CADTH Reimbursement Review

Cariprazine (Vraylar)

Sponsor: Allergan (an AbbVie Inc. company)

Therapeutic area: Bipolar disorder

This multi-part report includes:

Clinical Review

Pharmacoeconomic Review

Stakeholder Input

Clinical Review

Executive Summary

An overview of the submission details for the drug under review is provided in Table 1.

Introduction

Living with bipolar disorder is associated with a substantial burden of illness that people experience for approximately half of their lives.¹ The estimated average age of onset in Canada is 22.5 years² and the estimated lifetime prevalence of bipolar I disorder in Canada is 0.87%, affecting males and females equally.¹ Quality of life and psychosocial functioning, including the ability to maintain proper work, are severely impacted by bipolar disorder and more pronounced in those living with depressive symptoms, with multiple previous episodes or a longer duration of illness, and in those with cognitive decline.¹ In addition, suicide risk is substantially higher for patients living with bipolar disorder relative to the general population (10.7 deaths due to suicide per 100,000 people per year).¹

Bipolar I disorder is a mood disorder characterized by episodes of mania, hypomania, and major depression.³ Episodes of mania and depression present with significant changes in mood, energy, behaviour, sleep, and cognition. Mania also presents with change in activity. The Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) includes specifiers that describe characteristics of bipolar disorder and can be used to guide treatment decisions for acute mania and depression. The presence of mixed features is a specifier for patients experiencing at least 3 nonoverlapping symptoms from the opposing pole (mania or depression) during the majority of the days of the current episode.⁴

There are no disease-modifying treatments for bipolar disorder. Bipolar disorder is treated with the combination of pharmacologic, nonpharmacologic (e.g., electroconvulsive therapy), and psychosocial (e.g., psychotherapy) treatments. The clinical experts stated that the primary goal of current treatments is to improve acute symptoms and, by doing so, restore functioning. The prevention of recurrent mood episodes was described as the secondary goal. Long-term treatment efficacy is more uncertain as high-quality longitudinal studies of longer duration are limited. In Canada, pharmacologic treatments are usually managed by family physicians in milder to moderate cases or by psychiatrists in more severe cases, although this varies regionally. Medications include mood stabilizers (which include drugs from a variety of classes, such as second-generation antipsychotic drugs, anticonvulsants, and lithium), and antidepressants. The clinical experts stated that overall, there are many different pharmacologic options, and the choice of treatment is usually guided by evidence of efficacy and tolerability, characteristics of the patient (age, income, individual preferences), characteristics of the disorder (including, but not limited to, predominant symptomatology or polarity [mania, depression, or mixed features], medical and psychiatric comorbidities, and resistance to treatment) and characteristics of the treatments (safety and tolerability issues, availability, cost). The clinical experts indicated that the use of off-label drugs for mania is uncommon as there are several drugs available in Canada with adequate efficacy. In contrast, the experts indicated fewer treatments are available for the management of depressive episodes, and they are associated with relatively high rates of treatment failure due to nonresponse or poor tolerability. As a result, the use of drugs without approval, such as antidepressants, is common for bipolar depression.

Cariprazine is a second-generation antipsychotic drug that is approved by Health Canada for use in adults as monotherapy for the acute management of manic or mixed episodes associated with bipolar I disorder (bipolar mania) and the acute management of depressive episodes associated with bipolar I disorder (bipolar depression).⁵ The mechanism of action for cariprazine is unknown.⁵ The sponsor has requested reimbursement as per the indication.⁶ As such, the objective of this review is to perform a systematic review of the beneficial and harmful effects of cariprazine 1.5 mg, 3 mg, 4.5 mg, and 6 mg for use as monotherapy in adults for the acute management of manic or mixed episodes associated with bipolar I disorder (bipolar mania) and for the acute management of depressive episodes associated with bipolar I disorder (bipolar depression). Cariprazine has been previously reviewed by CADTH for the treatment of schizophrenia in adults and is approved by Health Canada for the acute and long-term maintenance treatment of schizophrenia in adults as well.

Stakeholder Perspectives

The information in this section is a summary of input provided by the patient groups that responded to CADTH’s call for patient input and from clinical experts consulted by CADTH for the purpose of this review.

Patient Input

This section was prepared by CADTH staff based on the input provided by patient groups.

Two responses to CADTH’s call for patient input were received for this review: a submission from the Institute for Advancements in Mental Health (IAM) and a submission from the Mood Disorders Society of Canada (MDSC). IAM and MDSC are organizations that support individuals living with mental illness such as bipolar disorder, including patients, their families, and service providers.

The information used to inform the IAM submission was based on 2 online surveys conducted in 2018 and in February 2022. Potential respondents to these online surveys included members of the IAM and Hope and Me — Mood Disorders Association of Ontario client networks. Among the respondents of the 2018 survey, 12% self-described as personally diagnosed, 50% were caregivers, 63% were family members or friends of someone diagnosed, and 18% worked in social services. Among the respondents of the 2022 survey, 33% identified as an individual living with symptoms of bipolar disorder, 58% were relatives, 8% were caregivers of someone with lived experience, and 1% did not specify. MDSC drew information from interviews with patients and family members, 2 national mental health surveys conducted in 2018 and in September 2021, and shared experiences that had been posted on the MDSC online discussion forum. The interviews were conducted with 5 patients with bipolar disorder, including semi-structured phone interviews with adults living with bipolar disorder in January 2021, and 3 family members.

Respondents indicated that living with bipolar disorder had impacted their mental health, social relationships, and day-to-day functioning at work and school. Patients can also experience a lack of insight into their illness, which impacts their ability and motivation to seek treatment and causes significant strain in relationships. Survey respondents described the advantages of taking medications for bipolar disorder, which include managing symptoms of bipolar disorder, experiencing fewer episodes of mental illness, and avoiding visits to the hospital. They also described disadvantages of taking medications for bipolar disorder; these include requiring frequent follow-ups with health care providers, needing to take the medication daily, and dealing with the symptoms — including bipolar depression, which was not being well controlled. Further, respondents described the benefits of an injectable formulation, which include convenience and not needing to remember to take it daily, while the difficulties were pain at injection site and frequent travel to clinics. The most common side effects of medications for bipolar disorder identified by respondents were drowsiness, dry mouth, restlessness, and weight gain. Respondents also identified the cost of medications as a significant barrier to access.

Survey respondents reported that treatment of bipolar disorder is individualized, as not every patient will respond to 1 medication. To find the right medication that enables the highest degree of functioning while minimizing side effects, patients with bipolar disorder often have to go through a trial-and-error process. This process involves taking a number of different medications and at different dosages until their goals of therapy have been achieved. This process can make it challenging for patients to adhere to their prescribed regimen and can be exacerbated by additional challenges, such as waiting to be approved for coverage by public drug programs and experiencing relapse. As a result, patients feel that outcomes can be improved by increasing equitable access to, and the selection of, medications that are reimbursable. According to respondents, antipsychotic medications could be improved by increasing their ability to control the symptoms of bipolar disorder, improving the side effect profile, and providing a greater range of strengths and dosages to lower the frequency of administration.

A copy of the patient input from IAM and MDSC is presented in Appendix 5.

Clinician Input

Input From Clinical Experts Consulted by CADTH

The clinical experts consulted by CADTH indicated numerous challenges with addressing treatment goals for patients living with bipolar I disorder using currently available treatments. Insufficient response to treatment was noted as being very common, particularly in bipolar depression. There are no disease-modifying treatments and the long-term effectiveness of current treatments is unclear. Bipolar disorder is typically a lifelong, persistent, and/or highly recurrent condition. Some key outcomes are not adequately addressed by current treatments. For example, the extent to which current treatments address cognition directly, instead of indirectly through mood symptoms, is unclear. Tolerability is a problem, particularly in bipolar depression, where the most effective drugs, like quetiapine and olanzapine, have well-known metabolic side effects; this is particularly concerning given the elevated, and independent, risk of metabolic and cardiovascular conditions in this population. Finally, adherence is also an issue — poor compliance is very common in real-world settings.

Based on the currently available clinical evidence, the clinical experts felt it was very unlikely that cariprazine would cause a shift in treatment paradigm, despite having a somewhat distinct pharmacologic profile. Mechanistically, cariprazine was described as very similar to currently available treatments and, therefore, very much within the current paradigm of symptomatic management. However, the experts noted that the combination of efficacy for both acute mania and depression, as well as an acceptable tolerability profile, may eventually result in cariprazine as a first-line treatment. The experts did not identify any reason to recommend that patients try other treatments before cariprazine. The clinical experts suggested cariprazine will probably be used in monotherapy and in combination with either lithium or anticonvulsants, which is a standard approach with second-generation antipsychotic drugs in bipolar disorder.

The experts indicated that it is challenging to identify and diagnose patients with bipolar disorder, and that misdiagnosis and/or delayed diagnosis are relatively common. Relatedly, they stated there are no diagnostic tools or tests that are useful, and the diagnosis is based on clinical assessment.

At this time, the clinical experts stated that it is not possible to identify patients who are most likely to exhibit a response to treatment with cariprazine. They noted that the reliable prediction of response has not been achieved even with sophisticated research methods (e.g., genomics, neuroimaging), and certainly cannot be done reliably with the most widely available clinical tools. Other than polarity of mood episode, there are no other validated indices to predict response to any given drug, according to the clinical experts. The clinical experts stated that presymptomatic patients should not be treated with cariprazine, or any other drug, mostly due to the low predictive power of current assessment tools. The clinical experts did not identify any patients who would be least suited for treatment with cariprazine, noting that within the population of adults with bipolar I disorder, there were no major contraindications unique to cariprazine.

The clinical experts identified the Young Mania Rating Scale (YMRS), the Montgomery–Åsberg Depression Rating Scale (MADRS), and the Hamilton Depression Rating Scale (HAM-D) as the most commonly used outcomes to assess response to treatment in research settings, but in the real world, patient-rated questionnaires such as the Patient Health Questionnaire and the Beck Depression Inventory are more common. The experts also noted that all of these outcomes, however, have very good concurrent validity. A reduction in the frequency or severity of symptoms, an improvement in symptoms, the stabilization (no deterioration) of symptoms, the ability to perform activities of daily living, and improved survival were all clinically meaningful according to the clinical experts. They described a reduction in the severity of symptoms (e.g., controlling physical agitation) and frequency of symptoms as the usual first goal of treatment. Subsequently, the experts stated that treatment aims to restore functioning — including improving cognitive functioning and coping skills — with a return to work, school, and/or daily activities being an important benchmark. The experts identified long-term goals, such as the prevention of relapses and recurrences, as well as the onset and progression of psychiatric comorbidities (e.g., anxiety disorders, substance abuse) and medical comorbidities (e.g., obesity, diabetes).

Regarding how often treatment response should be assessed, the clinical experts suggested that in the acute phase, response is usually assessed between 2 weeks to 4 weeks, depending on the severity. In the maintenance phase, it is recommended that patients be assessed at least every 2 months to 3 months. The clinical experts indicated that treatment discontinuation is determined by either lack of response or poor tolerability, noting that most guidelines recommend discontinuing a treatment if there is no response to very poor response within 4 weeks to 6 weeks. Further, they indicated that discontinuation due to tolerability depends on the severity and progression of specific side effects, particularly extrapyramidal symptom (EPS) and akathisia, which are the most common side effects of cariprazine and similar drugs. These side effects can be time-dependent; they tend to be worse during initial titration phases but improve with time. If side effects are moderate to severe and/or do not meaningfully improve in 1 weeks to 2 weeks, treatment should be discontinued.

The clinical experts reported that family physicians can and frequently do diagnose bipolar disorder and regularly prescribe similar drugs. The experts also noted that there are no special tests required for the diagnosis of bipolar disorder, nor for the prescription and monitoring of cariprazine. The clinical experts noted that given the high prevalence of bipolar disorder and the relative lack of psychiatrists across Canada, a significant proportion of patients is treated by family physicians; therefore, the experts recommended that requiring the involvement of specialists would significantly restrict the use of this medication.

Clinician Group Input

Two clinician groups provided input to this review: Canadian Network for Mood and Anxiety Treatments (CANMAT), a network of academic and clinical experts that produces treatment guidelines and educational material for health professionals and information for patients and families, and Western Canadian Clinical Advisory Network (WC-CAN), a network of senior psychiatrists across Alberta and British Columbia. One clinician on behalf of CANMAT and 6 clinicians with the WC-CAN contributed to these submissions. Both clinician groups recognized the unmet need for a medication that is effective in multiple phases of bipolar disorder, including bipolar depression, with low rates of adverse events (AEs) to minimize polypharmacy and improve adherence. The clinical experts consulted by CADTH identified additional unmet needs, which include the absence of disease-modifying drugs, the uncertainty regarding the long-term effectiveness and direct effects on cognition of currently available treatments, and the lack of depot alternatives for commonly used first-line pharmacologic options. Both clinician groups advocated for cariprazine as a first-line treatment option for patients with bipolar disorder in the treatment of acute mania and depression and for cariprazine to be used as monotherapy and possibly as combination therapy with other mood stabilizers.

Drug Program Input

Input was obtained from the drug programs that participate in the CADTH reimbursement review process. The following were identified as key factors that could potentially impact the implementation of a CADTH recommendation for cariprazine:

• considerations for the initiation of therapy

• considerations for generalizability.

Detailed drug program input is found in Table 6.

Clinical Evidence

Pivotal Studies and Protocol Selected Studies

Description of Studies

The systematic review of cariprazine included a total of 6 multi-centre, randomized, double-blind, placebo-controlled, parallel-group studies in adults with a primary diagnosis of bipolar I disorder. Of the 6 included randomized controlled trials (RCTs), 3 evaluated cariprazine 3 mg to 12 mg (flexible dose) for the treatment of acute manic or mixed episodes (study RGH-MD-31 [N = 238], study RGH-MD-32 [N = 312], and study RGH-MD-33 [N = 497] — frequently referred to as the bipolar mania studies in this report) and 3 evaluated cariprazine 1.5 mg and 3.0 mg (fixed dose) for the treatment of acute depressive episodes (study RGH-MD-56 [N = 578], study RGH-MD-53 [N = 493], and study RGH-MD-54 [N = 488] — frequently referred to as the bipolar depression studies in this report). The flexible-dosing regimen in the bipolar mania studies involved dose adjustments based on treatment response assessed by the investigator (study RGH-MD-31) or using the YMRS (study RGH-MD-32 and study RGH-MD-33) and tolerability, determined at the discretion of the investigator. The objective of each of the studies was similar: to evaluate the efficacy, safety, and tolerability of cariprazine monotherapy versus placebo for the treatment of acute manic or mixed episodes or acute depressive episodes associated with bipolar I disorder. The primary end point in all studies was the change in symptoms, measured using the YMRS total score in the studies of acute manic or mixed episodes and the MADRS total score in the studies of acute depressive episodes. The secondary end point was the change in the Clinical Global Impression–Severity of Illness (CGI-S) score; this was consistent across studies. Primary and secondary outcomes were assessed at week 3 in the acute mania studies and at week 6 in the acute depression studies. Other measures of changes in symptom severity (Clinical Global Impression–Improvement [CGI-I], Positive and Negative Syndrome Scale [PANSS], HAM-D, Hamilton Anxiety Rating Scale [HAM-A], and Quick Inventory of Depressive Symptomatology–Self-Rated [QIDS-SR]), functioning (Functioning Assessment Short Test [FAST]), suicidal ideation (Columbia–Suicide Severity Rating Scale [C-SSRS]), response rate, and remission rate were also reported.

