Combined aspartate aminotransferase level and PE-SARD score predict 1-month bleeding risk in acute pulmonary embolism

Background: Anti-coagulation is the cornerstone management of acute pulmonary embolism (PE), which is a double-edged sword, as it increases the risk of bleeding. Thus, predicting bleeding risk is necessary. The liver produces most coagulation factors to maintain the coagulation balance. However, the association between liver dysfunction markers and bleeding risk has not been thoroughly investigated.Methods: A single-center, retrospective analysis of patients with acute PE was performed. First, the authors studied the association between liver dysfunction indexes and the 1-month bleeding risk. Then, they investigated whether it is more effective to predict the bleeding risk using a new joint model, i.e., adding liver dysfunction indexes to the PE-SARD score, which is the first score to assess the bleeding risk of acute PE.Results: Among 469 patients with acute PE, 34 patients (7.2%) had bleeding events within 1 month after the onset. The levels of aspartate aminotransferase (AST) were higher in the bleeding group compared with the non-bleeding group (36.0 [18.25-90.0] vs. 23.0 [18.0-31.0], p = 0.008). Compared with AST<40, the odds ratios of 80=AST<120 and AST=120 were significant (8.825 [2.449-31.804] and 8.023 [2.543-25.315] respectively, p<0.01), even when adjusted for nine confounding factors (p<0.05). The area under the curve of PE-SARD combined with AST was significantly higher than that of the PE-SARD score (p = 0.02).Conclusions: In PE patients, AST is an independent factor in predicting the 1-month bleeding risk, and a novel joint model that combines AST and PE-SARD score improved the predictive efficiency for the 1-month bleeding risk.