Discontinuation trials with antipsychotics

The risk of relapse in discontinuation trials depends on many non-pharmacological, often poorly controllable factors, notably the expectations of the patients, doctors and nurses, other environmental factors, the duration of hospitalization and prior treatment, and the time interval since the last acute psychotic episode. On the basis of an analysis of 14 discontinuation trials, Kane and Lieberman (1987) found that the relapse rate varied greatly from study to study depending on the trial, relapse rates of 30 86% with clustering around 60 70% have been reported in the first 12 months after placebo substitution. According to Kane and Lieberman, this scatter is a result of the different inclusion criteria applied and the different definitions of relapse . 

They therefore proposed that a quantifiable increase in symptoms be used as the dependent variable instead of the relapse rate in future studies of this type. 

More important in practice than percentages would be an answer to the question of whether the relapse or exacerbation tendencies of individual patients can be estimated on the basis of specific features even before the planned withdrawal of antipsychotics. Unfortunately, the answer is largely negative. According to present knowledge there are no reliable clinical features that speak for or against the maintenance of antipsychotic therapy in individual cases (Moller, 1992). 

---

In an open trial of antipsychotics for outpatients with BPD, Teicher and colleagues ( 243) noted that sustained melancholic depression developed in three patients, necessitating removal from the trial and treatment with antidepressants. In this context, Gardner and Cowdry (248) also reported that melancholia developed in three patients on CBZ, necessitating drug discontinuance and treatment for depression. Flence, clinicians should anticipate that severe depression can develop in some patients with BPD and be prepared to treat with antidepressants. 

Wessels AM, Pollock BG, Anyama NG, Schneider LS, Lieberman JA, Marder SR, Bies RR. Association of 9-hydroxyrisperidone concentrations with risk of switching or discontinuation in the clinical antipsychotic trial of intervention effectiveness-Alzheimer s disease trial. J Clin Psychopharmacol 2010 30 683-7. 

Changes in the 10-year risk of coronary heart disease have been compared between treatment groups in 1125 patients followed for 18 months or until treatment discontinuation in the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) in schizophrenia [22 ]. The mean change in 10-year coronary heart disease risk differed significantly between treatments. Olanzapine was associated with a 0.5% increase and quetiapine with a 0.3% increase, while the risk fell in patients who used perphenazine (-0.5%), risperidone (—0.6%), and ziprasidone (—0.6%). The difference in 10-year coronary heart disease risk between olanzapine and risperidone was statistically significant. 

---