Meta-Analyses of the Efficacy of Asenapine for Acute Schizophrenia: Comparisons With Placebo and Other Antipsychotics

This work may not be copied, distributed, displayed, published, reproduced, transmitted, modified, posted, sold, licensed, or used for commercial purposes. By downloading this file, you are agreeing to the publisher’s Terms & Conditions.</.

Meta-Analyses of the Efficacy of Asenapine for Acute Schizophrenia: Comparisons With Placebo and Other Antipsychotics

Armin Szegedi, MD, PhD; Pierre Verweij, PhD; Wilbert van Duijnhoven, MSc; Mary Mackle, PhD; Pilar Cazorla, PhD; and Hein Fennema, PhD

In the last 2 decades, a variety of second-generation antipsychotics (SGAs) have been developed for schizophrenia. Asenapine has recently been added as a treatment option for schizophrenia for adults by the US Food and Drug Administration and as monotherapy or adjunctive therapy with lithium or valproate in the treatment of manic or mixed episodes associated with bipolar I disorder. Asenapine is also indicated in the European Union for the treatment of moderate to severe manic episodes associated with bipolar I disorder.1 Evaluating treatment options among SGAs is a complex task for a variety of reasons. Second-generation antipsychotics are not a pharmacologically homogeneous drug class,2 presumably resulting in different safety and tolerability profiles.3 Moreover, only a subset of schizophrenia patients will respond to any given SGA.4 Risk-benefit assessments across agents must be made on an individual patient basis, and response to an SGA will remain uncertain. Thus, it is valuable to have a new treatment option available.

Evaluating SGAs on the basis of risk-benefit is complex because even comparing “average” efficacy is not straightforward. First, even in large randomized clinical trials, observed treatment effects do not necessarily provide a homogeneous pattern. Asenapine data also illustrate this variability. Four active- and placebo-controlled short-term trials were conducted with asenapine in the effective dose range of 5 or 10 mg twice daily: 2 trials were robustly positive,5,6 1 was negative,7 and 1 failed.8 Second, placebo-controlled trials have shown efficacy for SGAs versus placebo, although this effect has been reduced over time,9,10 even when including compounds that traditionally showed substantial effect sizes. This reduction in efficacy increases the problem of comparing efficacy across SGAs.

Therefore, we used meta-analytic techniques to comprehensively evaluate efficacy data from all applicable randomized asenapine clinical trials. Specifically, we performed meta-analytic comparisons (vs placebo) of mean change in Positive and Negative Syndrome Scale (PANSS)11 total score for asenapine and other antipsychotics used as active controls in the same trials. To more clearly assess the clinical relevance of these treatment effects, PANSS response rates and associated odds ratios (ORs) for treatment response were assessed. Relative risk for treatment nonresponse with asenapine and other antipsychotics was examined by using data published by Leucht et al12 into which asenapine data were integrated. Consistent with previous reports,9,10 these efficacy analyses suggested that treatment effect might be influenced by publication year. Therefore, we further explored this finding by using a meta-regression to assess the impact of publication year on antipsychotic treatment effect versus placebo. Finally, to assess the relative efficacy of SGAs, a head-to-head network meta-analysis of PANSS total score change was conducted by using data from published head-to-head SGA trials,13 into which asenapine data were also integrated. These analyses comprehensively characterize the relative efficacy of asenapine within the SGA class, regardless of whether direct comparisons are available, and provide an update of the relative efficacy of the entire SGA class.

DATA SOURCES

Meta-Analyses of Asenapine Clinical Trial Program

Data for the meta-analysis included all randomized, placebo-controlled, 6-week studies of asenapine for acute schizophrenia that administered asenapine in the effective dosage range of 5 or 10 mg twice daily (0410046 [asenapine 5 mg, n = 58; placebo, n = 60; risperidone, n = 56], 0410217 [asenapine 5 mg, n = 102; asenapine 10 mg, n = 96; placebo, n = 93; olanzapine, n = 95], 0410228 [asenapine 5 or 10 mg, n = 85; placebo, n = 89; olanzapine, n = 85], 0410235 [asenapine 5 mg, n = 109; asenapine 10 mg, n = 105; placebo, n = 122; haloperidol, n = 112]). Each trial had similar inclusion and exclusion criteria and was designed to meet regulatory standards. The primary outcome measure in each trial was change from baseline PANSS total score to study end point in the intent-to-treat (ITT) population, with last observation carried forward (LOCF) as the primary method of imputing missing data; a mixed model for repeated measures (MMRM) was a prespecified secondary analysis in phase 3 trials.5,7,8

• Asenapine is an approved treatment for schizophrenia in the United States.
• The reported meta-analyses indicate that the efficacy of asenapine for acute schizophrenia is superior to placebo and comparable to several other second-generation antipsychotics.

