebo. The efficacy end-points were tested in the following sequential order at a two-sided Bonferroni-corrected significance level of 0.025:
As soon as the result for a given end-point was not significant at the 0.025 level, the testing procedure was stopped for all subsequent end-points. Nominal p-values, with no adjustment for multiplicity, were reported for end-points either not included in the hierarchy or reported after a result proved non-significant. The phrase 'separation from placebo' is used to describe findings with nominal p-values ≤ 0.05.
Analysis of Safety Variables. All AEs, vital signs, clinical laboratory values, physical examination results and electrocardiograms were descriptively summarised. Suicidal ideation and behaviour data collected using the C-SSRS were summarised at all time points using descriptive statistics.
Sexual dysfunction was defined as having an ASEX total score of ≥ 19, a score of ≥ 4 on any three items or a score of ≥ 5 on any one item. Rates of treatment-emergent sexual dysfunction (TESD) were measured for each treatment group in the subset of patients who did not meet these criteria for sexual dysfunction at baseline. Rates in each vortioxetine group were compared with those for placebo and duloxetine.
Methods
Study Design
This multicentre, randomised, double-blind, placebo-controlled, active-referenced, parallel-group, phase 3 study was conducted at 72 centres in the United States. The trial was initiated in June, 2008, and completed in February, 2009. The study protocol was approved at each study site either through a central institutional review board, or by individual study site institutional review board according to the requirements of the participating location. The study was conducted in compliance with the US Food and Drug Administration (FDA) code of Federal Regulations Part 21, the International Conference on Harmonisation Tripartite Guidelines for Good Clinical Practice and the World Medical Association Declaration of Helsinki. Eligible patients provided written informed consent prior to screening procedures.
Patients
Adult men and women (aged 18–65 years inclusive) were included if they had a primary diagnosis of GAD as defined by the Diagnostic and Statistical Manual of Mental Disorders, 4th edition Text Revision classification code 300.02, a HAM-A total score ≥ 20, a HAM-A score ≥ 2 on both item 1 (anxious mood) and item 2 (tension), and a Montgomery-Åsberg Depression Rating Scale (MADRS) total score ≤ 16 at screening and baseline. Patients were ineligible for study participation if they had received any investigational compound < 30 days or five half-lives prior to screening; had received vortioxetine in a previous clinical study; had a history of lack of response to duloxetine for the treatment of GAD; had any concurrent psychiatric disorder other than GAD or prior history of psychiatric disorders such as manic or hypomanic episode, schizophrenia or substance abuse disorder; had a significant suicide risk in the opinion of the investigator or a score of ≥ 5 on item 10 of the MADRS; or had previously failed to respond to adequate treatment with an SSRI or an SNRI.
Study Treatments
Eligible patients were randomised (1:1:1:1:1) to receive placebo, vortioxetine 2.5 mg, vortioxetine 5 mg, vortioxetine 10 mg or duloxetine 60 mg once daily during the 8-week double-blind treatment period using an interactive voice response system (IVRS) (Figure 1). Vortioxetine tablets (2.5, 5 or 10 mg) and duloxetine 60 mg (for treatment) or 30 mg (for tapering dose) capsules were enclosed in brownish-orange capsules; identical capsules containing lactose monohydrate/magnesium stearate filler were used for placebo. Patients were dispensed their double-blind study drug supplies (using IVRS) and were instructed to start taking the medication on the following day and daily thereafter, preferably in the morning, with or without food. No dose titration or escalation period for either vortioxetine or duloxetine was utilised.
(Enlarge Image)
Figure 1.
IVRS, interactive voice response system; QD, once daily
Efficacy Measures
Baseline measurements of HAM-A, Clinical Global Impression-Severity (CGI-S), patient-reported Hospital Anxiety and Depression (HAD) scale and the SDS were obtained prior to randomisation. The HAM-A and CGI-S were assessed at every evaluation visit. Other measures were evaluated on the following schedule: Clinical Global Impression-Improvement of illness (CGI-I) scale at weeks 1, 2, 4, 6 and 8; HAD at weeks 1, 4 and 8; and SDS at weeks 1, 2, 4 and 8.
