Response surface methodology (RSM), a collection of statistical and mathematical techniques, can empirically model multiple agent interactions. The use of RSM allows modeling of the interaction at all levels of each agent, rather than focusing on a single level of effect, thereby describing the changing interaction present. Further, this technique allows qualitative analysis of the resultant model of the combination exposure for existing interactions. RSM was used in the statistical analysis of sister chromatid exchanges (SCE) elicited by combination exposures of genotoxic agents, ethylnitrosouera (ENU) with cis-diaminedichloroplatinum (II) (DDP), and DDP with x-rays. ENU and DDP each exhibited curvilinear concentration-related increases in SCEs. X-ray exhibited a linear increase in SCEs. For the DDP-x-ray combination, RSM analysis indicates a less-than-additive interaction, suggested by the nonparallel concentration-response curves of one agent at varying concentrations of the other. For the DDP-ENU combination, an increasingly less-than-additive interaction was detected. This descriptive ability is shown to be useful in suggesting hypotheses about mechanisms of action. Employing a polynomial model assumes effect-addition as an underlying assumption of additivity. Models with other underlying assumptions of additivity also can be used.
All Science Journal Classification (ASJC) codes