Julie Fitzgerald, MD, PhD, FCCM
Children's Hospital of Philadelphia
Philadelphia, Pennsylvania
Disclosure information not submitted.
Athena Zuppa, MD
Director, Center for Clinical Pharmacology
Children's Hospital of Philadelphia
Medford, NJ, United States
Disclosure information not submitted.
Title: Population pharmacokinetic evaluation of vancomycin in critically ill children with sepsis
Introduction: Vancomycin in the most common antibiotic prescribed for critically ill children with suspected/known sepsis. Critically ill children with suspected/known sepsis are at high risk for acute kidney injury. They likely have highly variable and dynamically changing renal function and volume of distribution early in sepsis therapy during periods of fluid resuscitation and altered hemodynamics. Our objective is to develop an optimized pharmacokinetic (PK) model of vancomycin in this population.
Methods: Children >8 weeks and < 18 years of age hospitalized in a PICU with suspected sepsis treated with vancomycin and supported with invasive ventilation and/or vasoactive infusions were prospectively enrolled within 48 hours of intubation or start of vasoactive support. Within the same dosing interval and after ≥3 doses of vancomycin, up to 5 vancomycin concentrations were collected by whole blood microsampling. Whole blood concentrations were measured via tandem MS-HPLC and plasma concentrations were determined using a lab-established conversion factor. A population nonlinear mixed – effects modeling approach using NONMEM was used to characterize vancomycin pharmacokinetics. Both the bedside Schwartz and CKiD equations were tested as covariates on CL.
RESULTS A total of 22 subjects (average (sd) weight 33.2(17) kg and age 8.9(4.9) yrs) were used for this analysis. Pharmacokinetic parameters of vancomycin were estimated using a one-compartment disposition model with weight allometrically-scaled as a covariate on clearance (0.75) and volume of distribution (1). Vancomycin demonstrated a plasma drug clearance of 0.96 L/hr/70kg0.75 and a volume of distribution of 283 L/70kg. The bedside Schwartz equation performed better than CKiD to explain inter-subject variability in CL with the equation CL=0.96*(Bedside Schwartz)0.45.
Conclusions: Inclusion of eGFR using the bedside Schwartz equation performed better than CKiD in this initial cohort and improved the PK model of vancomycin in critically ill children with sepsis. The vancomycin CL was 3-5-fold lower and Vd was substantially higher than that reported in the literature for children not in an ICU.