Cardiometabolic Risks of Blonanserin and Perospirone in the Management of Schizophrenia: A Systematic Review and Meta-Analysis of Randomized Controlled Trials

Background The present study aimed to evaluate cardiometabolic risks [weight gain, blood lipid levels (total cholesterol and triglycerides), blood glucose levels, hemoglobin A1c (HbA1c) levels, and corrected QT interval (QTc) prolongation] associated with the use of blonanserin and perospirone versus other antipsychotics in the management of patients with schizophrenia. Method We conducted a systematic review and meta-analysis of patient data from randomized controlled trials comparing blonanserin or perospirone with other antipsychotics. Results In total, 4 blonanserin studies (n  = 1080) were identified [vs. risperidone (2 studies, n = 508); vs. haloperidol (2 studies, n = 572)]. Blonanserin produced less weight gain compared with risperidone (weighted mean difference = −0.86, 95% confidence intervals = −1.36 to −0.36, p = 0.0008; 2 studies, 480 patients). However, no significant differences were observed in blood lipid, glucose, and HbA1c levels or QTc prolongation between blonanserin and risperidone or haloperidol. For perospirone studies, 5 studies [562 adult patients with schizophrenia randomized to perospirone (n = 256), olanzapine (n = 20), quetiapine (n = 28), risperidone (n = 53), aripiprazole (n = 49), haloperidol (n = 75), or mosapramine (n = 81)] were identified. Perospirone did not differ from other antipsychotics with regard to weight gain and total cholesterol levels. Conclusions Our results suggest that blonanserin is associated with a lower of weight gain compared with other antipsychotics. Because the number of studies was small, additional controlled clinical trials with larger number of patients are indicated.

Recently, we published 2 meta-analyses of schizophrenia; 1 of them was a meta-analysis of blonanserin regarding efficacy and safety in schizophrenia [5]. The article reported following 2 primary results: (1) although blonanserin exhibits similar effects as risperidone and haloperidol with respect to the overall psychopathology of schizophrenia, blonanserin has more beneficial effects on negative symptoms compared with haloperidol; (2) blonanserin has a lower risk of hyperprolactinemia compared with other pooled antipsychotics. Although dizziness and akathisia occur significantly less often with blonanserin than with haloperidol, blonanserin has a higher risk of akathisia than risperidone. The second article was a meta-analysis regarding efficacy and safety of perospirone in schizophrenia [6]; the article reported following 3 primary results: (1) although perospirone did not differ from other pooled antipsychotics regarding the reduction in Positive and Negative Syndrome Scale (PANSS) negative and general subscale scores or regarding discontinuation because of any cause, inefficacy, and side effects, perospirone was inferior to other pooled antipsychotics in the reduction of PANSS total scores and positive subscale scores; (2) perospirone was superior to haloperidol in the reduction of PANSS negative subscale scores; (3) perospirone had lower scores related to extrapyramidal symptoms compared with other pooled antipsychotics. However, because we did not acquire available data with regard to cardiometabolic risks for meta-analyses in our previous articles, we did not perform meta-analyses addressing cardiometabolic risks of blonanserin and perospirone in patients with schizophrenia in our previous articles [5,6]. The present study aimed to evaluate cardiometabolic risks [weight gain, blood lipid levels (total cholesterol and triglycerides), blood glucose levels, hemoglobin A1c (HbA1c) levels, and corrected QT interval (QTc) prolongation] associated with the use of blonanserin and perospirone versus other antipsychotics in the management of patients with schizophrenia.

Inclusion Criteria and Search Strategy, Data Extraction, and Outcomes
We selected only randomized controlled trials (RCTs) using blonanserin or perospirone therapy in patients with schizophrenia. This meta-analysis was performed according to guidelines of preferred reporting items for systematic reviews and meta-analyses (PRISMA, 2009) [7]. We performed a systematic literature review according to the PICO strategy (Patients: schizophrenia, Intervention: blonanserin or perospirone, Comparator: other antipsychotics, and Outcome: cardiometabolic risks). We included openlabel and double-blinded RCTs comparing blonanserin or perospirone with other antipsychotics for patients with schizophrenia or schizophrenia-like psychoses. We allowed inclusion of nondouble-blinded studies to include more studies in the metaanalysis. To identify relevant studies, we searched PubMed, Cochrane Library databases, Google Scholar, and PsycINFO citations (without language restrictions) published up to January 2, 2013, using following key words: ''randomized,'' ''random,'' or ''randomly'' and ''blonanserin'' or ''perospirone'' and ''schizophrenia.'' Additional eligible studies were also sought by a hand search of reference lists from primary articles and relevant reviews.
Three authors (T.K., Y.M., and N.I.) assessed inclusion and exclusion criteria for each identified study. Discrepancies in different coding forms were resolved by discussions between authors (T.K., Y.M., and N.I.). Three authors (T.K., Y.M., and N.I.) independently extracted, assessed, and entered data into the Review Manager. When data required for the meta-analysis were missing, first/corresponding authors were contacted for additional information (including endpoint scores). Unpublished data were provided for this meta-analysis by Dr. Jaewon Yang, Dr. Seung-Hyun Kim, Dr. Sadanori Miura, Dr. Mitsukuni Murasaki, Dr. Yoshiteru Takekita, Bukwang Pharm. Co., Ltd., and Dainippon Sumitomo Pharma Co., Ltd. Moreover, we assessed the methodological quality of trials using Cochrane risk-of-bias criteria (http://bmg.cochrane.org/assessing-risk-bias-included-studies).

Outcomes
The primary outcome was weight gain. Secondary outcomes were changes in blood lipid levels (total cholesterol and triglycerides), blood glucose levels, HbA 1c levels, and QTc prolongation (Bazett method) on electrocardiography.

