Skip to main content
The Cochrane Database of Systematic Reviews logoLink to The Cochrane Database of Systematic Reviews
. 2023 Jul 28;2023(7):CD013866. doi: 10.1002/14651858.CD013866.pub2

Brexpiprazole versus placebo or other antidepressive agents for treating depression

Silviya Ralovska 1, Ivan Koychev 2, Petar Marinov 1, Toshi A Furukawa 3, Benoit Mulsant 4,5, Andrea Cipriani 2,6,
Editor: Cochrane Common Mental Disorders Group
PMCID: PMC10406422

Abstract

Background

Brexpiprazole is the newest antipsychotic drug approved as an adjunctive for treatment‐resistant depression. This systematic review provides comprehensive, high‐quality evidence for the effects of brexpiprazole in major depressive disorder (MDD) based on existing and emerging randomised controlled trials (RCTs).

Objectives

To evaluate the benefits and harms of brexpiprazole as monotherapy or adjunct treatment in the acute and longer‐term treatment of MDD compared to placebo or other antidepressive agents.

Search methods

We used standard, extensive Cochrane search methods. The latest search date was November 2021.

Selection criteria

We included RCTs in people with a history of non‐response to one or more antidepressant monotherapies comparing brexpiprazole as monotherapy or adjunct treatment with placebo or other antidepressive agents.

Data collection and analysis

We used standard Cochrane methods. Primary outcomes were 1. response to treatment, measured as number of participants who achieved 50% or more reduction of the score on a validated scale for depression, 2. dropouts due to any reason, and 3. dropouts due to adverse effects, all measured at eight weeks (acute treatment). Our secondary outcomes were 4. number of participants who responded to treatment after two and 18 weeks, number of participants who achieved remission after eight weeks, 5. total number of participants with any adverse events, 6. social functioning or adjustment, and 7. health‐related quality of life. We used GRADE to assess certainty of evidence for each outcome.

Main results

We retrieved nine studies with 3424 participants with treatment‐resistant depression, defined as lack of response to at least two trials of antidepressant monotherapy. The age of participants was 18 to 65 years in seven studies, 18 to 75 years in one study, and older than 65 years in one study. Inclusion criteria for all studies required a diagnosis of MDD with a current major depressive episode per Diagnostic and Statistical Manual of Mental Disorders Text Revision Fourth Edition criteria and additional criteria on minimal scores on Montgomery–Åsberg Depression Rating Scale (MADRS) and Hamilton Depression Rating Scale which varied across studies. Eight studies compared brexpiprazole as adjunctive treatment to an antidepressant versus placebo as adjunctive treatment to an antidepressant. One study added brexpiprazole or placebo to a combination of antidepressant and ketamine. Four studies used fixed dosages and five studies used flexible dosages of brexpiprazole. The manufacturer sponsored eight studies and one study was independently funded.

Brexpiprazole was superior in achieving response at eight weeks (odds ratio (OR) 1.47, 95% confidence interval (CI) 1.23 to 1.75; 8 studies, 3409 participants; high‐certainty evidence). More participants randomised to brexpiprazole dropped out due to any cause at eight weeks (OR 1.37, 95% CI 1.02 to 1.83; 7 studies, 2523 participants; high‐certainty evidence) or due to adverse effects (OR 2.88, 95% CI 1.37 to 6.05; 6 studies, 2472 participants; high‐certainty evidence). Brexpiprazole was superior to placebo when added to an antidepressant for achieving remission at eight weeks (OR 1.46, 95% CI 1.19 to 1.79; 7 studies, 3358 participants; high‐certainty evidence). When response was defined based on a change in the score on MADRS, brexpiprazole also demonstrated better efficacy than placebo (mean difference (MD) –1.39, 95% CI –1.96 to –0.82; 8 studies, 3263 participants; high‐certainty evidence). Compared to placebo, add‐on brexpiprazole is probably more likely to result in akathisia (OR 2.95, 95% CI 2.06 to 4.21; 7 studies, 3358 participants; moderate‐certainty evidence) and weight gain (OR 3.14, 95% CI 2.19 to 4.49; 9 studies, 3424 participants; moderate‐certainty evidence).

The risk of bias varied between low and unclear for most studies. Only one study was at high risk of selective reporting bias and other bias; however, this study had a small sample size and its impact on overall risk of bias was not considered significant. The certainty of the evidence on five of the main outcomes (response to treatment, measured as number of participants who achieved response and change in depressive symptoms, remission, dropouts due to any cause, dropouts due to adverse effects) was high. The certainty of evidence for akathisia and weight gain was downgraded one level to moderate due to indirectness.

Authors' conclusions

Our review supports the superior efficacy of add‐on brexpiprazole in comparison to placebo in adults with treatment‐resistant depression for the short‐term and acute treatment of depression. Our findings were limited by the small number of studies and the moderate‐certainty evidence for the most common adverse effects (akathisia and weight gain). There was insufficient evidence for the effects of brexpiprazole for depression in older people or children. In addition, we found no studies that compared brexpiprazole to other antidepressants or where brexpiprazole was added after lack of response to only one antidepressant. We also lacked sufficient data to establish the long‐term efficacy of brexpiprazole as adjunctive treatment. Active comparators, long‐term studies in different age groups and cost‐effect analyses are needed to more precisely establish the clinical role of brexpiprazole in the treatment of MDD.

Plain language summary

What are the benefits and harms of brexpiprazole, a new antipsychotic medication, for the treatment of depression when compared to dummy tablets or antidepressants?

Key messages

– Brexpiprazole added to an antidepressant is better at reducing the symptoms of depression than placebo (dummy tablet) added to an antidepressant.

– Brexpiprazole might be more commonly associated with weight gain and akathisia (a feeling of restlessness that causes a compelling need to move all the time).

What is depression?

Depression is a common mental health condition which leads to low mood, inability to experience pleasure, sleep disruption, weight loss, fatigue or low energy, slower speech and movements, feeling worthless or excessively guilty, being unable to concentrate and having thoughts of committing suicide. To be diagnosed with depression, a person should experience five or more of these symptoms, including low mood or loss of sense of pleasure for at least two weeks.

How is the condition treated?

Unless the symptoms of depression are mild, the first option for treatment is an antidepressant medication, but antidepressants only help in half of cases when prescribed for the first time. When an antidepressant is not enough to treat depression, one available option is to add another medication to help the effects of the antidepressant. Other options include changing the antidepressant to another antidepressant, talking therapy or stimulation of the brain through electrical pulses. Brexpiprazole is a new medication which can be added to an antidepressant that has not worked by itself. Brexpiprazole can be prescribed in doses from 0.5 mg to 3 mg for the treatment of depression. It has also been approved for the treatment of schizophrenia.

What did we want to find out?

We wanted to examine the effects of brexpiprazole in the treatment of depression when prescribed by itself or when added to another antidepressant and how its effects compare to the effects of antidepressants or placebo. To do so we explored several effects. First, we looked into how well brexpiprazole decreased the symptoms of depression. Second, we explored how many participants gave up taking part in the studies for any reason. Third, we investigated how many stopped participating because of the unwanted effects they experienced. Fourth, we explored what unwanted effects emerged with the use of brexpiprazole. We also wanted to know how brexpiprazole affected the quality of life of the participants and how it affected the way they functioned in their everyday lives.

What did we do?

We searched scientific databases for studies that reported the effects of brexpiprazole versus placebo or antidepressants for depression. We included only studies in which the participants were assigned randomly to one or the other treatment. Neither them nor the team that was evaluating the effects of the treatment knew what was prescribed and this helped to objectively evaluate the results without the knowledge of what was prescribed impacting on the outcome. We compared and summarised the results and evaluated how confident we were that the results of the studies represented the real effects of the medication. We based our confidence on different factors such as the size of the effects, the methods used, and what and how the results were reported.

What did we find and what were the limitations of our findings?

We found nine studies, which included 3424 participants; two‐thirds of them were women and one‐third were men.

We only found studies in which brexpiprazole was added after the participants had tried treatment with two or more antidepressants on their own. We have a high level of confidence that brexpiprazole added to an antidepressant reduced the symptoms of depression more after eight weeks of treatment than placebo added to an antidepressant. Brexpiprazole was also more likely to lead to no symptoms of depression than placebo when either of them was added to an antidepressant. We are moderately confident that brexpiprazole caused weight gain or subjective feelings of restlessness more often than placebo. We are less confident about the results around these unwanted effects as we could not find information about how they were measured and whether all studies used the same methods to measure them.

What did we fail to find?

We found no studies that explored how brexpiprazole would work for children or older people who have depression. We wanted to find how brexpiprazole would work when given for depression for longer periods of time, such as 18 to 24 weeks, but we found only one study and its results were not enough to draw a conclusion.

How up‐to‐date is this evidence?

Our evidence is up‐to‐date to 1 November 2021.

Summary of findings

Summary of findings 1. Brexpiprazole compared to antidepressive agents for the treatment of depression.

Brexpiprazole compared to other antidepressive agents for the treatment of depression
Patient or population: participants of any age with a primary diagnosis of major depressive disorder
Setting: any
Intervention: brexpiprazole
Comparison: other antidepressive agents
Outcomes Anticipated absolute effects* (95% CI) Relative effect
(95% CI) № of participants
(studies) Certainty of the evidence
(GRADE) Comments
Risk with antidepressive agentsa Risk with brexpiprazole
Response to treatment at 8 weeks Study population OR 1.47
(1.23 to 1.75) 3409
(8 RCTs) ⊕⊕⊕⊕
Highb Adjunctive brexpiprazole demonstrates better efficacy in achieving response to treatment at 8 weeks in comparison to adjunctive placebo.
176 per 1000 239 per 1000
(208 to 272)
Number of dropouts due to any cause at 8 weeks Study population OR 1.37
(1.02 to 1.83) 2523
(7 RCTs) ⊕⊕⊕⊕
High Participants receiving placebo as adjunctive treatment are less likely to leave the study for any reason in comparison with those who receive brexpiprazole.
74 per 1000 100 per 1000
(76 to 129)
Number of dropouts due to adverse effects at 8 weeks Study population OR 2.88
(1.37 to 6.05) 2472
(6 RCTs) ⊕⊕⊕⊕
High Brexpiprazole is more likely to lead to dropouts due to adverse effects at 8 weeks in comparison to placebo.
8 per 1000 25 per 1000
(12 to 50)
Remission rates at 8 weeks Study population OR 1.46
(1.19 to 1.79) 3358
(7 RCTs) ⊕⊕⊕⊕
High Brexpiprazole is more likely to lead to remission of MDD at 8 weeks in comparison to placebo.
119 per 1000 165 per 1000
(139 to 195)
Change in depressive symptoms at 8 weeks The change mean score in the control group ranged from –8.07 to 22.13 The change mean score in the control group ranged from –10.4 to 18.39 MD 1.39 lower
(1.96 lower to 0.82 lower) 3263
(8 RCTs)
⊕⊕⊕⊕
Highb Brexpiprazole leads to a slightly better reduction in depressive symptoms in comparison to placebo.
Adverse effects: akathisia (followed up until the end of the longest follow‐up period of the included trials) Study population OR 2.95
(2.06 to 4.21) 3358
(7 RCTs) ⊕⊕⊕⊝
Moderatec Brexpiprazole is probably more likely to cause akathisia than placebo.
27 per 1000 76 per 1000
(55 to 106)
Adverse effects: weight gain (followed up until the end of the longest follow‐up period of the included trials) Study population OR 3.14
(2.19 to 4.49) 3424
(9 RCTs) ⊕⊕⊕⊝
Moderated Brexpiprazole is probably more likely to cause gain in weight in comparison to placebo.
26 per 1000 76 per 1000
(54 to 105)
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: confidence interval; MD: mean difference; MDD: major depressive disorder; OR: odds ratio; RCT: randomised controlled trial.
GRADE Working Group grades of evidenceHigh certainty: we are very confident that the true effect lies close to that of the estimate of the effect.
Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different.
Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect.
Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.

aControl group risk estimates came from pooled estimates of control groups.
bOne study did not report efficacy results and was at high risk of selective reporting bias (NCT01837797). The study included only 15 participants out of 3409 included for this outcome and we decided not to downgrade the overall certainty of evidence.
cMissing information on the process of akathisia assessment. Presence of indirectness was highly likely. Downgraded one level.
dMissing information on the process of weight gain assessment and weight gain criteria. Presence of indirectness is highly likely due to high probability of differences in outcome measures. Downgraded one level.

Background

Description of the condition

Major depressive disorder (MDD) is a chronic mental health condition that affects over 163 million people worldwide; it is the third cause of disability in the global burden of disease (GBD 2017). Persistent low mood and diminished interest or pleasure in daily activities form the core symptoms of the disorder; they are accompanied by significant weight loss, fatigue or loss of energy, psychomotor retardation (slowing of mental and physical activity), feelings of worthlessness, excessive or inappropriate guilt, diminished ability to concentrate and suicidal ideation (APA 2022). According to the Diagnostic and Statistical Manual of Mental Disorders 5th Edition Text Revision (DSM‐5‐TR), five or more symptoms, including at least one of the main symptoms, and an episode duration of two or more weeks, are required for the diagnosis of MDD (APA 2022).

The treatment of depression includes pharmacological and psychological interventions, electroconvulsive therapy (ECT) and other treatment options. The choice of treatment for each person depends on various factors, such as the severity and duration of the depressive episode, response to treatment, risk–benefit ratio of the intervention and the person's preferences (NICE 2009). According to international guidelines, antidepressants are often prescribed as first‐line treatment for moderate‐to‐severe episodes of major depression (APA 2010; Kennedy 2016; NICE 2009). Based on slight differences in their pharmacological properties, antidepressants are grouped into classes, such as selective serotonin reuptake inhibitors (SSRI), serotonin noradrenaline reuptake inhibitors (SNRI), tricyclic antidepressants (TCA) and monoamine oxidase inhibitors (MAOI) (APA 2010). Overall, antidepressants are more efficacious than placebo for the acute treatment of depression in adults (Cipriani 2018). However, people with MDD vary substantially in their response rates (Kennedy 2016); when treated with a single antidepressant, as many as 50% of people with MDD do not reach a response threshold (Salanti 2018), measured as a 50% reduction of the score on a validated scale for depression (Bauer 2002). A variety of terms have been used to describe the degree of non‐responsiveness to treatment, including non‐response and partial response (Bauer 2002; Papakostas 2020). A person's depression is typically considered to be treatment resistant when it does not respond to a course of two or more antidepressants of adequate dosage and duration (Malhi 2016), although more specific criteria, such as additional lack of response to psychotherapy, have also been suggested (Fava 2020). Currently, in cases in which MDD does not respond to antidepressant monotherapy, switching to another antidepressant, psychotherapy or neurostimulation, and adjunctive treatment (i.e. adding another treatment to the primary treatment) have been recommended (Kennedy 2016; NICE 2009).

Description of the intervention

In July 2015, the US Food and Drug Administration (FDA) approved brexpiprazole as "monotherapy for schizophrenia in adolescents and adults", and "adjunctive treatment for adults with MDD" in the USA (FDA 2015a). In Europe, it is indicated only for the treatment of schizophrenia (EMA 2018). Brexpiprazole is a partial agonist of serotonin 5HT‐1A and dopamine D2 receptors, and an antagonist of serotonin 5HT‐2A receptors. It also has an antagonistic action with moderate affinity for the adrenergic α1/2‐R, histamine H1 receptors, and a low affinity for the muscarinic M receptors (Maeda 2014a; Otsuka 2020). Peak plasma concentration of brexpiprazole is achieved four hours after dose administration, with 95% bioavailability. Its terminal half‐life is 91 hours; the recommended regimen is once daily, 2 mg/day to 4 mg/day for schizophrenia, and 0.5 mg/day to 3 mg/day as add‐on treatment for depression (Otsuka 2020). Brexpiprazole is metabolised primarily by CYP2D6 and CYP3A4 enzymes, and, therefore, requires dosage adjustment when administered to people with poor CYP2D6 metabolism, or concomitantly with medications that inhibit CYP2D6, as well as with medications that induce or inhibit CYP3A4. (Otsuka 2020).

How the intervention might work

The mechanism by which brexpiprazole produces its antidepressant effects remains unclear. Preclinical studies compared brexpiprazole to aripiprazole (Maeda 2014a; Maeda 2014b; Stahl 2016), which has a similar pharmacological profile and has also been approved as adjunct treatment for major depression in adults (FDA 2007). Compared to aripiprazole, brexpiprazole has a lower intrinsic affinity for D2‐receptors, and a stronger affinity for serotonin 5HT‐1A and 5HT‐2A, and adrenergic α1B receptors, which should lead to a lower incidence of activation and extrapyramidal symptoms (Stahl 2016). Brexpiprazole has also a lower affinity for histamine H1 receptors, which should lead to lower rates of sedation and weight gain (Maeda 2014a; Stahl 2016). On the basis of these specific actions on dopamine D2, serotonin 5HT‐1A, serotonin 5HT‐2A, adrenergic α1B and histamine H1 receptors, it was suggested that brexpiprazole should be equally efficacious to aripiprazole as adjunctive treatment for major depression, while causing fewer adverse effects (Stahl 2016). One preclinical study of cognition in animal models of schizophrenia, sponsored by the manufacturer, found that brexpiprazole led to better cognitive performance in executive function domains than aripiprazole, and suggested this was due brexpiprazole's different affinities for dopamine D2, and serotonin 5HT‐2A and 5HT‐1A receptors (Maeda 2014b). Brexpiprazole as adjunct treatment led to improved cognition and functioning in one open‐label study in people with MDD (Fava 2017). To investigate the efficacy and safety of brexpiprazole as add on treatment for MDD in clinical settings, one short‐term Phase 3, double‐blind, randomised controlled trial (RCT), sponsored by Otsuka Pharmaceutical Co, randomised 379 people with MDD who did not respond to an eight‐week course of antidepressant monotherapy, to brexpiprazole or placebo as add‐on treatment for six weeks (Thase 2015a). Brexpiprazole 2 mg improved depression symptoms on the Montgomery–Åsberg Depression Rating Scale (MADRS) more than placebo (mean difference (MD) –3.12, 95% confidence interval (CI) –4.70 to –1.54; P < 0.0001). Those who received brexpiprazole more often had weight gain (brexpiprazole group: 7.98% versus placebo group: 3.14%), and more often had akathisia, a 'subjective feeling of motor restlessness manifested by a compelling need to be in constant movement' (Sadock 2015) (brexpiprazole group: 7.45% versus placebo group: 1.05%). Another short‐term Phase III RCT, with a similar design, suggested that the efficacy and tolerability of brexpiprazole in people with depression might be dose dependent (Thase 2015b). In that study, 1539 people were randomised to brexpiprazole 1 mg, brexpiprazole 3 mg or placebo, adjunctive to an antidepressant. Brexpiprazole 1 mg failed to reduce depression symptoms on the MADRS compared to placebo (MD –1.30, 95% CI –2.73 to 0.13; P < 0.0737). Brexpiprazole 3 mg was more efficacious than placebo on the MADRS (MD –1.95, 95% CI –3.39 to –0.51; P < 0.0079), but participants reported higher rates of akathisia (13.5%) than those who received brexpiprazole 1 mg (4.4%). These results suggested that higher dosages of brexpiprazole might be efficacious, but less tolerable than lower dosages, which in turn many not be efficacious.

Why it is important to do this review

In cases of an inadequate response to an antidepressant, recommendations for adjunct medications include an antipsychotic, another antidepressant, lithium, triiodothyronine and stimulants (Kennedy 2016). Brexpiprazole is the newest antipsychotic drug to receive approval as an adjunctive treatment for MDD, with several reviews of its efficacy and safety in depression already published (Citrome 2015; Kishi 2019; McIntyre 2016; Nelson 2018; Thase 2016; Thase 2019a; Yoon 2017). While informative, these reviews have a number of limitations. First, all existing reviews, except Kishi 2019, explored only short‐term effects (up to six‐week follow‐up), and were limited by the small and selective number of RCTs included. For instance, some reviews included only two RCTs (Citrome 2015; McIntyre 2016; Thase 2016; Yoon 2017), and Thase 2019a included four. Second, none of these reviews assessed the effect of brexpiprazole on depression in people under the age of 18 years. Third, some of these reviews mentioned a review protocol, but it was often not available, or was not easy to access. Fourth, the manufacturer of brexpiprazole sponsored several reviews (McIntyre 2016; Nelson 2018; Thase 2016; Thase 2019a). What is more, these reviews lacked a formal assessment of risk of bias. Finally, out of all of these reviews, only one evaluated the quality of the evidence using GRADE methodology (Yoon 2017). As there are more studies investigating the effect of brexpiprazole, we need a high‐quality review to properly assess the efficacy, tolerability, acceptability and safety of brexpiprazole as a treatment for MDD, and to guide clinical practice.

Objectives

To evaluate the benefits and harms of brexpiprazole as monotherapy or adjunct treatment in the acute and longer‐term treatment of MDD compared to placebo or other antidepressive agents.

Methods

Criteria for considering studies for this review

Types of studies

We included double‐blind RCTs. We decided to include cross‐over RCTs and cluster‐RCTs. Both published and unpublished studies were included. We decided to exclude quasi‐RCTs as allocation in this design is achieved by an approximation of randomisation (i.e. by alternate days, medical record number, etc.) (Furukawa 2016). Lack of randomisation in this study design poses a limitation to concluding a causal association between an intervention and an outcome (Schweizer 2016), and may introduce biases that may lead to loss of internal validity, the ability of a study to answer the research question correctly (Higgins 2011a; Schweizer 2016; Shadish 2002).

Types of participants

Diagnosis

We included participants with a primary diagnosis of MDD, according to Diagnostic and Statistical Manual of Mental Disorders 4th Edition (DSM‐IV; APA 1994), Diagnostic and Statistical Manual of Mental Disorders 4th Edition Text Revision (DSM‐IV‐TR; APA 2000), Diagnostic and Statistical Manual of Mental Disorders 5th Edition (DSM‐5; APA 2022), International Classification of Diseases 10th Revision (ICD‐10; WHO 1992), Feighner criteria (Feighner 1972), Research Diagnostic Criteria (Spitzer 1978), or any other standardised criteria for MDD.

We decided not to consider ICD‐9 (WHO 1978) for the diagnosis, because it lacked operationalised criteria. We excluded studies with more than 20% of participants with diagnosis of bipolar depression or psychotic depression (Furukawa 2016).

Characteristics

We included trials with participants of any age, of either sex and any ethnicity. We decided to include trials with people with MDD who did not respond to previous antidepressant treatment, and who were still experiencing acute depressive symptoms. For the purpose of the analysis, we divided participants into the following groups:

  • participants with history of an inadequate response (i.e. participants who reported a less than 50% reduction of the score on a standardised depression scale after treatment with one antidepressant as monotherapy, at adequate dose and duration (Papakostas 2020));

  • participants with treatment‐resistant depression (i.e. participants who did not respond to two or more trials with antidepressant monotherapy, at adequate dose and duration (Malhi 2016)).

Comorbidities

We excluded RCTs in which any participants had a concurrent primary diagnosis of another mental disorder, but we included trials in which participants had a secondary diagnosis of a mental disorder. We excluded RCTs including participants with serious medical illness (such as myocardial infarction, uncontrolled diabetes mellitus).

Setting

We included studies conducted in any setting (i.e. inpatients, general practice and outpatient specialist services, residents in care homes or nursing centres).

Types of interventions

Experimental intervention

The experimental intervention was brexpiprazole (OPC‐34712), either as monotherapy or add‐on therapy for the treatment of major depression. We included studies on brexpiprazole in the licensed dose range for the treatment of depression (0.5 mg to 3 mg); we also decided to include studies with higher doses, as higher doses of the medication (4 mg) were approved for another indication (i.e. schizophrenia) (FDA 2015a), and plasma concentrations may vary depending on differences in drug metabolism and comedications (Otsuka 2020). We included treatment arms with both fixed and flexible dosage regimens.

Comparator interventions

We included the following comparator interventions.

