COVID-19 infection-associated coagulopathy: Pathophysiology and clinical implications

Introduction

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is a novel single-stranded enveloped RNA virus that is the fifth endemic human coronavirus to be discovered. ¹ Termed COVID-19, the clinical syndrome caused by the virus was identified in December 2019 in Wuhan, Hubei Province, China, in a cluster of patients who presented with respiratory symptoms due to severe pneumonia, many progressing to death. ¹ The high infectious rate of the virus soon rendered it a worldwide pandemic with high global mortality and morbidity which, at the time of writing, continue to climb unchecked. The SARS-CoV-2 infection has been associated with a wide array of presentation profiles ranging from asymptomatic carrier status to overt upper respiratory infection and hypoxemia to acute respiratory distress syndrome (ARDS) and multi-organ failure. A particularly pervasive complication that both accompanies and at times outlives the primary symptom complex is the development of COVID-19 associated coagulopathy (CAC).

COVID-19 pathophysiology

The virulence of COVID-19 is thought to be, in part, due to the high affinity of its surface spike protein for the human angiotensin-converting enzyme 2 (ACE2) receptor.^2,3 Recent studies focusing on the distribution of ACE2 receptors not only found it to be primarily localized to the endothelial cells of the intestines and lungs, but also found the receptor across vascular beds and organ systems including the nasopharynx and oropharynx, liver, kidney, brain, heart, and testis.^2–4 Therefore COVID-19 not only causes endothelial damage that is a cornerstone of CAC pathophysiology, but often manifests as multi-organ involvement in non-survivors and is believed to result in increased upregulation of the renin–angiotensin–aldosterone system (RAAS). ⁵ Currently, the largest multicenter study of early Chinese COVID-19 patients found the most common comorbidities associated with the clinically manifest disease were hypertension, diabetes, and cardiovascular disease. ⁴ This specific mode of COVID-19 cellular entry and association with RAAS activation is believed to play an important role in the pathogenesis of CAC.

COVID-19 associated coagulopathy (CAC)

CAC encompasses a wide range of hemostatic derangements and mechanistic pathways with myriad clinical manifestations. Key patient-specific risk factors, diagnostic considerations, clinical manifestations, and broad management strategies in patients with CAC, are summarized in Table 1. The first studies conducted in Wuhan, China, and later confirmed with larger global cohorts, demonstrated that patients with COVID-19 have systemic inflammation leading to increased levels of pro-coagulant factors, and simultaneous inhibition of intrinsic anticoagulant mechanisms, leading to in situ thrombosis and thromboembolism.^6–12 Hemostatic derangements most commonly associated with COVID-19 include elevated D-dimer, lactate dehydrogenase, and fibrin degradation products. Additionally, patients may have prolonged prothrombin time, elevations of interleukin-6 (IL-6) and C-reactive protein (CRP).^{4,6,8,13–17} COVID-19 has also been variably associated with mild thrombocytopenia as well as lymphopenia with thrombocytosis in more severe diseases.^6,11,18,19

Table 1.

Summary of patient-specific risk factors, diagnosis, clinical manifestations, and broad management considerations in patients with CAC.

Pathogenesis of CAC	Diagnosis of CAC	Clinical manifestations	Treatment of CAC
Serologic factors: - ↑ D-dimer, fibrinogen, prothrombin time, lupus anticoagulant, IL-6 - Endothelial damage via ACE2 and other pathways - ↑ vWF, factor VIII, RAAS activation Patient-specific factors: - Age and comorbidities (malignancy, CHF, HTN, obesity, smoking, metabolic syndrome) Immobility Severity of illness: - ARDS, DIC, CVC related thrombosis	Laboratory data: D-dimer, fibrinogen, prothrombin time, platelets Additional imaging: - For suspected symptomatic VTE - For diagnosis of arterial thrombosis (MI, stroke, systemic thromboembolism) - In response to sudden changes in laboratory data and/or clinical status - Hypoxia out of proportion to lung pathology on previous imaging	Cardiac: - Type I (spontaneous) MI - Microvascular obstruction (?) Neurologic: - Stroke Pulmonary: - PE Vascular: - DVT - VTE Hematologic: - DIC Other: - CRRT filter clotting - Clotting of ECMO circuit	General recommendations: - Prophylaxis dose AC for all patients hospitalized with COVID-19 (LMWH preferred inpatient) - Treatment dose AC for all patients diagnosed with VTE Additional considerations: - Treatment or escalated dose of AC for patients undergoing CRRT or ECMO - Outpatient prophylaxis with rivaroxaban post-discharge in high-risk patients

