Schizophrenia Research Forum - A Catalyst for Creative Thinking

Meta-Analysis Finds Antipsychotics Help Prevent Relapse in Schizophrenia

16 May 2012. Antipsychotic drugs significantly reduce the likelihood that a patient with schizophrenia will relapse, according to a new meta-analysis appearing online May 3 in Lancet. The study, led by Stefan Leucht of Universität München, Germany, suggests that maintenance of antipsychotic therapy for up to a year after the illness has stabilized is beneficial to patients.

The findings are consistent with earlier conclusions about antipsychotic drugs, write Jim van Os of Maastricht University, The Netherlands, and Oliver Howes of King’s College London, U.K., in an accompanying commentary. They note also that, "[t]he present review breaks a trend, because systematic re-evaluations in the past two years have cast doubt on the perceived effectiveness of both pharmacological (e.g., antidepressant drugs) and non-pharmacological treatments (e.g., cognitive behavioral therapy in depression) in psychiatry…." However, van Os and Howes caution that the data, which were limited in terms of study durations and other "real-world" data, do not allow for a complete picture of the effects of antipsychotic drugs in schizophrenia.

Assessing benefits and risks
The five-year relapse rate of schizophrenia is around 80 percent (Robinson et al., 1999), meaning that one-fifth of patients won’t experience another episode during that time. Although the effect of antipsychotics on long-term survival is presently unclear, with both increases and decreases in mortality rates reported (see SRF related news story and SRF news story), neuroleptics are accompanied by severe side effects. Adverse effects include weight gain, sedation, and tardive dyskinesia, and can produce long-term health effects, create stigma, and be detrimental to quality of life. Determining the length of time patients should take antipsychotics after illness stabilization is an important and much-debated subject. In the present meta-analysis, researchers aimed to inform the debate by quantifying the effect of antipsychotics on relapse rate.

Leucht and colleagues combed through data from 65 clinical trials over the last 50 years that included a total over 6,000 patients, with a median of 47 patients per study and a median study duration of 26 weeks. A variety of definitions of relapse were used, with clinician judgment, need of medication, and rating scale score being the most common. Indeed, in their discussion, Leucht and colleagues write that, "Universally accepted definitions of relapse and stability of symptoms are urgently needed."

Nonetheless, the results of the meta-analysis were robust—fewer study participants on antipsychotics relapsed between seven and 12 months than those on placebo, and this was true for all time points studied. Some studies tracked relapse as well as hospital readmission (indicative of severe cases). In these trials, nearly 70 percent of the placebo group, but only one-quarter of those on antipsychotics, relapsed. Readmission to a hospital was recorded for 25 percent of placebo patients but only 10 percent of patients on antipsychotics. The length of time patients were stable before entering into the trial did not affect relapse rate, although the difference between the placebo and antipsychotic groups decreased with increasing study length, suggesting that antipsychotics may lose effectiveness over time. However, as noted by the authors, this effect could be due to decreased compliance over time, and thus, longer-lasting studies that more carefully monitor adherence are needed.

Antipsychotics seem to benefit other aspects of the illness besides relapse. At the end of the trials, there were fewer patients with unimproved or worse illness symptoms in the antipsychotic group than in the placebo, and drug treatment was associated with less aggressive behavior than placebo. In the small number of studies that looked at quality of life, an improvement was also seen in the neuroleptic group. But the effects of antipsychotics weren’t all positive. Not surprisingly, patients taking drugs experienced more side effects than those on placebo, including weight gain, sedation, and movement disorders (with the exception of dyskinesia, which was present more often in the placebo group than the antipsychotic group, possibly because of "rebound dyskinesia" from abrupt withdrawal of drugs). There was no overall effect on mortality, although the relatively short duration of follow-up would have made detection of such an effect difficult.

Parsing out the population
The authors also performed subgroup analyses to assess differences among drugs, finding that depot (long-lasting injection) preparations were better than oral forms at preventing relapse. However, in their discussion, the authors write that this finding is in question, and that direct comparisons between depot and oral versions of the same medicine are required.

No difference between first-generation and second-generation antipsychotics was found, and, interestingly, there was also no difference in relapse rates between tapered and abrupt drug withdrawal. This goes against the theory of supersensitivity psychosis (Chouinard et al., 1978), which suggests that long-term antipsychotic treatment increases dopamine receptor sensitivity such that sudden drug withdrawal causes a rebound psychosis and early relapse. In their commentary, van Os and Howes weigh in on the implications for supersensitivity psychosis, calling Leucht and colleagues’ data “reassuring.”

