Background: Antipsychotics are mainly used in the treatment of schizophrenia and other mental disorders, such as bipolar disorder or disruptive behaviour disorders (main indication for antipsychotics in children2). Discontinuation and switching the type of substance may lead to relevant symptoms interfering with the safety and adherence of psychiatric treatment. Especially rapid discontinuation of antipsychotics which functionally inhibit or stimulate receptors may lead to psychiatric and other somatic symptoms1,3.
Objective: Our goal is to systematically assess the full range of discontinuation symptoms.
Study Design: We plan to investigate the relationship between adverse events (AEs) and discontinuation of an antipsychotic by performing meta-analyses of individual participant data in the placebo groups of RCTs following patients with versus without previous medication.
Participants: Schizophrenia, bipolar disorder, schizoaffective disorder and children with disruptive behaviour disorders.
Main outcome measure: Our main outcome measure will be total AEs and recurrence of major symptoms (psychotic or manic symptoms or disruptive behaviour) in two placebo subgroups. The target group consists of patients who discontinued an antipsychotic just before receiving the placebo and the control group consists of patients who had not recently been taking antipsychotics before receiving the placebo.
Statistical analysis: The relationship between T0 and T1 scores for the two placebo subgroups will be examined with a mixed model of repeated measures and Kaplan-Meier estimator.
Antipsychotic drugs are a heterogeneous group of compounds with a wide range of receptor affinities and diverse functional effects4. These substances may cause a variety of side effects in patients5. Therefore, providing the appropriate antipsychotic substance is a complex process1. The process frequently includes discontinuation and switching of compounds and may be accompanied or even initiated by AEs comprising cholinergic, dopaminergic, serotonergic, histaminergic and adrenergic rebound phenomena3. During switching, AEs may be caused by the current drug but could also be related to the cessation of a prior drug. Differentiating the cause for the AEs requires knowledge of the discontinuation symptoms caused by the specific compound. Additionally, there is a large number of patients who show poor adherence of antipsychotic substances especially during stable phases of illness or at the beginning of relapse (e.g. 43% of schizophrenic patients had at least one year of poor adherence over four years)6. Therapeutic strategies and treatment adherence could be significantly optimized if clinicians and patients were well
informed about potential discontinuation symptoms. This study could have a major impact on health of patients as systematic analyses of discontinuation symptoms in antipsychotics could help to identify discontinuation symptoms and may help to promote the development of innovative therapeutic strategies and guidelines in this field. This would have very practical implications for the individual patient as rapid discontinuation of an antipsychotic without professional supervision is very frequent in clinical routine7. The importance of this study is highlighted by the lack of systematic assessment of discontinuation symptoms in RCTs after rapid and complete discontinuation of antipsychotic treatment1. This study will be a first step to implement further research into which factors are predictive for occurrence of discontinuation symptoms in an individual and in long term develop treatment strategies for discontinuation syndromes.
a. Evaluate whether discontinuation symptoms occur after rapid
discontinuation of the prestudy antipsychotic.
b. Evaluate whether discontinuation symptoms are linked to the
type of discontinued antipsychotic (e.g. olanzapine, amisulpride,
c. Evaluate whether discontinuation symptoms can be predicted by
specific receptor affinities (Ki values) of the discontinued
d. Evaluate whether discontinuation symptoms can be differentiated
from early recurrence of major symptoms (i.e. are certain AEs especially predictive before recurrence of psychotic or manic symptoms or disruptive behaviour)?)
a. Total AE rate and recurrence of major symptoms (psychotic or
manic symptoms or disruptive behaviour).
b. Association between type of discontinued antipsychotic and AE
rates and recurrence of major symptoms respectively.
c. Interdependence network8 between receptor affinities and AE
rates and recurrence of major symptoms respectively.
d. Most predictive AEs before recurrence of major symptoms.
We will merge data from the placebo groups in RCTs on antipsychotic treatment of patients with schizophrenia, bipolar disorder and schizoaffective disorder. Children with disruptive behaviour disorders will be included in the analysis as a separate group to investigate discontinuation symptoms in children9.
