Mar 3, 2015 - treatment-resistant depression, which showed OFC to be su- perior to fluoxetine alone.16-18 OFC has also demonstrated rapid and robust ...
NEW RESEARCH Olanzapine/Fluoxetine Combination in Children and Adolescents With Bipolar I Depression: A Randomized, Double-Blind, Placebo-Controlled Trial Holland C. Detke,
PhD,
Melissa P. DelBello,
Objective: To assess the efficacy and safety of olanzapine/fluoxetine combination (OFC) for the acute treatment of bipolar depression in children and adolescents. Method: Patients 10 to 17 years of age with bipolar I disorder (BP-I), depressed episode, baseline Children’s Depression Rating Scale–Revised (CDRS-R) total score 40, Young Mania Rating Scale (YMRS) total score 15, and YMRS-item 1 2 were randomized to OFC (6/25–12/50 mg/day olanzapine/fluoxetine; n ¼ 170) or placebo (n ¼ 85) for up to 8 weeks of double-blind treatment. The primary efficacy measure was mean change in CDRS-R using mixed-model repeated-measures methodology. Results: Baseline-to-week-8 least-squares mean change in CDRS-R total score was greater for OFC-treated patients than for placebo-treated patients (28.4 versus 23.4, p ¼ .003; effect size ¼ .46), with between-group differences statistically significant at week 1 (p ¼ .02) and all subsequent visits (all p < .01). Rates of and times to response and remission were statistically significantly greater for OFC- than for placebo-treated patients. The most frequent treatment-emergent adverse events in the OFC group were weight gain, increased appetite, and somnolence.
T
reatment of the depressive phase of bipolar disorder (BD) in children and adolescents remains a significant unmet medical need. Relative to those diagnosed in adulthood, patients diagnosed with BD during childhood or adolescence have significantly reduced quality of life1 and greater risk of poorer psychiatric, psychosocial, and healthrelated outcomes,2 including greater risk of suicidal behavior.3,4 Inadequate or delayed symptom control in these young patients may have serious impacts on social, emotional, and educational development that can leave this already-vulnerable population even further impaired and disadvantaged. Consequently, rapid and reliable intervention is critical. Unfortunately, the pharmacological treatment of bipolar depression poses a major therapeutic challenge. Unopposed antidepressant therapy may potentially precipitate manic symptoms,5 and although there are multiple medications effective for the treatment of mania, treatment of the
Clinical guidance is available at the end of this article.
JOURNAL OF THE AMERICAN ACADEMY OF C HILD & ADOLESCENT PSYCHIATRY VOLUME 54 NUMBER 3 MARCH 2015
MD,
John Landry,
MMath,
Roland W. Usher,
MS
Mean weight gain at patient’s endpoint was significantly greater for OFC- than for placebo-treated patients (4.4 kg versus 0.5 kg, p < .001). Treatment-emergent hyperlipidemia was very common among OFC-treated patients. Abnormal increases in hepatic analytes, prolactin, and corrected QT interval (QTc) were also common or very common but generally not clinically significant. Conclusion: In this study, OFC was superior to placebo, and has been approved by the US Food and Drug Administration (FDA) for the acute treatment of bipolar I depression in patients 10 to 17 years of age. Benefits should be weighed against the risk of adverse events, particularly weight gain and hyperlipidemia. Clinical trial registration information—A Study for Assessing Treatment of Patients Ages 10-17 with Bipolar Depression; http://clinicaltrials.gov; NCT00844857. Key Words: olanzapine fluoxetine combination, bipolar, depression J Am Acad Child Adolesc Psychiatry 2015;54(3):217–224.
depressive phase of the illness remains understudied in both children and adults, and many frequently used treatments, such as lithium, valproate, and lamotrigine, have limited or no evidence to support their use for the acute treatment of bipolar depression.6 At present, only 3 medications are approved for the acute treatment of bipolar depression in adults in the United States by the US Food and Drug Administration (FDA): quetiapine, lurasidone, and olanzapine/fluoxetine combination (OFC). With regard to testing these medications in children and adolescents, 2 randomized, double-blind studies of quetiapine were conducted in patients 10 to 17 years of age diagnosed with bipolar depression, but both studies failed to demonstrate a significant benefit of quetiapine versus placebo in that population.7,8 Pediatric trials of lurasidone are still pending. OFC is a fixed-dose combination capsule, available in generic preparation, which contains the second-generation antipsychotic (SGA) olanzapine and the selective serotonin reuptake inhibitor (SSRI) fluoxetine. Olanzapine and fluoxetine can also be separately co-administered to achieve the same approximate proportions as the fixed-combination OFC capsule.9,10 Olanzapine is an effective mood stabilizer11-13 that appears to enhance the antidepressant action
www.jaacap.org
217
DETKE et al.
