Elevated tissue factor pathway inhibitor delays thrombin generation in COVID-19 but is not associated with clinical outcomes

Abstract

Plasma levels of tissue factor pathway inhibitor (TFPI) are elevated in many patients with COVID-19 but the role of TFPI in COVID-19 coagulopathy remains elusive. We sought to determine the contribution of TFPI to thrombin generation in patients with COVID-19 and assess its association with thrombosis and other clinical outcomes. We used blood samples from an early COVID-19 clinical trial of adult patients hospitalized with acute COVID-19 from April 2020 to January 2021 (ClinicalTrials.gov identifier: NCT04360824). Plasma TFPI was measured by enzyme-linked immunosorbent assay, and thrombin generation potential was measured in the presence or absence of TFPI neutralizing antibodies. Thromboelastography was performed with whole-blood samples. We found that plasma TFPI was elevated in patients with COVID-19 compared with healthy individuals. Thrombin generation triggered by exogenous TF and phospholipids was increased in COVID-19, reflected by greater peak thrombin, velocity index, and endogenous thrombin potential; however, the time to initiation of thrombin generation (lag time) was delayed. Addition of a neutralizing anti-TFPI antibody significantly shortened the lag time in COVID-19 and normalized the difference in lag time between those with COVID-19 and healthy individuals. Plasma TFPI was positively associated with lag time, time to peak thrombin, and time to initial clot formation in thromboelastography. Multivariate analysis demonstrated that TFPI correlated with lag time and time to reach peak thrombin but not with 30-day mortality, thrombosis, or other adverse clinical outcomes. We conclude that elevated plasma TFPI delays the initiation of thrombin generation and clot formation but is not associated with thrombosis in patients hospitalized with COVID-19.

Graphical abstract

Figure 1.

Baseline TFPI concentration is associated with delayed initiation of thrombin generation in COVID-19 plasma. (A) PFP from patients with COVID-19 (red squares) and healthy controls (blue circles) was treated with heparinase, and thrombin generation was triggered by 1 pM TF and phospholipids. Data were collected for lag time, T_max, thrombin peak, velocity index, and ETP. Representative thrombin generation curves are presented, and the table shows comparisons of thrombin generation parameters between controls (n = 21) and patients with COVID-19 (n = 40). Data showing normal distribution are presented as mean ± standard error and were analyzed using the unpaired t test. Data not showing normal distribution are presented as median (interquartile range) and were analyzed using the Mann-Whitney U test. (B) Total TFPI levels in plasma from healthy controls (blue circles; n = 22) and patients with COVID-19 (red squares; n = 62) were measured by ELISA. Data are shown as fold change from the mean value of healthy controls, presented as mean ± standard error, and analyzed using the unpaired t test. ∗∗∗∗P < .0001. (C-G) Simple linear regression analysis of thrombin generation potential parameters and TFPI in COVID-19 plasma (n = 31).

Figure 2.

TFPI prolongs the initiation of thrombin generation in plasma from patients with COVID-19. PFP from controls (n = 5) or patients with COVID-19 (n = 8) were first treated with heparinase and then incubated with either a neutralizing anti-TFPI antibody or control immunoglobulin G (IgG) for 30 minutes before triggering thrombin generation with TF and phospholipids. (A) Representative thrombin generation curves are presented, and the table shows comparisons of thrombin generation parameters between paired samples treated with anti-TFPI antibody vs control IgG. (B-D) Individual data points for lag time (B), T_max (C), and ETP (D) are presented as mean ± standard error for all except the lag time for controls are presented as median (interquartile range). Data presented as mean ± standard error were analyzed using the paired t test, and those with median (interquartile range) were analyzed using the Wilcoxon matched-pairs signed-rank test. ns, P > .05; ∗P < .05; ∗∗∗∗P < .0001. ns, not significant.

Figure 3.

Multivariate analysis. A heat map showing multivariate analysis with Pearson correlation between plasma levels of TFPI, parameters of thrombin generation, demographic factors, and clinical outcomes. n = 81. Refer to supplemental Table 1 for R values and P values. BMI, body mass index; ICU, intensive care unit.

Figure 4.

TFPI is associated with initial clot formation in whole-blood TEG in COVID-19. (A-E) Simple linear regression analysis of plasma total TPFI vs whole-blood TEG parameters in COVID-19 (n = 60). (F) Simple linear regression analysis of TEG initial clot formation vs thrombin generation lag time (n = 38). (G) Plasma concentration of TF/annexin V double–positive EVs in healthy controls (n = 9) and patients with COVID-19 (n = 10). Data are presented as mean ± standard error and were analyzed using the unpaired t test. ∗∗∗∗P < .0001. max, maximum.

Figure 5.

Plasma TFPIα is elevated in COVID-19. The isoform distribution of plasma TFPI was determined by measuring total TFPI (A) and TFPIα (B) by ELISA using a monoclonal anti-TFPI-K2 antibody for capture and biotinylated monoclonal anti-TFPI-K1 (total TFPI) or anti-TFPI-K3 (TFPIα) antibodies in controls (n = 29) and patients with COVID-19 (n = 44). Data are presented as median (interquartile range) and were analyzed using the Mann-Whitney U test.

Figure 6.

Summary of effects of elevated plasma TFPI on thrombin generation in patients with COVID-19. COVID-19 enhances thrombin generation via multiple mechanisms, including EVs, histones, NETs, and aPL. Elevated plasma TFPI prolongs the lag time to thrombin generation but does not diminish peak thrombin. aPL, antiphospholipid antibody; FXa, factor Xa; FVIIa, factor VIIa; NET, neutrophil extracellular trap.