Safety Measures
Adverse events were documented at weeks 0, 1, 2, 4, 6 and 8 and follow up, and were coded by system organ class and preferred term using the Medical Dictionary for Regulatory Activities (MedDRA; International Federation of Pharmaceutical Manufacturers and Associations, Geneva, Switzerland) version 11.0. Vital signs were measured at all visits. Clinical laboratory values, weight and electrocardiograms were evaluated at weeks 0, 4 and 8, and physical examinations were conducted at screening and week 8. Suicidal ideation and behaviour were assessed as an exploratory outcome using the Columbia-Suicide Severity Rating Scale (C-SSRS) at all visits. Sexual functioning was evaluated using the Arizona Sexual Experiences Scale (ASEX).
Statistical Analysis
Analysis Sets. The safety set included all patients who received at least one dose of study medication. The full analysis set (FAS) comprised all randomised patients who received at least one dose of the study drug and had at least one valid postbaseline value for primary efficacy assessment (HAM-A).
Sample Size and Study Power. A standard deviation of 7.0 was assumed for the change from baseline in HAM-A total score (primary end-point), with a total of 750 evaluable subjects (150 per treatment group) estimated to be needed to achieve at least 85% power to detect a difference of 2.5 between the vortioxetine doses vs. placebo by a two-sample t-test with a 0.05 two-sided significance level.
Statistical Methods. Data analysis and descriptive statistics and inferential statistics tabulations were performed using Sas version 9.1.3 (SAS Institute, Inc, Cary, NC).
The primary efficacy end-point was the change from baseline in HAM-A total score at week 8. Comparisons between vortioxetine 5- and 10-mg treatment groups and placebo were performed using the mixed model for repeated measures (MMRM) on all available data.
Secondary Efficacy Analysis. Sensitivity analyses of the primary efficacy end-point – change from baseline in HAM-A total score – were made with an ANCOVA model, adjusting for baseline score, centre and treatment, using both LOCF and observed-case (OC) methods. CGI-S and CGI-I were analysed as continuous variables (based on FAS and both LOCF and OC) using an ANCOVA model, with the CGI-S baseline used as the covariate adjustment in the ANCOVA analysis of CGI-I. CGI-S and CGI-I were also analysed at the last visit using the Mantel–Haenszel test and stratifying by centre. Change from baseline in the HAD anxiety and depression subscales was analysed as continuous variables, similarly to the primary variable. HAM-A response (defined as a ≥ 50% decrease from baseline in HAM-A total score) and remission (defined as a HAM-A total score of ≤ 7) were analysed at all time points by logistic regression, adjusting for baseline score and treatment using both LOCF and OC methods. The CGI-S remission (CGI-S ≤ 2) and the CGI-I response (CGI-I ≤ 2) were analysed similarly.
To control for two-sided type I errors, a prespecified sequential testing procedure was applied to compare the vortioxetine 5- and 10-mg doses with placebo. The efficacy end-points were tested in the following sequential order at a two-sided Bonferroni-corrected significance level of 0.025:
Change from baseline in HAM-A total score at week 8 (MMRM)
Change from baseline in HAD anxiety subscore at week 8 (MMRM)
CGI-I at week 8 (MMRM)
Change from baseline in SDS total score at week 8 (MMRM)
HAM-A response rate at week 8 (logistic regression, LOCF)
Change from baseline in HAM-A total score at week 8 in patients with baseline HAM-A total score ≥ 25 (MMRM)
As soon as the result for a given end-point was not significant at the 0.025 level, the testing procedure was stopped for all subsequent end-points. Nominal p-values, with no adjustment for multiplicity, were reported for end-points either not included in the hierarchy or reported after a result proved non-significant. The phrase 'separation from placebo' is used to describe findings with nominal p-values ≤ 0.05.
Analysis of Safety Variables. All AEs, vital signs, clinical laboratory values, physical examination results and electrocardiograms were descriptively summarised. Suicidal ideation and behaviour data collected using the C-SSRS were summarised at all time points using descriptive statistics.
Sexual dysfunction was defined as having an ASEX total score of ≥ 19, a score of ≥ 4 on any three items or a score of ≥ 5 on any one item. Rates of treatment-emergent sexual dysfunction (TESD) were measured for each treatment group in the subset of patients who did not meet these criteria for sexual dysfunction at baseline. Rates in each vortioxetine group were compared with those for placebo and duloxetine.