AC: anticoagulation; ACE2: angiotensin-converting enzyme-2; ARDS: acute respiratory distress syndrome; CAC: COVID-19 associated coagulopathy; CHF: congestive heart failure; CRRT: continuous renal replacement therapy; CVC: central venous catheter; DIC: disseminated intravascular coagulation; DVT: deep vein thrombosis; ECMO: extracorporeal membrane oxygenation; HTN: hypertension; IL-6: interleukin 6; LMWH: low molecular weight heparin; MI: myocardial infarction; PE: pulmonary embolism; RAAS: renin-angiotensin-aldosterone system; VTE: venous thromboembolism; VWF: von Willebrand factor.

Endothelial damage is another component of COVID-19 infection that sets the stage for CAC-associated thrombosis and thromboembolism.^7,20 Binding of COVID-19 to ACE2 receptors allows for the entry of the virus into human cells causing cell death, endothelial damage, and the release of endothelial procoagulant factors such as von-Willebrand factor (vWF) and factor VIII.^2,16,20 Furthermore, binding of the ACE2 receptor leads to upregulation of the RAAS releasing increased levels of circulating angiotensin II, which in turn, promotes systemic vasoconstriction, the release of tissue factor and plasminogen activator inhibitor 1. ²¹

The risk of developing thromboembolism due to CAC is further compounded by patient-related and factors related to the overall severity of the disease. ⁴ Studies looking at risk factors for the development of COVID-19 related thromboembolism have been somewhat inconsistent.^11,22 Previous studies (in non-COVID-19 patients) have shown however that comorbidities such as malignancy, congestive heart failure, hypertension, obesity, smoking, and metabolic syndrome are independent risk factors for venous thromboembolism (VTE) and one may reasonably assume some relevance to COVID-19 patient cohorts as well.^21,23,24 Importantly, immobility associated with travel restrictions and stay-at-home orders widely imposed during the early phase of the pandemic is thought to be an independent driver of increased VTE incidence, inclusive of non-COVID-19 infected individuals. One retrospective study looking at the incidence of pulmonary embolism (PE) in patients being imaged with computed tomography (CT) angiography in the emergency departments of New York City hospitals during two similar time periods in 2019 versus 2020, found a two-fold increase in the incidence of PE in 2020 regardless of COVID-19 status. ²⁵ In another study comparing those admitted to the intensive care unit (ICU) with influenza in 2019, ICU patients admitted in 2020 with COVID-19, manifested a two-fold increase in the risk of PE. ²⁶

Certain disease states and pneumonia-related factors further exacerbate the clinical impact of CAC through disease progression to ARDS and disseminated intravascular coagulation (DIC), which are recognized risk factors for thromboembolism.^27,28 Recent data shows that one in three hospitalized patients developing ARDS and up to 71% of COVID-19 non-survivors developing DIC.^29,30 Furthermore, ICU admission and mechanical ventilation contribute to patient immobility and central venous catheters can serve as the nidus for thrombus formation in critically ill patients. COVID-19 pneumonia also leads to hypoxic vasoconstriction, which when combined with pulmonary endothelial damage and CAC, makes PE both common and devastating in the setting of severe COVID-19 infection.^16,31

Diagnosis of CAC

Clinicians should, in general, maintain a high index of suspicion for VTE in patients with COVID-19 infection irrespective of disease severity, however, there is no consensus as yet on how to triage patients for advanced diagnostic imaging. PE in the setting of a COVID-19 pneumonia can be difficult to discern given that shortness of breath, chest pain, hypoxia, tachycardia, and elevated D-dimer can all be manifestations of pneumonia alone. Although it has been observed that COVID-19 non-survivors often manifest significantly higher levels of D-dimer than survivors (up to 3–4-fold higher in non-survivors), it remains unclear how much of this is directly ascribable to VTE versus other pathways.^24,32,33 Some researchers have recommended various D-dimer cutoff values as a trigger for mandatory imaging for VTE, however, studies to support this approach have limited patient numbers, and further research is needed to help optimize the balance between enhanced VTE detection and resource utilization/issues of equipment contamination, etc.^33,34 Combining elevated CRP and IL-6 levels with D-dimer has also been proposed as a strategy to increase specificity for VTE diagnosis in COVID-19 but these prediction models have not yet been widely adopted.^35–37 Nevertheless, most consensus statements from prominent international groups have agreed on avoiding routine use of CT scans and lower extremity ultrasounds for asymptomatic COVID-19 patients.^31,38–41 The approach endorsed by the International Society on Thrombosis and Haemostasis (ISTH) and National Institutes of Health (NIH) is obtaining levels of D-dimer, fibrinogen, prothrombin time, and platelet levels on admission, and then to investigate for VTE if these labs are abnormal or the severity of disease hypoxia is out of proportion to lung injury on available imaging.^31,41