Van Os and Howes also raise important points about the other "real-world" implications of these data. They note that studies with lengthier follow-up periods are needed to provide a thorough cost-benefit analysis of long-term antipsychotic use. Moreover, they caution that “no evidence that other, more disabling, domains of psychopathology such as cognitive alterations or motivational impairment are similarly alleviated is available.”—Allison A. Curley.

Leucht S, Tardy M, Komossa K, Heres S, Kissling W, Salanti G, Davis JM. Antipsychotic drugs versus placebo for relapse prevention in schizophrenia: a systematic review and meta-analysis. Lancet. 2012 May 2. Abstract

Van Os J, Howes OD. Antipsychotic drugs for prevention of relapse. Lancet . 2012 May 2. Abstract

Comments on News and Primary Papers
Comment by:  William Carpenter, SRF Advisor (Disclosure)
Submitted 23 May 2012
Posted 23 May 2012

Since the John Davis meta-analysis in the 1970s (the second meta-analysis in medicine; see Davis, 1975), there is little doubt that drug beats placebo in relapse prevention. What is terrific here are the details on aspects of course that are addressed and more information on the risk side to complement benefit. I will comment on two findings. First, depot beats oral. Especially in the United States, we need to move depot into an attractive option and first-line approach (assuming oral administration of the same compound supports safety in the individual). Second, I have thought that there is little evidence for second-generation antipsychotics being superior to first-generation antipsychotics (clozapine excepted) in efficacy and effectiveness, and that the strongest case for second-generation antipsychotics was better relapse prevention. This meta-analysis failed to support this supposed advantage, and I have to adjust my view in this regard.

The van Os and Howes commentary merits a serious and thoughtful read. It gives emphasis to how early and modest the field is in discovery for the psychopharmacology of schizophrenia. In the 60 years since chlorpromazine, progress has been narrow, not very innovative, and not very robust.


Davis JM. Overview: maintenance therapy in psychiatry: I. Schizophrenia. Am J Psychiatry . 1975 Dec ; 132(12):1237-45. Abstract

View all comments by William CarpenterComment by:  Wolfgang Gaebel
Submitted 21 June 2012
Posted 22 June 2012
  I recommend the Primary Papers

Antipsychotics are (generally) effective in relapse prevention—however, evidence-based indicators for more individualized treatment are needed.

Leucht and colleagues (Leucht et al., 2012) conducted a new, highly comprehensive, but also very differentiated meta-analysis on the relapse preventive efficacy of antipsychotic drugs as compared to placebo, which is now published in The Lancet. In addition, detailed comments are provided by van Os and Howes (van Os and Howes, 2012) in the same issue. As in former reviews addressing this topic, the results clearly underline that antipsychotics are (generally) effective in preventing relapse in schizophrenia compared to placebo with average one-year relapse rates of 27 percent versus 64 percent, an estimated risk ratio of 0.4, and a number of 3 needed to treat (to benefit) patients. On the other hand, and despite this high-grade evidence, many patients with schizophrenia still show cognitive, affective, motivational, or "functional" deficits and impairments in the longer illness course. This indicates that antipsychotics are not so effective in all symptom domains and contributes to the notion by some researchers that schizophrenia is still a "poor outcome" disorder, in particular as compared to other mental disorders (Jobe and Harrow, 2005). Thus, further efforts are needed to develop drug and other treatment strategies to affect the illness course more comprehensively. In this regard, a more uniform and broadly accepted definition of "relapse" would be very helpful. For example, the consensus-based definition of "remission" (Andreasen et al., 2005) has stimulated a large amount of research and enables a better comparison of results across studies.

Some other aspects are noteworthy to address. First, it is still unclear how long antipsychotic treatment has to be maintained, especially after a first episode. A comparison of evidence-based treatment guidelines from different countries developed on highest-quality criteria yielded no or inconsistent recommendations regarding the duration of maintenance treatment (Gaebel et al., 2011). In the last years, different trials compared maintenance treatment (MT) with (stepwise) drug discontinuation (mostly supplemented by early drug intervention based on early signs for relapse, i.e., targeted intermittent treatment/IT) in the long-term treatment of first-episode patients. Whether drug treatment was discontinued after six months (Wunderink et al., 2007), after one year (Gaebel et al., 2011), or, as just published, after two years (Emsley et al., 2012; observational design after drug discontinuation), risk for relapse did noticeably increase after drug discontinuation (up to fivefold after one year compared to further MT). On the other hand, a considerable proportion of patients remains stable under intermittent treatment (in the Gaebel et al., 2011, trial, about 50 percent after MT in the first year). In addition, a relevant proportion of patients (about 20 percent) remain relapse free and develop only one illness episode also if drug treatment was suspended (e.g., Harrow et al., 2012).

