Vortioxetine for Major Depressive Disorder in Children: 12-Week Randomized, Placebo-Controlled Study

Abstract

Objective

To evaluate efficacy, safety, and tolerability of vortioxetine in children ages 7 to 11 years with major depressive disorder.

Method

Patients meeting criteria for incomplete improvement in depressive symptoms (Children’s Depression Rating Scale–Revised [CDRS-R] total score ≥40 plus <40% reduction and Parent Global Assessment Global Improvement score >2) after 4 weeks of single-blind lead-in treatment with a brief psychosocial intervention plus placebo were randomized 1:1:1:1 to 8-week double-blind treatment with brief psychosocial intervention and placebo, vortioxetine 10 mg/day, vortioxetine 20 mg/day, or fluoxetine 20 mg/day. Following preplanned interim analysis, enrollment to fluoxetine was stopped, and patients were randomized 1:1:1 to placebo, vortioxetine 10 mg, or vortioxetine 20 mg. The primary end point was change in CDRS-R total score from baseline to week 8 for average of vortioxetine 10-mg and 20-mg doses vs placebo.

Results

Of 683 patients enrolled in single-blind lead-in treatment, 540 were randomized to the double-blind period. The mean (SE) change from randomization to week 8 in CDRS-R total score for average of vortioxetine 10-mg and 20-mg doses vs placebo was −19.6 (1.2) and −17.5 (1.4), with a mean difference of −2.1 (1.2) (2-sided p = .0937). Overall, 47% of patients reported treatment-emergent adverse events in the double-blind period; nausea was the most common adverse event in the vortioxetine groups (11.1%-12.6%).

Conclusion

No statistically significant differences were observed in improvement in CDRS-R total score between placebo and vortioxetine; hence, efficacy of vortioxetine for treatment of major depressive disorder in children could not be confirmed. Safety and tolerability data were similar to that seen in adolescents and adults, with no outstanding safety concerns.

Clinical trial registration information

Active Reference (Fluoxetine) Fixed-dose Study of Vortioxetine in Paediatric Participants Aged 7 to 11 Years With Major Depressive Disorder (MDD); https://clinicaltrials.gov/study/NCT02709655.

Plain language summary

This global study evaluated the efficacy, safety, and tolerability of vortioxetine in children aged 7-11 years with major depressive disorder (MDD). Those who did not improve after 4 weeks of initial treatment with placebo and psychotherapy (540 children) were randomly assigned to a placebo, vortioxetine 10-mg/d, or vortioxetine 20-mg/d for 8 weeks and psychotherapy. The primary measure was the change in depressive symptoms using the Children’s Depression Rating Scale-Revised (CDRS-R); children on vortioxetine showed a slight improvement in CDRS-R total scores compared to those on placebo. Vortioxetine (average of 10-mg/d and 20-mg doses/d) was not more effective than placebo (mean difference of –2.1 points; p = 0.0937). Safety and tolerability measures were similar to previous studies of adults and adolescents treated with vortioxetine, with nausea being the most common adverse effect in the vortioxetine groups (11.1% among participants receiving vortioxetine 20-mg/d and 12.6% among participants receiving vortioxetine 10-mg/d).

Major depressive disorder (MDD) is one of the most common pediatric psychiatric disorders, with an estimated 4.4% of children and adolescents ages 3 to 17 years having been diagnosed with depression according to the National Survey of Children’s Health. In addition, according to a 2019 report, >15% of adolescents had at least 1 episode of MDD, and 10% of all pediatric patients (age range 3-17) received mental health services in the past year.1, 2, 3 MDD has significant consequences for children, including impaired school performance, poor social functioning, higher risk of substance abuse, exposure to negative events, interpersonal difficulties later in life, increased risk of other mental health disorders, and recurrent MDD episodes in adulthood.4, 5, 6

MDD poses a high disease burden in children and adolescents, with need for early screening, identification, and intervention to check disease progression.^7,8 In younger children, diagnosis can be particularly challenging because they may lack adequate vocabulary or insight to communicate their emotions⁹ and show higher levels of concomitant neurodevelopmental disorders.¹⁰ Thus, there is a need to better understand diagnosis and treatment of MDD in the pediatric population.

Treatment Gaps in Children With MDD

Treatment of MDD in children often centers around psychotherapy (specifically cognitive-behavioral therapy and interpersonal therapy), pharmacotherapy, or a combination of both, yet only 2 antidepressants have received regulatory approval for treatment of MDD in pediatric patients.³ Fluoxetine is approved for treatment of children and adolescents (age range 7-18) in the United States and European Union, and escitalopram is approved for treatment of adolescents (age range 12-18) in the United States.^3,8,11,12 Although fluoxetine is approved for use in young children, there is limited evidence of antidepressant use in children ages 7 to 12 years because most clinical studies include adolescents and children or adolescents only.^13,14 In addition, antidepressants that are effective in adults often do not show the same efficacy in children and adolescents.^3,15 As a result, data from adolescent or adult studies cannot be extrapolated to children, creating a knowledge gap about the use and effectiveness of antidepressants in that group.