The mean age of patients enrolled in the included studies ranged between 36 (standard deviation [SD] = 11) years and 44 (SD = 13) years and the mean weight of patients ranged from 70 (SD = 20) kg to 87 (SD = 25) kg. The population in the acute mania studies was slightly younger and had a lower body weight than patients in the acute depression studies. More patients in the acute mania studies were male (ranging from 54% to 68% across studies) and more patients in the acute depression studies were female (ranging from 57% to 65% across studies). The duration of bipolar I disorder was a mean of 10 (SD = 9) years to 16 (SD = 10) years and the age of onset ranged from 23 (SD = 8) years to 28 (SD = 11) years. The duration of and age of onset of bipolar I disorder was not reported in study RGH-MD-53 or study RGH-MD-54. During the trials for acute mania or mixed episodes, between 81% and |||||| of patients were currently experiencing a manic episode and between and 20% were experiencing a mixed episode. Between |||||| and |||||| of patients were diagnosed with a moderate episode, |||||| to |||||| with a severe episode without psychotic features, and |||||| to |||||| with a severe episode with psychotic features. During study RGH-MD-53 and study RGH-MD-54, between |||||| and |||||| of patients were currently experiencing a severe depressive episode and between |||||| and |||||| were experiencing a moderate depressive episode, and the duration of the current episode was a mean of 3.3 (SD = 2.3) months to 3.8 (SD = 2.5) months. Similar statistics were not reported in study RGH-MD-56.

Efficacy Results

A summary of key efficacy results for the studies in patients with acute manic or mixed episodes of bipolar disorder is presented in Table 2; in Table 3, a similar summary is presented for the studies in patients with acute depressive episodes of bipolar disorder. Of note, the bipolar studies used a flexible-dose regimen for cariprazine. The overall mean daily dose in study RGH-MD-31 and study RGH-MD-32 was 8.8 (SD = ||||||) mg and 7.5 (SD = ||||||) mg per day, respectively. The overall mean daily dose in study RGH-MD-33 was 4.8 (SD = ||||||) mg. The overall mean daily dose was not reported in the bipolar depression studies.

All 3 studies of acute mania demonstrated that treatment with flexible dosing of cariprazine was associated with a greater reduction of symptoms of acute mania relative to placebo, based on the YMRS total score at week 3. This was based on a difference in score between cariprazine 3 mg to 12 mg and placebo of –6.1 (95% confidence interval [CI], –8.9 to –3.3; P < 0.0001) in study RGH-MD-31, and –4.3 (95% CI, –6.7 to –1.9; P = 0.0004) in study RGH-MD-32. In study RGH-MD-33, the difference in score between cariprazine 3 mg to 6 mg and placebo was –6.1 (95% CI, –8.4 to –3.8; P < 0.001). The within-group change in the YMRS score was clinically meaningful for both the cariprazine and placebo treatment groups, based on a within-group minimal important difference (MID) of 6.6 points.⁷ All sensitivity analyses supported the results of the primary efficacy analyses. Subgroup analyses based on the pivotal trials did not suggest a difference in efficacy between patients experiencing a manic episode and patients experiencing a mixed episode. In the studies of acute bipolar depression, treatment with cariprazine 1.5 mg was associated with a greater reduction of symptoms of depression relative to placebo, based on a least squares mean (LSM) difference in the MADRS total score at week 6 of –2.5 (95% CI, –4.6 to –0.4; P = 0.0417) in study RGH-MD-53, –2.5 (95% CI, –4.6 to –0.4; P = 0.0331) in study RGH-MD-54, and –4.0 (95% CI, –6.3 to –1.6; P = 0.0030) in study RGH-MD-56. The comparison of cariprazine 3.0 mg to placebo did not consistently demonstrate a benefit with treatment. In study RGH-MD-54, the LSM difference was in favour of cariprazine 3.0 mg relative to placebo (LSM difference = –3.0; 95% CI, –5.1 to –0.9; P = 0.0103). The LSM difference in study RGH-MD-53 and study RGH-MD-56 was –1.8 (95% CI, –3.9 to 0.4; P = 0.1051) and –2.5 (95% CI, –4.9 to –0.1; P = 0.1122), respectively. A between-group MID of at least 2 points was identified for the MADRS total score.^8,9 Comparisons that were statistically significant were also clinically meaningful as per the MID, with the exception of cariprazine 3.0 mg in study RGH-MD-56, which was not statistically significant.

The secondary end point in all studies was the change in CGI-S, which is an outcome based on a global assessment of symptom severity relative to other patients that the assessor has observed. The CGI-S was validated through a comparison to the PANSS in patients with schizophrenia, but evidence of reliability and responsiveness were not identified. In the acute mania studies, the cariprazine treatment groups exhibited a greater change in overall severity based on the CGI-S at week 3 than the placebo treatment groups, which is consistent with the primary analysis. This was based on a difference in score between cariprazine 3 mg to 12 mg and placebo of –0.6 (95% CI, –1.0 to –0.3; P = 0.0001) in study RGH-MD-31, and –0.4 (95% CI, –0.7 to –0.1; P = 0.0027) in study RGH-MD-32. In study RGH-MD-33, the difference in score between cariprazine 3 mg to 6 mg and placebo was –0.6 (95% CI, –0.9 to –0.4; P < 0.001). In the acute depression studies, a benefit in terms of the change from baseline to week 6 in the CGI-S was demonstrated for comparisons of cariprazine 1.5 mg to placebo in study RGH-MD-53 (LSM difference = –0.3; 95% CI, –0.6 to –0.1; P = 0.0417) and study RGH-MD-56 (LSM difference = –0.4; 95% CI, –0.6 to –0.1; P = 0.0132). The LSM difference of the change from baseline in CGI-S at week 6 was –0.2 (95% CI, –0.4 to 0.1; P = 0.1370), –0.3 (95% CI, –0.5 to –0.0; P = 0.0662), and –0.3 (95% CI, –0.5 to –0.0; P = 0.1122) for study RGH-MD-53, study RGH-MD-54, and study RGH-MD-56, respectively. Similar to the primary analysis, cariprazine 3.0 mg did not demonstrate a benefit relative to placebo, nor did the 1.5 mg treatment group in study RGH-MD-54 (LSM difference = –0.2; 95% CI, –0.5 to 0.0; P = 0.0714). The suggested MID for the CGI-S is a difference of 1 point for both within-group and between-group analyses. Based on this threshold, a clinically meaningful within-group difference was observed for all treatment groups (except placebo in study RGH-MD-31); however, the between-group differences were not clinically meaningful.

Other assessments of the change in symptoms were reported in the included studies. A lack of control for multiplicity of testing, however, rendered additional efficacy outcomes as supportive only.

In the studies of acute mania, the overall improvement of symptoms using the CGI-I, the change in symptoms of depression using the MADRS total score, and the change in severity of psychotic symptoms using the PANSS were assessed. The results of the CGI-I at week 3 were supportive of the primary and secondary analyses. The LSM difference of the change from baseline in CGI-I at week 3 for cariprazine compared to placebo was –0.8 (95% CI, –1.2 to –0.5; P < 0.0001), –0.5 (95% CI, –0.7 to –0.2; P = 0.0004), and –0.7 (95% CI, –0.9 to –0.4; P < 0.001) in study RGH-MD-31, study RGH-MD-32, and study RGH-MD-33, respectively. An MID was not identified for the CGI-I, which made it difficult to interpret this outcome further. The results of the change in MADRS total score at week 3 were consistent with a small reduction of depressive symptoms, with no apparent difference between cariprazine and placebo. The clinical experts consulted by CADTH did not expect to see a change in depressive symptoms for patients experiencing an acute manic episode, but also suggested that not observing an increase in symptoms of depression is notable. A numerical reduction in the PANSS total score was observed at week 3 in all treatment groups (cariprazine and placebo) in the bipolar mania studies; however, limitations of the PANSS outcome led to difficulty with the interpretation of the results in the context of short-term treatment for acute mania associated with bipolar disorder.

The bipolar depression studies also assessed changes in severity of symptoms of depression using the HAM-D and QIDS-SR tools and symptoms of anxiety using the HAM-A tool. A clinically significant difference of 2 points or 3 points in the HAM-D has been suggested in the literature, although justification for this threshold was unclear and likely opinion-based.^8,10 The difference in the change in depressive symptoms based on the HAM-D at week 6 was inconsistent across studies, although a numerical decrease in HAM-D score (reduction of symptoms) was observed for all treatment groups. Also, study RGH-MD-56 reported similar results at week 6 and week 8. Neither cariprazine 1.5 mg nor 3.0 mg demonstrated a benefit relative to placebo based on the change in depressive symptoms at week 6 as per the QIDS-SR. A numerical reduction in the HAM-A score at week 6 indicating an improvement of symptoms of anxiety was observed in all treatment groups, with no difference observed between cariprazine 3.0 mg and placebo. The clinical relevance of these changes is unknown. Overall, the evidence in support of changes in the severity of symptoms associated with a depressive episode of bipolar disorder were inconsistent with respect to the difference between cariprazine 1.5 mg and placebo and did not support a benefit with cariprazine 3.0 mg relative to placebo.

The incidence of most severe suicidal ideation and most severe suicidal behaviour per the C-SSRS was reported in all studies except study RGH-MD-31. In all treatment groups of the bipolar mania studies, suicidal ideation was reported in 1% to 3% of patients. In the bipolar depression studies, suicidal ideation was reported in 3% to 11% of patients. The incidence by dose of cariprazine varied between studies. Suicidal behaviour was not reported for any patients in any of the included studies.

Daily functioning was also identified as an outcome of interest to patients and by clinicians. |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. Functioning was not assessed in any of the other included studies.

Response rates and remission rates based on reductions in the YMRS total score (acute mania studies) and MADRS total score (depressive studies) were also reported. These outcomes were also not controlled for multiplicity and, therefore, were considered supportive only. In the acute mania studies, a response was observed in 48% to 61% of patients receiving cariprazine and 25% to 44% of patients receiving placebo. Remission was observed in 42% to 52% of patients receiving cariprazine and 23% to 35% of patients receiving placebo. The clinical experts indicated that a trial duration of 3 weeks is likely too short to observe full remission, which may take up to 4 weeks to 6 weeks for an acute manic or mixed episode. In the acute depressive studies, 41% to 50% of patients and 43% to 52% of patients receiving cariprazine 1.5 mg and 3.0 mg, respectively, were considered responders as per the MADRS definition. The response rate among patients receiving placebo ranged from 32% to 40% across the trials. The benefit of treatment with cariprazine (1.5 mg and 3.0 mg) was inconsistently demonstrated across the studies based on this outcome. Similar results were reported for the analysis of MADRS remitters, where 26% to 37% of patients and 26% to 32% of patients receiving cariprazine 1.5 mg and 3.0 mg, respectively, and 20% to 23% of patients receiving placebo were considered to have a remission of depressive symptoms.

Outcomes related to health-related quality of life (HRQoL), hospitalization, cognitive impairment, and persistence with therapy were included in the CADTH systematic review protocol but not identified in the included studies.

Harms Results

In the bipolar mania studies, the percentage of patients who reported a treatment-emergent adverse event (TEAE) ranged from 78% to 86% for patients randomized to cariprazine (3 mg to 12 mg or 3 mg to 6 mg) and 61% to 79% for patients randomized to placebo. In the bipolar depression studies, the percentage of patients who reported an AE ranged from 50% to 62% for patients randomized to cariprazine 1.5 mg, 49% to 62% for patients randomized to cariprazine 3.0 mg, and 46% to 55% for patients randomized to placebo. The overall rate of AEs was higher in the acute mania trials than the acute depression trials, despite a shorter duration of treatment. This may be due to the use of a higher dose of cariprazine. Whether AEs were more likely to be detected in an inpatient setting or the result of the higher dose of cariprazine in the acute mania trials, or both, is unknown.

In all included studies, serious adverse events (SAEs) were infrequently reported and 1 death was reported in all of the included trials. In the bipolar mania studies, SAEs were reported by 3.2% to 4.2% of patients randomized to cariprazine (3 mg to 12 mg or 3 mg to 6 mg) and 1.9% to 4.2% of patients randomized to placebo. In the bipolar depression studies, the percentage of patients who reported at least 1 SAE in the cariprazine 1.5 mg and cariprazine 3 mg treatment groups ranged from 0.6% to 1.4% and 0 to 1.4%, respectively. In the placebo treatment groups, the percentage of patients who reported at least 1 SAE ranged from 1.3% to 3.4%. In the bipolar mania studies, patients who stopped treatment due to AEs (withdrawal due to adverse event [WDAE]) ranged from 9% to 14% of patients randomized to cariprazine and 5% to 10% of patients randomized to placebo. In the bipolar depression studies, WDAEs were reported by 3% to 8% of patients randomized to cariprazine 1.5 mg, 6% to 12% of patients randomized to cariprazine 3.0 mg, and 3% to 10% of patients randomized to placebo. The most common reasons for WDAEs were mania, akathisia, restlessness, and agitation. WDAEs tended to be slightly more frequent among patients randomized to cariprazine compared to placebo, but this was inconsistent across studies.

Of the notable harms identified in the CADTH systematic review protocol, hyperglycemia, weight gain, sexual dysfunction, tardive dyskinesia, and neuroleptic malignant syndrome were infrequently reported in all trials. Additionally, vomiting and EPS were infrequently reported in the bipolar depression studies. In the acute mania studies, the following notable harms were reported more frequently with cariprazine versus placebo: EPS (10% to 25% versus 2% to 10%), akathisia (17% to 22% versus 4% to 6%), vomiting (8% to 10% versus 3% to 5%), and restlessness (6% to 8% versus 1% to 5%). Similarly in the bipolar depression studies, akathisia (5% to 14% versus 1% to 3%) and restlessness (2% to 7% versus 3% to 4%) were more common in the cariprazine groups versus the placebo group and occurred more frequently with cariprazine 3.0 mg than cariprazine 1.5 mg (akathisia was 5% to 6% versus 6% to 14% and restlessness was 1% to 3% versus 6% to 7% for the 1.5 mg dose group versus the 3.0 mg dose group, respectively). The frequency of AEs during a short treatment period of 3 weeks in the acute mania studies and up to 8 weeks in the acute depression studies, and with the high discontinuation rates in the studies, is notable, although comparable to other treatments for bipolar disorder as indicated by the clinical experts consulted by CADTH.

Weight gain (an increase in body weight of at least 7%) was captured in the safety assessment of vital signs. While the duration of the trials may not have been long enough to sufficiently assess the impact of the treatment of cariprazine on weight gain, this outcome was still observed in the trials. However, there was only a slight difference in weight gain between the cariprazine and placebo treatment groups.