of fluoxetine,14,15 as demonstrated in adult studies of treatment-resistant depression, which showed OFC to be superior to fluoxetine alone.16-18 OFC has also demonstrated rapid and robust improvement of depressive symptoms in adults with bipolar I depression19,20 without increasing risk of switch to mania, and was superior to placebo,19 olanzapine monotherapy,19 and lamotrigine21 during acute treatment. Although OFC has not previously been studied in patients 1 point from the standard) before rating patients on those scales in the study. Safety measures included assessment of adverse events (AEs), vital signs, electrocardiograms, and laboratory analytes; EPS using the Barnes Akathisia Rating Scale,34 Abnormal Involuntary Movements Scale,35 and Simpson-Angus Scale36; and suicidality using the Columbia–Suicide Severity Rating Scale (C-SSRS).37
Statistical Analyses Analyses were based on a modified intention-to-treat (mITT) dataset (N ¼ 255) which excluded patients who did not receive study drug (n ¼ 4) and all patients from 2 sites that had significant data integrity violations such that data could not be considered credible or reliable (n ¼ 32). After notification by the sponsor, both sites were audited by the FDA, who agreed with the decision to use mITT as the primary analysis set for all efficacy and safety conclusions. Analyses were also conducted using the full ITT dataset but did not reveal any significant differences in findings from the mITT dataset. All results presented represent mITT analyses unless otherwise specified. Continuous efficacy analyses, including the primary analysis of mean change in CDRS-R, were based on mixed-model repeatedmeasures (MMRM) methodology using an unstructured correlation matrix with terms for baseline, country, treatment, visit, and treatment visit interaction. Effect size was calculated post hoc using Cohen’s d.38 Continuous safety data were analyzed using lastobservation-carried-forward (LOCF) methodology, with treatment comparisons based on analysis of covariance (ANCOVA) including terms for treatment and baseline. Mean weight change was analyzed using MMRM modeling with terms for baseline, treatment, visit, and treatment visit interaction. Means from the MMRM and ANCOVA models represent least-squares means. Times to response and remission were assessed using Kaplan–Meier methodology and the log-rank test based on the earliest time at which patients met the respective criteria. Categorical data were analyzed using Fisher’s exact test. All tests were 2-sided using a type-I error of 0.05, with no adjustments for multiple comparisons. Thus, only the primary efficacy measure comparison can be considered confirmatory; all other comparisons are considered supportive only.
Baseline Patient Characteristics
Variable Age (y), mean (SD) 10e12 years of age, n (%) 13e17 years of age, n (%) Sex, male, n (%) Race, white, n (%) Ethnicity, Hispanic, n (%) No. of previous episodes mania, median No. of previous episodes depression, median CDRS-R total score, mean (SD) YMRS total score, mean (SD) CGI-BP overall severity, mean (SD) Current comorbid ADHD diagnosis, n (%)
OFC (n ¼ 170) 14.6 29 141 84 119 38 1 2 54.6 6.1 4.4 45
(2.3) (17.1) (82.9) (49.4) (70.0) (22.4)
(10.0) (3.8) (0.7) (26.5)
Placebo (n ¼ 85) 15.0 10 75 46 61 23 2 2 53.7 6.2 4.3 17
(2.1) (11.8) (88.2) (54.1) (71.8) (27.1)
(8.1) (3.9) (0.7) (20.0)
Total (N ¼ 255) 14.7 39 216 130 180 61 1 2 54.3 6.1 4.4 62
(2.3) (15.3) (84.7) (51.0) (70.6) (23.9)
(9.4) (3.8) (0.7) (24.3)
Note: ADHD ¼ attention-deficit/hyperactivity disorder; CDRS-R ¼ Children’s Depression Rating ScaleeRevised; CGI-BP ¼ Clinical Global Impressions ScaleeBipolar Version; OFC ¼ olanzapine/fluoxetine combination; YMRS ¼Young Mania Rating Scale.
JOURNAL OF THE AMERICAN ACADEMY OF C HILD & ADOLESCENT PSYCHIATRY VOLUME 54 NUMBER 3 MARCH 2015
www.jaacap.org
219
DETKE et al.
