Methods of assessing toxic interactions in vitro: Experimental design and data analysis

The history, theory, statistics, and development of mathematical models that assess biological responses to combinations of chemicals are well represented in the literature. A relative paucity of literature focuses on the actual application of methods and experimental guidelines for assessing toxic interactions. The wide variety of methods in use-some complicated, many nonideal-confounds the comparison of results from different laboratories. This communication is not a comprehensive exposition of the methods appropriate for the study of toxic or pharmacologic interactions of combinations of chemical agents. The modular approach presented here is intended to offer a unifying framework for such studies and to provide a simple yet versatile procedure for new investigators in this area. This approach is modular in that the components, such as experimental designs and mathematical models, can be (re)selected to suit specific research questions and to test the generality of conclusions without altering the overall structure of the approach. A brief background is given with references to some of the most rigorous and recent theoretical developments and reviews of this topic. The strategy and steps of the modular approach are outlined, followed by a more detailed description of these steps, general considerations, and suggested guidelines. Each step (such as experimental design, data analysis) is illustrated by an example. Advantages and disadvantages as well as assumptions and limitations of selected methods are discussed. Also addressed are aspects of in vitro systems that make them particularly conducive to investigations of toxic interactions.