Critical Appraisal

Appropriate methods of randomization were used, although there was the potential for unblinding or knowledge of treatment received due to AEs — most notably EPS and akathisia, which were more common in the cariprazine treatment groups relative to placebo. Treatment groups were well balanced in terms of baseline characteristics. Discontinuation rates were high in all included studies, but generally aligned with expectations for clinical trials for acute episodes of bipolar disorder. Overall discontinuation rates were balanced between treatment groups in most studies, except study RGH-MD-54 and study RGH-MD-56 where discontinuation was notably higher among the cariprazine 3.0 mg group. Discontinuation was also higher for patients in the placebo treatment group compared to cariprazine 1.5 mg in study RGH-MD-56. The imbalance in discontinuation rates in study RGH-MD-56 appears to be driven by discontinuation due to AEs and withdrawal of consent. In the bipolar mania studies, there was an imbalance in the reason for discontinuation due to AEs (more frequent with cariprazine) and insufficient therapeutic response (more frequent with placebo), which may have biased the safety and efficacy results in favour of cariprazine. Missing data were handled using the last observation carried forward (LOCF) approach (study RGH-MD-31) and the mixed model of repeated measures (MMRM) approach (all other studies). Both methods rely on the assumption that data are missing at random, which is likely not the case given the imbalance in reasons for discontinuation that have been described. However, the sponsor conducted a pattern-mixture model analysis that relies on the assumption that data are not missing at random. The results of the pattern-mixture model analysis were consistent with the results of the primary analysis. All studies implemented methods to control for multiplicity up to the secondary end point, which was the change from baseline in the CGI-S at week 3 (bipolar mania studies) and week 6 (bipolar depression studies). Study RGH-MD-31 and study RGH-MD-32 used a closed testing procedure to control for the type I error rate; study RGH-MD-33 and the 3 bipolar depression studies used a matched parallel gatekeeping procedure. The analyses of the primary and secondary outcomes were the only outcomes controlled for multiplicity. Therefore, all other efficacy outcomes are at risk of type I error and viewed as supportive evidence only.

The clinical experts described the patient population included in the trials for cariprazine as typical for clinical trials, but lacking characteristics that are often seen in clinical practice such as the presence of comorbidities, rapid cycling, other diagnoses, comorbid substance use disorder, and elevated risk of suicide. The exclusion criteria can lead to a less severe and less complex sample relative to clinical practice. The clinical experts indicated that the baseline characteristics were suggestive of a patient population with bipolar I disorder of moderate severity. Overall, clinical experts did not have any major issues with the generalizability of the study populations to Canadian patients experiencing acute episodes associated with bipolar I disorder based on the baseline characteristics considered as a whole. In study RGH-MD-31 and study RGH-MD-32, the permitted dose of cariprazine ranged from 3 mg to 12 mg daily, which extends beyond the Health Canada–approved dose of up to 6 mg cariprazine daily. As such, specific conclusions regarding the effects of the Health Canada–approved dosing cannot be drawn based on these 2 studies. However, study RGH-MD-33 provided evidence for cariprazine at a dose that was consistent with the Health Canada indication (3 mg to 6 mg treatment group), which demonstrated a similar treatment effect to the higher doses used in study RGH-MD-31 and study RGH-MD-32. Dosing was consistent with the approved indication in all other studies included in this review. The studies for bipolar mania involved rapid titration that is inconsistent with how this drug is expected to be used by most patients treated in an outpatient setting. Generalizing the efficacy, safety, and tolerability outcomes observed in the trials to an outpatient setting for the treatment of acute mania is associated with some uncertainty. Another consideration for the intervention is that the clinical experts indicated it is unlikely for cariprazine to be used only for the acute management of episodes associated with bipolar I disorder. While the duration of the trials was considered adequate to observe a treatment effect on an acute episode, they were too short to properly assess safety and tolerability or efficacy beyond 3 weeks (bipolar mania studies) and 6 weeks to 8 weeks (bipolar depression studies). Lastly, all of the outcomes used in the studies are commonly used in clinical trials or research settings for bipolar I disorder. None of the included outcomes are designed for or typically used in clinical practice as per feedback from the clinical experts.

Indirect Comparisons

Description of Studies

One network meta-analysis (NMA) submitted by the sponsor¹⁷ and 2 published NMAs^18,19 that examined the comparative efficacy, safety, and tolerability of acute treatments for bipolar I disorder were included in this report. All of the NMAs conducted a systematic review of RCTs in adults either with acute bipolar mania (Kishi et al. [2021]),¹⁸ acute bipolar depression (Kadakia et al. [2021]),¹⁹ or both populations (sponsor-submitted NMA).¹⁷ The sponsor-submitted NMA performed separate analyses for the bipolar mania and depression populations. All of the included NMAs specified treatments for acute episodes of bipolar I disorder |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. Note that Kadakia et al. was limited to atypical antipsychotic drugs and recent publications (since May 2015). The sponsor-submitted NMA was conducted using a |||||||||||||||||||||||||||, the publication by Kishi et al. performed both pairwise and frequentist NMAs using a random effects (RE) model, and Kadakia et al. performed an NMA with a Bayesian approach where the base case was based on an RE model.

Efficacy Results

A total of ||||||||| studies were included in the sponsor-submitted NMA: ||||||||| studies in the manic or mixed bipolar I disorder patient population and ||||||||| studies in the depressive bipolar I disorder patient population. |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.

For the acute mania NMAs, ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. The analysis of change from baseline in the YMRS score ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| but this network was associated with evidence of inconsistency, statistical heterogeneity, and imprecision of the results, and, therefore, uncertainty about any conclusions that could be drawn.

For the acute depression NMAs, MADRS response |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. Further, the relative treatment effect in terms of |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. The analysis of MADRS remission ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. Of note, the presence of inconsistency for this network is unknown as inconsistency was not formally assessed. Additionally, it is unknown whether variability in the baseline MADRS score influenced the results of the NMA. The analysis of change from baseline in MADRS |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.

The results suggest weight gain (greater than 7%) |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. However, these results were considered highly uncertain due to potential heterogeneity, a lack of precision (mania NMA), and bias resulting from a small number of events. Therefore, there is uncertainty about the conclusion of ||||||||||||||||||||||||||| in terms of weight gain.

Overall, the sponsor-submitted NMA |||||||||||||||||||||||||||||||||||| for treatment with cariprazine relative to other comparators of interest for the treatment of acute manic or mixed episodes or acute depressive episodes.

Harms Results

In the bipolar mania NMAs, the analysis of the rate of EPS as well as sedation or somnolence |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. In the bipolar depression NMAs, ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||. This comparison was limited to aripiprazole and lurasidone as comparators. The analysis of the rate of sedation or somnolence suggested ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| in the rate of sedation or somnolence was reported for other comparisons of cariprazine 1.5 mg or cariprazine 3.0 mg and other treatments.

In the bipolar mania NMAs, all-cause discontinuation and discontinuation due to AEs were limited by evidence of inconsistency and substantial statistical heterogeneity. The results of the analysis suggest cariprazine ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.

Critical Appraisal

Studies that included patients with a dual diagnosis of substance use disorder, or that investigated an intervention not used as monotherapy, were excluded from the systematic review, which may have resulted in missing potentially relevant patients and interventions. The study design was limited to double-blind RCTs and the quality of the studies was assessed using the National Institute for Health and Care Excellence (NICE) checklist and the Jadad rubric. It is unclear how the quality assessment was used and the Jadad rubric is not considered to be a reliable tool for assessing study quality. An insufficient quality assessment may have resulted in the inclusion of low-quality trials. Additionally, a sensitivity analysis for the quality of data was not performed.

Variation in the health care setting — particularly among the studies for manic or mixed episodes and publication year, which ranged from 1991 to 2019 — are potential sources of heterogeneity among the included studies. Baseline patient characteristics, including age, sex, race, HAM-D score, MADRS score, YMRS score, CGI-S score, and time since diagnosis, lacked important details and were subject to a large amount of missing data, hindering the ability to conduct a robust assessment of heterogeneity in the study populations. As a result, no studies were excluded based on outliers in the baseline characteristics and it is unknown if the NMA was impacted by heterogeneity among the included patient populations. Inconsistency was assessed using a node-splitting approach, which is an appropriate statistical assessment of inconsistency, although it does not incorporate information from the entire network in the analysis. Evidence of inconsistency was identified in the following networks of the manic or mixed episodes studies: change from baseline in YMRS, all-cause discontinuation, and discontinuation due to AEs. For the NMAs of outcomes in depressive studies, the author of the sponsor-submitted report stated that inconsistency could not be assessed for the following networks: MADRS remission, weight gain, EPS, and all-cause discontinuation. The author reported that evidence of inconsistency was not identified for the remaining outcomes in the networks. However, very wide 95% CIs for the inconsistency factor of comparisons in the NMA for the following outcomes may suggest otherwise: sedation or somnolence (manic or mixed and depressive), all-cause discontinuation (manic or mixed), and discontinuations due to AEs (manic or mixed and depressive).

Missing data were an issue for certain outcomes, in part due to a small number of studies in the NMAs. This issue was compounded by the fact that there were few events per study for some outcomes, which was the case for the analysis of EPS (both the manic or mixed network and the depressive network) and weight gain (the depressive network). |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.

The 2 published NMAs summarized for this review^18,19 were limited by insufficiently reported data and details about the methodology used, as well as the low quality of evidence informing the networks. As a result, this summary has focused on the sponsor-submitted NMA. Briefly, the results of the published NMAs were generally consistent with the sponsor-submitted NMA, although subject to similar and additional limitations.

Other Relevant Evidence

Two studies were included as other relevant evidence for the review of cariprazine. One was study RGH-MD-36,²⁰ a long-term, open-label study that examined the long-term safety and tolerability of cariprazine 3 mg to 12 mg daily in patients with bipolar mania, and 1 was a post hoc analysis (McIntyre et al. [2019])²¹ that provided additional efficacy data on cariprazine in the subpopulation of patients with bipolar mania with mixed features.

Study RGH-MD-36

Description of Study

Patients were eligible to enrol in study RGH-MD-36 if they were not currently taking any treatment or had a documented history of intolerance or inadequate response to their current therapy. They were treated with a flexible dose of cariprazine for up to 16 weeks of treatment. During the screening period and for the first 2 weeks and up to 3 weeks of open-label treatment, all patients were hospitalized. At the end of week 3, all patients were discharged and followed as outpatients. Patients were discontinued from the study if they presented with clinical instability (by the end of week 3), with tolerability concerns, with worsening of symptoms, or with inadequate response, or if the investigator deemed it appropriate at any time to discontinue a patient.

All patients were required to have a total score of 18 or more on the YMRS and a total score of less than 18 on the MADRS. Further, a body mass index between ||||||||| kg/m² and ||||||||| kg/m², inclusive, was required. In comparison to the inclusion criterion regarding YMRS total score, the pivotal trials in bipolar mania (study RGH-MD-31,¹¹ study RGH-MD-32,¹² and study RGH-MD-33¹³) used a YMRS total score of 20 or more, and a score of 4 or more on 2 of the 4 following YMRS items: irritability, speech, content, and disruptive/aggressive behaviour.

Of the 403 patients who enrolled in the long term open-label study, a total of 402 patients received at least 1 dose of open-label cariprazine (safety population). The mean age of patients in the study was 41.4 (SD = 10.5) years. The majority of patients was male (57.2%) and White (51.2%). At baseline, the mean weight and body mass index was 86.5 (SD = 17.8) kg and 29.2 (SD = 5.3) kg/m², respectively. The mean age at onset and the known duration of bipolar I disorder was 27.9 (SD = 11.3) years and ||||||||| (SD = |||||||||) months, respectively. The duration of the current manic episode for the majority of patients (53.2%) was greater than 21 days.

Efficacy Results

A total of 132 (32.8%) patients completed the study and ||||||||| patients entered safety follow-up. The most frequently reported reason for discontinuation during the open-label treatment period was withdrawal of consent (19.7%), followed by AE (16.4%), and protocol violation (13.7%).

The mean change from baseline to week 16 in YMRS total score was –15.2 (SD = 9.2) and the mean change from baseline to week 16 in MADRS total score was –1.6 (SD = 7.5). At week 16, YMRS response criteria (≥ 50% reduction from baseline) was met by 64.2% of patients and YMRS remission criteria (total score ≤ 12) was met by 63.4% of patients.

Harms Results

TEAEs were reported in 335 (83.3%) patients during the open-label treatment. The most commonly reported AEs (frequency ≥ 10%) were akathisia (32.6%), headache (16.7%), constipation (10.7%), and nausea (10.4%). SAEs were reported in 30 (7.5%) patients. The following SAEs were reported in more than 1 patient: the worsening of mania in 9 (2.2%) patients, depression in 5 (1.2%) patients, akathisia in 3 (0.7%) patients, suicidal ideation in 2 (0.5%) patients, and suicide attempt in 2 (0.5%) patients. The most severe suicidal ideation and suicidal behaviour per the C-SSRS was reported in 35 (8.8%) patients and 3 (0.8%) patients, respectively. No deaths were reported in the safety population. Premature discontinuation due to at least 1 AE was reported in 66 (16.4%) patients during the open-label treatment. The most frequently cited reason was akathisia in 19 (4.7%) patients and depression in 6 (1.5%) patients.

The most commonly reported notable harms (frequency ≥ 5%) included akathisia in 131 (32.6%) patients, insomnia in 28 (7.0%) patients, EPS in 27 (6.7%) patients, restlessness in 26 (6.5%) patients, vomiting in 24 (6.0%) patients, sedation in 23 (5.7%) patients, and increase in weight in 23 (5.7%) patients. A total of ||||||||| patients with at least 1 AE was related to extrapyramidal symptoms during open-label treatment. During the open-label treatment, 129 (32.1%) patients required treatment for extrapyramidal symptoms, of which 74 (18.4%) patients used a beta-blocking drug (propranolol or propranolol hydrochloride), 64 (15.9%) patients used an anti-Parkinson drug (benztropine mesylate or biperiden), and ||||||||| patients used a psycholeptic drug (diphenhydramine hydrochloride, diphenhydramine, or zolpidem tartrate).

Critical Appraisal

In the absence of an active comparator or placebo group, the interpretation of the efficacy results from the long-term open-label study RGH-MD-36 is limited. This is compounded by the use of descriptive statistics only. The use of the LOCF approach could overestimate or underestimate the overall long-term treatment benefits, particularly given the very high rates of discontinuation in the open-labelled study. Patients were discontinued from the study if they presented with clinical instability by the end of week 3, with any tolerability concerns, with worsening of symptoms, or with inadequate response, or if the investigator deemed it appropriate at any time to discontinue a patient. Consequently, the resultant population may be more tolerant of cariprazine, which could potentially lead to an underreporting of AEs and a response bias as patients with an inadequate response (defined as an increase in the YMRS or MADRS total score by 30% or more at the end of week 2 or thereafter) were prematurely discontinued from the study.