  • Antidepressants (as monotherapy or add‐on therapy), including, but not limited to:

    • SSRIs – sertraline, paroxetine, citalopram, escitalopram, fluoxetine, fluvoxamine

    • SNRIs – venlafaxine, duloxetine, desvenlafaxine, milnacipran, levomilnacipran

    • TCAs – amitriptyline, amoxapine, clomipramine, desipramine, dosulepin or dothiepin, doxepin, imipramine, lofepramine, maprotiline, nortriptyline, protriptyline, trimipramine

    • MAOIs – phenelzine, isocarboxazide, tranylcypromine, moclobemide, brofaromine

    • other antidepressants – mirtazapine, agomelatine, trazodone, vortioxetine, bupropion, reboxetine, esketamine, vilazodone

  • Placebo (as monotherapy or add‐on therapy)

Types of outcome measures

Primary outcomes
  • Number of participants who responded to treatment after eight weeks; measured as at least 50% reduction of the total score on a standardised depression scale, in the following order: Hamilton Depression Rating Scale (HAM‐D; Hamilton 1960), MADRS (Montgomery 1979), Clinical Global Impression – Improvement (CGI‐I; Guy 1970), or any other validated depression scale (Furukawa 2016). The response rate was calculated based on the total number of randomised participants, also known as full or strict intention‐to‐treat (ITT) analysis which aims to include all participants randomised into a trial irrespective of what happened subsequently (Higgins 2011b).

  • Number of dropouts due to any cause after eight weeks.

  • Number of dropouts due to any adverse effect after eight weeks.

Secondary outcomes
  • Number of participants who responded to treatment after two and 18 weeks.

  • Number of participants who achieved remission after eight weeks, measured by these depression scales (in the following order, depending on availability):

    • 7 or fewer points on the HAM‐D17 (or 8 or fewer on any other HAM‐D scale);

    • 12 or fewer points on the MADRS;

    • score of 'not ill' or 'borderline mentally ill' (score 1 or 2) on the CGI – Severity;

    • change in depressive symptoms between baseline and eight weeks.

Safety

  • Total number of participants with any adverse events.

  • Total number of participants experiencing any of the following adverse effects: akathisia, anxiety, decreased blood cortisol, cardiovascular events (such as stroke or myocardial infarction), constipation, dizziness, fatigue, headache, increased appetite, weight gain, nasopharyngitis (inflammation of the nose and throat), leukopenia (reduction of the number of white blood cells below reference range), neutropenia (reduction of the number of neutrophils below reference range), agranulocytosis (a more severe reduction of granulocytes below reference range), metabolic changes (such as hyperglycaemia, diabetes mellitus, dyslipidaemia), mortality, suicidal thoughts and behaviour, neuroleptic malignant syndrome (a rare, life‐threatening reaction that can occur in response to neuroleptic or antipsychotic medication), pathological gambling and other compulsive behaviours, restlessness, somnolence, tardive dyskinesia (stiff, jerky, uncontrollable movements of one's face and body), tremor. In order to avoid missing any uncommon but important adverse effects, we reported any other adverse effects extracted from the included studies.

Adverse effects were considered from the beginning of the intervention until the end of the longest follow‐up period of the included trials.

  • Social functioning or adjustment, evaluated by Sheehan Disability Scale (SDS; Sheehan 1983).

  • Health‐related quality of life, evaluated by Quality of Life Enjoyment and Satisfaction Questionnaire – Short Form (Q‐LES‐SF) (Stevanovic 2011).

Timing of outcome assessment

At two weeks: we gave preference to two‐week data, but we also considered the outcome between weeks one and four (one or more and less than four weeks) if two‐week data were not available.

At eight weeks: we gave preference to eight‐week data, but we considered the outcome between weeks four and 12 (four or more and 12 weeks and less) if eight‐week data were not available.

At 18 weeks: we gave preference to 18‐week data, but we considered the outcome between weeks 12 and 24 (12 or more and 24 weeks or less) if 18‐week data were not available.

Search methods for identification of studies

Electronic searches

We ran searches on the following databases using relevant keywords, subject headings (controlled vocabularies) and search syntax, appropriate to each resource (Appendix 1). Searches were initially conducted on 17 September 2020 and updated on 2 November 2021.

  • Cochrane Common Mental Disorders Controlled Trials Register (CCMDCTR) (all available years)

  • Cochrane Central Register of Controlled Trials (CENTRAL; Issue 11, 2021) in the Cochrane Library

  • MEDLINE Ovid (1946 to 2 November 2021)

  • PubMed not MEDLINE (all years to 2 November 2021)

  • Embase Ovid (1974 to 2 November 2021)

  • PsycINFO Ovid (all years to November week 1, 2021)

We did not impose any restrictions on date, language or publication status.

We searched international trial registers (2 November 2021), including the US National Institutes of Health Trials Register ClinicalTrials.gov (clinicaltrials.gov) and the World Health Organization International Clinical Trials Registry Platform (ICTRP; apps.who.int/trialsearch/) to identify additional ongoing and unpublished studies.

Searching other resources

Reference lists

We checked the reference lists of all included studies and relevant systematic reviews to identify additional studies missed from the original electronic searches.

Correspondence

We contacted study authors and subject experts for information on unpublished or ongoing studies, or to request additional data.

Data collection and analysis

Selection of studies

Two review authors (SR, FDC or IK) independently screened titles and abstracts to include all potential studies identified by the search; we coded them as 'retrieve' (eligible or potentially eligible or unclear), or 'do not retrieve'. If there were any disagreements, a fourth review author (AC) arbitrated. We retrieved the full‐text study reports or publications, and two review authors (SR, FDC or IK) independently screened the full‐text reports and identified studies for inclusion; at this stage we identified and recorded reasons for exclusion of the ineligible studies. We resolved any disagreements that occurred through discussion and a fourth review author (AC) arbitrated. After retrieving all reports for inclusion, we identified and excluded duplicates, and collated multiple reports of the same study, so that each study, rather than each report, was the unit of interest in this review. We recorded the selection process in sufficient detail to complete a PRISMA flow diagram, and table describing the characteristics of excluded studies (Liberati 2009).

Data extraction and management

We created a data collection form adapted from Cochrane's data collection form for intervention reviews for RCTs only, and tested it on one eligible study; after discussion, all authors approved this. Two review authors (SR, FDC or IK) independently extracted study characteristics from included studies. We resolved disagreements by consensus, or by involving a fourth review author (AC). We collected the following data.

  • Methods: study design, total duration of study, number of study centres and location, study setting and date of study.

  • Participants: number of participants randomised, number of participants lost to follow‐up or withdrawn, number of participants analysed, mean age, age range, gender, severity of condition, diagnostic criteria, inclusion criteria and exclusion criteria.

  • Interventions: types of interventions, types of comparisons, concomitant medications and excluded medications.

  • Outcomes: outcomes specified and collected, and time points reported.

  • Notes: funding for trial, and notable conflicts of interest of trial authors.

One review author (SR) transferred data into Review Manager 5 (Review Manager 2020). We double‐checked that data were entered correctly by comparing the data presented in the systematic review with the data extraction form.

Assessment of risk of bias in included studies

Two review authors (SR, FDC or IK) independently assessed the risk of bias for each study to minimise errors in assessment and to ensure the judgement was not influenced by one review author's preconceptions (Boutron 2021), using the Cochrane RoB 1 tool, outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a). When additional information was needed, we contacted study authors to clarify incompletely reported information (Boutron 2021). The potential risk of bias for every study (or specific study outcomes when separate judgements for them were considered relevant) was categorised as 'low', 'high' or 'unclear' for each of the seven domains in the RoB 1 tool for every included study in the risk of bias table. The potential risk of bias, specifically blinding was also evaluated for 'efficacy' as a patient‐important outcome. Assessors provided clarification for every judgement provided as a quote from study report and a comment by the assessor. We resolved any disagreements through discussion or by involving another review author (AC). We assessed the risk of bias of the specific results of a trial according to the following domains.

  • Selection bias: random sequence generation; allocation concealment

  • Performance bias: blinding of participants and personnel

  • Detection bias: blinding of outcome assessment

  • Attrition bias: incomplete outcome data (attrition rate 30% or greater was considered high risk of bias)

  • Reporting bias: selective reporting

  • Other bias: other sources of bias, such as (see Section 8.15 of the Cochrane Handbook for Systematic Reviews of Interventions, Higgins 2011a): design‐specific risk of bias (for cluster‐RCTs or cross‐over RCTs), baseline imbalances, bias due to differential diagnostic activity, bias due to conduct of the study affected by interim results, deviation from the study protocol in a way that contradicted clinical practice, pre‐administration of an intervention that could moderate (enhance or diminish) the effect of the subsequent randomised intervention of interest; inappropriate administration of an intervention (co‐intervention), contamination, or overly wide inclusion criteria for participants; use of insensitive instruments to measure treatment outcomes, selective reporting of subgroups, fraud.

We also considered funding as a source of bias, as funding by people with vested interests in the results may directly imply the introduction of risk of bias (Higgins 2011a). However, as the risk of bias tool looks mostly into methodological aspects of trials, we included funding in the Characteristics of included studies table.

When considering treatment effects, we took into account the risk of bias for the studies that contributed to that outcome.

Measures of treatment effect

Dichotomous data

We calculated and reported odds ratio (OR) with 95% confidence interval (CI) on a logarithmic scale. The total number of events was calculated out of the total number of participants randomised (full ITT analysis).

Continuous data

For continuous data as effect measures, we included MD, least squares mean (LSM) or standardised mean difference (SMD) with 95% CI. When studies used the same measurement scale for an outcome, we used the MD or LSM, while we used the SMD when studies used different scales.

When results were reported with LSM (and standard errors (SE)), we decided to calculate the standard deviation (SD) as the multiplication of the SE and the square root of the number of participants in a group, as recommended in Martinez 2017. If results were presented as LSM or MD in different studies, we decided to pool them.

Unit of analysis issues

Cross‐over trials

If any cross‐over trials had been detected, we would have included only the first period of the trial (before cross‐over) and we would have evaluated the risk of bias using the standard RoB 1 tool, as the first part of the study is similar to a parallel‐design study. Although at least half of the data would have been lost with this decision, study population (people with acute depression) and the long wash‐out period of brexpiprazole (t½ = 93 hours, washout is approximately five times t½) may not have allowed for this study design or would have introduced a significant potential bias due to carry‐over effects (Higgins 2021).

Studies with multiple study groups

If a study included multiple treatment groups, we included only study groups that met the inclusion criteria in the review; however, we described study arms that did not meet these criteria in the Characteristics of included studies table; reasons for exclusion were also provided. For the assessment of the risk of bias in the specific case of multiple‐intervention studies we reported whether data were presented for each of the groups to which participants were randomised and whether the reports of the study might have suggested selective reporting of comparisons of interventions (Higgins 2021). For dichotomous outcomes, we decided to combine groups by summation of size samples and event counts for relevant intervention groups and compare them with the comparator group estimated following a standard approach for pair‐wise meta‐analysis. For continuous outcomes, we decided to use the same approach and combine results, using the formulae for MD and SMD, provided in Section 6.5.2.10 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2021).

Dealing with missing data

Missingness of data may be found on several levels of the review: whole studies may be missing from the review, an outcome may be missing from a study, summary data may be missing from the study and individuals may be missing from the summary data (Higgins 2011b).

Missing studies

Studies may be missing from the review if they were unpublished, the data were obscured and not accessible, if the search was not sufficiently comprehensive, etc. One possible explanation is publication bias. If 10 or more studies had been included in this review, we would have assessed publication bias with a funnel plot.

Missing outcome data

Studies may not have reported the specific outcomes (e.g. a serious adverse event); outcomes might not have been measured or selective reporting might have occurred.

Missing statistic measures

In case of missing statistic measures, we attempted to recalculate the missing data from the statistical data already provided in study reports. If this was not possible, we contacted study authors for additional information. Finally, if data were still not obtained, in the case of SD, we would have imputed external data from studies with similar designs, populations and interventions, and tested the appropriateness of this decision through a sensitivity analysis (Furukawa 2016).

Missing individual data
Dichotomous outcomes

For dichotomous outcomes in case of missingness of data, we decided to estimate effect size based on a full ITT analysis; in this case, we included all randomised participants as allocated at baseline as the sample size. Dropouts were regarded as non‐event cases.

Continuous outcomes

For continuous outcomes, we estimated the summary measures based on available‐case analysis, which meant we based our analysis only on data on those participants whose results were known. We considered the potential impact of the missing data on the results in the interpretation of the results of the review, depending on the degree of missingness, the pooled estimate of the treatment effect and the variability of the outcomes (Higgins 2011b).

Assessment of heterogeneity

We inspected forest plots visually to consider the direction and magnitude of effects and the degree of overlap between CIs. Lack of overlap between CIs suggests the presence of statistical heterogeneity. For a more formal evaluation of heterogeneity, we also considered the P value from the Chi² test. We used the I² statistic to quantify inconsistency amongst the trials in each analysis. We reported substantial heterogeneity and explored possible causes by prespecified subgroup analysis. We used the following thresholds for the interpretation of the statistics as suggested by the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2021):

  • 0% to 40%: might not be significant;

  • 30% to 60%: may represent moderate heterogeneity;

  • 50% to 90%: may represent substantial heterogeneity;

  • 75% to 100%: considerable heterogeneity.

We interpreted the size and importance for moderate, substantial and considerable heterogeneity of the I² statistic in the context of sample size, the magnitude and the direction of the treatment effects, and the strength of evidence for heterogeneity (e.g. P value from the Chi² statistic and the CIs for the I² statistic) (Deeks 2021).

Assessment of reporting biases

In most cases, selective non‐reporting or under‐reporting occurs when the results are unfavourable or non‐significant to the experimental intervention (Page 2021).

If a study protocol was available, we compared the reported outcome results with the outcomes planned in the protocol. If we had included more than 10 studies, we would have used a funnel plot which takes into account the effect size and number of participants in every study. Formal test for asymmetry of the plot revealing small‐study effect indicates the missingness of outcome measures from small studies with negative results which are more likely to be unpublished.

Data synthesis

We analysed the data using Review Manager 5 (Review Manager 2020). We undertook analysis when results were meaningful (e.g. the treatments, participants and the underlying clinical question were similar enough so that it was justified to pool them). We used random‐effects model for both dichotomous and continuous outcomes. The random‐effects model assumes that the effect follows a normal distribution and that the observed differences between studies are due to both chance and genuine variation between studies. For the synthesis of data for dichotomous outcomes, we used OR summary statistic with 95% CIs. For continuous outcomes, we used MD, LSM or SMD depending on the measurement scales with a corresponding 95% CI. We considered results with P < 0.05 and a 95% CI that did not cross the line of no effect as statistically significant.

Subgroup analysis and investigation of heterogeneity

We planned to carry out the following subgroup analyses using age and dosage as variables of interest because children and adolescents can respond differently to antidepressants than people aged 18 years or older (Cipriani 2016), and flexible dosage regimens versus fixed dosage regimens of antidepressants might differ in terms of efficacy (Furukawa 2020).

Age groups:

  • participants aged below 18 years;

  • participants aged 18 years or older.

Dosage groups:

  • fixed dosage – participants are assigned to a specific dose of a treatment; dosage adjustment is not allowed;

  • flexible dosage – dosage adjustments are allowed.

Sensitivity analysis

We planned to carry out the following sensitivity analyses:

  • excluding trials in which allocation concealment was not clearly described, or double‐blinding was unclear (Furukawa 2016);

  • excluding trials with dropout rates of more than 20% in one of the included treatment arms (Furukawa 2016);

  • excluding trials with imputed data (Furukawa 2005), and those for which the SD was missing, and we imputed it from other trials (Furukawa 2016);

  • excluding trials with one or more domains with high risk of bias.

Summary of findings and assessment of the certainty of the evidence

We created a summary of findings table using the following outcomes, including the two most frequent adverse events.

  • Number of participants who responded to treatment at eight weeks

  • Number of dropouts due to any cause at eight weeks

  • Number of dropouts due to any adverse effect at eight weeks

  • Number of participants who achieved remission at eight weeks

  • Change in depressive symptoms between baseline and eight weeks

  • Adverse effect: akathisia

  • Adverse effect: weight gain

We used the methods and recommendations described in Chapter 14 of the Cochrane Handbook for Systematic Reviews of Interventions (Schünemann 2021), and GRADEpro GDT software (GRADEpro GDT). We justified all decisions to downgrade the certainty of the evidence using footnotes and we made comments to aid reader's understanding of the review where necessary.

Factors that can downgrade the certainty of evidence include:

  • limitations in study design or execution (risk of bias): downgraded by one or two levels;

  • inconsistency of results: downgraded by one or two levels;

  • indirectness of evidence: downgraded by one or two levels;

  • imprecision: downgraded by one or two levels;

  • publication bias: downgraded by one or two levels.

Two review authors (SR, IK) independently assessed the certainty of the evidence, with disagreements resolved by discussion or involving a third review author (AC). Their judgements were justified, documented and incorporated into reporting of results for each outcome.

Results

Description of studies

Results of the search

The search yielded 814 records overall (691 from the initial search and 123 from the updated search in November 2021). After deduplication, we screened 473 reports for inclusion in the review. We excluded 442 records after screening of titles and abstracts for the following reasons: ineligible design, ineligible population, preclinical studies, ineligible interventions, duplicate publications and reports not available. As a result, 31 reports remained for full‐text review; we included nine studies (25 reports) in the qualitative and quantitative analyses, excluded one study (a duplicate, one report), two studies are awaiting classification (two reports) and three studies are ongoing (three reports) (Figure 1).

1.

1

Study flow diagram.

Included studies

The nine included studies are described in detail in the Characteristics of included studies table (Bauer 2019; Hobart 2018a; Hobart 2018b; NCT00797966; NCT01052077; NCT01837797; NCT03149991; Thase 2015a; Thase 2015b).

Studies
Design

All were double‐blind RCTs and had a similar design. The screening period was followed by a prospective phase where participants were treated with an antidepressant (with or without add‐on placebo). In the following double‐blind randomised phase, participants who did not respond to antidepressant treatment (response was measured as at least 50% reduction on a standardised rating scale) were allocated to either brexpiprazole or placebo as adjunctive treatment. Participants who responded to antidepressant treatment in the prospective phase were also followed up by the end of the study period.

One was a three‐armed trial comparing brexpiprazole versus quetiapine extended release (XR) versus placebo (Hobart 2018a); the eight remaining studies were two‐armed. NCT03149991 added either brexpiprazole or placebo to a combination of antidepressant and ketamine.

Two were Phase 2 studies (NCT00797966; NCT01052077), six were Phase 3 (Bauer 2019; Hobart 2018a; Hobart 2018b; NCT01837797; Thase 2015a; Thase 2015b), and one was a phase 4 study (NCT03149991).

Length

The duration of studies and follow‐up for monitoring adverse effects (after the end of the study) were: four weeks' duration and no follow‐up period for one study (NCT03149991); 14 weeks plus 30 days' follow‐up for six studies (Hobart 2018a; Hobart 2018b; NCT00797966; NCT01052077; Thase 2015a; Thase 2015b); 20 weeks plus four weeks' follow‐up for one study (NCT01837797); and 24 weeks plus four weeks' follow‐up for one study (Bauer 2019).

Sample size

We included data from 3424 participants in total (mean sample size: 378 participants). NCT01837797 had the smallest sample size (15 participants), as it was terminated prematurely for administrative reasons. By contrast, Bauer 2019 was the largest and longest study (686 participants).

Setting

Six studies recruited outpatients (Bauer 2019; Hobart 2018a; Hobart 2018b; NCT01837797; Thase 2015a; Thase 2015b). The remaining three studies did not report such information (NCT00797966; NCT01052077; NCT03149991).

Bauer 2019 included 112 sites across 16 countries in four continents (Republic of Korea, Bulgaria, Estonia, Finland, Germany, Latvia, Lithuania, Poland, Romania, Russia, Sweden, Ukraine, the UK, Mexico, Canada). Hobart 2018a was conducted in 75 sites in the US (55.0% of participants), Russia (17.0%), Poland (13.0%), France (6.0%), Serbia (4.5%), Germany (4.4%) and Canada. Hobart 2018b included 51 sites (60.5% of participants in the USA, 11.4% in Germany, 7.6% in Hungary, 11.4% in Poland, 9.2% in Slovakia). Thase 2015a included sites in the USA (74.9% of participants), Poland (9.7%), France (8.5%), Canada (4.8%) and Slovakia (2.1%). Thase 2015b included 71 sites in the US, Germany, Ukraine, Russia, Hungary, Canada and Romania. The remaining four studies were published only on ClinicalTrials.gov and were conducted only in the USA (NCT00797966 – 50 sites; NCT01052077 – 43 sites; NCT01837797 – number of sites not reported; NCT03149991 – 5 sites).

Participants
Age

Seven studies included participants aged between 18 and 65 years (Hobart 2018a; Hobart 2018b; NCT00797966; NCT01052077; NCT03149991; Thase 2015a; Thase 2015b), one study included participants aged between 18 and 75 years (Bauer 2019), and one study included participants older than 65 years (NCT01837797) (Table 2).

1. Age of participants.
Study Mean (years) SD (years)
Bauer 2019 47.1 12.1
Hobart 2018a 43.6 11.5
Hobart 2018b 43.0 12.7
NCT00797966 43.7 11.6
NCT01052077 43.5 11.8
NCT01837797 Not reporteda
NCT03149991 40.8 12.4
Thase 2015a 44.1 11.6
Thase 2015b 45.7 11.6

aAll participants aged 65 years or older.
SD: standard deviation.

Sex

All studies included both male and female participants with mean percentage of females of 67.5% (range 51% to 74%) (Table 3).

2. Sex: female.
Study Number %
Bauer 2019 608 69
Hobart 2018a 277 69
Hobart 2018b 291 74
NCT00797966 162 66
NCT01052077 253 68
NCT01837797 11 73
NCT03149991 26 51
Thase 2015a 267 70
Thase 2015b 460 68
Diagnosis

Participants in all nine studies had to have a diagnosis of MDD and a current major depressive episode as per DSM‐IV‐TR criteria. The diagnosis had to be confirmed by Mini‐International Neuropsychiatric Interview (MINI) in four studies (Bauer 2019; Hobart 2018a; Hobart 2018b; NCT01837797). Two studies required a MADRS Total score of 26 or greater (Bauer 2019; Hobart 2018a). Bauer 2019 had an additional criterion of Clinical Global Impression – Severity (CGI‐S) score of 4 or greater. Two studies required participants with a HAM‐D17 Total score of 18 or greater at screening and baseline visits (Hobart 2018b; Thase 2015a). Three studies had additional criteria about the characteristics of the depressive episode (e.g. duration of more than eight weeks) (Hobart 2018a; NCT00797966; NCT01052077). Eight studies had history of an inadequate response to antidepressants (at least one but no more than three) as an inclusion criterion (Bauer 2019; Hobart 2018a; Hobart 2018b; NCT00797966; NCT01052077; NCT01837797; Thase 2015a; Thase 2015b). Four studies defined inadequate response as less than 50% reduction in depressive symptom severity, assessed using the Antidepressant Treatment Response Questionnaire (ATRQ) (Hobart 2018a; Hobart 2018b; Thase 2015a; Thase 2015b). One study included participants with treatment‐resistant depression defined as people with a history of inadequate response to at least two antidepressants with duration of at least eight weeks for the current episode; this study also required moderate or severe depression as an inclusion criterion (NCT03149991).

Exclusion criteria

Six studies excluded participants with an inadequate response to more than three antidepressant treatments (Hobart 2018a; Hobart 2018b; NCT00797966; NCT01052077; Thase 2015a; Thase 2015b). In NCT03149991, the threshold for exclusion was failure to achieve response to more than seven courses of a therapeutic dose of antidepressant for at least eight weeks of duration. All studies excluded participants with Axis I diagnosis other than MDD or Axis II diagnosis of borderline, antisocial, paranoid, schizoid, schizotypal or histrionic personality disorder, or a combination of these. All studies excluded people with psychotic symptomatology in the current major depressive episode. People were excluded if they had received ECT for the current episode (Hobart 2018a; Hobart 2018b; Thase 2015b), or if they had history of inadequate response to ECT (Hobart 2018a; Hobart 2018b; Thase 2015b). Two studies excluded females who were breastfeeding or who had a positive pregnancy test (NCT00797966; NCT01052077). Three studies excluded people with significant suicide risk based on the investigator's judgement or according to Columbia Suicide Severity Rating Scale (C‐SSRS) (NCT00797966; NCT01837797; NCT03149991). One study had additional exclusion criteria such as MADRS Total score of less than 20 at the screen visit or the baseline visit (NCT03149991).

