In silico ADME and toxicity prediction of ceftazidime and its impurities

Abstract

To improve the quality control of drugs, we predicted the absorption, distribution, metabolism, excretion, and toxicity (ADMET) of ceftazidime (CAZ) and its impurities via in silico methods. We used three types of quantitative structure-activity relationship and docking software for precise prediction: Discovery Studio 4.0, OECD QSAR Toolbox 4.1, Toxtree, and the pkCSM approach. The pharmacokinetics and toxicity of ceftazidime and impurity A (Δ-2-CAZ) are similar. The biological properties of impurity B (CAZ E-isomer) are different from CAZ. Therefore, we focused on drug stability to analyze impurity B. Impurities D and I have strong lipophilicity, good intestinal absorption, and poor excretion in the body. Impurity D is particularly neurotoxic and genotoxic. It is important to control the content of impurity D. The toxicity of impurity F is low, but the toxicity is enhanced when it becomes the C-3 side chain of CAZ and forms a quaternary amine group. We conclude that the beta-lactam ring of nucleus, the quaternary amine group at the C-3 side chain, and the acetates at the C-7 side chain of CAZ are the main toxic functional groups. Impurities B and D may be the genetic impurity in CAZ and may also have neurotoxicity. This in silico approach can predict the toxicity of other cephalosporins and impurities.