For oral antipsychotics, the placebo group will be divided into two subgroups:
A. Target group: Patients who have just stopped taking the antipsychotic no longer than 3 days before entering the placebo group will be compared
B. Control group: All patients who have not been taking medication for
more than 1 month before entering the placebo group.
For long-acting injectable antipsychotics, the placebo group will also be divided into two subgroups:
C. Target group: Patients who should have had their last scheduled injection no longer than 1 week before entering the placebo group will be compared with D.
D. Control group: Patients who have not been receiving long-acting injectables in the last 3 month (and no oral antipsychotic for more than 1 month) before entering the placebo group.
Primary target: total AE rate during the first 12 weeks.
For oral application, the main outcome is change in the total AE rate and recurrence of major symptoms (psychotic or manic symptoms or disruptive behaviour) from baseline (T0) to four weeks (T1). All assessment time points in this timeframe will be included (i.e. the primary target is total AE rate and recurrence of major symptoms during the first 4 weeks).
For long-acting injectables, the main outcome is change in the total AE rate and recurrence of major symptoms from baseline (T0) to twelve weeks (T1). All assessment time points in this timeframe will be included (i.e. the primary target is total AE rate and recurrence of major symptoms during the first twelve weeks).
Oral and long-acting injectable antipsychotics will be calculated separately. All detected AEs will be included in the analyses. General AEs (e.g. vegetative dysregulation) and recurrence of psychotic or manic symptoms or disruptive behaviour will be calculated separately.
The main predictor will allow us to investigate the relationship between antipsychotic treatment (substance taken before entering the placebo group) and discontinuation symptoms. The main predictor is the rapid discontinuation of an oral or long-acting injectable antipsychotic.
Rapid discontinuation of the oral application will be defined as discontinuation less than three days before entering the placebo group10.
Rapid discontinuation of a long-acting injectable application will be defined as the next scheduled injection less than one week before entering the placebo group (e.g. 28 days after a four week depot antipsychotic)11.
These two groups will be compared to patients in the placebo group without rapid discontinuation of antipsychotics as described in the previous section “Data Source and Inclusion Criteria to be used to define the patient sample for your study”.
We will include additional variables/characteristics associated with occurrence of AEs and examine them for their possible confounding effect including age, sex, weight, duration and dose of antipsychotic application, previous medication, measures of psychopathology (PANSS/YMRS/Conduct Problem Subscale/etc.), duration of illness, duration of untreated psychosis, number of hospitalizations, etc. The Ki values were previously summarized (summary by Correll, p. 15, table 2)3 and will be implemented according to this study. Receptor types and corresponding rebound syndromes were also defined (definition by Correll, p. 18, table 3)3 and these definitions will be used in our study. The number needed to harm (NNH) and network analysis8 of receptor affinities will be calculated separately for both types of application (oral and long-acting injectable).
A mixed model of repeated measures (MMRM) and Kaplan-Meier estimator will be used to investigate the relationship between rapid discontinuation of an antipsychotic and the total AE rate and recurrence of major symptoms (psychotic or manic symptoms or disruptive behaviour) in an individual participant data meta-analysis. Baseline score (T0) of AE rates will be determined at the time when the participant is included in the study and the post-baseline score (T1) is determined at the last time point of the included timeline. Recurrence of psychotic or manic symptoms or disruptive behaviour will be determined between T0 and T1 (measured as change in score PANSS/YMRS/Conduct Problem Subscale/etc.) The within-subject factor is “time” and the between-subjects factor is “rapid discontinuation of an antipsychotic” (Yes/No) and the model will be tested adjusted and unadjusted for confounders (e.g. age, sex, duration of application, etc.).
The relationship between type of antipsychotic and the AE rates and recurrence of major symptoms will be assessed with multinominal logistic regression. The relationship between Ki values and the AE rate of the corresponding rebound syndrome and recurrence of major symptoms will be investigated with network analysis8 and ordinal logistic regression. Ki values will be treated as independent variables and the AE rate of the corresponding rebound syndrome and recurrence of major symptoms as dependent variable. The potential predictive value of an AE for a consecutive psychotic relapse will be investigated with multinominal logistic regression.