Meta-analyses were performed by using individual patient data, with asenapine (fixed-dose 5 mg twice daily, fixed-dose 10 mg twice daily, flexible-dose 5 or 10 mg twice daily) and active comparator data pooled across treatment regimens. For all analyses, data from small centers were pooled, as specified in the original statistical analysis plan for each protocol.

For LOCF analysis, meta-analyses were performed using analysis of covariance, with fixed factors for protocol, center (nested in protocol), and randomized treatment group; baseline PANSS total score was a covariate. This method was an extension of the primary analysis specified in each study protocol. The MMRM analysis included fixed factors for protocol, center (nested in protocol), randomized treatment group, visit (repeated measure), and visit-by-treatment interaction; the baseline PANSS total score was a fixed covariate. An unstructured variance-covariance structure was used to model within-patient errors. Denominator degrees of freedom were approximated using the Satterthwaite method14; estimates were derived by using the appropriate contrast at week 6.

In asenapine trials, PANSS responders were defined as patients with ≥ 30% decrease from baseline PANSS total score at end point. A meta-analysis on individual patient data was performed by using logistic regression for PANSS responders, with fixed factors for protocol, treatment, and center (nested in protocol); baseline PANSS total score was a covariate. This analysis provided ORs and 95% Wald-type confidence intervals (CIs) for asenapine and active controls versus placebo. The number needed to treat (NNT [the number of patients who must be treated with active drug to achieve 1 additional responder relative to placebo]) was derived from the inverse of the estimated difference in response rates from placebo established using PANSS total score meta-analyses, with binomial regression and the identity link function.

Meta-Analyses of Published Studies With Integrated Asenapine Data

To compare the asenapine treatment effect versus placebo with that of other antipsychotics, all placebo-controlled asenapine clinical trial results were added to a meta-analysis published by Leucht et al12; an additional asenapine trial (041002 [Merck; unpublished data on file; 2000]) supplied data for risperidone but not asenapine (the asenapine dose in this study was below the effective range). For each drug, meta-analysis on the standardized effect size using Hedges’ g was performed by method of DerSimonian and Laird,15 allowing for studies using scales other than the PANSS (eg, the Brief Psychiatric Rating Scale) to be included.

To compare PANSS responder results for asenapine with other antipsychotics, asenapine clinical data were combined with data from Leucht et al.12 However, Leucht et al12 expressed treatment differences as risk ratios for treatment nonresponse with 95% CIs rather than ORs for treatment response. For consistency, risk ratios for treatment nonresponse were calculated for asenapine.

Effect of Publication Year on Treatment Effect

Meta-regression was performed to assess the impact of publication year on antipsychotic treatment effect versus placebo by using a weighted regression analysis with the inverse of the squared standard error of each placebo comparison as a weight and publication year as a predictor of the standardized treatment effect. Although the years of the study period are more likely to impact treatment effect than the study publication year, publication years were used in these analyses because they simplify study sorting and because the publication year roughly correlates with the study periods.

Head-To-Head Network Meta-Analysis: Asenapine Versus Other SGAs

To assess the comparative efficacy of asenapine and other SGAs, a network meta-analysis combined data published by Leucht et al13 with all asenapine trials that included SGA active controls (studies 041004, 041021, 041022).6-8 In addition, data from the first 6 weeks of a 52-week asenapine safety trial (25517 [NCT00212784]: asenapine 5 or 10 mg, n = 869; olanzapine, n = 297)16 that used olanzapine as the active control were included. In total, data from 49 head-to-head SGA comparison studies were updated with data from 4 asenapine clinical trials (117 treatment groups; 14,861 patients). As in Leucht et al,13 we used change from baseline PANSS total score as the primary end point; LOCF accounted for imputed missing data.