There are other significant findings indicating that different patients respond differently to similar treatment strategies. Although efficacy of antipsychotics in general is proven, a distinct proportion of patients suffers from partial remission, symptom recurrence, or relapse despite antipsychotic treatment. This applies also to depot treatment, which minimizes adherence problems, although some patients do relapse, as once again shown in a recent placebo-controlled trial not yet included in the Leucht review (Kane et al., 2012). Similarly, the inconclusive results regarding abrupt versus stepwise drug discontinuation and the development of a "supersensitivity" psychosis represent (to a great extent) different effects in different patients. This corresponds to a finding of our own discontinuation trial, in which assured pre-treatment with antipsychotics and an excellent treatment response have been predictors for deterioration in the case of drug discontinuation (Gaebel et al., 2011). It seems that some patients do very well with antipsychotics (often mistaken by themselves as not being in need of drugs), and that they are the ones who should be recommended to rather maintain than to discontinue treatment.

Based on the Vulnerability-Stress-Coping-model for schizophrenia, many influencing factors contribute to illness development and course. Accordingly, schizophrenia represents a heterogeneous illness due to various pathogenetic factors. Thus, different effects in different patients to similar treatment strategies are to be expected. On the other hand, different antipsychotics do also show some differences in efficacy on different symptom domains (Leucht et al., 2009), as well as in side effect and receptor profiles (Correll, 2010). Thus, besides developing new drugs and other treatment strategies, research should also focus on matching the right patient to the right treatment strategy. This requires the availability of valid (i.e., evidence-based) indicators to enable effective individual treatment decisions. This could be clinical indicators (like symptom characteristics, response, adherence, etc.) or biological markers such as genetic variants, neurophysiological, chemical, or functional parameters (among others). For example, Mössner et al. (Mössner et al., 2009) analyzed data from the first-episode study incorporating the above-mentioned discontinuation trial (Gaebel et al., 2011) regarding genetic characteristics and drug response. They found that response in the negative symptom domain is different according to a common polymorphism (-1019C/G) located in the promoter region of the serotonin (5-HT) 1A receptor gene only for the SGA risperidone (as compared to haloperidol). Other observational trials show similar effects, but well-controlled RCTs are needed to provide sound evidence. Accordingly, we are now preparing a trial in which patients with primary negative symptoms are randomly allocated to different SGAs to investigate response differences in the negative symptom domain in relation to their (-1019C/G) polymorphisms.


Leucht S, Tardy M, Komossa K, Heres S, Kissling W, Salanti G, Davis JM (2012). Antipsychotic drugs versus placebo for relapse prevention in schizophrenia: a systematic review and meta-analysis. Lancet. 379:2063-71. Abstract

van Os J, Howes OD (2012). Antipsychotic drugs for prevention of relapse. Lancet. 2012; 379):2030-1. Abstract

Jobe TH, Harrow M. Long-term outcome of patients with schizophrenia: a review. Canadian Journal of Psychiatry. 2005; 50, 892-900. Abstract

Andreasen NC, Carpenter WT, Kane JM, Lasser RA, Marder SR, Weinberger DR. Remission in schizophrenia: proposed criteria and rationale for consensus. Am J Psychiatry 2005;62:441-449. Abstract

Gaebel W, Riesbeck M, Wobrock T (2011): Schizophrenia guidelines across the world: a selective review and comparison. Int Rev Psychiatry. 23: 379-87. Abstract

Wunderink L, Nienhuis FJ, Sytema S, Slooff CJ, Knegtering R, Wiersma D. Guided discontinuation versus maintenance treatment in remitted first-episode psychosis: relapse rates and functional outcome. J Clin Psychiatry, 2007, 68:654-61. Abstract

Gaebel W, Riesbeck M, Wölwer W, Klimke A, Eickhoff M, von Wilmsdorff M, Lemke M, Heuser I, Maier W, Huff W, Schmitt A, Sauer H, Riedel M, Klingberg S, Köpcke W, Ohmann C, Möller HJ; German Study Group on First-Episode Schizophrenia (2011): Relapse prevention in first-episode schizophrenia: Maintenance vs. intermittent drug treatment with prodrome-based early intervention. Results of a randomized controlled trial within the German Research Network on Schizophrenia. J Clin Psychiatry. 72: 205-18. Abstract

