MarÃa Patricia Hernández-Mitre (1), Melissa Ankravs (2,3), Hayoung Won (1), Jason A. Roberts (1,4,5,6), Steven C. Wallis (1), Lucy Sharrock (2), Nevin Walpola (2), Kathleen Byrne (2), Briannah Miles (2), Andrew Udy (7,8), Rinaldo Bellomo (2,3,8,9) and Adam Deane (2,3)
(1) The University of Queensland Centre for Clinical Research; (2) The Royal Melbourne Hospital; (3) Melbourne Medical School, The University of Melbourne; (4) Herston Infectious Diseases Institute; (5) Royal Brisbane & Women’s Hospital; (6) Nîmes University Hospital, University of Montpellier; (7) The Alfred Hospital; (8) Monash University; (9) The Austin Hospital
Introduction: The optimal drug and dosage for treating acute delirium lack consensus, especially among critically ill patients whose altered pharmacokinetics pose challenges. Unlike antimicrobials[1], data on antipsychotic drug use in this patient population remains scarce.
Objective: To characterise the population pharmacokinetic (PK) parameters of olanzapine and quetiapine in critically ill patients with delirium.
Methods: This single-centre prospective observational study enrolled adult patients (≥18 years) receiving immediate-release enteral olanzapine or quetiapine for acute delirium. Blood samples were collected via existing catheters at specific time points. For olanzapine, samples were collected at 1.5 h, 6 h, and immediately before the next dose, or at 24 h post-dose. For quetiapine, samples were collected at 1.5 h, 5 h, and immediately before the next dose, or at 12 h post-dose. For patients continuing the medication, additional trough samples were drawn at 24, 48 and 72 h after the initial trough event. Total plasma concentrations of these drugs were measured using ultra-high performance liquid chromatography-tandem mass spectrometry. Population PK analysis was conducted using the nonlinear mixed-effects model program NONMEM® 7.5.1 with the first-order conditional estimation with interaction (FOCE+I) method. Covariates screened included age, weight, albumin, bilirubin, international normalised ratio, and sex. For olanzapine, smoking status and concurrent administration of fluvoxamine, sertraline, or valproate were evaluated. For quetiapine, concurrent administration of erythromycin was assessed. Model evaluation was based on goodness-of-fit (GOF) plots and prediction corrected visual predictive check (pcVPC). Final models were internally validated through bootstrap analysis (n=1000).
Results: Twenty-seven patients contributed a total of 96 samples: 11 patients provided 36 samples for the olanzapine model, while 16 patients contributed 60 samples for the quetiapine model. A one-compartment model better described the data for olanzapine [estimate (relative standard error, RSE): apparent clearance (CL/F) = 19.7 L/h (28%); apparent volume of distribution (V/F) = 1080 L (18%); absorption rate constant (KA) = 0.49 h-1 FIX]. For quetiapine, a two-compartment model provided a better fit [estimate (RSE): CL/F = 24.8 L/h (16%); apparent central volume of distribution (Vc/F) = 34.7 L (25%); apparent intercompartmental clearance (Q/F) = 18.4 L/h (22%); apparent peripheral volume of distribution (Vp/F) = 69.9 L (23%); KA = 0.345 h-1 (16%)]. Both models incorporated first-order absorption and linear elimination, with exponential interindividual variability (IIV) [coefficient of variation (RSE) [shrinkage]: IIV_CL = 70.8% (27%) [18%]; IIV_V = 39.6% (21%) [20%] for olanzapine; IIV_CL = 67.4% (14%) [2%] for quetiapine], as well as proportional residual variability [%CV (RSE) [shrinkage]: 21% (0%) [20%] for olanzapine; 25.4% (0%) [12%] for quetiapine]. No covariates were retained in either final model. The GOF and pcVPC plots indicated acceptable fitting. The mean estimated PK parameters closely aligned with the median of the bootstrapped parameters, and fell within their respective 95% confidence intervals, demonstrating the stability and robustness of both models.
Conclusions: This study provides insights into the olanzapine and quetiapine population PK parameters in critically ill adults with delirium. High variability in drug clearance was observed, but limitations in the available data prevent the confident proposal of optimised dosing regimens at this time.
Grant acknowledgement: Society of Hospital Pharmacists of Australia National Translational Research Collaborative (Juno Pharmaceuticals Emerging Researcher Grant).
References:
[1] Roberts JA and Lipman J. Pharmacokinetic issues for antibiotics in the critically ill patient. Crit Care Med 2009; 37(3): 840-851.
Reference: PAGE 32 (2024) Abstr 11181 [www.page-meeting.org/?abstract=11181]
Poster: Drug/Disease Modelling - CNS