Skip to main content
Premium Trial:

Request an Annual Quote

MIT-Led Team Develops Genomics Platform to Support Heart Disease Treatment Decisions


NEW YORK – A group of investigators have developed a new platform for precision cardiology that is already being used in drug development and could eventually be harnessed to guide therapy selection in at-risk patients.

The work to develop the platform was led by researchers at the Massachusetts Institute of Technology, the Wellcome Sanger Institute, the University of Cambridge, and Lund University in Sweden. A paper describing the effort appeared in the journal npj Genomic Medicine in June.

The concept underlying the platform integrates diverse, wide-ranging phenotyping and genomic data on patients to better understand the mechanisms at play in inherited heart diseases. By doing so, clinicians can build models to lead them to treatment decisions, rather than prescribing therapies and gauging patients' responses after the fact.

In the study, the investigators showcased how their model could predict the impact of genetic variations on troponins, proteins involved in inherited heart diseases. They also analyzed data on 980 patients culled from 100 existing studies to look into the effect troponin T genetic variations had on interactions on proteins in heart muscle. This resulted in an intricate mapping of hot spots and corresponding clinical outcomes.

"The whole premise of this work is that the understanding of the genomics landscape of subcomplexes and intramolecular interactions is very complex," said Rameen Shakur, a clinician scientist in cardiology at MIT and an author on the paper. "What tends to happen is that the clinician says, 'This is the therapy we can give you, this is the management we can offer now,' and we fit our model around the understanding of what the biology should be like," Shakur said.

The aim of the recent study was to create a model that could support a better understanding of the links between genetic variations and clinical manifestations. By looking at a patient's data, clinicians could be able to understand what kinds of problems they could have in the future, even if they do not present with any symptoms, the researchers hypothesized.

"This is where precision and prognostication medicine needs to be," said Shakur. "You need to understand the biology, and you need to let the biology dictate where the clinical work is going, not the other way around," he said.

Shakur's work on the platform dates back several years to when he was a clinical researcher at the Wellcome Trust Sanger Institute in the UK. He received an Independent Wellcome Trust Fellowship to pursue his research interests, which gave him more or less carte blanche to embark on this scientific adventure. "I went in with a curiosity mindset," said Shakur. "I wasn't looking to find anything. But when you are in a position of not knowing, it can be so refreshing."

Shakur had faced challenges in his work as a clinician, especially working with patients who were flagged for having potentially pathogenic variations but otherwise presented normally. It indicated to him that there needed to be more dialogue between clinical geneticists, cardiologists, and other players.

"Clinically, I was seeing patients who had no problems on their heart size, or differences in screening, but they had a mutation where the genomics guys were telling me, 'This is a mutation that you need to be aware of,' but we couldn't make any clinical decision," said Shakur. "That's why I said, we need to talk to each other clinically."

As an example, he noted the recent case of Christian Eriksen, a Danish footballer who collapsed recently during a match with Finland during the 2020 UEFA European Football Championship and experienced a near-death cardiac arrest before being revived with CPR and defibrillation. "This is what this work is all about," said Shakur. "This is happening in the real world with patients where we don't see any changes in the heart, we don't see an enlarged heart, but we still have these electrical problems," he said.

Developing models to predict who is at greater risk of experiencing such cardiac events requires a change in perspective, though, Shakur said. "Biology is a spectrum," he said. "When you go to a physician, they will put you in one box or another. But that is not what biology is really doing."

Shakur's effort eventually expanded to include contributions from multiple institutions and was bankrolled by other funders, including the Medical Research Council in the UK, as well as the Swedish Research Council. It involved searching datasets to build models that could be used in precision oncology, which was no easy task.

"This has taken a long time, more than it should, because people don't share their data widely enough in the scientific community," noted Shakur. "We had 981 patients [in our study] because we scoured every paper out there."

