Post‐hoc analysis investigating the safety and efficacy of brexpiprazole in Japanese patients with schizophrenia who were switched from other antipsychotics in a long‐term study (Secondary Publication)

Abstract A post hoc analysis was performed using data obtained over eight weeks from 200 Japanese patients with schizophrenia who were switched to brexpiprazole monotherapy in a long‐term treatment study. The 8‐week period comprised of a 4‐week switching phase and a 4‐week post‐switch phase. For the antipsychotic switching schedule, brexpiprazole was first administered at 1 mg/day and increased to 2 mg/day by the end of week 4. Concurrently, the previous antipsychotic(s) was/were tapered gradually from the start of week 3 and discontinued by the end of week 4. Brexpiprazole could then be increased up to 4 mg/day according to the CGI‐I criteria. At week 8, 1.8%, 23.2%, 25.0%, and 50% of patients were administered daily brexpiprazole doses of 1, 2, 3, and 4 mg, respectively. The discontinuation rate at week 8 was 17.0%. The major reasons for discontinuation were consent withdrawal (9.5%), occurrence of adverse events (5.5%), and physician's decision (2.0%). Commonly reported adverse events were nasopharyngitis (13.5%), schizophrenia (9.0%), insomnia (6.5%), headache (5.5%), and akathisia (5.5%). The discontinuation rate was 4.9% for patients who were switched from aripiprazole as the primary antipsychotic and 25.4% for those who were switched from other antipsychotics. Owing to the serious adverse events that led to treatment discontinuation, careful switching to brexpiprazole is necessary in patients who previously used olanzapine as their primary antipsychotic.

In the treatment of psychiatric disorders such as schizophrenia, antipsychotic switches are routinely used to obtain better therapeutic responses with fewer side effects, or to reduce drug-related secondary negative symptoms or cognitive impairment by simplifying treatment regimens (reduced number of drugs and reduced doses) for which is used to reduce/manage the side effects. 3 However, antipsychotic switches may be associated with the withdrawal phenomena due to drug discontinuation or dose reduction of the previous antipsychotic(s), adverse events, insufficient effects, or deteriorated psychiatric symptoms attributable to replacement of new antipsychotics. Accordingly, antipsychotic switches should be performed with great caution. 3 Although there are recommended guidelines for switching for some antipsychotics, more experience is required in clinical practice to reach a consensus that is sufficient to establish guidelines.
Particularly for antipsychotics that are new in the market, careful evaluations of the data retrieved in clinical practice are important. [4][5][6] Here, we performed an exploratory post hoc analysis with Japanese patients who were switched from previous antipsychotics to the novel antipsychotic brexpiprazole in a long-term treatment study to assess the outcomes when a patient's previous antipsychotic was switched to brexpiprazole.

| Clinical study
The target population of the study was de novo patients who were eligible for a switch to brexpiprazole monotherapy and patients who had completed a placebo-controlled, double-blind, short-term treatment study with brexpiprazole. The long-term treatment study was conducted in compliance with the International Conference on Harmonisation Good Clinical Practice guidelines and Japanese Ministerial Ordinance on Good Clinical Practice. Each study site initiated patient enrollment after their institutional review boards (IRBs) reviewed and approved the study. All patients provided written informed consent before participation in the study. Written consent for minors was obtained from their legal representatives (parental authorities, spouses, guardians, or any person equivalent to them), but a personal signed informed consent was also provided by these patients. The results of the long-term treatment study have been published by Ishigooka et al. 7

| Study population and assessment period
The post hoc analysis was on de novo patients who were switched to brexpiprazole monotherapy in this long-term treatment study. 7 Since this post hoc analysis aimed to assess the influence of switches to brexpiprazole monotherapy, patients who did not use antipsychotics 30 days before informed consent retrieval and those who were F I G U R E 1 Switching procedure and analysis set withdrawn from the study due to deviations from the protocol were excluded. The assessment period was 8 weeks, which was composed of a 4-week protocol-defined switching phase and a 4-week postswitch phase, during which the influence of the switch to brexpiprazole monotherapy was observed.

