Interaction of cytotoxic agents: a rule-based system for computer-assisted cell survival analysis

The actual effectiveness of environmental noxious agents or anticancer drugs can be fully determined only by knowing if the effects (in the present case, the cytotoxic effects) induced by a given agent are enhanced by exposure to another (or other) agent(s). Given a certain combination of agents, it is possible to distinguish three types of interaction: (a) zero interaction or additivity; (b) positive interaction or synergism; and (c) negative interaction or antagonism. In this work, the methodological problems involved in evaluating the type and level of interaction between biologically active agents are discussed and an "intelligent" approach to the problem is proposed. In particular, a prototype of a computer-assisted rule based system, named CISA (Cytotoxic Interaction and Survival Analysis), designed in a KES environment (Knowledge Engineering System) and implemented on a personal computer, is described. By constructing isoboles based on experimental cell survival data and taking into account the relative confidence intervals, the system can indicate the appropriate combinations of dosages to be tested and finally determine the type and level of interaction. The system, which represents an attempt to administer "intelligently" the experimental data, is therefore able to identify the best strategy of analysis, to carry out the data processing and to offer suggestions to the investigator about the usefulness of the data and the planning of further experiments.