HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS Psychiatric Bulletin Advances in Psychiatric Treatment All RCPsych Journals		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Submit an eLetter View responses Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Related articles in BJP Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by El-Sayeh, H. G. Articles by Adams, C. E. Search for Related Content PubMed PubMed Citation Articles by El-Sayeh, H. G. Articles by Adams, C. E.

Aripiprazole for schizophrenia

Systematic review

Academic Unit of Psychiatry, University of Leeds, Leeds, UK

C. Morganti, MD

Ospedale Niguarda Ca'Granda, Milan, Italy

C. E. Adams, MRCPsych, MD

Academic Unit of Psychiatry, University of Leeds, Leeds, UK

Correspondence: Dr H.G. El-Sayeh, Briary Wing, Harrogate District Hospital, Lancaster Park Road, Harrogate HG2 7SX, UK. Tel: +44 (0)1423 553 682; fax: +44 (0)1423 553 379; email: hany.el-sayeh{at}chrd-pct.nhs.uk

Declaration of interest None.

Funding detailed in Acknowledgements.

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHOD RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Background Aripiprazole is an atypical antipsychotic that is reported to be effective in the treatment of schizophrenia.

Aims To investigate the effects of aripiprazole on patients with schizophrenia and schizophrenia-like psychoses by conducting a systematic review of randomised controlled trials (RCTs).

Method Database and manual searches and direct contact were used to identify relevant RCTs.

Results We included 10 randomised controlled studies (involving a total of 4125 patients), but study attrition waslarge and the standard of data reporting was poor. Compared with placebo, aripiprazole treatment was associated with a significant decrease in relapse rates, increased compliance with the study protocol, and a decrease in prolactin levels below the expected values. Compared with risperidone, aripiprazole caused less elevation of prolactin levels and less prolongation of the average QTc interval.

Conclusions Aripiprazole has been licensed despite the fact that few reliable data on this drug are publicly available. It may be effective for treatment of schizophrenia, but in terms of tolerability and global outcomes it shows little difference from existing antipsychotics.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHOD RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Atypical or second-generation antipsychotics are said to differ from conventional or typical antipsychotics in terms of their effects on the positive and negative symptoms of schizophrenia and on cognition, and in terms of their adverse effect profiles (Gelder et al, 2000). Aripiprazole is the prototype of a `third generation' of antipsychotics - the so-called dopamine-serotonin-system stabilisers (Rivas-Vasquez, 2003). It is claimed to be at least as effective as haloperidol in the treatment of positive and negative symptoms of schizophrenia, and it may cause fewer adverse effects. Aripiprazole is reported to be useful in all phases of schizophrenia, and to enhance cognitive function (Rivas-Vasquez, 2003). In 2002 the drug was granted approvable status by the US Food and Drug Administration (FDA) for the treatment of schizophrenia (Dubitsky et al, 2002). It has been included in recent guidelines on schizophrenia treatment (American Psychiatric Association Work Group on Schizophrenia, 2004) and it is licensed for use in several other countries, including the UK. We here report the findings of a systematic review of randomised controlled trials (RCTs) of the effects of aripiprazole.

	METHOD

TOP ABSTRACT INTRODUCTION METHOD RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Search strategy
We searched the Cochrane Schizophrenia Group's register (August 2004) and we also hand-searched relevant journals and conference proceedings, and used several grey literature sources (including pharmaceutical industry materials and non-systematic internet searches). In addition, we inspected the references cited in identified studies for further trials and we examined the FDA website. We also contacted relevant authors and the manufacturers of aripiprazole (Bristol-Myers Squibb and Otsuka pharmaceuticals). Full details have been published previously (El-Sayeh & Morganti, 2004).

