BNP levels have been shown in prior studies to have predictive value for atrial fibrillation. NT-proBNP is an inactive byproduct produced when cleaving BNP to its active form, and has been associated with an increased risk of morbidity and mortality in patients with and without cardiac disease. In this study, authors Roldán, et al have investigated the correlation of NT-proBNP levels with stroke, composite vascular events, and death in a population of patients with known afib, on anticoagulation (acenocoumarol), and with stable INRs (2-3, time in therapeutic range >70%).

1172 patients were included in the study with a mean age of 76. A CHA2DS2-VASc stroke risk score (median = 4) and HAS-BLED bleeding risk score was calculated at baseline, and patients were reassessed at a median follow-up of 1007 days. Primary end point was stroke or TIA, and secondary endpoints included a composite of cardiovascular events (stroke, TIA, systemic embolism, ACS, HF, cardiac death), as well as major bleeding and all-cause death.

On multivariate analysis, authors found that high NT-proBNP leves were significantly associated with poor prognosis, even after adjusting for the CHA2DS2-VASc score. A level of >822pg/nl had a HR of 2.71 for stroke; a level of >304 pg/nl had a HR of 1.85 for composite vascular events; and for all cause mortality an NT-proBNP level of >519pg/nl had a HR of 1.66. NT-proBNP levels did not predict increased bleeding risks.

Authors note limitations to this study, including the lack of a baseline echocardiogram, and that other factors such as age or renal dysfunction can affect NT-proBP levels. In this particular patient population, the big question is: how does this change management? Clearly, follow-up studies are needed. Do these results change at higher INR goals or on different oral anticoagulants? As more information in these areas come out in the future, it’s possible we may find NT-proBNP a useful tool in tailoring appropriate anticoagulation therapies to our patients.