Inclusion in double‐blind randomisation phase

Criteria for inclusion in the double‐blind phase included non‐response to antidepressant treatment, where non‐response was measured as less than 50% reduction in MADRS Total score between the start of prospective treatment and all visits in the prospective period (Bauer 2019; Hobart 2018a; Hobart 2018b). Criteria also included the severity of the major depressive episode (Bauer 2019; Hobart 2018a; Hobart 2018b); CGI‐I score of 3 or greater (minimally improved) at every visit (Bauer 2019; Hobart 2018a; Hobart 2018b; Thase 2015b). Additionally, some studies required less than 50% reduction in HAM‐D17 Total score from the start to the end of prospective treatment or HAM‐D17 score of 14 or greater at the end of the prospective treatment (Hobart 2018b; Thase 2015a). Two studies had their protocols amended with supplementary criteria for inclusion in the double‐blind phase with an additional criterion to ensure non‐response (Thase 2015a; Thase 2015b).

Interventions

In eight studies, the intervention was brexpiprazole and the comparator was placebo, as adjunctive treatment to oral antidepressant(s); NCT03149991 gave brexpiprazole and placebo adjunctive to a combination of antidepressants (oral) and ketamine (intranasal). One study included a third arm with an active comparator – quetiapine XR (Hobart 2018a); however, the quetiapine arm was not included in our analysis because we excluded atypical antipsychotics as per the review protocol (Ralovska 2021).

Only four studies provided specific information about the antidepressants used and involved the following drugs: duloxetine, escitalopram, fluoxetine, paroxetine, sertraline or venlafaxine XR (Hobart 2018a; Hobart 2018b; NCT01837797; Thase 2015a).

Dosage of study drugs

Brexpiprazole was administered either as fixed or flexible dose. The studies with fixed‐dose brexpiprazole were: Hobart 2018b (2 mg), NCT01837797 (1 mg and 3 mg), Thase 2015a (2 mg), Thase 2015b (1 mg). Studies with flexible dose were: Bauer 2019 (1 mg to 3 mg), Hobart 2018a (2 mg to 3 mg); NCT01052077 (1 mg to 3 mg); NCT03149991 (flexible dose up to 3 mg). NCT00797966 included four arms (three intervention arms and one control). We excluded the brexpiprazole 0.15 mg arm as it did not meet inclusion criteria for dosage as defined by the protocol for the review (Ralovska 2021). We excluded brexpiprazole 0.5 ± 0.25 mg as there was a lack of information whether the mean dose of brexpiprazole was greater than 0.5 mg as defined per the review protocol; we contacted study authors for additional information on the mean dose, but received no response. We included the third arm, brexpiprazole 1.5 mg ± 0.5 mg in the review.

Dosages of antidepressants varied across the studies, but were always within their therapeutic range: escitalopram (up to 20 mg/day), fluoxetine (up to 40 mg/day), paroxetine controlled release (CR) (up to 50 mg/day), sertraline (up to 200 mg/day), duloxetine (up to 60 mg/day) and venlafaxine XR (up to 225 mg/day).

In NCT03149991, the pattern of administration of intranasal ketamine was biweekly (40 mg) for two weeks, followed by weekly administration of intranasal ketamine (40 mg) for two weeks.

Primary outcomes

Seven studies had change in MADRS Total score at the end of the randomisation phase (after six weeks of treatment) as the primary outcome (Hobart 2018a; Hobart 2018b; NCT00797966; NCT01052077; NCT01837797; Thase 2015a; Thase 2015b). Bauer 2019 used number of participants who achieved full remission defined as a MADRS Total score of 10 or less and a decrease of 50% or greater from randomisation in MADRS Total score for at least eight consecutive weeks during randomised treatment. NCT03149991 had the primary outcome of change from baseline on Symptoms of Depression Questionnaire (SDQ). During the study they were assessed on days 0, 1, 2, 5, 8, 11, 14, 17, 21, 23 and 28 of randomisation.

Secondary outcomes

Eight studies reported response rate measured by 50% reduction on MADRS (Bauer 2019; Hobart 2018a; Hobart 2018b; NCT00797966; NCT01052077; NCT03149991; Thase 2015a; Thase 2015b); four studies reported this outcome at more than one time point (Bauer 2019; Hobart 2018a; NCT01052077; Thase 2015b). Eight studies reported all‐cause dropouts (as a measure of acceptability) and dropouts due to any reason (as a measure of tolerability) (Bauer 2019; Hobart 2018a; Hobart 2018b; NCT00797966; NCT01052077; NCT03149991; Thase 2015a; Thase 2015b). Five studies reported change in mean SDS score from beginning to the end of double‐blind phase as a key secondary outcome (Hobart 2018a; Hobart 2018b; NCT00797966; NCT01052077; Thase 2015b). Bauer 2019 reported full functional remission measured using the SDS Total score of 6 or less and all SDS domain scores of 2 or less was a key secondary outcome. NCT01837797 reported safety (number of participants who experienced any specific adverse effect) as the only outcome; all other included studies reported safety as a secondary outcome (Bauer 2019; Hobart 2018a; Hobart 2018b; NCT00797966; NCT01052077; NCT03149991; Thase 2015a; Thase 2015b).

Sponsorship

Otsuka and Lundbeck, the pharmaceutical companies that developed and marketed brexpiprazole, sponsored eight studies (Bauer 2019; Hobart 2018a; Hobart 2018b; NCT00797966; NCT01052077; NCT01837797; Thase 2015a; Thase 2015b). The Massachusetts General Hospital sponsored one study (NCT03149991).

Excluded studies

We excluded one report (NCT01052077b), which was a duplicate of an included study (NCT01052077).

Studies awaiting classification

Two studies are awaiting classification.

NCT01670279 (Phase 1 trial) had two cohorts of participants receiving brexpiprazole 3 mg plus antidepressant with different dosing regimens in terms of titration and a placebo plus antidepressant cohort. The study could not be included in the current review as we could not retrieve information around the number of antidepressant trials before study onset and could not clarify the group of participants with non‐response to one antidepressant or the one with treatment‐resistant depression.NCT03487198 was reported as completed; however, we did not find any published results. We contacted the authors/sponsors for further information on these studies, but we received no response.

Ongoing studies

Three studies are ongoing.

jRCT2080223986 is a Phase 2/3 RCT assessing the efficacy and safety of brexpiprazole as adjunctive therapy in people with MDD. It includes three intervention arms: brexpiprazole 1 mg plus antidepressant, brexpiprazole 2 mg plus antidepressant and placebo plus antidepressant. NCT03538691 is a Phase 3 double‐blind RCT comparing flexible‐dose brexpiprazole (2 mg to 3 mg) and placebo as add‐on treatment to antidepressants. NCT03697603 is a Phase 2/3 RCT comparing brexpiprazole 1 mg plus antidepressant, brexpiprazole 2 mg plus antidepressant and placebo plus antidepressant.

Risk of bias in included studies

Overall, the evaluation of risk of bias on all six domains varied between low and unclear risk of bias, except for NCT01837797, which had high risk of bias in two domains (Figure 2; Figure 3).

2.

2

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

3.

3

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Allocation

Random sequence generation: all studies were at low risk of bias for this domain. All published studies reported that assignment to treatment was based on computer‐generated random number tables (Bauer 2019; Hobart 2018a; Hobart 2018b; Thase 2015a; Thase 2015b). NCT03149991 outlined that participants were allocated "individual treatment codes". The remaining three studies reported that allocation was "randomised" (NCT00797966; NCT01052077; NCT01837797).

Allocation concealment: it was not possible to evaluate the risk of bias for five studies because this information was not provided (NCT00797966; NCT01052077; NCT01837797; Thase 2015a; Thase 2015b). The remaining four studies were at low risk of bias (Bauer 2019; Hobart 2018a; Hobart 2018b; NCT03149991).

Blinding

All nine studies were at low risk of bias for blinding of participants and personnel. However, six studies were at unclear risk in terms of blinding of outcome assessors due to lack of sufficient information (Bauer 2019; NCT01052077; NCT01837797; NCT03149991; Thase 2015a; Thase 2015b).

Incomplete outcome data

The risk of attrition bias was low in six studies (Hobart 2018a; Hobart 2018b; NCT00797966; NCT01052077; Thase 2015a; Thase 2015b). Attrition bias was high either due to a dropout rate greater than 30% for all participants or dropout rate in one of the treatment arms was at least two times higher than another treatment arm. When attrition between study arms was unbalanced, we judged that attrition was likely and evaluated risk of bias as unclear (Bauer 2019; NCT03149991). NCT01837797 did not report attrition (unclear risk).

Selective reporting

Eight studies had low risk of reporting bias as outcomes were reported as outlined in an available protocol (Bauer 2019; Hobart 2018a; Hobart 2018b; NCT00797966; NCT01052077; NCT03149991; Thase 2015a; Thase 2015b). NCT01837797 was at high risk of bias; it published only safety results although other outcomes, such as efficacy, were also listed in the method section.

Other potential sources of bias

NCT01837797 was at high risk of other bias as the study was prematurely terminated, and no firm conclusions could be drawn regarding safety and efficacy. The risk of other bias of the remaining studies was unclear due to insufficient data to inform whether important risk existed or not (Higgins 2011a).

Effects of interventions

See: Table 1

Primary outcomes

Number of participants who responded to treatment after eight weeks

There was high‐certainty evidence that participants were more likely to achieve response to treatment with adjunctive brexpiprazole than with adjunctive placebo after eight weeks (OR 1.47, 95% CI 1.23 to 1.75; P < 0.0001, I²= 0%; 8 studies, 3409 participants; Analysis 1.1).

1.1. Analysis.

1.1

Comparison 1: Brexpiprazole versus placebo, Outcome 1: Response to treatment at 8 weeks

Number of dropouts due to any cause at eight weeks

High‐certainty evidence demonstrated that participants receiving adjunctive placebo were less likely to drop out for any reason than participants receiving adjunctive treatment with brexpiprazole. However, the magnitude and clinical significance of this effect is uncertain as the lower range of the 95% CI is close to the line of no difference (OR 1.37, 95% CI 1.02 to 1.83; P = 0.04; I²= 0%; 7 studies, 2523 participants; Analysis 1.2).

1.2. Analysis.

1.2

Comparison 1: Brexpiprazole versus placebo, Outcome 2: Acceptability (dropouts due to any cause) at 8 weeks

Number of dropouts due to any adverse effect after eight weeks

There was high‐certainty evidence that participants receiving adjunctive treatment with brexpiprazole were more likely to drop out of studies due to adverse effects than participants receiving placebo. However, the magnitude of this effect ranged widely (OR 2.88, 95% CI 1.37 to 6.05; P = 0.005, I²= 0%; 6 studies, 2472 participants; Analysis 1.3).

1.3. Analysis.

1.3

Comparison 1: Brexpiprazole versus placebo, Outcome 3: Tolerability (dropouts due to adverse effects) at 8 weeks

Secondary outcomes

Number of participants who responded to treatment after two and 18 weeks

There was evidence that more participants allocated to brexpiprazole were likely to reach a response threshold in two weeks when compared to participants allocated to placebo (OR 1.65, 95% CI 1.04 to 2.63; P = 0.003, I²= 0%; 3 studies, 1452 participants). One study reported long‐term response (Bauer 2019), and there was no difference between groups (OR 0.88, 95% CI 0.68 to 1.15; P = 0.35, 886 participants; Analysis 1.4).

1.4. Analysis.

1.4

Comparison 1: Brexpiprazole versus placebo, Outcome 4: Response at different time points (secondary outcome)

Number of participants who achieved remission after eight weeks

There was high‐certainty evidence that brexpiprazole as adjunctive treatment was more likely to lead to remission of MDD when compared to placebo (OR 1.46, 95% CI 1.19 to 1.79; P < 0.001, I²= 0%; 6 studies, 3358 participants; Analysis 1.5).

1.5. Analysis.

1.5

Comparison 1: Brexpiprazole versus placebo, Outcome 5: Remission

Change in depressive symptoms between baseline and eight weeks

There was an improvement in depressive symptoms using the MADRS after eight weeks with adjunctive brexpiprazole compared with adjunctive placebo (MD –1.39, 95% CI –1.96 to –0.82; P < 0.001, I²= 27%; 8 studies, 3263 participants; Analysis 1.6).

1.6. Analysis.

1.6

Comparison 1: Brexpiprazole versus placebo, Outcome 6: Change in depressive symptoms at 8 weeks

Total number of participants with any adverse events

Brexpiprazole caused a greater total number of adverse effects than placebo (OR 1.63, 95% CI 1.29 to 2.06; P < 0.0001, I²= 46%; 9 studies, 3422 participants; Analysis 1.7).

1.7. Analysis.

1.7

Comparison 1: Brexpiprazole versus placebo, Outcome 7: Safety – total number of adverse effects

Total number of participants experiencing specific adverse effects

We reported the data for the adverse events where there was a difference between groups below.

Adverse effect: akathisia

There was moderate‐certainty evidence that add‐on brexpiprazole was about three times more likely to be related to akathisia compared to placebo (OR 2.95, 95% CI 2.06 to 4.21; P < 0.001, I²= 32%; 7 studies, 3358 participants; Analysis 1.9). However, it was not clear how akathisia was objectively measured.

1.9. Analysis.

1.9

Comparison 1: Brexpiprazole versus placebo, Outcome 9: Safety – akathisia

Adverse effect: constipation

Brexpiprazole was related to higher rates of constipation compared with placebo (OR 3.18, 95% CI 1.08 to 9.41; P < 0.04, I²= 0%; 2 studies, 423 participants; Analysis 1.12).

1.12. Analysis.

1.12

Comparison 1: Brexpiprazole versus placebo, Outcome 12: Safety – constipation

Adverse effect: increased appetite

Brexpiprazole was related to increased appetite compared with placebo, although only three of the nine studies reported this adverse effect (OR 3.98, 95% CI 1.54 to 10.32; P < 0.004, I²= 0%; 3 studies, 826 participants; the magnitude of the effect was uncertain due to the very wide CIs).

Adverse effect: insomnia

Brexpiprazole was more likely to increase insomnia compared with placebo (OR 1.69, 95% CI 1.09 to 2.62; P = 0.02, I²= 40%; 8 studies, 3372 participants; Analysis 1.18).

1.18. Analysis.

1.18

Comparison 1: Brexpiprazole versus placebo, Outcome 18: Safety – insomnia

Adverse effect: restlessness

Brexpiprazole caused more restlessness than placebo (OR 2.26, 95% CI 1.35 to 3.80; P < 0.002, I²= 83%; 5 studies, 2739 participants; Analysis 1.24).

1.24. Analysis.

1.24

Comparison 1: Brexpiprazole versus placebo, Outcome 24: Safety – restlessness

Adverse effect: somnolence

Brexpiprazole was more likely to cause somnolence than placebo (OR 3.56, 95% CI 2.02 to 6.28; P < 0.001, I²= 27%; 5 studies, 2739 participants; Analysis 1.26).

1.26. Analysis.

1.26

Comparison 1: Brexpiprazole versus placebo, Outcome 26: Safety – somnolence

Adverse effect: weight gain

Moderate‐certainty evidence demonstrated that brexpiprazole caused more weight gain than placebo (OR 3.14, 95% CI 2.19 to 4.49; P < 0.001, I²= 1%; 9 studies, 3424 participants; Analysis 1.29).

1.29. Analysis.

1.29

Comparison 1: Brexpiprazole versus placebo, Outcome 29: Safety – weight gain

All adverse effects reported in only one study are listed in Table 4.

3. Adverse events.
Adverse event Study Brexpiprazole Placebo Odds ratio 95% CI
Event Total Event Total
Abdominal pain Thase 2015a 1 188 0 191 2.04 0.12 to 6.27
Agitation Bauer 2019 3 444 0 441 6.00 0.36 to 9.97
Alteration in taste NCT03149991 4 25 2 26 2.29 0.38 to 4.08
Asthenia Hobart 2018a 4 197 0 206 8.54 0.51 to 12.53
Atrial fibrillation Thase 2015a 0 188 1 191 0.51 0.01 to 3.54
Bilateral hand pain NCT03149991 1 25 0 26 2.17 0.13 to 6.50
Blurred vision NCT03149991 0 25 1 26 0.50 0.01 to 3.58
Circulatory collapse Bauer 2019 1 444 0 441 1.99 0.12 to 6.19
Comminuted fracture Thase 2015a 1 188 0 191 2.04 0.12 to 6.27
Chronic obstructive pulmonary disease NCT00797966 1 121 0 126 2.10 0.13 to 6.36
Depression NCT01052077 0 185 1 187 0.50 0.01 to 3.54
Diplopia NCT03149991 0 25 1 26 0.50 0.01 to 3.58
Dissociation/detachment NCT03149991 2 25 3 26 0.67 0.10 to 2.55
Dry mouth NCT03149991 0 25 1 26 0.50 0.01 to 3.58
Dyspnoea NCT01837797 1 9 0 6 1.50 0.08 to 5.62
Elevated alanine transaminase NCT03149991 1 25 0 26 2.17 0.13 to 6.50
Euphoria NCT03149991 1 25 1 26 1.04 0.06 to 3.87
False‐positive investigation result Bauer 2019 0 444 1 441 0.50 0.01 to 3.53
Family stress Bauer 2019 1 444 0 441 1.99 0.12 to 6.19
Forearm fracture Bauer 2019 0 444 1 441 0.50 0.01 to 3.53
Gait disturbance NCT01837797 1 9 0 6 1.50 0.08 to 5.62
Gastric distress NCT03149991 2 25 0 26 4.52 0.26 to 8.74
Hot flushes NCT03149991 1 25 0 26 2.17 0.13 to 6.50
Hypotension NCT01052077 0 185 1 187 0.50 0.01 to 3.54
Increased salivation postdose NCT03149991 1 25 0 26 2.17 0.13 to 6.50
Influenza Bauer 2019 0 444 1 441 0.50 0.01 to 3.53
Insomnia/sleep disturbance NCT03149991 2 25 3 26 0.67 0.10 to 2.55
Intentional overdose Bauer 2019 0 444 2 441 0.25 0.01 to 3.24
Irregular/early menses NCT03149991 1 25 1 25 1.00 0.06 to 3.83
Lethargy NCT03149991 1 25 0 26 2.17 0.13 to 6.50
Light headedness NCT03149991 1 25 0 26 2.17 0.13 to 6.50
Loss of consciousness Bauer 2019 1 444 0 441 1.99 0.19 to 6.19
Lower extremity oedema, bilateral NCT03149991 0 25 1 26 0.50 0.01 to 3.58
Major depression Bauer 2019 1 444 1 441 0.99 0.06 to 3.77
Muscle aches NCT03149991 0 25 1 26 0.50 0.01 to 3.58
Obesity Bauer 2019 1 444 0 441 1.99 0.12 to 6.19
Ovarian cyst Bauer 2019 0 444 1 441 0.50 0.01 to 3.53
Pancreatitis Bauer 2019 1 444 0 441 1.99 0.12 to 6.19
Panic episode NCT03149991 1 25 0 26 2.17 0.13 to 6.50
Parkinson's disease NCT01837797 1 9 0 6 1.50 0.08 to 5.62
Pneumonia Thase 2015a 0 188 1 191 0.51 0.01 to 3.54
Prolactin 3 above limit of normal: male Thase 2015b 1 456 1 221 0.48 0.03 to 3.26
Prolactin 3 times above limit of normal: female Thase 2015b 2 456 0 221 2.43 0.12 to 5.47
Renal mass NCT00797966 0 121 1 126 0.52 0.01 to 3.55
Restless legs NCT03149991 0 25 1 26 0.50 0.01 to 3.58
Ruptured cerebral aneurism Bauer 2019 0 444 1 441 0.50 0.01 to 3.53
Sciatica Bauer 2019 0 444 1 441 0.50 0.01 to 3.53
Seizure Bauer 2019 1 444 0 441 1.99 0.12 to 6.19
Serious injurious behaviour Bauer 2019 0 444 1 441 0.50 0.01 to 3.53
Shakiness NCT03149991 1 25 0 26 2.17 0.13 to 6.50
Sinus bradycardia NCT01052077 0 185 3 187 0.17 0.01 to 3.11
Suicide attempt Bauer 2019 0 444 1 441 0.50 0.01 to 3.53
Syncope Thase 2015a 1 188 0 191 2.04 0.12 to 6.27
Tinnitus NCT03149991 0 25 1 26 0.50 0.01 to 3.58
Twitching of left‐side lower lip NCT03149991 1 25 0 26 2.17 0.13 to 6.50
Ulna fracture NCT00797966 1 121 0 126 2.10 0.13 to 6.36
Unsteadiness of gait NCT03149991 1 25 0 26 2.17 0.13 to 6.50
Urinary hesitancy NCT03149991 1 25 1 26 1.04 0.06 to 3.87
Vertigo Hobart 2018b 1 192 0 202 2.12 0.13 to 6.38
Vivid dreams NCT03149991 0 25 1 26 0.50 0.01 to 3.54
Shift: serum glucose normal to high (< 100 mg/dL to ≥ 126 mg/dL) Bauer 2019 9 444 2 441 4.54 0.98 to 6.08
Shift: serum glucose impaired to high (≥ 100 mg/dL and < 126 mg/dL to ≥ 126 mg/dL) Bauer 2019 15 444 7 441 2.17 0.88 to 3.07
Shift: high‐density lipoprotein‐cholesterol normal to low (≥ 40 mg/dL to < 40 mg/dL) Bauer 2019 27 444 21 441 1.29 0.72 to 1.88
Shift: triglycerides normal to high (< 150 mg/dL to ≥ 200 mg/dL and < 500 mg/dL) Bauer 2019 36 444 27 441 1.35 0.81 to 1.87
Shift: triglycerides normal/borderline to high (< 200 mg/dL to ≥ 200 mg/dL and < 500 mg/dL) Bauer 2019 70 444 46 441 1.61 1.08 to 2.01
High‐density lipoprotein‐cholesterol normal to low (40 mg/dL to < 40 mg/dL) Hobart 2018a 11 197 5 206 2.38 0.81 to 3.45
Triglycerides normal to high (< 150 mg/dL to 200 mg/dL and < 500 mg/dL) Hobart 2018a 11 197 8 206 1.46 0.58 to 2.40
Triglycerides normal/borderline to high (< 200 mg/dL to 200 mg/dL and < 500 mg/dL) Hobart 2018a 24 197 13 206 2.06 1.02 to 2.77
Prolactin > 2 times upper limit: female Hobart 2018a 1 197 0 206 3.15 0.13 to 6.36
Prolactin > 2 times upper limit: male Hobart 2018a 4 197 3 206 1.40 0.31 to 2.91
Prolactin > 3 times upper limit: female Hobart 2018a 2 197 2 206 1.05 0.15 to 3.02
Prolactin > 3 times upper limit: male Hobart 2018a 1 197 0 206 3.15 0.13 to 6.36

CI: confidence interval.

Social functioning or adjustment

After eight weeks of treatment, there was an improvement in social functioning with brexpiprazole compared to placebo, measured using the SDS (MD –0.34, 95% CI –0.50 to –0.17; P < 0.0001, I²= 0%; 7 studies, 3217 participants; Analysis 1.30). Bauer 2019 explored this effect at 18 weeks and found no clear differences between brexpiprazole and placebo regarding improvement the functioning (MD 1.20, 95% CI –0.74 to 3.14; P = 0.23; 1 study, 694 participants; Analysis 1.30).

1.30. Analysis.

1.30

Comparison 1: Brexpiprazole versus placebo, Outcome 30: Social functioning and adjustment

Health‐related quality of life

There was no difference in health‐related quality of life measured using the Q‐LES‐SF between brexpiprazole and placebo at eight or 18 weeks (8 weeks: MD 0.08, 95% CI –0.04 to 0.20; P = 0.93, I²= 0%; 2 studies, 1076 participants; 18 weeks: MD –0.11, 95% CI –0.26 to 0.03; P = 0.13; 1 study, 694 participants; Analysis 1.31).

