Predictive Value for Increased Factor XIa and Plasma Kallikrein Activity in Acute Venous Thromboembolism

Venous thromboembolism (VTE) is associated with increased coagulation activity, which in part can be attributed to the contact pathway of coagulation. Evidence from pre-clinical and epidemiological studies suggests that deficiency in factors of contact activation (e.g. coagulation factors (F) XI and FXII) protects against VTE. However, limited information exists regarding the activation of the contact system in the setting of acute VTE.

In the current study, patients with confirmed VTE events (n=321) from the VTEval study and controls (n=300) from the population-based PREVENT-it pilot study were included. Plasma samples were collected from patients after confirmed VTE events or controls upon inclusion in the study. FXI as well as FXIa and plasma kallikrein (PKa) levels were assessed in plasma samples from all subjects using an activated thromboplastin time-based assay (FXI:c), a thrombin generation-based assay (CAT:FXIa) and by measuring inhibitory complexes (FXIa:antithrombin (AT), FXIa:alpha-1-antitrypsin (α1AT), FXIa:C1 esterase inhibitor (C1Inh) and PKa:C1Inh) using enzyme-linked immunoassay (ELISA). After a 2-year follow up period, a composite endpoint of recurrent VTE or death was determined.

Increased FXI:c levels were determined in VTE patients compared to control individuals (124.08 ± 37.48% vs. 113.55 ± 27.99%), whereas CAT:FXIa levels were reduced in VTE patients (0 pM [IQR, 0-0.56] vs 0.56 pM [IQR, 0-0.88]). Levels of FXIa:α1AT and FXIa:AT inhibitory complexes were increased in VTE patients compared to controls (median[IQR]; 311.8 pM [238.2-424.0] vs. 202.5 pM [143.7 - 287.5] and 29.1 pM [23.4-38.3] vs 23.2 pM [19.7-29.8], respectively). Considering that 86% of the VTE patients were already on anticoagulant treatment (Table 1), investigation of their possible effect on the biomarkers revealed that only the CAT:FXIa was influenced by the presence of anticoagulants. Logistic regression models revealed a good discriminatory value for FXI:c and FXIa:α1AT (AUC=0.64 [0.6/0.69] and AUC=0.67 [0.62/0.71], respectively) to distinguish VTE from controls, whereas the other biomarkers were not able to distinguish between groups. The outcome recurrent VTE or death could be predicted by the inhibitory complexes, but not by the FXI(a) levels (Figure 1). Only the FXIa:α1AT complexes were able to both detect the presence of VTE (OR per SD [95%CI]: 1.28 [1.01-1.63], p=0.04) and predict recurrent VTE or death (HR per SD [95%CI]: 1.40 [1.2-1.62], p<0.0001).

In summary, acute VTE is associated with both elevated FXI:c levels and increased activation of FXI and plasma prekallikrein, the latter specifically indicating contact activation. The generation of FXIa during acute VTE and its association with recurrent VTE suggests an important risk contribution of FXI activation. This study has added evidence favouring the utility of FXIa inhibition in the setting of acute VTE.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal