Get access

Pharmacokinetic–pharmacodynamic modeling and simulation for in vivo bactericidal effect in murine infection model

Authors

  • Takayuki Katsube,

    1. Biostatistics Department, Shionogi & Co., Ltd., Sagisu 5-12-4, Fukushima-ku, Osaka 553-0002, Japan
    Search for more papers by this author
  • Yoshinori Yamano,

    1. Discovery Research Laboratories, Shionogi & Co., Ltd., Futaba-cho 3-1-1, Toyonaka-shi, Osaka 561-0825, Japan
    Search for more papers by this author
  • Yoshitaka Yano

    Corresponding author
    1. Biostatistics Department, Shionogi & Co., Ltd., Sagisu 5-12-4, Fukushima-ku, Osaka 553-0002, Japan
    2. Graduate School of Pharmaceutical Sciences, Kyoto University, Yoshida-Shimoadachi 46-29, Sakyo-ku, Kyoto 606-8501, Japan
    • Biostatistics Department, Shionogi & Co., Ltd., Sagisu 5-12-4, Fukushima-ku, Osaka 553-0002, Japan. Telephone: +81-75-753-9254; Fax: +81-75-753-4502
    Search for more papers by this author

Abstract

A pharmacokinetic (PK)/pharmacodynamic (PD) modeling strategy to simulate in vivo bactericidal effects for three carbapenem antibiotics, doripenem (DRPM), meropenem (MEPM)/cilastatin (CS), and imipenem (IPM)/CS, against a Pseudomonas aeruginosa (P. aeruginosa) strain is proposed. The PD model we have already developed to explain in vitro time-kill profiles was modified to incorporate the growth rate, bactericidal activities, and PK profiles in murine lung infection models. Plasma concentration data and bacterial time-kill data for each antibiotic consist of six and eight time points, respectively, at one dose regimen (four or five mouse/point). In vivo time-kill curves could be well simulated for each antibiotic by the PK/PD model. Simulated bacterial counts at 24 h and PK/PD indices derived from total drug concentrations (time above the minimum inhibitory concentration (MIC) (T > MIC), Cmax/MIC, and AUC/MIC) for various dose regimens were examined for MEPM/CS and IPM/CS. Simulated bacterial counts correlated only with T > MIC (correlation coefficient: 0.951 for MEPM/CS, 0.982 for IPM/CS) and T > MIC values to achieve a bacteriostatic effect and a 2-log killing effect for both antibiotics were estimated to be approximately 15 and 20%, respectively, which are similar to previously reported results. These findings suggested that the proposed PK/PD model is a good tool for predicting in vivo bactericidal effects. © 2007 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 97:1606–1614, 2008

Ancillary