Pathophysiology
No coherent explanation for these protean manifestations is on offer; that is, no encompassing pathophysiology for this disorder has been advanced that accounts for such widespread damaging effects. We propose that, in severely ill patients, Covid-19 produces these diverse, malign outcomes by partial expropriation of the clotting cascade. Other than the signs and symptoms of a viral pneumonitis, we believe that it is this expropriation of clotting and the sequelae caused by this anomalous sort of clotting that accounts for almost all of the findings in Covid-19 that are unexpected in a straightforward viral pneumonitis. In corroboration of this proposed mechanism, heparin administration, which has become routine over the last few months, has significantly decreased Covid-19 lethality. Furthermore, almost all severely ill patients show a marked rise in D-dimers which is unequivocal evidence that clotting has occurred somewhere in the body.
Several days into a severe Covid-19 infection, “exudate” macrophages that have assembled in response to a SARS-CoV-2 infection of alveolar lung epithelium attack and destroy the virally infected cells. In mouse models of lethal influenza virus pneumonia this “attack of the macrophages” occurs between the fifth and the seventh day of infection.3 There is good evidence that, in Covid-19, a similar process takes place. That “good evidence” consists of a bloodstream blizzard of oligomeric fibrin that in severe Covid-19 pneumonia typically makes an appearance seven to ten days into the viremia. It is significant to note that in the animal viral pneumonitis model, where the macrophage attack occurs between day five and day seven, the attack takes place two to three days after the influenza virus titer has peaked; it is, therefore, not a direct response to viral overload. However, once the “bloodstream blizzard” of oligomeric fibrin appears, the pathophysiology of Covid pneumonia is no longer primarily that of a viral disease; it is now largely that of a clotting disorder.
Fibrin results from the enzymatic action of thrombin on fibrinogen. Where can this thrombin be coming from? Given the destruction of respiratory epithelium that Covid-19 produces, the likeliest source is the dead and dying alveolar lining cells under attack by thromboplastin-rich alveolar macrophages.4 Thromboplastin, also referred to as Tissue Factor (TF) initiates the extrinsic clotting pathway. Once initiated by TF, the extrinsic clotting pathway can produce thrombin in only a few seconds. If this is, indeed the source of the thrombin, it raises another question: why does this thrombin produce a “blood-stream blizzard“ of oligomeric (short-chain and hence soluble) fibrin rather than the usual large, vessel-occluding clot?
This atypical clotting is, very likely the result of minute quantities of thrombin entering the bloodstream again and again—for repeatedly introducing minute quantities of thrombin into vigorously stirred, anticoagulated blood does produce fibrin oligomers of short enough chain length that they remain soluble. If the process is continued, a “blood-stream blizzard” of oligomeric fibrin in quantities similar to that seen in Covid-19 patients can be recreated on the laboratory bench. From this evidence it seems likely that, as each alveolus succumbs under Covid-19 attack, a minute amount of thrombin will result from activation of clotting initiated by thromboplastin-rich macrophages and other cellular debris. As this thrombin—and likely some thromboplastin as well—enters the pulmonary circulation, the small amount of the enzyme (thrombin) will result in only a few molecules of the product (fibrin) before it is diluted by ongoing blood flow.
This anomalous type of clotting has drawn scant attention thus far, for at least two reasons. 1. While it is underway there are often no immediate associated clinical symptoms. The lack of clinical symptoms occurs because most of the newly produced fibrin molecules remain, like the antecedent fibrinogen molecules, in solution. 2. When the earliest microclots do appear, the occlusion of arteriolar and/or capillary-sized vessels in widely scattered regions throughout the body will produce few localizing symptoms.
That microclots form, circulate and occlude vessels has been confirmed by supravital capillaroscopy in ventilator dependent Covid-19 patients.5 As short-chain fibrin molecules lengthen by polymerization they become too long to remain in solution. It is microclots of this size—on the borderline between soluble and insoluble fibrin—that will first appear. Microclots of this size are too small to be detected by routinely available, non-invasive, diagnostic techniques.
Because the fibrin binding sites on short-chain fibrin oligomers are complexed to complementary sites on native fibrinogen molecules, they are not readily accessible to other short-chain fibrin molecules. This inaccessibility slows the growth of fibrin protofibrils and the formation of fibrin clots. These molecular aggregates between short-chain fibrin and native fibrinogen molecules are known as Soluble Fibrin Monomer Complexes6 and, for convenience will be referred to as Soluble Fibrin (SF). Test methods for detection of SF are not readily available, and the few methods that are available are not offered by most clinical laboratories.
Routine clotting assays such as those for quantifying fibrinogen levels will make no distinction between native fibrinogen and SF as the differences between intact fibrinogen molecules and SF molecules are minute. However, these tiny molecular changes result in a profound difference in molecular behavior. Without further modification, if the bloodstream concentration of SF is increased it will polymerize and precipitate out of solution, randomly forming clinically undetectable microclots throughout the body. The other molecular species (fibrinogen) is a normal protein constituent of blood and unless it is further modified it will circulate for days before being degraded and replaced.