The clinical experts consulted by CADTH stated that the exclusion of patients with, for example, rapid cycling and active substance use disorder can lead to patients with complex cases who are seen in clinical practice being missed in the study. However, the clinical experts recognized that clinical trials will typically use said exclusion criteria to avoid confounding variables. There was a notable discontinuation rate of greater than 50%, which decreases the certainty and generalizability of the efficacy and safety results. According to the clinical experts, a discontinuation rate of approximately 35% is typically anticipated for clinical trials in bipolar mania. Further, some patients in the study received a dose higher than the Health Canada–recommended daily dose of cariprazine, which is up to 6 mg per day.

Post Hoc Analysis

Description of Study

Data from 3 pivotal trials of cariprazine in adult patients with acute manic or mixed episodes associated with bipolar I disorder were pooled and used for the post hoc analysis. The objective of the post hoc analysis was to determine the effect of cariprazine on manic and depressive symptoms versus placebo in the subpopulation of patients with mania and subsyndromal depressive features.

A total of 1,037 patients were pooled from the pivotal trials. The number of patients who met the DSM-5 criteria for mixed state (≥ 3 depressive symptoms) and the 2 proxy definitions for mixed episode (≥ 2 depressive symptoms and a MADRS total score ≥ 10) was 141 (13.6%), 269 (25.9%), and 453 (43.7%), respectively.

Results

The pooled placebo and active treatment groups showed an improvement in the mean YMRS total score at week 3 relative to baseline. The difference between cariprazine and placebo in change in mean YMRS total score was –3.79 (standard error [SE] = not reported [NR]; P = 0.0248), –2.91 (SE = NR; P = 0.0207), and –5.49 (SE = NR; P < 0.0001) in patients with mixed features as defined by 3 or more depressive symptoms, 2 or more depressive symptoms, and a MADRS total score of 10 or more, respectively, in favour of cariprazine.

The results of the change in mean MADRS total score at week 3 relative to baseline were inconsistent based on the definition used for patients with mixed features. There was a benefit with cariprazine based on the MADRS total score of 10 or more and no difference was observed using the other 2 definitions. The difference between cariprazine and placebo in change in mean MADRS total score was –1.59 (SE = NR; P < 0.0082) in patients with mixed features as defined by a MADRS total score of 10 or more, in favour of cariprazine.

The proportion of responders (≥ 50% improvement from baseline in the YMRS total score) was higher for cariprazine in the 2 or more depressive symptoms group (47%; P = 0.0483) and the MADRS total score of 10 or more group (57%; P =  < 0.0001) than in the placebo group (34% and 31%, respectively). There was no difference between the cariprazine and placebo treatment groups (P = 0.2608), based on the 3 or more depressive symptoms definition for patients with mixed features.

The proportion of remitters (YMRS total score ≤ 12) was higher for cariprazine in the 2 or more depressive symptoms group (39%; P = 0.0462) and the MADRS total score of 10 or more group (44%; P =  < 0.0001) than in the placebo group (27% and 23%, respectively). There was no difference between the cariprazine and placebo treatment groups (P = 0.1224) based on the 3 or more depressive symptoms definition for patients with mixed features.

Critical Appraisal

The pooled, post hoc analysis was summarized to supplement the evidence for patients experiencing mixed episodes associated with bipolar I disorder. The pooled analysis is subject to the same limitations of the bipolar mania studies included in the systematic review, in addition to a small sample size and a lack of power to detect a difference between treatment groups. Moreover, given the subgroups of interest were not included as stratification variables at randomization, differences in baseline characteristics between the groups would be expected to introduce bias into the results observed. Overall, the results of this analysis should be considered exploratory.

Conclusions

Six RCTs informed the systematic review of cariprazine. These included the 3 pivotal trials for the acute treatment of mania or mixed episodes of bipolar I disorder (study RGH-MD-31, study RGH-MD-32, and study RGH-MD-33) and 3 pivotal trials for the acute treatment of depressive episodes of bipolar I disorder (study RGH-MD-53, study RGH-MD-54, and study RGH-MD-56). The studies of acute mania or mixed episodes evaluated a flexible dose of cariprazine ranging from 3 mg to 12 mg once daily, and the studies of acute depressive episodes evaluated a fixed dose of cariprazine 1.5 mg and 3.0 mg once daily.

All 3 studies of acute mania demonstrated that treatment with cariprazine was associated with a statistically significant and clinically meaningful greater reduction in symptoms of acute mania relative to placebo, based on the YMRS total score after 3 weeks of treatment. This was supported by the results of the secondary end point, which demonstrated a greater reduction in overall disease severity after 3 weeks based on the CGI-S. The studies of acute bipolar depression demonstrated that treatment with cariprazine 1.5 mg was associated with a statistically significant and clinically meaningful greater reduction in symptoms of depression relative to placebo, based on the change in the MADRS total score at week 6. A dose response was not observed as cariprazine 3.0 mg did not consistently demonstrate a benefit relative to placebo, since the primary end point was not met in 2 of the 3 pivotal trials. The results of the secondary end point — change from baseline to week 6 in the CGI-S — were consistent with the primary analysis, except for 1 study (study RGH-MD-54) that did not demonstrate a benefit with cariprazine 1.5 mg relative to placebo. Additional efficacy outcomes were not controlled for multiplicity and should only be considered as supportive evidence. Additional outcomes were generally consistent with the primary and secondary analyses.

The frequency of AEs due to EPS, akathisia, vomiting, and restlessness were considered relevant safety concerns, although consistent with the safety profile of similar treatments. Both safety and efficacy outcomes may have been impacted by the high discontinuation rates across studies due to AEs, insufficient therapeutic response, and withdrawal of consent, which may have overestimated the reported results. Additionally, the short duration of the trials limited the ability to sufficiently evaluate certain outcomes such as weight gain. Direct comparisons to other available treatments were not identified, but uncertain evidence from 3 NMAs suggests that treatment with cariprazine offers ||||||||| relative to other second-generation antipsychotic drugs and relevant treatments as per YMRS and MADRS outcomes. Cariprazine may exhibit a different tolerability profile, but a conclusion about this cannot be drawn due to the uncertainty of the available evidence.

Introduction

Disease Background

Living with bipolar disorder is associated with a substantial burden of illness that people experience for approximately half of their lives.¹ Bipolar disorder frequently manifests in late adolescence and young adulthood, with an overall average age of onset of 25 years.¹ The typical age of onset differs geographically; it tends to be earlier in the US (18 years to 20 years) and Canada (22.5 years) and later in European countries (29 years).^1-3 The estimated lifetime prevalence of bipolar I disorder in Canada is 0.87%, affecting males and females equally.¹ Quality of life and psychosocial functioning (e.g., the ability to maintain proper work) are severely impacted by bipolar disorder and more pronounced in those living with depressive symptoms, in those with multiple previous episodes or a longer duration of illness, and in those with cognitive impairment.¹ In addition, suicide risk is substantially higher for patients living with bipolar disorder relative to the general population (10.7 deaths due to suicide per 100,000 per year).¹

Bipolar disorder is a mood disorder characterized by episodes of mania, hypomania, and major depression,³ and can be divided into 2 main diagnostic subtypes — namely, bipolar I disorder and bipolar II disorder. Bipolar I disorder involves the presence of at least 1 full manic episode, with features that include inflated self-esteem, the decreased need for sleep, pressured speech, racing thoughts, distractibility, psychomotor agitation, and risky behaviour.¹ Bipolar II disorder involves a hypomanic episode, in which manic symptoms are brief or less severe and do not significantly impact functional impairment, hospitalization, or psychosis.

Episodes of mania and depression present with significant changes in mood, energy, behaviour, sleep, and cognition. Mania also presents with change in activity. To diagnose mania as per the DSM-5 criteria, abnormal and persistent elevated, expansive, or irritable mood, and abnormal and persistent increased activity or energy must present for most of the day, every day for 1 week minimum. In addition, the episode must have at least 3 of the symptoms of mania listed in Table 4. Also, the mood disturbance must lead to marked impairment in functioning, require hospitalization, or be accompanied by psychotic features.^1,22 Episodes of bipolar depression are characterized by a minimum of 2 weeks of depressed mood and/or the inability to feel pleasure, and at least 4 of the symptoms listed in Table 4, as per the DSM-5.^1,22 The DSM-5 includes specifiers that describe characteristics of bipolar disorder and can be used to guide treatment decisions for acute mania and depression. The presence of mixed features is a specifier, which describes a complex presentation of bipolar disorder.¹ To qualify for the mixed features specifier, a patient must be experiencing at least 3 nonoverlapping symptoms from the opposing pole (mania or depression) during the majority of the days of the current episode.⁴

Table 4: Symptoms of Bipolar Mania and Depression

The diagnosis of bipolar disorder is challenging for many reasons, including a lack of insight from the individual into the psychopathology and functional impairment they are experiencing, the frequent onset of depressive states, and high rates of comorbidity, which all can lead to variation in help-seeking behaviour for hypomanic or manic periods.^1,22 Misdiagnosis and delayed diagnosis are relatively common. Often, patients are not accurately diagnosed and may be treated for other disorders, especially unipolar depression, for up to 10 years from the onset of symptoms.¹ It is very common for patients diagnosed with bipolar disorder to have comorbid psychiatric diagnoses, which increases the complexity of the illness and difficulty of making an accurate diagnosis. Common comorbidities include disorders of substance use, impulse control, anxiety, and personality.¹ The clinical experts consulted by CADTH noted there are no diagnostic tools or tests that are useful, and the diagnosis is based on a clinical assessment centred on the criteria that has been described.

Standards of Therapy

There are no disease-modifying treatments for bipolar disorder. The clinical experts stated that the primary goal of current treatments is to improve acute symptoms and, by doing so, restore functioning. The prevention of recurrent mood episodes was described as the secondary goal, which can be achieved over the short-term to medium term (e.g., 6 months to 12 months) with current treatments, although more often for manic episodes than depressive episodes. Long-term efficacy is more unknown as high-quality longitudinal studies of longer duration are limited.

Bipolar disorder is treated with the combination of pharmacologic, nonpharmacologic (e.g., electroconvulsive therapy), and psychosocial (e.g., psychotherapy) treatments. In Canada, pharmacologic treatments are usually managed by family physicians in milder to moderate cases (which tend to be predominantly chronically depressed patients), or by psychiatrists in more severe cases (more presentations of mania or rapid cycling). However, the experts noted that there is a lot of regional variability as psychiatrists tend to be concentrated in urban settings.

Medications include mood stabilizers (which include drugs from a variety of classes, such as second-generation antipsychotic drugs, anticonvulsants, and lithium) and antidepressants. The 2018 CANMAT and ISBD guidelines list recommendations for the treatment of mania with considerations for safety and tolerability concerns, as efficacy is comparable among first-line options.¹ For example, lithium is recommended unless there is a reason not to, such as mixed features, comorbid substance use, or previous nonresponse to lithium.¹ The guidelines do not indicate a preference for monotherapy, noting that about 20% more patients respond to combination therapy based on clinical trial data than with monotherapy. The clinical experts described the use of lithium with other drugs as common in clinical practice. The 2018 CANMAT guidelines provide more guidance for the optimization of therapy for patients experiencing bipolar depression. The guidance takes into consideration current and prior medication use and response, personal preference, safety and tolerability, and clinical features that may influence prognosis. The clinical experts stated that overall, there are many different options of medications, with no consensual first or best option in clinical practice. The choice is usually guided by characteristics of the patient, including, but not limited to, predominant symptomatology or polarity (mania, depression, or mixed features), age, income, and medical and psychiatric comorbidities. Of note, the 2018 CANMAT guidelines list cariprazine as the last or second-last option for first-line monotherapy treatment of acute mania and acute depression, while also noting that safety concerns have a limited impact and tolerability concerns have a minor impact on the selection of cariprazine for treatment. The clinical experts suggested that the lack of evidence for the use of cariprazine as maintenance therapy is a limitation of its use. However, they also noted that a benefit of cariprazine is the potential to treat both acute mania and acute depression. Additionally, while the 2021 CANMAT guidelines specific to bipolar disorder with mixed presentations do not include any first-line treatment options due to insufficient evidence in this population, cariprazine is recommended as a second-line treatment option for acute mania with mixed features.⁴

The clinical experts indicated that the use of off-label drugs for mania is less common as there are several drugs approved for use in Canada with level 1 evidence of efficacy for the treatment of mania. For reference, the 2018 CANMAT guidelines define level 1 data as evidence based on a meta-analysis with a narrow CI or replicated, double-blind RCT that includes a placebo or active control comparison with at least 30 participants in each active treatment arm.¹ In contrast, the experts indicated that there are only a few drugs with Health Canada approval and/or level 1 evidence of efficacy for the management of depressive episodes, and they are associated with relatively high rates of treatment failure due to insufficient response or poor tolerability. As a result, the use of drugs without approval for bipolar depression is common. For example, the experts noted that antidepressants are widely used in the real world, despite limited evidence of their efficacy and the fact that most treatment guidelines only recommend them as second-line or third-line options. Given the overall low response and remission rates of pharmacologic therapy in bipolar disorder, most patients use combinations of 2 to 3 different drugs — usually of mechanistically distinct drugs (e.g., an antipsychotic, lithium, and an antidepressant). Monotherapy is preferred, but harder to sustain in the real world.

Drug

Cariprazine is an atypical antipsychotic drug that has been approved by Health Canada for use in adults as monotherapy for the acute management of manic or mixed episodes associated with bipolar I disorder (bipolar mania) and the acute management of depressive episodes associated with bipolar I disorder (bipolar depression).⁵ The mechanism of action for cariprazine is unknown; however, it may be mediated through partial agonist activity at the central dopamine D3 and D2 receptors, and serotonin 5-HT1A receptors. Cariprazine also has antagonist activity at serotonin 5-HT2A receptors.⁵ Cariprazine forms 2 major metabolites, desmethyl cariprazine and didesmethyl cariprazine, that have in vitro receptor binding profiles similar to the parent drug. The drug and its active metabolites have an extended half-life of 2 days to 4 days for cariprazine, 1 day to 2 days for desmethyl cariprazine and 1 week to 3 weeks for didesmethyl cariprazine.⁵

Cariprazine is available as 1.5 mg, 3 mg, 4.5 mg, and 6 mg oral capsules administered once daily.⁵ The recommended dosage of cariprazine depends on the type of bipolar disorder — bipolar mania or bipolar depression. For bipolar mania, the suggested initial dosage is 1.5 mg to 6 mg once daily. Depending on clinical response and tolerability, subsequent dose increases may be made in 1.5 mg increments to a maximum of 6 mg daily for bipolar mania, although the lowest effective dose should be used.⁵ For bipolar depression, the recommended dosage is 1.5 mg once daily, which may be increased to 3 mg daily on day 15 depending on clinical response and tolerability. The lowest effective dose should be used and the maximum recommended dosage for bipolar depression is 3 mg daily.⁵ Due to the long half-life of the drug and its active metabolites, changes in dosage will not be fully reflected in the plasma for several weeks; thus, treatment response and the occurrence of adverse effects may be delayed following initiation and after a dosage change.⁵

The sponsor has requested reimbursement as per the indication.⁶ Cariprazine has been previously reviewed by CADTH for the treatment of schizophrenia in adults. Cariprazine is also approved by Health Canada for the treatment of schizophrenia in adults.