Statistical Analysis
We pooled data for side effects. We based our analysis on intention-to-treat (ITT) or modified ITT data (i.e., at least 1 dose or at least 1 follow-up assessment); no observed case data were included. The meta-analysis was performed using the Review Manager (RevMan), version 5.1, for Windows (http://ims. cochrane.org/revman). To combine studies, we used random effects model by DerSimonian and Laird [8]. We used this conservative model to address the possibility that underlying effects that differ across studies and populations would be heterogeneous. We used weighted mean difference (WMD) and 95% CIs for continuous data. In cases where I-squared values were .50%, sensitivity analyses were conducted to seek reasons for the heterogeneity [9]. However, we could not determine any significant heterogeneity in all outcomes between both treatment groups. Finally, funnel plots were visually inspected to explore the possibility of publication bias.

Blonanserin
Study characteristics. The PubMed search yielded 11 studies using abovementioned key words for screening (FIGURE 1-a). We excluded 2 duplicate studies across 3 databases as well as 3 studies based on the title or abstract review. Additional 4 full-text articles were excluded because they were either review papers (N = 3) or data were based on the same sample (N = 1), yielding 2 eligible studies [10,11] (FIGURE 1-a). In addition, 2 other studies [12,13] were identified from the review [14][15][16]. In total, we identified 4 studies involving 1080 patients with schizophrenia that reported data from double-blinded, randomized controlled trials of blonanserin for schizophrenia, thus meeting our inclusion criteria (vs. risperidone: n = 508; vs. haloperidol: n = 572). The mean study duration was 7.5 weeks; 3 trials lasted 8 weeks, and 1 trial lasted 6 weeks. Sample sizes ranged from 206 to 307 patients. The mean age of the study population was 41.0 years. Two studies [11,13] used the SGA risperidone, and the other 2 [10,12] used the FGA haloperidol as a comparator. All studies were sponsored by the pharmaceutical industry. Two studies were published in English and the other 2 in Japanese. Two studies were conducted in Japan, 1 in Korea, and the remaining 1 was conducted in the United States and Europe. All studies were of high methodological quality based on Cochrane Risk of Bias Criteria because all blonanserin studies were doubleblinded RCTs and mentioned required details of the study design (FIGURE S1 and S2). Characteristics of studies included have been presented in our previous paper [16].
Cardiometabolic risks. Comparing blonanserin with other pooled antipsychotics, no significant differences were observed in weight gain, blood lipid levels (total cholesterol and triglycerides), blood glucose levels, HbA 1c levels, or QTc prolongation (FIGURE 2 and 3). However, blonanserin apparently produced less weight gain compared with risperidone (WMD = 20.86, 95% CI = 21.36 to 20.36, p = 0.0008; 2 studies, 480 patients;    presence of publication bias, but the sensitivity was limited by the small number of 4 studies.
Cardiometabolic risks. Perospirone did not differ from other antipsychotics with regard to weight gain and total cholesterol. Because there were no available data regarding other cardiometabolic outcomes, we did not conduct that meta-analysis (FIGURE S3). Visual inspection of the funnel plot for weight gain in perospirone vs. other antipsychotics did not suggest the presence of publication bias, but the sensitivity was limited by the small number of 5 studies.

Discussion
To our knowledge, this is the first comprehensive meta-analysis of cardiometabolic risks of blonanserin and perospirone for patients with schizophrenia. Although previous studies did not report that blonanserin was associated with a lower risk of weight gain, trends in this direction have been reported. Our metaanalysis enabled us to obtain greater statistical power compared with that in past studies (Cochrane Handbook for Systematic Reviews of Interventions, http://handbook.cochrane.org/); therefore, we were able to establish that blonanserin had a lower risk of weight gain compared with risperidone. Antipsychotic affinity for the histamine H 1 receptor is most closely linked to increased weight gain, although affinity for dopamine D 2 , serotonin 1A , serotonin 2C and a 2 -adrenergic receptors may also be involved [23][24][25]. Blonanserin lacks significant affinity for histamine H 1, serotonin 1A , a 2 -adrenergic receptors [16]. Haloperidol also lacks significant affinity for histamine H 1 , serotonin 1A , serotonin 2C and a 2 -adrenergic receptors [16]. On the other hand, risperidone has a high affinity for histamine H 1 , dopamine D 2 , serotonin 2C and a 2 -adrenergic receptors [16]. Olanzapine and clozapine also have a high affinity for histamine H 1 and serotonin 2C receptors [16]. These mechanisms may be involved in the different effects seen among the antipsychotic groups. Although several studies have reported a difference in the degree of weight gain induced by antipsychotics between patients with antipsychotic-naïve schizophrenia and those with chronic schizophrenia (i.e., patients with antipsychotic-naïve schizophrenia have revealed more weight gain than those with chronic schizophrenia), and although some have revealed a varying degree of weight gain, depending on the antipsychotic dosage (e.g., risperidone dosage was associated with an increased degree of weight gain) [25,26], our results indicate that blonanserin may be the safest antipsychotic with regard to weight gain. However, because the number of studies was small, additional controlled clinical trials with larger number of patients with antipsychotic-naïve schizophrenia are indicated. There are other limitations to our findings. Although blonanserin was significantly associated with a lower weight gain risk compared with other antipsychotics, the significant result detected may be driven by the Miura study data [13]. Moreover, although no obvious publication bias could be identified, the possibility of the bias would be excluded from the significant result based on a subgroup analysis with only two studies.

Conclusion
Our results suggest that blonanserin is associated with a lower weight gain risk compared with other antipsychotics. However, because this meta-analysis includes only 4 blonanserin RCTs and 5 perospirone RCTs, future research should generate more safety data using larger samples.