key: cord-0816100-r644rxe3
authors: Thachil, Jecko
title: Forum article: The protective rather than prothrombotic fibrinogen in COVID‐19 and other inflammatory states
date: 2020-06-03
journal: J Thromb Haemost
DOI: 10.1111/jth.14942
sha: d05e6bffab82c4e2e99b4e01577ab409892ce02b
doc_id: 816100
cord_uid: r644rxe3

Hypercoagulability has been recognised as a common complication of COVID‐19. Exact mechanisms for this extreme coagulation activation have not yet been elucidated. However, one of the consistent laboratory finding is the increase in fibrinogen, in some cases, marked elevation. High circulating levels of fibrinogen have been linked to thrombosis for years and for this reason, hyper‐fibrinogenemia is considered one of the mechanisms for COVID‐19 coagulopathy. In this forum article, instead of the prothrombotic role, a protective function for fibrinogen is discussed. Fibrinogen, like the other well‐known acute phase reactants, is increased in COVID‐19 possibly to protect the host.

COVID-19 continues to cause significant mortality all around the world. One of the key pathogenic features observed in COVID-19 is the intense inflammation induced by the severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) with the development of a cytokine storm in the most severe cases. From the coagulation perspective, colleagues from different parts of the world have noted several-fold increase in fibrinogen levels in many patients who require critical care support. 1, 2, 3 Why may the fibrinogen increase so high?

Fibrinogen is a glycoprotein produced in the body's major synthetic machinery, the liver, which is also an overlooked anti-infective organ. This is clearly evidenced in sepsis being one of the commonest causes of death in patients with liver cirrhosis. 4 Liver, in its anti-infective role, releases several acute phase reactants including fibrinogen, ferritin, C-reactive protein (CRP), and a plethora of cytokines. Several reports have confirmed marked elevation of these acute phase proteins in patients requiring hospitalisation for COVID-19 suggesting the liver function is in overdrive. 2, 3, 5 These different molecules including ferritin, CRP and fibrinogen play key roles in body's defence against invading pathogens.

Ferritin and the regulatory molecule, hepcidin are crucial in the iron-infection axis. 6 Every micro-organism needs iron for their survival. In infectious states, iron is sequestered away from the circulation inside ferritin; a process regulated by hepcidin, thus limiting the supply to the pathogens. 6 Clinically, this important role of ferritin was observed in the negative trials where routine iron supplementation for preschool children in a malaria-endemic population resulted in increased risk of severe illness and death. 7, 8 Another acute phase reactant, CRP also plays protective role in pro-inflammatory states with high likelihood of adult respiratory distress syndrome (ARDS). In patients with trauma, CRP was shown to defend the human body from histone-induced endothelial cell damage. 9 Mice experiments by this group showed CRP forms a complex with histones and protects the host from endothelial leakage/damage which would have resulted in lung

This article is protected by copyright. All rights reserved oedema, and pulmonary thrombosis (both commonly noted in COVID-19). 9 A third member of this beneficial group of proteins, haptoglobin works as an antioxidant, binds circulating extracellular haemoglobin in inflammatory states and can stimulate the monocyte/macrophage system. 10

Fibrinogen's role among these protective warriors is likely to be two-fold -regulating the antimicrobial function of the immune cells and secondly, clot formation thus limiting spread of the pathogen. Flick et al, showed fibrinogen to be a physiologically relevant ligand for the leukocyte integrin, Mac-1 and thus important in regulating the inflammatory response independent of the clotting function. 11 In relation to viral infections, Mac-1 is a surface receptor for extracellular double stranded RNA (SARS Co-V 2 is an RNA virus). 12 It is possible that high levels of fibrinogen saturate Mac-1 and therefore reduce the harmful effects from the virus. 13, 14 As an acute phase protein, increased concentrations of soluble fibrinogen antagonise leukocyte recruitment, and can contribute to inflammation resolution. 15 The several host defence roles of fibrinogen is summarised in a review where two main mechanisms are detailed, assisting host protective immune function and forming fibrin matrices which serve as a protective barrier. 16 