The network meta-analysis proceeded in 2 steps. First, for each pair of SGAs for which data were available, a meta-analysis using the DerSimonian and Laird15 random effects model was performed to estimate relative efficacy and associated variances, taking into account possible heterogeneity. Second, these meta-analytic results were entered into a single linear regression.17 The design matrix was specified with indicators for each separate comparison (by study), and fixed variance components were used that included the DerSimonian and Laird15 random effects contribution. By designating 1 SGA as a reference, relative efficacy and associated 95% CIs versus all other SGAs were estimated with maximum likelihood techniques by using the relative efficacy of all available trials.

RESULTS

Data from the ITT populations of each study, as defined in the study protocols, were included in the meta-analyses.

Meta-Analyses of the Asenapine Clinical Trial Program

Mean change in PANSS total score: asenapine and active controls versus placebo. Meta-analysis of change from baseline PANSS total score indicated that asenapine was superior to placebo when using LOCF (-3.6; P = .002) and MMRM (-4.1; P = .001; Figure 1).5-8 The magnitude of the pooled asenapine effect was comparable to the pooled effect versus placebo for active controls used in the same trials (LOCF, -4.0, P = .002; MMRM, -4.8, P = .001; Figure 1).

PANSS response rates: asenapine and active controls versus placebo. The PANSS response rates with asenapine and active controls numerically exceeded placebo in 3 of 4 trials (see Supplementary eTable 1 at PSYCHIATRIST.COM). Asenapine was statistically superior to placebo in 2 trials (study 0410046: 5 mg twice a day, P = .029; study 0410235: 5 mg twice a day, P < .01, and 10 mg twice a day, P = .016); active controls were statistically superior to placebo in 1 trial (study 0410046: olanzapine 15 mg every day, P = .011). When study noncompleters (ie, those who did not complete a study for any reason) were categorized as PANSS nonresponders, PANSS response rates for asenapine were statistically superior to placebo in 1 study (study 0410235: 5 mg twice a day, P = .001, and 10 mg twice a day, P = .012), and none of the active controls were superior to placebo (see Supplementary eTable 1).

Treatment response ORs: asenapine and active controls versus placebo. Odds ratios for treatment response with asenapine and active controls generally exceeded 1.0 across individual trials (see Supplementary eTable 1), with 95% CIs above 1.0 in those cases in which PANSS response rates were statistically higher than those for placebo.

Meta-analysis of the ORs for treatment response reported statistical superiority of asenapine over placebo (P < .001; Figure 2)5-8,18; the corresponding NNT was 10.2. Comparable findings were obtained for pooled active controls (P = .002; Figure 2); the NNT was 12.0. When study noncompleters (ie, those who did not complete a study for any reason) were considered nonresponders, similar ORs were reported for asenapine (1.6 [95% CI, 1.2-2.3]; P = .003; NNT = 14.1) and active controls (1.5 [95% CI, 1.1-2.2]; P = .019; NNT = 17.3).

Meta-Analyses of Published Studies With Integrated Asenapine Data

Mean change in PANSS total score: asenapine and SGAs versus placebo and the effects of publication year. To illustrate the effect size of publication year, asenapine data were added to the findings that were based on Leucht et al.12 Figure 35-8,18-50 summarizes the standardized effect sizes (Hedges’ g) versus placebo. Asenapine and active-control effect sizes versus placebo in asenapine clinical trials were somewhat lower than what has been historically observed for other SGAs. The meta-regression confirmed that effect size versus placebo has decreased over time (P = .01; see Supplementary eFigure 1).

Treatment nonresponse risk ratio: asenapine and SGAs versus placebo. Individual trial risk ratios for treatment nonresponse for asenapine and other SGAs did not significantly differ from placebo. When study results were pooled, risk ratios tended to favor active drug over placebo (Figure 4).

Head-To-Head Network Meta-Analysis: Asenapine Versus Other SGAs

The network of head-to-head SGA comparisons can be schematically depicted by a line diagram, with the width of each line reflecting the size of the database available for direct comparison between 2 drugs (see Supplementary eFigure 2; thicker lines indicate larger numbers of patients). Estimates are based on all connections and reveal that relative PANSS total score reductions were highest with olanzapine and lowest with ziprasidone (see Supplementary eTable 2). Relative efficacy was more favorable for asenapine compared with clozapine, sertindole, quetiapine, aripiprazole, and ziprasidone; effect size with asenapine was less favorable compared with olanzapine, risperidone, and amisulpride, with estimated differences versus asenapine ranging from 3.9 points more than ziprasidone to 2.9 points less than olanzapine (Figure 5). Asenapine ranked fourth among the 8 agents in this analysis, while olanzapine ranked first and ziprasidone ranked last. The differences between asenapine and other drugs, with the exception of ziprasidone, were not statistically significant.