Selection and inclusion criteria
We reliably selected RCTs that compared aripiprazole at any dose (the recommended target dose is 10-15 mg/day, range 10-30 mg/day) with any other antipsychotics or placebo in the treatment of people with schizophrenia or schizophrenia-like psychoses. Our primary outcome was relapse, but we also investigated a number of other outcomes, including death, mental state, cognitive functioning, adverse effects and quality of life. Before we viewed the data, we stipulated that outcome measures were to be categorised as short-term (up to 12 weeks), medium-term (13-26 weeks) or long-term (over 26 weeks). We assessed study quality using the criteria described in the Cochrane Reviewers' Handbook 4.2.0 (Clarke & Oxman, 2003).

Data analysis
We analysed the data using RevMan version 4.2.3 (Cochrane Collaboration, Oxford, UK; see http://www.cc-ims.net/RevMan/current.htm), and we calculated random-effects relative risk (RR) and its 95% confidence interval by intention-to-treat analysis. Where possible, we calculated the number needed to treat (NNT) and the number needed to harm (NNH) (see http://www.nntonline.net). On the condition that more than 60% of participants were accounted for with respect to any given study outcome, everyone allocated to the intervention was counted, whether they completed the follow-up or not. It was assumed that those individuals who dropped out had a negative outcome (other than death). Continuous data were synthesised using weighted mean difference. Statistical heterogeneity was assessed by inspecting the relevant graph and was supplemented using the I-squared statistic (Higgins et al, 2003). If inconsistency was high (>75%), the data were not pooled, but were presented separately and the reasons for heterogeneity were investigated.

	RESULTS

TOP ABSTRACT INTRODUCTION METHOD RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
We identified over 400 citations, of which 54 reported 10 relevant studies (Table 1). All of these 10 studies were randomised, and all but one (Kern et al, 2001) were double-blind. However, none of them made the method of randomisation explicit or tested masking. Consequently they all carry a moderate risk of bias and may therefore overestimate the positive effects of aripiprazole (Clarke & Oxman, 2003). Six of the studies reported data up to 12 weeks (Carson et al, 2000, 2002; Daniel et al, 2000; Adson et al, 2002; Kane et al, 2003; Potkin et al, 2003), three reported data up to 6 months (Kern et al, 2001; Carson et al, 2002; McQuade et al, 2003) and one reported data at 1 year (Kujawa et al, 2002).

Most of the studies were conducted in North America or Europe and involved between 103 (Csernansky et al, 2002) and 1294 (Kujawa et al, 2002) participants. Most of the participants were male inpatients in their thirties or forties. The majority had well-defined diagnoses of schizophrenia with little comorbidity. Such individuals represent a minority of patients in everyday care. However, we would also like to point out that there is variation in the clinical condition of the patients who were randomised in the different studies. Although the majority of the participants were noted to have had an acute relapse of schizophrenia, in some studies the participants had chronic stable schizophrenia or treatment-resistant schizophrenia. Any interpretation of the findings of the meta-analyses must take into account this clinical heterogeneity, and the fact that it could make our findings more rather than less applicable to everyday care.

Aripiprazole was compared with placebo in 6 studies (encompassing a total of 1628 patients), with haloperidol in 4 studies (n=1913), with perphenazine in 1 study (n=300), with olanzapine in 2 studies (n=573) and with risperidone in 1 study (n=301). Aripiprazole was given over a wide range of doses (2-30 mg/day) (Dubitsky et al, 2002). Two studies reported deaths (Carson et al, 2000; Adson et al, 2002) but did not supply usable outcomes, although data were available on the FDA website. We found no usable data on service outcomes, general functioning, behaviour, engagement with mental health services, satisfaction with treatment, economic outcomes or cognitive functioning. Although relapse was the primary outcome measure for this review, only one study that compared aripiprazole treatment with placebo (Carson et al, 2002) provided data on this outcome, and relapse in that study was defined by changes in rating scale scores, not by re-hospitalisation rates as are commonly used. Seven of the 10 studies that were included reported data in terms of both a last-observation-carried-forward (LOCF) analysis and an observed-cases analysis (where observed cases are defined as those who completed the trial). We could not use the LOCF data because of the high drop-out rates reported in the studies as well as the tendency to report mean figures without providing a measure of variance.

