Data Rescaling for Evaluation Anti-Tumor Effect of Drug Combination from In Vitro Experiments

• Activity Number: 699
• Type: Contributed
• Date/Time: Thursday, August 13, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Qin Liu* and Xiangfan Yin and Dario Altieri
• Companies: The Wistar Institute and The Wistar Institute and The Wistar Institute
• Keywords: drugs combination ; Synergistic effect ; preclinical ; Loewe additivity ; Bliss independence ; interaction index

Abstract

Drug combination effect is usually tested first in vitro at preclinical stage. A common way to show synergistic effect of drug combination is a heat map with table format based on the percent excess Loewe additivity or Bliss independence rule. Such heat map is not efficient in the drug screening process since it won't give a statistical inference on synergistic effect. An Interaction index (II) is more appreciated and can be estimated by response surface modeling approaches. When data have replications, individual II can be estimated with its 95% confidence interval at specific dose levels. A response surface model requests the response to be from 0 to 100%. The experimental data normalized by averaged readouts from cells in DSMO and in Doxycycline frequently have values beyond the required range. In order to take into account out of range data, we proposed to rescale the data and then calculate II. With real data as examples, we first compared the results from heat maps with and without data rescaling. Our data analysis showed that results from response surface modeling with rescaled data are consistent with heat maps, but easier to show a clear conclusion.