DISCUSSION

These meta-analytic results support the efficacy of asenapine in the treatment of acute schizophrenia, showing superiority over placebo in the change from baseline PANSS total score (the primary end point for asenapine schizophrenia trials). The effect size of asenapine was comparable to that of active controls.

The average effect sizes versus placebo for asenapine and active control were lower than what has been historically observed for SGAs and also illustrated by a “failed” study8 in which neither asenapine nor olanzapine were significantly different from placebo. However, the finding of reduced efficacy is not unexpected; as demonstrated by the meta-regression of publication year on effect size, treatment effects versus placebo in clinical trials have decreased over time.9,10 Placebo response also tended to be larger than historically expected, a finding consistent with previous reports of increased placebo response in recently conducted clinical trials.10

Given the issues surrounding decreased effect size and increased placebo response over time, direct head-to-head comparisons of agents used in the same trial are likely to be the best means of judging the relative efficacy of different agents. The primary disadvantage of relying on head-to-head comparisons among antipsychotics is the inability to perform comparisons among all agents owing to a lack of sufficient data. However, this disadvantage can be overcome to a large extent with network meta-analysis.

By using network meta-analysis to establish relative efficacy for all agents for which data are available, and the precision of that relative efficacy, comparisons for which direct head-to-head data are not available can be made. For example, comparisons with olanzapine are available for many drugs. By comparing those relative efficacies, conclusions can also be drawn regarding the relative efficacy of drugs that were not directly compared. As a result, fair conclusions can be made from the relative abundance of some head-to-head comparisons. Therefore, what this publication adds, in addition to the assessment of the efficacy of asenapine, is a method for estimating relative efficacy across a complete SGA network.

Similar analyses have been conducted for major depressive disorder51; however, our approach is preferable because maximum-likelihood-based estimates are used instead of Markov chain Monte Carlo simulations.51 Because the maximum-likelihood-based estimates technique is relatively new, it is important to note the key assumption that is required for interpretation of the analyses. That is, it is assumed that differences in mean change from baseline in PANSS total score across studies are comparable, despite differences in patient population, treatment duration, time the study was conducted, and other factors. In essence, this assumption exists in any meta-analysis, and, ultimately, meta-analyses are only as good as the underlying studies on which they are based. No additional assumptions are required for a network meta-analysis, which can therefore be considered as the appropriate method for quantitatively integrating individual study results across drugs.

Although mean change from baseline PANSS total score is widely accepted as the primary efficacy end point in clinical trials and has been used in the current analysis, additional efficacy measures can be made with Brief Psychiatric Rating Scale or Clinical Global Impressions scales. However, the clinical relevance of an average treatment effect remains difficult to interpret. Analysis of individual treatment response rates may be more easily understood from a clinical perspective. Within the asenapine program, PANSS responder rates with asenapine were superior to placebo and comparable to active controls. When expressed as NNT, 10.2 patients need to be treated with asenapine (12.0 for pooled comparators) to achieve 1 additional responder compared with placebo. To compare the effect of asenapine versus placebo with that of other antipsychotics, we used data from Leucht et al,12 in which responder rates were expressed as risk ratios for nonresponse. This analysis revealed that antipsychotics were not superior to placebo in many individual trials, possibly due to variations in the response to treatment across trials, but all were superior to placebo when pooled with meta-analytic techniques. When we used comparison with placebo as the reference for relative efficacy, the treatment nonresponse risk ratio for asenapine was comparable to that for olanzapine, ziprasidone, and aripiprazole and numerically higher than that for risperidone, zotepine, and amisulpride.

Although these analyses focus on the “average” efficacy response, individual responses can vary considerably across patients. The clinician is continually challenged to select the appropriate antipsychotic even though an individual patient’s response cannot be reliably predicted before treatment. The same is true for safety and tolerability responses, with some patients having markedly better tolerance for one drug than another. Although we have included only efficacy measures in the present analysis, it should be noted that efficacy together with safety and tolerability predicts treatment outcome of SGAs. In essence, drug selection occurs by trial and error, with clinicians switching among agents based on the patient’s response to and tolerance of treatment. In that sense, we show that asenapine offers clinicians another alternative, with efficacy comparable to that of most other SGAs.

