Inflammation Drives Coagulopathies in Sars-Cov-2 Patients

Background:A hypercoagulable state has been consistently reported in patients with severe Coronavirus Disease 2019 (COVID-19), caused by Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), characterized by elevated D-dimer, prolonged PT, and mild thrombocytopenia, though the mechanism is unclear. We have previously shown that human immunodeficiency virus (HIV) infection causes depletion of the anticoagulant protein S and virus-mediated platelet activation. Based on early reports, we hypothesized that a similar process contributed to COVID-19-associated thrombosis.

Aim:To probe platelet activation and coagulation factor activity in SARS-CoV-2-infected patients.

Methods:Blood was collected from consenting patients with differing COVID-19 severity: outpatients (15), hospitalized inpatients (15), and healthy controls (8). Platelet-leukocyte aggregate (PLA) formation and monocyte profiling were measured by flow cytometry. Coagulation factors were assessed by enzymatic assays. PS, von Willebrand Factor (vWF), PC, cytokines, and anti-S-Protein (viral spike protein) IgG were measured by ELISAs.

Results:Ninety percent of SARS-CoV-2⁺ out-patients and in-patients had circulating anti-S-Protein IgG, but plasma IL-6 and TNFα were only elevated in three in-patients, consistent with reports that systemic inflammation is relatively rare in this population. Immune response did not correlate with disease severity. Unlike in HIV1+/AIDS patients, total PS was not reduced in SARS-CoV-2+ patients. However, the anticoagulant pool of PS ("free PS") was reduced in plasma samples from in-patients compared to controls (47.2%±23.3% vs. 100.8±42.6%, p=0057), while out-patients had an intermediate concentration (73.1%±28.9%). Specific loss of free PS is likely mediated by an increase in C4-binding protein (C4bp), which binds PS. In-patients also had a trend toward elevated plasma tissue factor (TF) compared to controls (79.5±121.4 fM vs. 37.8±39.7 fM, p = 0.32). Endothelial cells and monocytes can express TF under inflammatory conditions. We evaluated endothelial damage and dysfunction by measuring E-Selectin, which was unchanged in either in-patients or out-patients, and von Willebrand Factor (vWF), which was elevated in in-patients compared to controls (143±29.8 ng/mL vs. 56.2±41.9 ng/mL, p=0.0023). Plasma from in-patients also had elevated myeloperoxidase (524±187 ng/mL vs. 127±35 ng/mL, p=0.0026) and had a trend toward increased platelet-leukocyte aggregates (14.6±11.7% vs. 5.2±3.7%, p=0.24), indicating platelet and leukocyte stimulation. Unlike in the HIV1+/AIDS patients, no virus was detectable in any of the SARS-CoV-2+ patient plasmas. Consistent with a lack of direct platelet-virus interaction, plasma PF4 and platelet Akt phosphorylation were unchanged in the patient samples. We also observed a trend toward increased TF on TF+/CD64+/CD11b+ monocytes from in-patients compared to controls (MFI = 3244±2340 vs. 1741±382, p=0.18). Two inpatients were followed until they were SARS-CoV-2-negative. In both, PLAs, IL-6, vWF, and plasma TF remained elevated out to 28 days and PS remained reduced, suggesting that hemostatic dysregulation persists after SARS-CoV-2 is undetectable.

Conclusions:We propose that localized inflammation in SARS-CoV-2+ patients results in a decrease in anticoagulant PS, through a shift of the free and C4bp-bound forms. At the same time, this inflammation causes stimulation of endothelial cells, which secrete procoagulant vWF, monocytes, which express TF and release it into plasma on microvesicles, and platelets, which form platelet-leukocyte aggregates. These changes may not return to baseline post-infection, suggesting that long-term monitoring of thrombotic risk may be necessary for SARS-CoV-2+ patients.