In particular, the researchers focused their work on the cardiac troponin T variations, which are known to be associated with increased risk of cardiac death or familial cardiomyopathy. However, the outcomes of patients who carry these variations and their responses to therapies vary. In the study, the researchers interrogated the molecular interactions across the heart's actin thin filament complex to better determine their prognosis and to guide patient care decisions.

The work resulted in a model that can be used to survey the functional, structural, and chemical consequences of carrying certain troponin T variations. They then integrated this model with clinical data of familial cardiomyopathy cases from 106 publications describing 136 disease-causing variations from 981 cases. The troponin T variations showed different hot spots for dilated and hypertrophic cardiomyopathies, and the researchers discovered there were worse survival outcomes for patients carrying certain variations in certain regions.

The researchers were also able to group patients into high and low survival groups based on where they happened to carry troponin T variations. This led them to claim in the paper that their integrative genomic, structural model from genotype to clinical data integration has "implications for enhancing clinical genomics methodologies to improve risk stratification."

They cautioned at the end of the paper, however, that analyzing clinical outcomes of variations could be complicated by variable penetrance in patients with the same variation. Environmental, lifestyle, and other genetic factors could also influence outcomes. "Accounting for these features will require extensive large-scale longitudinal clinical genomics studies," they wrote in the paper.

Still, they believe their approach could lay the groundwork for the use of combining systems biology, clinical genomics, and phenotyping data to better understand how troponin T variations in a patient might impact outcome and to select better treatments based on that information.

The premise was embraced by Mark Caulfield, chief scientific officer at Genomics England, who commented publicly on the publication, though Genomics England was not involved in the work.

"This new tool is a significant advancement in precision cardiology, with real potential to benefit patients with inherited heart disease and influence genomic healthcare," Caulfield said in a statement. "The study demonstrates the importance of collaboration and knowledge sharing in driving innovation with real world outcomes."

For Shakur, the new approach provides a glimpse of how patients could be managed in the future. "We can look at the biology and say maybe we should do better surveillance if the person has variations in a region of significance," he said. "The current status quo for clinical management is that everyone should fit the whole shoe," he added. "That is not true."

By looking at the underlying genomic landscape, Shakur believes clinicians​, scientists, and patients should have a shared understanding on assessing both the biological and clinical consequences of variations to look at a future of proactive and synergistic management.

Next steps

Much of the hard work of sifting through publications and datasets was carried out by Health in Code, a private company based in Galicia, the northwesternmost community of Spain on the Atlantic seaboard. Health in Code focused on clinical evaluation, building empirical curves, and collecting information about patients identified in papers, as well as interpretation of results.

According to Lorenzo Monserrat, a cardiologist at the University of A Coruña and cofounder of Health in Code who served as its CEO until last year, this extensive pooling of data is something that is rarely accomplished in clinical genetics today.

"The most important contribution [of this paper] is the combination of clinical details on a large number of patients and families with basic structural, functional, and molecular biology," Monserrat said. "Most studies in genetics, if you read the publications, discuss whether variants are pathogenic or nonpathogenic," he said. "We think we have to go farther and determine what is the phenotype, the outcome, and the prognosis of individuals carrying these genetic variants."

Armed with that information, it will be easier for clinicians to manage cases, Monserrat maintained. "Current recommendations are really poor," he said. "We receive information about new therapies that might be used to modify the clinical presentation of disease, but it's clear that these therapies will not work for all patients."

In the future, information on variation could be folded into individual reports to establish prognosis, anticipate outcome, and direct follow-up of patients, he added.

Shakur said he envisions a similar vehicle for introducing such data into routine clinical use. But, he noted, the work is still in its early stages. The researchers will seek additional funding to expand the work, potentially bringing in new collaborators from the Far East, South America, and Europe. They also plan to investigate inherited variations in other subcomplexes using their model, looking at myosin thick filament and its interaction with actin thin filament. "In essence, the process works, and this can be reproduced for other systems," said Shakur.

There is also, of course, the application of the model in drug development, which Shakur said is already underway. He said that a trial with an academic partner began a year and a half ago, where a patient received a drug specific for his mutation. The trial will close by the end of this year.