Pharmacokinetics of Ceftazidime in Children and Adolescents with Obesity

Abstract

Purpose: The aim of this study was to evaluate ceftazidime pharmacokinetics (PK) in a cohort that includes a predominate number of children and adolescents with obesity and assess the efficacy of competing dosing strategies.

Methods: A population PK model was developed using opportunistically collected plasma samples. For each dosing strategy, model-based probability of target attainment (PTA) estimates were computed for study participants using empirical Bayes estimates. In addition, the effects of body size and renal function on PTA were evaluated using stochastic model simulations with virtually generated subjects.

Results: Twenty-nine participants, 24 of whom were obese, contributed data towards the analysis. The median (range) age, body weight, and body mass index of participants were 12.2 years (2.3-20.6), 59.2 kg (8.4-121), and 25.2 kg/m² (13.8-42.9), respectively. Administration of 50 mg/kg intravenously (IV) every 8 hours (q8h; max 6 g/day) or 40 mg/kg IV q6h (max 6 g/day) resulted in PTA values of ≥ 90% (minimum inhibitory concentration 8 mg/L) for the subset of obese participants with estimated glomerular filtration rates (GFR) ≥ ~ 80 mL/min/1.73 m². However, for both regimens, stochastic model simulations denoted lower PTA values (< 90%) with increasing body weight for virtual subjects with GFR ≥ 120 mL/min/1.73 m². Alternatively, permitting for a maximum daily dose of 8 g/day using a 40 mg/kg IV q6h regimen provided PTA values that were near or above target (90%) for virtual subjects between 10 to 120 kg with GFR ≥ 80 mL/min/1.73 m².

Conclusion: Our analysis suggests administration of 40 mg/kg IV q6h (max 8 g/day) maximizes PTA in children and adolescents with obesity and GFR ≥ 80 mL/min/1.73 m².

Trial registration: Clinicaltrials.gov Identifier: NCT01431326.

Figure 1.

Ceftazidime concentration vs. time plots. Red circles depict concentrations that were originally reported to be collected during contaminant dose-infusions. Concentration-time values based on their originally reported sample collection times are display in subplot (a); whereas, subplot (b) depicts the timing of collected samples relative to the last administered dose for the modified dataset that was used for model development. Numerical participant ID numbers are annotated for reference. Filled circles depict participants for whom only a single concentration was collected; whereas, open-triangles represent participants who contributed 2 concentrations towards the analysis.

Figure 2.

Bayesian posthoc clearance estimates (normalized to 70 kg) from the base model versus glomerular filtration rate (GFR, Equation 2). Typical clearance estimates from the conventional GFR power model (GFR power; red solid line) and the saturation GFR model (GFR saturation; green solid line) are denoted. A locally estimated scatterplot smoothing (loess) curve was fit to the data and displayed for reference (blue dotted line)

Figure 3.

Goodness-of-fit plots for the final PopPK model. Observed vs. population predicted concentrations (a); observed vs. individual predicted concentrations (b); conditional weighted residuals vs. population predicted concentrations (c); and conditional weighted residuals vs. time after last dose administration (d). In A and B, the line of identity is depicted by solid lines. In all subplots, dotted lines depict locally estimated scatterplot smoothing (loess) curves.

Figure 4.

Probability of target attainment (PTA) dosing simulations for study participants. PTA simulations were based on individualized Bayesian PK parameter estimates from the final PopPK model. Simulations excluded 1 participant with renal impairment (estimated GFR <50 mL/min/1.73m²). PTA was defined as the proportion of participants that achieved a fT%>MIC exceeding/or equal to 50% of the dosing interval at steady-state. PTA simulations were summarized for all study participants (i.e., obese and non-obese; N=28) in subplot (a) and obese study participants (N=23) in subplot (b). Grey shaded regions depict the ideal target attainment range (90–100%). The clinical breakpoint point MIC for Pseudomonas aeruginosa (i.e., 8 mg/L) is denoted by vertical dashed lines.

Figure 5.

Probability of target attainment (PTA) visual predictive checks. Five-hundred stochastic trial simulations were generated using the final PopPK model. Each simulated trial included a set of virtual subjects whose demographics were congruent to the subset of study participants with estimated GFR values ≥50 mL/min/1.73m² (N=28). PTA was defined as the proportion of participants that achieved a fT%>MIC exceeding/or equal to 50% of the dosing interval at steady-state. PTA evaluations were computed for three different dosing regimens: 30 mg/kg IV q8h (a), 50 mg/kg IV q8h (b), and 40 mg/kg IV q6h (c). For each dosing regimen, PTA results from stochastic model simulations were compared results generated based on individualized Bayesian PK estimates for study participants with estimated GFR ≥ 50 mL/min/1.73m². Dashed lines depict 50^th percentiles from stochastic model simulations. Grey shaded regions portray 90% PTA predictions intervals generated from stochastic model simulations. Filled symbols denote PTA results from study participants based on individualized Bayesian PK estimates.

Figure 6.

Probability of target attainment (PTA) dosing evaluations based on stochastic model simulations. Each point (filled circles) summarizes the PTA for a set of 100 virtual subjects with the same weight and renal function (i.e., GFR). Three separate dosing strategies were evaluated in each set of virtual subjects: 50 mg/kg IV q8h [max 6 g/day] (a and d), 40 mg/kg IV q6h [max 6 g/day] (b and e), and 40 mg/kg IV q6h [max 8 g/day] (c and f). Each dose was administered via a 30 minute intermittent infusion. All PTA evaluations were conducted at a MIC of 8 mg/L. PTA was defined as the proportion of virtual subjects that achieved a fT%>MIC exceeding/or equal to 50% (a, b, and c) or 100% (d, e, and f) of the dosing interval at steady-state. Evaluated GFR levels (80, 100, and ≥120 mL/min/1.73m²) were reflective of the range of values observed among study participants, excluding one participant with a GFR of ~32 mL/min/1.73m². Grey shaded regions depict the ideal target attainment range (90–100%).