key: cord-1025709-4szbkfn6
authors: Dolby, Heather W; Potey, Philippe M D; Wilder-Smith, Annika; Clohisey, Sara; Millar, Jonathan E; Baillie, J Kenneth; Dorward, David A; Lucas, Christopher D; Russell, Clark D
title: Histological evidence of pulmonary micro-thrombosis and vasculitis in life-threatening respiratory virus diseases
date: 2020-12-28
journal: Open Forum Infect Dis
DOI: 10.1093/ofid/ofaa640
sha: 055fa9bac4fe38afe28638634172008ec3ffc868
doc_id: 1025709
cord_uid: 4szbkfn6

Pulmonary micro-thrombosis and vasculitis occur in fatal COVID-19. To determine if these processes occur in other life-threatening respiratory virus infections we identified autopsy studies of fatal influenza(n=455 patients), SARS(n=37), MERS(n=2), adenovirus(n=34) and RSV(n=30). Histological evidence of thrombosis was frequently present in adults with fatal influenza and SARS, with vasculitis also reported.

Thrombotic complications occur with high frequency in coronavirus disease 2019 (COVID- 19) , involving pulmonary thrombo-emboli (PTE), deep vein thrombosis and catheter-related thrombosis, despite thromboprophylaxis with low molecular weight heparin (LMWH). 1 Histologically, pulmonary micro-thrombi are frequent autopsy findings in fatal COVID-19, even in the absence of macrovascular PTE, and in a sub-group of patients are likely due to immunothrombosis distinct from conventional PTE. 2 This occurs irrespective of receipt of LMWH thromboembolism prophylaxis. 2 There is also mounting evidence of a spectrum of pulmonary vasculitis in COVID-19: neutrophilic capillaritis 3, 4 , lymphocytic endotheliitis 5 , lympho-plasma cellular arterial vasculitis 6 , obliterating endarteritis involving C5aR1 + macrophages 7 and MRP8 + mononuclear cell vasculitis. 2 Together, these findings suggest pulmonary vasculitis and immunothrombosis could be treatable traits contributing to respiratory failure in sub-groups of patients. 8 The established clinical benefit of corticosteroids in COVID-19 supports a causal role for inflammation in severe disease, and trials of anticoagulation are ongoing (NCT04344756, NCT04406389). 9 We reviewed human autopsy data from other lifethreatening respiratory virus infections to determine if thrombosis and pulmonary vasculitis occur in other viral infections. 

Diffuse alveolar damage (DAD), alveolar haemorrhage and bronchiolitis were the most commonly reported findings in fatal influenza, with neutrophilic bronchopneumonia suggestive of secondary bacterial infection also common (Supplementary Data Table 1 ). DAD was specifically reported in every study, present in 75% of patients. Where data were available, 56% of patients received invasive mechanical ventilation (IMV; 150/270). The presence or absence of histological evidence of thrombosis was reported in 14/24 studies, with thrombosis present in patients from 12 of these studies and 66/317 (21%) patients where quantified. These were most commonly described as fibrin microthrombi. Pulmonary vascular inflammation was less commonly sought by investigators but was present in 5/6 studies where it was specified (15/79 cases where quantified, 19%). In four studies this was described as perivasculitis and where examined, inflammatory cells involved were CD8 + Tcells (n=7) or mononuclear cells (n=1). One study reported inflammatory infiltrate in the intima of medium vessels and endotheliitis (inflammatory cell type not stated). Vascular involvement in fatal influenza A H1N1 and Covid-19 was compared directly in one study 10 . Although prevalence within A c c e p t e d M a n u s c r i p t 6 the cohort was not reported, an infiltrate of CD3 + T-cells associated with pre/post-capillary walls was described in both diseases. Capillary micro-thrombi were seen in all cases from both diseases, but were substantially more prevalent in Covid-19.

DAD was present in all fatal cases and where stated all patients had received IMV. In SARS, thrombosis was reported in 3/4 studies and where quantified was present in 18/26 patients (specified as involving small veins/vessels in 17/18 patients). Vasculitis was reported as being present in two studies, though its prevalence was not stated (Supplementary Data Table 2 ). One study described a small vessel pulmonary vasculitis (inflammatory cell type not stated), and another reported vasculitis of small pulmonary veins, involving monocytes, neutrophils and lymphocytes, with fibrinoid necrosis (in addition to vasculitis of small veins in the heart, liver, kidney, adrenal gland and muscle, involving monocytes, lymphocytes and plasma cells). Only two case reports of autopsies in fatal MERS were identified, one of which reported a CD4 + lymphocytic pulmonary artery vasculitis.

There were no reports of pulmonary thrombosis in four studies of infants with fatal RSV infection, often in the context of sudden infant death (Supplementary Data Table 3 ). Vascular inflammation was described in one study, with a predominantly mononuclear and occasionally eosinophilic infiltrate surrounding bronchial arteries.