1.31. Analysis.

1.31

Comparison 1: Brexpiprazole versus placebo, Outcome 31: Health‐related quality of life

Sensitivity analysis

Excluding trials in which allocation concealment was not clearly described, or double‐blinding was unclear

Seven studies were at unclear allocation concealment or blinding and were excluded in this sensitivity analysis (Bauer 2019; NCT00797966; NCT01052077; NCT01837797; NCT03149991; Thase 2015a; Thase 2015b); only two studies remained (Hobart 2018a; Hobart 2018b). When measuring efficacy at eight weeks, there were no differences between groups (OR 0.76, 95% CI 0.53 to 1.09; P = 0.59, I²= 0%; 2 studies, 797 participants; Analysis 1.32).

1.32. Analysis.

1.32

Comparison 1: Brexpiprazole versus placebo, Outcome 32: Sensitivity analysis – excluding studies where allocation concealment or blinding was unclear

Excluding trials with dropout rates of more than 20% in one of the included treatment arms

One study included a study arm with more than 20% attrition (Bauer 2019). For efficacy at eight weeks results remained consistent with the original analysis (OR 1.63, 95% CI 1.33 to 1.99; 7 studies, 2523 participants; I² = 0%; Analysis 1.33).

1.33. Analysis.

1.33

Comparison 1: Brexpiprazole versus placebo, Outcome 33: Sensitivity analysis – excluding trials where attrition was > 20%

Excluding trials with imputed data, and those for which the standard deviation was missing and we imputed it from other trials

All included studies reported continuous outcomes as LSM measures, where SE instead of SD was reported and SD had to be calculated. None of the reported analyses included imputed data and sensitivity analysis was not required.

Excluding trials with one or more domains at high risk of bias

Only NCT01837797 was at high risk of bias in two domains. This study did not report efficacy outcomes.

Subgroup analysis and investigation of heterogeneity

Age groups: our search did not retrieve any studies that included participants below the age of 18 years and, therefore, we could not perform this subgroup analysis.

Fixed versus flexible dose: in terms of efficacy, there were no differences between the studies using a fixed‐dose regimen and the studies using a flexible‐dose regimen, both for the dichotomous and continuous outcomes (dichotomous: test for subgroup differences: Chi² = 1.28, degrees of freedom (df) = 1 (P = 0.26), I² = 22.1%; Analysis 1.35; continuous: test for subgroup differences: Chi² = 0.01, df = 1 (P = 0.94), I² = 0%; Analysis 1.34).

1.35. Analysis.

1.35

Comparison 1: Brexpiprazole versus placebo, Outcome 35: Subgroup analysis – fixed versus flexible dose – efficacy, measured as response to treatment at 8 weeks

1.34. Analysis.

1.34

Comparison 1: Brexpiprazole versus placebo, Outcome 34: Subgroup analysis – fixed versus flexible dose – change in depressive symptoms at 8 weeks

Discussion

Summary of main results

Our review on brexpiprazole for depression included nine RCTs and 3424 participants. The main data are presented in Table 1 and outcome results can be found in the Data and Analyses sections as well as in Table 4 for adverse events that were reported in only one study and did not allow for data synthesis.

All included studies compared adjunctive brexpiprazole to adjunctive placebo as an add‐on treatment to antidepressants in participants with treatment‐resistant depression. At eight weeks, participants who received brexpiprazole were more likely to achieve response than participants who were randomised to placebo, thus suggesting better efficacy of brexpiprazole than placebo. When efficacy of brexpiprazole was explored as a continuous outcome, the mean reduction of the score on MADRS was just above one point. At the same acute effect time point, participants who received brexpiprazole were more likely to drop out due to any cause, and due to adverse effects.

Based on the results of three studies, in a short‐term period of treatment, brexpiprazole demonstrated better efficacy than placebo when added to an antidepressant; however, we were unable to establish its long‐term efficacy due to the very limited evidence available.

The most common adverse effects associated with brexpiprazole were weight gain and akathisia. In comparison, the most commonly reported adverse effects in RCTs of aripiprazole, which has a similar pharmacological profile, against placebo as add‐on treatment for refractory depression were akathisia, restlessness and headache (Berman 2007).

Overall completeness and applicability of evidence

There are several limitations in the overall completeness of our findings. All studies had similar design which required lack of response to at least two other antidepressants before randomisation to either brexpiprazole or placebo resulting in our review including only participants with treatment‐resistant depression. First, failure to respond to another antidepressant was an inclusion criterion for each study and the numbers varied between one and seven antidepressants depending on the study. Once the trial had started, the participants were included in an initial phase during which they received an antidepressant either with or without placebo depending on the study. Those who failed to respond were later randomised to either brexpiprazole or placebo in the next phase of the placebo. As all studies followed the abovementioned design, our search failed to identify studies where participants received brexpiprazole as adjunctive treatment after non‐response to only one antidepressant trial. Moreover, we were unable to find a study that used brexpiprazole as monotherapy for MDD.

The similarity of study design might have contributed to the overall low heterogeneity. However, besides the difference of previous treatment failures to other antidepressants, median duration of the depressive episode was also not specified and described in the published results of studies. Both these factors are amongst the predictors for treatment‐resistant depression (Bartova 2019), and can contribute to differences in severity and chronicity of treatment resistance (Kraus 2020). It remains unclear the extent to which such variability in the number previous treatment failures to other antidepressants and depressive episode duration, if established, contributed to the size of measured effects of brexpiprazole and if participants differing in these characteristics benefited differently. The included studies used different rating scales and cut‐off criteria (e.g. HAM‐D17 score 18 or greater; MADRS score 26 or greater; CGI‐S score 4 or greater) for the assessment of the severity of the current depressive episode. Real‐world data studies demonstrate that clinicians use the severity of depressive symptoms for the choice of augmentation medication for treatment‐resistant depression, however it does not necessarily justify the choice of one option over the other (Dold 2018). Supporting evidence from individual participant data on antidepressant efficacy demonstrates that variability in response rate is not associated with baseline depression severity (Furukawa 2018). The significance of baseline severity of symptoms of depression for the effect of medications such as antipsychotics and brexpiprazole on response and other outcomes of interest requires further clarification; the studies excluded participants with psychotic depression. This may limit the applicability of the findings from this review to the real‐world population of patients as the prevalence of major depressive episode with psychotic features is approximately 0.4% and prevalence of psychotic features in patients with major depression varies between 10.92% (Dold 2019) and 18.5% (Ohayon 2002), thus representing a significant proportion of participants with major depression.

Three studies did not report detailed information around random sequence generation (selection bias) and included only the description "randomised" (NCT00797966; NCT01052077; NCT01837797). These studies were all conducted by the pharmaceutical company, followed the same study design and used similar methodology to the other five RCTs conducted by Otsuka. Therefore, even when the description of the randomisation process was not detailed, we assumed these RCTs to be of low risk of bias in the process of random allocation of participants to separate treatment groups.

Our review identified only one study on the long‐term efficacy of brexpiprazole for depression and there was no difference in its results for efficacy, with very wide CIs, so we could not draw any conclusion around the long‐term efficacy of the intervention. All studies in the systematic review included placebo as a comparator as we failed to identify antidepressive agents as comparators for brexpiprazole. Only one study included low‐dose quetiapine as an active comparator; however, atypical antipsychotics were not included in our study protocol and were not subject to this review (Ralovska 2021).

Quality of the evidence

The overall certainty of the evidence for the efficacy of brexpiprazole as adjunctive treatment for the acute treatment of depression both as a dichotomous and continuous outcome as well as for remission was high, thus allowing us to conclude with greater certainty of better efficacy of brexpiprazole than placebo adjunctive to an antidepressant for the treatment of MDD. The certainty of evidence was moderate for the incidence of the main adverse effects, akathisia and weight gain, and, therefore, these effect estimates should be viewed with caution. Evidence was lacking on how these effects were measured and, therefore, indirectness of the evidence was suspected. Another possible factor that could have impacted the certainty of the evidence is that participants in all studies were receiving different antidepressive agents and, while they were randomised to either brexpiprazole or placebo as add‐on treatments, aggregate data did not allow us to specify the impact of this difference on the overall safety and efficacy results.

One of the difficulties of the review process was drawing conclusions from studies involving only elderly people. Only one study included exclusively participants above the age of 65 years and it had a high risk of bias as it had prespecified efficacy results but did not report on any. This study had very small number of participants, so its external validity was also severely compromised. Another study including only elderly people is still awaiting classification as it did not report its inclusion criteria sufficiently. However, the study sample of this study is also very small, and it is likely that it will not provide sufficient evidence to draw sufficient conclusions around the effects of brexpiprazole in this age group.

High‐certainty evidence contributed to the effect estimates of remission, dropouts due to any cause and dropouts due to adverse effects and conclusions around these outcomes can be drawn with higher level of certainty.

Potential biases in the review process

Our search was comprehensive, however, we cannot rule out the possibility for missing studies. We were unable to retrieve all data from included studies as well as results from two studies awaiting classification. We contacted four study authors for more details, but we either did not receive responses or the authors were unable to provide information as they no longer had access to it. Due to the small number of studies (fewer than 10), we were also unable to explore publication bias (Page 2021).

Eight studies were sponsored by the manufacturer and only one study was independently funded (NCT03149991). The impact of sponsorship on the results of this review was unclear and results on different outcomes of interest might have been differently affected. In general, inclusion of industry‐funded studies might result in more favourable efficacy results (Lundh 2017). Industry funding might also produce different effects on different outcomes of interest (e.g. safety results are less likely to be affected by sponsorship bias) (Lundh 2017). However, the association between sponsorship and the increase in effect size for efficacy outcome measures of antidepressants in particular was not confirmed by a more recent publication (Cipriani 2018).

Brexpiprazole is a relatively new medication, and the results of the included studies may be impacted by "novelty bias" or "wish bias", which is a magnification of the effectiveness of the novel agents in comparison to the older agents (Barbui 2004; Salanti 2010). The factors contributing to this bias might include optimism bias and conflicts of interest and can be impacted by the quality of the included studies; the exact mechanisms that define it remain unclear (Salanti 2010). All included studies in the current review lacked active comparators. Whether their results were subject to novelty bias when the outcome measures involve comparisons to placebo might require separate investigation, more studies, and comparison of direct and indirect data through multiple‐treatment meta‐analysis (Salanti 2010).

Two studies changed their inclusion criteria for the double‐blind randomisation phase by an amendment in the study protocol (Amendment 3) (FDA 2015b), while the studies were ongoing (Thase 2015b; Thase 2015a); initial criteria for inclusion in the double‐blind randomisation phase allowed for participants who varied between response and non‐response to treatment during the initial study phase to be classified as non‐responders and be included in the double‐blind randomisation phase as long as they demonstrated lack of response at the last visit before randomisation; however, this variation in response was reconsidered from a clinical perspective after the randomisation had already started (FDA 2015b), and the criteria for inclusion in the double‐blind phase were reviewed. This led to the introduction of Amendment 3 and according to it non‐response should have been present on every visit of the initial phase. One issue arising with this amendment is that it was applied retrospectively, after randomisation had already begun, and it may have introduced bias due to change in criteria for selection of participants (FDA 2015b). However, it can be argued that Amendment 3 was justified as it revealed participants who were consistent in their lack of response to antidepressant and, therefore, the efficacy of brexpiprazole during the double‐blind phase would not be confounded by fluctuation in presentation and variability in the efficacy of the antidepressant. This is particularly relevant for the establishment of the short‐term efficacy of brexpiprazole. Therefore, in our analysis we included only participants who met inclusion criteria as per this amendment out of all participants in study groups.

Agreements and disagreements with other studies or reviews

Three relatively recent reviews have explored the effects of brexpiprazole as adjunctive treatment for depression (Beyer 2016; Kishi 2019; Yoon 2017). Four pooled post‐hoc analyses have also been published (Hobart 2018c; McIntyre 2016; Nelson 2018; Thase 2019a). All these publications confirmed our findings of better efficacy of brexpiprazole in comparison to placebo as add‐on treatment for depression. Differences in effect size arose most likely from the lesser number of the included studies in the previous reviews, as most of them are outdated with several RCTs being published after them. Similarly to our review, akathisia was the most common adverse effect. In some reviews, headache was a more common adverse effect than weight gain (Thase 2019a; Thase 2019b), and Thase 2019b found restlessness was the third most common adverse effect; these differences can again be justified by the different number of studies and participants included. The most comprehensive and recent review is Kishi 2019, while the remaining reviews included significantly fewer RCTs. Kishi 2019 included nine studies, eight of which overlapped with studies included in the current review. One study was listed in our review as a study awaiting classification; the rationale for this decision was the lack of information on treatment with antidepressants prior to study inclusion (NCT01670279). Our review also included NCT03149991, which was ongoing at the time of publication of Kishi 2019. The primary and secondary outcomes were defined similarly to the ones outlined in our review; however, more frequent time points were specified – every week from week one to week six. Efficacy measures from Kishi 2019 are consistent with findings from our review with brexpiprazole demonstrating better efficacy in achieving response in comparison to placebo and changes in depressive symptoms at six weeks. Another difference between both reviews is that Kishi 2019 divided the studies into two groups (studies including doses of brexpiprazole of 2 mg or less and studies including doses greater than 2 mg) as 2 mg was the FDA recommended dose of brexpiprazole for depression. While reporting on dose–response effect has good clinical and practical implications, the methodology needs to allow for more change points on the dose range and to prevent against presumptuous division of doses (Furukawa 2019). Providing evidence for this notion, one recent systematic review and dose–effect meta‐analysis confirmed that lower doses of brexpiprazole (1 mg to 2 mg) optimally balanced between efficacy, acceptability and tolerability of the medication for participants with MDD; however, the review was limited by the small number of studies (Furukawa 2022). An earlier review, relying on similar methodology confirmed such dose–effect relationship for aripiprazole, suggesting that lower range of the licensed dose (2 mg to 5 mg) proved to be most effective and that increases of the dose of aripiprazole beyond 10 mg proved to not be more beneficial in case of inadequate response to antidepressants (Furukawa 2021). Both our review and Kishi 2019 included subgroup analysis for fixed and flexible doses (Furukawa 2020). Kishi 2019 also explored the number needed to treat (NNT), the mean number of participants needed to treat to obtain one response due to treatment. While useful, the methodology used relied on the Lapaucis method for estimation, which is an unadjusted NNT method and does not take into account explanatory variables such as treatment week or flexible versus fixed dose; therefore, other adjusted methods might be more applicable for the benefit of more precise estimates (Vancak 2021).

Authors' conclusions

Implications for practice.

In our systematic review on brexpiprazole for the treatment of major depressive disorder (MDD), we aimed to outline the clinical applicability of this new medication in MDD. To achieve more comprehensive results, we searched for evidence for the effects of brexpiprazole compared to placebo or other antidepressants in participants from all age groups with a history of non‐response to one antidepressant or history of treatment‐resistant depression. Our findings were limited and should be interpreted only for adults with treatment‐resistant depression. They should be interpreted with caution due to the low number of included studies and varying certainty of evidence (from moderate to high certainty) for several primary outcomes.

We can conclude with high certainty that brexpiprazole as adjunctive therapy can be beneficial in effectively reducing the symptoms of depression. This effect was established for both short‐term (up to two weeks) and acute treatment (eight weeks). The mechanisms through which brexpiprazole produces its short‐term effects are unclear, but short‐term efficacy of any treatment in depression might be very beneficial due to the initial delayed onset of action of antidepressants, and it is worth further exploring if this effect would be preserved if brexpiprazole is given as monotherapy. We were unable to establish whether brexpiprazole would remain efficacious in the long‐term treatment (up to 24 weeks) of depression due to lack of evidence. Brexpiprazole might also be effective for achieving remission after eight weeks of treatment in adults who have had at least two previous unsuccessful treatments with another antidepressant monotherapy. Due to the small number of studies in our review, we were unable to compare the efficacy of brexpiprazole in fixed or flexible dosage regimens. Brexpiprazole has a relatively long half‐life and the dosage regimen might impact both its efficacy and tolerability; given the nature of major depression, it is worth exploring how brexpiprazole should be titrated in clinical settings.

Brexpiprazole probably caused a higher incidence of adverse effects, including akathisia and weight gain; however, these findings were also limited by the moderate certainty of the evidence. Adverse effects were reported as aggregate data for the entire study period and through follow‐up which did not allow us to specify when they emerged and if and when they resolved. As we included only aggregate data, we were also unable to establish a difference in the effects of brexpiprazole when added to different antidepressants; the adverse effects of the first‐line antidepressant treatment might need to be considered due to cumulative risks of adverse drug–drug interactions.

Implications for research.

Further accumulation of evidence for brexpiprazole in depression might contribute to increased clarity regarding its effects, with several areas for research already outlined. Our review failed to provide any evidence for the effects of brexpiprazole in children and adolescents; brexpiprazole is also not currently approved for the treatment of depression in this age group. We also did not find sufficient evidence to establish the effects of brexpiprazole in elderly people with depression. We failed to retrieve any studies that included antidepressants as active comparators or studies that included participants with lack of response to only one antidepressant.

On a wider scope in the field of major depression, guidelines offer different options for enhancing response to treatment to antidepressant monotherapy in moderate‐to‐severe depression when it has not been initially achieved; such options may include lithium, second‐generation antipsychotic drugs, liothyronine and stimulants (Malhi 2021). Newer pharmacological treatments for depression, such as pramipexole, are also emerging. Reports on pramipexole's efficacy and safety, both as monotherapy and add‐on therapy, have already been published (Tundo 2019); more ongoing studies, such as the PAX‐D study, are to explore its efficacy and acceptability in treatment‐resistant depression (Au‐Yeung 2021). With so many options available and new treatments being marketed, clinicians' choices of the best treatment for a person with depression should be based on robust and substantive evidence.

Commonly, conclusions about efficacy of pharmacological agents are drawn by comparing different active treatments with placebo when they are not sufficient to inform clinical decisions of one available treatment option over the other (Cipriani 2020; Naci 2020; Singh 2020). Instead, such decisions should rely more closely on evidence from comparative efficacy studies and should be based on the best available methodology for this purpose, such as network meta‐analysis (NMA) (Naci 2020). One recent NMA explored the comparative efficacy and acceptability of pharmacological augmentation for depression, including brexpiprazole, other atypical antipsychotic drugs, antidepressants, stimulants, thyroid hormones, lithium and placebo (Nuñez 2022). There were only two statistically significant results for brexpiprazole and they demonstrated better efficacy than placebo and worse acceptability than T4 thyroid hormone. This analysis had several limitations, such as very broad inclusion criteria with no threshold for percentage of participants with bipolar disorder and psychotic depression, as well as no differentiation between the number of failed antidepressant treatments before inclusion. However, research in this field is highly needed as it will easily outline lack of comparative effectiveness randomised controlled trials and would provide evidence of how the already existing studies position different therapeutic options in comparison to others. Another important area for research is the individual response and preference of treatment, which can be explored through individual patient data. Strategies to approach non‐response, such as augmentation or switching can also be compared (Kato 2018), and comparisons can be made between pharmacological, non‐pharmacological treatments and their combination (Cuijpers 2020).

Finally, our review did not include cost‐effectiveness analysis of brexpiprazole for depression and further research on the potential of brexpiprazole to reduce both the clinical and economic burden for patients who are in jeopardy to transition to treatment‐resistant depression or are already experiencing treatment‐resistant depression is needed. In the US, almost half of the economic costs for depression can be attributed to treatment‐resistant depression. Effective interventions should also be identified in the context of cost‐effectiveness relation as treatment for treatment‐resistant depression might be continuous. Atypical antipsychotic drugs can result in higher mean adjusted pharmacy costs and higher mean mental health‐related costs than antidepressant switching (Hassan 2016); however, this can expectedly vary in different settings and sites.

History

Protocol first published: Issue 2, 2021

Acknowledgements

Cochrane Common Mental Disorders (CCMD) supported the authors in the development of this review.

The following people conducted the editorial process for this article.

  • Sign‐off Editor (final editorial decision): Nick Meader, Deputy Co‐ordinating Editor, CCMD and University of Newcastle.

  • Managing Editor (selected peer reviewers, collated peer‐reviewer comments, provided editorial guidance to authors, edited the article): Jessica Hendon, CCMD, Centre for Reviews and Dissemination, University of York.

  • Information Specialist (search review, provided editorial guidance to authors, edited the article): Sarah Dawson, CCMD and University of Bristol.

  • Peer reviewers (provided comments and recommended an editorial decision): Simon Davies, Department of Psychiatry, University of Toronto, Canada and Centre for Addiction and Mental Health, Toronto, Canada (clinical/content review); Verity Westgate, University of Oxford (consumer review); Nuala Livingstone, Associate Editor, Cochrane Evidence Production and Methods Directorate (methods review); David Marshall, University of York (methods review); Gillian Worthy, Adelphi Values (methods review).

  • Copy Editor (copy‐editing and production): Anne Lawson, Cochrane Central Production Service.

The review authors are grateful to the peer reviewers for their time and comments. We are grateful to Dr Franco De Creszenzo for his support with data collection, study appraisal and critical evaluation of the protocol and review. We would also like to thank Cochrane Copy Edit Support for the team's help.

Cochrane Group funding acknowledgement: the UK National Institute for Health Research (NIHR) is the largest single funder of the Cochrane Common Mental Disorders Group.

Disclaimer: the views and opinions expressed herein are those of the review authors and do not necessarily reflect those of the NIHR, National Health Service (NHS), or the Department of Health and Social Care.

Appendices

Appendix 1. Database search strategies

Searches were initially conducted on 17 September 2020 and updated 2 November 2021.

Cochrane Central Register of Controlled Trials 
(CENTRAL; Issue 11 of 12, 2021) in the Cochrane Library
((Brexpiprazole or OPC‐34712 or OPC34712 or Rexulti or Rxulti) AND (“affective disorder*” or “affective symptom*” or mood* or mental or depress* or MDD))

Cochrane Common Mental Disorders Controlled Trials Register (CCMDCTR) 
(all years to June 2016) (archived register)
(Brexpiprazole or OPC‐34712 or OPC34712 or Rexulti or Rxulti) n=13

PubMed not MEDLINE (all years to 2 November 2021)
(Brexpiprazole or OPC‐34712 or OPC34712 or Rexulti or Rxulti) AND pubmednotmedline[sb] n=30

Ovid MEDLINE(R) ALL (1946 to 2 November 2021)
(Brexpiprazole or OPC‐34712 or OPC34712 or Rexulti or Rxulti).mp. not (exp Animals/ not Humans/)

Ovid Embase (1974 to 2 November 2021)
1 Brexpiprazole/ 
2 (Brexpiprazole or OPC‐34712 or OPC34712 or Rexulti or Rxulti).mp. 
3 (1 or 2)
4 exp Depression/ 
5 (affective disorder* or affective symptom* or mood* or mental or depress* or MDD).mp.
6 (4 or 5) 
7 (3 and 6)
8 Randomized Controlled Trial/
9 Randomization/
10 Controlled Clinical Trial/ 
11 Placebo/
12 placebo.ti,ab.
13 trial.ti.
14 ((singl* or doubl* or trebl* or tripl*) adj3 (blind* or mask* or dummy)).mp.
15 (RCT or random*).mp.
16 exp Clinical Trial/
17 or/8‐16
18 (7 and 17) 
19 ((Animal/ or Nonhuman/) not Human/)
20 (18 not 19)

Ovid APA PsycINFO (all years to November Week 1 2021)
(Brexpiprazole or OPC‐34712 or OPC34712 or Rexulti or Rxulti).mp.

WHO‐ICTRP
Search‐1 (17 September 2020) Brexpiprazole and MDD OR Brexpiprazole and depression OR Brexpiprazole and depressive OR Brexpiprazole and depressed OR Brexpiprazole and mood OR Brexpiprazole and mental OR Brexpiprazole and affective disorder OR Brexpiprazole and affective disorders OR Brexpiprazole and affective symptom OR Brexpiprazole and affective symptoms OR OPC‐34712 and MDD OR OPC‐34712 and depression OR OPC‐34712 and depressive OR OPC‐34712 and depressed OR OPC‐34712 and mood OR OPC‐34712 and mental OR OPC‐34712 and affective disorder OR OPC‐34712 and affective disorders OR OPC‐34712 and affective symptom OR OPC‐34712 and affective symptoms OR OPC34712 and MDD OR OPC34712 and depression OR OPC34712 and depressive OR OPC34712 and depressed OR OPC34712 and mood OR OPC34712 and mental OR OPC34712 and affective disorder OR OPC34712 and affective disorders OR OPC34712 and affective symptom OR OPC34712 and affective symptoms OR Rexulti and MDD OR Rexulti and depression OR Rexulti and depressive OR Rexulti and depressed OR Rexulti and mood OR Rexulti and mental OR Rexulti and affective disorder OR Rexulti and affective disorders OR Rexulti and affective symptom OR Rexulti and affective symptoms OR Rxulti and MDD OR Rxulti and depression OR Rxulti and depressive OR Rxulti and depressed OR Rxulti and mood OR Rxulti and mental OR Rxulti and affective disorder OR Rxulti and affective disorders OR Rxulti and affective symptom OR Rxulti and affective symptoms

Search‐2 Update (2 November 2021) 
Drug terms only, records posted >September 2021
Brexpiprazole or OPC‐34712 or OPC34712 or Rexulti or Rxulti

ClinicalTrials.gov (2 November 2021)
Brexpiprazole OR OPC‐34712 OR OPC34712 OR Rexulti OR Rxulti

***************************

Data and analyses

Comparison 1. Brexpiprazole versus placebo.

