An exploratory pharmacokinetic-pharmacodynamic analysis of ceftazidime-avibactam and various blood biomarkers in patients with hospital-acquired pneumonia or ventilator-associated pneumonia caused by Klebsiella pneumoniae
International Journal of Antimicrobial Agents, 2026
O’Jeanson A., Nielsen E., Athanassa Z., Ginosyan A., Ioannidis K., Manioudaki S., Petsa I., Giamarellou H., Bader I., Skarmoutsou N., Mylona E., Sakagianni A., Loryan I., Karaiskos I., Friberg L.
| Disease area | Application area | Sample type | Products |
|---|---|---|---|
Infectious Diseases | Pathophysiology | Plasma |  Olink Target 96 |
Abstract
Purpose
To characterise the pharmacokinetics (PK) of ceftazidime-avibactam (CAZ-AVI) and explore the dynamics of a broad blood immune biomarker panel in critically ill patients with hospital-acquired (HAP) or ventilator-associated pneumonia (VAP) caused by Klebsiella pneumoniae.
Methods
Population PK (PopPK) models were developed for CAZ and AVI individually and jointly, evaluating covariate effects and correlations at interindividual (IIV) and residual unexplained variability (RUV) levels. Biomarker turnover dynamics in blood were modelled, and exposure-biomarker relationships were assessed.
Results
The dataset included 266 drug plasma concentration measurements and 48-87 observations per biomarker, collected from ten critically ill patients. CAZ and AVI PK were well-described by two-compartment models, with estimated creatinine clearance (eCrCL) as a key covariate on clearance. Joint modelling revealed significant correlations between the two drugs at both IIV and RUV levels, supporting coherent patient-specific simulations. Turnover dynamics were characterised for 32 biomarkers. The most pronounced biomarker changes over the course of treatment were decreases in IL6, CRP, and AREG, and increases in DCBLD2, LAMP3, and TRIM21. A significant exposure-response relationship was identified between CAZ plasma concentrations and CRP turnover, with higher CAZ levels associated with inhibition of CRP production.
Conclusions
The joint PopPK model confirmed eCrCL as a key covariate for clearance. Changes in 32 immune response biomarkers were quantified, with CRP showing a clear exposure-response relationship. These findings provide a foundation for prioritising biomarkers for future studies to guide therapeutic strategies in critically ill patients with HAP or VAP.