| Antipsychotic switching procedure and the doses of brexpiprazole
The switch to brexpiprazole monotherapy was achieved by crosstitration over 4 weeks, that is, brexpiprazole was administered alongside the previous antipsychotic(s) for the first 2 weeks, followed by tapering of the previous antipsychotic(s) and, ultimately, discontinuation of the previous drug(s) in the latter 2 weeks.
The initial dose of brexpiprazole was 1 mg/day, and if tolerable, it was increased to 2 mg/day by the end of week 4. After week 5 (1 week after switching was successfully achieved), brexpiprazole was increased up to 4 mg/day according to the Clinical Global Impression-Improvement (CGI-I) criteria, if no issues with tolerability occurred. The previous antipsychotic(s) were gradually tapered over 2 weeks from the beginning of week 3 and discontinued by the end of week 4 ( Figure 1).

| Prior antipsychotic and antiparkinsonian treatments
In this post hoc analysis, antipsychotics that were used 30 days before informed consent retrieval were defined as previous antipsychotics to reflect the treatments provided to each patient before enrollment in this study. The doses of the previous antipsychotics were converted to chlorpromazine (CP)-equivalent doses. 8,9 Parenterals, as-needed drugs, and vegetamin (a combination product, domestic medicine in Japan) were excluded from CP-equivalent doses. The primary antipsychotic for each patient was defined as an antipsychotic that was administered at the highest CP-equivalent dose 30 days before informed consent retrieval. Similarly, the doses of antiparkinsonian drugs administered 30 days before informed consent retrieval were converted to biperiden-equivalent doses. 8 CGI-S score) were summarized using descriptive statistics. The use of antipsychotic(s) 30 days before informed consent retrieval (primary antipsychotic(s), number of antipsychotics, total daily CP-equivalent dose of antipsychotics, and total daily CP-equivalent dose of antipsychotics by primary antipsychotic) was summarized using frequencies (number and percentage of patients) and descriptive statistics. The use of antiparkinsonian drugs administered 30 days before informed consent retrieval (number of antiparkinsonian drugs and the daily total biperiden-equivalent dose of the drugs) was summarized using frequencies (number and percentage of patients) and/or descriptive statistics. Baseline data were values observed before the antipsychotic switch (ie, before starting brexpiprazole therapy).

| Analysis methods
The numbers and percentages of patients exposed to brexpiprazole were summarized by daily dose at the start of brexpiprazole and every 2 weeks thereafter.
The numbers and percentages of patients who discontinued or completed treatment during the 4-week switching phase, 4-week post-switch phase, and 8-week assessment period (4-week switching phase + 4-week post-switch phase) were calculated. The number and percentage of patients who discontinued treatment were summarized by reason for discontinuation.
The numbers and percentages of patients who discontinued or completed treatment over the 8-week assessment period (4-week switching phase + 4-week post-switch phase) were categorized based on patient characteristics and primary antipsychotic(s). The number and percentage of patients who discontinued treatment were summarized by primary antipsychotic(s) and reason for discontinuation. For patients who completed or discontinued treatment, patient characteristics and total daily CP-equivalent dose of the antipsychotic(s) administered 30 days before informed consent retrieval were summarized separately using descriptive statistics.
We performed univariate logistic regression with treatment discontinuation or completion over the 8-week assessment period (4-week switching phase + 4-week post-switch phase) using continuous variables of age, baseline PANSS total score, number of previous antipsychotics, and the total daily CP-equivalent dose of prior antipsychotics, as well as the type of primary antipsychotic(s) (aripiprazole or non-aripiprazole, olanzapine or non-olanzapine, risperidone or non-risperidone, and paliperidone or non-paliperidone) as explanatory variables. For each parameter, an estimate, standard error, Wald test P value, and odds ratio with its 95% confidence interval (two-sided) were determined. No significance test was performed, and P values were presented for reference.
Adverse events with an incidence of ≥ 5%, serious adverse events, and adverse events leading to treatment discontinuation during the 4-week switching phase, the 4-week post-switch phase, and the 8-week assessment period (4-week switching phase + 4-week postswitch phase) were summarized using frequencies (number and percentage of patients). Adverse events that led to the discontinuation of brexpiprazole therapy are listed by the day of onset, causal relationship to the trial drug, severity, and seriousness.
For PANSS total score, changes from baseline over the 8-week assessment period (4-week switching phase + 4-week post-switch phase) were summarized using descriptive statistics. Similar statistical analysis was performed for patients who used aripiprazole or non-aripiprazole antipsychotics as primary antipsychotic.
Imputations were not conducted for missing data.

