Dynamics of troponins and 30-day mortality in hospitalized patients with pulmonary embolism

Abstract

Background

In patients with pulmonary embolism (PE), the impact of repeated troponin I or T (TnI/TnT) measurements remains unclear.

Methods

Using Danish national registries, we identified PE patients (≥18 years) hospitalized between 2013 and 2018 with initial TnI or TnT measurement within −1/+1 day from admission and >1 repeated measurement within three days. Trajectories of TnI and TnT were identified using latent class trajectory modeling. Hazard ratios for 30-day mortality were compared across trajectories via multivariable Cox regression.

Results

Among 1539 patients with TnI measurements and 1323 with TnT measurements, three distinct trajectories were identified. Trajectory I (n_TnI = 286, n_TnT = 472) exhibited consistently low TnI/TnT concentrations, trajectory II (n_TnI = 1076, n_TnT = 724) demonstrated initial elevated TnI/TnT decreasing within 24 h, and trajectory III (n_TnI = 177, n_TnT = 127) was characterized by elevated index TnI/TnT increasing within 10 h. 30-day mortality rates were higher in trajectory II and III compared to I in both the TnI (3 %, 7 % and 18 % across trajectory I to III) and the TnT (1 %, 9 % and 20 % across trajectory I to III) cohort. After adjustment hazard ratio of 30-day mortality for trajectory II vs. I was 7.42 (95 % CI 1.00–54.84, p = 0.04, TnI) and 2.93 (95 % CI 1.17–7.33, p = 0.02 TnT); and for trajectory III vs. I, 16.42 (95 % CI 2.42–127.29, p = 0.007, TnI) and 8.21 (95 % CI 2.78–24.19, p < 0.001, TnT).

Conclusion

A steep increase in TnI or TnT concentration within 10 h of PE diagnosis significantly escalates 30-day mortality risk indicating that early serial sampling may enhance risk stratification of PE patients.