Quantifying intraindividual variations in plasma clozapine levels: a population pharmacokinetic approach

Clozapine has strong recommendations for therapeutic drug monitoring. While factors that influence interindividual variation in plasma clozapine levels have been extensively reported, intraindividual variation remains poorly studied. We employed a population pharmacokinetic approach to assess intraindividual variations in plasma levels of both clozapine and N-desmethylclozapine, as well as the impact of smoking on this variability.

Patients who were initiated on clozapine from January 2009 to December 2010 and who provided at least 2 plasma samples were included in this study. The observed concentrations of clozapine and N-desmethylclozapine were applied in a Bayesian pharmacokinetic modeling approach by using a previously published pharmacokinetic model from an independent sample to compute a predicted concentration. The predicted concentrations of clozapine and N-desmethylclozapine were then compared with the observed concentrations in the form of a ratio: predicted-to-observed concentration ratio (Cpred/Cobs). The coefficient of variation of the Cpred/Cobs ratios was taken as a measure of intraindividual variation.

A total of 723 plasma levels from 61 patients were included in this analysis. The coefficient of variation of Cpred/Cobs ratios for clozapine and N-desmethylclozapine were 29.8% (SD = 17.2%) and 27.4% (SD = 16.4%), respectively. Though values were higher, smoking did not have a significant effect on coefficients of variation of clozapine (33.5% vs 26.3%, P = .184) or N-desmethylclozapine (30.7% vs 24.2%, P = .100).

Clinicians need to be aware of intraindividual variability and not assume that plasma levels are static. If plasma levels are used to guide dosing of clozapine, serial measurements rather than a single level might be necessary to make an informed clinical decision. The clinical implications of intraindividual variability in plasma clozapine levels need further study.