Limitation

As with all meta-analyses, our findings are indirect comparisons, as the data for other SGAs were extrapolated from different studies with all the inherent limitations specific to those individual studies. Some of these limitations include differences in patient population, treatment duration, and time the study was conducted.

In conclusion, these meta-analyses provide additional support for the superiority of asenapine over placebo for acute schizophrenia. Further, the network meta-analysis suggests that the efficacy of asenapine is comparable to or nonsignificantly better than that of several other SGAs in treating acute schizophrenia.

Drug names: aripiprazole (Abilify), asenapine (Saphris), clozapine (FazaClo, Clozaril, and others), haloperidol (Haldol and others), lithium (Lithobid and others), olanzapine (Zyprexa and others), quetiapine (Seroquel and others), risperidone (Risperdal and others), ziprasidone (Geodon and others).

Author affiliations: Merck, Rahway, New Jersey (Drs Szegedi and Mackle, and Cazorla); and Merck Sharp & Dohme, Oss, The Netherlands (Drs Verweij and Fennema and Mr van Duijnhoven).

Author contributions: Dr Szegedi was involved in the oversight of the original trials, in the analysis and interpretation of these analyses, and in the preparation and finalization of the manuscript. Drs Mackle and Cazorla participated in the design of processes and standards for data acquisition during the original trials and in the harmonization of protocol interpretation across sites involved in the original trials, contributed to the interpretation of data, and contributed to the development and critical review of the intellectual content of the manuscript. Drs Verweij and Fennema and Mr van Duijnhoven conducted the literature searches and statistical analyses, with Dr Fennema also contributing substantially to the interpretation of the data. All authors approved of the submission of the final draft of the manuscript.

Potential conflicts of interest: Drs Szegedi and Mackle are full-time employees of and stock shareholders in Merck. Drs Verweij and Fennema are employees of Merck Sharp & Dohme. Dr Cazorla and Mr van Duijnhoven were employees of Merck and Merck Sharp & Dohme at the time of the study.

Funding/support: This work was supported by funding from Merck, Whitehouse Station, New Jersey.

Role of sponsor: The study sponsor (Merck) was responsible for the study design; the collection, analysis, and interpretation of the data; the writing of the report; and the decision to submit the paper for publication.

Previous presentation: Part of the data were presented at American Psychiatric Nurses Association 25th Annual Conference; October 19-22, 2011; Anaheim, California.

Acknowledgments: Advice and assistance were provided by Siem Heisterkamp, PhD (formerly of Merck Sharp & Dohme, currently at Grünenthal GmbH, Aachen, Germany), and Sabine Braat, PhD, of Merck Sharp & Dohme. Medical writing and editorial assistance were provided by Karen Zimmermann, PhD, and Craig Slawecki, PhD, of Complete Healthcare Communications, Chadds Ford, Pennsylvania, and PAREXEL, Hackensack, New Jersey, both funded by Merck Sharp & Dohme Corp, a subsidiary of Merck & Co, Inc, Whitehouse Station, New Jersey.

Supplementary material: Available at PSYCHIATRIST.COM.

REFERENCES

1. Saphris [prescribing information]. Whitehouse Station, NJ: Merck & Co, Inc. http://www.spfiles.com/pisaphrisv1.pdf. Updated May 2012. Accessed September 10, 2012.

2. Shahid M, Walker GB, Zorn SH, et al. Asenapine: a novel psychopharmacologic agent with a unique human receptor signature. J Psychopharmacol. 2009;23(1):65-73. PubMed doi:10.1177/0269881107082944

3. Edwards SJ, Smith CJ. Tolerability of atypical antipsychotics in the treatment of adults with schizophrenia or bipolar disorder: a mixed treatment comparison of randomized controlled trials. Clin Ther. 2009;31(pt 1):1345-1359. PubMed doi:10.1016/j.clinthera.2009.07.004

4. Glick ID, Bossie CA, Alphs L, et al. Onset and persistence of antipsychotic response in patients with schizophrenia. J Clin Psychopharmacol. 2009;29(6):542-547. PubMed doi:10.1097/JCP.0b013e3181befa2a

