Population pharmacokinetic modeling and Monte Carlo simulation of varying doses of intravenous metronidazole

Kelly A. Sprandel, George L. Drusano, David W. Hecht, John C. Rotschafer, Larry H. Danziger, Keith A. Rodvold

Research output: Contribution to journalArticle

30 Scopus citations

Abstract

Population pharmacokinetic modeling and Monte Carlo simulation (MCS) are approaches used to determine probability of target attainment (PTA) of antimicrobial therapy. The objectives of this study were 1) to determine a population pharmacokinetic model (PPM) using metronidazole and hydroxy-metronidazole concentrations from healthy subjects and critically ill patients, and 2) to determine the probability of attaining the pharmacodynamic target area under the plasma concentration (AUC)/MIC ratio ≥70 against 218 clinical isolates of Bacteroides fragilis using MCS. Eighteen healthy subjects were randomized to 3 dosages of intravenous metronidazole (500 mg every 8 h, 1000 mg day-1, 1500 mg day-1) in an open-label 3-way crossover fashion. Serial blood samples were collected over 25.5 h on the 3rd day of each study period. An additional of 8 critically ill patients received intravenous metronidazole 500 mg every 8 h. Serial blood samples were collected over 8 h after the 2nd day of dosing. Plasma metronidazole and hydroxy-metronidazole concentrations were analyzed using a high-performance liquid chromatographic assay. The 834 plasma concentrations from 62 data sets were simultaneously modeled with Non-Parametric Adaptive Grid population modeling program. A 4-compartment model with a metabolite and zero-order infusion into the central compartment was used. The mean parameter vector and covariance matrix from PPM were inserted into the simulation module of ADAPT II. A 10 000-subject MCS was performed to determine the probability of PTA for a total drug AUC to MIC ratio ≥70 against 218 isolates of B. fragilis (MIC range, 0.125-2.0 mg L-1). Mean parameter values were CLnon-OH, 3.08 L h-1; Vc, 35.4 L; KOH, 0.04 h-1; CLOH, 2.78 L h-1; and VOH, 9.66 L. The regression values of the observed versus predicted concentrations (r2) of metronidazole and hydroxy-metronidazole were 0.972 and 0.980, respectively. The PTA for metronidazole 1500 mg day-1 or 500 mg every 8 h (taken together) and 1000 mg day-1 were 99.9% and 99.8%, respectively, over the reported MIC distribution range. For an MIC of 4 mg L-1, the predicted PTA decreased to 80.0% and 28.5%, respectively. A PPM was determined by comodeling metronidazole and hydroxy-metronidazole concentrations from healthy subjects and critically ill patients. Based on this model, attainment of the target pharmacodynamic parameter (AUC/MIC ratio ≥70) against B. fragilis isolates is >99% when MICs are <2 mg L-1, irrespective of the dosing interval of 24 h.

Original languageEnglish (US)
Pages (from-to)303-309
Number of pages7
JournalDiagnostic Microbiology and Infectious Disease
Volume55
Issue number4
DOIs
StatePublished - Aug 1 2006

Fingerprint Dive into the research topics of 'Population pharmacokinetic modeling and Monte Carlo simulation of varying doses of intravenous metronidazole'. Together they form a unique fingerprint.

  • Cite this