It has long been understood that patients with atrial fibrillation are much more likely to experience heart failure than members of the general population. However, tools to predict which of these patients are at greatest risk are not yet available.
Now, a multicenter team led by investigators from Brigham and Women’s Hospital has used patient data compiled from three large clinical trials to identify potential prognostic biomarkers that could be used to determine patients’ levels of risk.
“There is an interest in risk-stratifying patients with atrial fibrillation according to their heart failure risk,” says Paul Haller, MD, PhD, a research fellow in the Thrombolysis in Myocardial Infarction (TIMI) Study Group at the Brigham. “If we can figure out a way to determine who is at highest risk, this may open up an avenue to identify patients who might benefit from early preventive treatment.”
Dr. Haller is the first author of the study, which was published online in August 2024 in the Journal of the American College of Cardiology. He also presented the research at the recent European Society of Cardiology Congress 2024.
A Comprehensive Database of Patients With Atrial Fibrillation
The data used in the study came from three randomized clinical trials that compared various non-vitamin K antagonist oral anticoagulants with warfarin for the prevention of stroke and other systemic embolism events in patients with atrial fibrillation. These were the ARISTOTLE, ENGAGE AF-TIMI 48, and RE-LY trials.
The pooled data are part of the COMBINE AF cohort, which was made possible through the efforts of the trial’s lead investigators, including Brigham cardiologist Robert P. Giugliano, MD, ScM. COMBINE AF is one of the largest and most comprehensive databases of its kind, containing detailed information on more than 71,000 patients with atrial fibrillation treated with an anticoagulant.
“These data give us a valuable assessment of patients’ blood levels of important biomarkers in a large subset of patients at the time they were enrolled in one of these trials,” Dr. Haller says. “They also give us systematic follow-up data on the patients—most importantly, what the outcomes were.”
The composite endpoint for the study was hospitalization for heart failure or cardiovascular death. Secondary endpoints were heart failure–related hospitalization and death.
Three Biomarkers Linked to Increased Heart Failure Risk
Using these data, the investigators report detailed information on three biomarkers that are linked to increased heart failure risk:
- N-terminal pro-B-type natriuretic peptide (NTproBNP), which is commonly used in the diagnostic management of heart failure but is more challenging to interpret in patients with atrial fibrillation
- High-sensitivity cardiac troponin T (hs-cTnT), which is commonly used to detect damage to heart muscle—in particular, for the diagnosis of heart attack
- Growth-differentiation factor 15 (GDF-15), which is a biomarker of stress and inflammation
Using weighted quantile sum regression analysis, a novel statistical method to weigh the importance of predictors, the researchers determined that the contribution to risk assessment was similar for NTproBNP and hs-cTnT for cardiovascular death or heart failure–related hospitalization (38% and 41%, respectively) in patients with atrial fibrillation. GDF-15 provided a statistically significant but lesser contribution to risk assessment.
“GDF-15 can be considered as a general measure of inflammation and is only used for research purposes in the United States,” Dr. Haller explains. “But doctors here are quite familiar with NTproBNP and hs-cTnT, although hs-cTnT is used in a different clinical context. The important finding this paper suggests is that a dual-marker strategy that includes both of these biomarkers could be of greater value for predicting heart-failure events in patients with atrial fibrillation than only considering NTproBNP.”
Next Steps for Validating Biomarkers’ Value in Patient Management
The investigators acknowledge that prospective trials are needed to understand whether preventive treatments started on the basis of such biomarker testing are warranted.
“Historically, it’s been tough for biomarkers to be implemented into clinical routine because of the lack of clinical trials confirming their utility,” Dr. Haller says. “Ideally, we would have a trial where we apply a biomarker strategy randomly to a group of patients and compare it with standard of care.”
Dr. Haller adds that because of the costs associated with both measuring biomarkers and managing patients with medical interventions, it will also be important to determine whether the benefits outweigh both the costs and potential risks. However, the availability of two of these three biomarkers in clinical routine presents a favorable opportunity for implementation if supported by further evidence.