| Patient disposition and characteristics
Among 208 Japanese patients who were switched to brexpiprazole monotherapy in a long-term treatment study, two patients did not use antipsychotics 30 days before providing informed consent and six were withdrawn from the analysis due to protocol deviations.
The post hoc analysis was therefore performed with the data re-

| Use of antiparkinsonian drugs
The proportion of patients receiving antiparkinsonian drugs 30 days before informed consent retrieval was 29.0% (58 of 200 patients), and this proportion decreased to 12.4% (21 of 170 patients) at week 8 (ie, 4 weeks after the switch). For the mean daily biperiden-equivalent dose of the antiparkinsonian drugs, a similar value was found at 30 days before informed consent retrieval (2.7 mg/day) and week 8 (ie, 4 weeks after the switch; 2.8 mg/day).
However, the total daily CP-equivalent dose of previous antipsychotics exceeded 2400 mg in one patient who discontinued treatment. high as 20 to 30%, regardless of the primary antipsychotics (Table 3).
When patients who had used aripiprazole as primary antipsychotic and those using non-aripiprazole were compared by univariate logistic regression with discontinuation rate over the 8-week assessment period as the response variable, odds ratio of 6.648 (P = .0006) was confirmed. The odds ratio of discontinuation rate by mean total daily CP-equivalent dose of previous antipsychotics was 1.000 (P = .3970; Table 4).
Eleven patients discontinued treatment due to adverse events (MedDRA/J version 16.0 preferred terms): with schizophrenia in seven patients, akathisia in two patients, and thirst and oculogyric crisis, each in one patient ( Table 6). Four of the 11 patients discontinued treatment due to serious adverse events (ie, schizophrenia in three patients and akathisia in one patient). Three of these patients received olanzapine, and one patient was administered haloperidol (Table 6). In the three patients previously treated with olanzapine, the onset of adverse events was day 28 (schizophrenia: stupor), day 29 (schizophrenia: soliloquy, hallucination auditory, delusional mood, feeling irritated, and confusion), and day 37 (akathisia: a Drug-Induced Extrapyramidal Symptoms Scale [DIEPSS] akathisia score of 4, resulting in difficulty managing physical conditions), and in the patient previously treated with haloperidol was day 57 (schizophrenia: auditory hallucination, delusion, and silly smile). Therefore, the onset of serious adverse events in the four patients who discontinued due to a serious adverse event occurred around the end of the protocol-specified antipsychotic switching (week 4) or 4 weeks after switching completion (week 8; Table 6).

| Psychiatric symptoms
The PANSS total score (mean ± SD) at the start of antipsychotic switching (baseline) was 69.8 ± 22.1 (N = 200), and mean change in PANSS total score from baseline to week 8 was −5.2 ± 10.9 (N = 168, observed cases). The change in PANSS total score from baseline to week 8 was compared between patients who had used aripiprazole as primary antipsychotic and those using another antipsychotic. In patients using aripiprazole, baseline PANSS total score was 64.6 ± 19.0 (N = 82) and the change from baseline to week 8 was −5.7 ± 10.0 (N = 78, observed case). In patients using non-antipsychotic, baseline PANSS total score was 73.3 ± 23.5 (N = 118) and the change from baseline to week 8 was −4.7 ± 11.7 (N = 90, observed case). Therefore, there was no substantial change in PANSS total score from baseline to week 8 whether aripiprazole was used as the primary antipsychotic or not.