Vortioxetine for Treatment of MDD in Children

Vortioxetine is approved for the treatment of MDD in adults in more than 88 countries worldwide, including the United States and European Union, and has been shown to be efficacious and well tolerated in adults including elderly adults in both short- and long-term studies.16, 17, 18 In pediatric patients, pharmacokinetic analysis of vortioxetine at doses of 5 to 20 mg/day demonstrated concentrations generally proportional with dosage and aided in identifying the suitable dose range for further studies.¹⁹

To evaluate the efficacy, safety, and tolerability of vortioxetine (10 mg/day and 20 mg/day) for acute treatment of adolescents and children with MDD, 2 short-term, placebo-controlled, fluoxetine-referenced trials, 1 in adolescents and 1 in children, were conducted. The results of the study in adolescents showed a reduction in the Children’s Depression Rating Scale–Revised (CDRS-R) total score across all treatment groups, but no statistically significant difference was observed between average treatment effect of the 2 vortioxetine groups and placebo; the difference between fluoxetine and placebo was significant. Vortioxetine demonstrated an acceptable safety profile in adolescents, with nausea being the most commonly reported adverse event (AE), similar to in adults.²⁰ Here, we report results from the study in children ages 7 to 11 with a diagnosis of MDD according to DSM-5.

Method

Study Design

This 12-week, interventional, fixed-dose, placebo-controlled, active-reference (fluoxetine) study was conducted from May 2016 to January 2022 at 86 sites in 18 countries. The study consisted of a 5- to 15-day screening period, 4-week single-blind lead-in period, 8-week double-blind treatment period, and 4-week safety follow-up period after the last dose of treatment (Figure S1, available online). In the single-blind lead-in period, patients received treatment with 3 sessions of a standardized brief psychosocial intervention with placebo capsules to ensure blinding until randomization. Patients showing an incomplete improvement in depressive symptoms at the end of this period (defined as <40% reduction in CDRS-R total score from baseline, CDRS-R total score of ≥40, and Parent Global Assessment [PGA] Global Improvement score of >2) were enrolled into the double-blind treatment period and randomized (via interactive response technology, stratified by site) in a 1:1:1:1 ratio into 4 treatment groups: placebo, vortioxetine 10 mg/day or 20 mg/day, or fluoxetine 20 mg/day. All patients in these 4 groups continued to receive a brief psychosocial intervention (2 sessions) with the treatment. The comprehensive study design and rationale were described previously in the adolescent study.²⁰

This study was conducted in compliance with Good Clinical Practice guidelines and the ethical principles of the Declaration of Helsinki. Study protocols and amendments were approved by the relevant independent ethics committees in accordance with local requirements. Written assent from patients when possible and written informed consent from patients’ parents or legal representatives were obtained before the start of any study activities. All study personnel who collected data from study participants were qualified health care professionals trained in administering rating scales and experienced with pediatric patients with MDD. This study is registered at ClinicalTrials.gov (NCT02709655).

Study Participants

This study enrolled male and female outpatients aged 7 to 11 years with a primary diagnosis of MDD (according to the DSM-5 criteria and confirmed using the Schedule for Affective Disorders and Schizophrenia for School-Age Children [K-SADS]),²¹ a CDRS-R total score ≥45, and a Clinical Global Impressions–Severity (CGI-S) scale score ≥4 at screening and baseline visit of the single-blind lead-in period. Patients with comorbid anxiety disorders were allowed to participate except for those with posttraumatic stress disorder or obsessive-compulsive disorder.

Patients with intellectual disability (IQ <70 or based on clinical evidence), a current psychiatric disorder other than MDD established as the primary diagnosis, moderate to severe head trauma, neurological conditions, or a known history of bipolar disorder in a first-degree relative were excluded. Patients with attention-deficit/hyperactivity disorder were excluded if they required pharmacological treatment and could not be maintained on a stable dose of an appropriate stimulant medication for a minimum of 4 weeks before the baseline visit. In addition, patients receiving cognitive-behavioral therapy or psychotherapy planned to be intensified were also excluded. Patients were excluded if they had attempted suicide or were at a significant risk for suicide (either in the opinion of the investigator or if they responded “yes” to suicidal ideation question 4 or 5 or “yes” to suicidal behavior on the Columbia-Suicide Severity Rating Scale [C-SSRS] within the last 12 months). Patients on disallowed medications, such as antipsychotics and anxiolytics, were asked to discontinue the medication or be excluded if they required it.

Study Assessments

The primary efficacy assessment evaluated improvement in severity of MDD symptoms based on the change in CDRS-R total score from randomization to week 8.²² The CDRS-R is a clinician-rated scale that measures severity of depression in children and adolescents that consists of 17 items with scores ranging from 17 (normal) to 113 (severe depression). CDRS-R total scores were assessed at screening; enrollment; weeks 2 and 3 of the single-blind lead-in period; and weeks 2, 4, and 6 of the double-blind treatment period.

