High-sensitivity troponin-T concentrations and risk of heart failure in patients with acute pulmonary embolism
C Byrne, M Afzal, J M C Frary, F S Davidovski, T Biering Soerensen, C Torp Pedersen, K H Kragholm, D L Bhatt, M PareekAbstract
Background
High-sensitivity cardiac troponins are the gold-standard biomarkers for diagnosing myocardial injury. We have previously shown that elevated high-sensitivity troponin-T (hsTnT) is associated with both short- and long-term mortality in patients with acute pulmonary embolism (PE). However, it is unknown whether hsTnT may reflect concurrent or developing heart failure (HF) in this setting.
Purpose
To determine the early and late risk of HF in individuals with acute PE according to their hsTnT concentration at presentation.
Methods
Retrospective cohort study based on Danish nationwide administrative registries. We identified all patients discharged from the hospital with a PE from 2013 through 2020 (International Classification of Diseases, Tenth Revision diagnostic codes: I260, I269, and I269A) and merged them with all records of at least one hsTnT (Elecsys® Troponin T hs, Roche Diagnostics; 99th percentile upper reference limit, 13.5 ng/l) measurement obtained during the same hospitalization. Subjects with known HF were excluded. Persons were stratified into quartiles (Q1-Q4) based on their first hsTnT measurement. Standardized absolute and relative risks (RR) for HF at days 0-30 and days 31-365 were calculated through multivariable Cox regression with average treatment effect modeling. Models were standardized for demographic and clinical features, including hemoglobin concentration and estimated glomerular filtration rate.
Results
A total of 5,323 individuals were included. Median age (25th-75th percentile) was 70.2 (58.0-78.7) years, 2,651, (50.2%) were women, and 129 (2.4%) had a prior history of venous thromboembolism. Median baseline hsTnT was 27 (25th-75th percentile, 14-72) ng/l, and 3,996 (75.1%) had an elevated concentration. Seventy-one (1.3%) individuals had been diagnosed with HF at 30 days; 14 within the first day of admission, and 57 from days 2 through 30 (Figure 1). Only 2 HF events occurred in Q1, 16 in Q2, 26 in Q3, and 27 in Q4. Accordingly, the standardized absolute risk of early HF was negligible in Q1 (0.13%) and highest in Q4 (2.64%). The RR of developing HF within 30 days after acute PE was significantly higher in hsTnT Q3 and Q4 vs. Q1 and Q2 (RR: 3.17; 95% CI, 1.35-4.99; P=0.02). Seventy-two (1.5%) additional HF diagnoses were registered between days 31-365, with 5 events in Q1, 22 in Q2, 28 in Q3, and 17 in Q4 (Figure 2). The standardized absolute risk of HF was 0.52% in Q1, and 1.85% in Q4. The RR of HF 31-365 days after PE tended to be higher for individuals in hsTnT Q3 and Q4 vs. Q1 and Q2, but did not reach significance (RR: 1.72; 95% CI, 0.87-2.56; P=0.10).
Conclusions
The risk of HF in individuals with acute PE, without a history of HF, was very low. A large proportion of early HF cases appeared concomitantly with the PE event, and individuals with a hsTnT concentration >=28 ng/l had a significantly higher risk of HF within 30 days of their acute PE.Figure 1Figure 2