key: cord-0772305-yxky7ycm
authors: Obi, Andrea T.; Barnes, Geoffrey D.; Napolitano, Lena M.; Henke, Peter K.; Wakefield, Thomas W.
title: Venous thrombosis epidemiology, pathophysiology, and anticoagulant therapies and trials in severe acute respiratory syndrome coronavirus 2 infection
date: 2020-09-08
journal: J Vasc Surg Venous Lymphat Disord
DOI: 10.1016/j.jvsv.2020.08.030
sha: 61e6b64cdcba92d412b2f83c1aab369b7762d333
doc_id: 772305
cord_uid: yxky7ycm

OBJECTIVE: Infection with the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) virus confers a risk of significant coagulopathy, with the resulting development of venous thromboembolism (VTE), potentially contributing to the morbidity and mortality. The purpose of the present review was to evaluate the potential mechanisms that contribute to this increased risk of coagulopathy and the role of anticoagulants in treatment. METHODS: A literature review of coronavirus disease 2019 (COVID-19) and/or SARS-CoV-2 and cell-mediated inflammation, clinical coagulation abnormalities, hypercoagulability, pulmonary intravascular coagulopathy, and anticoagulation was performed. The National Clinical Trials database was queried for ongoing studies of anticoagulation and/or antithrombotic treatment or the incidence or prevalence of thrombotic events in patients with SARS-CoV-2 infection. RESULTS: The reported rate of VTE among critically ill patients infected with SARS-CoV-2 has been 21% to 69%. The phenomenon of breakthrough VTE, or the acute development of VTE despite adequate chemoprophylaxis or treatment dose anticoagulation, has been shown to occur with severe infection. The pathophysiology of overt hypercoagulability and the development of VTE is likely multifactorial, with evidence supporting the role of significant cell-mediated responses, including neutrophils and monocytes/macrophages, endothelialitis, cytokine release syndrome, and dysregulation of fibrinolysis. Collectively, this inflammatory process contributes to the severe pulmonary pathology experienced by patients with COVID-19. As the infection worsens, extreme D-dimer elevations, significant thrombocytopenia, decreasing fibrinogen, and prolongation of prothrombin time and partial thromboplastin time occur, often associated with deep vein thrombosis, in situ pulmonary thrombi, and/or pulmonary embolism. A new phenomenon, termed pulmonary intravascular coagulopathy, has been associated with morbidity in patients with severe infection. Heparin, both unfractionated heparin and low-molecular-weight heparin, have emerged as agents that can address the viral infection, inflammation, and thrombosis in this syndrome. CONCLUSIONS: The overwhelming inflammatory response in patients with SARS-CoV-2 infection can lead to a hypercoagulable state, microthrombosis, large vessel thrombosis, and, ultimately, death. Early VTE prophylaxis should be provided to all admitted patients. Therapeutic anticoagulation therapy might be beneficial for critically ill patients and is the focus of 39 ongoing trials. Close monitoring for thrombotic complications is imperative, and, if confirmed, early transition from prophylactic to therapeutic anticoagulation should be instituted. The interplay between inflammation and thrombosis has been shown to be a hallmark of the SARS-CoV-2 viral infection.

Impaired pulmonary function and coagulation abnormalities have been shown to be hallmarks of severe infections with coronavirus disease 2019 (COVID- 19) , caused by severe acute respiratory syndrome coronavirus 2 (SAR-S-CoV-2). The link among infection, inflammation, and hypercoagulability has been highlighted by the three clinical stages associated with COVID-19 infection recently described by Thachil et al 1 in primarily hospitalized patients: Stage 1: patients with nonsevere symptoms localized to the lungs and associated with a D-dimer level two to three times greater than normal but normal prothrombin times (PTs) and partial thromboplastin times (PTTs) and normal or elevated platelet counts and fibrinogen levels. These patients can be treated in the hospital or at home. Stage 2: patients with more severe symptoms requiring critical care associated with D-dimer levels three to six times greater than normal with a decreased platelet count and minor prolongation of the PT. These patients can develop deep vein thrombosis in the lower limbs and have filling defects on computed tomography imaging due to pulmonary thrombi or emboli. Stage 3: patients with worsening clinical symptoms and requiring higher level critical care (eg. extracorporeal membrane oxygenation [ECMO] support). These patients can develop venous thromboembolism (VTE) and multisystem organ failure with ischemia in various organs. The D-dimer levels will be six times greater than normal or more, with significant thrombocytopenia, decreasing fibrinogen levels, and marked prolongation of the PT and PTT. Systemic thrombosis, including a disseminated intravascular coagulation (DIC)-like picture, and extensive pulmonary thrombi or emboli can develop. These changes can result in eventual mortality. 1 Within the present review, we have examined the epidemiology of venous thrombosis in patients with COVID-19 infection, the role of the cellular and molecular mechanisms underlying coagulation abnormalities, and the data regarding anticoagulant use for these patients. We have highlighted the ongoing trials that could be of interest to vascular specialists.