Secondary efficacy measures were evaluated during the 8-week treatment period and involved assessment of global impression, depressive symptoms, functionality, and health-related quality of life (HRQoL). The CGI–Improvement (CGI-I) scale was used to assess overall clinical condition and improvement and provided the patient’s condition relative to baseline on a 7-point scale ranging from 1 (very much improved) to 7 (very much worse). The CGI-S rated the patient’s current state of mental illness on a 7-point scale ranging from 1 (normal—not at all ill) to 7 (among the most extremely ill).²³ Additionally, to evaluate severity of the patient’s depressive symptoms, a parent-rated variation of the CGI-I, the PGA,²⁴ with scores ranging from 1 (very much improved) to 7 (very much worse), and 2 versions of the General Behavior Inventory (GBI) 10-item depression scale (1 child rated and 1 parent rated), with scores ranging from 0 to 30 (higher scores indicate greater pathology), were used.

Functionality was assessed using the clinician-rated Children’s Global Assessment Scale (CGAS)²⁵ and patient-rated Pediatric Quality of Life Inventory (PedsQL)²⁶ scale. The CGAS scores range from 1 (most functionally impaired) to 100 (healthiest), with a score >70 indicating normal function. The PedsQL was used to assess present-moment functioning in patients by means of the total score (average of 6 items: anxiety, sadness, anger, worry, fatigue, and pain using visual analog scales) and the emotional distress summary score (mean of anxiety, sadness, anger, and worry items). HRQoL was measured using the Pediatric Quality of Life Enjoyment and Satisfaction Questionnaire (PQ-LES-Q),²⁷ a patient-rated scale consisting of 15 items, with each item rated on a 5-point scale from 1 (very poor) to 5 (very good), and a total score range of 1 to 14, with higher scores indicating greater satisfaction. Tanner staging was performed at the screening visit to assess physical development and sexual maturity using the 5 stages. In male patients, the 5 stages of maturation were assessed by comparing pubic hair, penis, and testicle growth to reference images; in female patients, breast development and pubic hair were evaluated against reference images. To assess treatment adherence, plasma concentrations of vortioxetine were evaluated at 4 and 8 weeks of the double-blind period.

Multidimensional Anxiety Scale for Children–short version (MASC-10) total score was assessed as an exploratory measure to evaluate the effect of vortioxetine on comorbid symptoms. Safety was assessed at each study visit, including evaluation of AEs according to the Medical Dictionary for Regulatory Activities version 24.1 and the Pediatric Adverse Event Rating Scale (PAERS), a validated scale to assess AEs in pediatric patients on psychotropic medication; vital sign measurements; and the C-SSRS to measure suicidal ideation and behavior. Clinical safety laboratory test results, vital signs, and electrocardiograms, or results of other physical and neurological examinations, were collected during the 8-week period. These were recorded as AEs if considered clinically significant by the investigator. Both the child- and parent-rated versions of the GBI 10-Item Mania Subscale²⁸ were used to screen for manic symptoms in patients, where a score ≥18 was indicative of potential risk of mania.

Statistical Analysis

Efficacy analyses were based on the full analysis set, composed of all patients who took ≥1 dose of vortioxetine and had a valid assessment and ≥1 valid postrandomization assessment of the CDRS-R total score. The primary efficacy end point of the change from randomization in CDRS-R total score at week 8 was evaluated using a restricted maximum likelihood based on a mixed model for repeated measures (MMRM), with fixed effects of treatment, country, and week, and the continuous covariates of CDRS-R total score at randomization, treatment-by-week interaction, and CDRS-R total score at randomization-by-week interaction.

Primary comparison was the average (weights of 0.5 and 0.5) of the effect of vortioxetine at doses of 10 mg/day and 20 mg/day (AVG vortioxetine) vs placebo at week 8 in the double-blind treatment period tested at a 1-sided test on the level obtained for the final analyses taking the alpha spending for the interim into account to maintain a 2.5% overall significance level. This new α level was calculated after unblinding. Under the design assumption of a similar effect for vortioxetine doses, evaluation of average effect was a more efficient strategy than testing doses separately, allowing for a decrease in sample size and subsequent testing of each dose individually. If the primary comparison was statistically significant, the 2 pairwise comparisons of the 2 vortioxetine doses vs placebo were performed separately based on a 1-sided test on the same α level as above. The interim analysis used an error-spending approach based on the Kim-DeMets method with a ρ of 2 applied on the outcome from the MMRM. The α was adjusted to .02266 1-sided after the interim analysis, and interim boundaries were updated, which were then combined with the results of MMRM analysis to be used in the evaluation of primary end points.

Continuous secondary and exploratory end points were analyzed using an MMRM as done for the primary end point, with comparisons from the same model used for all time points. For dichotomous outcomes (CDRS-R response with a ≥50% decrease in CDRS-R total score; CDRS-R remission with CDRS-R total score ≤28; and CGI-S remission with CGI-S score of 1 or 2), the primary methodology for analysis each week during the double-blind treatment period (full analysis set, last observation carried forward) was logistic regression with treatment as a factor and score at randomization as a covariate. In addition to tablet count, adherence to vortioxetine treatment was assessed based on plasma concentration data and at a population level by comparing the distribution of individual oral clearance values with the distribution seen in healthy volunteers treated under well-controlled conditions. Patients were regarded as noncompliant if they had plasma drug concentrations below the lower limits of quantification at any visit or an estimated oral clearance >120 L/hour for vortioxetine.

