BackgroundThis therapeutic drug monitoring (TDM) study aimed to determine the role of olanzapine (OLZ) and N-desmethyl-OLZ (DMO) levels in the therapeutic efficacy of OLZ in patients with schizophrenia.MethodPlasma concentrations of OLZ (COLZ) and DMO (CDMO) in schizophrenic patients 12 hours post-dose were assessed. The correlations of COLZ and CDMO with the various scores of the Positive and Negative Syndrome Scale (PANSS) were evaluated. A receiver operating characteristic curve (ROC) was utilized to identify the threshold COLZ and COLZ/CDMO ratio for maintenance of satisfactory efficacy.ResultsA total of 151 samples from patients with schizophrenia were analyzed for individual COLZ and CDMO levels. The mean COLZ and CDMO levels were 37.0 ± 25.6 and 6.9 ± 4.7 ng/mL, respectively, and COLZ was ~50% higher in female or nonsmokers (p<0.01). In all patients, the daily dose of OLZ was positively correlated with COLZ and CDMO. Linear relationships between COLZ and OLZ dose were observed in both nonsmokers and smokers (rs = 0.306, 0.426, p<0.01), although CDMO was only correlated with OLZ dose in smokers (rs = 0.485, p<0.01) and not nonsmokers. In all patients, COLZ was marginally negatively correlated with the total PANSS score. The total PANSS score was significantly negatively correlated with the COLZ/CDMO ratio (p<0.005), except in smokers. The ROC analysis identified a COLZ/CDMO ratio ≥2.99 or COLZ ≥22.77 ng/mL as a predictor of maintenance of an at least mildly ill status (PANSS score ≤58) of schizophrenia in all patients.ConclusionsA significantly negative correlation between the steady-state COLZ/CDMO ratio and total PANSS score was observed in Taiwanese schizophrenic patients. TDM of both OLZ and DMO levels could assist clinical practice when individualizing OLZ dosage adjustments for patients with schizophrenia.