Missing data will be treated as recommended by Little et al.12 We will register if reasons for missing data were documented and develop a primary set of assumptions about the cause for missing data12. The primary set of assumptions will be followed by a matching statistically valid analysis (e.g. estimating-equation methods) and robustness tested with a sensitivity analysis12.
Avoiding the recurrence of major symptoms and rebound phenomena after discontinuation or switching of antipsychotics is a key factor when planning a safe and successful therapy. Rebound phenomena and recurrence of major symptoms like psychotic or manic symptoms or disruptive behaviour are among the known risks when discontinuing antipsychotics but the systematic evaluation have been scarcely studied1.
We intend to assess the complete spectrum of discontinuation symptoms in patients with schizophrenia, schizoaffective disorder, bipolar disorder and children with disruptive behaviour disorders treated with antipsychotics in the placebo group of randomized controlled trials.
Immediately after the data is available the project will start and the study plan will be published online (8/2017). The analysis will be completed six months later (2/2018). The manuscript will be drafted and submitted after four months (06/2018). The publication is planned for 08/2018. The YODA project will be informed about the completion of each milestone and reports will be made available.
To benefit both health professionals and patients we will present the study at internationally accredited conferences (e.g. symposia at the WPA) and make the study available in major medical journals (e.g. JAMA Psychiatry, American Journal of Psychiatry, Lancet Psychiatry). Based on our results we will develop and validate a questionnaire to assess the risk of discontinuation symptoms. Patients will be directly affected as national and international treatment guidelines will be influenced.
1. Cerovecki, A. et al. Withdrawal Symptoms and Rebound Syndromes Associated with Switching and Discontinuing Atypical Antipsychotics: Theoretical Background and Practical Recommendations. CNS Drugs 27, 545–572 (2013).
2. Penfold, R. B. et al. Use of Antipsychotic Medications in Pediatric Populations: What do the Data Say? Curr Psychiatry Rep 15, 13–16 (2013).
3. Correll, C. U. From receptor pharmacology to improved outcomes:
individualising the selection, dosing, and switching of antipsychotics.
European Psychiatry 25, S12–S21 (2010).
4. Leucht, S. et al. Comparative efficacy and tolerability of 15 antipsychotic
drugs in schizophrenia: a multiple-treatments meta-analysis. Lancet 382,
5. Hasan, A. et al. World Federation of Societies of Biological Psychiatry
(WFSBP) Guidelines for Biological Treatment of Schizophrenia, Part 2: Update 2012 on the long-term treatment of schizophrenia and management of antipsychotic-induced side effects. The World Journal of Biological Psychiatry 14, 2–44 (2013).
6. Valenstein, M. et al. Antipsychotic adherence over time among patients receiving treatment for schizophrenia: A retrospective review. J Clin Psychiatry 67, 1542–1550 (2006).
7. Fava, G. A., et al. Withdrawal Symptoms after Selective Serotonin Reuptake Inhibitor Discontinuation: A Systematic Review. Psychother Psychosom 84, 72–81 (2015).
8. Barabási, A. L., Gulbahce, N. & Loscalzo, J. Network medicine: a network- based approach to human disease. Nature Reviews Genetics (2011). doi:10.1038/nrg2918
9. Lu, H. & Rosenbaum, S. Developmental pharmacokinetics in pediatric populations. J Pediatr Pharmacol Ther 19, 262–276 (2014).
10. Kane, J. et al. Treatment of schizophrenia with paliperidone extended- release tablets: A 6-week placebo-controlled trial. Schizophrenia Research 90, 147–161 (2006).
11. Hough, D. et al. Paliperidone palmitate maintenance treatment in delaying the time-to-relapse in patients with schizophrenia: A randomized, double- blind, placebo-controlled study. Schizophrenia Research 116, 107–117 (2010).
12. Little, R. J. et al. The Prevention and Treatment of Missing Data in Clinical Trials. N Engl J Med 367, 1355–1360 (2012).