The safety analysis for the single-blind lead-in period was based on all patients enrolled who took ≥1 dose of single-blind study drug, and safety analysis for the double-blind treatment period was based on all patients randomized who took ≥1 dose of double-blind study drug. The overall AEs, including treatment-emergent adverse events (TEAEs) and serious adverse events (SAEs), clinical safety laboratory test values, vital signs, body measurements (height, weight, body mass index), electrocardiogram parameters, and C-SSRS, PAERS, and GBI mania subscale scores, were summarized using descriptive statistics. The data were analyzed using SAS version 9.4 statistical software (SAS Institute Inc, Cary, North Carolina).

The sample size required for a power of 85% with a 1-sided significance level of 2.5% was 102 patients in a nonsequential approach. To maintain the 85% power after interim analysis, sample size was increased by a factor of 1.045 to correct for the loss of power due to the sequential approach. As a result, 107 randomized patients per group (vortioxetine 10 mg/day, vortioxetine 20 mg/day, placebo) were required in the final analysis. Taking a 15% withdrawal rate in the double-blind period into account, 126 randomized patients per group were required. Because of recruitment difficulties, an interim analysis based on a sequential approach with nonbinding stopping rules for efficacy and futility was performed as an adaptive measure to potentially terminate the study. In addition, the fluoxetine arm as active reference was removed because the comparison was not part of primary testing strategy, and a smaller allocation to the active reference arm would not have altered the ability to compare vortioxetine with placebo. A blinded sample size recalculation was performed before interim analysis to ensure that the study was adequately powered. Because of higher variability than originally assumed, the overall sample size was increased to 539 randomized patients (157 each in the placebo, vortioxetine 10 and 20 mg/day groups and 68 patients in the fluoxetine group before interim analysis) to achieve 85% power for the study.

Results

Patient Disposition

In this study, 840 patients were screened, of whom 677 were treated in the single-blind lead-in period, and 540 were randomized in the double-blind period, in which 89% (n = 484) completed the treatment. Some of the most frequent reasons for withdrawal across all treatment groups were withdrawal of consent (2.4%) and noncompliance with study drug (2.0%). The proportion of patients who withdrew due to AEs was 0.7% in the placebo group, 1.3% in the vortioxetine 10 mg/day group, 2.0% in the vortioxetine 20 mg/day group, and 0.0% in the fluoxetine group. The full analysis set consisted of 530 patients, with 153 in the placebo group, 148 each in the vortioxetine 10 mg/day and vortioxetine 20 mg/day groups, and 81 in the fluoxetine group (Figure 1). The interim analysis was performed on data from 271 randomized patients, and because efficacy or futility criteria were not met, the study continued until the target sample size was reached.

Figure 1.

Patient Disposition

Note:AEs = adverse events; APTS = all-patients-treated set; BPI = brief psychosocial intervention; DB = double-blind; FAS = full analysis set; IMP = investigational medicinal product; PBO = placebo; SB = single-blind.^aPrimary reason.^bThe preplanned interim analysis was performed, and the study was continued based on a priori assumptions.^cThe fluoxetine group was stopped due to recruitment challenges.

Demographic and Baseline Characteristics

Baseline demographic and disease characteristics were comparable across the 4 treatment groups (Table 1). Among the randomized patients in the double-blind treatment period, there were marginally more boys (55%, n = 297) than girls (45%, n = 243), with a mean age of 9 years. Overall, the demographics, height, weight, and body mass index at randomization were similar to that seen at enrollment. At enrollment, according to the Tanner staging, 56% of girls and 65% of boys were prepubertal and 44% of girls and 35% of boys were pubertal (Tanner stage II-IV). At randomization, the baseline efficacy variables indicated moderate to marked illness, with a mean CDRS-R total score of 60.6 points and CGI-S score of 4.6 points (Table 1), comparable to the mean CDRS-R total score (63.4 points) and CGI-S score (4.8 points) for patients at enrollment.

Table 1.

Baseline Demographic and Disease Characteristics and Efficacy Scores at Randomization of Patients in Double-Blind Treatment Period

Characteristics	Placebo		Vortioxetine 10 mg/day		Vortioxetine 20 mg/day		Fluoxetine 20 mg/daya
Demographic characteristics
No. of patients	153		151		153		83
	n	(%)	n	(%)	n	(%)	n	(%)
Sex
Female	62	(40.5)	72	(47.7)	70	(45.8)	39	(47.0)
Male	91	(59.5)	79	(52.3)	83	(54.2)	44	(53.0)
	Mean	(SD)	Mean	(SD)	Mean	(SD)	Mean	(SD)
Age, y	9.3	(1.36)	9.4	(1.53)	9.3	(1.39)	9.3	(1.43)
	n	(%)	n	(%)	n	(%)	n	(%)
Race or ethnicity
Asian	0	(0)	0	(0)	2	(1.3)	0	(0)
Black or African American	21	(13.7)	20	(13.2)	21	(13.7)	13	(15.7)
Other	60	(39.2)	56	(37.1)	52	(34.0)	22	(26.5)
White	70	(45.8)	74	(49.0)	76	(49.7)	45	(54.2)
Not reported	2	(1.3)	1	(0.7)	2	(1.3)	3	(3.6)
Baseline disease characteristics
	Mean	(SD)	Mean	(SD)	Mean	(SD)	Mean	(SD)
Number of previous episodes of MDD	0.3	(0.52)	0.3	(0.51)	0.2	(0.51)	0.3	(0.50)
Duration of current episode, wk	29.3	(33.02)	32.8	(36.90)	30.9	(36.76)	35.3	(40.61)
	n	(%)	n	(%)	n	(%)	n	(%)
Received pharmacotherapy
No	123	(80.4)	128	(84.8)	136	(88.9)	72	(86.7)
Yes	30	(19.6)	23	(15.2)	17	(11.1)	11	(13.3)
Efficacy scores at randomization
No. of patients	153		148		148		81
	Mean	(SD)	Mean	(SD)	Mean	(SD)	Mean	(SD)
CDRS-R total score	60.1	(10.19)	60.7	(9.34)	60.4	(9.41)	61.1	(9.52)
CGI-S score	4.6	(0.7)	4.6	(0.69)	4.6	(0.64)	4.7	(0.64)

