von Willebrand factor/ADAMTS13 axis and venous thromboembolism in moderate‐to‐severe COVID‐19 patients

Abstract

Patients with severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2)- have been reported to develop endothelium injury, an excessive inflammatory response and subsequent marked hypercoagulability, leading to a high prevalence of venous thromboembolic events (VTE), up to c. 50% in severe cases.1, 2 von Willebrand factor (VWF) is known as a major cause of thrombo-inflammation via mechanisms including endothelial activation, enhanced VWF secretion, assembly of hyper adhesive VWF multimers, defective VWF cleavage by its specific protease a-disintegrin-like-metalloprotease-with-a-thrombospondin-type-1-motif-member 13 (ADAMTS13) and adhesion/deposition of VWF-platelet thrombi in the vasculature.2, 3 Platelet or ultra large VWF multimer complexes have been shown to promote pulmonary vascular microthrombosis, contributing to respiratory failure.4 Interestingly, previous studies have shown the involvement of the VWF/ADAMTS13 axis in the pathological the VTE process,3, 5 thus supporting the view that imbalance in this axis may lead to subsequent thrombi, in both small and large vessels.4 Surprisingly, the VWF/ADAMTS13 axis has been poorly investigated in coronavirus disease-2019 (COVID-19) patients. Only limited series with up to 88 patients have been published.6-12 Therefore, we conducted an exploratory study in moderate-to-severe COVID-19 patients managed in hospital to analyse the implication of the VWF/ADAMTS13 axis in VTE (i.e., deep venous thrombosis (DVT) or pulmonary embolisms (PE)). We conducted a single-centre, prospective study including all consecutive COVID-19 adult patients admitted from 17 March 2020 to 11 April 2020 to our intensive care unit (referred as ‘critically ill patients’) and medicine ward (referred as ‘non-critically ill patients’). This study was part of the ICU-COVID and French-COVID cohort registries approved by our institutional ethics committee (IDRCB 2020-A00256-33; CPP 11-202020.02.04.68737). SARS-CoV-2 infection was diagnosed on admission using reverse transcription-polymerase chain reaction (Cobas-SARS-CoV-2 Kits, Roche, France). To diagnose lower extremity DVT, duplex ultrasound was performed weekly in critically ill patients13 and upon clinical suspicion in non-critically ill patients. If suspected, PE was confirmed using computed-tomography/angiography. Laboratory parameters were measured within 48 h post-admission. Regarding patients transferred from the medical ward to the intensive care unit (ICU), only data from ICU blood sampling was included in the analysis and those patients were considered as critically ill. ADAMTS13 activity was measured using an in-house fluorescence resonance energy transfer (FRETS)- VWF73 assay with the recombinant VWF73 peptide (Peptide Institute, Osaka, Japan).14 VWF antigen (VWF:Ag) was measured using enzyme-linked immunosorbent assay (Asserachrom®, Stago, France). Quantitative variables are expressed as medians (25th–75th centiles) and categorical variables as percentages. Pearson’s correlation tests were used to assess the relationships between quantitative parameters. Comparisons were performed using Mann-Whitney U and Fisher’s exact tests, as appropriate. P-values <0·05 were considered significant. One-hundred-and-thirty-three consecutive COVID-19 patients (median age 65 years [56–75]; male/female ratio, 2·7), including 68 critically ill and 65 non-critically ill, were recruited. Clinical and laboratory parameters are summarised in Table I. Overall, VTE occurred in 38 patients including isolated DVT in 24 patients (63%; 13 distal and 11 proximal), isolated PE in nine patients (24%) and both PE/DVT in five patients (13%). Patients with VTE (n = 38) Patients without VTE (n = 95) D-Dimer was significantly higher [6010 ng/ml (3022–13 752) vs. 1630 ng/ml (800–2980), P < 0·0001], whereas ADAMTS13 activity was significantly lower in VTE versus non-VTE patients [59·0 IU/l (38·8–70·5) vs. 68·5 IU/l (52·0–87·5), P = 0·005] Fig 1. This association remained significant in both