MedPath

Individualized Cefepime Dosing Study

Not Applicable
Completed
Conditions
Antimicrobial Treatment
Interventions
Other: Blood sampling
Other: Urine sampling
Other: Determination of renal markers
Other: Population pharmacokinetic modeling
Other: Covariate screening
Other: Monte Carlo simulations
Registration Number
NCT02680600
Lead Sponsor
Onze Lieve Vrouw Hospital
Brief Summary

Several population pharmacokinetic (PK) models for cefepime in critically ill patients have been described, all indicating that variability in renal clearance is the main determinant of observed variability in exposure. The main objective of this study was hence to determine which renal marker best predicts cefepime clearance.

Detailed Description

Timely and appropriate antibiotic therapy, sufficient to guarantee adequate antibiotic concentrations in blood and tissues, is one of the most important interventions in critically ill patients with infections.1,2 Cefepime is a fourth generation cephalosporin with broad spectrum activity against Gram-negative bacteria that is used as empirical and directed therapy for severe infections like sepsis and pneumonia. Nevertheless, administration of adequate antibiotic doses is a real challenge in critically ill patients because the pharmacokinetics (PK) of these drugs may be influenced by the complex pathophysiological changes that occur during sepsis.2 Recent reviews described the enormous pharmacokinetic variability of beta-lactam antibiotics in critically ill patients.3,4 Therefore, strategies for dose individualization are explored in an attempt to better control a patient's exposure to the antibiotic, thereby potentially improving the prognosis of critically ill patients with infection. On the one hand, several smaller studies have already shown that better outcomes for critically ill patients can be expected with higher drug exposures, at least for less severely ill patients.5,6 This conclusion was supported by the DALI study, a large multi-center prospective study.7 On the other hand, it was shown that insufficient antibiotic exposure may lead to the development of antibiotic resistance.8 This link was initially shown with inappropriately low quinolone exposures, but more recently with other classes of antibiotics including beta-lactams.9,10 In addition to ensuring that plasma levels are high enough for optimal antimicrobial activity and suppressing emergence of resistance, individualized dosing might offer a perspective to prevent potential side-effects originating from toxic plasma levels. This seems particularly relevant for cefepime, a beta-lactam antibiotic, as it was shown that cefepime is an underappreciated cause of neurotoxicity, especially in intensive care unit (ICU) patients,11,12 patients with impaired renal function,13-16 and patients with brain disorders.17 Population pharmacokinetic models provide a quantitative view of the effect of particular individual factors on the plasma concentration time profile of a drug. Population PK models thereby help to establish individual treatment regimen in patients, depending on the specific patient covariates that were included in the model. As cefepime is a hydrophilic compound, drug elimination is mainly determined by renal clearance and to a lesser extent by non-renal clearance. Therefore, renal markers have been explored as the main determinant to predict cefepime variability in population PK models.18-24 However, none of the published PK models was developed using both plasma and urinary data, though having access to both matrices may be an advantage to identify clinically relevant covariates. Moreover, only creatinine-based markers were used as covariates and, up to now, it was unclear whether the newer markers to assess renal function (e.g. cystatine C, uromodulin and Kidney Injury Moleclure-1 (KIM-1)) are more accurate to predict cefepime clearance.

In this study, a clinical trial was conducted to develop a population PK model for cefepime in critically ill patients assessing renal and non-renal clearance separately, based on both plasma and urinary cefepime concentrations. This model then served as a tool to compare the adequacy of six different renal markers as predictors for renal cefepime clearance. After integrating the most adequate predictor into the PK model, the final model was used to evaluate current dose recommendations for cefepime.

Recruitment & Eligibility

Status
COMPLETED
Sex
All
Target Recruitment
20
Inclusion Criteria
  • Patient age 18 years or more
  • Hospitalized in the ICU of OLV hospital Aalst
  • Elected by the treating physician to receive cefepime,irrespectively of the study
  • Presence of arterial or central line for blood sampling
Exclusion Criteria
  • Exact time of cefepime administration or blood sampling unknown
  • No written informed consent by the patient or his/her (legal) representative

Study & Design

Study Type
INTERVENTIONAL
Study Design
SINGLE_GROUP
Arm && Interventions
GroupInterventionDescription
Study armCefepime dosing* Cefepime dosing * Blood sampling * Urine sampling * Determination of renal markers * Population pharmacokinetic modeling * Covariate screening * Monte Carlo simulations
Study armBlood sampling* Cefepime dosing * Blood sampling * Urine sampling * Determination of renal markers * Population pharmacokinetic modeling * Covariate screening * Monte Carlo simulations
Study armUrine sampling* Cefepime dosing * Blood sampling * Urine sampling * Determination of renal markers * Population pharmacokinetic modeling * Covariate screening * Monte Carlo simulations
Study armPopulation pharmacokinetic modeling* Cefepime dosing * Blood sampling * Urine sampling * Determination of renal markers * Population pharmacokinetic modeling * Covariate screening * Monte Carlo simulations
Study armCovariate screening* Cefepime dosing * Blood sampling * Urine sampling * Determination of renal markers * Population pharmacokinetic modeling * Covariate screening * Monte Carlo simulations
Study armDetermination of renal markers* Cefepime dosing * Blood sampling * Urine sampling * Determination of renal markers * Population pharmacokinetic modeling * Covariate screening * Monte Carlo simulations
Study armMonte Carlo simulations* Cefepime dosing * Blood sampling * Urine sampling * Determination of renal markers * Population pharmacokinetic modeling * Covariate screening * Monte Carlo simulations
Primary Outcome Measures
NameTimeMethod
Median absolute predictive error (MdAPE) of population PK model without covariatesEvaluation during a maximum follow-up period of 5 days
Median absolute predictive error (MdAPE) of population PK model with different renal markers incorporatedEvaluation during a maximum follow-up period of 5 days
Secondary Outcome Measures
NameTimeMethod
The estimated probability of target attianment (%) for the different categories of the Sanford guideBased on data from a maximum follow-up period of 5 days
The estimated probability of toxic levels (%) for the different categories of the Sanford guideBased on data from a maximum follow-up period of 5 days
© Copyright 2025. All Rights Reserved by MedPath