Outcome or subgroup title No. of studies No. of participants Statistical method Effect size
1.1 Response to treatment at 8 weeks 8 3409 Odds Ratio (M‐H, Random, 95% CI) 1.47 [1.23, 1.75]
1.1.1 Response to treatment at 8 weeks 8 3409 Odds Ratio (M‐H, Random, 95% CI) 1.47 [1.23, 1.75]
1.2 Acceptability (dropouts due to any cause) at 8 weeks 7 2523 Odds Ratio (M‐H, Random, 95% CI) 1.37 [1.02, 1.83]
1.2.2 Acceptability (all‐cause dropouts) 7 2523 Odds Ratio (M‐H, Random, 95% CI) 1.37 [1.02, 1.83]
1.3 Tolerability (dropouts due to adverse effects) at 8 weeks 6 2472 Odds Ratio (M‐H, Random, 95% CI) 2.88 [1.37, 6.05]
1.3.3 Tolerability (dropouts due to adverse events) 6 2472 Odds Ratio (M‐H, Random, 95% CI) 2.88 [1.37, 6.05]
1.4 Response at different time points (secondary outcome) 4   Odds Ratio (M‐H, Fixed, 95% CI) Subtotals only
1.4.1 Early response at 2 weeks (ranging between ≥ 1 and < 4 weeks) 3 1452 Odds Ratio (M‐H, Fixed, 95% CI) 1.65 [1.04, 2.63]
1.4.2 Long‐term response at 18 weeks (ranging between > 12 and ≤ 24 weeks) 1 886 Odds Ratio (M‐H, Fixed, 95% CI) 0.88 [0.68, 1.15]
1.5 Remission 7   Odds Ratio (M‐H, Fixed, 95% CI) Subtotals only
1.5.2 At 8 weeks (ranging between ≥ 4 and < 12 weeks) 7 3358 Odds Ratio (M‐H, Fixed, 95% CI) 1.46 [1.19, 1.79]
1.6 Change in depressive symptoms at 8 weeks 8   Mean Difference (IV, Random, 95% CI) Subtotals only
1.6.2 At 8 weeks (ranging between ≥ 4 and ≤ 12 weeks) 8 3263 Mean Difference (IV, Random, 95% CI) ‐1.39 [‐1.96, ‐0.82]
1.7 Safety – total number of adverse effects 9 3422 Odds Ratio (M‐H, Random, 95% CI) 1.63 [1.29, 2.06]
1.8 Safety – accidental overdose 2 900 Odds Ratio (M‐H, Fixed, 95% CI) 1.02 [0.59, 1.76]
1.9 Safety – akathisia 7 3358 Odds Ratio (M‐H, Fixed, 95% CI) 2.95 [2.06, 4.21]
1.10 Safety – anxiety 4 2336 Odds Ratio (M‐H, Fixed, 95% CI) 1.02 [0.55, 1.88]
1.11 Safety – bruxism 2 66 Odds Ratio (M‐H, Fixed, 95% CI) 0.85 [0.12, 6.09]
1.12 Safety – constipation 2 423 Odds Ratio (M‐H, Fixed, 95% CI) 3.18 [1.08, 9.41]
1.13 Safety – diarrhoea 3 998 Odds Ratio (M‐H, Fixed, 95% CI) 1.02 [0.58, 1.80]
1.14 Safety – dizziness 4 1354 Odds Ratio (M‐H, Fixed, 95% CI) 1.10 [0.49, 2.45]
1.15 Safety – fatigue 6 3111 Odds Ratio (M‐H, Fixed, 95% CI) 1.28 [0.82, 2.01]
1.16 Safety – headache 7 3030 Odds Ratio (M‐H, Fixed, 95% CI) 0.81 [0.61, 1.08]
1.17 Safety – increased appetite 3 826 Odds Ratio (M‐H, Fixed, 95% CI) 3.98 [1.54, 10.32]
1.18 Safety – insomnia 8 3372 Odds Ratio (M‐H, Fixed, 95% CI) 1.69 [1.09, 2.62]
1.19 Safety – irritability 2 454 Odds Ratio (M‐H, Fixed, 95% CI) 1.66 [0.43, 6.41]
1.20 Safety – nasopharyngitis 6 2979 Odds Ratio (M‐H, Fixed, 95% CI) 1.21 [0.86, 1.70]
1.21 Safety – nausea 2 298 Odds Ratio (M‐H, Fixed, 95% CI) 0.78 [0.17, 3.53]
1.22 Safety – panic attack 2 650 Odds Ratio (M‐H, Fixed, 95% CI) 0.76 [0.25, 2.32]
1.23 Safety – radius fracture 2 1132 Odds Ratio (M‐H, Fixed, 95% CI) 3.07 [0.32, 29.61]
1.24 Safety – restlessness 5 2739 Odds Ratio (M‐H, Fixed, 95% CI) 2.26 [1.35, 3.80]
1.25 Safety – sedation 3 824 Odds Ratio (M‐H, Fixed, 95% CI) 1.65 [0.42, 6.40]
1.26 Safety – somnolence 5 2739 Odds Ratio (M‐H, Fixed, 95% CI) 3.56 [2.02, 6.28]
1.27 Safety – suicidal ideation 2 937 Odds Ratio (M‐H, Fixed, 95% CI) 0.56 [0.12, 2.63]
1.28 Safety – tremor 2 692 Odds Ratio (M‐H, Fixed, 95% CI) 1.52 [0.65, 3.53]
1.29 Safety – weight gain 9 3424 Odds Ratio (M‐H, Fixed, 95% CI) 3.14 [2.19, 4.49]
1.30 Social functioning and adjustment 7   Mean Difference (IV, Fixed, 95% CI) Subtotals only
1.30.1 At 8 weeks (ranging between ≥ 4 and ≤ 12 weeks) 7 3217 Mean Difference (IV, Fixed, 95% CI) ‐0.34 [‐0.50, ‐0.17]
1.30.2 At 18 weeks (ranging between > 12 and ≤ 24 weeks) 1 694 Mean Difference (IV, Fixed, 95% CI) 1.20 [‐0.74, 3.14]
1.31 Health‐related quality of life 2   Std. Mean Difference (IV, Fixed, 95% CI) Subtotals only
1.31.1 At 8 weeks (ranging between ≥ 4 and ≤ 12 weeks) 2 1076 Std. Mean Difference (IV, Fixed, 95% CI) ‐0.01 [‐0.13, 0.11]
1.31.2 At 18 weeks (ranging between > 12 and ≤ 24 weeks) 1 694 Std. Mean Difference (IV, Fixed, 95% CI) ‐0.11 [‐0.26, 0.03]
1.32 Sensitivity analysis – excluding studies where allocation concealment or blinding was unclear 2 797 Odds Ratio (M‐H, Random, 95% CI) 0.76 [0.53, 1.09]
1.33 Sensitivity analysis – excluding trials where attrition was > 20% 7 2523 Odds Ratio (M‐H, Random, 95% CI) 1.63 [1.33, 1.99]
1.34 Subgroup analysis – fixed versus flexible dose – change in depressive symptoms at 8 weeks 8   Mean Difference (IV, Random, 95% CI) Subtotals only
1.34.1 Fixed dose 3 1401 Mean Difference (IV, Random, 95% CI) ‐2.01 [‐3.40, ‐0.62]
1.34.2 Flexible dose 5 1862 Mean Difference (IV, Random, 95% CI) ‐1.14 [‐1.73, ‐0.55]
1.35 Subgroup analysis – fixed versus flexible dose – efficacy, measured as response to treatment at 8 weeks 8   Odds Ratio (M‐H, Random, 95% CI) Subtotals only
1.35.1 Fixed dose 3 1450 Odds Ratio (M‐H, Random, 95% CI) 1.51 [1.16, 1.97]
1.35.2 Flexible dose 5 1959 Odds Ratio (M‐H, Random, 95% CI) 1.49 [1.12, 1.97]

1.8. Analysis.

1.8

Comparison 1: Brexpiprazole versus placebo, Outcome 8: Safety – accidental overdose

1.10. Analysis.

1.10

Comparison 1: Brexpiprazole versus placebo, Outcome 10: Safety – anxiety

1.11. Analysis.

1.11

Comparison 1: Brexpiprazole versus placebo, Outcome 11: Safety – bruxism

1.13. Analysis.

1.13

Comparison 1: Brexpiprazole versus placebo, Outcome 13: Safety – diarrhoea

1.14. Analysis.

1.14

Comparison 1: Brexpiprazole versus placebo, Outcome 14: Safety – dizziness

1.15. Analysis.

1.15

Comparison 1: Brexpiprazole versus placebo, Outcome 15: Safety – fatigue

1.16. Analysis.

1.16

Comparison 1: Brexpiprazole versus placebo, Outcome 16: Safety – headache

1.17. Analysis.

1.17

Comparison 1: Brexpiprazole versus placebo, Outcome 17: Safety – increased appetite

1.19. Analysis.

1.19

Comparison 1: Brexpiprazole versus placebo, Outcome 19: Safety – irritability

1.20. Analysis.

1.20

Comparison 1: Brexpiprazole versus placebo, Outcome 20: Safety – nasopharyngitis

1.21. Analysis.

1.21

Comparison 1: Brexpiprazole versus placebo, Outcome 21: Safety – nausea

1.22. Analysis.

1.22

Comparison 1: Brexpiprazole versus placebo, Outcome 22: Safety – panic attack

1.23. Analysis.

1.23

Comparison 1: Brexpiprazole versus placebo, Outcome 23: Safety – radius fracture

1.25. Analysis.

1.25

Comparison 1: Brexpiprazole versus placebo, Outcome 25: Safety – sedation

1.27. Analysis.

1.27

Comparison 1: Brexpiprazole versus placebo, Outcome 27: Safety – suicidal ideation

1.28. Analysis.

1.28

Comparison 1: Brexpiprazole versus placebo, Outcome 28: Safety – tremor

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Bauer 2019.

Study characteristics
Methods Phase 3, multicentre, randomised, double‐blind, parallel group, placebo‐controlled, flexible‐dose, long‐term study
Participants Outpatients; men and women aged ≥ 18 and ≤ 75 years
Key inclusion criteria: primary diagnosis of MDD according to DSM‐IV‐TR criteria (current MDE) confirmed using the MINI; MADRS Total score ≥ 26, CGI‐S score ≥ 4 at screening visit and start of prospective treatment period; participants had an insufficient response to ≥ 1 and < 3 adequate trials of treatment with an antidepressant.
For inclusion in double‐blind randomised phase (Phase B):
  • < 50% reduction in MADRS Total score between start of prospective treatment and all visits in prospective period;

  • MADRS Total score ≥18 at randomisation visit;

  • CGI‐I score ≥ 3 (minimally improved) at every visit in prospective period.


Key exclusion criteria: DSM‐IV‐TR Axis I diagnosis other than MDD; presenting with suicidal ideation or behaviour; substance abuse or dependence within past 180 days; current Axis II (DSM‐IV‐TR) diagnosis of borderline, antisocial, paranoid, schizoid, schizotypical or histrionic personality disorder; experienced/experiences hallucinations, delusions or any psychotic symptomatology in the current episode.
Interventions
  • Brexpiprazole 1–3 mg + antidepressant, 444 participants randomised

  • Placebo + antidepressant, 442 participants randomised


Antidepressants: escitalopram 10–20 mg/day; fluoxetine 20–40 mg/day; paroxetine IR 20–40 mg/day; sertraline 50–200 mg/day; duloxetine 60 mg/day; venlafaxine 75–225 mg/day
Outcomes Primary outcome: full remission defined as MADRS Total score ≤ 10 and ≥ 50% decrease from randomisation in MADRS Total score for ≥ 8 consecutive weeks during randomised treatment.
Key secondary outcomes: full functional remission using SDS Total score ≤ 6 and all SDS domain scores ≤ 2 observed for ≥ 8 consecutive weeks during randomised treatment period; change from randomisation to week 6 and week 24 in MADRS Total score; response at week 6 and week 24 during randomised treatment period; acceptability; tolerability, safety, change from randomisation to week 6 and week 24 in SDS Total score during randomised treatment period.
Notes  
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Low risk Quote: "Treatment assignments were based on a computer generated randomisation code provided by the study sponsor."
Allocation concealment (selection bias) Low risk Quote: "Double‐blind study medications were provided by the sponsor, in 2‐week wallet cards, and were assigned by the IVRS/IWRS [Interactive Voice Response Systems/Interactive Web Response Systems]. The tablets for brexpiprazole and placebo were identical in appearance."
Blinding of participants and personnel (performance bias)
All outcomes Low risk Quote: "Masking: double (participant, investigator)."
Blinding of outcome assessment (detection bias)
All outcomes Unclear risk Quote: "At each visit, the MADRS Total score and the CGI‐S and CGI‐I scores were entered into, and analysed by, an interactive voice/ web‐response system (IVRS/IWRS) in order to maintain the blinding."
Incomplete outcome data (attrition bias)
All outcomes Unclear risk Discrete reported attrition rates according to protocol criteria (156/886 (17.6%) participants). Unbalanced between groups: 61/442 (13.8%) in placebo arm and 95/444 (21.4%) in brexpiprazole arm. Attrition bias was likely.
Selective reporting (reporting bias) Low risk Outcomes reported as listed in protocol.
Other bias Unclear risk No information on other bias available.

Hobart 2018a.

Study characteristics
Methods Phase 3, multicentre, randomised, double‐blind, placebo and active comparator‐controlled trial, flexible dose
Participants Outpatients; men or women aged 18–75 years
Key inclusion criteria: diagnosis of MDD and in a current MDE, as defined by DSM‐IV‐TR criteria and confirmed by both the MINI and an adequate clinical psychiatric evaluation; current MDE ≥ 8 weeks in duration; history for the current MDE of an inadequate response to ≥ 1 and < 3 adequate antidepressant treatments (an inadequate response defined as < 50% reduction in depressive symptom severity, as assessed by the ATRQ); MADRS Total score ≥ 26 at screening and baseline visits.
Key exclusion criteria: inadequate response (< 50% reduction in depressive symptom severity) to > 3 antidepressant treatments during the current MDE at a therapeutic dose (as defined by the ATRQ); received ECT for the current MDE; inadequate response to ECT; current Axis I (DSM‐IV‐TR) diagnosis of another disorder; current Axis II (DSM‐IV‐TR) diagnosis of borderline, antisocial, paranoid, schizoid, schizotypal or histrionic personality disorder, or mental retardation.
Interventions
  • Brexpiprazole 2–3 mg + antidepressant, 197 participants randomised

  • Placebo + antidepressant, 206 participants randomised

  • Quetiapine XR 150–300 mg/day + antidepressant; 100 participants


(Randomisation ratio 2:2:1)
Antidepressants: escitalopram (10–20 mg/day); fluoxetine (20–40 mg/day); paroxetine CR (25–50 mg/day); sertraline (50–200 mg/day); duloxetine (30–60 mg/day); venlafaxine XR (37.5–225 mg/day)
Outcomes Primary outcome: change in MADRS Total score in week 6 of randomisation.
Key secondary outcomes: response rate: 50% reduction in MADRS Total score (week 6); CGI‐I response rate (week 6); MADRS Remission Rate at week 6 (MADRS Total score ≤ 10 and 50% reduction in MADRS Total score); acceptability; tolerability; safety; change in SDS score in week 6.
Notes Brexpiprazole + antidepressant – number of prior antidepressant failures: 1 antidepressant: 168 (85.3%) participants; 2 antidepressants: 26 (13.2%) participants; 3 antidepressants: 3 (1.5%) participants; 4 antidepressants: 0 (0%) participants.
Placebo + antidepressant: number of prior antidepressant failures: 1 antidepressant: 171 (83.0%) participants; 2 antidepressants: 28 (13.6%) participants; 3 antidepressants: 5 (2.4%) participants; 4 antidepressants: 1 (0.5%) participant.
The study included a quetiapine + antidepressant arm, which was excluded from the review as it did not meet the predefined protocol inclusion criteria for comparator intervention.
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Low risk Quote: "Treatment assignments were based on a central, computer‐generated randomisation code provided by the study sponsor."
Allocation concealment (selection bias) Low risk Quote: "Double‐blind study medications were provided by the sponsor in numbered, weekly blister cards, and were assigned by the IVRS/IWRS."
Blinding of participants and personnel (performance bias)
All outcomes Low risk Quote: "… to reduce potential bias, all study site personnel and patients were blinded to the phases of the study, the allocation of adjunctive treatments, the randomization criteria, the timing of randomization, and the timing of the efficacy endpoints."
Quote: "To maintain blinding, patients randomized to brexpiprazole tablets also took quetiapine XR placebo capsules, patients randomized to quetiapine XR capsules also took brexpiprazole placebo tablets, and patients randomized to placebo took both brexpiprazole and quetiapine XR placebos."
Quote: "Masking: quadruple (participant, care provider, investigator, outcomes assessor)."
Blinding of outcome assessment (detection bias)
All outcomes Low risk Quote: "… to reduce potential bias, all study site personnel and patients were blinded to the phases of the study, the allocation of adjunctive treatments, the randomization criteria, the timing of randomization, and the timing of the efficacy endpoints."
Quote: "To maintain blinding, patients randomized to brexpiprazole tablets also took quetiapine XR placebo capsules, patients randomized to quetiapine XR capsules also took brexpiprazole placebo tablets, and patients randomized to placebo took both brexpiprazole and quetiapine XR placebos."
Quote: "Masking: quadruple (participant, care provider, investigator, outcomes assessor)."
Incomplete outcome data (attrition bias)
All outcomes Low risk We found 60 discontinuations out of 503 participants randomised (11.9%). 10% discontinuations for placebo arm, 13% for brexpiprazole arm, 14% for quetiapine arm; fairly balanced between groups.
Selective reporting (reporting bias) Low risk Outcomes reported as listed in protocol.
Other bias Unclear risk No information on other bias available.

Hobart 2018b.

Study characteristics
Methods Phase 3, multicentre, randomised, double‐blind, placebo‐controlled trial, fixed‐dose
Participants Outpatients; men and women aged 18–65 years
Key inclusion criteria: diagnosis of MDD and in a current MDE, as defined by DSM‐IV‐TR criteria and confirmed by both the MINI and an adequate clinical psychiatric evaluation; the current MDE ≥ 8 weeks in duration; history for the current MDE of an inadequate response to ≥ 1 and < 3 adequate antidepressant treatments (an inadequate response defined as < 50% reduction in depressive symptom severity, as assessed by the ATRQ); HAM‐D17 Total score ≥ 18 at screening and baseline visits.
Key exclusion criteria: inadequate response (< 50% reduction in depressive symptom severity) to > 3 monotherapy antidepressant treatments during the current MDE; received ECT for current MDE; inadequate response to ECT; current Axis I (DSM‐IV‐TR) diagnosis; current Axis II (DSM‐IV‐TR) diagnosis of borderline, antisocial, paranoid, schizoid, schizotypal or histrionic personality disorder.
Interventions
  • Brexpiprazole 2 mg + antidepressant, 192 participants

  • Placebo + antidepressant, 202 participants


Antidepressants: escitalopram tablets 10 mg/day or 20 mg/day: 35 (18.2%); fluoxetine capsules 20 mg/day or 40 mg/day: 38 (19.8%); paroxetine CR 25 mg/day, 37.5 mg/day or 50 mg/day: 29 (15.1%); sertraline tablets 50 mg/day, 100 mg/day, 150 mg/day or 200 mg/day: 24 (12.5%); duloxetine delayed‐release capsules 30 mg/day, 40 mg/day or 60 mg/day: 27 (14.1%); venlafaxine XR capsules 37.5 mg/day, 75 mg/day, 150 mg/day or 225 mg/day: 39 (20.3%)
Outcomes Primary outcome: change in MADRS at week 6
Key secondary outcomes: change from end of Phase A to end of Phase B in MADRS Total score for the subpopulations with anxious distress; change from baseline on CGI‐S, HAM‐D17 scores, MADRS – Response, CGI‐I – Response; remission as MADRS ≥ 50% reduction from baseline and MARDS score ≤ 10; acceptability; tolerability; safety; change in SDS score.
Notes Brexpiprazole + antidepressant: number of prior antidepressant failures: 1 antidepressant: 157 (81.8%) participants; 2 antidepressants: 31 (16.1%) participants; 3 antidepressants: 4 (2.1%) participants.
Placebo + antidepressant: number of prior antidepressant failures: 1 antidepressant: 162 (80.2%) participants; 2 antidepressants: 35 (17.3%) participants; 3 antidepressants: 5 (2.5%): participants.
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Low risk Quote: "Treatments were assigned by the IWRS based fixed‐block, computer‐generated randomisation code provided by the study sponsor, stratified by study center."
Allocation concealment (selection bias) Low risk Quote: "Treatment assignments were blinded to patients, investigators, and sponsor personnel, including those involved in the analysis."
Blinding of participants and personnel (performance bias)
All outcomes Low risk Quote: "Brexpiprazole and matching placebo tablets were provided by the sponsor, packaged in numbered, weekly blister cards."
Quote: "Masking: quadruple (participant, care provider, investigator, outcomes assessor)."
Blinding of outcome assessment (detection bias)
All outcomes Low risk Quote: "Brexpiprazole and matching placebo tablets were provided by the sponsor, packaged in numbered, weekly blister cards."
Quote: "Masking: quadruple (participant, care provider, investigator, outcomes assessor)."
Incomplete outcome data (attrition bias)
All outcomes Low risk Low reported attrition rates according to protocol criteria (21/394 (5.3%) participants). Slightly unbalanced between groups (6/202 participants in placebo arm and 15/192 in brexpiprazole arm). Attrition bias was unlikely.
Selective reporting (reporting bias) Low risk Outcomes reported as listed in protocol.
Other bias Unclear risk No information on other bias available.

NCT00797966.