VTE is a common hospital-acquired condition, occurring in medical and surgical intensive care unit (ICU) patients at rates historically ranging from 6.6% to 30%. [2] [3] [4] [5] Modern thromboprophylaxis has effectively reduced the risk of VTE development in the ICU patient population to 5% to 15%. [5] [6] [7] However, the initial reports of critically ill patients with COVID-19 in Wuhan, China, demonstrated a VTE rate of 25%, greater than expected in a population with a baseline low incidence of both VTE and thromboprophylaxis usage. 8, 9 Similarities were quickly found across multiple other patient populations. The incidence of VTE in critically ill Dutch patients with COVID-19 was 31% 10 and was 32% in Swiss patients. 11 An autopsy study of 12 consecutive patients from Hamburg, Germany, found previously unsuspected deep vein thrombosis in 7 of the 12 patients (58%), with pulmonary embolism (PE) considered the direct cause of death for 4 patients. 12 Pulmonary infection has been established as a risk factor for VTE in multiple studies. 13, 14 A mild elevation in VTE risk continues for #1 year after recovery from the initial infection. 13 Data are lacking regarding the VTE risk from other coronavirus pneumonias. The SARS pandemic in 2003 was associated with a 23% to 33% risk of VTE in the critically ill; however, no findings of an association between VTE and Middle East respiratory syndrome were found in a review of the reported data. 15, 16 In contrast, the exaggerated risk described in the early studies of patients with COVID-19 more closely approximated the severe coagulopathy associated with the 2009 H1N1 viral pneumonia pandemic. In that pandemic, the VTE incidence was estimated at 37% among ICU patients. 17 The remarkable similarities between the COVID-19 and 2009 H1N1 pandemics extend to a high incidence of breakthrough VTE, defined as the occurrence of VTE despite aggressive thromboprophylaxis or therapeutic anticoagulation. 3 In 150 patients infected with SARS-CoV-2 who had developed adult respiratory distress syndrome (ARDS) in France, 64 (42.7%) thrombotic complications were reported. The thrombotic complications included PE in 16.7%, continuous renal replacement therapy circuit clotting in 28 of 29 patients, and ECMO circuit clotting in 3 of 12 patients receiving ECMO support, although all had receiving thromboprophylaxis (70% a prophylactic dose and 30% a therapeutic dose). 18 Another report of 107 ICU patients with COVID-19 pneumonia confirmed the presence of PE in 22 (21%) despite prophylactic dose anticoagulant therapy for 20 patients and a therapeutic dose for 2 patients. 19 In a cohort of French critically ill patients, the incidence of VTE was 100% for those treated with prophylactic anticoagulation and 56% for those who had received therapeutic anticoagulation. 20 In a single-center study from Amsterdam of 198 hospitalized patients with COVID-19 (all of whom had received thromboprophylaxis), the cumulative VTE incidence at 7, 14, and 21 days was 16%, 33%, and 42%, respectively. The cumulative incidence of VTE was greater in the ICU patients (n ¼ 75; 26%, 47%, and 59%) than in the non-ICU patients (n ¼ 123; 5.8%, 9.2%, and 9.2%) at 7, 14, and 21 days, respectively. However, none of the 19 patients (0%) who had continued therapeutic anticoagulation therapy for other indications had developed VTE compared with 39 of 179 of the remaining patients (22%; Fisher's exact test, P ¼ .03). VTE was associated with death (adjusted hazard ratio, 2.4; 95% confidence interval, 1.02-5.5). 21 In contrast, SARs-CoV-2eassociated arterial thrombosis and thromboembolism has occurred at a lower frequency. In a study of three Dutch hospitals evaluating 184 ICU patients with proven COVID-19 pneumonia, arterial events, which included ischemic stroke, myocardial infarction, or systemic arterial embolism, occurred in 3.7%. 10 Of 829 ICU patients with a high prescription of chemical thromboprophylaxis, 18.6% had developed arterial thrombotic events, including myocardial infarction, ischemic stroke, and other systemic thromboembolism. 22 Thus, patients infected with the novel coronavirus who develop severe pneumonia will exhibit higher than expected rates of VTE compared with patients without such severe viral pneumonia. Hypercoagulability has been shown to be a manifestation of severe COVID-19 pneumonia. 23, 24 Overt, pathological venous thrombosis might be an important contributor to the morbidity and mortality of critically ill patients with COVID-19, prompting multiple societies to recognize the importance of thromboprophylaxis for this patient population. [25] [26] [27] Further understanding of the cellular and molecular mechanisms responsible for this hypercoagulable state are necessary for the development of effective therapies.