More participants who were allocated to aripiprazole completed the studies compared with those allocated placebo (n=1658, 6 RCTs, RR (leaving study for any reason)=0.68, 95% CI 0.55-0.86; NNT=4, 95% CI 6-11). However, aripiprazole showed no significant advantage over typical antipsychotics (n=2213, 5 RCTs, RR (leaving study for any reason by 12 weeks)=0.90, 95% CI 0.76-1.05; see Fig. 1). In total, 52% of participants left these 5 studies early. If an LOCF analysis were to have been used, this would have meant that large assumptions would have to be made about the outcomes of over half the participants. Before we viewed the data, we had stated that making such assumptions for over 40% of participants rendered the outcomes impossible to interpret. In the comparison with the other atypical antipsychotic medications, 53% of the patients who were allocated aripiprazole treatment left the studies before the end of the trial, compared with 58% of the patients in the comparison groups (n=618, 2 RCTs, RR (leaving for any reason)=1.05, 95% CI 0.93-1.19). When drop-out was due to adverse effects, again we found no significant difference between aripiprazole and other atypical antipsychotics (n=618, 2 RCTs, 5% v. 8%, RR=0.78, 95% CI 0.42-1.42).

View larger version (28K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1 Comparison of aripiprazole with placebo in participants who left the study early for any reason.

Aripiprazole had a favourable effect compared with placebo on a range of adverse effects, including headache (n=615, 2 RCTs, RR=1.04, 95% CI 0.76-1.43), anxiety (n=615, 2 RCTs, RR=0.86, 95% CI 0.53-1.39), weight gain (n=615, 2 RCTs, RR=2.64, 95% CI 0.70-9.95), extrapyramidal side-effects (n=615, 2 RCTs, RR=1.63, 95% CI 0.15-17.55) and changes in QTc interval (aripiprazole 20 mg) (n=204, 1 RCT, 95% CI 73.19 to 9.19). Aripiprazole did appear to be significantly superior to placebo in terms of the number of patients who showed a rise in serum prolactin level to at least 23 ng/ml in one short study (n=305, RR=0.32, 95% CI 0.13-0.81; NNT=14, 95% CI 11-50) (Potkin et al, 2003). Because of high drop-out rates and underreporting in the included studies, we could only derive data on adverse effects for aripiprazole compared with typical antipsychotics from a single study which used a perphenazine control (Kane et al, 2003). The results suggest that there is little significant difference in specific adverse effects between aripiprazole and this typical antipsychotic (Fig. 2), apart from the finding that patients who were allocated aripiprazole required less antiparkinsonian medication (NNT=4, 95% CI 3-5) and more often experienced insomnia (NNH=4, 95% CI 3-9). No outcomes were available to allow comparison of aripiprazole with typical antipsychotic medication with regard to changes in the QTc interval.

View larger version (27K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 2 Comparison of aripiprazole with typical antipsychotics with regard to adverse effects.

View larger version (27K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 3 Comparison of aripiprazole with other atypical antipsychotics with regard to adverse effects.

Eight people who were allocated aripiprazole are known to have died in open-label extension arms of two of the studies (total n=834) (Carson et al, 2000; Adson et al, 2002). However, the authors note that the causes of these eight deaths included suicide. The people who were randomised in these two trials were experiencing an acute relapse of schizophrenia, and this may partly explain the observed mortality figures. None of these deaths occurred in the randomised controlled phase of these trials.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHOD RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Limitations of data on deaths
Despite the eight deaths reported in the results section above, this research finding has not been widely disseminated. Also, on the FDA website, a number of additional deaths are reported in those known to be allocated to aripiprazole in various trials. However, we have been informed by the pharmaceutical company in question that these deaths did not occur in trials relevant to this review. We are in continued dialogue with Bristol-Myers Squibb, and hope to gain further clarification on these and other data. Not disseminating clear information regarding these people's outcome, we argue, breaks that unspoken contract that occurs between researchers and trial participants at the point of gaining informed consent. Formalising the contract of informed consent, public registration of all future trials before randomisation and clearer dissemination of trial data are all essential steps in rectifying this situation.

