Therapeutic drug monitoring of clozapine and relapse - a retrospective study of routine clinical data

Objective: Therapeutic drug monitoring (TDM) of the atypical antipsychotic drug clozapine is recommended. Clinical studies have indicated a therapeutic window for clozapine serum levels in schizophrenic and schizo-affective patients during acute treatment, i.e. for patients who do not respond to treatment with typical antipsychotics. However, despite the frequent use of clozapine also in maintenance treatment, very few data are available showing the relationship between serum levels of clozapine and the prevention of relapse. Thus, the primary objective of the study was to investigate the relationship between serum levels of clozapine and relapse during maintenance treatment. Methods: A retrospective study of routine TDM-data was conducted. Samples obtained on an acute treatment ward from patients with < 4 days hospitalization (recent admissions) were regarded as samples associated with relapse. Samples which can be attributed to an intoxication were identified as described in the TDM-form. The serum level of clozapine, as well as age, gender, smoking habits, concurrent drugs, psychiatric diagnosis and dose of clozapine were evaluated. Data analysis was performed on individual samples and, alternatively, on multiple samples from a single patient which are summarized according to a typical clinical situation. Results: 404 serum levels were measured in 86 patients. After exclusion of patients receiving acute treatment, 65 relevant clinical situations were identified in 50 patients: 12 relapses, 8 intoxications (a total of 20 situations with poor outcome) and 45 situations involving patients with good maintenance outcome. Samples involving relapse had serum levels of 198 +/- 211 ng/ml (10 - 624), intoxications had serum levels of 1,969 +/- 705 ng/ml (900 - 2,900) and those with good outcome had serum levels of 384 +/- 255 ng/ml (56 - 1,028) (mean +/- SD (range)). By means of sensitivity of receiver operating characteristic curves (ROC) a lower limit of the therapeutic window can be estimated at about 50 - 250 ng/ml and an upper limit at about 745 - 1,050 ng/ml. The frequencies of good and poor outcome were significantly different within and outside these ranges, e.g. chi(2) = 11.8 and p < 0.001 for 250 to 745 ng/ml. Comparison of only good outcome and relapse provided a significant difference in the serum level of clozapine (Student's t-test p = 0.024). However, 67% of relapses were predicted in a model of logistic regression only if the variables serum level and concurrent treatment with other psychotropic drugs were included simultaneously as independent variables. Neither variable was able to predict relapse if used as a single variable in separate models. Finally, it was found that serum levels of clozapine were increased in women, in aged patients and in nonsmokers. Conclusions: It is tentatively concluded that serum levels of clozapine < 50 ng/ml are related to relapse irrespective of concurrent psychotropic drugs. In cases where there are no concurrent psychotropic drugs, serum levels of clozapine < 250 ng/ml are associated with relapse. The risk of relapse is low for serum levels of clozapine > 250 ng/ml irrespective of concurrent psychotropic drugs. The risk of intoxication is increased with serum levels > 750 ng/ml. The TDM of clozapine is recommended during maintenance treatment.