The usual cause of death and the need for ventilatorassisted hospitalization from COVID-19 has been the progression to ARDS. The hallmark of ARDS is the influx of neutrophils (polymorphonuclear [PMN] cell family) into the lungs after extravasation from the bloodstream. PMN degranulation releases a wide variety of inflammatory mediators, resulting in a positive feedback loop that results in a further influx of PMN cells, vascular leakage, a buildup of fluid in the lungs, and the inability to exchange oxygen. Alveolar neutrophil and platelet accumulation results in thromboinflammation. Platelets, in particular, have the ability to trigger deployment of neutrophil extracellular traps (NETs), which correlate with the alveolarecapillary and epithelial barrier disruption in ARDS. 28 In patients with severe infection of SARS-CoV-2 viral pneumonia progressing to respiratory failure, serum cell-free DNA, myeloperoxidase (MPO)-DNA, and citrullinated histone H3 levels will be elevated; the latter are two highly specific markers for NETs. Cell-free DNA and MPO-DNA correlate strongly with the absolute neutrophil count. 29 Cell-free DNA and MPO-DNA were more elevated in hospitalized patients receiving mechanical ventilation compared with hospitalized patients breathing room air, suggesting that the severity of the illness is associated with the degree of neutrophil response. 29 Additionally, serum from individuals infected with SARS-CoV-2 triggered NET release from control neutrophils in experiments performed in vitro. 29 In addition, a recent retrospective case-control study of hospitalized patients with COVID-19 reported that the patients who had developed thrombosis had had significantly greater blood levels of NET remnants (cell-free DNA, MPO-DNA complexes, and citrullinated histone H3) and neutrophil activation (calprotectin) compared with those without thrombosis and were associated with increased D-dimer levels. 30 Taken together, these data provide evidence that SARS-CoV-2 infection induces a pro-NETotic state in hospitalized patients. These NETs could be partially responsible for the thrombotic phenotype of severely ill patients with COVID-19 (Table I) . Thirteen studies are listed in ClinicalTrials.gov under COVID and NETs, pointing to the potentially important role of activated neutrophils in COVID-19 illness.

In addition to neutrophils, pathologic studies have demonstrated a low to moderate number of mononuclear cells in the lungs of patients who died of COVID-19. 31 The direct role they play is not entirely clear but is likely multifactorial because of the diverse roles monocytes and macrophages (Mo/MF) play in lung physiology. Broadly, Mo/MF are classified by their inflammatory or anti-inflammatory functions. [32] [33] [34] For example, interleukin (IL)-1 and IL-12 secretion and cell surface Ly6C hi , CCR2 þþ , and CX3CR1 þ antigen expression characterize classically activated proinflammatory Mo/MF. In contrast, IL-10 secretion, Nr4a1 transcription factor dependency, and cell surface Ly6C lo , CCR2 À , and CX 3 CR1 þþ antigen expression characterize alternatively activated Mo/MF with prohealing and inflammation-resolving activities. It has been shown that the number of monocytes is the same in patients infected with COVID-19 and controls. However, in patients from China with COVID-19, significant morphologic and functional differences were seen in the monocytes, with the differences more pronounced in those who had required ICU admission and had a prolonged length of stay. These monocytes were larger on flow cytometry and were positive for CD11b, CD14, CD16, CD68, CD80, CD163, and CD206. They also secreted IL-6, IL-10, and tumor necrosis factor-a, an indication of both an inflammatory and anti-inflammatory phenotype. 35, 36 Humans can manifest a significant cytokine storm with severe COVID-19 pneumonia that can lead to death, 37, 38 and many of the responsible mediators are derived from Mo/MF. [39] [40] [41] These include proinflammatory cytokines such as IL-1b, IL-8, IL-6, and interferon (IFN) type I. These cells then directly interact with the other major leukocytes that could be involved in the development of VTE. 42, 43 The direct viral infection might promote the release of monocyte chemoattractants by alveolar epithelial cells, alveolar macrophages, and stromal cells and result in a sustained recruitment of monocytes into the lungs, creating a vicious circle. 44 These monocytes then differentiate into proinflammatory macrophages. Consistently, inflammatory Mo/MØ have been shown to mediate the pathogenesis of experimental SARS-CoV-2 pneumonia in mice. 45 Table I . Cell-mediated molecular mechanisms of hypercoagulability Neutrophils NETs Monocytes/macrophages Type I interferons, IL-1b, IL-6, IL-8, TF   Endothelial cells  Increased TF, decreased TFPI, AT,  protein C AT, Antithrombin; IL, interleukin; NETs, neutrophil extracellular traps; TF, tissue factor; TFPI, tissue factor pathway inhibitor.

