A cancer drug atlas enables synergistic targeting of independent drug vulnerabilities

Abstract

Personalized cancer treatments using combinations of drugs with a synergistic effect is attractive but proves to be highly challenging. Here we present an approach to uncover the efficacy of drug combinations based on the analysis of mono-drug effects. For this we used dose-response data from pharmacogenomic encyclopedias and represent these as a drug atlas. The drug atlas represents the relations between drug effects and allows to identify independent processes for which the tumor might be particularly vulnerable when attacked by two drugs. Our approach enables the prediction of combination-therapy which can be linked to tumor-driving mutations. By using this strategy, we can uncover potential effective drug combinations on a pan-cancer scale. Predicted synergies are provided and have been validated in glioblastoma, breast cancer, melanoma and leukemia mouse-models, resulting in therapeutic synergy in 75% of the tested models. This indicates that we can accurately predict effective drug combinations with translational value.

Reanalysis of in vivo drug synergy validation study rules out synergy in most cases

“…Narayan et al 1 claim the successful in vivo validation of synergy in 5 separate experiments, which they present in Figure 5. Experiments 5A and B concern a U87 glioblastoma model, 5C a triple negative breast cancer model, 5D a melanoma model and 5E a chronic myeloid leukemia model.…”

“…With a similar experimental design using multiple dose levels and sufficiently sized groups, it can also be applied to in vivo data 3 . Narayan et al 1 used relatively small sample sizes, without multiple dosages for the drug and/or its combinations. This means that, in this case, the CI cannot adequately model drug interaction, as the physicochemical massaction law is not fully measured.…”

“…The drug combination studies performed by Narayan et al 1 collected data for various time points, for the same individual mice. Analyzing data per time point ignores this aspect, and typically has less power to find effects than if data for all available data points were used.…”

“…Notably, 5 mice per group sufficed as the variability within groups was modest. Narayan et al 1 , however, have used a shortcut by testing single agents and the combination at just 1 dose level, disobeying the requirement to create dose-effect curves.. Moreover, in some cases, the selected dose level for a single drug treatment was at the extreme end of the response curve.…”

“… Abstract In a recent paper in Nature Communications, Narayan et al 1 described the development of an in silico method to predict synergy of drug combinations, validated in vivo using 5 different models for 5 different drug pairs. The published analysis of the in vivo experimental data used the Chou and Talalay Combination Index, which ignores important aspects of the study design, including, but not limited to the variability between individual mice and the longitudinal nature of the data.…”