Other data limitations
Many of the data used in this review were obtained from conference proceedings and posters, making extraction difficult and double-counting likely. No serious published attempt was made to give each study a unique identifier. A total of 16 relevant studies, including a number of Japanese phase II and phase III studies, were only available on the FDA website and could not be included because the data were incomplete (Dubitsky et al, 2002). Therefore we cannot include these studies without the express assistance of the pharmaceutical companies who own the material. Multiple requests for further information on the highlighted FDA-identified trials have been made by telephone, by e-mail and in person. It is unlikely that patients who gave their informed consent would have understood that their results would remain undisclosed and would therefore not help to inform the care of other people with schizophrenia.

These studies were not designed to provide results of great relevance to everyday care. They were designed in line with the stipulations of the drug regulatory authorities. The majority of trials included well-defined study participants with little comorbidity. The typical antipsychotic drugs of comparison reported in this review were occasionally of such a nature or used at such a dose that they distanced these trials even further from everyday practice. The outcomes are remarkably few in number, of limited duration and poorly reported, and they take little account of the CONSORT statement (Moher et al, 2001), or else they carry such assumptions as to render them meaningless. Accordingly, findings from these studies are difficult to translate into meaningful decisions about patient care. Where complete data on adverse effects are available, the decision to report only events which occur in at least 5-10% of participants means that rare serious adverse events are not recorded. Large amounts of data could not be used for this review, partly because of the poor quality of reporting. Many studies failed to provide standard deviations when reporting mean changes in a particular outcome measure. Other studies failed to report outcomes in more than 40% of randomised patients. In accordance with our protocol (see above section on data analysis), we believe that including data from this population would involve making too many assumptions about final outcomes, and that these data should not be used until further information has been obtained.

Recommendations and implications
We recommend that there should be greater compliance with CONSORT guidance in future studies. The allocation of unique study identifier numbers to minimise confusion, performing an intention-to-treat analysis on all outcomes, and clear presentation of all study data are critical to this process.

This systematic review suggests that aripiprazole does not differ significantly from some typical or other atypical antipsychotics in terms of several global outcomes and adverse effects. However, it does not appear to cause hyperprolactinaemia, an adverse effect that is commonly seen with the typical antipsychotics and even with some of the other atypical antipsychotics. There appeared to be no significant difference in prolongation of the QTc interval compared with placebo, but there was less change in the QTc interval in one study that compared aripiprazole with other atypical antipsychotics (risperidone, n=200 patients). No data were available on the effect on QTc interval compared with typical antipsychotics. Aripiprazole may cause more insomnia than typical antipsychotics, but is perhaps also associated with less need for antiparkinsonian medication. However, the authors acknowledge that because of the lack of available evidence, and the limited numbers of comparator drugs that were used in these trials, further studies using a wider range of comparator drugs may be required before the results can be generalised to an antipsychotic class (either typical or atypical) as a whole.

Aripiprazole is an interesting compound with a novel mechanism of antipsychotic action, but its real effects are unclear, partly as a consequence of the requirements of both the regulatory authorities and the pharmaceutical industry. This review effectively demonstrates why large, long, well-designed, well-conducted and adequately reported pragmatic RCTs should be part of the regulatory authority's requirements. It also illustrates the way in which clinicians, recipients of care, policy makers and even those who work in the pharmaceutical industry are compromised by the limitations of using explanatory trials as the sole basis for allowing a drug to be given a national licence.