Clinical manifestations of CAC

It is now well-established that CAC increases the risks of both venous and arterial thrombosis. Early studies from Wuhan, China cited concerning high rates of VTE in hospitalized patients with COVID-19, ranging from 25% to 46%.^8,11,15 A more recent study from New York City reviewed 3334 inpatients with COVID-19, only found a 16% incidence of combined VTE and arterial thrombosis (ischemic stroke, myocardial infarction, and systemic arterial thromboembolism). ⁴² A similar but smaller study with 1114 patients from Boston-area hospitals highlighted differences in VTE and arterial thrombosis occurrence between treatment settings. Anticoagulated COVID-19 patients who were admitted to the ICU had a higher risk of arterial thrombosis (35%) and VTE (46%) when compared to patients hospitalized in non-ICU care (3% and 6%, respectively). ¹² Outpatients, on the other hand, had no arterial or venous thrombosis within 30 days of a COVID-19 diagnosis. ¹² Studies have also shown the rate of stroke in COVID-19 to be about 2% and myocardial injury/infarction to be around 9%. ⁴² Nevertheless, the incidence of VTE and arterial thrombosis has varied greatly between different countries and treatment settings. Despite the severe coagulopathy that can develop with COVID-19, clinically relevant bleeding during the course of illness remains a rare occurrence, but clinicians should nevertheless remain vigilant especially in patients who are being therapeutically anticoagulated. ³¹

In addition to the aforementioned CAC-associated manifestations of vascular thrombosis, COVID-19 infection is also associated with myocardial injury manifested by elevated troponin levels and seen in up to 20% to 30% of patients hospitalized with COVID-19.^12,43,44 Mechanistically, this may be related to microvascular occlusion or may be the result of myocardial inflammation and cell death. Indeed, it has been noted that COVID-19 infection can lead to viral myocardial infiltration and even fulminant myocarditis in rare cases. ⁴⁵ It has been speculated that irrespective of the mechanism, the true incidence of myocardial injury is even higher than current estimates. Although there was a 30% to 40% reduction in global volumes of acute coronary syndromes (ACS)/acute myocardial infarction (AMI) during the first wave of the pandemic, there were also commensurate rises in reported rates of out of hospital cardiac arrest, signaling that perhaps patients were suffering ischemic events at home, unable or perhaps unwilling to present for emergency medical care.^46–48

Candidate therapies

Although therapeutic-dose anticoagulation is widely recommended for diagnosed cases of COVID-19-associated thrombosis/thromboembolism, no study to date has shown the differential benefit of one anticoagulant over another. ³⁸ Key pharmacologic differences between the commonly used classes of antithrombotic medications, along with accepted clinical applications, are highlighted in Table 2. The duration of treatment has also been debated, however, current anticoagulation guidelines for provoked thrombosis lend guidance in this regard with a recommended minimum of 3 months of anticoagulation. ³⁸ Both direct oral anticoagulants (DOACs), as well as vitamin K antagonists, have been used on an outpatient basis, while inpatients are preferentially treated with parenteral agents such as unfractionated heparin (UFH) and low molecular weight heparin (LMWH). UFH has been shown in preclinical models to reduce sepsis-induced acute lung injury while LMWH has been found to limit IL-6 activity, thus reducing systemic inflammation.^49,50 In the setting of UFH use, some authors have suggested that anti-Xa levels, rather than activated partial thromboplastin time, may provide a more reliable and reproducible index of anticoagulant effect in patients with CAC. ⁵¹ Some centers have empirically increased the intensity of prophylactic anticoagulation, however, there is no set criteria or evidence to support which patients should receive this intensified treatment.^39,41 Finally, drug–drug interactions should be monitored as several antiviral therapies currently being used or investigated can accelerate the enzymatic activity of DOACs and antiplatelet therapies. ⁴¹

Table 2.