Mo/MØ also activate and are activated by natural killer cells and T cells, further promoting the recruitment of these monocyte-derived macrophages through the production of granulocyte-macrophage colony-stimulating factor, tumor necrosis factor, and IFN-g. 44, 46 In the lungs, oxidized phospholipids can accumulate and can also activate these Mo/MØ through a number of pathways. 44 Virus-sensing pathways such as toll-like receptor 7, which senses singlestranded RNA recognition, and IFN type I can induce the expression of entry receptors, allowing the virus to enter into the cytoplasm of the macrophages, activating the NLRP3 inflammasome. 44 This can then lead to the secretion of mature IL-1b and IL-18, which then amplify the activation of the macrophages through paracrine or autocrine effects and might have procoagulant effects. 29 Oxidative stress, pathogen-associated molecular patterns, and damage-associated molecular patterns (DAMPs) also activate Mo/MØ via the TLR4-TRIF-TRAF6-NFkb pathway, resulting in surface tissue factor expression. 44, 47 Additionally, activated inflammatory cells (neutrophils and Mo/MØ) can directly bind to platelets, and these neutrophileplatelet and plateletemonocyte complexes and activated platelets can induce and amplify thrombosis.

Endothelial cell infection has been implicated as a cause of the multiple organ involvement associated with severe COVID-19 infection. The induction of endotheliitis occurs as a direct consequence of viral involvement (as noted by the presence of viral bodies) and the host inflammatory response. 48 Other proposed mechanisms of endothelial injury include the induction of apoptosis and pyropotosis. 48 Endotheliitis could explain the systemic impaired microcirculatory function in different vascular beds (including the kidneys and lungs) and their clinical sequelae (ie, acute kidney injury, ARDS) in patients with COVID-19 infection. During inflammation, natural anticoagulants such as the tissue factor pathway inhibitor, antithrombin (AT), and protein C will also be downregulated on the endothelial cell surface, with tissue factor upregulated, contributing to thrombosis. 43, 49 Thus, a therapy that targets the endothelial cells might play an important protective role in the treatment of COVID-19. 48 One potential candidate could be selectin inhibitors, because selectins are critical to the interactions among inflammatory cells, platelets, and the endothelium.

