(T-087) Population pharmacokinetics of cefepime in critically ill children and young adults

Tomoyuki Mizuno, PhD – Associate Professor, Cincinnati Children's Hospital; Kelli Paice, MD – Clinical Fellow, Cincinnati Children's Hospital; Kathryn Pavia, MD – Clinical Fellow, Cincinnati Children's Hospital; Peter Tang, PhD – Professor, University of Cincinnati; Abigayle Abigayle Gibson, BSc – Clinical Research Coordinator, Cincinnati Children's Hospital; Calise Curry, BA – Clinical Research Coordinator, Cincinnati Children's Hospital; Jennifer Kaplan, MD, PhD – Professor, Cincinnati Children's Hospital; Sonya Girdwood, MD, PhD – Professor, Cincinnati Children's Hospital

Objectives: Cefepime is a fourth-generation cephalosporin antibiotic with broad-spectrum activity widely prescribed for severe and nosocomial infections. Achieving optimal antibiotic exposure with cefepime in critically ill children and young adults is challenging due to significant pharmacokinetic (PK) variability. Suboptimal dosing may lead to treatment failure or toxicity. We aim to develop a parametric population PK model of cefepime for critically ill pediatric and young adult patients, identify factors influencing drug disposition, and conduct Monte Carlo simulations to determine the optimal initial dosing regimens.

Methods: Pediatric intensive care unit patients receiving cefepime 30-50 mg/kg Q8h as a 30-minute infusion were prospectively followed for clinical data collection and scavenged opportunistic plasma sampling.[1] Cefepime concentrations were measured using high-performance liquid chromatography. Nonlinear mixed effects modeling was conducted using NONMEM v7.5 interfaced with Pirana. Allometric body weight scaling was included a priori with fixed exponents to account for body size differences. Monte Carlo simulations were performed using the final model to determine the initial dosing regimen that maximizes the probability of maintaining free concentration above the minimum inhibitory concentration (MIC) of Pseudomonas aeruginosa (8 mg/L breakpoint) for the entire dosing interval (100% fT>MIC).

Results: Data from 100 patients, aged 1 month to 30 years, with 510 samples, were described using a two-compartment model with first-order elimination. Inter-individual variability was estimated for clearance (CL) and central volume of distribution (V1), but could not be estimated for intercompartmental clearance (Q) or peripheral volume (V2). Estimated glomerular filtration rate (eGFR) and cumulative percentage of fluid balance were identified as covariates on CL and V1, respectively. The final parameters were estimated with good precision: CL 6.4 L/h/70 kg0.75, Q 3.7 L/h/70 kg0.75, V1 15 L/70 kg, and V2 8.9 L/70 kg. The goodness-of-fit plots showed no model misspecification. The Monte Carlo simulations suggested that the usual dose of 50 mg/kg requires a longer duration of infusion (3 hours) or a shorter dosing interval (every 6 hours) to ensure target attainment. The dose regimen of 50 mg/kg Q8h as a 30-minute infusion only had high probability of target attainment (>90%) in patients with renal impairment.

Conclusions: A cefepime PK model for critically ill pediatric patients was successfully developed accounting for patients' renal function, fluid status, and body size, using real world data and an opportunistic sampling approach. The model was internally validated with good precision and stability and may be used for model-informed precision dosing. The initial cefepime dosing regimen may require modifications such as a longer duration of infusion or a shorter dosing interval to ensure target attainment against Pseudomonas aeruginosa.

Citations: [1] Tang Girdwood SC. et al., J Clin Pharmacol, 61(4), 565, 2020.