Table 5: Key Characteristics of Treatments for Bipolar I Disorder

BD = bipolar disorder; BD1 = bipolar I disorder; DSM-III-R = Diagnostic and Statistical Manual of Mental Disorders, Third Edition, Revised.

^aHealth Canada–approved indication for acute treatment.

Source: Product monographs for cariprazine,⁵ aripiprazole,²³ asenapine,²⁴ lurasidone,²⁵ olanzapine,²⁶ paliperidone,²⁷ quetiapine,²⁸ risperidone,²⁹ ziprasidone,³⁰ lithium,³¹ divalproex,³² carbamazepine,³³ haloperidol,³⁴ clonazepam,³⁵ and chlorpromazine.³⁶

Stakeholder Perspectives

The information in this section is a summary of input provided by the patient groups and clinician groups that responded to CADTH’s call for patient input and from clinical experts consulted by CADTH for the purpose of this review.

Patient Group Input

This section was prepared by CADTH staff based on the input provided by patient groups.

Two responses to CADTH’s call for patient input were received for this review: a submission from IAM and a submission from MDSC. IAM and MDSC are organizations that support individuals living with mental illness such as bipolar disorder, including patients, their families, and service providers.

The information used to inform the IAM submission was based on 2 online surveys conducted in 2018 and February 2022. Potential respondents to these online surveys included members of the IAM and Hope and Me – Mood Disorders Association of Ontario client networks. Among the respondents of the 2018 survey, 12% self-described as personally diagnosed, 50% were caregivers, 63% were family members or friends of someone diagnosed, and 18% worked in social services. Among the respondents of the 2022 survey, 33% identified as an individual living with symptoms of bipolar disorder, 58% were relatives, 8% were caregivers of someone with lived experience, and 1% did not specify. MDSC drew information from interviews with patients and family members, 2 national mental health surveys conducted in 2018 and September 2021, and shared experiences that have been posted on the MDSC online discussion forum. The interviews were conducted with 5 patients with bipolar disorder, including semi-structured phone interviews with adults living with bipolar disorder in January 2021, and 3 family members.

Respondents indicated that living with bipolar disorder has impacted their mental health, social relationships, and day-to-day functioning at work and school. Patients can also experience a lack of insight into their illness, which impacts their ability and motivation to seek treatment and causes significant strain in relationships. Survey respondents described the advantages of taking medications for bipolar disorder, which include managing symptoms of bipolar disorder, experiencing fewer episodes of mental illness, and avoiding visits to the hospital. They also described disadvantages of taking medications for bipolar disorder; these include requiring frequent follow-ups with health care providers, needing to take the medication daily, and dealing with the symptoms — including bipolar depression, which was not being well controlled. Further, respondents described the benefits of an injectable formulation, which include convenience and not needing to remember to take it daily, while the difficulties were pain at injection site and frequent travel to clinics. The most common side effects of medications for bipolar disorder identified by respondents were drowsiness, dry mouth, restlessness, and weight gain. Respondents also identified the cost of medications as a significant barrier to access.

Survey respondents reported that treatment of bipolar disorder is individualized, as not every patient will respond to 1 medication. To find the right medication that enables the highest degree of functioning while minimizing side effects, patients with bipolar disorder often have to go through a trial-and-error process. This process involves taking a number of different medications and at different dosages until their goals of therapy have been achieved. This process can make it challenging for patients to adhere to their prescribed regimen and can be exacerbated by additional challenges such as waiting to be approved for coverage by public drug programs and experiencing relapse. As a result, patients feel that outcomes can be improved by increasing equitable access to, and the selection of, medications that are reimbursable. According to respondents, antipsychotic medications can be improved by increasing their ability to control the symptoms of bipolar disorder, improving the side effect profile, and providing a greater range of strengths and dosages to lower the frequency of administration.

A copy of the patient input from IAM and MDSC are presented in Appendix 5.

Clinician Input

Input From Clinical Experts Consulted by CADTH

All CADTH review teams include at least 1 clinical specialist with expertise regarding the diagnosis and management of the condition for which the drug is indicated. Clinical experts are a critical part of the review team and are involved in all phases of the review process (e.g., providing guidance on the development of the review protocol, assisting in the critical appraisal of clinical evidence, interpreting the clinical relevance of the results, providing guidance on the potential place in therapy). The following input was provided by 2 clinical specialists with expertise in the diagnosis and management of bipolar I disorder.

Unmet Needs

The clinical experts indicated that there are a number of challenges with addressing treatment goals for patients living with bipolar I disorder using currently available treatments. Insufficient response to treatment was noted as being very common, particularly in bipolar depression. Estimates vary, but around 50% of patients do not respond sufficiently well to the first option, and about 30% to 40% of patients will go on to develop treatment resistance (e.g., failure to respond to at least 3 drugs). As mentioned earlier, there are no disease-modifying treatments for bipolar disorder and the long-term effectiveness of current treatments is unclear. More often than not, bipolar disorder is a lifelong, persistent and/or highly recurrent condition. There are key outcomes that are not adequately addressed by current treatments. For example, the extent to which current treatments address cognition directly, instead of indirectly through mood symptoms, is unclear. Tolerability is a problem, particularly in bipolar depression, where the most effective drugs, like quetiapine and olanzapine, have well-known metabolic side effects, which are particularly concerning given the elevated, and independent, risk of metabolic and cardiovascular conditions in this population. Finally, adherence is also an issue; poor compliance is very common in real-world settings. All common first-line options for bipolar depression (e.g., quetiapine, lurasidone, lithium, lamotrigine) are daily oral medications, with no depot alternatives.

Place in Therapy

Based on the currently available clinical evidence, the clinical experts felt it was very unlikely that cariprazine would cause a shift in treatment paradigm, despite having a somewhat distinct pharmacologic profile. The experts indicated that the overall effect sizes, response and remission rates, and frequency of side effects of cariprazine are broadly similar to currently available treatments. Nonetheless, the experts noted that given the combination of efficacy for both acute mania and depression, as well as an acceptable tolerability profile, it is expected that cariprazine will eventually be used as a first-line treatment. Additionally, the experts did not identify any reason to recommend that patients try other treatments before cariprazine. The clinical experts suggested cariprazine will probably be used in monotherapy and in combination with either lithium or anticonvulsants, which is a standard approach with the use of second-generation antipsychotic drugs in bipolar disorder. In addition, the clinical experts noted that cariprazine was not developed based on a model of the pathophysiology of bipolar disorder; it is mechanistically very similar to currently available treatment and, therefore, very much within the current paradigm of symptomatic management.

Patient Population

At this time, the clinical experts stated that it is not possible to identify patients who are most likely to exhibit a response to treatment with cariprazine. They noted that reliable prediction of response has not been achieved even with sophisticated research methods (e.g., genomics, neuroimaging), and certainly cannot be done reliably with the most widely available clinical tools. Other than polarity of mood episode, there are no other validated indices to predict response to any given drug, according to the clinical experts. They also referred to several studies that have assessed potential clinical and biologic predictors of response to pharmacologic treatment, but noted the findings are mostly inconsistent and nonspecific. Further, the experts indicated that even the presence of psychotic symptoms is not a reliable predictor, as patients with no history of psychosis are known to potentially respond to antipsychotic drugs. In practice, the experts stated that the choice of treatment is guided by patients’ characteristics and preference but is fundamentally a trial-and-error approach. Overall, the experts had not identified anything that would indicate specific subpopulations are more or less likely to respond to cariprazine based on the currently available clinical data.

The experts indicated that it is challenging to identify and diagnose patients with bipolar disorder, and that misdiagnosis and/or delayed diagnosis are relatively common. Further, the experts noted that it is well-documented that most patients are initially treated as patients with unipolar depression — even for years, in some situations — before receiving the diagnosis of bipolar disorder. Relatedly, they stated there are no diagnostic tools or tests that are useful, and the diagnosis is based on clinical assessment.

The clinical experts stated that presymptomatic patients should not be treated with cariprazine, or any other drug, mostly due to the low predictive power of current assessment tools (i.e., it is not possible to predict the development of bipolar disorder with a high degree of confidence — even in a person with a confluence of well-known risk factors, such as first-degree family history, early onset of symptoms, or history of trauma).

The clinical experts did not identify any patients who would be least suited for treatment with cariprazine, noting that within the population of adults with bipolar I disorder, there were no major contraindications (other than the usual for this population) for cariprazine.

Assessing Response to Treatment

The clinical experts identified YMRS, MADRS, and HAM-D as the most commonly used outcomes to assess response to treatment in research settings, largely due to the fact that they are clinician-administered. They also reported that in the real world, patient-rated questionnaires like the Patient Health Questionnaire and the Beck Depression Inventory are more common. The experts noted that all of these outcomes, however, have very good concurrent validity.

A reduction in the frequency or severity of symptoms, improvement in symptoms, stabilization (no deterioration) of symptoms, ability to perform activities of daily living, and improved survival were all clinically meaningful according to the clinical experts. They described a reduction in the severity of symptoms (e.g., controlling physical agitation) and frequency of symptoms as the usual first goal of treatment. The experts described research studies that typically use a threshold of improvement of 25% to 50%, but noted that most clinicians rely on subjective reports; therefore, the magnitude that is considered meaningful does vary a lot across physicians. Subsequently, the experts stated that treatment aims to restore functioning — including improving cognitive functioning and coping skills — with a return to work, school, and/or daily activities being an important benchmark. The experts described long-term goals, such as the prevention of relapses and recurrences, as well as the onset and progression of psychiatric comorbidities (e.g., anxiety disorders, substance abuse) and medical comorbidities (e.g., obesity, diabetes).

Regarding how often treatment response should be assessed, the clinical experts suggested that in the acute phase, response is usually assessed between 2 weeks and 4 weeks, depending on the severity. In the maintenance phase, it is recommended that patients be assessed at least every 2 months to 3 months.

Discontinuing Treatment

The clinical experts indicated that treatment discontinuation is determined by either a lack of response or poor tolerability, noting that most guidelines recommend discontinuing a treatment if there is no response to very poor response within 4 weeks to 6 weeks. Further, they indicated that discontinuation due to tolerability depends on the severity and progression of specific side effects — particularly EPS and akathisia, which are the most common side effects of cariprazine and similar drugs. These side effects can be time-dependent; they tend to be worse during initial titration phases but improve with time. If side effects are moderate to severe and/or do not meaningfully improve in 1 week to 2 weeks, treatment should be discontinued.

Prescribing Conditions

The clinical experts reported that family physicians can and frequently do diagnose bipolar disorder and regularly prescribe similar drugs. The experts also noted that there are no special tests required for the diagnosis of bipolar disorder, nor for the prescription and monitoring of cariprazine. Given the high prevalence of bipolar disorder and the relative lack of psychiatrists in Canada, a significant proportion of patients are treated by family physicians; therefore, requiring the involvement of specialists would significantly restrict the use of this medication.

Clinician Group Input

This section was prepared by CADTH staff based on the input provided by patient groups.

Two clinician groups provided input to this review: CANMAT and WC-CAN. One clinician on behalf of CANMAT and 6 clinicians with the WC-CAN contributed to these submissions. Both clinician groups recognized the unmet need for a medication that is effective in multiple phases of bipolar disorder, including bipolar depression, with low rates of AEs to minimize polypharmacy and improve adherence. Both clinician groups advocated for cariprazine as a first-line treatment option for patients with bipolar disorder in the treatment of acute mania and depression and to be used as monotherapy and possibly as combination therapy with other mood stabilizers.

Drug Program Input

The drug programs provide input on each drug being reviewed through CADTH’s reimbursement review processes by identifying issues that may impact their ability to implement a recommendation. The implementation questions and corresponding responses from the clinical experts consulted by CADTH are summarized in Table 6.

Table 6: Summary of Drug Plan Input and Clinical Expert Response

CDEC = CADTH Canadian Drug Expert Committee; HAM-D = Hamilton Depression Rating Scale; MADRS = Montgomery–Åsberg Depression Rating Scale; YMRS = Young Mania Rating Scale.

Clinical Evidence

The clinical evidence included in the review of cariprazine is presented in 3 sections. The first section, the systematic review, includes pivotal studies provided in the sponsor’s submission to CADTH and Health Canada, as well as those studies that were selected according to an a priori protocol. The second section includes indirect evidence from the sponsor and indirect evidence selected from the literature that met the selection criteria specified in the review. The third section includes sponsor-submitted long-term extension studies and additional relevant studies that were considered to address important gaps in the evidence included in the systematic review.

Systematic Review (Pivotal and Protocol Selected Studies)

Objectives

To perform a systematic review of the beneficial and harmful effects of cariprazine 1.5 mg, 3 mg, 4.5 mg, and 6 mg for use as monotherapy in adults for the acute management of manic or mixed episodes associated with bipolar I disorder (bipolar mania) and for the acute management of depressive episodes associated with bipolar I disorder (bipolar depression)

Methods

Studies selected for inclusion in the systematic review included pivotal studies provided in the sponsor’s submission to CADTH and Health Canada, as well as those meeting the selection criteria presented in Table 7. Outcomes included in the CADTH review protocol reflect outcomes considered to be important to patients, clinicians, and drug plans.

Of note, the systematic review protocol presented as follows was established before the granting of a Notice of Compliance from Health Canada.

Table 7: Inclusion Criteria for the Systematic Review

AE = adverse event; HRQoL = health-related quality of life; RCT = randomized controlled trial; SAE = serious adverse event; WDAE = withdrawal due to adverse event.

^aMay be used alone or in combination with other treatments (e.g., lithium, divalproex).

The literature search for clinical studies was performed by an information specialist using a peer-reviewed search strategy according to the PRESS Peer Review of Electronic Search Strategies checklist.³⁷

Published literature was identified by searching the following bibliographic databases: MEDLINE All (1946–) via Ovid, Embase (1974–) via Ovid, and APA [American Psychological Association] PyscINFO (1806‒) via Ovid. All Ovid searches were run simultaneously as a multifile search. Duplicates were removed using Ovid deduplication for multifile searches, followed by manual deduplication in Endnote. The search strategy consisted of both controlled vocabulary, such as the US National Library of Medicine’s MeSH (Medical Subject Headings), and keywords. The main search concept was Vraylar (cariprazine). Clinical trials registries were searched: the US National Institutes of Health’s ClinicalTrials.gov, the WHO’s International Clinical Trials Registry Platform search portal, Health Canada’s Clinical Trials Database, and the European Union Clinical Trials Register.

No filters were applied to limit the retrieval by study type. Retrieval was not limited by publication date or by language. Conference abstracts were excluded from the search results. Refer to Appendix 1 for the detailed search strategies.

The initial search was completed on March 8, 2022. Regular alerts updated the search until the meeting of the CADTH Canadian Drug Expert Committee on June 22, 2022.