Emsley R, Oosthuizen PP, Koen L, Niehaus DJ, Martinez G (2012). Symptom recurrence following intermittent treatment in first-episode schizophrenia successfully treated for 2 years: a 3-year open-label clinical study. J Clin Psychiatry. 73: e541-e547. Abstract

Harrow M, Jobe TH, Faull RN. Do all schizophrenia patients need antipsychotic treatment continuously throughout their lifetime? A 20-year longitudinal study. Psychol Med. 2012 Feb 17:1-11. Abstract

Kane JM, Sanchez R, Perry PP, Jin N, Johnson BR, Forbes RA, McQuade RD, Carson WH, Fleischhacker WW. Aripiprazole Intramuscular Depot as Maintenance Treatment in Patients With Schizophrenia: A 52-Week, Multicenter, Randomized, Double-Blind, Placebo-Controlled Study. J Clin Psychiatry 2012;73(5):617-624. Abstract

Leucht S, Corves C, Arbter D, Engel RR, Li C, Davis JM (2009). Second-generation versus first-generation antipsychotic drugs for schizophrenia: a meta-analysis. Lancet; 373: 31-41. Abstract

Correll CU (2010). From receptor pharmacology to improved outcomes: individualising the selection, dosing, and switching of antipsychotics. Eur Psychiatry. 25 Suppl 2:S12-21. Abstract

Mössner R, Schuhmacher A, Kühn KU, Cvetanovska G, Rujescu D, Zill P, Quednow BB, Rietschel M, Wölwer W, Gaebel W, Wagner M, Maier W (2009). Functional serotonin 1A receptor variant influences treatment response to atypical antipsychotics in schizophrenia. Pharmacogenet Genomics;19:91-4. Abstract

View all comments by Wolfgang Gaebel

Comments on Related News

Related News: Clozapine: The Safest Antipsychotic?

Comment by:  John McGrath, SRF Advisor
Submitted 23 July 2009
Posted 23 July 2009
  I recommend the Primary Papers

The results of this study are surprising. In those with schizophrenia, those on clozapine had by far the lowest relative risk of death (compared to patients on other antipsychotics). Compared to older medications, atypical antipsychotics, to date, do not seem to be impacting on the relative risk of death.

I congratulate the authors on this impressive study. The study is another reminder of the utility of population-based record linkage studies. Thank heavens for the Nordic countries' health registers.

A few years ago we wondered if the differential mortality rate for schizophrenia was worsening over time (Saha et al., 2007). In addition to differential access to health care, we worried that the adverse effects of atypical antipsychotics might be a “ticking time bomb” for worsening mortality in the decades to come. The new Finnish study shows a more nuanced picture emerging.

While the results are thought provoking, let’s not forget about the main game. We all agree that there is still much more work to be done in optimizing the general physical health of people with schizophrenia.


Saha S, Chant D, McGrath J. A systematic review of mortality in schizophrenia: is the differential mortality gap worsening over time? Arch Gen Psychiatry . 2007 Oct 1 ; 64(10):1123-31. Abstract

View all comments by John McGrath

Related News: Clozapine: The Safest Antipsychotic?

Comment by:  Francine Benes, SRF Advisor
Submitted 4 November 2009
Posted 4 November 2009

Clozapine: A First-Line Antipsychotic?
Tiihonen et al., of the University of Kuopio in Finland, compared mortality rates in over 66,000 patients with schizophrenia with the entire population of Finland and concluded that clozapine should be used as a first-line drug in the treatment of this disorder. Clozapine is a very effective antipsychotic, and for patients who have received it for several years, the improvement in clinical status can be quite remarkable (Lindstrom, 1988; Agid et al., 2008). Additionally, the improved mortality rate of patients on clozapine may be attributable, at least in part, to the close monitoring of their white blood cell count (WBC).

The stipulation that weekly or biweekly blood samples must be drawn is not an issue that can be viewed lightly, because approximately 1-2 percent of patients on clozapine may show significant decreases in their WBC. This may be a harbinger of agranulocytosis, a potentially lethal form of morbidity in which the bone marrow loses its ability to generate leukocytes; death remains a significant risk for patients taking this drug (Taylor et al., 2009). To some, this may seem like a small price to pay for an improved quality of life. For others, however, it represents an unacceptable degree of risk. Additionally, many patients consider the requirement for frequent blood drawing as intrusive and/or painful and refuse to have it done (personal observation).