RESULTS
statistically significantly greater improvement relative to placebo at week 1 (p ¼ .02) and at all subsequent visits (all p values < .01) (Figure 2). Analysis using the full ITT dataset also found superiority of OFC to placebo on mean change in CDRS-R from baseline to week 8 (p ¼ .003). There was no significant interaction between treatment effect and country on the CDRS-R. More OFC-treated patients achieved response at endpoint (78.2%) than placebo-treated patients (59.2%; p ¼ .003), and time to response was faster with OFC (median of w3 weeks) than placebo (median of w5 weeks, p ¼ .001). More OFC-treated patients also achieved remission (59.0%) than placebo-treated patients (43.4%; p ¼ .035), and time to remission was faster with OFC than placebo (p ¼ .028). Median time to remission was w7 weeks for OFC-treated patients; too few placebo patients achieved remission to calculate a median. Mean changes in secondary efficacy measures are also presented in Table 2. OFC-treated patients demonstrated greater improvement than placebo on all measures of depressive symptomatology and overall illness severity (all p values 1 patient were aggression and suicidal ideation (both 1.2% in each group). SAEs occurring in 1 patient each were agitation, BD, homicidal ideation, self-injurious behavior, suicide attempt, and
Efficacy On the primary efficacy analysis, the OFC group had a 5-point greater mean decrease on the CDRS-R at week 8 than did the placebo group (t ¼ 3.07, p ¼ .003; Table 2). Effect size for OFC at week 8 was 0.46. The OFC group showed TABLE 2
Changes in Efficacy Measures From Baseline to Week 8 (Mixed-Model Repeated Measures) OFC n ¼ 170 Mean Baseline (SD)
CDRS-R total score BDRS total score CGI-BP overall severity CGI-BP severity of depression CGI-BP severity of mania YMRS total score
OFC n ¼ 170 LS Mean Change (SE)
Placebo n ¼ 85 Mean Baseline (SD)
Placebo n ¼ 85 LS Mean Change (SE)
Between-Group Difference (95% CI)
Between-Group p Value
54.6 25.1 4.4 4.5
(10.0) (5.7) (0.7) (0.7)
28.4 16.5 2.2 2.4
(1.1) (0.8) (0.1) (0.1)
53.7 25.1 4.3 4.4
(8.2) (5.4) (0.7) (0.7)
23.4 13.4 1.8 2.0
(1.5) (1.1) (0.2) (0.2)
5.0 3.0 0.4 0.4
(8.3, (5.5, (0.7, (0.7,
1.8) 0.6) 0.0) 0.0)
.003 .016 .030 .038
1.4 6.1
(0.7) (3.8)
0.1 2.0
(0.1) (0.5)
1.4 6.2
(0.7) (3.9)
0.0 1.6
(0.1) (0.6)
0.1 0.5
(0.3, (1.9,
0.2) 1.0)
.607 .527
Note: BDRS ¼ Bipolar Depression Rating Scale; CDRS-R ¼ Children’s Depression Rating ScaleeRevised; CGI-BP ¼ Clinical Global Impressions Scale-eBipolar Version; LS ¼ least-squares; OFC ¼ olanzapine/fluoxetine combination; SE ¼ standard error; YMRS ¼ Young Mania Rating Scale.
JOURNAL 220
www.jaacap.org
OF THE
AMERICAN ACADEMY OF C HILD & ADOLESCENT PSYCHIATRY VOLUME 54 NUMBER 3 MARCH 2015
OLANZAPINE/FLUOXETINE COMBINATION IN PEDIATRIC BIPOLAR DEPRESSION
FIGURE 2 Visitwise mean change in Children’s Depression Rating Scale–Revised (CDRS-R) total score (mixed-model repeated measures). Note: LS ¼ least-squares; OFC ¼ olanzapine/ fluoxetine combination.