Note: CDRS-R = Children’s Depression Rating Scale–Revised; CGI-S = Clinical Global Impressions–Severity; MDD = major depressive disorder; wk = week; y = year.

^aThe fluoxetine group was stopped due to recruitment challenges.

Efficacy Outcomes

Primary Analysis

In the double-blind treatment period, the mean CDRS-R total score decreased from randomization to week 8 in all treatment groups, indicating improvement in depressive symptoms (Figure 2). The mean (SE) change in CDRS-R total score from randomization to week 8 was −17.5 (1.4) in the placebo group and −19.6 (1.2) in the AVG vortioxetine groups, and the mean (SE) difference (−2.1 points [1.2] in favor of AVG vortioxetine) was not statistically significant (p = .094). Hence, the primary end point was not met, and subsequent p values were considered nominal. Neither of the single doses of vortioxetine separated from placebo; the mean (SE) differences from placebo for the vortioxetine 10 mg/day and 20 mg/day groups were −1.7 (1.4) (p = .234) and −2.5 (1.4) (p = .088), respectively. The mean (SE) difference from placebo for the fluoxetine 20 mg/day group was –3.3 (1.7) (p = .053). However, as recruitment for the fluoxetine group was stopped at the time of the interim analysis, the comparison is based on a reduced fluoxetine group, whereas the placebo group recruited patients for the entire study duration.

Figure 2.

Change From Randomization to Week 8 in Children’s Depression Rating Scale–Revised Total Score in Double-Blind Treatment Period

Note:AVG vortioxetine = average of 10 mg/day and 20 mg/day vortioxetine doses; DB = double-blind.

Secondary and Exploratory Analyses

The outcomes of CDRS-R total score, CDRS-R subscores (Table S1, available online), CGI-S, CGI-I, GBI-10, CGAS, PedsQL Present Functioning Visual Analogue Scale (PedsQL VAS), and PQ-LES-Q score showed no effect of vortioxetine over placebo and were comparable with results of the primary analyses (Table 2). Improvements in depressive symptoms, global clinical impression, functionality (CGAS), and HRQoL were observed in all treatment groups through week 8 in the double-blind treatment period. In general, no significant differences were observed between placebo and AVG vortioxetine groups from randomization to week 8 (Table 2) except for functionality, where the mean change from randomization (difference from placebo) for PedsQL VAS total average score (−1.5 [−0.3]; p = .047) and emotional distress score (−1.6 [−0.4]; p = .018) showed significant improvement.

Table 2.

Secondary and Exploratory Efficacy Variables Showing Mean Change From Randomization to Week 8 (Mixed Model for Repeated Measures)

