Research sets sights on heart disease discoveries through precision cardiology and bioinformatics

Heart with big data visualization

Globally, heart disease is the leading cause of death in men and women and is projected to affect almost half of all adults by 2030. Nevertheless, cardiovascular research lags when it comes to harnessing technology, big data and funding – critical elements for precision cardiovascular medicine to make the sort of dramatic advances achieved in cancer research and care.

“Cardiovascular disease as a field hasn’t seen very much computational firepower coming to the rescue to do something about it, at least relative to precision oncology,” said Bohdan Khomtchouk, PhD, an instructor in the Section of Computational Biomedicine and Biomedical Data Science at UChicago Medicine’s Institute for Genomics and Systems Biology. “We need more people to help us fight heart disease with computation.”

Khomtchouk recently co-authored a review article in the journal Briefings in Bioinformatics, summarizing the current state of cardiovascular disease literature and cardioinformatics, the emerging field for administering and analyzing data created in precision cardiovascular medicine.

We need more people to help us fight heart disease with computation. 

Using large-scale text mining and natural language processing methods, the researchers analyzed existing published literature for cancer and heart disease research and found that although cardiovascular research has increased since 2000, the number of studies tapping bioinformatics and computational biology techniques remains modest compared to those for oncology. Analysis of data from the National Institutes of Health revealed cardiovascular research has received approximately $500 million less annually in funding than cancer research since 2008.

“The field of cardiovascular disease has always been data and technology driven, but when it comes to taking advantage of the vast amount of information that already exists and developing a systematic approach for analyzing it, the field has fallen short,” said James K. Liao, MD, Chief of the Section of Cardiology at UChicago Medicine. “The emergence of computational biology and bioinformatics in the cardiovascular field will provide insights and potential therapeutic avenues for the treatment of cardiovascular patients.”

Positive trends in software, wearables

The researchers note some positive trends in developing cardioinformatics, such as the recent partnership between the American Heart Association Institute for Precision Cardiovascular Medicine and Amazon Web Services to fund grants for developing artificial intelligence that improve understanding of precision cardiovascular medicine data. Several academic institutions, non-profits, companies and government entities are creating software platforms like HeartBioPortal and HeartBD2K to organize and integrate biomedical data with the hopes of spurring more discoveries in cardiovascular health and disease.

Wearable devices and sensors capable of collecting real-time data also make it increasingly possible to monitor an individual’s exposure to chemical compounds and micro-organisms, as well as their early symptoms of cardiovascular disease. Software that classifies images of food and drink consumption in social media timeline photos could even help track the diet and lifestyle habits of patients with atherosclerosis. These technical advances in measuring the environmental factors and exposures that influence cardiovascular disease mean that gene-environment interactions can now be studied in greater detail than previously possible.

Nevertheless, cardiovascular researchers face unique challenges when tapping bioinformatics due to the peculiarities of these enormously complex diseases. Unlike cancer, which progresses rapidly and often has a clear before-and-after disease state, cardiovascular disease develops over decades and may transition from one diagnosis (such as hypertension) to another (atherosclerosis). Frequently, cardiovascular disease patients also have other illnesses, such as diabetes or cancer.

“Monitoring the causes and development of this complicated, progressive disease is an incredibly computationally challenging task,” said Khomtchouk, who previously served as an American Heart Association Postdoctoral Fellow in Stanford University’s Department of Biology and Department of Medicine (Division of Cardiovascular Medicine) prior to joining the University of Chicago. “We need new, innovative data integration approaches for risk stratification and surveillance at both an individual and population level for different race and ethnic groups.”

The researchers propose prioritizing three strategies to advance cardioinformatics: making more biomedical data accessible to researchers while protecting patient privacy, integrating different types of multi-omics data to improve research, and developing artificial intelligence and machine learning models while respecting the role of human expertise in medicine.

The study, "Cardioinformatics: the nexus of bioinformatics and precision cardiology," was supported by the American Heart Association and the National Science Foundation. Additional authors of the article include Diem-Trang Tran of the University of Utah, Kasra Vand of Quiltomics, Matthew Might of the University of Alabama at Birmingham and Themistocles Assimes and Or Gozani of Stanford University. Liao was not involved in the review article.