5. Kane JM, Cohen M, Zhao J, et al. Efficacy and safety of asenapine in a placebo- and haloperidol-controlled trial in patients with acute exacerbation of schizophrenia. J Clin Psychopharmacol. 2010;30(2):106-115. PubMed doi:10.1097/JCP.0b013e3181d35d6b

6. Potkin SG, Cohen M, Panagides J. Efficacy and tolerability of asenapine in acute schizophrenia: a placebo- and risperidone-controlled trial. J Clin Psychiatry. 2007;68(10):1492-1500. PubMed doi:10.4088/JCP.v68n1004

7. Efficacy and safety of asenapine with placebo and olanzapine (41021) (P05933) (completed). Clinicaltrials.gov Web site. http://clinicaltrials.gov/ct2/show/NCT00156117?term=NCT00156117&rank=1. Updated August 29, 2012. Accessed October 31, 2012.

8. Efficacy and safety of asenapine with placebo and olanzapine (41022) (completed). Clinicaltrials.gov Web site. http://clinicaltrials.gov/ct2/show/study/NCT00151424?term=NCT00151424&rank=1. Updated August 29, 2012. Accessed October 31, 2012.

9. Julious S, Wang S. How biased are indirect comparisons, particularly when comparisons are made over time in controlled trials? Drug Inf J. 2008;42(6):625-633. doi:10.1177/009286150804200610

10. Kemp AS, Schooler NR, Kalali AH, et al. What is causing the reduced drug-placebo difference in recent schizophrenia clinical trials and what can be done about it? Schizophr Bull. 2010;36(3):504-509. PubMed doi:10.1093/schbul/sbn110

11. Kay SR, Fiszbein A, Opler LA. The Positive and Negative Syndrome Scale (PANSS) for schizophrenia. Schizophr Bull. 1987;13(2):261-276. PubMed

12. Leucht S, Arbter D, Engel RR, et al. How effective are second-generation antipsychotic drugs? a meta-analysis of placebo-controlled trials. Mol Psychiatry. 2009;14(4):429-447. PubMed

13. Leucht S, Komossa K, Rummel-Kluge C, et al. A meta-analysis of head-to-head comparisons of second-generation antipsychotics in the treatment of schizophrenia. Am J Psychiatry. 2009;166(2):152-163. PubMed doi:10.1176/appi.ajp.2008.08030368

14. Satterthwaite FE. An approximate distribution of estimates of variance components. Biometrics. 1946;2(6):110-114. PubMed doi:10.2307/3002019

15. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7(3):177-188. PubMed doi:10.1016/0197-2456(86)90046-2

16. Schoemaker J, Naber D, Vrijland P, et al. Long-term assessment of asenapine vs olanzapine in patients with schizophrenia or schizoaffective disorder. Pharmacopsychiatry. 2010;43(4):138-146. PubMed doi:10.1055/s-0030-1248313

17. van Houwelingen HC, Arends LR, Stijnen T. Advanced methods in meta-analysis: multivariate approach and meta-regression. Stat Med. 2002;21(4):589-624. PubMed doi:10.1002/sim.1040

18. Modell S, Werner C, Spevakne-Gorocs T, et al. Efficacy and safety of lower doses of aripiprazole. Pharmacopsychiatry. 2005;38:A160. doi:10.1055/s-2005-918782

19. Arato M, O’ Connor R, Meltzer HY; ZEUS Study Group. A 1-year, double-blind, placebo-controlled trial of ziprasidone 40, 80 and 160 mg/day in chronic schizophrenia: the Ziprasidone Extended Use in Schizophrenia (ZEUS) study. Int Clin Psychopharmacol. 2002;17(5):207-215. PubMed doi:10.1097/00004850-200209000-00001

20. Arvanitis LA, Miller BG. Multiple fixed doses of “Seroquel” (quetiapine) in patients with acute exacerbation of schizophrenia: a comparison with haloperidol and placebo. The Seroquel Trial 13 Study Group. Biol Psychiatry. 1997;42(4):233-246. PubMed doi:10.1016/S0006-3223(97)00190-X

21. Bai YM, Yu SC, Lin CC. Risperidone for severe tardive dyskinesia: a 12-week randomized, double-blind, placebo-controlled study. J Clin Psychiatry. 2003;64(11):1342-1348. PubMed doi:10.4088/JCP.v64n1110