ovulation disorder in the OFC group and depression, psychotic disorder, anaphylactic reaction, and pyoderma in the placebo group. No deaths occurred. Weight and Vital Signs. Mean baseline-to-endpoint weight change (LOCF) was significantly greater for OFCtreated patients (4.4 kg, standard error [SE] ¼ 0.2) than for placebo-treated patients (0.5 kg, SE ¼ 0.3; p < .001). Using MMRM methodology to model expected weight gain if all patients completed the 8-week study, mean weight change from baseline to week 8 was 5.1 kg (SE ¼ 0.3) for OFC-treated versus 0.6 kg (SE ¼ 0.4) for placebo-treated patients (p < .001). Baseline-to-endpoint mean change in body mass index was also significantly greater for the OFC group (1.5 kg/m2, SE ¼ 0.1) than for the placebo group (0.1 kg/m2, SE ¼ 0.1; p < .001). Incidence of gaining 7% of baseline body weight was significantly greater for the OFC group relative to placebo (52% versus 4%, p < .001), as was incidence of gaining 15% of baseline body weight (14% versus 0%, p < .001). No placebo patients and 2 OFCtreated patients gained 25% of their baseline. There were no statistically significant between-group differences in baseline-to-endpoint mean changes in blood pressure, pulse, or body temperature. Orthostatic hypotension did not occur statistically significantly more often with OFC than with placebo. Lipids and Glucose. OFC-treated patients had mean baseline-to-endpoint increases in fasting total cholesterol (16.3 mg/dL, SE ¼ 1.9), LDL cholesterol (9.7 mg/dL, SE ¼ 1.5), and triglycerides (35.4 mg/dL, SE ¼ 5.3), whereas placebo-treated patients had small mean decreases (all between-group p values 20 ng/mL) relative to placebo (58.1% versus 15.6%, p < .001). Five female OFC-treated patients developed AEs potentially related to prolactin elevation: galactorrhea, dysmenorrhea, and ovulation disorder (possible mittelschmerz). Regarding liver enzymes, OFC-treated patients experienced greater baseline-to-endpoint mean increases in alanine aminotransferase (ALT ¼ 7.6 U/L [SE ¼ 1.5]) and g-glutamyl transferase (GGT, 2.3 U/L [SE ¼ 0.6]) compared with the placebo group, which showed little or no change. Although a significant number of OFC- versus placebo-treated patients shifted from within normal limits at baseline to 1 ULN at any time postbaseline on ALT (48% versus 3%) and aspartate transaminase (AST, 37% versus 10%), most of these changes were transient and not clinically significant. Five OFCtreated patients had ALT shifts from normal to 3 ULN. No patients met Hy’s rule (ALT 3 ULN and total bilirubin 2 ULN), and no patients developed hepaticrelated symptomatology. Electrocardiographic Findings. Mean baseline-to-endpoint increase in heart rate was greater for OFC-treated patients (4.5 beats/min, SE ¼ 0.8) than for placebo-treated (1.0 beats/ min, SE ¼ 1.2, p ¼ .013). Mean baseline-to-endpoint increase in Fridericia-corrected QT interval (QTcF) was also greater for OFC-treated patients (8.2 milliseconds, SE ¼ 1.0) than for the placebo-treated patients (1.1 milliseconds, SE ¼ 1.4, p < .001). A total of 12.0% of OFC-treated patients experienced QTcF increase 30 milliseconds from baseline compared to 1.3% of placebo-treated patients (p ¼ .005). Only 1 patient developed QTcF 460 milliseconds (maximum of 464 milliseconds), and no patients increased by 60 milliseconds from baseline at any time. Extrapyramidal Symptoms. Mean changes on the EPS scales were near 0, with no statistically significant differences between treatment groups. Rates of treatmentemergent akathisia, parkinsonism, and dyskinesia were all 1.3% in both groups and were not statistically different between groups. Suicidality. Rates of suicidal ideation and behavior were low throughout the study, with no statistically significant difference between groups. Based on the C-SSRS, treatmentemergent suicidal ideation occurred in 9.4% of OFC-treated patients and 13.1% of placebo-treated patients, and treatment-emergent suicidal behavior occurred in 1.8% of OFC-treated patients and 2.4% of placebo-treated patients.
DISCUSSION
This was the first controlled clinical trial to evaluate OFC in a pediatric population. The study met its primary objective, as OFC was found to be statistically superior to placebo in the
www.jaacap.org
221
DETKE et al.
TABLE 3 Treatment-Emergent Adverse Events in 5% of Olanzapine/Fluoxetine Combination (OFC)–Treated Patients OFC n ¼ 170
Any adverse event Weight gain Appetite increase Headache Somnolence Tremor Blood triglycerides increase Fatigue Vomiting Sedation
Placebo n ¼ 85
n
(%)
n
(%)
Between-Group p Value
125 34 28 27 27 15 12
(73.5) (20.0) (16.5) (15.9) (15.9) (8.8) (7.1)
49 1 1 12 2 1 2
(57.6) (1.2) (1.2) (14.1) (2.4) (1.2) (2.4)
.015