Grey literature (literature that is not commercially published) was identified by searching relevant websites from the Grey Matters: A Practical Tool For Searching Health-Related Grey Literature checklist.³⁸ Included in this search were the websites of regulatory agencies (US FDA and European Medicines Agency). Google was used to search for additional internet-based materials. Refer to Appendix 1 for more information on the grey literature search strategy.

These searches were supplemented by contacting the sponsor of the drug for information regarding unpublished studies.

Two CADTH clinical reviewers independently selected studies for inclusion in the review based on titles and abstracts, according to the predetermined protocol. Full-text articles of all citations considered potentially relevant by at least 1 reviewer were acquired. Reviewers independently made the final selection of studies to be included in the review, and differences were resolved through discussion.

Findings From the Literature

A total of 388 studies were identified from the literature for inclusion in the systematic review (Figure 1). The included studies are summarized in Table 8. No studies were excluded (Appendix 2).

Figure 1: Flow Diagram for Inclusion and Exclusion of Studies

Table 8: Details of Included Studies for Bipolar Mania

CGI-I = Clinical Global Impression–Improvement; CGI-S = Clinical Global Impression–Severity of Illness; DB = double-blind; DSM-IV-TR = Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision; ECT = electroconvulsive therapy; EPS = extrapyramidal symptom; MADRS = Montgomery–Åsberg Depression Rating Scale; NMS = neuroleptic malignant syndrome; PANSS = Positive and Negative Syndrome Scale; RCT = randomized controlled trial; TSH = thyroid-stimulating hormone; UDS = urine drug screen; YMRS = Young Mania Rating Scale.

Note: Three additional reports were included — Durgam et al. (2015),³⁹ Sachs et al. (2015),⁴⁰ and Calabrese et al. (2015).⁴¹

^a3 of the study centres were discontinued before any patients had been consented and screened; 29 centres participated in the study.

^bDSM-IV-TR Axis I diagnoses include delirium, dementia, amnestic disorder, and other cognitive disorders, as well as schizophrenia, schizoaffective disorder, and other psychotic disorders.

Source: Clinical Study Reports for study RGH-MD-31, study RGH-MD-32, and study RGH-MD-33.^11-13

Table 9: Details of Included Studies for Bipolar Depression

BMI = body mass index; BD = bipolar disorder; CGI-S = Clinical Global Impression–Severity of Illness; DSM-IV-TR = Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision; DSM-5 = Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition; ECT = electroconvulsive therapy; FAST = Functioning Assessment Short Test; HAM-A = Hamilton Anxiety Rating Scale; HAM-D = Hamilton Depression Rating Scale; MADRS = Montgomery–Åsberg Depression Rating Scale; MINI = Mini International Neuropsychiatric Interview; QIDS-SR = Quick Inventory of Depressive Symptomatology–Self-Rated; YMRS = Young Mania Rating Scale.

Note: Three additional reports were included — Earley et al. (2020),⁴² Earley et al. (2019),⁴³ and Durgam et al. (2016).⁴⁴

^aExcluded from report.

Source: Clinical Study Reports for study RGH-MD-53, study RGH-MD-54, and study RGH-MD-56.^14-16

Description of Studies

A summary of the details of included studies for bipolar mania and bipolar depression is available in Table 8 and Table 9, respectively.

Three pivotal trials for the treatment of acute manic or mixed episodes were identified for this review. These included 1 phase II study (study RGH-MD-31 [N = 238]) and 2 phase III studies (study RGH-MD-32 [N = 312] and study RGH-MD-33 [N = 497]). The objective of each of the 3 bipolar mania studies was to evaluate the efficacy, safety, and tolerability of cariprazine monotherapy versus placebo for the treatment of acute manic or mixed episodes associated with bipolar I disorder. Each of the studies was a multi-centre, randomized, double-blind, placebo-controlled, parallel-group, flexible-dose study in adults with a primary diagnosis of bipolar I disorder. Patients were enrolled in study RGH-MD-31 between 2007 and 2008, and in study RGH-MD-32 and study RGH-MD-33 between 2010 and 2011. None of the pivotal trials included study centres located in Canada.

Patients were randomized at a 1:1 ratio to 1 of 2 treatment groups in study RGH-MD-31 and study RGH-MD-32 — cariprazine 3 mg to 12 mg (flexible dose) or placebo. In study RGH-MD-33, patients were randomized at a 1:1:1 ratio to 1 of 3 treatment groups — cariprazine 3 mg to 6 mg, cariprazine 6 mg to 12 mg, or placebo. Randomization was not stratified in any of the studies. Data have not been presented for the cariprazine 6 mg to 12 mg group of study RGH-MD-33 as the dosage above 6 mg per day is not aligned with the Health Canada–approved indication.⁵

Each of the studies included a no-drug washout period during the screening phase (up to 4 days to 7 days) followed by a 3-week double-blind treatment period and a 2-week follow-up period. All patients started hospitalization during the screening phase and remained hospitalized for a minimum of 2 weeks (14 days) following the start of the double-blind treatment period. After 2 weeks, patients were permitted to be discharged and followed as outpatients if the following criteria were met: the severity of illness was mild or less (based on a score of 3 or less on the CGI-S), there was no significant risk of suicide or violent behaviour (based on investigator clinical judgment), and the patient was ready for discharge (based on investigator opinion). Patients who were not eligible for discharge could be re-evaluated any time after day 14 and before day 21. Additionally, following transitioning to an outpatient setting, patients could be re-hospitalized or placed in a day treatment program if necessary.

Three pivotal trials for the treatment of acute depressive episodes were identified for this review. These included 1 phase IIb study (study RGH-MD-56 [N = 578]) and 2 phase III studies (study RGH-MD-53 [N = 493] and study RGH-MD-54 [N = 488]). The objective of each of the 3 bipolar depression studies was to evaluate the efficacy, safety, and tolerability of cariprazine 1.5 mg per day and 3 mg per day versus placebo in patients with bipolar I depression. Each of the studies was a multi-centre, randomized, double-blind, placebo-controlled, parallel-group, fixed-dose study in adults with a primary diagnosis of bipolar I disorder. Patients were enrolled in study RGH-MD-31 between 2007 and 2008, and in study RGH-MD-32 and study RGH-MD-33 between 2010 and 2011. None of the study centres in study RGH-MD-53 or study RGH-MD-54 were located in Canada, while 3 centres in study RGH-MD-56 were located in Canada.

Patients were randomized at a 1:1:1 ratio to 1 of 3 treatment groups in study RGH-MD-53 and study RGH-MD-54 — cariprazine 1.5 mg per day, cariprazine 3.0 mg per day, or placebo. In study RGH-MD-56, patients were randomized at a 1:1:1:1 ratio to 1 of 4 treatment groups — cariprazine 0.75 mg per day, cariprazine 1.5 mg per day, cariprazine 3.0 mg per day, or placebo. Randomization was not stratified in any of the studies. Data have not been presented for the cariprazine 0.75 mg group of study RGH-MD-56 as the dosage is not aligned with the Health Canada–approved indication.⁵

Each of the studies included a 2-week no-drug washout period during the screening phase, followed by a 6-week or, in the case of study RGH-MD-56 only, 8-week double-blind treatment period and a 1-week follow-up period. In study RGH-MD-53 and study RGH-MD-54, patients may have been hospitalized for up to 7 days during the screening period or during the first week of the double-blind treatment period. In study RGH-MD-56, patients may have been hospitalized for up to 2 weeks following the start of the double-blind treatment period.

Populations

Inclusion and Exclusion Criteria

A list of inclusion and exclusion criteria for bipolar mania and bipolar depression is available in Table 8 and Table 9, respectively.

All studies enrolled adult patients (18 to 65 years of age) with bipolar I disorder with or without psychotic symptoms as per the criteria of the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR) (studies for bipolar mania) or without psychotic features as per the DSM-5 (studies for bipolar depression). Diagnosis was confirmed by the administration of either the Structured Clinical Interview for DSM [Diagnostic and Statistical Manual of Mental Disorders] or the Mini International Neuropsychiatric Interview. In the bipolar mania studies, patients with comorbid diagnoses such as conduct disorder, obsessive-compulsive disorder, anxiety disorders, and substance abuse were eligible for enrolment.

In the bipolar mania studies, patients were required to have a YMRS total score of at least 20, a score of at least 4 on 2 of the 4 YMRS items (irritability, speech, content, and disruptive/aggressive behaviour), and a MADRS total score of less than 18. Patients were also required to be voluntarily hospitalized with a primary diagnosis of mania. In the bipolar depression studies, patients were required to have a HAM-D total score of at least 20, a HAM-D item 1 score of at least 2, and a CGI-S score of at least 4. Patients also needed to have had a previous manic or mixed episode, verified by either treatment with an antipsychotic medication with an approved indication for mania, hospital or medical records, or a patient report corroborated by a caretaker or a previous or current treating clinician. Additionally, patients had be in the process of being treated as an outpatient, with a current major depressive episode of at least 4 weeks and not exceeding 12 months in duration.

All studies excluded patients if they had had another DSM-IV-TR or DSM-5 Axis I diagnosis or Axis II diagnosis severe enough to interfere with participation, or alcohol or substance abuse or dependence within the prior 3 months (studies for bipolar mania) or prior 6 months (studies for bipolar depression), suicide risk, or significant medical conditions that could interfere with the study or endanger the patient’s well-being. Additionally, patients who had received electroconvulsive therapy within 3 months of screening or a depot neuroleptic within 3 months of screening (studies for bipolar mania) or 6 months of screening (studies for bipolar depression) were excluded.

In the bipolar mania studies, patients were excluded if they were experiencing their first manic episode or met the criteria for rapid cycling. In the bipolar depression studies, patients were excluded if they had a YMRS total score of greater than 12 (or greater than 10 in study RGH-MD-56), 4 or more episodes of a mood disturbance (depression, mania, hypomania, or mixed state) within 12 months, or a history of nonresponse to 2 or more antidepressant trials of adequate dose and treatment duration with or without mood stabilizers in the current depressive episode. Additionally, patients were excluded from study RGH-MD-53 and study RGH-MD-54 if they experienced treatment failure of quetiapine, lurasidone, or Symbyax (fluoxetine and olanzapine) in the treatment of bipolar depression during the current depressive episode. Patients were excluded from study RGH-MD-56 if they had a history of nonresponse in the current depressive episode to 2 or more adequate treatment trials with Symbyax (fluoxetine and olanzapine), quetiapine (including monotherapy), lithium (including monotherapy), or a mood stabilizer (lithium, valproate, lamotrigine, carbamazepine, or oxcarbazepine) in combination with an antidepressant.

Baseline Characteristics

A summary of baseline characteristics in the bipolar mania studies is provided in Table 10.

In the bipolar mania studies, the mean age of patients was between 36 (SD = 11) years and 43 (SD = 12) years, between 54% and 68% of patients were male, and the mean weight of patients ranged from 70 (SD = 20) kg to 82 (SD = 16) kg. The distribution of patients by race varied between the 3 studies. In study RGH-MD-31, 40% to 47% of patients were White, 26% to 31% of patients were Black, and 24% to 25% of patients were Asian. In study RGH-MD-32, 57% to 58% of patients were Asian, and in study RGH-MD-33, 70% to 71% of patients were White. Patients with ongoing metabolic disorders ranged from 9% to 18% in the bipolar mania studies. In study RGH-MD-33, ongoing metabolic disorders were reported in 18% of patients randomized to cariprazine and 10% of patients randomized to placebo. The duration of bipolar I disorder was a mean of 10 (SD = 9) years to 15 (SD = 10) years and the age of onset ranged from 23 (SD = 8) years to 28 (SD = 10) years. During the trials, between 81% and 91% of patients were currently experiencing a manic episode and between 10% and 20% of patients were experiencing a mixed episode. The mean baseline efficacy scores are presented in Table 11. At baseline, the mean YMRS total score ranged from 30 (standard error of the mean = 0.5) to 33 (SD = 6), and the mean CGI-S score ranged from 4.6 (SD = 0.6) to 4.8 (SD = 0.7). Additionally, the mean MADRS total score at baseline ranged from 8.2 (SD = 4.1) to 9.6 (SD = 3.7) and the PANSS mean score at baseline ranged from 60.2 (standard error of the mean = 1.3) to 63.0 (SD = 14.9).

A summary of baseline characteristics in the bipolar depression studies is provided in Table 12. In the bipolar depression studies, the mean age of patients was between 41 (SD = 11) years and 44 (SD = 13) years, between 57% and 65% of patients were female, the mean weight of patients ranged from 80 (SD = 17) kg to 87 (SD = 25) kg, and 72% to 78% of patients were White. Between ||||||||| and ||||||||| of patients had a history of metabolic disorders. The duration of and age of onset of bipolar I disorder was not reported in study RGH-MD-53 or study RGH-MD-54. In study RGH-MD-56, the duration of bipolar I disorder ranged from 14.6 (SD = 9.5) years to 15.5 (SD = 10.3) years and the age of onset ranged from 25.4 (SD = 10.2) years to 28.4 (SD = 11.4) years. During study RGH-MD-53 and study RGH-MD-54, between 11% and 28% of patients were currently experiencing a severe depressive episode and between 71% and 89% were experiencing a moderate depressive episode, and the duration of the current episode was a mean of 3.5 (SD = 2.5) months to 3.9 (SD = 2.6) months. The proportion of patients with mild, moderate, or severe depressive episode (current or most recent) were not reported in study RGH-MD-56. The mean efficacy scores at baseline are presented in Table 13. At baseline, the mean MADRS total score ranged from 30.3 (SD = 4.4) to 31.5 (SD = 4.8), the mean CGI-S score ranged from 4.4 (SD = 0.5) to 4.5 (SD = 0.6), and the mean HAM-D score at baseline ranged from 23.9 (SD = 3.2) to 24.9 (SD = 2.9). Where available, the HAM-A score, QIDS-SR score, and FAST total score were also summarized (Table 13).

A summary of relevant prior medications used by patients in the bipolar mania studies and bipolar depression studies is available in Table 14 and Table 15, respectively.

Interventions

Study interventions cariprazine and placebo were administered orally, as a single dose, once daily at approximately the same time in all studies. As the studies were double-blind, cariprazine and placebo were available as capsules identical in appearance. Patients received treatment for 3 weeks in the bipolar mania studies, for 6 weeks in study RGH-MD-53 and study RGH-MD-54, and for 8 weeks in study RGH-MD-56. All investigational products were dispensed in blister packs, with 1 provided for each week of the treatment period. All blister packs included enough capsules for 7 days of treatment plus 3 extra days (up to 10 days of treatment). In the bipolar mania studies, blister packs were dispensed on day 0, day 7, and day 14. In the bipolar depression studies, 1 to 2 blister packs were dispensed at day 0, and at the end of week 1, week 2, and week 4. In the 8-week RGH-MD-56 study, 2 blister packs were also dispensed at week 6. In all studies, patients were instructed to return all unused study drugs at each study visit.

