Comprehensive Early Toxicology Profiling of lead compounds is increasingly being considered as a novel strategy to reduce attrition rates in drug development. The principal safety concerns are usually in the area of cardiosafety, genetic toxicology and target organ toxicity for which standard experimental procedures are well established in low throughput format. The challenge is to identify predictive in vitro assays and technologies, and how to implement these assays into an automated screening environment. Results will be presented from high-throughput cell-based assays for hERG, Micronuclei formation and liver toxicity prediction. Examples will be shown how this data is being used for early decision making in the drug discovery process.

Predicting carcinogenesis is not an exact science. Even proposed mechanisms for genotoxicity, one of the more common properties of carcinogens, can be unknown, genuinely multiple or disputed. The effectiveness of genotoxicity screening methods is assessed by their ability to identify well-known genotoxins and/or carcinogens. Before the synthesis of a new compound, in silico methods based on the recognition of alerting structures are increasingly used to design out future problems. Mammalian germ cell mutagenesis and carcinogenesis certainly are essentially death warrants for compounds in development. In between the conception and death of a carcinogen lie the bioassays for genotoxicity. Data will be presented for both well-characterized compounds and proprietary compounds, to reveal how different screening methods, both in vitro and in silico, vary in their coverage of different compound collections.

In vitro cell toxicity studies are required as part of the early phase assessment of the safety of drug candidates. A change in the dynamic behavior of cells or in cell morphology is a powerful indicator of cellular responses which current methods cannot measure. Here we have studied the possibility of using a machine vision technology to follow and automatically analyze changes in the dynamic behavior of living cells during exposure to test compounds. We demonstrated several different patterns of cellular death dynamics that could be used to profile compounds. We also noted that a decrease in cell migration and changes in cell morphology (e.g. the surface area of neurites, appearance of cytoplasmic vacuoles) are more sensitive indicators of cell toxicity than end points such as total ATP. Thus, the machine vision technology permits simultaneous analysis of multiple parameters in a cell population without the use of labels or assay kits. The technology permits novel and sensitive ways of profiling compounds based on their effects on cell behavior and morphology.