Circulating biomarkers are associated with atrial fibrillation burden in patients with paroxysmal atrial fibrillation

Abstract Background Biochemical markers of inflammation, coagulation, and myocardial stress have been associated with prevalent and incident atrial fibrillation (AF) but their association with AF burden is not known. Purpose To investigate the association of blood biomarkers with AF burden and -recurrence status after one year of continuous rhythm monitoring. Methods We included patients with paroxysmal AF from The Reappraisal of Atrial Fibrillation: Interaction between HyperCoagulability, Electrical Remodelling, and Vascular Destabilisation in the Progression of AF (RACE V) study. To enable continuous rhythm monitoring, patients underwent implantation of a Reveal LINQ implantable loop recorder (Medtronic, USA) at baseline. Peripheral blood samples were collected from all patients. We evaluated a panel of 92 candidate biomarkers by multiplex immunoassays using the Olink Cardiovascular III panel (Uppsala, Sweden). Outcomes assessed in this study were AF burden and -recurrence after one year of monitoring. AF burden was defined as the total percentage of time spent in AF over one year, and AF recurrence was defined as ≥0.1% time in AF. Logistic regression with and without adjustment for preselected covariates was applied to estimate the association between each biomarker and AF recurrence. For AF burden, we used a negative binomial regression model to account for overdispersion of zeros in the AF data. Following covariates were included in the multivariable analyses: age, sex, history of heart failure, hypertension, renal insufficiency, diabetes, coronary artery disease, heart failure, body mass index and use of antiarrhythmic drugs. Results Blood samples from 404 patients from the RACE V study were analysed. A total of 90 out of 92 biomarkers displayed differential expression with cumulative AF burden in the unadjusted analyses, but only three remained statistically significant after multivariable adjustment for classic cardiovascular risk markers: NT-proBNP, spondin-1 (SPON-1), and pulmonary surfactant-associated protein D (PSP-D) (Figure 1 and 2). For AF recurrence, we identified four candidate biomarkers in the multivariable analyses: NT-proBNP, myeloblastin (PRTN-3), elafin (PI-3), and azurocidin (AZU-1) (Figure 1). Increased levels of NT-pro-BNP was the strongest predictor of both outcomes (incidence rate ratio 11.9, 95% CI 5.7-24.7, for cumulative AF burden, odds ratio 4.7, 95% CI 1.8-12.1, for AF recurrence). Conclusions In patients with paroxysmal AF, NT-proBNP was the strongest and most consistent biochemical marker of AF burden and -recurrence. The clinical utility of NT-proBNP for AF monitoring as well as its predictive ability in terms of disease burden needs to be investigated in larger and prospective studies.