Background:
With mortality rates of 15–40% despite optimal therapy, infective endocarditis (IE) remains one of the deadliest infectious diseases. Gram-positive bacteria are the leading pathogens, and the aortic valve is most frequently affected. The early pathophysiology critically depends on a platelet–fibrin scaffold that stabilizes bacterial adhesion and vegetation formation, positioning thrombin (factor II) as a central node linking coagulation and bacterial colonization. Whether anticoagulant classes are functionally interchangeable with respect to IE susceptibility remains unclear.
Hypothesis:
We hypothesize that direct factor II (thrombin) inhibition reduces susceptibility to IE, an effect that cannot be reproduced by upstream factor Xa inhibition.
Methods:
In vivo, IE was induced in our established ultrasound-guided murine model by intravenous injection of S. epidermidis or S. aureus 72 hours after wire-induced endothelial injury of the aortic valve. Mice were pretreated with the direct factor II inhibitor Dabigatran, the factor Xa inhibitor Apixaban, or vehicle. Vegetation size (mm²) was quantified histologically (Gram staining) and by echocardiography. In vitro, primary human valvular interstitial cell (VIC)–coated ibidi µ-slides were perfused with human whole blood inoculated with S. aureus under defined shear stress, with or without the direct thrombin inhibitor hirudin, and surface-adherent bacteria were quantified by image-based analysis.
Results:
In S. epidermidis IE, Dabigatran significantly reduced valvular vegetation size at 1 day, from 3693 mm² (control) to 2519 mm² (mean difference 1175; 95% CI 494–1856; p=0.0001), whereas the factor Xa inhibitor Apixaban had no effect (3571 mm²; difference 122; p=0.97). In the more virulent S. aureus IE, Dabigatran again markedly reduced vegetation size (mean difference vs control 2735 mm²; 95% CI 925–4544; p<0.0001), whereas Apixaban showed no significant effect (difference −38 mm²; p>0.99). Consistent with these in vivo findings, in vitro functional thrombin depletion (hirudin) reduced relative bacterial adhesion in human whole blood from 99.5 ± 3.4 to 61.8 ± 9.4 (mean ± SD; n=6 per group; Mann–Whitney p=0.0022), confirming a thrombin-dependent contribution to bacterial adhesion.
Conclusion:
Factor II inhibition confers protection against IE in both S. epidermidis and S. aureus murine models and reduces bacterial adhesion under defined flow conditions, whereas upstream factor Xa inhibition does not. These findings support a thrombin/fibrin-specific, host-protective mechanism and suggest a clinically relevant, class-dependent difference among oral anticoagulants. Population-based registry data should be analyzed to confirm this hypothesis.
