A Population pharmacokinetic structural model for Vancomycin in preterm neonates

ANDRÉS PORTA, MÓNICA TRAVAGLIANTI, GRACIELA CASTRO, NIEVES LICCIARDONE, PAULO CÁCERES GUIDO, VIVIANA NISELMAN, PAULA SCHAIQUEVICH

Sttutgart

Congreso; 12th International Congress of Therapeutic Drug Monitoring & Clinical Toxicology; 2011

IATDMCT

Background Preterm neonates are susceptible to nosocomial infections during their stay in the neonatal intensive care unit. Vancomycin is a frequently used antibiotic for treatment of nosocomial bacterial pathogens. Low plasma concentrations may result in non effective therapy while high levels are reported to be associated with nephrotoxicity and ototoxicity. Besides, vancomycin clearance shows a high inter-individual variability in the pharmacokinetic parameters in neonates. The aim of this work was to develop a population pharmacokinetic structural model based on non-linear mixed effects models as a basis for further model development to optimize the dose regimen of vancomycin in preterm neonates. Methods The study population consisted of 118 preterm newborn infants hospitalized at the neonatal intensive care unit at our Hospital (tertiary-level hospital, Argentina) between 2000 and 2011. A total of 269 plasma vancomycin levels (prospective data) obtained before the next dose and 1-h after the end of vancomycin infusion and associated covariates (postmenstrual age; postnatal age; weight; height; clearance of creatinine; bilirubin; apgar score; associated pathologies; concomitant medication) were recorded. The model was developed using the stochastic approximation EM-algorithm implemented in Monolix. Model selection was performed by means of graphical and statistical information criteria also provided by Monolix. Results We selected a one-compartment linear disposition model with zero order input and first-order elimination. The residual variability was modeled using a mixed error model (additive and proportional). Mean population estimate for volume of distribution was 905 ml (CV 13%) with a between-subject variability 81.6%. Mean population estimate for clearance was 96.3 ml/h (CV 10%) and between-subject variability was 95.4%. The parameters of the residual error model ( were: a= 4.17(CV 17%), b=0.222(CV 28 %). Conclusion The population parameters estimated with Monolix, which is a free and based on well mathematically studied algorithms software, are consistent with those reported in the literature and obtained using other PopPK programs. This is the first model developed for vancomycin in neonates in a South American country population. Covariates will be next studied so as to explain the high between-subject variability observed in the pharmacokinetic parameters. This analysis will allow us to obtain an individualized dose regimen of vancomycin in the studied population.