Study characteristics
Methods Phase 2, multicentre, randomised, double‐blind, placebo‐controlled study; parallel design, flexible dose
Participants Men or women aged 18–65 years
Key inclusion criteria: DSM‐IV diagnosis of MDD, current depressive episode ≥ 8 weeks in duration, history for the current episode of an inadequate response to ≥ 1 and < 3 adequate antidepressant treatments.
Key exclusion criteria: breastfeeding; positive pregnancy test result prior to receiving study drug; inadequate response to ≥ 3 adequate trials of antidepressant treatments during current depressive episode at a therapeutic dose for an adequate duration; current Axis I (DSM‐IV‐TR) diagnosis; clinically significant current Axis II (DSM‐IV‐TR) diagnosis of borderline, antisocial, paranoid, schizoid, schizotypal or histrionic personality disorder.
Interventions Intervention: brexpiprazole + antidepressant flexible dose, 3 intervention groups
  • Brexpiprazole 0.15 mg + antidepressant, 62 participants

  • Brexpiprazole 0.5 mg ± 0.25 mg + antidepressant, 120 participants

  • Brexpiprazole 1.5 mg ± 0.5 mg + antidepressant, 121 participants


Comparison
  • Placebo + antidepressant, 126 participants

Outcomes Primary outcome: change in score from baseline in MADRS to week 6 of randomisation phase.
Secondary outcomes: change in score from baseline CGI‐S, CGI‐I, HAM‐D17, IDS‐SR (reported for each week for MADRS, IDS‐SR, CGI‐I; reported for week 6 HAM‐D17), response rate (MADRS); remission: MADRS ≤ 10 and ≥ 50% reduction from baseline; acceptability; tolerability; safety; change from end of Phase A (week 8 visit) to end of Phase B (week 14 visit) in SDS mean score.
Notes 2 arms of this study were not included in the review: brexpiprazole 0.15 mg + antidepressant clearly did not meet the inclusion criteria for dosage as defined in the protocol for the review; for the brexpiprazole 0.5 mg ± 0.25 mg, the mean dosage of brexpiprazole was not reported in the study report. We tried to contact the study authors for the information, but received no response. We decided on a more conservative approach and excluded the arm from the review due to lack of information whether the mean dosage of brexpiprazole was > 0.5 mg.
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Low risk Quote: "Allocation: randomised."
Allocation concealment (selection bias) Unclear risk Not reported.
Blinding of participants and personnel (performance bias)
All outcomes Low risk Quote: "Masking: triple (participant, investigator, outcomes assessor)."
Blinding of outcome assessment (detection bias)
All outcomes Low risk Quote: "Masking: triple (participant, investigator, outcomes assessor)."
Incomplete outcome data (attrition bias)
All outcomes Low risk Moderately low attrition rate (66/429 (15.4%) participants). Fairly balanced between groups (11/62 (17.7%) in the brexpiprazole 0.15 mg + antidepressant group; 18/120 (15%) in the brexpiprazole 0.5 mg ± 0.25 mg + antidepressant group; 21/121 (17.36%) in the brexpiprazole 1.5 mg ± 0.5 mg + antidepressant group; 16/126 (12.70%) in the placebo + antidepressant group). Attrition bias was unlikely.
Selective reporting (reporting bias) Low risk Outcomes reported as listed in protocol.
Other bias Unclear risk No information on other bias available.

NCT01052077.

Study characteristics
Methods Phase 2, multicentre, randomised, double‐blind, placebo‐controlled study, flexible dose
Participants Men and women aged 18–65 years
Key inclusion criteria: MDD as defined by DSM‐IV‐TR criteria; current depressive episode ≥ 8 weeks in duration; history for the current depressive episode of an inadequate response to ≥ 1 and < 3 adequate antidepressant treatments; participants randomised after a prospective phase if they had an incomplete response.
Key exclusion criteria: breastfeeding; positive pregnancy test result prior to receiving study drug; inadequate response to > 3 adequate trials of antidepressant treatments during current depressive episode at a therapeutic dose for an adequate duration; current Axis I (DSM‐IV‐TR) diagnosis of: delirium, dementia, amnestic or other cognitive disorder, schizophrenia, schizoaffective disorder, or other psychotic disorder, bipolar I or II disorder; clinically significant current Axis II (DSM‐IV‐TR) diagnosis of borderline, antisocial, paranoid, schizoid, schizotypal or histrionic personality disorder.
Interventions
  • Brexpiprazole 1–3 mg + antidepressant, 185 participants

  • Placebo + antidepressant, 187 participants


Antidepressants: an FDA‐approved antidepressant
Outcomes Primary outcome: change from the end of Phase A (week 8 visit) to the end of Phase B (week 14 visit) in MADRS.
Key secondary outcomes: change from end of Phase A (week 8 visit) to Phase B by study week in CGI‐S score; IDS‐SR Total score; HAM‐D17 Total score; number of participants with MADRS Response during Phase B relative to end of Phase A (week 8) visit; number of participants with CGI‐I Response during phase B relative to end of Phase A (week 8); number of participants with MADRS Remission during Phase B relative to end of Phase A (week 8) visit; acceptability; tolerability (at week 14); safety; change from end of Phase A (week 8) to Phase B in SDS score.
Notes  
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Low risk Quote: "Allocation: randomised."
Allocation concealment (selection bias) Unclear risk No details provided.
Blinding of participants and personnel (performance bias)
All outcomes Low risk Quote: "Masking: triple (participant, care provider, investigator)."
Blinding of outcome assessment (detection bias)
All outcomes Unclear risk Quote: "Masking: triple (participant, care provider, investigator)."
Incomplete outcome data (attrition bias)
All outcomes Low risk Low reported attrition rate according to protocol criteria (32/372 (8.6%) participants). Balanced attrition in both groups: 18/185 (9.7%) in brexpiprazole group, 16/187 (8.6%) in placebo group. Attrition bias was unlikely.
Selective reporting (reporting bias) Low risk Outcomes reported as listed in protocol.
Other bias Unclear risk No information on other bias available.

NCT01837797.

Study characteristics
Methods Phase 3, interventional, randomised, double‐blind, parallel‐group, placebo‐controlled, fixed‐dose
Participants Outpatients; men or women aged ≥ 65 years
Key inclusion criteria: recurrent MDD diagnosed according to DSM‐IV‐TR; current MDE should be confirmed using the MINI; had ≥ 1 previous MDE before age of 60 years; moderate‐to‐severe depression and an insufficient response to ≥ 1 and < 3 adequate antidepressants treatments; women must have had their last natural menstruation ≥ 24 months prior to screening visit; men must agree to protocol‐defined use of effective contraception if his female partner is of childbearing potential.
Key exclusion criteria: current psychiatric disorder or Axis I disorder (DSM‐IV‐TR criteria), established as the primary diagnosis, other than MDD; current Axis II (DSM‐IV‐TR) diagnosis of borderline, antisocial, paranoid, schizoid, schizotypal or histrionic personality disorder.
Interventions
  • Brexpiprazole 1 mg + antidepressant, 3 participants

  • Brexpiprazole 3 mg + antidepressant, 6 participants

  • Placebo + antidepressant, 6 participants


Antidepressants: duloxetine 60 mg/day, capsules, orally; escitalopram 5 mg/day or 10 mg/day, tablets, orally; fluoxetine 20 mg/day or 40 mg/day, capsules, orally; paroxetine 20 mg/day, 30 mg/day or 40 mg/day, orally, tablets; sertraline 50 mg/day, 100 mg/day or 150 mg/day; venlafaxine 75 mg/day, 150 mg/day or 225 mg/day, capsules, orally
Outcomes Primary outcome: change from randomisation in depressive symptoms on MADRS during the randomised treatment.
Secondary outcomes: change from randomisation in CGI‐S score; response during randomised treatment; remission (MADRS); acceptability; tolerability, safety.
Notes  
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Low risk Quote: "Allocation: randomised."
Allocation concealment (selection bias) Unclear risk No details provided.
Blinding of participants and personnel (performance bias)
All outcomes Low risk Quote: "Masking: double (participant, investigator)."
Blinding of outcome assessment (detection bias)
All outcomes Unclear risk No details provided.
Incomplete outcome data (attrition bias)
All outcomes Unclear risk No details provided.
Selective reporting (reporting bias) High risk Only safety results reported, no efficacy results provided.
Other bias High risk Quote: "The study was prematurely terminated and no firm conclusions can be drawn regarding safety and efficacy."

NCT03149991.

Study characteristics
Methods Phase 4, double‐blind, randomised, placebo‐controlled study, fixed‐flexible dose escalation
Participants Inpatients; men or women aged 18–65 years
Key inclusion criteria: diagnosed with MDD, single or recurrent, and currently experiencing an MDE of ≥ 8 weeks in duration, prior to screening; history of treatment‐resistant depression during the current MDE, as assessed by investigator and remote centralised rater using the Massachusetts General Hospital ATRQ. Treatment‐resistant depression defined as failure to achieve a satisfactory response (e.g. < 50% improvement of depression symptoms), as perceived by participant, to ≥ 2 "treatment courses" during current episode of a therapeutic dose of an antidepressant therapy of ≥ 8 weeks' duration (including the current antidepressant); MADRS Total score > 20 at both screening visit and baseline visit (day –7/–28 and day 0); in good general health, as ascertained by medical history, physical examination (including measurement of supine and standing vital signs), clinical laboratory evaluations.
Key exclusion criteria: history during the current MDE of failure to achieve satisfactory response (e.g. < 50% improvement of depression symptoms) to > 7 treatment courses of a therapeutic dose of an antidepressant therapy of ≥ 8 weeks' duration; MADRS Total score < 20 at screening visit or baseline visit; current Axis I disorder, diagnosed at screening using SCID‐5 AXIS I Disorders; history of bipolar disorder, schizophrenia or schizoaffective disorders, or any history of psychotic symptoms in the current or previous depressive episodes; history of anorexia nervosa, bulimia nervosa or eating disorder not otherwise specified, within 5 years of screening; any Axis I or Axis II disorder that at screening was clinically predominant to their MDD or had been predominant to their MDD at any time within 6 months prior to screening.
Interventions
  • Brexpiprazole (up to 3 mg) + ketamine + antidepressant: 25 participants

  • Placebo + ketamine + antidepressant: 26 participants


Ketamine: biweekly administration of intranasal ketamine 40 mg for 1 week, followed by weekly administration of intranasal ketamine 40 mg for 2 weeks
Antidepressant: standard therapy (oral treatment)
Outcomes Primary outcome: change from baseline on SDQ assessed on days 0, 1, 2, 5, 8, 11, 14, 17, 21, 23 and 28.
Key secondary outcomes: long‐term sustained response, as measured by achieving a 50% reduction on the MADRS on day 28; efficacy on secondary outcome variables; acceptability; tolerability; safety.
Notes  
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Low risk Quote: "The randomisation schedule for blinding of randomised treatment will be maintained by MGH CTNI or representative and will not be disclosed until after database lock."
Quote: "Allocation: randomised."
Allocation concealment (selection bias) Low risk Quote: "Individual treatment codes, indicating the treatment randomisation for each randomized patient, will be available to the investigator(s) or pharmacists from MGH CTNI or its designee.
The treatment code must not be broken except in medical emergencies when the appropriate management of the patient necessitates knowledge of the treatment randomisation. The investigator documents and reports the action to MGH CTNI or representative, without revealing the treatment given to the patient to MGH CTNI or representative's staff."
Blinding of participants and personnel (performance bias)
All outcomes Low risk Quote: "The investigator, patient, and study staff will be blinded. Packaging and labelling of the study drugs will be performed in a way to ensure blinding throughout the study."
Blinding of outcome assessment (detection bias)
All outcomes Unclear risk Quote: "The investigator, patient, and study staff will be blinded. Packaging and labelling of the study drugs will be performed in a way to ensure blinding throughout the study."
Incomplete outcome data (attrition bias)
All outcomes Unclear risk Low attrition rate according to protocol criteria (6/51 (11.8%) participants). Unbalanced attrition rate in both study groups (4/25 (16%) in brexpiprazole group; 2/26 (7.69%) in placebo group). Attrition bias risk was likely.
Selective reporting (reporting bias) Low risk Outcomes reported as stated in study protocol.
Other bias Unclear risk No information on other bias available.

Thase 2015a.

Study characteristics
Methods Phase 3, parallel, randomised, double‐blind, placebo‐controlled, fixed‐dose
Participants Outpatients, men or women aged 18–65 years
Key inclusion criteria: met DSM‐IV‐TR criteria for a single or recurrent non‐psychotic episode of MDD of ≥ 8 weeks' duration; during the current episode, must have reported inadequate response, defined as < 50% reduction in symptoms via participant self‐reports to an adequate trial of 1–3 antidepressants, including their most recent drug treatment; HAM‐D17 Total score ≥ 18 both at screening and on first day of prospective treatment; outpatients aged 18–65 years who met DSM‐IV‐TR criteria for a single or recurrent non‐psychotic episode of MDD of ≥ 8 weeks' duration; during the current episode, must have reported an inadequate response, defined as < 50% reduction in symptoms via participant self‐reports on the Massachusetts General Hospital ATRQ, to an adequate trial of 1–3 antidepressants, including their most recent drug treatment; HAM‐D17 Total score ≥ 18 both at screening and on the first day of prospective treatment.
Inclusion criteria after week 8: (protocol amendment): < 50% reduction in HAM‐D17 Total score between baseline of Phase A and the week 8 visit; HAM‐D17 Total score ≥ 14 at week 8 visit; < 50% reduction in MADRS Total score between baseline of Phase A and scheduled visits at weeks 2, 4, 6 and 8; and CGI‐I score of ≥ 3 at scheduled visits at weeks 2, 4, 6 and 8.
Key exclusion criteria: inadequate response to > 3 adequate trials of antidepressant treatments during current depressive episode at a therapeutic dose for an adequate duration; current Axis I (DSM‐IV‐TR) diagnosis; clinically significant current Axis II (DSM‐IV‐TR) diagnosis of borderline, antisocial, paranoid, schizoid, schizotypal or histrionic personality disorder.
Interventions
  • Brexpiprazole 2 mg + antidepressant, 188 participants

  • Placebo + antidepressant, 191 participants


Antidepressants: escitalopram (10–20 mg/day), fluoxetine (20–40 mg/day), paroxetine CR (37.5–50 mg/day), sertraline (100 mg/day, 150 mg/day or 200 mg/day), duloxetine (40–60 mg/day), venlafaxine XR (75 mg/day, 150 mg/day or 225 mg/day)
Outcomes Primary outcome: change from baseline (end of Phase A (week 8)) to week 14 in MADRS.
Key secondary outcomes: mean CGI‐I scale score (end of Phase A (week 8)) to week 14; change from baseline (end of Phase A (week 8)) to week 14 in CGI‐S; change from baseline (end of Phase A (week 8)) to week 14 in the IDS‐SR; change from baseline (end of Phase A (week 8)) to week 14 in HAM‐D; change from baseline (end of Phase A (week 8)) to week 14 in HAM‐A Total score; percentage of participants with MADRS response at week 14 relative to baseline; percentage of participants with CGI‐I scale response rate at week 14 relative to baseline; percentage of participants with MADRS Remission at week 14 relative to baseline; acceptability; tolerability; safety; mean change from baseline to week 14 in SDS score; change from baseline (end of Phase A (week 8)) in SDS item scores.
Notes  
Risk of bias
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Low risk Quote: "Randomisation was conducted via an interactive voice or web response system using a fixed‐block, computer‐generated randomization schedule with a block size of 4 and stratified by study center."
Allocation concealment (selection bias) Unclear risk No details provided.
Blinding of participants and personnel (performance bias)
All outcomes Low risk Quote: "Masking: triple (participant, care provider, investigator)."
Blinding of outcome assessment (detection bias)
All outcomes Unclear risk Quote: "Masking: triple (participant, care provider, investigator)."
Incomplete outcome data (attrition bias)
All outcomes Low risk Low attrition rate according to study protocol (27/379 (7.12%) of all participants); balanced attrition in both groups (14/188 (7.44%) in brexpiprazole group; 13/191 (6.81%) of placebo group). Attrition bias was unlikely.
Selective reporting (reporting bias) Low risk Outcomes reported as listed in protocol.
Other bias Unclear risk No information on other bias available.

Thase 2015b.

Study characteristics
Methods Phase 3, parallel, randomised, double‐blind, placebo‐controlled, multicentre, fixed dose
Participants Outpatients; men and women aged 18–65 years
Key inclusion criteria: diagnosis of MDD and in a current MDE, as defined by DSM‐IV‐TR criteria and confirmed by both the MINI and an adequate clinical psychiatric evaluation; current MDE must be ≥ 8 weeks in duration; history for current MDE of an inadequate response to ≥ 1 and < 3 adequate antidepressant treatments; if participant showed ≥ 50% improvement on any antidepressant treatment in current episode, then they must have had an inadequate response to a subsequent adequate antidepressant treatment of another antidepressant drug prior to entry into trial; for the most recent antidepressant treatment, participant must not report ≥ 50% improvement; HAM‐D17 Total score ≥ 18 at screening and baseline visits.
For inclusion in randomisation phase: HAM‐D17 score ≥ 14; < 50% reduction from baseline in HAM‐D17; < 50% reduction in MADRS Total score between start of prospective treatment and each scheduled visit; CGI‐I score ≥ 3 at each scheduled visit to be eligible for randomisation.
Key exclusion criteria: inadequate response (< 50% reduction in depressive symptom severity) to > 3 monotherapy antidepressant treatments during current MDE, ECT during current episode or inadequate response to ECT, current Axis I (DSM‐IV‐TR) diagnosis; current Axis II (DSM‐IV‐TR) diagnosis of borderline, antisocial, paranoid, schizoid, schizotypal or histrionic personality disorder.
Interventions
  • Brexpiprazole 1 mg + antidepressant, 226 participants

  • Brexpiprazole 3 mg + antidepressant, 230 participants

  • Placebo + antidepressant, 221 participants

Outcomes Primary outcome: mean change from the end of Phase A (week 8 visit) to Phase B (week 14 visit).
Secondary outcomes: mean change from end of Phase A (week 8 visit) to Phase B by study week in CGI‐S; IDS‐SR Total score; HAM‐D17 Total score; HAM‐A; CGI‐I score; percentage of participants with a MADRS response during Phase B Relative to end of Phase A; CGI‐I Response during Phase B relative to end of Phase A (week 8 visit); percentage of participants with a MADRS Remission during Phase B relative to end of Phase A (week 8); acceptability; tolerability (at 8 weeks); safety; mean change from end of Phase A (week 8 visit) to end of Phase B (week 14 visit) in SDS mean scores; change from baseline (end of Phase A (week 8)) in SDS Item scores.
Notes
  • Brexpiprazole 1 mg + antidepressant number of prior antidepressant failures: 1 antidepressant: 177 (78.7%) participants, 2 antidepressants: 42 (18.7%) participants; 3 antidepressants: 6 (2.7%) participants.

  • Brexpiprazole 1 mg + antidepressant number of prior antidepressant failures: 1 antidepressant: 184 (81.4%) participants, 2 antidepressants: 34 (15.0%) participants; 3 antidepressants: 7 (3.1%) participants.

  • Placebo + antidepressant number of prior antidepressant failures: 1 antidepressant: 170 (78.0%) participants, 2 antidepressants: 44 (20.2%) participants; 3 antidepressants: 4 (1.8%) participants.

Risk of bias
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Low risk Quote: "An interactive voice or web response system was used for assigning treatments using a fixed‐block, computer‐generated randomized schedule stratified by study center."
Allocation concealment (selection bias) Unclear risk No details provided.
Blinding of participants and personnel (performance bias)
All outcomes Low risk Quote: "Masking: triple (participant, care provider, investigator)."
Blinding of outcome assessment (detection bias)
All outcomes Unclear risk Quote: "Masking: triple (participant, care provider, investigator)."
Incomplete outcome data (attrition bias)
All outcomes Low risk Low reported attrition rate (43/677 (6.35%) of all participants). Balanced attrition rate in study arms (10/226 (4.42%) of brexpiprazole 1 mg group; 10/230 (4.35%) of brexpiprazole 3 mg group; 13/221 (5.88%) of placebo group). Attrition bias was unlikely.
Selective reporting (reporting bias) Low risk Outcomes reported as listed in protocol.
Other bias Unclear risk No information on other bias available.

ATRQ: Antidepressant Treatment Response Questionnaire; CGI‐I: Clinical Global Impression – Improvement; CGI‐S: Clinical Global Impression – Severity; CR: controlled release; DSM‐IV: Diagnostic and Statistical Manual of Mental Disorders 4th Edition; DSM‐IV‐TR: Diagnostic and Statistical Manual of Mental Disorders 4th Edition Text Revision; ECT: electroconvulsive therapy; FDA: Food and Drug Administration; HAM‐A: Hamilton Anxiety Rating Scale; HAM‐D: Hamilton Depression Rating Scale; HAM‐D17: Hamilton Depression Rating Scale 17‐item version; IDS‐SR: Inventory of Depressive Symptomatology – Self‐Report; IR: immediate release; MADRS: Montgomery–Åsberg Depression Rating Scale; MDD: major depressive disorder; MDE: major depressive episode; MINI: Mini‐International Neuropsychiatric Interview; SCID‐5: Structured Clinical Interview for DSM‐5; SDQ: Symptoms of Depression Questionnaire; SDS: Sheehan Disability Scale; XR: extended‐release.

Characteristics of excluded studies [ordered by study ID]

Study Reason for exclusion
NCT01052077b A duplicate of an included study (NCT01052077)

Characteristics of studies awaiting classification [ordered by study ID]

NCT01670279.

Methods Phase 1, multicentre, randomised, double‐blind, sequential cohort, placebo‐controlled trial, ascending dose
Participants Men or women aged 70–85 years
Key inclusion criteria: normal or clinically stable findings on physical examination, medical history, clinical laboratory determinations, ECGs in relation to age, people with stable psychiatric diagnosis of MDD.
Key exclusion criteria: vagus nerve stimulation device implanted or who have received ECT within 6 months of screening; current Axis I (DSM‐IV‐TR) diagnosis, clinically significant current Axis II (DSM‐IV‐TR) diagnosis of borderline, antisocial, paranoid, schizoid, schizotypal or histrionic personality disorders; active suicidal ideation with some intent to act and who's most recent episode occurred within last 6 months; have met DSM‐IV‐TR criteria for substance abuse or dependence within past 180 days; hypothyroidism or hyperthyroidism or an abnormal result for free T4 at screening; currently have clinically significant neurological, hepatic, renal, metabolic, haematological, immunological, cardiovascular, pulmonary or gastrointestinal disorders; insulin‐dependent diabetes mellitus; uncontrolled hypertension (diastolic blood pressure > 95 mmHg) or symptomatic hypotension; epilepsy, a history of epilepsy, or a history of seizures; positive drug screen for cocaine or other drugs of abuse.
Interventions
  • Brexpiprazole up to 3 mg (cohort 1) + antidepressant, 6 participants

  • Brexpiprazole up to 3 mg (cohort 2) + antidepressant, 7 participants

  • Placebo + antidepressant, 5 participants

Outcomes Primary outcomes: number of participants who tolerated brexpiprazole; number of adverse effects reported; incidence of laboratory values of potential clinical significance; incidence of vital signs of potential clinical significance; incidence of vital signs of potential clinical significance; incidence of physical examination evaluation of potential clinical significance; mean change from baseline to study completion in Simpson‐Angus Scale Total score; mean change from baseline to study completion in Barnes Akathisia Global Score; mean change from baseline to study completion in Abnormal Involuntary Movement Scale rating score.
Notes  

NCT03487198.

Methods Phase 3, multicentre, randomised, double‐blind, placebo‐controlled trial, flexible dose
Participants Outpatients; men or women aged 18–65 years
Key inclusion criteria: diagnosis of MDD, and in a current MDE, according to the diagnostic criteria in DSM‐IV‐TR and confirmed using the MINI; the current MDE must be ≥ 8 weeks in duration; history for the current MDE of an inadequate response to ≥ 1 and < 3 standard antidepressant treatments; for the most recent antidepressant treatment, the participant must not report ≥ 50% improvement; standard treatment defined as: an antidepressant treatment for ≥ 6 weeks in duration (≥ 3 weeks if combined treatment) at a minimum effective dose (or higher) according to prescription drug labels; ≥ once of 1–3 standard treatments is monotherapy for > 6 weeks; inadequate response defined as: < 50% reduction in depressive symptom severity, as measured by self‐report score on Visual Analogue Scale; HAM‐D17 Total score ≥ 18 at screening.
Key exclusion criteria: ECT treatment history at any time, or a vagus nerve stimulation or deep brain stimulation device implanted for management of treatment‐resistant depression; current need for hospitalisation or hospitalisation within 4 weeks prior to screening for current MDE; current Axis I (DSM‐IV‐TR) diagnosis; current Axis II (DSM‐IV‐TR) diagnosis of borderline, antisocial, paranoid, schizoid, schizotypal or histrionic personality disorder; receiving new‐onset psychotherapy; present a serious risk of suicide; medical history or clinical evidence shows the health condition may cause an undue adverse event or interfere with assessments of safety/efficacy during course of trial.
Interventions
  • Brexpiprazole 2–3 mg + antidepressant

  • Placebo + antidepressant

Outcomes Primary outcome: change from the end of Phase A (week 8 visit) to the end of Phase B (week 14 visit) in the MADRS Total score.
Key secondary outcomes: change from the end of Phase A (week 8 visit) in MADRS Total score, CGI‐S, CGI‐I score for every trial week visit in Phase B other than week 14 visit; change from the end of Phase A (week 8 visit) to the end of Phase B (week 14 visit) in HAM‐D17 Score; QIDS‐SR16 score at week 8 and week 14; MADRS Response Rate, CGI‐S Response Rate; QIDS‐SR16 Response Rate at every trial week visit in Phase B; QIDS‐SR16 Complete Remission Rate (recovery) at every trial week visit in Phase B.
Notes No study results posted.