Bipolar Mania Studies

In study RGH-MD-31 and study RGH-MD-32, patients were randomized to 1 of 2 treatment groups: cariprazine 3 mg to 12 mg or placebo. In study RGH-MD-33, patients were randomized to 1 of 3 treatment groups: cariprazine 3 mg to 6 mg, cariprazine 6 mg to 12 mg, or placebo. All products (cariprazine and placebo) were identical in appearance to maintain blinding. The daily dose for each patient ranged from 1 capsule to 4 capsules in study RGH-MD-31 and study RGH-MD-32 or 1 capsule to 3 capsules in study RGH-MD-33. In all studies, treatment was administered to patients by hospital staff while the patients were hospitalized. For patients who were discharged early, the patient or a caregiver was responsible for ensuring the study drug was taken as directed.

In the bipolar mania studies, the intervention was based on a flexible-dose regimen. Patients in the cariprazine treatment groups received cariprazine 1.5 mg per day on day 0 and cariprazine 3 mg per day on day 1, each provided as 1 capsule. The dose could be increased to 2 3 mg capsules (a total of 6.0 mg per day) on day 2 if the patients were experiencing an inadequate response without tolerability problems, based on the judgment of the investigator. Dose increases differed slightly between study RGH-MD-31 and study RGH-MD-32 beginning on day 3. In study RGH-MD-31, the dose could be increased by 1 capsule to 3 3 mg capsules (9.0 mg per day) on day 3, and starting on day 4, the dose could be increased by 1 capsule to a maximum of 4 3 mg capsules (12.0 mg per day), depending on response and tolerability. In study RGH-MD-32 (Figure 22), the dose could be increased to 6.0 mg per day on day 2 due to inadequate response without tolerability problems, based on investigator judgment. Patients remained on the dose administered at day 2 for 2 days (day 2 and day 3). On day 4, inadequate responders were identified using the YMRS (less than 50% improvement in YMRS from day 2 to day 4). The dose was increased to 9.0 mg per day on day 4 for 3 days (day 4 to day 6) for patients with an inadequate response without tolerability issues. Additional increases at increments of 3 mg per day were determined using the YMRS assessment of improvement to a maximum of 12 mg per day. In study RGH-MD-33 (Figure 23), all patients randomized to receive cariprazine were administered cariprazine 1.5 mg per day on day 0 and cariprazine 3 mg per day on day 1. Those in the 3 mg per day to 6 mg per day group remained on the 3 mg per day dosage on day 2, while patients in the 6 mg to 12 mg per day treatment group received 6.0 mg on day 2. Beginning on day 3, the approach to dosing titration was similar to study RGH-MD-32 in that dose increases were made based on an assessment of inadequate response as per the YMRS and in the absence of tolerability issues. Dose increases were considered at day 5, day 7, and day 10. Dose increases were made at increments of 1.5 mg per day in the cariprazine 3 mg per day to 6 mg per day group (up to 1 or 2 additional 1.5 mg capsules, corresponding to a total of 4.5 mg or 6.0 mg per day, respectively), and 3 mg per day in the cariprazine 6 mg per day to 12 mg per day group (up to 1 or 2 additional 3 mg capsules, corresponding to a total of 9 mg per day or 12 mg per day, respectively).

In all studies, dose increases or decreases were made in increments of 1 capsule as described. A decrease in the dose was considered if there were tolerability problems. Alternatively, patients could skip the dose for 1 day to 2 days (study RGH-MD-31) or 1 day to 3 days (study RGH-MD-32 and study RGH-MD-33). In study RGH-MD-32 and study RGH-MD-33, dose adjustments (an increase or decrease except for a temporary drug discontinuation for 1 day to 3 days) were not permitted after day 14. Each of the studies indicated that frequent switching was not allowed; additional detail was not provided. Of note, adjustments to the number of capsules administered (dose adjustments) were assessed and implemented for patients in the placebo groups in a similar manner to the cariprazine groups.

Bipolar Depression Studies

In study RGH-MD-53 and study RGH-MD-54, patients were randomized to 1 of 3 treatment groups: cariprazine 1.5 mg, cariprazine 3.0 mg, or placebo. In study RGH-MD-56, patients were randomized to 1 of 4 treatment groups: cariprazine 0.75 mg (not shown), cariprazine 1.5 mg, cariprazine 3.0 mg, or placebo.

In study RGH-MD-53 and study RGH-MD-54, all patients randomized to cariprazine dose groups received cariprazine 1.5 mg per day for 2 weeks, from day 1 through day 14. For patients randomized to the cariprazine 3 mg per day group, the dose was increased to 3 mg per day on day 15. In study RGH-MD-56, all patients randomized to cariprazine treatment groups received cariprazine 0.5 mg per day on day 1 and day 2. The dose was increased to 0.75 mg per day on day 3, then 1.0 mg per day on day 5, and 1.5 mg per day on day 8. Patients randomized to the cariprazine 3 mg treatment group had another dose increase to 3 mg on day 15 of treatment.

If there were tolerability issues, temporary discontinuation of treatment for up to 3 consecutive days was permitted at the discretion of the investigator. Patients unable to tolerate the fixed dose of treatment or patients off drug for 4 or more consecutive days were prematurely discontinued from the study.

Concomitant Therapy and Rescue Medications

Concomitant therapy and rescue medications that were permitted for use in the included studies are described in Table 16. Additional psychotropic medications, including psychostimulants, were not allowed — with some exceptions. Patients were permitted to use the following for insomnia: zolpidem, zolpidem extended release, zaleplon, chloral hydrate, and eszopiclone. For EPS that emerged or worsened during the study, patients could be treated with diphenhydramine, benztropine or equivalent, or propranolol for the treatment of akathisia. For agitation, restlessness, irritability, and hostility, patients could be treated with lorazepam, oxazepam, or diazepam.

Outcomes

A list of efficacy end points identified in the CADTH review protocol that were assessed in the clinical trials included in this review is provided and further summarized in Table 17. A detailed discussion and critical appraisal of the outcome measures is provided in Appendix 4.

Table 16: Concomitant and Rescue Medications Permitted During the Trials

AE = adverse event; AIMS = Abnormal Involuntary Movement Scale; BARS = Barnes Akathisia Rating Scale; EPS = extrapyramidal symptom; SAS = Simpson-Angus Scale.

^aMaximum of 2 mg per day in study RGH-MD-31 and maximum of 3 mg per day in study RGH-MD-32 and study RGH-MD-33.

Source: Clinical Study Reports for study RGH-MD-31, study RGH-MD-32, study RGH-MD-33, study RGH-MD-53, study RGH-MD-54, and study RGH-MD-56.^11-16

Table 17: Summary of Outcomes of Interest Identified in the CADTH Review Protocol

CGI-I = Clinical Global Impression–Improvement; CGI-S = Clinical Global Impression–Severity of Illness; FAST = Functioning Assessment Short Test; HAM-A = Hamilton Anxiety Rating Scale; HAM-D = Hamilton Depression Rating Scale; MADRS = Montgomery–Åsberg Depression Rating Scale; NA = not applicable; PANSS = Positive and Negative Syndrome Scale; QIDS-SR = Quick Inventory of Depressive Symptomatology–Self-Rated; YMRS = Young Mania Rating Scale.

Source: Clinical Study Reports for study RGH-MD-31, study RGH-MD-32, study RGH-MD-33, study RGH-MD-53, study RGH-MD-54, and study RGH-MD-56.^11-16

Young Mania Rating Scale (Change in Symptom Severity, Response Rate, and Remission Rate)

The YMRS is an 11-item, clinician-administered rating scale used to assess the severity of manic symptoms.⁴⁵ A severity rating is assigned to each item, based on the patient’s subjective report of his or her condition over the past 48 hours and the clinician’s assessment of the patient’s behaviour. Item 5, item 6, item 8, and item 9 (irritability, speech, content, and disruptive/aggressive behaviour) are graded on a 0 to 8 scale while the remaining 7 items (elevated mood, increased motor activity-energy, sexual interest, sleep, language-thought disorder, appearance, and insight) are graded on a 0 to 4 scale, based on increasing severity. The YMRS total score ranges from 0 to 60 where higher scores indicate more severe mania; thus, a negative change (or decrease) from baseline indicates a reduction (or improvement) in manic symptoms.⁴⁵

In the bipolar mania studies, the YMRS was administered by an experienced and trained rater using the YMRS Scripted Interview Guide developed by Concordant Rater Systems. Assessments and ratings were to have been made by the same rater at approximately the same time of day. The YMRS total score was used to inform multiple outcomes in the bipolar mania studies — namely, the change from baseline to week 3 in the YMRS total score, the remission rate at week 3, and the response rate at week 3. The response rate was defined by the proportion of patients with an improvement in the YMRS total score of at least 50%. The remission rate was defined by the proportion of patients with a YMRS total score of 12 or less.

In patients with acute manic or mixed bipolar disorder, the minimal clinically significant difference was 6.6 points.⁷ Also, there is evidence to support a 50% reduction on the YMRS score as a clinically meaningful definition of response in acute treatment of manic or mixed episodes in pediatric patients with bipolar I disorder.⁴⁶

Montgomery–Åsberg Depression Rating Scale (Change in Symptom Severity, Response Rate, Remission Rate)

The MADRS is a 10-item, clinician-rated scale used to assess the severity of depressive symptoms during the past week.^47,48 Each item is rated on a scale of 0 (item not present or normal) to 6 (severe or continuous presence of the symptoms), resulting in a maximum total score of 60 points, in which higher scores indicate greater depressive symptomology.⁴⁷

Similar to the YMRS, the MADRS was used to inform multiple outcomes in the bipolar depression studies. This included the change from baseline to week 6 in MADRS total score, MADRS response, and MADRS remission where response was defined as at least a 50% reduction from baseline in the MADRS total score and remission was defined as a MADRS total score of 10 or less. The MADRS response, remission, and total score were reported by visit. In the bipolar mania studies, the MADRS total score was evaluated as the change from baseline to week 3.

The psychometric properties of MADRS are based on studies in major depressive disorder. There is evidence to support an improvement of at least 2 points on the MADRS as clinically relevant.^8,9 Response to treatment is usually defined as at least a 50% reduction on the MADRS total score from baseline.⁴⁹ No consensus was reached for a cut-off score on the MADRS for defining remission in clinical trials.⁵⁰ The criterion score to identify remission has ranged from 4 through 12 in various trials in depression.^49,51,52 In bipolar disorder, a cut-off score of 5 and 10 has been associated with remission and a softer definition of remission, respectively.⁵³

Clinical Global Impression–Severity of Illness and Clinical Global Impression–Improvement (Change in Symptom Severity)

The CGI-S is a 7-point scale that measures the overall severity of the illness in comparison with the severity of other patients the physician has observed. CGI-S assesses the overall severity of mental disorders at the time of the assessment on a 7-point scale ranging from 1 (normal, not at all ill) to 7 (extremely ill).⁵⁴

The CGI-I is a 7-point scale that measures the change in the overall severity of illness for the individual patient. The 7-point scale of the CGI-I ranges from 1 (very much improved) to 7 (very much worse).⁵⁴

The bipolar mania studies indicated that both the CGI-S and the CGI-I were assessed by a psychiatrist. The bipolar depression studies indicated that the CGI-S was administered by the investigator or a sub-investigator with extensive professional training and experience in assessing mental illness. In the bipolar mania studies, the CGI-S and CGI-I were evaluated as a change from baseline to week 3. In the bipolar depression studies, the CGI-S was evaluated as a change from baseline to week 6 (study RGH-MD-53 and study RGH-MD-54) or week 7 (study RGH-MD-56). The CGI-S score by visit was also reported in study RGH-MD-53.

A 1-point improvement has been considered as a clinical improvement in clinical studies in schizophrenia and based on expert opinion for within-group differences.^{55,56, 57,58} Evidence for an MID in the CGI-I scale was not found.

Positive and Negative Syndrome Scale (Change in Symptom Severity)

The PANSS is a 30-item rating scale that was developed to assess the severity of positive and negative symptoms within the past week in patients with schizophrenia. The PANSS is administered as a standardized Structured Clinical Interview. It was reported that information is obtained from the patient and supporting clinical information from family, hospital staff, and other reliable informants. It consists of 3 subscales — namely, positive symptoms, negative symptoms, and general psychopathology. The PANSS total score is informed by 30 items: 7 positive symptoms, 7 negative symptoms, and 16 general psychopathology symptoms, each rated on a 7-point scale from 1 (absent) to 7 (extreme). The total score ranges from 30 to 210 and higher scores indicate more severe symptoms or psychopathology.^59,60 The PANSS total score was reported as the change from baseline to week 3 in the bipolar mania studies. Evidence of an MID for the PANSS in patients with bipolar disorder was not identified. Evidence of an MID for the PANSS in patients with schizophrenia that was identified is opinion-based and associated with uncertainty.

Hamilton Depression Rating Scale (Change in Symptom Severity, Remission Rate)

The 17-item HAM-D is a clinician-rated scale used to assess the severity of symptoms and address both somatic and psychological symptoms of depression.^61-63 The items are either rated on a 5-point scale (a 0 to 4 spectrum) or a 3-point scale (a 0 to 2 spectrum), where increasing scores represent increasing severity of symptoms.^64,65 Scores are summed to obtain a total score out of 52 or 53.^10,66 The psychometric properties of HAM-D are based on studies of patients with major depressive disorder. NICE recommended a 3-point difference between the drug and placebo groups as a criterion for clinical significance.¹⁰ A separate report suggested a 2-point difference between an antidepressant and placebo might be clinically relevant.⁸ Remission was defined as a score of 7 or less on the HAM-D total score by a consensus panel in 1991.^62,67 Suggestions for an optimal cut-off score to define remission has ranged from 2 to 7.⁶² The generalizability of the psychometric assessment of the HAM-D to patients with bipolar disorder is unclear.

Hamilton Anxiety Rating Scale (Change in Symptom Severity)

The HAM-A is a 14-item, clinician-rated scale used to assess somatic and psychic anxiety symptoms.^68,69 The items are rated on a 5-point scale (a 0 to 4 spectrum), where increasing scores indicate higher levels of symptom severity, and are summed to yield a total score.¹⁵ Evidence for the MID of HAM-A was not found. The use of the HAM-A as an indicator of the severity of anxiety states in depressive disorders has been criticized⁷⁰; however, there is some controversy regarding this conclusion, since the HAM-A is commonly used as an outcome measure in trials with patients with depression. The generalizability of the HAM-A to patients with bipolar disorder is unknown.

Quick Inventory of Depressive Symptomatology–Self-Rated (Change in Symptom Severity)

The QIDS-SR is a 16-item, self-reported tool that measures depressive symptom severity.⁶⁶ Items included assess Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) diagnostic symptoms for major depressive disorder. The recall period is 7 days as patients are asked to rate their symptoms during this period. The responses are converted from the 16 items into 9 DSM-IV symptom criterion domains: sleep, weight, psychomotor changes, depressed mood, decreased interest, fatigue, guilt, concentration, and suicidal ideation. The items are used to characterize a major depressive episode. Each item is scored on a scale of 0 to 3, with higher scores representing greater symptom severity. The total score ranges from 0 to 27.⁶⁶ Evidence for an MID in the QIDS-SR scale was not identified.

