Aims: Low levels of brain natriuretic peptides on admission identify low-risk patients with acute pulmonary embolism (APE) through their high NPV for mortality. However, increased natriuretic peptide values on admission are less helpful for identifying high-risk patients due to their low PPV. The aim of the study was to test whether the PPV for mortality can be improved by performing serial NT-proBNP measurements on admission, at 12 h, and at 24 h.
Methods and results: We prospectively included 113 consecutive patients with APE (mean age 63+/-18 years), of whom 10 had clinically massive APE. Thirty-day mortality was 15% (95% CI: 8%-22%). In survivors, median NT-proBNP levels decreased within 24 h from 1895 ng/L (range: 16-33,340) to 1007 ng/L (range: 9-33,243) (p<0.001). In non-survivors, median NT-proBNP levels at baseline (11,491 ng/L, range: 618-60,958) remained elevated at 24 h (8139 ng/L, range: 35-70,018; p=NS). The 30-day mortality rate in the group of 18 patients with NT-proBNP >7500 ng/L and less than 50% decrease of NT-proBNP within 24 h was 61% (95% CI: 39%-84%). PPV and NPV of NT-proBNP >7500 ng/L on admission and less than 50% decrease of NT-proBNP within 24 h were 61% and 94%, respectively.
Conclusion: Persistent elevation of plasma NT-proBNP levels within 24 h after APE diagnosis indicates ongoing right ventricular dysfunction with a poor prognosis. These critically ill patients may be candidates for rapid aggressive intervention, including thrombolysis, catheter thrombectomy, or surgical embolectomy.