DSM‐IV‐TR: Diagnostic and Statistical Manual of Mental Disorders 4th Edition Text Revision; ECG: electrocardiogram; ECT: electroconvulsive therapy; HAM‐D17: Hamilton Depression Rating Scale 17‐item version; MADRS: Montgomery–Åsberg Depression Rating Scale; MDD: major depressive disorder; MDE: major depressive episode; MINI: Mini‐International Neuropsychiatric Interview; QIDS‐SR16: Quick Inventory of Depressive Symptomatology Self‐Report 16 item.

Characteristics of ongoing studies [ordered by study ID]

jRCT2080223986.

Study name A multi‐center, randomized, double‐blind, placebo‐controlled, parallel group‐comparison trial to assess the efficacy and safety of brexpiprazole as adjunctive therapy in patients with major depressive disorder
Methods Phase 2–3, multicentre, randomised, double‐blind, placebo‐controlled, parallel group‐comparison, fixed dose
Participants Outpatients, or inpatients at the time of informed consent whose treatment status can be successfully shifted to outpatient status before enrolment in the antidepressant treatment period; men and women aged 20–64 years (at the time of informed consent).
Key inclusion criteria: level of comprehension sufficient to allow written informed consent to all observation, examination, evaluation items specified in protocol, and understand the content of trial; DSM‐5 classification‐based diagnosis of "major depressive disorder, single episode" or "major depressive disorder, recurrent episode," and whose current episode has persisted ≥ 8 weeks.
Key exclusion criteria: treatment history showing that all antidepressants (including those not used for the current MDE) cannot be tolerated; history of ECT; diagnosis of any of the following diseases according to DSM‐5: neurocognitive disorders; schizophrenia spectrum and other psychotic disorders; bipolar and related disorders; feeding and eating disorders; obsessive‐compulsive disorder; panic disorder; posttraumatic stress disorder.
Interventions
  • Brexpiprazole 1 mg + antidepressant

  • Brexpiprazole 2 mg + antidepressant

  • Placebo + antidepressant

Outcomes Primary outcome: MADRS
Secondary outcomes: CGI‐I, CGI‐S, HAM‐D, SDS, MADRS‐S
Starting date 25 July 2018
Contact information Otsuka Pharmaceutical Co., Ltd. Drug Information Center, Tel: +81‐3‐6361‐7314
Notes  

NCT03538691.

Study name A trial to evaluate the efficacy, safety & tolerability of brexpiprazole in the maintenance treatment of adults with major depressive disorder
Methods Phase 3, multicentre, randomised, double‐blind, placebo‐controlled trial, flexible dose
Participants Men or women aged 18–65 years
Key inclusion criteria: both a diagnosis of recurrent MDD and in a current MDE ≥ 8 weeks in duration, as defined by DSM‐5 and confirmed by both MINI and an adequate clinical psychiatric evaluation; history for the current MDE of an inadequate response to 1 or 2 adequate antidepressant treatments, and must currently be taking a protocol‐mandated antidepressant treatment at an adequate dose and duration, and most not have reported ≥ 50% improvement; HAM‐D17 Total score ≥ 18 at screening visit, and Phase A baseline visits.
Key exclusion criteria: inadequate response to ECT at any time or who have had a vagus nerve stimulation or deep brain stimulation device implanted at any time for management of treatment‐resistant depression; who have had transcranial magnetic stimulation during current MDE; primary DSM‐5 diagnosis of schizophrenia spectrum and other psychotic disorders, bipolar and related disorders, obsessive compulsive disorders, feeding and eating disorders, neurocognitive disorders, panic disorder, post‐traumatic stress disorder; current DSM‐5 diagnosis of borderline, antisocial, paranoid, schizoid, schizotypal, or histrionic personality disorder or intellectual disability.
Interventions
  • Brexpiprazole 2–3 mg + antidepressant

  • Placebo + antidepressant


Antidepressants: citalopram hydrobromide 20 mg/day or 40 mg/day; escitalopram 10 mg/day or 20 mg/day; fluoxetine 20 mg/day or 40 mg/day; paroxetine, oral CR tablet 37.5 mg/day or 50 mg/day; sertraline 100 mg/day, 150 mg/day or 200 mg/day; duloxetine, oral delayed‐release capsule 40 mg/day or 60 mg/day; venlafaxine XR: daily XR capsule; 75 mg/day, 150 mg/day or 225 mg/day
Outcomes Primary outcome: time to relapse by any criteria as defined in blinded addendum.
Key secondary outcomes: percentage of participants who relapse, percentage of participants maintaining remission, mean change from randomisation at week 20 in MADRS Total score at week 46, mean change from randomisation at week 20 in CGI‐S score at week 46; mean change from randomisation at week 20 in each of the SDS individual item scores at week 46; change from randomisation at week 20 in SDS mean Total score at week 46; time to functional relapse based on SDS criteria.
Starting date 13 July 2018
Contact information Otsuka Pharmaceutical Development & Commercialization, Inc.
Notes  

NCT03697603.

Study name A study of brexpiprazole in patients with major depressive disorder
Methods Phase 2–3, multicentre, randomised, double‐blind, placebo‐controlled, parallel group, comparison trial, fixed dose
Participants Outpatients, or inpatients at time of informed consent whose treatment status can be successfully shifted to outpatient status before enrolment in antidepressant treatment period; men and women aged 20–65 years (at the time of informed consent).
Key inclusion criteria: level of comprehension sufficient to allow them to give written informed consent to all of the observation, examination, evaluation items specified in protocol, and understand contents of trial; DSM‐5 classification‐based diagnosis of "major depressive disorder, single episode" or "major depressive disorder, recurrent episode," and whose current episode has persisted for ≥ 8 weeks.
Key exclusion criteria: treatment history showing that all antidepressants (including those not used for the current MDE) cannot be tolerated; history of ECT; diagnosis of any of the following diseases according to DSM‐5: neurocognitive disorders, schizophrenia spectrum and other psychotic disorders, bipolar and related disorders, feeding and eating disorders, obsessive‐compulsive disorder, panic disorder, post‐traumatic stress disorder.
Interventions
  • Brexpiprazole 1 mg + antidepressant

  • Brexpiprazole 2 mg + antidepressant

  • Placebo + antidepressant

Outcomes Primary outcome: mean changes from baseline in MADRS Total scores at last visit of double‐blind period.
Secondary outcome: proportion of participants scoring 1 or 2 on the CGI‐I scale at week 14.
Starting date 30 July 2020
Contact information Otsuka Pharmaceutical Co., Ltd.
Notes  

CGI‐I: Clinical Global Impression – Global Improvement; CGI‐S: Clinical Global Impression – Severity; DSM‐5: Diagnostic and Statistical Manual of Mental Disorders 5th Edition; ECT: electroconvulsive therapy; HAM‐D: Hamilton Depression Rating Scale; MADRS: Montgomery–Åsberg Depression Rating Scale; MADRS‐S: Montgomery–Åsberg Depression Rating Scale – Self‐assessment; MDE: major depressive episode; SDS: Sheehan Disability Scale; XR: extended release.

Differences between protocol and review

We made the following changes from the review protocol (Ralovska 2021).

For continuous outcomes we initially planned to represent mean differences when studies used the same scale or standardised mean differences when studies used different scales. All but one study had used the method of least squares mean for the purpose of presenting change of scores and we decided that this measurement takes into account variables that could possibly impact the outcome (such as age, sex, etc.) and balances the groups.

Contributions of authors

AC conceived the review.

SR selected the studies, appraised their quality and extracted data.

SR entered the data into Review Manager 5 and carried out the analyses.

SR and AC drafted the manuscript.

All review authors critically reviewed the text.

Sources of support

Internal sources

  • University of Oxford, UK

    The University of Oxford has supported the study in terms of salary costs for some of the authors and in‐kind contributions (e.g. access to libraries and journals).

External sources

  • National Institute for Health Research (NIHR), UK

    AC is supported by the National Institute for Health Research (NIHR) Oxford Cognitive Health Clinical Research Facility, by an NIHR Research Professorship (grant RP‐2017‐08‐ST2‐006), by the NIHR Oxford and Thames Valley Applied Research Collaboration and by the NIHR Oxford Health Biomedical Research Centre (grant BRC‐1215‐20005).

  • NIHR Oxford Cognitive Health Clinical Research Facility, UK

    AC is supported by the NIHR Oxford Cognitive Health Clinical Research Facility

  • NIHR Oxford and Thames Valley Applied Research Collaboration, UK

    AC is supported by the NIHR Oxford and Thames Valley Applied Research Collaboration

  • NIHR Oxford Health Biomedical Research Centre, UK

    AC and FDC are supported by the NIHR Oxford Health Biomedical Research Centre (grant BRC‐1215‐20005)

Declarations of interest

SR: none.

IK declares funding through National Institute of Health Research (personal award), Oxford Health Biomedical Research Centre, Medical Research Council (Dementias Platform UK grant). He is a paid medical advisor receiving retainer fees or stock options (or both) to digital technology developing solutions for detecting or intervening (or both) in cognitive impairment (Five Lives SAS, Cognetivity Ltd and Mantrah Ltd). IK has received grant funding as well as speaker fees from Novo Nordisk.

PM has received consultancy fees from Angelini Pharma and Lundbeck LLC.

TAF reports personal fees from Kyoto University Original, grants and personal fees from Mitsubishi‐Tanabe, personal fees from SONY, grants and personal fees from Shionogi, outside the submitted work. In addition, TAF has patents 2020‐548587 and 2022‐082495 pending, and intellectual properties for Kokoro‐app licensed to Mitsubishi‐Tanabe.

BM holds and receives support from the Labatt Family Chair in Biology of Depression in Late‐Life Adults at the University of Toronto. He currently receives or has received with the past five years research support from Brain Canada, the Canadian Institutes of Health Research, the CAMH Foundation, the Patient‐Centered Outcomes Research Institute (PCORI), the US National Institute of Health (NIH), Capital Solution Design LLC (software used in a study founded by CAMH Foundation), Eli Lilly (medications for a NIH‐funded clinical trial), HAPPYneuron (software used in a study founded by Brain Canada), and Pfizer (medications for a NIH‐funded clinical trial). He has also been an unpaid consultant to Myriad Neuroscience.

AC has received research and consultancy fees from INCiPiT (Italian Network for Paediatric Trials), CARIPLO Foundation and Angelini Pharma.

New

References

References to studies included in this review

Bauer 2019 {published data only}

  1. 2012-001380-76. Interventional, randomised, double-blind, parallel-group, placebo-controlled, flexible-dose long-term study to evaluate the maintenance of efficacy and safety of 1 to 3 mg/day of brexpiprazole as adjunctive treatment in patients with major depressive disorder with an inadequate response to antidepressant treatment (ADT). www.clinicaltrialsregister.eu/ctr-search/trial/2012-001380-76/results (first received 28 May 2013).
  2. Bauer M, Hefting N, Lindsten A, Josiassen MK, Hobart M. A randomised, placebo-controlled 24-week study evaluating adjunctive brexpiprazole inpatients with major depressive disorder. Acta Neuropsychiatrica 2019;31:27-35. [DOI] [PubMed] [Google Scholar]
  3. NCT01838681. Brexpiprazole as adjunctive treatment in patients with major depressive disorder with an inadequate response to antidepressant treatment [Interventional, randomised, double-blind, parallel-group, placebo-controlled, flexible-dose long-term study to evaluate the maintenance of efficacy and safety of 1 to 3 mg/day of brexpiprazole as adjunctive treatment in patients with major depressive disorder with an inadequate response to antidepressant treatment]. clinicaltrials.gov/ct2/show/study/NCT01838681 (first received 24 April 2013).

Hobart 2018a {published data only}

  1. 2012-003948-67. A phase 3, multicenter, randomized, double-blind, placebo- and active comparator-controlled trial of flexible-dose brexpiprazole (OPC-34712) as adjunctive therapy in the treatment of adults with major depressive disorder, the Delphinus trial. www.clinicaltrialsregister.eu/ctr-search/trial/2012-003948-67/results (first received 26 November 2012).
  2. Hobart M, Skuban A, Zhang P, Josiassen MK, Hefting N, Augustine C, et al. Efficacy and safety of flexibly dosed brexpiprazole for the adjunctive treatment of major depressive disorder: a randomized, active-referenced, placebo-controlled study. Current Medical Research and Opinion 2018;34(4):633-42. [DOI] [PubMed] [Google Scholar]
  3. NCT01727726. A study of flexible-dose brexpiprazole as adjunctive therapy in the treatment of adults with major depressive disorder, the Delphinus trial [A phase 3, multicenter, randomized, double-blind, placebo and active comparator controlled trial of flexible-dose brexpiprazole (OPC-34712) as adjunctive therapy in the treatment of adults with major depressive disorder, the Delphinus trial]. clinicaltrials.gov/ct2/show/NCT01727726 (first received 16 November 2012).

Hobart 2018b {published data only}

  1. 2014-000062-22. A phase 3, multicenter, randomized, double-blind, placebo-controlled trial of the safety and efficacy of fixed-dose brexpiprazole (OPC-34712) as adjunctive therapy in the treatment of adults with major depressive disorder with and without anxious distress. www.clinicaltrialsregister.eu/ctr-search/trial/2014-000062-22/results (first received 8 May 2013).
  2. A phase 3, multicenter, randomized, double-blind, placebo-controlled trial of the safety and efficacy of fixed-dose brexpiprazole (OPC-34712) as adjunctive therapy in the treatment of adults with major depressive disorder with and without anxious distress. drks.de/search/en/trial/DRKS00007103 (first received 29 October 2014).
  3. DRKS00007103. A Phase 3, multicenter, randomized, double-blind, placebo-controlled trial of the safety and efficacy of fixed-dose brexpiprazole (OPC-34712) as adjunctive therapy in the treatment of adults with major depressive disorder with and without anxious distress. trialsearch.who.int/Trial2.aspx?TrialID=DRKS00007103 (first received 14 November 2014).
  4. Hobart M, Skuban A, Zhang P, Augustine C, Brewer C, Hefting N, et al. A randomized, placebo-controlled study of the efficacy and safety of fixed-dose brexpiprazole 2 mg/d as adjunctive treatment of adults with major depressive disorder. Journal of Clinical Psychiatry 2018;79:4. [DOI] [PubMed] [Google Scholar]
  5. Hobart M, Skuban A, Zhang P, Yang Y, Augustine C, Brewer C, et al. A randomized, double-blind, placebo-controlled study of brexpiprazole as adjunctive therapy in the treatment of adults with major depressive disorder. Neuropsychopharmacology 2016;41:S533. [Google Scholar]
  6. NCT02196506. Study of the safety and efficacy of fixed-dose brexpiprazole (OPC-34712) as adjunctive therapy in the treatment of adults with major depressive disorder with and without anxious distress [A Phase 3, multicenter, randomized, double-blind, placebo-controlled trial of the safety and efficacy of fixed-dose brexpiprazole (OPC-34712) as adjunctive therapy in the treatment of adults with major depressive disorder with and without anxious distress]. clinicaltrials.gov/ct2/show/NCT02196506 (first received 22 July 2014).

NCT00797966 {published data only}

  1. NCT00797966. Study of the safety and efficacy of OPC-34712 as adjunctive therapy in the treatment of patients with major depressive disorder [A phase 2, multicenter, randomized, double-blind, placebo-controlled study of the safety and efficacy of OPC-34712 as adjunctive therapy in the treatment of patients with major depressive disorder]. clinicaltrials.gov/ct2/show/NCT00797966 (first received 25 November 2008).
  2. Thase M, Fava M, Hobart M, Skuban A, Zhang P, McQuade R, et al. Efficacy of adjunctive OPC-34712 across multiple outcome measures in major depressive disorder: a phase II, randomized, placebo-controlled study. Neuropsychopharmacology 2011;36:S302-24. [Google Scholar]

NCT01052077 {published data only}

  1. NCT01052077. Study of the safety and efficacy of OPC-34712 as adjunctive therapy in the treatment of adults with major depressive disorder (STEP-D222) [A phase 2, multicenter, randomized, double-blind, placebo-controlled study of the safety and efficacy of OPDC-34712 (1 to 3 mg/Day) as adjunctive therapy in the treatment of adults with major depressive disorder]. clinicaltrials.gov/ct2/show/NCT01052077 (first received 20 January 2010).

NCT01837797 {published data only}

  1. 2012-001361-32. Interventional, randomised, double-blind, parallel-group, placebo-controlled, fixed-dose study to evaluate the efficacy and safety of brexpiprazole (1 and 3 mg/day) as adjunctive treatment in elderly patients with major depressive disorder with an inadequate response to antidepressant treatment. www.clinicaltrialsregister.eu/ctr-search/trial/2012-001361-32/results (first received 15 April 2013).
  2. NCT01837797. Efficacy and safety of brexpiprazole as adjunctive treatment in elderly patients with major depressive disorder with an inadequate response to antidepressant treatment [Interventional, randomised, double-blind, parallel-group, placebo-controlled, fixed-dose study to evaluate the efficacy and safety of brexpiprazole (1 and 3 mg/day) as adjunctive treatment in elderly patients with major depressive disorder with an inadequate response to antidepressant treatment]. clinicaltrials.gov/ct2/show/NCT01837797 (first received 23 April 2013).

NCT03149991 {published data only}

  1. NCT03149991. A study of brexpiprazole plus ketamine in treatment-resistant depression (TRD) [A double-blind, placebo-controlled study of brexpiprazole plus ketamine in treatment-resistant depression (TRD)]. clinicaltrials.gov/ct2/show/NCT03149991 (first received 24 February 2017).

Thase 2015a {published data only}

  1. 2011-001350-28. A phase 3, multicenter, randomized, double-blind, placebo-controlled trial of the safety and efficacy of fixed-dose OPC-34712 as adjunctive therapy in the treatment of adults with major depressive disorder, the Pyxis trial. www.clinicaltrialsregister.eu/ctr-search/trial/2011-001350-28/results (first received 28 September 2011).
  2. NCT01360645. Study of the safety and efficacy of fixed dose OPC-34712 as adjunctive therapy in the treatment of adults with major depressive disorder (the Pyxis Trial) (Pyxis) [A phase 3, multicenter, randomized, double-blind, placebo-controlled trial of the safety and efficacy of fixed dose OPC-34712 as adjunctive therapy in the treatment of adults with major depressive disorder, the Pyxis Trial]. clinicaltrials.gov/ct2/show/NCT01360645 (first received 25 May 2011).
  3. Thase M, Youakim J, Skuban A, Hobart M, Zhang P, Legacy S, et al. Efficacy and safety of adjunctive brexpiprazole (OPC-34712) in major depressive disorder: results: of two pivotal clinical studies. Neuropsychopharmacology 2014;39:S357. [Google Scholar]
  4. Thase ME, Youakim JM, Skuban A, Hobart M, Augustine C, Zhang P, et al. Efficacy and safety of adjunctive brexpiprazole 2 mg in major depressive disorder: a phase 3, randomized, placebo-controlled study in patients with inadequate response to antidepressants. Journal of Clinical Psychiatry 2015;76(9):1224-31. [DOI] [PubMed] [Google Scholar]

Thase 2015b {published data only}

  1. 2011-001349-33. A phase 3, multicenter, randomized, double-blind, placebo-controlled trial of the safety and efficacy of two fixed doses of OPC-34712 as adjunctive therapy in the treatment of adults with major depressive disorder, the Polaris trial. www.clinicaltrialsregister.eu/ctr-search/trial/2011-001349-33/results (first received 25 June 2011).
  2. NCT01360632. Study of the safety and efficacy of two fixed doses of OPC-34712 as adjunctive therapy in the treatment of adults with major depressive disorder (the Polaris trial) (Polaris) [A phase 3, multicenter, randomized, double-blind, placebo-controlled trial of the safety and efficacy of two fixed doses of OPDC-34712 as adjunctive therapy in the treatment of adults with major depressive disorder, the Polaris trial]. clinicaltrials.gov/ct2/show/NCT01360632 (first received 25 May 2011).
  3. Thase M, Youakim J, Skuban A, Hobart M, Zhang P, Legacy S, et al. Efficacy and safety of adjunctive brexpiprazole (OPC-34712) in major depressive disorder: results: of two pivotal clinical studies. Neuropsychopharmacology 2014;39:S357. [Google Scholar]
  4. Thase ME, Youakim JM, Skuban A, Hobart M, Zhang P, McQuade RD, et al. Adjunctive brexpiprazole 1 and 3 mg for patients with major depressive disorder following inadequate response to antidepressants: a phase 3, randomized, double-blind study. Journal of Clinical Psychiatry 2015;76(9):1232-40. [DOI] [PubMed] [Google Scholar]

References to studies excluded from this review

NCT01052077b {published data only}

  1. NCT01052077. Study of the safety and efficacy of OPC-34712 as adjunctive therapy in the treatment of adults with major depressive disorder (STEP-D222) [A phase 2, multicenter, randomized, double-blind, placebo-controlled study of the safety and efficacy of OPDC-34712 (1 to 3 mg/day) as adjunctive therapy in the treatment of adults with major depressive disorder]. clinicaltrials.gov/ct2/show/NCT01052077 (first received 10 January 2010).

References to studies awaiting assessment

NCT01670279 {published data only}

  1. NCT01670279. Phase 1 study to assess the safety/​tolerability of brexpiprazole as adjunctive therapy in elderly subjects with major depressive disorder [A phase 1, multicenter, randomized, double-blind, sequential cohort, placebo-controlled trial to assess the safety and tolerability of ascending multiple oral doses of brexpiprazole as adjunctive therapy in the treatment of elderly subjects with major depressive disorder]. www.clinicaltrials.gov/study/NCT01670279 (first received 7 August 2012).

NCT03487198 {published and unpublished data}

  1. NCT03487198. The safety and efficacy of brexpiprazole as adjunctive therapy in the treatment of major depressive disorder [A phase 3, multicenter, randomized, double-blind, placebo-controlled trial evaluating the safety and efficacy of brexpiprazole as adjunctive therapy in the treatment of major depressive disorder]. www.clinicaltrials.gov/study/NCT03487198 (first received 23 March 2018).

References to ongoing studies

jRCT2080223986 {published data only}

  1. jRCT2080223986. A study of brexpiprazole in patients with major depressive disorder [A multi-center, randomized, double-blind, placebo-controlled, parallel group-comparison trial to assess the efficacy and safety of brexpiprazole as adjunctive therapy in patients with major depressive disorder]. jrct.niph.go.jp/en-latest-detail/jRCT2080223986 (first received 25 July 2018).

NCT03538691 {published data only}

  1. NCT03538691. A trial to evaluate the efficacy, safety & tolerability of brexpiprazole in the maintenance treatment of adults with major depressive disorder [A phase 3, multicenter, randomized, double-blind, placebo-controlled trial to evaluate the efficacy, safety, and tolerability of brexpiprazole as adjunctive therapy in the maintenance treatment of adults with major depressive disorder]. www.clinicaltrials.gov/study/NCT03538691 (first received 13 April 2018).

NCT03697603 {published data only}

  1. NCT03697603. A study of brexpiprazole in patients with major depressive disorder [A multi-center, randomized, double-blind, placebo-controlled, parallel group-comparison trial to assess the efficacy and safety of brexpiprazole as adjunctive therapy in patients with major depressive disorder]. clinicaltrials.gov/study/NCT03697603 (first received 5 October 2018).

Additional references

APA 1994

  1. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders (DSM-IV). 4th edition. Washington (DC): American Psychiatric Association, 1994. [Google Scholar]

APA 2000

  1. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR) 4th Text Revised. Washington (DC): American Psychiatric Association, 2000. [Google Scholar]

APA 2010

  1. Gelenberg AJ, Freeman MP, Markowitz JC, Rosenbaum JF, Thase ME, Trivedi MJ, et al, Work Group on Major Depressive Disorder. American Psychiatric Association practice guideline for the treatment of major depressive disorder. 3rd Edition. psychiatryonline.org/pb/assets/raw/sitewide/practice_guidelines/guidelines/mdd-1410197717630.pdf (accessed 10 July 2023).