Efficacy Variables	Placebo	AVG Vortioxetine	Vortioxetine 10 mg/day	Vortioxetine 20 mg/day	Fluoxetine 20 mg/daya
Secondary efficacy variables
GBI depression score (child)
No. of patients	134	—	131	132	75
Mean baseline score	1.6	—	1.2	1.4	1.7
Change from randomization	−5.3	−5.8	−6.1	−5.5	−5.5
Difference from placebo	—	−0.5	−0.9	−0.2	−0.2
95% CI	—	−1.63, 0.54	−2.12, 0.39	−1.48, 1.03	−1.70, 1.29
p	—	.3226	.1759	.7211	.7888
GBI depression score (parent)
No. of patients	136	—	132	134	78
Mean baseline score	1.4	—	1.0	1.3	1.7
Change from randomization	−5.8	−6.5	−6.5	−6.5	−6.6
Difference from placebo	—	−0.7	−0.7	−0.6	−0.7
95% CI	—	−1.77, 0.42	−1.97, 0.56	−1.91, 0.61	−2.25, 0.74
p	—	.2237	.2749	.3120	.3232
CGI-S score
No. of patients	137	—	132	134	78
Mean baseline score	4.6	—	4.6	4.6	4.7
Change from randomization	−1.3	−1.4	−1.4	−1.4	−1.6
Difference from placebo	—	−0.1	−0.1	−0.1	−0.3
95% CI	—	−0.33, 0.11	−0.35, 0.16	−0.39, 0.12	−0.56, 0.04
p	—	.3230	.4844	.3136	.0897
PedsQL VAS total average score
No. of patients	136	—	130	134	78
Mean baseline score	3.7	—	3.7	3.8	4.0
Change from randomization	−1.2	−1.5	−1.5	−1.5	−1.5
Difference from placebo	—	−0.3	−0.4	−0.3	−0.4
95% CI	—	−0.67, −0.00	−0.77, 0.01	−0.68, 0.09	−0.82, 0.09
p	—	.0473	.0563	.1278	.1173
PedsQL Emotional Distress summary score
No. of patients	136	—	130	134	78
Mean baseline score	3.6	—	3.7	3.8	4.0
Change from randomization	−1.2	−1.6	−1.8	−1.5	−1.7
Difference from placebo	—	−0.4	−0.6	−0.3	−0.4
95% CI	—	−0.79, −0.08	−1.00, −0.17	−0.70, 0.13	−0.93, 0.04
p	—	.0177	.0062	.1741	.0738
CGAS score
No. of patients	136	—	132	134	78
Mean baseline score	55.0	—	54.7	55.4	52.7
Change from randomization	12.7	13.1	13.0	13.2	16.3
Difference from placebo	—	0.4	0.3	0.5	3.6
95% CI	—	−1.81, 2.60	−2.28, 2.83	−2.03, 3.05	0.63, 6.65
p	—	.7257	.8338	.6910	.0179
PQ-LES-Q total score
No. of patients	136	—	132	134	78
Mean baseline score	41.1	—	40.8	41.4	38.8
Change from randomization	6.2	7.1	7.1	7.1	6.8
Difference from placebo	—	0.9	0.9	0.9	0.6
95% CI	—	−0.73, 2.51	−1.02, 2.73	−0.95, 2.79	–1.64, 2.78
p	—	.2813	.3691	.3337	.6134
PQ-LES-Q overall score
No. of patients	136	—	132	134	78
Mean baseline score	3.1	—	3.0	3.1	2.8
Change from randomization	0.4	0.5	0.5	0.5	0.4
Difference from placebo	—	0.1	0.1	0.1	0.04
95% CI	—	−0.06, 0.25	−0.10, 0.26	−0.07, 0.28	−0.17, 0.25
p	—	.2333	.3711	.2427	.6942
Exploratory efficacy variable
MASC-10 (anxiety)
No. of patients	132	—	129	130	76
Mean baseline score	12.4	—	12.8	12.6	11.8
Change from randomization	−1.9	−1.9	−1.8	−2.1	−2.3
Difference from placebo	—	−0.1	0.1	−0.2	−0.4
95% CI	—	−1.04, 0.93	−1.05, 1.22	−1.34, 0.93	−1.72, 0.95
p	—	.9094	.8799	.7273	.5669

Note: AVG vortioxetine = average of 10 mg/day and 20 mg/day vortioxetine doses; CDRS-R = Children’s Depression Rating Scale–Revised; CGAS = Children’s Global Assessment Scale; CGI-S = Clinical Global Impressions–Severity; GBI = General Behavior Inventory; MASC-10 = Multidimensional Anxiety Scale for Children–short version; PedsQL = patient-rated Pediatric Quality of Life Inventory; PedsQL VAS = PedsQL Present Functioning Visual Analogue Scale; PQ-LES-Q = Pediatric Quality of Life Enjoyment and Satisfaction Questionnaire.

^aThe fluoxetine group was stopped because of recruitment challenges.

At week 8, 36.5% of patients in the vortioxetine 10 mg/day group (p = .2451), 39.9% in the vortioxetine 20 mg/day group (p = .0834), 30.7% in the placebo group, and 46.9% in the fluoxetine group (p = .0098) showed response to treatment as assessed by CDRS-R total score, without statistically significant differences between the vortioxetine groups and placebo group. Remission as assessed by CDRS-R total scores (15.5% in vortioxetine 10 mg/day [p = .9192], 19.6% in vortioxetine 20 mg/day [p = .3168], 15.7% in placebo, and 25.9% in fluoxetine [p = .0368]) and CGI-S scale scores (22.3% in vortioxetine 10 mg/day [p = .9128], 20.9% in vortioxetine 20 mg/day [p = .6697], 22.9% in placebo, and 29.6% in fluoxetine [p = .2293]) did not show statistically significant differences between vortioxetine and placebo groups. At the end of the double-blind treatment period, no apparent relation was found between efficacy variables and vortioxetine plasma exposure. Results of exploratory analysis showed no significant differences in the AVG or individual vortioxetine treatment groups compared with placebo for CDRS-R item scores or MASC-10 comorbid symptoms scores.

Adherence With Treatment

A majority of patients (≥95%) across treatment groups demonstrated ≥80% adherence based on tablet count. However, 77 (28%) of the 273 patients treated with vortioxetine showed nonadherence based on plasma concentrations in terms of estimated oral clearance >120 L/hour or pharmacokinetic samples below the first lower limits of quantification at any visit.

Safety Outcomes

Adverse Events

In the single-blind lead-in period, 169 (25.0%) patients reported 297 TEAEs. The most common TEAEs were headache (5.8% of patients), nausea (3.4%,), and abdominal pain (2.2%). A total of 6 (0.9%) patients reported 7 SAEs in the single-blind lead-in period, and 4 (0.6%) patients reported 5 TEAEs leading to withdrawal.