the ICU [50·0 (37·5–69·5) vs. 62·5 (49·3–74·0), P < 0·001] and medical ward patients [63·0 (44·0–74·3) vs. 74·0 (55·5–94·5), P = 0·002]. Eighteen VTE patients (47%) vs. 17 non-VTE patients (18%) had ADAMTS13 activity below the normal range (50–150 IU/dl). Noteworthy, none of the patients exhibited ADAMTS13 activity of <10 IU/dl, thus excluding a thrombotic thrombocytopenic purpura (TTP) diagnosis. All patients had strikingly high VWF:Ag levels that were significantly higher in VTE versus non-VTE [522 IU/dl (411–672) vs. 473 IU/dl (311–589) respectively, P = 0·05]. Death occurred in 23 patients (17%). D-Dimer was significantly higher [3370 ng/ml (2090–7515) vs. 1900 ng/ml (890–3890), P = 0·03], whereas ADAMTS13 activity was significantly lower [50·0 IU/dl (32·8–55·5) vs. 69·0 IU/dl (53·0–88·0), P < 0·0001] in non-survivors versus survivors Fig 1. No association was observed between death and VWF:Ag level. We found negative but weak relationships with substantial inter-individual variability between ADAMTS13 activity and D-Dimer (r = −0·272, P = 0·006) as well as ADAMTS13 activity and VWF:Ag levels (r = −0·183, P = 0·04). In this exploratory study, we demonstrated that ADAMTS13 activity and VWF:Ag levels were significantly associated with VTE onset. Moreover, ADAMTS13 activity was significantly associated with survival. In SARS-CoV and Middle East respiratory syndrome-coronavirus (MERS-CoV) infections, endotheliopathy-associated vascular microthrombotic disease has been described.4 SARS-CoV-2 is known to provoke an excessive local and systemic inflammatory response-denominated cytokine storm in the most severe cases.1 Microthrombi and thrombi were also reported in COVID-19-related lung histopathological findings in addition to diffuse alveolar damage and inflammatory changes.15 Interestingly, VWF expression was shown to be more important in the endothelial cells of lung small vessels with upregulated transcription, under hypoxic conditions.3 Moreover, VWF/ADAMTS13 axis involvement in thrombi formation in microcirculation was established, as illustrated in TTP.3 Together, these observations suggest a potential role of VWF/ADAMTS13 in COVID-19. In our patients, we found a markedly imbalanced VWF/ADMTS13 axis with extremely high circulating VWF concentrations up to 1470 IU/dl, contrasting with normal or decreased ADAMTS13 activities, partly due to consumption by its substrate VWF and/or inhibition by high interleukin-6 levels, as previously shown.3 We, therefore, hypothesized that high D-Dimer and low ADAMTS13 activities may be related to microthrombosis in the respiratory system12 and represent surrogate markers of death, as reported with ADAMTS1311 or D-Dimer in COVID-1916 and VWF in early acute lung injury.17 Of note, we did not observe TTP, in agreement with other reports (Table S1). Increasing evidence supports the view that inflammation can cause thrombosis in large vessels mediated by VWF.3 In COVID-19, specific pathways promoting clot formation have been suggested involving SARS-CoV-2 binding to angiotensin-converting enzyme-2 with subsequent angiotensin-2 accumulation, vasoconstriction and VWF release from endothelial cells.2 Our findings linking VWF and ADAMTS13 to VTE support such hypotheses. Our study has limitations including the single-centre setting and the short study period. DVT was not systematically screened in non-critically ill patients, underestimating its prevalence in this group. To conclude, our data highlights the role of the VWF/ADAMTS13 axis at the crossroads between lung system microthrombosis and venous macrothrombosis in COVID-19. VS, BM and AS designed the study. MD, VS, MN, BJ, NB, BM and AS collected the data. MD, VS, BM and AS analysed the data. DS, BC, SV, PB and BM performed patient management. MD, VS, BM and AS wrote the first draft of the report. All authors read and approved the final manuscript. The authors thank Caren Brumpt for helpful data collecting and Agnès Veyradier for critical reading. The authors report no conflicts of interests. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

Topics & Concepts