Skip to main content
Premium Trial:

Request an Annual Quote

Duke Study IDs Predictive Gene Expression Signature for Aspirin Response in Coronary Artery Disease


Researchers from Duke University have developed a gene-expression signature that is associated with how patients with coronary artery disease respond to preventive aspirin therapy.

The group published its results online in the Journal of the American College of Cardiology earlier this month.

Though the results are preliminary, the team believes the discovery could potentially lead to a clinical blood test doctors could use to predict patient outcomes and guide therapeutic decision making by distinguishing which patients are responding well to aspirin and which might benefit from higher doses or alternative platelet inhibiting drugs like Bristol-Myers Squibb's Plavix (clopidogrel) or its newer competitors.

Deepak Voora, the study's first author told PGx Reporter this week that it has been long apparent to clinicians that some patients respond well to aspirin therapy while others remain at risk for strokes or heart attacks even when treated with the drug. Some tests for aspirin response based on platelet function currently exist, but are technically cumbersome and only available at large medical centers.

Voora said the Duke team's research was driven by the need for an alternate biomarker or signature to assess aspirin response that could be used more easily and more broadly in clinical practice.

"People have used traditional GWAS approaches to try to identify genes that might be regulating this response," he said. "But there hasn’t been much that has come out of the literature."

"Our approach was to use whole-genome gene expression profiling as a way to find new markers of what's underlying this variation in drug response."

In the study, Voora and his colleagues recruited two 50-subject cohorts of healthy volunteers and another set — in collaboration with George Washington University — of patients on low-dose aspirin for CHD.

The team measured gene expression using whole-genome microarrays in the first healthy cohort before, during, and after a four-week course of aspirin treatment at a dose of 325 mg, and correlated the expression data with measurements of platelet response to develop the predictive signature.

Then the researchers validated the predictive power of the signature in the second healthy cohort and in the CHD cohort, eventually settling on a set of 60 co-expressed genes that were the most highly correlated with platelet response to aspirin therapy.

According to the study authors, the group was also able to show that this "aspirin response signature" was specifically associated with platelet function on aspirin therapy and not with individual's un-treated platelet function.

The team then turned to the Catheterization Genetics (CATHGEN) biorepository to measure whether the signature also correlated with patient outcomes. CATHGEN stores samples from patients who underwent cardiac catheterization at Duke University Medical Center and who were followed to record whether they suffered heart attack or died.

"We had access to 600 [samples from patients] who had undergone gene expression profiling," Voora said. "So, we linked that data with their outcome data … as a way to link the gene expression to an adverse event that aspirin should be preventing.”

According to Voora, the aspirin response signature was predictive of death or myocardial infarction in the CATHGEN cohort independent of traditional risk factors, gender, and race, and of clinical findings during cardiac catheterization like the number of blocked blood vessels. The signature also added prognostic ability above and beyond traditional risk factors, the study authors found.

The team has applied for a patent covering the signature "in case the opportunity materializes" to develop a commercial test, according to Voora.

"We are primarily interested in the research, but I think this work clearly lays the foundation for a simple diagnostic test that could be done in the outpatient setting," he added.

"Currently doctors don’t have a good way of measuring how patients respond to aspirin in the office, in striking contrast to pretty much every other drug we use in cardiology," said Voora. "With statins we can measure cholesterol, for ACE inhibitors we measure blood pressure … but with aspirin we give it and just assume that it works."

While tests exist that measure platelet function directly, these require that samples be processed within hours of a blood draw and require specialized equipment and training, Voora said. "If you are in a practice outside a major medical center you are not going to have access" to such tests.

"The really nice thing about a gene expression-based test is that it eliminates all these technical difficulties," he added. "Blood RNA can be kept at room temperature for several days without degrading and there is equipment already available to automate extraction and PCR of the transcripts you are interested in, so there is already a really well-established workflow for these types of tests in widespread use."

A main caveat to the group's finding, according to Voora, is that it's not yet clear what doctors can do with information about whether a patient is likely responding to aspirin or not.

Beyond providing prognostic information, it would be ideal if the signature could also help inform treatment decision making, especially considering the availability of other platelet inhibiting drugs like Plavix, Eli Lilly's Effient (prasugrel) and AstraZeneca's Brilinta (ticagrelor).

"Right now doctors really don’t have a good way of identifying in whom those medications should potentially be added or substituted for aspirin. So, the goal here is to get a biomarker into physicians' hands to help them make those decisions," Voora said.

The Duke team, meanwhile, is planning to follow up the study with a wider look at the predictive ability of the new signature, under an NIH grant Voora said he was awarded last week.

"We will be exploring how the signature performs across the full range of aspirin doses, and we will also be comparing other antiplatelet agents to aspirin – to see if the signature can be affected by other drugs," said Voora.

Voora and his colleagues also aim to test the signature to see if it is linked to clinical events in different populations. "In the paper we focus on the CATHGEN cohort, people coming in for an invasive cardiac procedure who are not very reflective of the general person who takes aspirin,” he said. "So with the new grant we will be expanding to datasets of patients who are more low risk and are taking aspirin in the outpatient setting."