APA 2022

  1. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision (DSM-5-TR). Washington (DC): American Psychiatric Association Publishing, 2022. [Google Scholar]

Au‐Yeung 2021

  1. Au-Yeung SK, Griffiths J, Roberts S, Edwards C, Yu LM, Bogacz R, et al. PAX-D: study protocol for a randomised placebo-controlled trial evaluating the efficacy and mechanism of pramipexole as add-on treatment for people with treatment resistant depression. Evidence-Based Mental Health 2021;25(2):77-83. [DOI] [PMC free article] [PubMed] [Google Scholar]

Barbui 2004

  1. Barbui C, Cipriani A, Brambilla P, Hotopf M. "Wish bias" in antidepressant drug trials? Journal of Clinical Psychopharmacology 2004;24(2):126-30. [DOI] [PubMed] [Google Scholar]

Bartova 2019

  1. Bartova L, Dold M, Kautzky A, Fabbri C, Spies M, Serretti A, et al. Results of the European Group for the Study of Resistant Depression (GSRD) – basis for further research and clinical practice. World Journal of Biological Psychiatry 2019;20(6):427-48. [DOI] [PubMed] [Google Scholar]

Bauer 2002

  1. Bauer M, Whybrow PC, Angst J, Versiani M, Moller HJ, World Federation of Societies Biological Psychiatry Task Force on Treatment Guidelines for Unipolar Depressive Disorders. World Federation of Societies of Biological Psychiatry (WFSBP) guidelines for biological treatment of unipolar depressive disorders. Part 1. Acute and continuation treatment of major depressive disorder. World Journal of Biological Psychiatry 2002;3(1):5-43. [DOI] [PubMed] [Google Scholar]

Berman 2007

  1. Berman RM, Marcus RN, Swanink R, McQuade RD, Carson WH, Corey-Lisle PK, et al. The efficacy and safety of aripiprazole as adjunctive therapy in major depressive disorder: a multicenter, randomized, double-blind, placebo-controlled study. Journal of Clinical Psychiatry 2007;68(6):843-53. [DOI] [PubMed] [Google Scholar]

Beyer 2016

  1. Beyer JL, Weisler RH. Adjunctive brexpiprazole for the treatment of major depressive disorder. Expert Opinion in Pharmacotherapy 2016;17:2331-9. [DOI] [PubMed] [Google Scholar]

Boutron 2021

  1. Boutron I, Page MJ, Higgins JP, Altman DG, Lundh A, Hróbjartsson A. Chapter 7: Considering bias and conflicts of interest among the included studies. In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 6.2 (updated February 2021). Cochrane, 2021. Available from training.cochrane.org/handbook.

Cipriani 2016

  1. Cipriani A, Zhou X, Del Giovane C, Hetrick SE, Qin B, Whittington C, et al. Comparative efficacy and tolerability of antidepressants for major depressive disorder in children and adolescents: a network meta-analysis. Lancet 2016;388(10047):881-90. [DOI] [PubMed] [Google Scholar]

Cipriani 2018

  1. Cipriani A, Furukawa TA, Salanti G, Chaimani A, Atkinson LZ, Ogawa Y, et al. Comparative efficacy and acceptability of 21 antidepressant drugs for the acute treatment of adults with major depressive disorder: a systematic review and network meta-analysis. Lancet 2018;16(4):420-9. [DOI] [PMC free article] [PubMed] [Google Scholar]

Cipriani 2020

  1. Cipriani A, Ioannidis JA, Rothwell PM, Glasziou P, Li T, Hernandez AF, et al. Generating comparative evidence on new drugs and devices after approval. Lancet 2020;395(10228):998-1010. [DOI] [PubMed] [Google Scholar]

Citrome 2015

  1. Citrome L. Brexpiprazole for schizophrenia and as adjunct for major depressive disorder: a systematic review of the efficacy and safety profile for this newly approved antipsychotic – what is the number needed to treat, number needed to harm and likelihood to be helped or harmed? International Journal of Clinical Practice 2015;69(9):978-97. [DOI] [PubMed] [Google Scholar]

Cuijpers 2020

  1. Cuijpers P, Noma H, Karyotaki E, Vinkers CH, Cipriani A, Furukawa TA. A network meta-analysis of the effects of psychotherapies, pharmacotherapies and their combination in the treatment of adult depression. World Psychiatry 2020;19(1):92-107. [DOI] [PMC free article] [PubMed] [Google Scholar]

Deeks 2021

  1. Deeks JJ, Higgins JP, Altman DG. Chapter 10: Analysing data and undertaking meta-analyses. In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 6.2 (updated February 2021). Cochrane, 2021. Available from training.cochrane.org/handbook.

Dold 2018

  1. Dold M, Bartova L, Kautzky A, Serretti A, Porcelli S, Souery D, et al. Clinical factors associated with augmentation treatment with second-generation antipsychotics and lithium in major depression – results from a European multicenter study. European Neuropsychopharmacology 2018;28(12):1305-13. [DOI] [PubMed] [Google Scholar]

Dold 2019

  1. Dold M, Bartova L, Kautzky A, Porcelli S, Montgomery S, Zohar J, et al. Psychotic features in patients with major depressive disorder: a report from the European group for the study of resistant depression. Journal of Clinical Psychiatry 2019;80(1):17m12090. [DOI] [PubMed] [Google Scholar]

EMA 2018

  1. European Medicines Agency. Rxulti (brexpiprazole): summary of opinion (initial authorisation). www.ema.europa.eu/en/documents/smop-initial/chmp-summary-positive-opinionrxulti_en.pdf (accessed 8 June 2020).

Fava 2017

  1. Fava M, Okame T, Matsushima Y, Perry P, Weiller E, Baker RA. Switching from inadequate adjunctive or combination treatment options to brexpiprazole adjunctive to antidepressant: an open-label study on the effects on depressive symptoms and cognitive and physical functioning. International Journal of Neuropsychopharmacology 2017;20(1):22-30. [DOI] [PMC free article] [PubMed] [Google Scholar]

Fava 2020

  1. Fava GA, Cosci F, Guidi J, Rafanelli C. The deceptive manifestations of treatment resistance in depression: a new look at the problem. Psychotherapy and Psychosomatics 2020;89(5):265-73. [DOI] [PubMed] [Google Scholar]

FDA 2007

  1. US Food and Drug Administration. Abilify (aripiprazole):NDA application approval. www.accessdata.fda.gov/drugsatfda_docs/appletter/2007/021436s018,%20021866s005,%20021729s005,%20021713s013ltr.pdf (accessed 4 April 2020).

FDA 2015a

  1. US Food and Drug Administration. Rexulti (brexpiprazole): NDA application approval. www.accessdata.fda.gov/drugsatfda_docs/appletter/2015/205422Orig1Orig2s000ltr.pdf (accessed 4 April 2020).

FDA 2015b

  1. US Food and Drug Administration. Rexulti (brexpiprazole): Medical review(s). www.accessdata.fda.gov/drugsatfda_docs/nda/2015/205422Orig1Orig2s000MedR.pdf (accessed 1 September 2021).

Feighner 1972

  1. Feighner JP, Robins E, Guze SB, Woodruff RA, Winokur G, Munoz R. Diagnostic criteria for use in psychiatric research. Archives of General Psychiatry 1972;26(1):57-63. [DOI] [PubMed] [Google Scholar]

Furukawa 2005

  1. Furukawa TA, Cipriani A, Barbui C, Brambilla P, Watanabe N. Imputing response rates from means and standard deviations in meta-analysis. International Clinical Psychopharmacology 2005;20(1):49-52. [DOI] [PubMed] [Google Scholar]

Furukawa 2016

  1. Furukawa TA, Salanti G, Atkinson LZ, Leucht S, Ruhe HG, Turner EH. Comparative efficacy and acceptability of first-generation and second-generation antidepressants in the acute treatment of major depression: protocol for a network meta-analysis. BMJ Open 2016;6(7):e010919. [DOI] [PMC free article] [PubMed] [Google Scholar]

Furukawa 2018

  1. Furukawa TA, Maruo K, Noma H, Tanaka S, Imai H, Shinohara K, et al. Initial severity of major depression and efficacy of new generation antidepressants: individual participant data meta-analysis. Acta Psychiatrica Scandinavica 2018;137:450-8. [DOI] [PubMed] [Google Scholar]

Furukawa 2019

  1. Furukawa TA, Cipriani A, Cowen PJ, Leucht S, Egger M, Salanti G. Optimal dose of selective serotonin reuptake inhibitors, venlafaxine, and mirtazapine in major depression: a systematic review and dose-response meta-analysis. Lancet Psychiatry 2019;6(7):601-9. [DOI] [PMC free article] [PubMed] [Google Scholar]

Furukawa 2020

  1. Furukawa TA, Salanti G, Cowen PJ, Leucht S, Cipriani A. No benefit from flexible titration above minimum licensed dose in prescribing antidepressants for major depression: systematic review. Acta Psychiatrica Scandinavica 2020;141(5):401-9. [DOI] [PubMed] [Google Scholar]

Furukawa 2021

  1. Furukawa Y, Hamza T, Cipriani A, Furukawa TA, Salanti G, Ostinelli EG. Optimal dose of aripiprazole for augmentation therapy of antidepressant-refractory depression: preliminary findings based on a systematic review and dose-effect meta-analysis. British Journal of Psychiatry 2021;15:1-8. [DOI] [PubMed] [Google Scholar]

Furukawa 2022

  1. Furukawa Y, Oguro S, Obata S, Hamza T, Ostinelli EG, Kasai K. Optimal dose of brexpiprazole for augmentation therapy of antidepressant-refractory depression: a systematic review and dose-effect meta-analysis. Psychiatry and Clinical Neurosciences 2022;76(9):416-22. [DOI] [PubMed] [Google Scholar]

GBD 2017

  1. GBD 2017 Disease and Injury Incidence and Prevalence Collaborators. Global, regional, and national incidence, prevalence, and years lived with disability for 354 diseases and injuries for 195 countries and territories, 1990–2017: a systematic analysis for the Global Burden of Disease Study 2017. Lancet 2017;392(10159):1789-858. [DOI] [PMC free article] [PubMed] [Google Scholar]

GRADEpro GDT [Computer program]

  1. GRADEpro GDT. Hamilton (ON): McMaster University (developed by Evidence Prime), accessed 15 March 2023. Available at gradepro.org.

Guy 1970

  1. Guy W, Bonato RR. Manual for the ECDEU Assessment Battery. 2nd edition. Chevy Chase (MD): U.S. Department of Health, Education, and Welfare, National Institute of Mental Health, 1970. [Google Scholar]

Hamilton 1960

  1. Hamilton M. A rating scale for depression. Journal of Neurology, Neurosurgery and Psychiatry 1960;23:56-62. [DOI] [PMC free article] [PubMed] [Google Scholar]

Hassan 2016

  1. Hassan AK, Farmer KC, Brahm NC, Keast S, Nesser N, Neas BR. Does the use of atypical antipsychotics as adjunctive therapy in depression result in cost savings? Comparing healthcare costs and utilization between second-line treatment options. Journal of Mental Health 2016;25(6):486-91. [DOI] [PubMed] [Google Scholar]

Higgins 2011a

  1. Higgins JP, Altman DG, Sterne JA. Chapter 8: Assessing risk of bias in included studies. In: Higgins JP, Green S, editor(s). Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. Available from training.cochrane.org/handbook/archive/v5.1/.

Higgins 2011b

  1. Higgins JP, Deeks JJ, Altman DG. Chapter 16: Special topics in statistics. In: Higgins JP, Green S, editor(s), Cochrane Handbook for Systematic Reviews of Interventions. Version 5.1.0 (updated March 2011). The Cochrane Collaboration, 2011. Available from training.cochrane.org/handbook/archive/v5.1/.

Higgins 2021

  1. Higgins JP, Eldridge S, Li T. Chapter 23: Including variants on randomized trials. In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 6.2 (updated February 2021). Cochrane, 2021. Available from www.training.cochrane.org/handbook 2021.

Hobart 2018c

  1. Hobart M, Zhang P, Weiss C, Meehan SR, Eriksson H. Adjunctive brexpiprazole and functioning in major depressive disorder: a pooled analysis of six randomized studies using the Sheehan disability scale. International Journal of Neuropsychopharmacology 2018;22:173-9. [DOI] [PMC free article] [PubMed] [Google Scholar]

Kato 2018

  1. Kato T, Furukawa TA, Mantani A, Kurata K, Kubouchi H, Hirota S, et al. Optimising first- and second-line treatment strategies for untreated major depressive disorder – the sun☺d study: a pragmatic, multi-centre, assessor-blinded randomised controlled trial. BMC Medicine 2018;16(1):103. [DOI] [PMC free article] [PubMed] [Google Scholar]

Kennedy 2016

  1. Kennedy SH, Lam RW, McIntyre RS, Tourjman SV, Bhat V, Blier P, et al. Canadian network for mood and anxiety treatments (CANMAT) 2016 clinical guidelines for the management of adults with major depressive disorder: section 3. Pharmacological treatments. Canadian Journal of Psychiatry 2016;61(9):540-60. [DOI] [PMC free article] [PubMed] [Google Scholar]

Kishi 2019

  1. Kishi T, Sakuma K, Nomura I, Matsuda Y, Mishima K, Iwata N. Brexpiprazole as adjunctive treatment for major depressive disorder following treatment failure with at least one antidepressant in the current episode: a systematic review and meta-analysis. International Journal of Neuropsychopharmacology 2019;22(11):698-709. [DOI] [PMC free article] [PubMed] [Google Scholar]

Kraus 2020

  1. Kraus C, Kadriu B, Lanzenberger R, Zarate CA Jr, Kasper S. Prognosis and improved outcomes in major depression: a review. FOCUS the Journal of Lifelong Learning in Psychiatry 2020;18(2):220-35. [DOI] [PMC free article] [PubMed] [Google Scholar]

Liberati 2009

  1. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, Ioannidis JP, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLOS Medicine 2009;6(7):e1000100. [DOI] [PMC free article] [PubMed] [Google Scholar]

Lundh 2017

  1. Lundh A, Lexchin J, Mintzes B, Schroll JB, Bero L. Industry sponsorship and research outcome. Cochrane Database of Systematic Reviews 2017, Issue 2. Art. No: MR000033. [DOI: 10.1002/14651858.MR000033.pub3] [DOI] [PMC free article] [PubMed] [Google Scholar]

Maeda 2014a

  1. Maeda K, Sugino H, Akazawa H, Amada N, Shimada J, Futamura T, et al. Brexpiprazole I: in vitro and in vivo characterization of a novel serotonin-dopamine activity modulator. Journal of Pharmacology and Experimental Therapeutics 2014;350(3):589-604. [DOI] [PubMed] [Google Scholar]

Maeda 2014b

  1. Maeda K, Lerdrup L, Sugino H, Akazawa H, Amada N, McQuade RD, et al. Brexpiprazole II: antipsychotic-like and procognitive effects of a novel serotonin-dopamine activity modulator. Journal of Pharmacology and Experimental Therapeutics 2014;350(3):605-14. [DOI] [PubMed] [Google Scholar]

Malhi 2016

  1. Malhi GS, Byrow Y. Is treatment-resistant depression a useful concept? Evidence-Based Mental Health 2016;19(1):1-3. [DOI] [PMC free article] [PubMed] [Google Scholar]

Malhi 2021

  1. Malhi GS, Bell E, Bassett D, Boyce P, Bryant R, Hazell P, et al. The 2020 Royal Australian and New Zealand College of Psychiatrists clinical practice guidelines for mood disorders. Australian & New Zealand Journal of Psychiatry 2021;55(1):7-117. [DOI] [PubMed] [Google Scholar]

Martinez 2017

  1. Martinez MN, Bartholomew MJ. What does it "mean"? A review of interpreting and calculating different types of means and standard deviations. Pharmaceutics 2017;9(2):14. [DOI] [PMC free article] [PubMed] [Google Scholar]

McIntyre 2016

  1. McIntyre RS, Weiller E, Zhang P, Weiss C. Brexpiprazole as adjunctive treatment of major depressive disorder with anxious distress: results from a post-hoc analysis of two randomised controlled trials. Journal of Affective Disorders 2016;201:116-23. [DOI] [PubMed] [Google Scholar]

Montgomery 1979

  1. Montgomery SA, Asberg M. A new depression scale designed to be sensitive to change. British Journal of Psychiatry 1979;134:382-9. [DOI] [PubMed] [Google Scholar]

Naci 2020

  1. Naci H, Salcher-Konrad M, Kesselheim AS, Wieseler B, Rochaix L, Redberg RF, et al. Generating comparative evidence on new drugs and devices before approval. Lancet 2020;395(10228):986-97. [DOI] [PubMed] [Google Scholar]

Nelson 2018

  1. Nelson JC, Weiller E, Zhang P, Weiss C, Hobart M. Efficacy of adjunctive brexpiprazole on the core symptoms of major depressive disorder: a post hoc analysis of two pooled clinical studies. Journal of Affective Disorders 2018;227:103-8. [DOI] [PubMed] [Google Scholar]

NICE 2009

  1. National Institute for Health and Care Excellence. Depression in adults: recognition and management. Clinical guideline CG90. www.nice.org.uk/guidance/cg90 (accessed 12 October 2020).

Nuñez 2022

  1. Nuñez NA, Joseph B, Pahwa M, Kumar R, Resendez MG, Prokop LJ, et al. Augmentation strategies for treatment resistant major depression: a systematic review and network meta-analysis. Journal of Affective Disorders 2022;302:385-400. [DOI] [PMC free article] [PubMed] [Google Scholar]

Ohayon 2002

  1. Ohayon M, Schatzberg A. Prevalence of depressive episodes with psychotic features in the general population. American Journal of Psychiatry 2002;159(11):1855-61. [DOI] [PubMed] [Google Scholar]

Otsuka 2020

  1. Otsuka Pharmaceutical Co Ltd. Highlights of prescribing information for Rexulti. www.otsuka-us.com/media/static/Rexulti-PI.pdf (accessed 4 April 2020).

Page 2021

  1. Page MJ, Higgins JP, Sterne JA. Chapter 13: Assessing risk of bias due to missing results in a synthesis In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA, editor(s). Cochrane Handbook for Systematic Reviews of Interventions Version 6.2 (updated February 2021). Cochrane, 2021. Available fromwww.training.cochrane.org/handbook.

Papakostas 2020

  1. Papakostas GI, Jackson WC, Rafeyan R, Trivedi MH. Inadequate response to antidepressant treatment in major depressive disorder. Journal of Clinical Psychiatry 2020;81(3):OT19037COM5. [DOI] [PubMed] [Google Scholar]

Review Manager 2020 [Computer program]

  1. Review Manager 2020. Version 5.4. Copenhagen: The Cochrane Collaboration, 2020.

Sadock 2015

  1. Sadock BJ, Sadock VA, Ruiz P. Kaplan & Sadock's Synopsis of Psychiatry: Behavioral Sciences/Clinical Psychiatry. Eleventh edition. Philadelphia (PA): Lippincott Williams and Wilkins, 2015. [Google Scholar]

Salanti 2010

  1. Salanti G, Dias S, Welton NJ, Ades AE, Golfinopoulos V, Kyrgiou M, et al. Evaluating novel agent effects in multiple-treatments meta-regression. Statistics in Medicine 2010;29(23):2369-83. [DOI] [PubMed] [Google Scholar]

Salanti 2018

  1. Salanti G, Chaimani A, Furukawa TA, Higgins JP, Ogawa Y, Cipriani A, et al. Impact of placebo arms on outcomes in antidepressant trials: systematic review and meta-regression analysis. International Journal of Epidemiology 2018;47(5):1454-64. [DOI] [PMC free article] [PubMed] [Google Scholar]

Schweizer 2016

  1. Schweizer ML, Braun BI, Milstone AM. Research methods in healthcare epidemiology and antimicrobial stewardship-quasi-experimental designs. Infection Control & Hospital Epidemiology 2016;37(10):1135-40. [DOI] [PMC free article] [PubMed] [Google Scholar]

Schünemann 2021

  1. Schünemann HJ, Higgins JP, Vist GE, Glasziou P, Akl EA, Skoetz N, et al. Chapter 14: Completing 'Summary of findings' tables and grading the certainty of the evidence. In: Higgins JP, Thomas J, Chandler J, Cumpston MS, Li T, Page MJ, et al, editor(s), Cochrane Handbook for Systematic Reviews of Interventions Version 6.2 (updated February 2021). Cochrane 2021. Available from www.training.cochrane.org/handbook.

Shadish 2002

  1. Shadish WR, Cook TD, Campbell DT. Experimental and Quasi-Experimental Designs for Generalized Causal Inference. Boston (MA): Houghton Mifflin, 2002. [Google Scholar]

Sheehan 1983

  1. Sheehan DV. The Anxiety Disease. New York (NY): Charles Scribner and Sons, 1983. [Google Scholar]

Singh 2020

  1. Singh I, Naci H, Miller J, Caplan A, Cipriani A. Ethical implications of poor comparative effectiveness evidence: obligations in industry-research partnerships. Lancet 2020;395(10228):926-8. [DOI] [PubMed] [Google Scholar]

Spitzer 1978

  1. Spitzer RL, Endicott J, Robins E. Research diagnostic criteria: rationale and reliability. Archives of General Psychiatry 1978;35(6):773-82. [DOI] [PubMed] [Google Scholar]

Stahl 2016

  1. Stahl SM. Mechanism of action of brexpiprazole: comparison with aripiprazole. CNS Spectrums 2016;21(1):1-6. [DOI] [PubMed] [Google Scholar]

Stevanovic 2011

  1. Stevanovic D. Quality of Life Enjoyment and Satisfaction Questionnaire-short form for quality of life assessments in clinical practice: a psychometric study. Journal of Psychiatric and Mental Health Nursing 2011;18(8):744-50. [DOI] [PubMed] [Google Scholar]

Thase 2016

  1. Thase ME, Zhang P, Skuban A, Hobart M, Weiss C, Weiller E, et al. Efficacy of adjunctive brexpiprazole in patients with major depressive disorder: a clinical overview. Current Psychiatry Reviews 2016;12:291-301. [Google Scholar]

Thase 2019a

  1. Thase ME, Weiller E, Zhang P, Weiss C, McIntyre RS. Adjunctive brexpiprazole in patients with major depressive disorder and anxiety symptoms: post hoc analyses of three placebo-controlled studies. Neuropsychiatric Disease and Treatment 2019;15:37-45. [DOI] [PMC free article] [PubMed] [Google Scholar]

Thase 2019b

  1. Thase ME, Zhang P, Weiss C, Meehan SR, Hobart M. Efficacy and safety of brexpiprazole as adjunctive treatment in major depressive disorder: overview of four short-term studies. Expert Opinion in Pharmacotherapy 2019;20(15):1907-16. [DOI] [PubMed] [Google Scholar]

Tundo 2019

  1. Tundo A, Filippis R, De Crescenzo F. Pramipexole in the treatment of unipolar and bipolar depression. A systematic review and meta-analysis. Acta Psychiatrica Scandinavica 2019;140(2):116-25. [DOI] [PubMed] [Google Scholar]

Vancak 2021

  1. Vancak V, Goldberg Y, Levine SZ. Guidelines to understand and compute the number needed to treat. Evidence-Based Mental Health 2021;24(4):131-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

WHO 1978

  1. World Health Organization. The Ninth Revision of the International Classification of Diseases and Related Health Problems (ICD-9). Geneva: World Health Organization, 1978. [Google Scholar]

WHO 1992

  1. World Health Organization. The Tenth Revision of the International Classification of Diseases and Related Health Problems (ICD-10). Geneva: World Health Organization, 1992. [Google Scholar]

Yoon 2017

  1. Yoon S, Jeon SW, Ko YH, Patkar AA, Masand PS, Pae CU, et al. Adjunctive brexpiprazole as a novel effective strategy for treating major depressive disorder: a systematic review and meta-analysis. Journal of Clinical Psychopharmacology 2017;37:46-53. [DOI] [PubMed] [Google Scholar]

References to other published versions of this review

Ralovska 2021

  1. Ralovska S, Koychev I, De Crescenzo F, Marinov P, Cipriani A. Brexpiprazole versus placebo or other antidepressive agents for treating depression. Cochrane Database of Systematic Reviews 2021, Issue 2. Art. No: CD013866. [DOI: 10.1002/14651858.CD013866] [DOI] [Google Scholar]

Articles from The Cochrane Database of Systematic Reviews are provided here courtesy of Wiley

RESOURCES