The pathophysiology of COVID-19eassociated hypercoagulability has many unique aspects (Table II) . Severe COVID-19 hypercoagulopathy is associated with features of both DIC and thrombotic microangiopathy, conferring a procoagulant state associated with VTE risk. This hypercoagulable state appears to be related to the severity of illness and resultant thromboinflammation and not to intrinsic viral activity. Systemic coagulopathy is a marker of adverse outcomes for these patients, and elevated Ddimer levels at admission and increasing D-dimer levels after admission have been associated with significantly increased mortality. 50 Coagulation abnormalities have been shown to be the hallmark of severe COVID-19 infection, with more marked abnormalities developing as patients progress from stage 1 to 3. 1 A study that had prospectively collected blood coagulation values from 94 admitted patients and 40 healthy controls in Wuhan, China, reported the following results (patients vs controls): lower AT values, elevated D-dimer levels, elevated fibrin degradation products (FDPs), and elevated fibrinogen (P < .001 for all). Greater D-dimer elevations (not age adjusted) and FDP values correlated with more severe or critical disease. 51 In a retrospective study of 183 patients (Wuhan, China), elevated PT, D-dimer, and FDP values on hospitalization were associated with patient death. 52 The investigators had also analyzed the values for a period of 14 days. They noted that a further increase in D-dimer and FDP levels with a decrease in fibrinogen and AT activity was observed in the patients who had eventually died compared with the survivors after 7 to 10 days. 52 D-dimer elevation has been consistently associated with disease severity, mortality, and need for assisted ventilation. 53, 54 These serologic and coagulation markers should be measured and monitored throughout the course of the infection. Whether a change in serologic markers, such as further increases in D-dimer and FDP levels and a decrease in fibrinogen and AT activity, can be used as a method to determine when to switch from prophylactic to therapeutic anticoagulation remains to be determined. However, some experts have recommend transitioning to therapeutic anticoagulation at D-dimer levels of three to six times the upper limit of normal. 1 Only approximately one third of 1099 patients with COVID-19 from 31 provinces in China developed clinical thrombocytopenia. 53 In 30 patients infected with SARS-CoV-2, the thromboelastography parameters were consistent with a state of hypercoagulability as shown by decreased reaction times and kinetic values, increased angle and maximum amplitude values, and a small decrease in Ly30. Other laboratory abnormalities included increased fibrinogen, D-dimer, C-reactive protein, factor VIII, protein C, and von Willebrand factor (n ¼ 11), suggesting the presence of hypercoagulability with a severe inflammatory state. 55 These results differ from the hallmark characteristics often seen with acute DIC, which will typically include consumptive coagulopathy with significantly prolonged PT/activated PTT, thrombocytopenia, and low fibrinogen clotting activity. 55 It is only at the most severe stage of infection (stage 3) that the laboratory abnormalities in COVID-19 will appear more consistent with DIC, such as significant thrombocytopenia, marked prolongation in the PT and PTT, and decreasing fibrinogen levels. 1 Recently, soluble urokinase has been identified as a predictor of the development of severe respiratory failure in the setting of SARS-CoV-2 infection. 56 Consistently, mice infected with murine-adapted related SARS-CoV-2 developed more severe lung pathology in the setting of a dysregulated urokinase pathway. Similar changes in the urokinase, coagulation, and fibrinolytic pathway expression signatures have been noted in those with highly pathogenic SARS-CoV-2 and influenza virus infections, arguing for a conserved role for these pathways in virus-induced end-stage lung disease, similar to that in acute lung injury and ARDS. 57 A recent review has suggested that elevated plasmin or plasminogen might be a common risk factor for COVID-19 susceptibility. The rationale for this includes that elevated plasmin or plasminogen has been common in patients identified as having a high risk of COVID-19, that plasmin cleaves the SARS-CoV-2 spike protein, which might enhance the action of the virus, and that the very high D-dimer levels (which have been related to mortality and severity) might relate to increased fibrinolysis as mediated by plasmin. 58 Thus, another approach to treatment could be the use of antiprotease compounds.

Data are emerging that COVID-19 infection causes diffuse bilateral pulmonary inflammation associated with a novel pulmonary-specific vasculopathy, which has recently been termed pulmonary intravascular coagulopathy (PIC). PIC is distinct from DIC in its early stages. [59] [60] [61] Lung pathologic examinations from autopsy of patients with COVID-19 have confirmed significant microvascular thrombosis and hemorrhage, in addition to the extensive alveolar and interstitial inflammation causing diffuse alveolar damage. The common pulmonary vascular findings in these postmortem reports of patients with COVID-19 pneumonia included vascular thrombosis, fibrin thrombi, small vessel occlusion, blood vessel wall edema, immune cell infiltration, and pulmonary infarction. 31, 62, 63 Infection of the type II pneumocytes with COVID-19 virus induces pulmonary macrophage and neutrophil recruitment (similar to macrophage activation syndrome) triggering diffuse pulmonary inflammation and immunothrombosis. Pulmonary microthrombi formation will stimulate local fibrinolytic activity in the lung, which might not be adequate. Thus, ultimately, the local immunothrombosis induced by COVID-19 could result in pulmonary infarction and hemorrhage, with resultant PIC-induced pulmonary hypertension. A recent small study compared the lungs from patients who had died of COVID-19 and patients who had died of influenza A (H1N1) infection. 64 They reported three distinctive pulmonary vascular features of COVID-19 infections, including (1) severe endothelial injury associated with intracellular virus and disrupted cell membranes; (2) widespread vascular thrombosis with microangiopathy and occlusion of alveolar capillaries; and (3) significant new vessel growth through a mechanism of intussusceptive angiogenesis. Also, the alveolar capillary microthrombi were nine times greater in the COVID-19 lungs compared with the influenza A lungs, and vascular angiogenesis was 2.7 times greater. 64 