In the double-blind treatment period, incidence of TEAEs was similar across groups, ranging from 49% (vortioxetine 10 mg/day) to 43.1% (placebo). A low and similar incidence of TEAEs leading to withdrawal was observed in the vortioxetine groups and placebo group, ranging from 3 (2%) to 1 (0.7%) (Table 3). The 6 TEAEs leading to withdrawal included 1 occurrence each of major depression, mania, vomiting, blood thyroid-stimulating hormone increase, nausea, and suicide attempt. Severe TEAEs occurred in 8 patients: 2 (1.3%), 3 (2.0%), 2 (1.3%), and 1 (1.2%) in the placebo, vortioxetine 10 mg, vortioxetine 20 mg, and fluoxetine groups, respectively. SAEs were reported by 7 patients overall, with similar incidence across groups. In 1 patient from the vortioxetine 20 mg/day group, SAEs of major depression and mania (parent scale 6.0, child scale 0) were considered to be related to study drug. The details of AEs and TEAEs with an incidence of >2% in the double-blind treatment period are summarized in Table 3.

Table 3.

Patients With Adverse Events Reported in the Double-Blind Treatment Period

Preferred Terms	Placebo		Vortioxetine 10 mg/day		Vortioxetine 20 mg/day		Fluoxetine 20 mg/day
	n	(%)	n	(%)	n	(%)	n	(%)
No. of patients	153	—	151	—	153	—	83	—
Patient-years of exposure	22	—	21	—	22	—	12	—
Patients with TEAEs	66	(43.1)	74	(49.0)	72	(47.1)	40	(48.2)
Patients with SAEs	3	(2.0)	1	(0.7)	2	(1.3)	1	(1.2)
Patients with TEAEs leading to withdrawal	1	(0.7)	2	(1.3)	3	(2.0)	0	(0)
Patients with TEAEs with an incidence >2%	45	(29.4)	55	(36.4)	42	(27.5)	28	(33.7)
Nausea	7	(4.6)	19	(12.6)	17	(11.1)	5	(6.0)
Headache	17	(11.1)	14	(9.3)	14	(9.2)	4	(4.8)
Vomiting	3	(2.0)	14	(9.3)	10	(6.5)	3	(3.6)
Abdominal pain	2	(1.3)	9	(6.0)	6	(3.9)	2	(2.4)
Dizziness	5	(3.3)	7	(4.6)	5	(3.3)	3	(3.6)
Illness	0	(0)	0	(0)	5	(3.3)	0	(0)
Nasopharyngitis	5	(3.3)	6	(4.0)	4	(2.6)	3	(3.6)
Upper abdominal pain	4	(2.6)	4	(2.6)	3	(2.0)	3	(3.6)
Weight increased	4	(2.6)	1	(0.7)	3	(2.0)	2	(2.4)
Decreased appetite	2	(1.3)	1	(0.7)	2	(1.3)	3	(3.6)
Diarrhea	4	(2.6)	5	(3.3)	1	(0.7)	3	(3.6)
Dry mouth	4	(2.6)	4	(2.6)	1	(0.7)	0	(0)
Weight decreased	0	(0)	0	(0)	1	(0.7)	2	(2.4)
Epistaxis	0	(0)	1	(0.7)	0	(0)	2	(2.4)
Forearm fracture	0	(0)	0	(0)	0	(0)	2	(2.4)
Viral infection	0	(0)	1	(0.7)	0	(0)	2	(2.4)

Note: SAEs = serious adverse events; TEAEs = treatment-emergent adverse events.

Suicidality

Three patients in the single-blind period and 2 in the double-blind treatment period had suicide-related TEAEs. Regarding the 2 patients in the double-blind period, a suicide attempt (considered an SAE) was reported by 1 patient in the placebo group, and suicidal ideation was reported by 1 patient in the vortioxetine 10 mg/day group. No patients had suicidal ideation during the single-blind period as captured by the C-SSRS score. In the double-blind period, nonspecific active suicidal thoughts were reported in 5 patients, 1 each in the placebo and vortioxetine 20 mg/day groups and 3 in the fluoxetine group.

Mania Assessment

Overall, the mean changes from randomization to week 8 in the 10-item GBI Mania subscale score, as assessed by either the parent or the child, were small and similar across groups. A GBI Mania subscale score ≥18 points, indicating a potential risk of mania, was reported by 15 patients (2, 3, 3, and 7 patients in the placebo, vortioxetine 10 mg/day, vortioxetine 20 mg/day, and fluoxetine 20 mg/day groups, respectively), with no clinically relevant difference across groups and none reported as TEAEs. One patient in the vortioxetine 20 mg/day group withdrew from the study because of a high GBI Mania subscale score of 20 as assessed using the child GBI scale.

Laboratory and Vital Sign Findings

The mean changes from randomization in all clinical safety laboratory tests, vital signs, weight, body mass index, height, and electrocardiogram parameters were small, comparable between treatment groups, and determined not to be clinically relevant. Overall, the proportion of patients with potentially clinically significant values for these variables after randomization was low and similar across treatment groups.

