Page - (000088) - in Biomedical Chemistry: Current Trends and Developments

mathematical models have been employed to predict the pharmacokinetic attributes of compounds, preventing the synthesis and screening of compounds with obvious liabilities. Based on mathematical and statistical relationships between the physicochemical and molecular characteristics of the NCEs and their pharmacokinetic properties, recent in silico models are commercialized as software that predict pharmacokinetic parameters: Meteor®, MetabolExpert® and MetaSite® are examples of commercial metabolic fate/stability predictors; GastroPlus® and IDEA® perform pharmacokinetic simulations of the rate and extent of absorption in gastrointestinal tract (Parrott, 2002; Kuentz, 2006). Although in silico models have higher throughput ADME screening ability, saving more time and effort than in vitro and in vivo models, they can hardly be accurate enough to replace real circumstances (Lavè, 2007; Burton, 2010; Gallo, 2010). Thus, during the phases prior to lead characterization where the throughput is high, the potential pharmacokinetic features are mainly addressed by in vitro tools. On the other hand, in the last phase of drug discovery, more comprehensive in vitro and in vivo models may provide a definitive assessment of overall drug disposition of the best compounds previously screened. Table 2.1.2 summarizes the major in vitro model systems most frequently employed for ADME/pharmacokinetic analysis in the early stages of drug discovery programs and their respective objectives. At the initial stages of drug discovery, ADME screening is conducted in a high-throughput mode and includes, but it is not limited to, determinations of apparent permeability [using parallel artificial membrane permeability assays (PAMPA) and/or cell lines such as human colon adenocarcinoma (Caco- 2) or Madin-Darby canine kidney cells (MDCK)], plasma protein binding, human or animal liver microsomal stability, and identification of the reactive metabolites. Most of these in vitro models are generally amenable to miniaturization and automation, improving the evaluation capacity of HTS assays (Mayr, 2009). For ADME/pharmacokinetic analysis in the late stage of drug discovery, major methodologies include protein binding (as well as blood– plasma partitioning and plasma stability), hepatocyte stability, metabolic enzyme phenotyping and human cytochrome P450 enzyme (CYP) inhibition/induction.