Heparin serves a multifaceted role in the setting of acute infection. In addition to its function as an anticoagulant, it has a role as a negative regulatory factor for smooth muscle cell proliferation and migration and shifts the localization of urokinase-type plasminogen activator from the cell layer to the medium, depending on the mitogen used. 65 Heparin alters fibrinolysis by potentiating the effect of AT to inhibit the plasminogen system and plasmin. 66, 67 Heparin has also exhibited significant anti-inflammatory effects by neutralizing DAMPs to protect the endothelial cells and reducing the toxicity of histones on endothelial tight junctions, resulting in decreased lung edema and vascular leakage. 68, 69 Cell-surface heparin sulfate has been implicated as a viral binding site. 70 Specifically, blocking this virus binding site with heparin inhibited viral infection in a cellbased assay. 71 A direct antiviral effect of heparin has been reported for SARS-CoV-2. 72 Binding of the viral spike protein to the cell surface heparin sulfate was reduced in the presence of heparin, inhibiting viral infection into cells. 73 A mutation in the spike protein residue 614 from aspartic acid to glycine has recently been identified as a possible fitness advantage, increasing the transmission of the virus in Europe and the Americas. 74 Whether this mutation (or future mutations in the spike protein) will significantly affect heparin binding is an important translational question.

Anticoagulants, including unfractionated heparin and low-molecular-weight heparin, possess a number of general anti-inflammatory effects. These include (1) inhibition of the activation and function of neutrophils; (2) prevention of the expression of inflammatory mediators, which initiate and drive the activation of the innate immune system, reducing many cytokines; (3) inhibition of the proliferation of vascular smooth muscle cells; and (4) inhibition of thrombosis with a decrease in thrombin and a subsequent decrease in inflammation. 75 These effects appear independent of achieving therapeutic anticoagulation. 76 Heparin might serve a dual role in critically ill patients with severe COVID-19, inhibiting viral cell infection and decreasing the proinflammatory effects of the cytokine storm, including chemokines and danger signals (DAMPs). It might be possible to design heparin-like molecules to enhance the antiviral and anti-inflammatory effects and lessening the anticoagulant effect, which would decrease bleeding. 77 Heparin administered at prophylactic doses 78 and at just less than therapeutic doses has shown beneficial effects on COVID-19 mortality and influenza A H1N1 mortality. 17 A recent Chinese study compared heparin prophylaxis in 99 patients with severe COVID-19 infection and 350 patients without heparin prophylaxis. 78 They reported that the 28-day mortality rate was significantly decreased from 64.2% to 40% with heparin prophylaxis (P ¼ .029) in patients with increased sepsis-induced coagulopathy scores of $4 or D-dimer levels more than sixfold the upper limit of normal (mortality, 32.8% vs 52.4%; P ¼ .017). 78 Additional studies are underway to validate these important findings.

The results from two other studies have suggested potentially beneficial effects of treatment-dose heparin anticoagulation in patients with COVID-19. In a study from Brazil, 27 consecutive patients with COVID-19 had been treated with therapeutic heparin anticoagulation. 79 The investigators reported that the partial pressure of arterial oxygen/fractional inspired oxygen ratio had increased significantly during the 72 hours after the start of anticoagulation. Also, 81% of the patients could be discharged home within 11 days. 79 The patients who had required intubation and mechanical ventilation had been extubated within 12.5 days, and no bleeding complications had occurred. 79 Additionally, in a study from New York, from March 14 to April 11, 2020, 2773 patients had been hospitalized within the Mount Sinai Health System in New York City. These patients had been given treatment-dose systemic anticoagulation. 80 After adjustment for demographics and comorbid disease, mortality was assessed. For the patients who had required mechanical ventilation, the mortality in the hospital had decreased from 62.7% to 29.1%, with a median survival of 21 days compared with 9 days for those who had not received anticoagulation. 80 No survival benefit was seen for the non-ICU patients. This observational study had not accounted for immortal time bias, which could have contributed to the reported association between therapeutic anticoagulation and beneficial outcomes. 80 At our institution, because of the high observed incidence of VTE events and previous clinical success with heparin infusion for severe influenza, we have endorsed heparin infusion for critically ill patients with SARs-CoV-2 infection (Fig) . Duplex ultrasound scanning is performed only for patients considered at high risk of bleeding and thrombosis by the bedside clinician, with the aid of clinical prediction models. 81 Widespread duplex ultrasound screening is not feasible during pandemic circumstances owing to sheer number of patients, amount of time required for disinfection of the ultrasound machine, COVID-19 risk to the technician, and often prone positioning of the patients. At our institution, patients with a contraindication to anticoagulation and clinical imaging of acute VTE and those with thrombus found in ECMO cannulas will undergo inferior vena cava filter placement. The relative success or failure of this protocol compared with others remains to be seen. In particular, a pressing need exists for additional data regarding the risks of major bleeding for patients treated with full-dose anticoagulants. In a recent study, the development of major hemorrhage occurred in 4.8% of severely ill patients with COVID-19. Of these 21 patients, 7 had been receiving therapeutic full-dose anticoagulation at the occurrence of hemorrhage. 82 Anticoagulation has been the genesis of many studies around the world in the treatment paradigm of COVID-19. We found 37 studies listed under COVID-19 and anticoagulation in the ClinicalTrials.gov website as of June 2, 2020. These studies are investigating heparin and low-molecular-weight heparin at various doses, alone and combined with other agents, such as direct oral anticoagulation agents and platelet inhibitors. A portion of these studies, plus others, are listed in Table III .