Thrombus formation limiting the spread of the invading pathogen has long been known to be a host defence mechanism. Different components of the coagulation pathway are highly relevant in the antiinfective process. These include the central figure, thrombin, anticoagulant proteins like protein C, and thrombomodulin, the contact pathway constituents like factor XII and of, course, fibrinogen. 17 The Flick lab attempted to distinguish between fibrinogen-and fibrin-dependent antimicrobial function in vivo by creating Fib AEK mice which lacked the capacity for fibrin polymer formation. 18 These animal models retained the functional capacity for platelet interactions but were unable to form polymers and clear intraperitoneal Staphylococcus aureus inoculum suggesting thrombus formation is also important in the antimicrobial process. 18 The localised thrombus formation in the lungs is most likely an attempt by the coagulation system to limit the spread of the SARS Co-V 2. Recent autopsy reports of patients with COVID-19 demonstrated this localised micro-thrombi in the lung specimens. 19

Elevated plasma fibrinogen in inflammatory states like trauma and pregnancy has been linked to an increased thrombotic risk, but without evidence for definite causation. In these clinical scenarios, increased coagulation activation including higher levels of fibrinogen is necessary for various physiological reasons, but primarily to limit ongoing (trauma) or impending (pregnancy) haemorrhage. However, despite the high levels of fibrinogen, all pregnant females do not develop thrombosis suggesting the hyper-fibrinogenemia is Accepted Article unlikely to be prothrombotic. Similarly, not all trauma or postoperative cases are associated with thrombosis. Indeed, fibrinogen has never been shown to play a direct, causative role in causing thrombosis in these conditions. 20 The Wolberg lab conducted animal experiments to demonstrate a link between high levels of fibrinogen and thrombosis and observed that hyper-fibrinogenemia did not cause spontaneous thrombosis in vivo. 20 They also quoted past observations where injecting human fibrinogen into mice was unable to cause spontaneous fibrin deposition. The conclusion from the study was that multiple hits were required to cause thrombosis in addition to elevated fibrinogen levels. 20 Certainly, in trauma patients, there can be several hits which can contribute to thrombosis, while in pregnant females; additional complications like pre-eclampsia or hyperemesis-induced dehydration could act as additional hits to trigger thrombosis.

If fibrinogen behaves as a protective acute phase reactant on one part and as a prothrombotic molecule on another part, could it be coming from two different sites to serve the two different roles? The "two-layer thrombus" model demonstrated by Stalker et al may provide clues to this duality. 21 Their experiments demonstrated a hierarchical organization for thrombus with an inner core of tight thrombus from activated platelets and an outer loose, plasma-permeable shell, less dependent on activated platelets. 21 Interestingly, in the inner core, the activated platelets are packed together by the integrin αIIbβ3, which is the platelet receptor for fibrinogen. The fibrinogen for the inner core of thrombus comes from the platelet α-granules which also store several coagulation proteins (factor V, XI and XIII) involved in secondary haemostasis. 22 At the same time, the circulating fibrinogen is possibly executing its efforts as an acute phase protein.

Although this dual role of fibrinogen has not been examined in detail, that of platelet and circulating Von Willebrand factor (VWF), another acute phase reactant, has been studied. Platelet α-granules contain 20% of the total VWF protein and are the high molecular weight multimeric forms. 22, 23 Using pig bone marrow transplantation model, it was shown that platelet VWF markedly ameliorates bleeding in severe Von Willebrands disease and in its presence, much lower levels of plasma VWF was required for haemostasis. 24 This suggests that the platelet VWF is the predominant form involved in thrombus formation; while the circulating VWF similar to fibrinogen, may be more of a participant in the host defence function in inflammatory conditions.

How can all the above information be put together? In a patient with COVID-19, the fibrinogen (and VWF) level increases as part of the host defence. In the initial stages, the main role is to regulate the body's intense inflammatory response and the exceedingly high fibrinogen level may not be a disadvantage. At this stage, fibrinogen's acute phase function is predominant to its role in thrombi formation, which is occurring at low-grade level (reflecting in mild increase in D-dimers). This scenario is similar to many other clinical situations where an acute phase response is seen like pregnancy, trauma and postoperative states,