Nine studies of fatal adenovirus infection were identified with infrequent reports of thrombosis (2 patients) and vasculitis (2 patients), which was described as necrotising pulmonary vein vasculitis in one case and due to giant cells in the other (Supplementary Data Table 4) . and survive compared to patients who die. 18 Similar D-dimer changes occur in patients with H1N1 influenza A infection; which, combined with the increased risk of radiologically-diagnosed PTE, supports thrombosis being a relevant process in pathogenesis and not a non-specific post-mortem artefact. [19] [20] [21] Biological mechanisms further support the contribution of these processes to pathogenesis. Engulfment of influenza virions by platelets activates TLR7 signalling, leading to pro-thrombotic neutrophil DNA release and aggregation. 22 Platelet degranulation and neutrophil pro-thrombotic proteomic signatures have been identified in blood from patients with COVID-19 ARDS and low density neutrophils from these samples aggregate with platelets. 23 TLR7 is a ssRNA sensor, also involved in the host response to SARS-CoV-2 24 , and activation is likely to occur during infection with other single-stranded RNA viruses. A TLR7/8 agonist upregulates healthy neutrophil Mac-1 platelet binding complex, as see on neutrophils from COVID-19 ARDS patients. 23 Platelet-endothelial adhesion also occurs in vitro in response to influenza and in people with COVID-19, circulating platelets display a hyper-reactive transcriptional response and aggregate with leucocytes. 25, 26 In ARDS, platelets interact with endothelial cells, immune cells, neutrophil extracellular traps and pathogens, and their activation can lead to immunothrombosis. 27 Evidence from multiple sources supports the role of myeloid recruitment to the lung in COVID-19 which could link inflammation, vasculitis and immunothrombosis and identify therapeutic targets. In a genome wide association study of critical illness in COVID-19, a CCR2 variant predicted to increase expression in the lung was identified. 28 Similarly, the chemoattractant C5a and myeloid growth factor GM-CSF are associated with COVID-19 severity 7, 29 . The C5a-C5aR1 axis has also been associated with H1N1 influenza. 30 Immunothrombosis and vasculitis were identified in cohorts of adult patients whereas they were infrequent in children. Increased innate immune activation in older adults could contribute to A c c e p t e d M a n u s c r i p t 9 susceptibility to these processes, or alternatively it may relate to clinical differences: most of the children had died suddenly whereas most adults had received IMV and would have had a more protracted illness. 31 The majority of the identified studies, including COVID-19 and those from the 2009 H1N1 pandemic, have been conducted at a time when critically ill patients routinely receive thromboembolism prophylaxis with LMWH. 32 Thrombosis could therefore be occurring largely independently of the intrinsic pathway in some patients. Whilst the results of trials of therapeutic anticoagulation in COVID-19 are awaited, we suggest that a sub-group of patients with immunothrombosis, especially with vasculitis, may be more responsive to immunomodulatory therapy, distinct from conventional PTE responsive to therapeutic anticoagulation alone. An alternative or potentially complementary approach would be to therapeutically protect the endothelium, as recently discussed in the context of COVID-19. 33 

This study did not include any work requiring patient consent. 

Incidence of thrombotic complications in critically ill ICU patients with COVID-19

Tissue-specific Immunopathology in Fatal COVID-19

Targeting potential drivers of COVID-19: Neutrophil extracellular traps

Postmortem examination of COVID-19 patients reveals diffuse alveolar damage with severe capillary congestion and variegated findings in lungs and other organs suggesting vascular dysfunction

Endothelial cell infection and endotheliitis in COVID-19

Early postmortem mapping of SARS-CoV-2 RNA in patients with COVID-19 and correlation to tissue damage

Association of COVID-19 inflammation with activation of the C5a-C5aR1 axis

Treatable traits and therapeutic targets: Goals for systems biology in infectious disease

Dexamethasone in Hospitalized Patients with Covid-19 -Preliminary Report

Severe acute respiratory syndromecoronavirus infection in aged nonhuman primates is associated with modulated pulmonary and systemic immune responses

Replication and pathogenesis of the pandemic (H1N1) 2009 influenza virus in mammalian models

MyD88 is required for protection from lethal infection with a mouse-adapted SARS-CoV

Acute Respiratory Distress Syndrome and Diffuse Alveolar Damage. New Insights on a Complex Relationship

Intravascular coagulation associated with the adult respiratory distress syndrome

The pulmonary vascular lesions of the adult respiratory distress syndrome

Diffuse alveolar damage associated mortality in selected acute respiratory distress syndrome patients with open lung biopsy

Elevated antiviral, myeloid and endothelial inflammatory markers in severe COVID-19. medRxiv

Serum D-dimer changes and prognostic implication in 2009 novel influenza A(H1N1)

Chest radiographic and CT findings in novel swineorigin influenza A (H1N1) virus (S-OIV) infection

Thromboembolic events in patients with severe pandemic influenza A/H1N1

The role of platelets in mediating a response to human influenza infection

Proteomics identifies a type I IFN, prothrombotic hyperinflammatory circulating COVID-19 neutrophil signature distinct from non-COVID-19 ARDS. medRxiv

Systematic review and metaanalysis identifies potential host therapeutic targets in COVID-19. medRxiv

Platelet gene expression and function in patients with COVID-19. Blood

Influenza Virus Infection Induces Platelet-Endothelial Adhesion Which Contributes to Lung Injury

Immunothrombosis in Acute Respiratory Distress Syndrome: Cross Talks between Inflammation and Coagulation

Genetic mechanisms of critical illness in Covid-19. medRxiv

Targeting GM-CSF in COVID-19 Pneumonia: Rationale and Strategies

The role of C5a in acute lung injury induced by highly pathogenic viral infections. Emerg Microbes Infect

Age-dependent dysregulation of innate immunity

Venous thromboembolism in the ICU: main characteristics, diagnosis and thromboprophylaxis. Critical Care

Pulmonary Endothelial Dysfunction and Thrombotic Complications in COVID-19 Patients

A c c e p t e d M a n u s c r i p t 11 A c c e p t e d M a n u s c r i p t