Infection with COVID-19 clearly induces overlapping pathways that link the inflammatory response to the coagulation response. The overwhelming inflammatory response in patients with COVID-19 infection can lead to a hypercoagulable state, microthrombosis, large vessel thrombosis, and, ultimately, death. Thus, early VTE prophylaxis should be provided to all admitted patients, along with other agents that will interrupt the hyperimmune response that occurs with infection with the virus. 83 Close monitoring for thrombotic complications is imperative, and, if confirmed, early transition from prophylactic to therapeutic anticoagulation should be instituted. Therapeutic anticoagulation could be required in certain clinical situations, and the transition from prophylactic to therapeutic anticoagulation for many patients could be required when diagnostic imaging studies are not available. 81 Further guidelines will depend on the results of clinical trials. 84 We have no better example of the close interaction between inflammation and thrombosis, suggesting that inflammation augments thrombosis and that thrombosis augments inflammation. In such situations, it is likely that a combination of anticoagulation agents, anti-inflammatory agents, and antiviral agents will be necessary to treat this deadly disease. The specific anticoagulant and its formulation, dosage, and timing and the specific antiinflammatory agent and antiviral agent to be used are important topics for further timely research.

Conception and design: AO, PH, TW Analysis and interpretation: AO, GB, LN, PH, TW 

A proposal for staging COVID-19 coagulopathy

Venous thromboembolism prophylaxis in critically ill patients

Validation of the Caprini venous thromboembolism risk assessment model in critically ill surgical patients

Deep vein thrombosis during prolonged mechanical ventilation despite prophylaxis

A comparison of enoxaparin with placebo for the prevention of venous thromboembolism in acutely ill medical patients. Prophylaxis in Medical Patients with Enoxaparin Study Group

Low-molecular-weight heparin thromboprophylaxis in medical-surgical critically ill patients: a systematic review

Saudi Critical Care Trials Group. Adjunctive intermittent pneumatic compression for venous thromboprophylaxis

Incidence of venous thromboembolism in Asian populations: a systematic review

Prevalence of venous thromboembolism in patients with severe novel coronavirus pneumonia

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

Venous thromboembolism in critically ill patients with COVID-19: results of a screening study for deep vein thrombosis

Autopsy findings and venous thromboembolism in patients with COVID-19: a prospective cohort study

Acute infections and venous thromboembolism

Risk factors and clinical impact of postoperative symptomatic venous thromboembolism

Large artery ischaemic stroke in severe acute respiratory syndrome (SARS)

Critically ill patients with severe acute respiratory syndrome (SARS) in a designated national SARS ICU: clinical features and predictors for mortality

Empirical systemic anticoagulation is associated with decreased venous thromboembolism in critically ill influenza A H1N1 acute respiratory distress syndrome patients

CRICS TRIGGERSEP Group (Clinical Research in Intensive Care and Sepsis Trial Group for Global Evaluation and Research in Sepsis). High risk of thrombosis in patients with severe SARS-CoV-2 infection: a multicenter prospective cohort study

Lille ICU Haemostasis COVID-19 Group. Pulmonary embolism in COVID-19 patients: awareness of an increased prevalence

High incidence of venous thromboembolic events in anticoagulated severe COVID-19 patients

Incidence of venous thromboembolism in hospitalized patients with COVID-19

Thrombosis in hospitalized patients with COVID-19 in a New York City health system

Coagulopathy of coronavirus disease 2019

COVID-19 and its implications for thrombosis and anticoagulation

Attention should be paid to venous thromboembolism prophylaxis in the management of COVID-19

Thromboembolism and anticoagulant therapy during the COVID-19 pandemic: interim clinical guidance from the anticoagulation forum

COVID-19 and thrombotic or thromboembolic disease: implications for prevention, antithrombotic therapy, and follow-up

Acute respiratory distress syndrome

Neutrophil extracellular traps in COVID-19

Neutrophil extracellular traps and thrombosis in COVID-19

Pulmonary and cardiac pathology in African American patients with COVID-19: an autopsy series from New Orleans