22. Beasley CM Jr, Tollefson G, Tran P, et al. Olanzapine versus placebo and haloperidol: acute phase results of the North American double-blind olanzapine trial. Neuropsychopharmacology. 1996;14(2):111-123. PubMed doi:10.1016/0893-133X(95)00069-P

23. Beasley CM Jr, Sanger T, Satterlee W, et al. Olanzapine versus placebo: results of a double-blind, fixed-dose olanzapine trial. Psychopharmacology (Berl). 1996;124(1-2):159-167. PubMed doi:10.1007/BF02245617

24. Beasley CM Jr, Sutton VK, Hamilton SH, et al; Olanzapine Relapse Prevention Study Group. A double-blind, randomized, placebo-controlled trial of olanzapine in the prevention of psychotic relapse. J Clin Psychopharmacol. 2003;23(6):582-594. PubMed doi:10.1097/01.jcp.0000095348.32154.ec

25. Borison RL, Pathiraja AP, Diamond BI, et al. Risperidone: clinical safety and efficacy in schizophrenia. Psychopharmacol Bull. 1992;28(2):213-218. PubMed

26. Borison RL, Arvanitis LA, Miller BG; US SEROQUEL Study Group. ICI 204,636, an atypical antipsychotic: efficacy and safety in a multicenter, placebo-controlled trial in patients with schizophrenia. J Clin Psychopharmacol. 1996;16(2):158-169. PubMed doi:10.1097/00004714-199604000-00008

27. Canuso CM, Dirks B, Carothers J, et al. Randomized, double-blind, placebo-controlled study of paliperidone extended-release and quetiapine in inpatients with recently exacerbated schizophrenia. Am J Psychiatry. 2009;166(6):691-701. PubMed doi:10.1176/appi.ajp.2009.08040613

28. Chouinard G, Jones B, Remington G, et al. A Canadian multicenter placebo-controlled study of fixed doses of risperidone and haloperidol in the treatment of chronic schizophrenic patients. J Clin Psychopharmacol. 1993;13(1):25-40. PubMed doi:10.1097/00004714-199302000-00004

29. Cooper SJ, Tweed J, Raniwalla J, et al. A placebo-controlled comparison of zotepine versus chlorpromazine in patients with acute exacerbation of schizophrenia. Acta Psychiatr Scand. 2000;101(3):218-225. PubMed doi:10.1046/j.0902-4441.2000.ap90086.x

30. Cooper SJ, Butler A, Tweed J, et al. Zotepine in the prevention of recurrence: a randomised, double-blind, placebo-controlled study for chronic schizophrenia. Psychopharmacology (Berl). 2000;150(3):237-243. PubMed doi:10.1007/s002130000452

31. Daniel DG, Zimbroff DL, Potkin SG, et al; Ziprasidone Study Group. Ziprasidone 80 mg/day and 160 mg/day in the acute exacerbation of schizophrenia and schizoaffective disorder: a 6-week placebo-controlled trial. Neuropsychopharmacology. 1999;20(5):491-505. PubMed doi:10.1016/S0893-133X(98)00090-6

32. Danion JM, Rein W, Fleurot O; Amisulpride Study Group. Improvement of schizophrenic patients with primary negative symptoms treated with amisulpride. Am J Psychiatry. 1999;156(4):610-616. PubMed

33. Fabre LF Jr, Arvanitis L, Pultz J, et al. ICI 204,636, a novel, atypical antipsychotic: early indication of safety and efficacy in patients with chronic and subchronic schizophrenia. Clin Ther. 1995;17(3):366-378. PubMed doi:10.1016/0149-2918(95)80102-2

34. Honigfeld G, Patin J, Singer J. Clozapine: antipsychotic activity in treatment-resistant schizophrenics. Adv Ther. 1984;1:77-97.

