Background
COVID-19 is an infection induced by the SARS-CoV-2 coronavirus affecting mostly adults. This viral disease is mild in many patients but is characterized in symptomatic forms by an atypical interstitial inflammatory lung disease [
1]. In addition, severe forms are associated with an extreme inflammatory reaction related to a “cytokine storm,” with lung involvement that can lead to acute respiratory distress syndrome (ARDS) requiring intensive care unit (ICU) management.
In critically ill patients, especially those with hypoxemia, coagulation changes reflecting inflammation are generally observed, with increased D-dimer and fibrinogen levels [
2] and, more rarely, a consumptive coagulopathy associated with a poor prognosis [
2].
The pathophysiology of SARS-Cov2 infection is still poorly defined, but major inflammation and hypoxemia associated with a prothrombotic state are significant features of severe forms. Chinese, Italian, North American, and French cohorts have consistently reported that severe forms affect more often elderly patients with comorbidities (hypertension, diabetes, cardiovascular or pulmonary pathology), with high mortality in those requiring ICU admission [
3‐
6].
More recently, the impact of obesity, often associated with other comorbidities, has been highlighted in severe forms of COVID-19 [
7]. A French study has also confirmed that almost half of the patients admitted to ICU are obese (with BMI > 30 kg/m
2) and require mechanical ventilation more often [
8].
A Chinese study reported frequent venous thrombotic episodes in severe COVID-19 [
9], and survival was improved with heparin thromboprophylaxis [
2]. In another report, venous thromboembolic events occurred in 27% of 187 Dutch patients with COVID-19 hospitalized in the ICU [
10]. Further, reports from Italy, France, and Switzerland have observed frequent venous thromboembolic complications in COVID-19 with a risk that appears particularly high in patients requiring ICU admission and/or with obesity, and frequent clotting of indwelling catheters, dialysis filters, ECMO oxygenators, and arterial thrombotic events including acute limb ischemia or stroke. In addition, pulmonary embolism has recently been identified as the most common thrombotic event occurring despite thromboprophylaxis [
11,
12].
However, no study has formally documented an increased thrombotic risk in COVID-19 compared to other severe infections, nor demonstrated that this risk was associated with a poor prognosis. Nevertheless, some pathophysiological features (major inflammation in particular) and the populations affected by this pathology (with comorbidities, particularly obesity) lead to further debates on the specific thromboprophylaxis treatment modalities for COVID-19.
Although low molecular weight heparin (LMWH) is the therapy of choice in patients with severe infections, current dosing strategies [
13,
14] may be inadequate in patients with increased inflammatory responses and obesity and critically ill in the ICU. In addition, we are confronted with a highly challenging medical situation, with a large influx of severe patients in ICUs and in whom it is much more difficult to confirm a diagnosis of thrombosis, which explains why the implementation of curative anticoagulant treatment can sometimes be discussed on a probabilistic basis.
In this context, given the paucity of available data, the Groupe d’intérêt en Hémostase Périopératoire (GIHP) and the Groupe Français d’études sur l’Hémostase et la Thrombose (GFHT) have developed a guidance document on the prevention of thrombosis and monitoring of hemostasis in hospitalized patients with COVID-19 in order to provide support for clinical management.
These proposals, developed by a large group of reviewers, members of the GIHP and the GFHT, are organized into 4 objectives and will be modified according to the evolution of our knowledge on COVID-19.
Discussion
With these practical proposals, we aimed to assist clinicians in the therapy and monitoring of anticoagulation to prevent thrombosis in hospitalized patients with COVID-19. As a result, we did not address the management of a consumptive coagulopathy, which can occur in severe patients, nor certain situations such as ambulatory patients, pregnancy, or patients with underlying diseases (e.g., sickle cell disease, congenital hemostatic defects), as well as the role of thrombolysis, antiplatelet drugs, and the management of arterial thrombotic events.
We first proposed defining the level of thrombotic risk according to the severity of the disease, i.e., the presence or absence of hypoxemia and BMI (Table
1). Other additional risk factors for thromboembolism such as recent thrombosis or active cancer must also be considered.
