A Multi-Stage, Model-Free Test of Synergy in Drug Combinations

• Activity Number: 233
• Type: Contributed
• Date/Time: Monday, August 3, 2009 : 2:00 PM to 3:50 PM
• Sponsor: Biopharmaceutical Section
• Author(s): Colleen Kelly*+ and Ping Feng and Toana Kawashima and Anja Wilmes and John Miller
• Companies: Exponent, Inc. and National Base for Drug Clinical Trials and Fred Hutchinson Cancer Research Center and Victoria University of Wellington and Victoria University of Wellington
• Keywords: drug interactions ; dose-response modeling ; response surface ; additivity

Abstract

Most methods of assessing drug interactions involve fitting dose-response curves or surfaces, using an assumed functional form. An exception is Laska et al.'s (1994) model-free test (MFT) which only assumes the potency ratio of the two drugs. We aim to improve the power and robustness of this method. The MFT uses the overly conservative monotonic Min test, which guards against inflated Type I errors when the potency ratio estimate is inaccurate. We develop a more powerful non-monotonic alternative, the multi-stage MFT (MSMFT), that tests the accuracy of the potency ratio and updates it, if necessary. In simulations, the MSMFT demonstrated similar power to the MFT when the potency ratio estimate was accurate and greater power when the potency ratio estimate was inaccurate. We illustrate the methodology in a real example that demonstrates synergy between two chemotherapeutic agents.