Classes of anticoagulants and antiplatelets potentially used in the treatment of thrombotic conditions occurring in COVID-19 patients, coupled with mechanisms of action and common clinical applications.

Drug (route of administration)	Drug class	Mechanism of action	clinical use
UFH (SC and IV)	Indirect antithrombin	Binds antithrombin (antithrombin III) to inactivate factor Xa and reduce the conversion of prothrombin (factor II) into thrombin (factor IIa)	Used for inpatient prophylaxis and anticoagulation for VTE. May also be used to treat arterial thrombosis across a variety of disease states and vascular beds.
Low molecular weight heparin (SC)	Indirect antithrombin	Binds antithrombin to increase its inactivation of factor Xa; more homogeneous molecular size/ lower molecular weight than UFH resulting in greater anti-Xa:anti-IIa activity	Used for inpatient and outpatient prophylaxis and anticoagulation for VTE.
Fondaparinux (SC)	Synthetic pentasaccharide factor Xa inhibitor	Binds with antithrombin to selectively inactivate factor Xa	Used for inpatient prophylaxis and anticoagulation. Reduced risk of HIT compared to heparinoids.
Rivaroxaban (PO), apixaban (PO), edoxaban (PO), and betrixaban (PO)	Direct oral factor Xa inhibitors	Directly binds and inhibits factor Xa without requiring other enzymes or cofactors	Used primarily for inpatient/outpatient prophylaxis and anticoagulation in a variety of venous thrombotic conditions including critically medically ill patients (rivaroxaban). Prophylactic rivaroxaban may be used following discharge in medically critically ill patients
Dabigatran (PO), argatroban (IV), and bivalirudin (IV)	Direct thrombin (IIa) inhibitors	Directly binds and inactivates thrombin (factor IIa) without requiring other enzymes or cofactors	Dabigatran may be used for outpatient anticoagulation in VTE. Argatroban is mainly used in patients at risk of or diagnosed with HIT.
Abciximab (IV), eptifibatide (IV), and tirofiban (IV)	Platelet glycoprotein IIb/IIIa inhibitors	Bind platelet surface GpIIb/IIIa receptors to prevent fibrinogen-mediated platelet aggregation	Used primarily to treat coronary thrombosis in the context of acute coronary syndromes/acute myocardial infarction and in conjunction with percutaneous coronary intervention.
Streptokinase (IV) and tissue plasminogen activator (IV)	Fibrinolytics	Catalyzes cleavage of plasminogen to plasmin with disruption of clot-bound fibrin molecules	Used in ischemic stroke and acute myocardial infarction if immediate catheter-based intervention is not available.

Note: Some FDA-approved indications for the medications listed and certain available agents have been purposely omitted in the interest of brevity and as they do not directly apply to COVID-19 patients.

HIT: heparin-induced thrombocytopenia; IV: intravenous; PO: per os (oral); SC: subcutaneous; UFH: unfractionated heparin; VTE: venous thromboembolism.

Consensus holds that all hospitalized COVID-19 patients should receive some form of prophylactic anticoagulation. ⁴¹ There is currently insufficient evidence to support the routine use of prophylactic anticoagulation in stable outpatients with COVID-19, an observation upheld in the NIH COVID-19 treatment guidelines. ⁴¹ It is known, however, that increased risk of VTE persists for many weeks post-discharge following hospitalization for any critical medical illness including COVID-19.^41,52,53 In 2019, the Food and Drug Administration (FDA) approved oral rivaroxaban (10 mg once daily for up to 39 days) for post-discharge in patients with a high risk of VTE and low bleeding risk, an indication which may be considered in inpatients being discharged following COVID-19 infection. ⁵³ However, the decision to prophylactically treat for VTE needs to be an individualized determination with due consideration for the likelihood of VTE and risk of bleeding. ⁴¹