Functioning Assessment Short Test (Functioning)

The FAST is a 24-item, clinician-rated scale used to assess functional impairment in patients with mental disorders. The items are divided into 6 areas of functioning: autonomy, occupational functioning, cognitive functioning, financial issues, interpersonal relationships, and leisure time. Items are rated on a 4-point scale (0 = no difficulty; 3 = severe difficulty). The total score is the sum of all 24 items, with higher scores indicating greater difficulty.⁷¹ There is evidence to suggest the minimum clinically important difference ranges from an 8-point to 9-point change from baseline.⁷²

Columbia–Suicide Severity Rating Scale

The C-SSRS is a clinician-rated instrument that reports the severity of both suicidal ideation and behaviour. Suicidal ideation is rated using a 5-item scale ranging from 1 (wish to be dead) to 5 (active suicidal ideation with specific plan and intent). The C-SSRS also captures information about the intensity of ideation — specifically, the frequency, duration, controllability, deterrents, and reasons for the most severe types of ideation. Suicidal behaviour is also rated using a 5-item scale, ranging from 0 (no suicidal behaviour) to 4 (actual attempt). The C-SSRS was reported in all studies except study RGH-MD-31 and an assessment was completed at all study visits by the investigator or designee with extensive professional training and experience in assessing mental illness.⁵

Statistical Analysis

Primary Outcome of the Studies

The primary outcome in the bipolar mania studies was the change from baseline to week 3 in the YMRS total score. The primary outcome in the bipolar depression studies was the change from baseline to week 6 in the MADRS total score.

Power Calculation

Power calculations for the primary end points are described as follows. Power calculations did not account for secondary end points.

In study RGH-MD-31, a sample size of approximately 120 patients in each treatment group would have provided at least 80% power to detect an effect size (treatment group difference relative to pooled SD) of 0.363 at a 2-sided significance level of 0.05.

In study RGH-MD-32, a sample size of 160 randomized patients in each of the 2 treatment groups provided 90% power to detect effect sizes (treatment group difference relative to pooled SD) of 0.40 between the placebo group and the cariprazine treatment group in the primary efficacy parameter, based on an MMRM using simulations. The simulation assumed a correlation of 0.7 between the repeated measures and effect sizes of 0.16, 0.17, 0.27, 0.32, 0.36, and 0.40 for the cariprazine treatment group for the planned postbaseline visits.

In study RGH-MD-33, adjustments were made for multiple comparisons involving the 2 efficacy end points and 2 cariprazine flexible-dose groups by using the matched parallel gatekeeping procedure. A sample size of 165 randomized patients in each of the 3 treatment groups provided 92% power to detect effect sizes (treatment group difference relative to pooled SD) of 0.430 between the placebo and cariprazine groups (6 mg per day to 12 mg per day and 3 mg per day to 6 mg per day) for the primary efficacy parameter, respectively, based on an MMRM model using simulations. The simulation assumed a correlation of 0.7 between the repeated measures and effect sizes of 0.43 for both the cariprazine 6 mg per day to 12 mg per day treatment group and the 3 mg per day to 6 mg per day treatment group for the primary and secondary parameters.

The power calculation for study RGH-MD-31 did not account for loss to follow-up. The power calculations for study RGH-MD-32 and study RGH-MD-33 assumed a 35% dropout rate.

The power calculations used in study RGH-MD-53 and study RGH-MD-54were identical. For the comparison of the primary end point in RGH-MD-53 and RGH-MD-54, the sample size of 160 patients per arm provided approximately 82% statistical power to show a statistically significantly higher effect in each dose of cariprazine versus placebo. The study had statistical power of 90% to show that at least 1 of the 2 cariprazine doses was statistically significantly more efficacious than placebo in the primary end point. These calculations assumed an effect size of 0.36 (treatment group difference relative to SD). All statistical powers presented in this section were calculated adjusting for multiple comparisons using a matched parallel gatekeeping procedure, with the familywise type I error rate being controlled at a 0.05 level (2-sided). The dropout rate was assumed to be 22% at week 6. The within-person correlation for both the primary and secondary end points was assumed to be 0.6, as was correlation between the 2 end points (primary and secondary). Assumptions of effect sizes, correlation coefficients, and dropout rate were based on the RGH-MD-56 study.

In study RGH-MD-56, a sample size of 150 randomized patients in each of the 4 treatment groups provided approximately 90% power to detect at least 1 of the 3 effect sizes at a 2-sided significance level of 0.05 at week 6 for the primary efficacy parameter, adjusted for multiple comparisons. This was based on assumed effect sizes (treatment group difference relative to pooled SD) of at least 0.30, 0.36, and 0.4 for the 3 cariprazine dosage groups (0.75 mg per day, 1.5 mg per day, and 3 mg per day) at week 6, a correlation coefficient of 0.66 for within-patient assessments for both the primary and secondary efficacy parameters, a correlation coefficient of 0.60 between the primary and secondary efficacy parameters, and a 25% patient dropout rate at week 6, for the primary efficacy parameter, adjusted for multiple comparisons. The correlation coefficients for within-patient assessments and the dropout rate were based on study RGH-MD-52.

Statistical Test or Model

In study RGH-MD-31, the primary analyses were performed using the LOCF approach. The comparison between cariprazine and placebo was performed by means of an analysis of covariance (ANCOVA) model, with treatment group and study centre as factors and baseline YMRS total score as a covariate. The normality assumption in the primary efficacy analysis model was checked by examining the normal probability plot of the residuals from the ANCOVA model. In study RGH-MD-32 and study RGH-MD-33, the primary analysis was performed using an MMRM with treatment group, study centre, visit, and treatment-group-by-visit interaction as fixed effects and the baseline value and baseline-by-visit interaction as the covariates. An unstructured covariance matrix was used to model the covariance of within-patient scores. The Kenward-Roger approximation was used to estimate denominator degrees of freedom.

The analysis of the primary end point in the bipolar depression studies was similar to that which was described for study RGH-MD-32 and study RGH-MD-33. In study RGH-MD-56, pairwise tests of no difference of the change from baseline between the 3 cariprazine dose groups and placebo were conducted; the LSM difference with 95% CI is reported for each comparison. In addition, for each treatment group, the fitted mean change (least squares) is provided. The study was considered positive if at least 1 dose group of cariprazine was statistically superior to placebo at week 6.

In the bipolar mania studies, the primary efficacy analysis was based on the intention-to-treat (ITT) population. All statistical tests were 2-sided hypothesis tests performed at the 5% level of significance for main effects. All CIs were 2-sided 95% CIs, unless stated otherwise. The baseline for efficacy was defined as the values recorded at visit 2 before the first dose of the double-blind study drug.

Descriptive statistics were presented by visit and treatment group for all efficacy parameters based on the LOCF and the observed cases approaches in study RGH-MD-31. In study RGH-MD-32, study RGH-MD-33, and study RGH-MD-56, by-visit analyses based on an MMRM for all continuous efficacy parameters using observed cases were also performed.

All efficacy analyses were based on the ITT population, unless stated otherwise. The baseline for each specific efficacy end point was defined as the value recorded at visit 2 (week 0). If this value was not available, the last available value before the first dose was to be used as the baseline value. Efficacy results were considered to be statistically significant after considering control for multiplicity, described under the summary of secondary outcomes of the studies as follows. All statistical hypothesis tests were performed at the 2-sided 5% significance level for main effects. All CIs were 2-sided 95% CIs, unless stated otherwise. By-visit analyses, which were done using the LOCF approach, were presented for all efficacy parameters.

Data Imputation Methods

In study RGH-MD-31, data were imputed using the LOCF approach for missing postbaseline total scores. Data were only imputed if at least 1 nonmissing postbaseline total score was available for that patient. Individual item scores were not carried forward (not imputed).

In study RGH-MD-32, study RGH-MD-33, study RGH-MD-53, study RGH-MD-54, and study RGH-MD-56, the primary analysis was performed based on all postbaseline scores using only the observed cases without imputation of missing values.

Additionally, in the depression studies, the baseline value was carried forward only for the intermittent missing values immediately after baseline. If all the postbaseline values were missing, the baseline value was not carried forward.

Subgroup Analyses

Subgroup analyses by demographic factors and baseline bipolar mania were performed for the primary analysis in the bipolar mania studies — the change from baseline at week 3 in the YMRS total score. The following subgroups were explored: age group (< 55 years, ≥ 55 years), sex (male, female), race group (White, all other races), geographic region (US, non-US), and baseline bipolar I type (mania, mixed). All subgroup analyses were reported descriptively by treatment group.

Sensitivity Analyses

In study RGH-MD-31, 2 sensitivity analyses were performed on the primary efficacy parameter (change from baseline to week 3 in the YMRS total score). Between-treatment group comparisons at week 3 were performed on the observed cases, using the ANCOVA model described for the primary analysis, and using an MMRM with treatment group, study centre, visit, and treatment-group-by-visit interaction as fixed effects and baseline YMRS total score as a covariate. An unstructured covariance matrix was used to model the covariance of within-patient measurements. This analysis was performed on the changes from baseline at all postbaseline time points during the double-blind treatment phase using only the observed cases.

In study RGH-MD-32 and study RGH-MD-33, a sensitivity analysis using a pattern-mixture model based on nonfuture dependent missing value restrictions was performed to assess the robustness of the primary MMRM results to the possible violation of the missing-at-random assumption.

The analysis of the primary end point in the depression studies was similar to that which was described for study RGH-MD-32 and study RGH-MD-33.

Secondary Outcomes of the Studies

In the bipolar mania studies, the secondary efficacy outcome was the change from baseline to week 3 in the CGI-S score. In the bipolar depression studies, the secondary efficacy outcome was the change from baseline to week 6 in the CGI-S score. In each study, the secondary outcome was analyzed using a similar approach to the primary analysis (ANCOVA model or MMRM model).

In study RGH-MD-31 and study RGH-MD-32, the analysis of the secondary end point was carried out inferentially at a 2-sided 5% significance level only if the result of the primary efficacy end point was significant at the 0.05 level. This is a closed testing procedure that controls the studywise type I error rate.

In study RGH-MD-33, a matched parallel gatekeeping procedure was used to control the overall type I error rate for multiple comparisons of the 2 active doses versus placebo across the primary and secondary hypotheses. The primary hypotheses consisted of 2 null hypotheses, corresponding to the comparisons of cariprazine 3 mg to 6 mg and cariprazine 6 mg to 12 mg, respectively, with placebo in regard to the primary efficacy parameter. Similarly, the secondary hypotheses include comparisons of cariprazine 3 mg to 6 mg and cariprazine 6 mg to 12 mg, respectively, with placebo in regard to the secondary efficacy parameter. Rejection of at least 1 of the primary hypotheses was required to examine the secondary hypotheses; significance in the secondary end point for a dose level could only be claimed if the corresponding primary hypothesis was found to be significant.

Similar to study RGH-MD-33, study RGH-MD-53 and study RGH-MD-54 implemented a parallel gatekeeping procedure to control the overall type I error rate for multiple comparisons. The primary hypotheses consisted of 4 null hypotheses, corresponding to comparisons of cariprazine 1.5 mg and cariprazine 3.0 mg to placebo for the primary efficacy analysis (MADRS total score) and the same comparisons for the secondary efficacy analysis (CGI-S score). The hypotheses corresponding to the primary efficacy analyses served as a parallel gatekeeper for the secondary efficacy analyses.

Table 18: Statistical Analysis of Efficacy End Points

ANCOVA = analysis of covariance; CGI-S = Clinical Global Impression–Severity of Illness; MADRS = Montgomery–Åsberg Depression Rating Scale; MMRM = mixed model of repeated measures; OC = observed case; PMM = pattern-mixture model, YMRS = Young Mania Rating Scale.

Source: Clinical Study Reports for study RGH-MD-31, study RGH-MD-32, study RGH-MD-33, study RGH-MD-53, study RGH-MD-54, and study RGH-MD-56.^11-16

Analysis Populations

Three analysis sets were described for all included studies, which comprised the randomized population, the safety population, and the ITT population.

The randomized population was defined as consisting of all patients who were screened and randomized, which is consistent with the definition of an ITT population.

The safety population was defined as consisting of all patients who were randomized and who received at least 1 dose of the double-blind investigational product.

The ITT population defined by the sponsor consisted of all patients in the safety population who had at least 1 postbaseline assessment of the YMRS total score (mania studies) or MADRS total score (depression studies). As this is not a true definition of an ITT population, it has been referred to as a modified intention-to-treat (mITT) population herein.

Results

Patient Disposition

A summary of the patient disposition in the bipolar mania studies and bipolar depression studies is provided in Table 19 and Table 20, respectively.

In the bipolar mania studies, 68%, 66%, and 64% of screened patients were randomized to study RGH-MD-31, study RGH-MD-32, and study RGH-MD-33, respectively. Discontinuation rates were similar between treatment groups and ranged from 23% to 38% across studies. The most common reasons for discontinuation across studies were AEs (5% to 14%), insufficient therapeutic response (1% to 26%), and withdrawal of consent (9% to 17%). In general, more patients randomized to placebo discontinued due to insufficient therapeutic response compared to patients randomized to cariprazine. Discontinuation due to AEs was more common among patients randomized to cariprazine relative to placebo. Discontinuation due to withdrawal of consent was similar between treatment groups, with the exception of study RGH-MD-32 (16.5% versus 11.0% for cariprazine versus placebo).

In the bipolar depression studies, 57%, 62%, and 58% of screened patients were randomized to study RGH-MD-53, study RGH-MD-54, and study RGH-MD-56, respectively. The proportion of patients who discontinued from study ranged from 15% to 36% across the studies, and an imbalance in discontinuation rates between treatment groups was observed in study RGH-MD-54 and study RGH-MD-56. In both studies, the proportion of patients who discontinued was higher in the cariprazine 3 mg treatment group relative to cariprazine 1.5 mg or placebo. The most common reasons for discontinuation included AEs (3% to 12%), lost to follow-up (3% to 7%), and withdrawal of consent (2% to 10%), and were generally similar between treatment groups with the exception of AEs in study RGH-MD-53 (7% for cariprazine 3 mg versus 3% for each of cariprazine 1.5 mg and placebo) and withdrawal of consent in study RGH-MD-56 (3%, 10%, and 8% for cariprazine 1.5 mg, cariprazine 3 mg, and placebo, respectively).

Exposure to Study Treatments

A summary of exposure to study treatments for the bipolar mania studies and bipolar depression studies is provided in Table 21 and Table 22. Overall, the mean duration of treatment was similar between treatment groups in all studies. Of note, the bipolar studies used a flexible-dose regimen for cariprazine. The overall mean daily dose in study RGH-MD-31 and study RGH-MD-32 was 8.8 (SD = |||||||||) mg and 7.5 (SD = |||||||||) mg per day, respectively. The overall mean daily dose in study RGH-MD-33 was 4.8 (SD = |||||||||) mg. The overall mean daily dose was not reported in the bipolar depression studies.