EXPERIMENTAL DESIGNS FOR COMBINATION DOSE-RESPONSE STUDIES IN PRECLINICAL RESEARCH

Preclinical combinatorial drug studies face significant challenges due to the vast number of combinations and conditions. Combination MCP-Mod was recently developed to address a methodological gap, extending an existing method for analyzing two-drug treatments. This approach enables within-dose level contrast analysis, prediction of advantageous dosages for future studies, and sample size calculation. Our results demonstrate that Combination MCP-Mod outperforms traditional Multiple Comparison Procedures (MCP) in identifying underlying dose-response models, capturing the dose-response signal, and detecting the target dose set across various experimental designs. Notably, Combination MCP-Mod requires a substantially smaller sample size to achieve adequate power. We assess the impact of experimental design factors like grid fulfillment, sample size per dose level, and number of dose levels. The method framework maintained strong performance, allowing for the utilization of a larger number of dose levels within combinatorial dose-ranging studies by leaving out certain combinations within the grid, meanwhile keeping the total sample size relatively small. Keywords: Dose-response, Combination MCP-Mod, Grid fulfillment, Proof of Concept.