Protective and pathogenic functions of macrophage subsets

Macrophage activation and polarization: nomenclature and experimental guidelines

Macrophage biology in development, homeostasis and disease

COVID-19 infection induces readily detectable morphological and inflammation-related phenotypic changes in peripheral blood monocytes, the severity of which correlate with patient outcome

Aberrant pathogenic GM-CSF þ T cells and inflammatory CD14 þ CD16 þ monocytes in severe pulmonary syndrome patients of a new coronavirus

Cytokine release syndrome in severe COVID-19

Plasma inflammatory cytokines and chemokines in severe acute respiratory syndrome

Inhibition of NF-kappaB-mediated inflammation in severe acute respiratory syndrome coronavirus-infected mice increases survival

Targeting potential drivers of COVID-19: neutrophil extracellular traps

The host immune response in respiratory virus infection: balancing virus clearance and immunopathology

Immune factors in deep vein thrombosis initiation

Mechanisms of venous thrombosis and resolution

Pathological inflammation in patients with COVID-19: a key role for monocytes and macrophages

Dysregulated type I interferon and inflammatory monocyte-macrophage responses cause lethal pneumonia in SARS-CoV-infected mice

Pathogenic T-cells and inflammatory monocytes incite inflammatory storms in severe COVID-19 patients

Identification of oxidative stress and toll-like receptor 4 signaling as a key pathway of acute lung injury

Endothelial cell infection and endotheliitis in COVID-19

Inflammation and thrombosis

Webinar: thrombosis, thromboprophylaxis & coagulopathy in COVID-19 infections. isth Academy

Prominent changes in blood coagulation of patients with SARS-CoV-2 infection

Abnormal coagulation parameters are associated with poor prognosis in patients with novel coronavirus pneumonia

China Medical Treatment Expert Group for Covid-19. Clinical characteristics of coronavirus disease 2019 in China

Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study

Hypercoagulability of COVID-19 patients in intensive care unit: a report of thromboelastography findings and other parameters of hemostasis

Soluble urokinase plasminogen activator receptor (suPAR) as an early predictor of severe respiratory failure in patients with COVID-19 pneumonia

Mechanisms of severe acute respiratory syndrome coronavirus-induced acute lung injury

Elevated plasmin(-ogen) as a common risk factor for COVID-19 susceptibility

The role of cytokines including interleukin-6 in COVID-19 induced pneumonia and macrophage activation syndrome-like disease

COVID19 coagulopathy in Caucasian patients

Why the immune mechanisms of pulmonary intravascular coagulopathy in COVID-19 pneumonia are distinct from macrophage activation syndrome with disseminated intravascular coagulation

Clinical pathology of critical patient with novel coronavirus pneumonia

Pulmonary vascular endothelialitis, thrombosis, and angiogenesis in COVID-19

Regulation of baboon arterial smooth muscle cell plasminogen activators by heparin and growth factors

The action of antithrombin III on plasmin and activators of plasminogen

Role of heparin in the inactivation of thrombin, factor Xa, and plasmin by antithrombin III

The versatile heparin in COVID-19

Unfractionated heparin alleviates sepsis-induced acute lung injury by protecting tight junctions

Cell surface heparan sulfate and its roles in assisting viral infections

Initial interaction of herpes simplex virus with cells is binding to heparan sulfate

Inhibition of SARS pseudovirus cell entry by lactoferrin binding to heparan sulfate proteoglycans

The 2019 coronavirus (SARS-CoV-2) surface protein (Spike) S1 receptor binding domain undergoes conformational change upon heparin binding

Tracking changes in SARS-CoV-2 spike: evidence that D614G increases infectivity of the COVID-19 virus

More than an anticoagulant: do heparins have direct anti-inflammatory effects?

Low-dose low-molecular-weight heparin is anti-inflammatory during venous thrombosis

Using heparin molecules to manage COVID-2019

Anticoagulant treatment is associated with decreased mortality in severe coronavirus disease 2019 patients with coagulopathy

Heparin therapy improving hypoxia in COVID-19 patientsda cases series

Association of treatment dose anticoagulation with in-hospital survival among hospitalized patients with COVID-19

Practical diagnosis and treatment of suspected venous thromboembolism during COVID-19 pandemic

COVID-19 and coagulation: bleeding and thrombotic manifestations of SARS-CoV-2 infection

Webinar: thrombotic and hemostatic issues in critical care units managing COVID-19

Evidence-based practical guidance for the antithrombotic management in patients with coronavirus disease (COVID-19) in 2020