How to use meropenem in pediatric patients undergoing CKRT? Integrated meropenem pharmacokinetic model for critically ill children

Abstract

Standard dosing could fail to achieve adequate systemic concentrations in ICU children or may lead to toxicity in children with acute kidney injury. The population pharmacokinetic analysis was used to simultaneously analyze all available data (plasma, prefilter, postfilter, effluent, and urine concentrations) and provide the pharmacokinetic characteristics of meropenem. The probability of target fT > MIC attainment, avoiding toxic levels, during the entire dosing interval was estimated by simulation of different intermittent and continuous infusions in the studied population. A total of 16 critically ill children treated with meropenem were included, with 7 of them undergoing continuous kidney replacement therapy (CKRT). Only 33% of children without CKRT achieved 90% of the time when the free drug concentration exceeded the minimum inhibitory concentration (%fT > MIC) for an MIC of 2 mg/L. In dose simulations, only continuous infusions (60-120 mg/kg in a 24-h infusion) reached the objective in patients <30 kg. In patients undergoing CKRT, the currently used schedule (40 mg/kg/12 h from day 2 in a short infusion of 30 min) was clearly insufficient in patients <30 kg. Keeping the dose to 40 mg/kg q8h without applying renal adjustment and extended infusions (40 mg/kg in 3- or 4-h infusion every 12 h) was sufficient to reach 90% fT > MIC (>2 mg/L) in patients >10 kg. In patients <10 kg, only continuous infusions reached the objective. In patients >30 kg, 60 mg/kg in a 24-h infusion is sufficient and avoids toxicity. This population model could help with an individualized dosing approach that needs to be adopted in critically ill pediatric patients. Critically ill patients subjected to or not to CKRT may benefit from the administration of meropenem in an extended or continuous infusion.

Keywords: acute kidney injury; continuous kidney replacement therapy; continuous renal replacement therapy; critically ill children; dose individualization; meropenem; population pharmacokinetics.

Fig 1

Left panels (1.a, b, c, d): goodness-of-fit plots. Red, blue, green, and purple solid circles represent concentrations of meropenem in plasma/prefilter, postfilter, effluent, and urine, respectively. 1.a: DV, dependent variable (observed measurements); PRED, typical population model predictions; 1.b: IPRED, individual model predictions; 1.c: |IWRES|, absolute individual weighted residuals; 1.d: CWRES, conditional weighted residuals; and TAD, time after dose. Solid lines represent the perfect fit. Right panels (2.a, b, c, d): population-corrected visual predictive checks. Colored solid circles, normalized observations. Solid and dashed lines represent the median and 2.5th and 97.5th percentiles, respectively, of the normalized observations. Colored areas cover the 95% prediction intervals of the 2.5th and 97.5th (gray) and 50th (red) percentiles obtained after simulating 1,000 thousand clinical studies.

Fig 2

Individual model predictions. Each panel shows the (i) individual model predicted plasma concentration vs time profiles (blue) considering the full dosing scheme, and (ii) observed concentrations in red.

Fig 3

Impact of selected covariates on the typical plasma concentration vs time profiles of meropenem. Dosing regimen consisted of 150 mg of meropenem infused intravenously during 30 min every 8 h, except in the upper right panel, where the dose levels administered were 100, 150, and 200 mg for patients weighting 5, 7.5, and 10 kg, respectively.

Fig 4

Target attainment. Each panel shows the percentage target attainment (%TA) per total body weight group for patients without (upper panels) and with (lower panels) continuous kidney replacement therapy. Each colored profile represents one of the dosing schedules listed on the right. %TA is calculated as the mean% time above minimum inhibitory concentrations during the first dosing interval of the second day of treatment.

Fig 5

Full plasma concentration vs time profiles in patients with continuous renal replacement therapy. Colored areas cover the 95% prediction intervals of the concentrations generated from 1,000 virtual patients for each dosing scenario and weight group. The horizontal solid lines highlight the plasma values of 45 mg/mL (red) and 65 mg/mL (black).