An Application of a Hill-based Response Surface Model for a Drug Combination Experiment on Lung Cancer

摘要

Combination chemotherapy with multiple drugs has been widely applied to cancer treatment due to enhanced e cacy and reduced drug resistance. For drug combination experiment analysis, response surface modeling has been commonly adopted. In this paper, we introduce a Hill-based global response surface model and provide an application of the model to a 512-run drug combination experiment with three chemicals, namely AG490, U0126, and indirubin-3’-monoxime (I-3-M), on lung cancer cells. The results demonstrate generally improved goodness of fit of our model from the traditional polynomial model, as well as the original Hill model based on fixed-ratio drug combinations. We identify different dose-effect patterns between normal and cancer cells based on our model, which indicates the potential effectiveness of the drug combination in cancer treatment. Meanwhile, drug interactions are analyzed both qualitatively and quantitatively. The distinct interaction patterns between U0126 and I-3-M on two types of cells uncovered by the model could be a further indicator of the efficacy of the drug combination.