Population Pharmacokinetics of Cefoxitin as a Prophylactic Antimicrobial Agent
Isla A (1), Vázquez S (2), Trocóniz IF (3), López de Tejada I (4), Canut A (2), Muriel López J (4), Solinís MA (1), Gascón AR (1)
(1) Laboratory of Pharmacy and Pharmaceutical Technology, Faculty of Pharmacy, University of the Basque Country; Vitoria-Gasteiz, Spain. (2) Microbiology Unit. Santiago Hospital: Vitoria-Gasteiz, Spain. (3) Department of Pharmacy and Pharmaceutical Technology, Faculty of Pharmacy, University of Navarra; Pamplona, Spain. (4) General and Digestive Surgery Unit, Santiago Hospital; Vitoria-Gasteiz, Spain.
Objectives: The aim of this study was to develop a population pharmacokinetic (PPK) model of cefoxitin (FOX) as a prophylactic agent in patients undergoing elective rectal or colon surgery in order to quantify the degree of inter-patient variability and identify the patient characteristics responsible for such variability.
Methods: Plasma concentration-time data were obtained from 56 patients who received 2 g FOX q2h during surgery. The PPK model was developed using NONMEM VII and the FOCE estimation method with INTERACTION. Selection between models was based on the value of the objective function, the precision of parameter estimates and the goodness of fit plots. Once a base model was selected, patient characteristics including demographic, clinical, laboratory and surgical data were explored for influence on PK parameters. For model evaluation, parameter precision was evaluated computing the 2.5th, 50th, and 97.5th percentiles obtained from the analysis of 1000 bootstrap datasets . Visual and numerical predictive checks were used to explore model performance of the selected model . In both procedures 1000 datasets with the same study design characteristics as the original dataset were simulated.
Results: A one-compartment model best fitted the data. Creatinine clearance (CLCR) was found to have a significant correlation with the total clearance of the drug. The population clearance was expressed as 11.5x(CLCR/77)0.52 and the apparent volume of distribution was 12 L. The percentage of η-, and ε-shrinkage  was greater than 25%, and therefore the use of goodness-of-fit based on the normalized prediction distribution errors (npde)  was justified. The selected model was capable to capture the mean tendency and dispersion of the data during the first two administrations involving 90% of the observations available. The percentiles for Cmax and AUClast obtained from the simulated dataset during the performance of the numerical predictive check were in agreement with raw data, with median values of 159 mg/L, and 201 mgxh/L, respectively.
Conclusions: This study has quantified the influence of CLCR on the PK of FOX in patients undergoing colorectal surgery. Considering that the main objective of antimicrobial prophylaxis is to maintain free-drug plasma levels above the MIC for common contaminating pathogens during the surgery, the developed PPK model will be helpful to redefine dose regimens of FOX for surgical prophylaxis in patients with high CLCR.
. Lindbom L, et al. Comput Methods Programs Biomed 2004; 75:85-94
. Karlsson MO, et al. PAGE 17 (2008) Abstr 1434 [www.page-meeting.org/?abstract=1434].
. Karlsson MO, et al. Clin Pharmacol Ther 2007; 82:17-20.
. Comets E, et al. Comput Methods Programs Biomed 2008; 90:154-166.