where the rise in fibrinogen is physiological and an increase in D-dimers is noted without clinical evidence of thrombi. However, if the underlying inflammatory condition or host's inflammatory response progress unabated, the haemostatic system answers by causing widespread thrombi to limit dissemination of the harmful microbes or damage-associated proteins. Marked thrombi formation results in significant increase in D-dimers but also leads to exhaustion of platelet granules. Since fibrinogen is no more released from the platelets, its level starts to decrease in tandem with increasing D-dimers. This has clearly been shown in one of the most quoted papers in COVID-19 hypercoagulability literature from Tang et al. 1 In this report, a clear distinction between survivors and non-survivors were identified by the significantly elevated D-dimer and lower fibrinogen but not high fibrinogen levels. 1 The D-dimer/fibrinogen ratio has been explored in the past as a thrombotic marker. Kucher et al showed higher D-dimer values and lower fibrinogen levels in patients with pulmonary embolism and even suggested the use of D-dimer/fibrinogen ratio of greater than 10 3 as being highly specific for pulmonary clots. 25 The authors also commented on the possibility of decreasing fibrinogen levels correlating with increasing pulmonary occlusion rate. 25

What is the clinical relevance of the protective fibrinogen hypothesis? Monitoring fibrinogen levels along with D-dimers in critically patients with (and without) COVID-19 is clearly important. If the D-dimers do not increase in tandem with the fibrinogen, it may be hypothesized that the protective role of fibrinogen is dominating (acute phase response >>> thrombosis). However, if the D-dimers start increasing and the fibrinogen levels start decreasing, this is the time when marked coagulation activation and thrombi formation is occurring (Figure 1) . Unfortunately, clinicians only notice the fibrinogen levels when it's markedly reduced and the patient may have started bleeding, which may sometimes be an unsalvageable scenario. An interesting dilemma here is whether fibrinogen replacement should be given to maintain the higher levels rather than considering it when the levels have started decreasing, when such transfusions can be harmful. Unfortunately, most of the clinical studies on fibrinogen have focussed on bleeding outcomes with limited considerations looking at its benefits outside haemostasis.

So far, the protective role of fibrinogen as an acute phase reactant remains a hypothesis. Some basic science research suggests this role although clinical studies overlooked any function outside thrombosis causation, that too without definite proof. It is unusual for an evolutionary mechanism like an acute phase reaction to prove harmful to the host, especially when it occurs in scenarios like sepsis, and trauma which have affected humans from the early ages. It is time to start examining the non-haemostatic functions of fibrinogen in future research studies. In a similar way, it may be useful to study other coagulation proteins like the Von Willebrand factor for a possible protective role in the acute phase.

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

Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China

Clinical Characteristics of Coronavirus Disease 2019 in China

Management of Sepsis in Patients With Cirrhosis: Current Evidence and Practical Approach

Clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus-infected pneumonia in Wuhan, China

Hepcidin and the iron-infection axis

Effects of routine prophylactic supplementation with iron and folic acid on admission to hospital and mortality in preschool children in a high malaria transmission setting: community-based, randomised, placebo-controlled trial

Effect of routine prophylactic supplementation with iron and folic acid on preschool child mortality in southern Nepal: community-based, cluster-randomised, placebo-controlled trial

Human CRP defends against the toxicity of circulating histones

Haptoglobin, inflammation and disease

Leukocyte engagement of fibrin(ogen) via the integrin receptor alphaMbeta2/Mac-1 is critical for host inflammatory response in vivo

CD11b/CD18 (Mac-1) is a novel surface receptor for extracellular doublestranded RNA to mediate cellular inflammatory responses

The coagulation system in host defense

A unique recognition site mediates the interaction of fibrinogen with the leukocyte integrin Mac-1 (CD11b/CD18)

Acute-phase concentrations of soluble fibrinogen inhibit neutrophil adhesion under flow conditions in vitro through interactions with ICAM-1 and MAC-1 (CD11b/CD18)

Accepted Article This article is protected by copyright. All rights reserved

Fibrinogen Is at the Interface of Host Defense and Pathogen Virulence in Staphylococcus aureus Infection

Innate immunity and coagulation

Mice expressing a mutant form of fibrinogen that cannot support fibrin formation exhibit compromised antimicrobial host defense

Pathological evidence of pulmonary thrombotic phenomena in severe COVID-19

Causal relationship between hyperfibrinogenemia, thrombosis, and resistance to thrombolysis in mice

Hierarchical organization in the hemostatic response and its relationship to the platelet-signaling network

Platelet alpha-granules: basic biology and clinical correlates

Platelet von Willebrand factor: comparison with plasma von Willebrand factor

Transplantation of normal bone marrow into a pig with severe von Willebrand's disease