Limits in using brain natriuretic peptide (BNP) as a biomarker of acute right ventricular dysfunction in pulmonary embolism

Background. Risk stratification and an appropriate therapeutic approach could be lifesaving in acute pulmonary embolism (PE). Echocardiographic (ECHO) acute right ventricular dysfunction (RVD) is the actual “gold standard” in risk evaluation of PE. We previously demonstrated that plasma BNP levels were significantly higher in patients with PE and acute RVD on ECHO vs. patients with normal RV function on ECHO. Aim. Evaluation of the limits of plasma BNP in signalling acute RVD in PE. Methods. 70 patients with PE were prospectively investigated: 42(60.0%) men, mean ± SD(standard deviation) age 52.51±8.82. BNP was measured on admission using a quantitative fluorescence immunoassay (TriageBNP). ECHO evaluation of the RV function was performed in the first hour after admission. Study protocol was approved by local Ethical Committee. Patients were divided into two groups: group 1-with acute RVD on ECHO, n=24(34.3%) patients; group 2 - without acute RVD on ECHO, n=46(65.7%). Patients from group 1 were further divided into two subgroups: subgroup 1A-admitted in 12 hours after the onset of PE symptoms, n=12(50.0%) patients. Results. BNP proved good in discriminating between patients with and without acute RVD (AUC=0.88, P<0.0001). The cut-off level of plasma BNP=50 pg/mL showed the best sensitivity=0.86 and specificity=0.82 in identifying acute RVD. BNP levels were significantly lower in subgroup 1A (admitted soon) compared to subgroup 1B (admitted later than 12 hours): medians 45.25 pg/mL vs. 344.50 pg/mL, P<0.0001. Eight patients from subgroup 1A, all admitted soon after the onset of their PE symptoms, and all experiencing at least one syncopal episode showed BNP under the cut-off level. In subgroup 1A BNP did not correlate with RV end-diastolic diameter (R=0.23, P=NS), while in subgroup 1B BNP and RV end-diastolic diameter showed a consistent positive correlation (R=0.91, P<0.0001). In subgroup 1A BNP correlated significantly, but negatively, with RV systolic