Abstract
AIMS: Pulmonary embolism (PE) is a common and potentially life-threatening condition requiring emergent diagnostic work-up. Despite wide use of biomarkers, little is known how they predict long-term prognosis of patients evaluated for suspected PE. METHODS AND RESULTS: We sought to assess the predictive performance of N-terminal pro-brain natriuretic peptide (NT-proBNP), C-reactive protein, fibrin D-dimer (FIDD), and cardiac troponin T (cTnT) in patients who underwent computed tomography pulmonary angiography (CTPA) for clinical suspicion of PE. The analysis involved 1001 patients, with 222 (22.2%) receiving a PE diagnosis at index imaging. Mean ages of patients with and without PE were 65.0 ± 17.1 and 64.5 ± 17.7 years, respectively. Median follow-up time was 3.9 years (interquartile range 2.9-4.9). Mortality was relatively high among both patients with and without documented PE (24.8% vs. 31.7%, P = 0.047). In patients with PE, only elevated NT-proBNP > 1000 ng/L and C-reactive protein > 50 mg/L levels at hospital admission were associated with higher mortality in an adjusted Cox regression model, but receiver operating characteristic (ROC) analysis showed no improved prediction compared to clinical variables. Among patients without PE, elevated NT-proBNP > 1000 ng/L, C-reactive protein > 10 mg/L, cTnT > 50 ng/L, and FIDD > 1.0 mg/L all predicted mortality. In an ROC analysis among patients without PE, models including NT-proBNP, cTnT, or C-reactive protein provided improved predictive performance. CONCLUSION: Patients evaluated for clinical suspicion of PE have high long-term mortality. Commonly used biomarkers provide long-term prognostic value in patients without PE. Given the relatively young age, it is vital to identify these high-risk patients and perform differential diagnosis work-up for alternative life-threatening conditions, and manage them as appropriate.