Discussion

In this study, children receiving vortioxetine (10 mg/day and 20 mg/day), placebo, and fluoxetine (20 mg/day) showed improvement in CDRS-R total score. However, vortioxetine was not significantly superior to placebo in the primary or secondary efficacy end points at the end of 8 weeks of treatment. In the previously published adolescent study of vortioxetine by Findling et al.,¹⁹ no statistically significant difference between the vortioxetine groups and placebo group was found, whereas fluoxetine showed a significant difference from placebo. In the current study, fluoxetine showed a 3.3-point difference from placebo (p = .053), but this result should be interpreted with caution, as the fluoxetine arm was stopped early, and thus populations in the placebo and fluoxetine groups may not be directly comparable. In pediatric trials with antidepressants, the response to active drug is comparable across trials, and a varied placebo response rate often determines the effect sizes, rendering a trial positive or negative.²⁹ It is also very common for an active drug to not show differentiation from placebo in pediatric studies because of a high placebo response rate.³⁰ Bridge et al.³¹ reported that the placebo response rate was higher in MDD trials as opposed to trials of other psychiatric disorders, particularly in children <12 years of age compared with adults and adolescents. Use of a greater number of study sites is often a predictor of high placebo response rate, as patients with low disease severity are also recruited. It is noteworthy that a systematic review by Mossman et al.³² reported that industry-funded studies with a high placebo response rate and low drug-placebo differences are essentially not negative, but rather failed trials.

Study designs for pediatric MDD trials have traditionally led to negative outcomes because of high placebo response rates. This study was carefully designed to account for this; use of a placebo run-in period³³ during the single-blind phase of the study was a key feature and specifically used to check the high placebo response rate by filtering patients with less severe MDD or who responded to brief psychosocial intervention. Despite the robust study design, a confluence of the above-mentioned factors may have contributed to vortioxetine not differentiating significantly from placebo.

Another probable contributing factor leading to no differences between placebo and vortioxetine groups could be the low adherence to medication observed in this study. Pharmacokinetic analysis of vortioxetine plasma concentrations showed that more than one-third of patients were nonadherent to the medication schedule. Nonadherence may be influenced by factors such as lack of family support, negative attitudes toward antidepressants, severity or chronicity of illness, regimen, frequency of medication, AEs, and length of time required for optimal response.³⁴

Vortioxetine was generally tolerated well and showed a safety profile in pediatric participants similar to adolescents and adults.^18,20 The most common AE among children, adolescents, and adults was nausea. Use of antidepressants is often linked to an increase in suicidal ideation in children and adolescents compared with placebo.³⁵ However, in this study, vortioxetine did not appear to be associated with an increase in suicidal ideation and risk of suicide in children, and suicidal ideation was reported by a lower percentage of patients in the current study (0.7%-3.7%) compared with the adolescent study (8%-10%).²⁰ Selective serotonin reuptake inhibitors may be associated with an increased risk of mania and hypomania, especially in young children.³⁶ Vortioxetine did not appear to induce mania in children, as indicated in the GBI Mania subscale. Among the few patients with a parent-reported GBI Mania subscale score ≥18 points, no hypomania/mania was confirmed.

Historically, most studies of antidepressants for treatment of MDD in a pediatric population have focused on adolescents, with only some involving young children between the ages of 7 and 11.^33,37,38 Furthermore, most studies analyze the pediatric population as a whole, with data for younger children limited to subgroup analyses. Considering the gap in data, this study offers context regarding the differences in MDD characteristics and the use of antidepressants, such as vortioxetine, specifically in young children. In addition, understanding the safety and tolerability of antidepressants in young children will pave the way for further research in this population.

Although this study was a failed trial, as it did not meet the primary efficacy end point, it is important to consider some of the limitations. This was a relatively short study with a treatment duration of 8 weeks, which may have contributed to inadequate response in some patients. Despite efforts to improve study design, the short duration of the placebo lead-in period may have resulted in a high placebo response rate. Additionally, the multiple exclusion criteria may limit generalizability to the wider population of children with MDD treated in clinical practice.

Improvements in the primary end point of CDRS-R total score from randomization to week 8 were shown for both the vortioxetine groups and placebo group, and there were no statistically significant differences between these groups among pediatric patients. Similar findings were observed for secondary efficacy measures, with none of them significantly different from placebo. The safety and tolerability of vortioxetine in children were in line with that previously established in adolescent and adult MDD populations.

CRediT authorship contribution statement

Michael Huss: Writing – review & editing, Validation, Supervision, Project administration, Methodology, Investigation, Conceptualization. Robert L. Findling: Writing – review & editing, Methodology, Investigation, Conceptualization. Melissa P. DelBello: Writing – review & editing, Visualization, Methodology, Investigation, Conceptualization. Oscar Necking: Writing – review & editing, Supervision, Project administration. Maria L. Petersen: Writing – review & editing, Project administration, Conceptualization. Simon Nitschky Schmidt: Writing – review & editing, Visualization, Validation, Software, Methodology, Formal analysis, Data curation, Conceptualization. Monika Rosen: Writing – review & editing, Supervision, Methodology, Conceptualization.