There are several specific clinical scenarios that have arisen over the past year of COVID-19 associated VTE without consensus on the approach to anticoagulation. First is the increased risk of thromboembolism with ARDS, which is a well-documented clinical entity prior to COVID-19. Some very small series have trialed intensified anticoagulation and/or thrombolysis in ARDS with the goal of avoiding the downward spiral of pulmonary vascular thrombosis. ²⁷ Data from one case series of three patients showed some transient improvement in respiratory status in ARDS after treatment with thrombolysis, however, only one patient had a durable response to treatment, and more trials are required before this becomes an accepted treatment for severe ARDS. ⁵⁴ The second scenario is clotting associated with continuous renal replacement therapy (CRRT), which is utilized in up to 3.2% of patients admitted with COVID-19. ⁵⁵ It is well established that CRRT is associated with significant rates of circuit clotting and many centers empirically anticoagulate all patients undergoing CRRT.^51,56 However, despite anticoagulation, up to 83% of patients with COVID-19 can have thrombosis of their CRRT filter within a median time of 6.5 h, leading to blood loss, worsening acidosis, volume overload, and electrolyte abnormalities while off CRRT. ⁵¹ Given this high rate of filter thrombosis, it has been widely accepted that patients with COVID-19 undergoing CRRT should be therapeutically anticoagulated. ⁵⁴ Similar considerations may extend to COVID-19 patients placed on extracorporeal membrane oxygenation (ECMO) and other forms of mechanical circulatory support. ⁴¹ The third scenario is the navigation of VTE in patients with concomitant DIC, which can occur in nearly 10% of COVID-19 patients. ²⁹ ISTH recommends transfusing non-bleeding DIC patients to a platelet count above 25 × 10⁹/l. In bleeding DIC patients, the recommendation stands to transfuse blood products to maintain platelets >50 × 10⁹/l, fibrinogen >1.5 g/l, and prothrombin time ratio below 1.5 s, in addition to discontinuing anticoagulation. ³¹ Finally, mechanical thromboprophylaxis for VTE should be a mainstay of treatment for all patients hospitalized during this pandemic.

Implications for vascular interventional procedures

As detailed in the above passages, patients with COVID-19 infection, with or without concomitant CAC, often suffer thrombotic events with evidence of end-organ jeopardy or damage. Particularly as it relates to acute myocardial infarction, stroke, and to a lesser extent, PE, catheter-based therapies may be acutely required for reperfusion of the thrombosed vascular bed. Procedural considerations may be broadly divided into the management of antecedent anticoagulant/antiplatelet therapy leading up to the procedure, choice of antithrombotic regimen during the intervention and mitigation of bleeding risk after the procedure. Although it is beyond the scope of this review to explore these nuanced and complex issues in detail, it goes without saying that anticoagulant and often, antiplatelet therapies, are necessarily intensified in the periprocedural period, enhancing the risks of both access site and non-access site bleeding, especially in critically ill patients who have already been instrumented. In the context of percutaneous coronary intervention for ACS and AMI, it has been anecdotally observed by the authors and by others that the burden of vascular thrombus encountered in COVID-19 patients is often more extensive and qualitatively different/more resistant than that seen with similar cases in uninfected patients. Use of platelet glycoprotein IIb/IIIa inhibitors and potent P2Y₁₂ inhibitors may be considered in such cases. Finally, proceduralists must take into consideration the thrombocytopenia and DIC that may ensue over the course of severe COVID-19 illness, vis-a-vis tolerance to antiplatelet therapy after placement of arterial stents.

Conclusions

CAC is increasingly recognized as a source of morbidity and mortality in hospitalized patients with COVID-19 infection. Guidance documents issued by the NIH and ISTH call for monitoring of D-dimer, prothrombin time, fibrinogen, and platelets and investigating for VTE if there are changes in these parameters. LMWH is the preferred agent for inpatient prophylaxis of most inpatients with COVID-19. Vigilance for the development of VTE should continue well into the post-discharge period with due consideration for prophylaxis with rivaroxaban in patients who have recovered from critical medical illness. Although increased thrombosis risk has not been credibly linked to mild COVID-19 illness being managed on an outpatient basis, it bears recognition that our collective knowledge in this field is rapidly evolving and there may yet be associations we are currently unaware of. It should also be noted that immobility, whether related to a period of convalescence or due to stay-at-home orders affecting all individuals, irrespective of COVID-19 status, may increase the risk of VTE. If a diagnosis of VTE is made, a minimum of 3 months of anticoagulation should be prescribed with longer durations of treatment at the discretion of the treating clinician and with careful consideration of bleeding risk. Full dose (not prophylactic) anticoagulation should be considered in all COVID-19 patients receiving CRRT. Finally, proceduralists should be cognizant of qualitative and quantitative differences in vascular thrombosis and should make therapeutic choices with deference to these issues as well as to tolerance for antithrombotic therapies, vis-a-vis future risks of thrombocytopenia, DIC, and bleeding.