Influence of dose, cigarette smoking, age, sex, and metabolic activity on plasma clozapine concentrations: a predictive model and nomograms to aid clozapine dose adjustment and to assess compliance in individual patients.

Abstract

The measurement of plasma clozapine concentrations is useful in assessing compliance, optimizing therapy, and minimizing toxicity. We measured plasma clozapine and norclozapine (N-desmethylclozapine) concentrations in samples from 3782 patients (2648 male, 1127 female). No clozapine was detected in 291 samples (227 patients, median prescribed dose 300 mg/d). In 4963 (50.2 %) samples (2222 patients); plasma clozapine concentration ranged from 10 to 350 ng/mL.Step-wise backward multiple regression analysis (37 % of the total samples) of log10 plasma clozapine concentration against log10 clozapine dose (mg/d), age (year), sex (male = 0, female = 1), cigarette smoking habit (nonsmokers = 0; smokers = 1), body weight (kg), and plasma clozapine/norclozapine ratio (clozapine metabolic ratio, MR) showed that these covariates explained 48% of the observed variation in plasma clozapine concentration (C = ng/mL x 10-3) (P < 0.001) according to the following equation: log 10 (C) = 0.811 log 10 (dose) + 0.332 (MR) + 69.42 X 10 (-3) (sex) + 2.263 x 10 (-3) (age) + 1.976 x 10(-3) (weight) - 0.171 (smoking habit) - 3.180. This model and its associated confidence intervals were used to develop nomograms of plasma clozapine concentration versus dose for male and female smokers and nonsmokers. Predicted plasma clozapine changes by +48% in nonsmokers, +17% in females, +/-8 % for every 0.1 change in MR (reference 1.32), +/-4% for every 5 years (reference 40 years), and +/-5 % for every 10 kg body weight (reference 80 kg). The nomograms can be used (i) to individualize dosage to achieve a given target plasma clozapine concentration, and (ii) for quantitative evaluation of adherence by estimating the likelihood of an observed concentration being achieved by a given dosage regimen. The model has been validated against published data.