	ACKNOWLEDGMENTS

TOP ABSTRACT INTRODUCTION METHOD RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
We thank Gill Rizzello, Mark Fenton and John Rathbone for help with the literature search and with the preparation and editing of this document.

The work received the intramural support of the University of Leeds and Ospedale Niguarda Ca'Granda.

	REFERENCES

TOP ABSTRACT INTRODUCTION METHOD RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Adson, D., Bari, B., Bona, J., et al (2002) Abilify (Aripiprazole) Tablets. Medical Review. Part 1. Rockville, MD: US Food and Drug Administration. http://www.fda.gov/cder/foi/nda/2002/21-436_Abilify.htm [date accessed 01/10/04].

Allebeck, P. (1989) Schizophrenia: a life-shortening disease. Schizophrenia Bulletin, 15, 81-89.[Medline]

American Psychiatric Association Work Group on Schizophrenia (2004) Practice guideline for the treatment of patients with schizophrenia. American Journal of Psychiatry, 161, 15-18.

Carson, W. H., Ali, M., Saha, A. R., et al (2000) A double-blind, placebo-controlled trial of aripiprazole and haloperidol in patients with schizophrenia or schizoaffective disorder. Abstract: 39th Annual Meeting of the American College of Neuropsychopharmacology, Dec 10-14, San Juan, Puerto Rico. Available on request from: Bristol-Myers Squibb Ltd Medical Information (tel: 0800 7311 736).

Carson, W. H., Pigott, T. A., Saha, A. R., et al (2002) Aripiprazole vs. placebo in the treatment of chronic schizophrenia. International Journal of Neuropsychopharmacology, 5 (suppl. 1), 187.

Clarke, M. & Oxman, A. D. (2003) Cochrane Reviewer's Handbook 4.2.0. Oxford: Cochrane Collaboration. http://www.cochrane.dk/cochrane/handbook/hbook.htm

Csernansky, J., Garbutt, J., Goff, D. C., et al (2002) In Dubitsky, G. M., Harris, R., Laughren, T., Hardeman, S. Abilify (Aripiprazole) Tablets, Medical Review, Part 1. Rockville, MD: US Food and Drug Administration. http://www.fda.gov/cder/foi/nda/2002/21-436_Abilify.htm,1-50 [date accessed 01/10/04].

Daniel, D. G., Saha, A. R., Ingenito, G., et al (2000) Aripiprazole, a novel antipsychotic: overview of a phase II study result. International Journal of Neuropsychopharmacology, 3 (suppl. 1), 157.

Dubitsky, G., Harris, R. D., Laughren, T., et al (2002) Abilify (Aripiprazole) Tablets. Medical Review Part 4. Rockville, MD: US Food and Drug Administration. http://www.fda.gov/cder/foi/nda/2002/21-436_Abilify_medr_P4.pdf [date accessed 01/10/04].

El-Sayeh, H. G. & Morganti, C. (2004) Aripiprazole for schizophrenia (Cochrane Review). Cochrane Library, issue 2. Chichester: John Wiley & Sons.

Gelder, M., Lopez-Ibor Jr, J. J. & Andreason, N. C. (2000) Antipsychotic and anticholinergic drugs. In New Oxford Textbook of Psychiatry, Volume 2.

Guy, W. (1976) ECDEU Assessment Manual for Psychopharmacology. Revised DHEW Pub. (ADM). Rockville, MD:National Institute for Mental Health.

Higgins, J. P., Thompson, S. G., Deeks, J. J., et al (2003) Measuring inconsistency in meta-analyses. BMJ, 327, 557 -560.[Free Full Text]

Kane, J., Carson, W., Kujawa, M., et al (2003) Aripiprazole vs perphenazine in treatment-resistant schizophrenia. Abstract: 156th Annual Meeting of the American Psychiatric Association, May 17-22, San Francisco, USA. Available on request from: Bristol-Myers Squibb Ltd Medical Information (tel: 0800 7311 736).

