Abstract 4245: Drug response metrics and pharmacological profiling using the OncoPanel™ cell-based profiling service

High-throughput, cell-based drug screening platforms are essential to guiding drug discovery programs and can rapidly generate large amounts of pharmacological information about a candidate drug. Drug response data combined with genomic and phenotypic information enables an exploratory analysis that may reveal potential associative biomarkers and deepen the pharmacological profile of a drug. The choice of metric is an important consideration as different metrics will reflect different underlying biological mechanisms. Normalized growth rate inhibition (GR) metrics offer an alternative summary of drug response, and may eliminate bias associated with experimental factors that contribute to variation in growth rate. In these sets of experiments, we used drug response data, generated with the Eurofins OncoPanel cell-based profiling service, to investigate a variety of standard of care compounds against a cell line panel selected for variation in growth rate, cancer indication, cancer hallmark mutational status, and baseline apoptosis rate. Seeding density and incubation times were systematically varied to investigate the relationship between end-point confluence, kinetics, genotype, apoptotic priming, and the robustness of different dose response metrics. Based on these results, select cell lines and standard of care compounds were included in a series of drug-drug combination studies to further investigate the biological underpinning of synergy, at the same time exploring how alternative metrics influence pharmacological profiling experiments. Compounds were added to cells in 10 half-log dilutions, in triplicate, and allowed to incubate for 1, 3, 5, or 10 days. Initial cell density conditions were varied to allow for a final targeted confluence of 30, 80, or 100%. A time-zero plate was fixed at compound addition to allow for GR calculations. Following incubation, the cells were fixed, stained with DAPI, and imaged with a high-content imaging system. Cell proliferation dose response curves were fitted using a 4-parameter log-logistic model with custom curve-fitting software. These data were re-analyzed using the GRmetrics package in R to determine the GR metrics. Summary parameters, including IC50, GR50, and GRmax, were compared in subsequent analyses. For the combination analyses, a 9 x 9 concentration matrix design was used, and Bliss analysis was conducted to assess synergy. Single parameters are typically used to summarize drug responses in cell lines, whi