Effects of hemoglobin concentration and creatinine clearance in pro-B-type natriuretic peptide-based left ventricular filling pressure prediction in patients with preserved left ventricular systolic function.

Abstract

Although N-terminal pro-B-type natriuretic peptide (pro-BNP) has been shown to correlate with left ventricular (LV) filling pressure, pro-BNP-based LV filling pressure prediction has some limitations due to several factors that affect pro-BNP. The aim of this study was to evaluate the effects of blood hemoglobin concentration and creatinine clearance (CCr) on pro-BNP-based LV filling pressure prediction in patients with preserved LV systolic function. A total of 421 consecutive patients referred for coronary angiography underwent LV pressure measurement by fluid-filled catheters. Patients with plasma creatinine levels>or=1.5 mg/dl and LV ejection fractions<50% were excluded. LV diastolic pressures and echocardiographic parameters were compared with pro-BNP levels in 281 patients. Blood hemoglobin levels and CCr were measured simultaneously. Log pro-BNP was independently correlated with hemoglobin concentration (beta=-0.261, p<0.001), CCr (beta=-0.230, p<0.001) and LV pre-A-wave pressure (beta=0.384, p<0.001). A pro-BNP level of 124 pg/ml was the optimal cutoff for LV pre-A-wave pressure>15 mm Hg in all patients (sensitivity 67%, specificity 67%, p<0.001). However, by subclassification according to hemoglobin and CCr tertiles, optimal cut-off values varied significantly, and their predictive accuracies could be improved (from 89 to 331 pg/dl, with diagnostic accuracy up to 79%). In conclusion, in the pro-BNP-based prediction of elevated LV filling pressure, subclassification on the basis of hemoglobin concentration and CCr should be considered in patients with preserved LV systolic and renal function.