Kay, S. R., Fiszbein, A. & Opler, L. A. (1987) The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophrenia Bulletin, 13, 261 -276.[Medline]

Kern, R. S., Cornblatt, B., Carson, W. H., et al (2001) An open-label comparison of the neurocognitive effects of aripiprazole versus olanzapine in patients with stable psychosis (abstract). Schizophrenia Research, 1-2, 234.[CrossRef]

Kujawa, M., Saha, A. R., Ingenito, G. G., et al (2002) Aripiprazole for long-term maintenance treatment of schizophrenia. International Journal of Neuropsychopharmacology, 5 (suppl. 1), 186.

McQuade, R. D., Jody, D. N., Kujawa, M. J., et al (2003) Long-term weight effects of aripiprazole vs olanzapine. Abstract: 156th Annual Meeting of the American Psychiatric Association, May 17-22, San Francisco, USA. Available on request from: Bristol-Myers Squibb Ltd Medical Information (tel: 0800 7311 736).

Moher, D., Schulz, K. F. & Altman, D. (2001) The CONSORT statement: revised recommendations for improving the quality of reports of parallel-group randomized trials. JAMA, 285, 1987 -1991.[Abstract/Free Full Text]

Office for National Statistics (2001) Age-Standardised Mortality Rates by Sex and Cause. London: Office for National Statistics. www.statistics.gov.uk/StatBase/Expodata/Spreadsheets/D7739.xls

Potkin, S. G., Saha, A. R., Kujawa, M. J., et al (2003) Aripiprazole, an antipsychotic with a novel mechanism of action, and risperidone vs placebo in patients with schizophrenia and schizoaffective disorder. Archives of General Psychiatry, 60, 681 -690.[Abstract/Free Full Text]

Rivas-Vasquez, R. A. (2003) Aripiprazole: a novel antipsychotic with dopamine-stabilising properties. Professional Psychology: Research and Practice, 34, 108 -111.[CrossRef]

Received for publication June 21, 2004. Revision received June 16, 2005. Accepted for publication June 20, 2005.

Related articles in BJP:

From the Editor's desk

Peter Tyrer
BJP 2006 189: 200. [Full Text]  

This article has been cited by other articles:

				C. Veiga-Martinez, M. Perez-Alvarez, and J. M. Garcia-Montes Acceptance and Commitment Therapy Applied to Treatment of Auditory Hallucinations Clinical Case Studies, April 1, 2008; 7(2): 118 - 135. [Abstract] [PDF]

				R. Hodgson, C. Bushe, and R. Hunter Measurement of long-term outcomes in observational and randomised controlled trials The British Journal of Psychiatry, August 1, 2007; 191(50): s78 - s84. [Abstract] [Full Text] [PDF]

				G. A. Silver, R. Gutierrez-Esteinou, and R. D. McQuade Aripiprazole - data on efficacy and associated mortality The British Journal of Psychiatry, July 1, 2007; 191(1): 86 - 86. [Full Text] [PDF]

				P. Lepping Antipsychotics preferred by mental health professionals Psychiatr. Bull., July 1, 2007; 31(7): 272 - 272. [Full Text] [PDF]

				T. J. Crow Aripiprazole - data on efficacy and associated mortality The British Journal of Psychiatry, July 1, 2007; 191(1): 86 - 87. [Full Text] [PDF]

				P. A. M Mortimer Review: aripiprazole for schizophrenia--more high quality research required Evid. Based Ment. Health, February 1, 2007; 10(1): 14 - 14. [Full Text] [PDF]

eLetters:

This Article Abstract Full Text (PDF) Submit an eLetter View responses Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Related articles in BJP Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by El-Sayeh, H. G. Articles by Adams, C. E. Search for Related Content PubMed PubMed Citation Articles by El-Sayeh, H. G. Articles by Adams, C. E.