Table 1Thrombotic risk levels in patients with COVID-19 according to BMI, requirement of O2 or mechanical assistance, and other risk factors of thrombosis
Non-hospitalized patient with BMI < 30 kg/m2 and no added risk factors for thromboembolism (such as active cancer, recent history of thrombosis) | BMI < 30 kg/m2, without the need for high-flow nasal oxygen therapy or mechanical ventilation, with or without added risk factors for thromboembolism | - BMI < 30 kg/m2, under high-flow nasal oxygen therapy or mechanical ventilation, with or without risk factors for thromboembolism - BMI > 30 kg/m2 without high-flow nasal oxygen therapy or mechanical ventilation, but with added risk factors for thromboembolism - BMI > 30 kg/m2 with high-flow nasal oxygen therapy or mechanical ventilation, and without added risk factors for thromboembolism | - BMI > 30 kg/m2 with added risk factors for thromboembolism, AND high-flow nasal oxygen therapy or mechanical ventilation - ECMO (venovenous or veno-arterial) - Unexplained catheter thrombosis - Dialysis filter thrombosis - Marked inflammatory syndrome and/or hypercoagulability (e.g., fibrinogen > 8 g/l (800 mg/dl) and/or D-dimers > 3 μg/ml or 3000 ng/ml) |
Antithrombotic prophylaxis is not required in case of low risk, but is necessary in all hospitalized and immobile patients, as for any severe acute infection [
13,
14], with preference given to heparins. Interestingly, a specific beneficial effect of heparin on COVID-19 is also suggested, due to its pleiotropic actions (cytokine binding, inhibition of chemotaxis, leukocyte migration, and complement activation, sequestration of inflammatory proteins) [
15].
In the less severe forms and if the BMI is < 30 kg/m
2 (intermediate risk), it is recommended, in accordance with the most recent guidelines [
14], to prescribe fondaparinux or a LMWH at a standard dose (e.g., enoxaparin 4000 IU/24 h SC).
But in obese patients (BMI > 30 kg/m
2), who present a higher thrombotic risk, we suggest increased doses of LMWH, in accordance with ESC proposals in other scenarios [
16]. However, in the absence of data on the bleeding risk associated with higher doses of LMWH in the context of COVID-19, it is preferable not to initially exceed 10,000 IU/12 h of enoxaparin SC.
In ICU patients, the thrombotic risk is frequently high [
17], especially in case of sepsis [
18], and prophylaxis at usual doses may be ineffective, and more importantly in case of obesity [
19]. Heparin therapy at therapeutic doses had been proposed during the H1N1 influenza, due to an increased thrombotic risk, particularly in patients with ARDS [
20]. Moreover, the incidence of severe hemorrhagic accidents is rather low in COVID-19 patients [
21,
22], whereas the thrombotic risk is high despite anticoagulation, as recently documented in 2 French cohorts [
11,
12]. We therefore propose increased or even therapeutic doses (if additional thromboembolic risk factors are present) of LMWH or unfractionated heparin in COVID-19 patients hospitalized in ICUs. Thus, in obese patients with well-identified thromboembolic risk factors (active cancer or recent thrombosis in particular), therapeutic doses of LMWH adjusted according to actual weight [
23] will be preferentially administered. Therapeutic dosing of heparin is also indicated in cases of thrombosis or during ECMO. However, it should also be considered in cases of a major inflammatory syndrome associated with severe pulmonary involvement, which often results in elevated fibrinogen and/or D-dimer concentrations, also indicating a severe prothrombotic state, regardless of BMI. The increase in D-dimer levels in COVID-19 is more pronounced in the most critically ill patients [
24], with a concentration often greater than 3 μg/ml [
2]. In addition, mortality was lower in patients whose D-dimer levels were above this threshold, when treated with heparin [
25]. Therefore, these variables should be measured regularly, and a rapid increase in the level of D-dimers (outside of DIC) and/or fibrinogen should prompt to search for a thrombotic event and discuss the administration of therapeutic doses of heparin. In practice, an intensification of heparin treatment should be considered based on clinical and biological considerations of severity, but such a proposal is debated [
26,
27], and clinical trials are mandatory. However in practice, when evaluation of thrombosis is not feasible, such as in critically ill ventilated patients, a probabilistic antithrombotic intensification should be considered when there is a high thrombotic risk.
Biologically, we propose to monitor prothrombin time, APTT, fibrinogen, and D-dimers at least every 48 h, combined in patients treated with UFH with monitoring of platelet count and anti-Xa activity (if curative doses and/or renal failure). This regular monitoring of hemostasis is crucial in all COVID-19 patients hospitalized with two main objectives as follows:
1.To evaluate the evolution of the disease using predictive markers of complications
2.To screen for heparin overdose, associated with a risk of bleeding, and heparin-induced thrombocytopenia (HIT) in UFH-treated patients
Anti-Xa activity, although not correlated with the occurrence of thrombosis or bleeding, especially in case of obesity [
28], should be measured in patients receiving increased doses of LMWH, but is only intended to eliminate overdose (Fig.
1). Dosage adjustment to achieve a target value is therefore not recommended. In patients treated with UFH, dose adjustment based on anti-Xa activity is recommended [
29]. Thus, we propose to maintain anti-Xa activity between 0.3 and 0.5 IU/ml during enhanced prophylactic treatment and between 0.5 and 0.7 IU/ml during therapeutic treatment.
