NT-proBNP in severe aortic valve stenosis: necessity for higher cut-off values regarding congestive heart failure detection in atrial fibrillation

Abstract Background NT-proBNP is an established biomarker for the determination of fluid overload. Patients suffering from atrial fibrillation seem to have higher serum levels of NT-proBNP than patients in sinus rhythm, independent from fluid overload. However, there is no established alternative cut-off for patients with atrial fibrillation, especially in patients with severe aortic valve stenosis. Aim The aim of the current study was to evaluate the levels of serum NT-proBNP for patients with atrial fibrillation (AF) and sinus rhythm (SR) for the determination of fluid overload in patients with severe aortic valve stenosis. Methods 167 Patient suffering from severe aortic valve stenosis undergoing evaluation for TAVR at our hospital were included in this study. Upon recruitment blood samples from every patient were collected and an ECG was documented for each patient. Serum NT-proBNP was measured and volume overload was objectivated by calculation of plasma volume status (PVS), an already published method to estimate fluid overload or depletion by using a formula including weight and haematocrit. Results 37 patients (22.2%) suffered from AF upon recruitment. 47 patients (28.1%) showed fluid overload, indicated by increased PVS (PVS > 0, SR n=37, AF n=10). 94 patients (56.3%) were male and mean age was 80 years. Patients with and without AF showed no significant difference in age, sex or prevalence of cardiovascular diseases, except for gradients across the aortic valve (SR 47mmHg vs. AF 38mmHg; p<0.05) and left ventricular ejection fractions (SR 57% vs. AF 55%; p<0.05). Patients suffering from atrial fibrillation showed significantly higher serum levels of NT-proBNP (4843 ng/L vs. 1226 ng/L, p<0.001). ROC-analyses in patients with SR and severe aortic valve stenosis, showed satisfying results for NT-proBNP (AUC 0.72; sensitivity 0.81; specificity 0.58) to detect fluid overload. Comparable results were evident in patients with AF (AUC 0.77; sensitivity 0.70; specificity 0.78), but cut-off-values in patients suffering from atrial fibrillation were more than six times as much as in patients with sinus rhythm (6575 ng/L vs. 1053 ng/L). Conclusion NT-proBNP proved itself as a sufficient marker to determine volume overload in patients with severe aortic valve stenosis. Cut-offs for accurate determination of fluid overload deviate greatly in patients with sinus rhythm and atrial fibrillation, therefore different cut-offs might be required.