In only a minority of COVID-19 patients do the high levels of SF lead to the formation of large, branching, three-dimensional, fibrin polymers—polymers so large that they show up in the vasculature as clinically recognized macroscopic clots. More often—even though some portion of the SF may have already achieved the size of fibrin protofibrils7—macroscopically visible clots do not form. The miniscule clots that do form are fragile, and, for the most part, rapidly dissolved by plasmin. The only evidence of their transient existence may be a rise in D-dimers over the next several hours. However, if a “bloodstream blizzard” of SF has supervened, the subsequent rise in D-dimers will not be subtle. Often it will be so extreme as to dramatically exceed the upper limits of the usually-reportable range for the D-dimer assay. In Covid-19 patients, D-dimers can remain at extremely high levels for 100 hours or more. Once this has occurred, death is extremely likely. However, it may be delayed for days or weeks while tissue downstream from occluded arterioles/capillaries in organ systems throughout the body undergoes coagulation necrosis. Death, when it does supervene, is often so far removed in time that it may appear unrelated to the episode of anomalous clotting that initiated the rapid rise in SF.
At least two clinical situations are known where soluble fibrin or some near equivalent is produced but where visible clots do not form. The first clinical situation arises because the goal is precisely defibrination. The intent is to deplete the body’s entire supply of fibrinogen for therapeutic purposes—for anticoagulation. The agent employed is Ancrod8, a procoagulant extract of the Malayan pit-viper venom. When this extract is administered (very slowly and in very carefully limited amounts) to adult humans, it—like thrombin—transforms fibrinogen into fibrin. However, the venom extract exposes active fibrin binding sites by cleaving only fibrinopeptide A from the fibrinogen molecules (as opposed to thrombin’s cleavage of both fibrinopeptides A and B). Furthermore, Ancrod does not activate Factor XIII, so stabilization of fibrin clots does not take place. Consequently, the clots that do form are fragile and break up almost immediately. Although the goal of defibrination is readily achieved there is scant evidence that Ancrod-induced defibrination is therapeutically useful.9,10
A second clinical situation involving SF is the disseminated intravascular coagulation—DIC—that can develop (fortunately briefly) in patients during liver transplantation surgery. If SF is going to appear at all, it will usually make its appearance just after reperfusion of the transplanted liver, when residual dead and dying cells from the transplanted liver are washed into the transplant recipient’s vasculature. (This is occasionally and catastrophically accompanied by the sudden appearance of large clots in the heart and great vessels.) If the newly transplanted liver is healthy and begins functioning immediately, the SF levels will generally drop back towards the normal range over the next 60 minutes or so.
A similar mechanism appears to be responsible for the generation of SF in patients with severe Covid-19 pneumonia. SARS-CoV-2 is a respiratory virus; it attacks respiratory epithelium. It typically causes necrosis of the infected cells as the cellular machinery is diverted to viral replication and the host cell dies. Should hemorrhage into the damaged alveolus occur (at autopsy in severe Covid-19 pneumonia alveolar hemorrhage is widespread), dead and dying alveolar cells, along with TF-rich macrophages will activate the extrinsic clotting pathway on contact. However, instead of a massive infusion of dead and dying cells as occurs in liver transplantation, this thrombin generation and/or infusion of necrotic cells into the pulmonary circulation will occur, one dying alveolus at a time.
The Covid-19 virus, having coopted an anomalous variant of clotting to produce SF, can now extend its damaging effects to the entire body. Wherever the SF molecules encounter other fibrin oligomers—due to conditions such as cooling in the extremities, vascular narrowing, roughened endothelium or other factors that result in turbulence or non-laminar flow, or simply extremely high levels of SF—the short chains of oligomeric fibrin can encounter one another, polymerize, lengthen and form two-stranded protofibrils 0.5–0.6 µm in length. These protofibrils correspond to ~ 20–25 monomers11 and are no longer soluble. Prior to this development the monomers and shorter polymers have been held in soluble form while complexed to carrier fibrinogen molecules. Now, however, as the protofibrils lengthen they become long enough to self-interact and aggregate laterally. A sol to gel transition occurs and microclots form.12 Even after gelation, new fibers and branching points continue to develop.13 As this process continues, the previously soluble SF now forms microclots, capable of occluding small blood vessels throughout the body. As already noted, most such occlusions are rapidly cleared by clot lysis or by remodeling of the fibrin microclot prior to stabilization by Factor XIII.14 But if those microclots persist for more than several minutes, the tissue supplied by the occluded blood vessel, may die.
A clotting process, initiated in the lungs of Covid-19 patients, that distributes SF throughout the body is not merely a hypothetical construct. High levels of SF exist in the blood of severely ill Covid-pneumonia patients and can be readily demonstrated with an appropriate assay. Even though it is likely formed only in the lungs, it is present in, and can be precipitated from, blood samples withdrawn from the arm veins or central catheters of such patients. It therefore exists throughout the entire bloodstream.
We report measurements of these oligomeric fibrin molecules in patients with severe Covid-19 pneumonia and document that the process of SF formation can result in whole body defibrination even though macroclots are not clinically or radiologically identified.