| D ISCUSS I ON
For patients who were switched to the antipsychotic brexpiprazole (a dopamine D 2 receptor partial agonist), the discontinuation rate during the 8 weeks after treatment initiation was 4.9% for switches from a dopamine D 2 receptor partial agonist (aripiprazole) and 25.4% for switches from dopamine D 2 receptor antagonists (non-aripiprazole). This result suggests that more clinical experience is needed to derive the best method for use when switching from dopamine D 2 receptor antagonists to brexpiprazole.
Several studies have reported their results for the switch from dopamine D 2 receptor antagonists to aripiprazole, a dopamine D 2 receptor partial agonist like brexpiprazole. In the US-BETA study conducted in the United States over an 8-week assessment period, the rate of discontinuation for any cause was 35.0% and the rate of discontinuation due to adverse events was 17.1%. The adverse events with an occurrence of ≥ 10% were insomnia (24%), nausea (16%), headache (11%), and anxiety (10%). 10 In the EU-BETA study conducted in Europe over an 8-week assessment period, the rate of discontinuation for any cause was 27.2% and the rate of discontinuation due to adverse events was 12.6%. Similarly, the adverse events with an occurrence of ≥ 10% incidence in the EU-BETA study were insomnia (16.0%) and nausea (10.2%). 11 Because of the differences in switching rules used in the present post hoc analysis (ie, for switching to brexpiprazole monotherapy), and the US-BETA and EU-BETA studies (ie, for switching to aripiprazole monotherapy), a limitation exists when comparing the results in the different studies.
Nonetheless, we observed lower discontinuation rate and incidence  of adverse events in the present study than those found in the above studies, suggesting that switching to brexpiprazole might be safer than switching to aripiprazole.
The discontinuation rate when switching to brexpiprazole may be further lowered when the following two points, (a) and (b), are considered.

| (a) Taper period for the previous antipsychotic(s)
Brexpiprazole, a dopamine D 2 receptor partial agonist, has a high affinity for the dopamine D 2 receptor, serotonin 5-HT 1A receptor, and  approved brexpiprazole dose is 1 to 2 mg/day. Therefore, further evaluations are warranted to establish practical switching procedures that are consistent with this approved dose. As well, switching period of the study was 4 weeks. It is shorter than common switching strategy. In the further evaluation, switching period should be longer than this protocol. Regarding the observation period, the 4-week duration of post-switching phase is relatively short for the assessment of switching strategy. The results of discontinuation rate can reflect "initial" discontinuation, and longer duration is needed to assess outcome of switching to brexpiprazole particularly for those who received olanzapine before the switching to brexpiprazole.

CO N FLI C T O F I NTE R E S T
JI received honorarium from Otsuka Pharmaceutical Co., Ltd. for providing advices or lectures as a medical advisor for this long-term treatment study with brexpiprazole in Japanese patients with schizophrenia. TU, SI, YK, and SM are employees of Otsuka Pharmaceutical Co., Ltd.

AUTH O R CO NTR I B UTI O N S
JI, TU, YK, and SM contributed to the study design. YK contributed to the establishment of the analysis plan and the implementation of data analysis. JI, TU, and SI wrote a first draft of the manuscript. All authors contributed to the interpretation of data, critically revised the manuscript, and approved the final manuscript for submission.

E TH I C A L A PPROVA L
Each study site initiated patient enrollment after the institutional review boards (IRBs) reviewed and approved the study.

I N FO R M ED CO N S ENT
All patients provided written informed consent before participation in the study. Written consent for minors/dependents was obtained from their legal representatives (parental authorities, spouses, guardians, or any person equivalent to them), but a personal signed informed consent was also provided by these patients.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are proprietary in nature and are not publicly available. However, the data are available on request from the corresponding author upon reasonable request.