A new method of evaluating antitumor activity from measured tumor volumes

In studies of antitumor activity, measurement of tumor growth delay is commonly used to evaluate treatment effects. However, this measurement sometimes provides no information. We propose an approach for comparing tumor response by using nonparametric regression. In a xenograft model, we found that a significant difference between two treatment effects was undetectable by comparison of tumor growth delays. The results of our finding suggest that measurements used to evaluate the treatment effects should be carefully selected in analyses of antitumor activity.