Abstract 15727: Cancer and Atherosclerosis Share Numerous Pathogenic Drivers Representing Novel Translational Targets for Mutual Benefit

Richard Baylis,Hua Gao,Fudi Wang,Caitlin Bell,Lingfeng Luo,Jeanette Erdmann,Johan Bjorkegren,Derek Klarin,Nick J. Leeper

DOI: https://doi.org/10.1161/circ.146.suppl_1.15727

IF: 37.8

2022-01-01

Circulation

Abstract:Cancer and cardiovascular disease make up nearly half of all deaths in the United States. Despite knowing a great deal about each of these diseases individually, surprisingly little is known about how these diseases interact within patients or the common dysregulated genes and pathways that these diseases share. These diseases are known to share numerous risk factors and emerging epidemiologic data suggests that the presence of coronary artery disease (CAD) increases the risk of developing cancer. In addition, studies like the CANTOS trial have shown that targeting mutually detrimental pathways could yield benefit for both diseases. To understand the relationship of cancer and CAD, we tested two hypotheses, 1) the similarities between cancer and CAD are explained, in part, by shared genetic factors and 2) these diseases share dysregulated processes/pathogenic drivers and that targeting these processes could result in mutual clinical benefit. To test these hypotheses, we assessed the genetic correlation of CAD and each cancer subtype in the UK biobank. This identified a surprising genetic overlap between prostate cancer and CAD. However, the other cancers lacked significant genetic correlation with CAD, suggesting that common genetic factors were unlikely to be mediating their interaction. To identify common pathogenic drivers, we used transcriptomic data that had been correlated with either cancer survival or CAD disease severity/outcomes to generate transcriptional risk profiles for each disease. We then intersected these data to identify mutually detrimental processes, which included several known pathways including inflammation and hypoxia but also numerous novel pathways. Interestingly, we found that many cancers were indistinguishable from CAD based on their transcriptomic risk profile. Next, we used the drug prediction algorithm, OCTAD, to identify novel drug targets for each disease state and then validated our predictions using a pharmacovigilance approach with EMR data. Taken together, our study suggests that the interaction of cancer and CAD is primarily driven by shared pathogenic drivers rather than shared genetic predisposition and that targeting these processes may allow for mutual clinical benefit.