Perhaps the greatest source of resistance to using clozapine as a “first-line” drug is the psychiatrist who is faced with this decision. In general, most believe that they would be exposing their patient to unnecessary risk and prefer to look toward other, more “benign” antipsychotic drugs (APDs) for treatment options. In practice, however, the second-generation atypical APDs are not necessarily better candidates for “first-line” use, because they may be even more likely to cause excessive weight gain, diabetes mellitus, and cardiovascular disease (Wehring et al., 2003; Henderson et al., 2005) and result in increased mortality (Meatherall and Younes, 2002). In addition to the risk of agranulocytosis, clozapine may also cause unacceptable amounts of sedation, drooling, and weight gain. Typical APDs, on the other hand, are associated with other side effects that can be quite debilitating. These include extrapyramidal movement disorders, such as 1) akathisia, a condition that may cause a worsening of symptoms as a result of agitation; 2) drug-induced Parkinsonism, in which hypokinesia usually complicates the negative symptoms of schizophrenia; and 3) tardive dyskinesia, a syndrome in which there are involuntary movements of the tongue and lips that can result in significant disability and even disfigurement (Peacock et al., 1996).

In considering the choice of an APD for a “first-episode” patient with schizophrenia, all of these factors must be considered. It is impossible to know how a particular patient with no prior history of having taken an APD will respond to any given drug. What may be an excellent “first-line” drug for one patient may not be so for another. So, the choice of a “first-line” drug requires that the doctor and patient work together to identify the APD that is most appropriate at a particular time in the course of the illness, particularly if the patient has a treatment-sensitive or treatment-resistant form of schizophrenia (Wang et al., 2004).


Agid O, Kapur S, Remington G. Emerging drugs for schizophrenia. Expert Opin Emerg Drugs. 2008;13:479-95. Abstract

Henderson DC, Nguyen DD, Copeland PM, Hayden DL, Borba CP, Louie PM, Freudenreich O, Evins AE, Cather C, Goff DC. Clozapine, diabetes mellitus, hyperlipidemia, and cardiovascular risks and mortality: results of a 10-year naturalistic study. J Clin Psychiatry. 2005;66:1116-21. Abstract

Lindstrom LH. The effect of long-term treatment with clozapine in schizophrenia: a retrospective study in 96 patients treated with clozapine for up to 13 years. Acta Psychiatr Scand. 1988;77:524-9. Abstract

Meatherall R, Younes J. Fatality from olanzapine induced hyperglycemia. J Forensic Sci. 2002;47:893-6. Abstract

Peacock L, Solgaard T, Lublin H, Gerlach J . Clozapine versus typical antipsychotics. A retro- and prospective study of extrapyramidal side effects. Psychopharmacology (Berl). 1996; 124:188-96. Abstract

Taylor DM, Douglas-Hall P, Olofinjana B, Whiskey E, Thomas A. Reasons for discontinuing clozapine: matched, case-control comparison with risperidone long-acting injection. Br J Psychiatry. 2009;194:165-7. Abstract

Wang PS, Ganz DA, Benner JS, Glynn RJ, Avorn J. Should clozapine continue to be restricted to third-line status for schizophrenia?: a decision-analytic model. J Ment Health Policy Econ. 2004;7:77-85. Abstract

Wehring HJ, Kelly DL, Love RC, Conley RR. Deaths from diabetic ketoacidosis after long-term clozapine treatment. Am J Psychiatry. 2003;160:2241-2. Abstract

View all comments by Francine Benes

Related News: Clozapine: The Safest Antipsychotic?

Comment by:  Edward Orton (Disclosure)
Submitted 18 November 2009
Posted 18 November 2009
  I recommend the Primary Papers

Dr. Benes notes that clozapine is "...a very effective antipsychotic, and...improvement in clinical status can be quite remarkable." The mortality figures reported by Tihonen et al. have proved quite striking to schizophrenia researchers. The perception within the psychiatry community that clozapine is too risky for first-line therapy needs further assessment and discussion. Only about 5 percent of schizophrenics in the U.S. receive clozapine (Lieberman, 2009), leaving the vast majority of patients undermedicated because of this perception. The major issue with starting a patient on clozapine is WBC monitoring. I would like to call upon the NIMH to establish a major study in which schizophrenics are introduced to clozapine on an inpatient basis for 30-60 days to establish safety. It is well known that most WBC events associated with clozapine occur in the first few weeks of treatment. Also, I note that current prescribing practice with clozapine actually allows for monthly blood monitoring after 12 months of continuous clozapine use. Thus, the burden of monitoring diminishes sharply after one year.


Lieberman J. A Beacon of Hope: Prospects for Preventing and Recovering from Mental Illness. NARSAD Research Quarterly 2 (1), Winter 2009.

View all comments by Edward Orton