Matthew Fidler1, Steven E Kern. 1. Department of Pharmaceutics and Pharmaceutical Chemistry, University of Utah, Salt Lake City, Utah 84108, USA.
Abstract
BACKGROUND: Minto et al. (Anesthesiology 2000) described a mathematical approach based on response surface methods for characterizing drug-drug interactions between several intravenous anesthetic drugs. To extend this effort, the authors developed a flexible interaction model based on the general Hill dose-response relation that includes a set of parameters that can be statistically assessed for interaction significance. METHODS: This new model was developed to identify pharmacologically meaningful interaction-related parameters and address mathematical limitations in previous models. The flexible interaction model and the model of Minto et al. were compared in their assessment of additivity using simulated sample data sets. The flexible interaction model was also compared with the Minto model in describing drug interactions using data from several other clinical studies of propofol, opioids, and benzodiazepines from Short et al. (Anesthesiology 2002) and Kern et al. (Anesthesiology 2004). RESULTS: The flexible interaction model was able to accurately classify an additive interaction based on the classic definition proposed by Loewe, with at most an 8% difference between the two surfaces. Also, the proposed model fit the clinical interaction data as well or slightly better than that of Minto et al. CONCLUSIONS: The new model can accurately classify additive and synergistic drug interactions. It also can classify antagonistic interactions with biologically rational surfaces. This has been a problem for other interaction models in the past. The statistically assessable interaction parameters provide a quantitative manner to assess the interaction significance.
BACKGROUND: Minto et al. (Anesthesiology 2000) described a mathematical approach based on response surface methods for characterizing drug-drug interactions between several intravenous anesthetic drugs. To extend this effort, the authors developed a flexible interaction model based on the general Hill dose-response relation that includes a set of parameters that can be statistically assessed for interaction significance. METHODS: This new model was developed to identify pharmacologically meaningful interaction-related parameters and address mathematical limitations in previous models. The flexible interaction model and the model of Minto et al. were compared in their assessment of additivity using simulated sample data sets. The flexible interaction model was also compared with the Minto model in describing drug interactions using data from several other clinical studies of propofol, opioids, and benzodiazepines from Short et al. (Anesthesiology 2002) and Kern et al. (Anesthesiology 2004). RESULTS: The flexible interaction model was able to accurately classify an additive interaction based on the classic definition proposed by Loewe, with at most an 8% difference between the two surfaces. Also, the proposed model fit the clinical interaction data as well or slightly better than that of Minto et al. CONCLUSIONS: The new model can accurately classify additive and synergistic drug interactions. It also can classify antagonistic interactions with biologically rational surfaces. This has been a problem for other interaction models in the past. The statistically assessable interaction parameters provide a quantitative manner to assess the interaction significance.