35. Kane JM, Carson WH, Saha AR, et al. Efficacy and safety of aripiprazole and haloperidol versus placebo in patients with schizophrenia and schizoaffective disorder. J Clin Psychiatry. 2002;63(9):763-771. PubMed doi:10.4088/JCP.v63n0903

36. Keck P Jr, Buffenstein A, Ferguson J, et al. Ziprasidone 40 and 120 mg/day in the acute exacerbation of schizophrenia and schizoaffective disorder: a 4-week placebo-controlled trial. Psychopharmacology (Berl). 1998;140(2):173-184. PubMed doi:10.1007/s002130050755

37. Kryzhanovskaya LA, Jeste DV, Young CA, et al. A review of treatment-emergent adverse events during olanzapine clinical trials in elderly patients with dementia. J Clin Psychiatry. 2006;67(6):933-945. PubMed doi:10.4088/JCP.v67n0610

38. Lecrubier Y, Quintin P, Bouhassira M, et al. The treatment of negative symptoms and deficit states of chronic schizophrenia: olanzapine compared to amisulpride and placebo in a 6-month double-blind controlled clinical trial. Acta Psychiatr Scand. 2006;114(5):319-327. PubMed doi:10.1111/j.1600-0447.2006.00887.x

39. Marder SR, Meibach RC. Risperidone in the treatment of schizophrenia. Am J Psychiatry. 1994;151(6):825-835. PubMed

40. Möller HJ, Riedel M, Müller N, et al. Zotepine versus placebo in the treatment of schizophrenic patients with stable primary negative symptoms: a randomized double-blind multicenter trial. Pharmacopsychiatry. 2004;37(6):270-278. PubMed doi:10.1055/s-2004-832683

41. Pigott TA, Carson WH, Saha AR, et al; Aripiprazole Study Group. Aripiprazole for the prevention of relapse in stabilized patients with chronic schizophrenia: a placebo-controlled 26-week study. J Clin Psychiatry. 2003;64(9):1048-1056. PubMed doi:10.4088/JCP.v64n0910

42. Potkin SG, Saha AR, Kujawa MJ, et al. Aripiprazole, an antipsychotic with a novel mechanism of action, and risperidone vs placebo in patients with schizophrenia and schizoaffective disorder. Arch Gen Psychiatry. 2003;60(7):681-690. PubMed doi:10.1001/archpsyc.60.7.681

43. Potkin SG, Gharabawi GM, Greenspan AJ, et al. A double-blind comparison of risperidone, quetiapine and placebo in patients with schizophrenia experiencing an acute exacerbation requiring hospitalization. Schizophr Res. 2006;85(1-3):254-265. PubMed doi:10.1016/j.schres.2006.03.027

44. Small JG, Hirsch SR, Arvanitis LA, et al; Seroquel Study Group. Quetiapine in patients with schizophrenia: a high- and low-dose double-blind comparison with placebo. Arch Gen Psychiatry. 1997;54(6):549-557. PubMed doi:10.1001/archpsyc.1997.01830180067009

45. Center For Drug Evaluation and Research. Application 21-436. US Food and Drug Administration Medical Review(s). http://www.accessdata.fda.gov/drugsatfda_docs/nda/2002/21-436_Abilify_medr_P1.pdf. Accessed September 10, 2012.

46. Center for Drug Evaluation and Research. Approval Package for Application Number: 20588/S002 and 20272/S007. US Food and Drug Administration Medical Review(s). http://www.accessdata.fda.gov/drugsatfda_docs/nda/97/20588.pdf. Accessed September 10, 2012.

47. van Kammen DP, McEvoy JP, Targum SD, et al. A randomized, controlled, dose-ranging trial of sertindole in patients with schizophrenia. Psychopharmacology (Berl). 1996;124(1-2):168-175. PubMed doi:10.1007/BF02245618

48. Zborowski J, Schmitz P, Staser J, et al. Efficacy and safety of sertindole in a trial of schizophrenic patients. Biol Psychiatry. 1995;37(9):661-662. doi:10.1016/0006-3223(95)94656-H

49. Zimbroff DL, Kane JM, Tamminga CA, et al; Sertindole Study Group. Controlled, dose-response study of sertindole and haloperidol in the treatment of schizophrenia. Am J Psychiatry. 1997;154(6):782-791. PubMed

50. Corrigan MH, Gallen CC, Bonura ML, et al; Sonepiprazole Study Group. Effectiveness of the selective D4 antagonist sonepiprazole in schizophrenia: a placebo-controlled trial. Biol Psychiatry. 2004;55(5):445-451. PubMed doi:10.1016/j.biopsych.2003.10.004

51. Cipriani A, Furukawa TA, Salanti G, et al. Comparative efficacy and acceptability of 12 new-generation antidepressants: a multiple-treatments meta-analysis. Lancet. 2009;373(9665):746-758. PubMed doi:10.1016/S0140-6736(09)60046-5