Platelet count should be monitored every 48 h between days 4 and 14 of treatment with UFH to detect HIT [
30]. A relative heparin resistance has been described in COVID-19 [
24], but infrequently. Rarely associated with a modest decrease in antithrombin level [
24], it can be resolved in most cases by increasing heparin doses. Antithrombin supplementation, which can increase the risk of bleeding in ICU patients [
31,
32], is not recommended [
33], as well as argatroban, which should be reserved only for patients with HIT [
30,
34].
In case of unexplained and/or atypical thrombosis, antiphospholipid antibodies can also be detected [
35] and could be frequent in COVID-19 [
11]. However, these antibodies are often non-pathogenic in severe infections, and their persistence should therefore be checked after COVID-19 recovery.
Finally, monitoring the hemostasis of COVID-19 patients is essential for the diagnosis of DIC, which complicates severe forms [
2], and should be suspected with classic manifestations including thrombocytopenia, prolongation of APTT and PT, decrease in fibrinogen levels, elevation of D-dimers, and the presence of fibrin monomers.
Acknowledgements
The authors gratefully thank all following members of the GIHP and the GFHT for their help in defining these proposals.
Members of the “Groupe d’intérêt en hémostase périopératoire” (GIHP): P. Albaladejo (Anesthesia and Critical Care, Grenoble France), N. Blais (Hematology-Hemostasis, Montréal, Canada), F. Bonhomme (Anesthesia and Critical Care, Geneva, Switzerland), A. Borel-Derlon (Hematology-Hemostasis, Caen, France), A. Cohen (Cardiology, Paris, France), J.-P. Collet (Cardiology, Paris, France), E. de Maistre (Hematology-Hemostasis, Dijon, France), P. Fontana (Hematology-Hemostasis, Geneva, Switzerland), D. Garrigue Huet (Anesthesia and Critical Care, Lille, France), A. Godier (Anesthesia and Critical Care, Paris, France), Y. Gruel (Hematology-Hemostasis, Tours, France), A. Godon (Anesthesia and Critical Care, Grenoble, France), B. Ickx (Anesthesia and Critical Care, Brussels, Belgium), S. Laporte (Clinical Pharmacology, Saint-Etienne, France), D. Lasne (Hematology-Hemostasis, Paris, France), J. Llau (Anesthesia and Critical Care, Valencia, Spain), G. Le Gal (Vascular Medicine, Ottawa, Canada), T. Lecompte (Hematology-Hemostasis, Geneva, Switzerland), S. Lessire (Anesthesia and Critical Care, Namur, Belgium), J.H. Levy (Anesthesia and Critical Care, Durham, USA), D. Longrois (Anesthesia and Critical Care, Paris, France), S. Madi-Jebara (Anesthesia and Critical Care, Beyrouth, Lebanon), A. Mansour (Anesthesia and Critical Care, Rennes, France), M. Mazighi (Neurology, Paris, France), P. Mismetti (Clinical Pharmacology, Saint-Etienne), P.E. Morange (Hematology-Hemostasis, Marseille, France), S. Motte (Vascular Medicine, Brussels, Belgium), F. Mullier (Hematology-Hemostasis, Namur, Belgium), N. Nathan (Anesthesia and Critical Care, Limoges, France), P. Nguyen (Hematology-Hemostasis, Reims, France), G. Pernod (Vascular Medicine, Grenoble, France), N. Rosencher (Anesthesia and Critical Care, Paris, France), S. Roullet (Anesthesia and Critical Care, Bordeaux, France), P.M. Roy (Emergency Medicine, Angers, France), S. Schlumberger (Anesthesia and Critical Care, Suresnes, France), P. Sié (Hematology-Hemostasis, Toulouse, France), A. Steib (Anesthesia and Critical Care, Strasbourg, France), S. Susen (Hematology-Hemostasis, Lille, France), C.A. Tacquard (Anesthesia and Critical Care, Strasbourg, France), S. Testa (Hematology, Cremona, Italy), A. Vincentelli (Cardiac Surgery, Lille, France), and P. Zufferey (Anesthesia and Critical Care, Saint- Etienne, France).
Members of the “Groupe Français d’études en Hémostase et Thrombose” (GFHT):
A. Borel-Derlon (Hematology-Hemostasis, Caen), E Boissier (Hematology-Hemostasis, Nantes), B Dumont (Hematology-Hemostasis, Paris), E. de Maistre (Hematology-Hemostasis, Dijon), Y. Gruel (Hematology-Hemostasis, Tours, France), C James (Hématology, Bordeaux), D. Lasne (Hematology-Hemostasis, Paris), T. Lecompte (Hematology-Hemostasis, Genève, Suisse), P.E. Morange (Hematology-Hemostasis, Marseille), P. Nguyen (Hematology-Hemostasis, Reims, France), P. Sié (Hematology-Hemostasis, Toulouse), V. Siguret (Hematology-Hemostasis, Paris), and S. Susen (Hematology-Hemostasis, Lille, France).