Thromboembolic Disease and Hypercoagulable States

Thromboembolic Disease

The formation of blood clots to seal injured vessels is an essential protective mechanism in animals and is referred to as coagulation. Once these clots have served their purpose, they are degraded by various enzymatic processes (fibrinolysis). The body also has in place several mechanisms to prevent the aberrant formation of blood clots (anticoagulation). Together, these processes form the complex but elegant system of hemostasis. Abnormalities in any of the components of hemostasis can result in thromboembolic disease with serious consequences. In fact, thromboembolic disease is one of the leading causes of death among critically ill people [1,2].

As a point of clarification, the term "thrombus" refers to an aggregation of platelets and other blood components that causes partial or complete vascular obstruction. An "embolus" is a clot or other plug (e.g., fat or fibrocartilaginous material) that breaks off from an originating site and lodges in a distant vessel. Thromboembolic disease encompasses all types of thrombi and emboli.

Hypercoagulable or prothrombotic states are clinical conditions that predispose affected patients to blood clot emboli. The pathophysiology of thrombosis depends on three major factors: alterations in the vessel wall (endothelial injury), impairment of blood flow (blood stasis), and in abnormalities within the coagulation system (hypercoagulability). The interactions of these three processes culminating in thromboembolic disease are known as Virchow's triad.

Vascular injury leads to the exposure of subendothelial vessel wall components, such as collagen, resulting in platelet adhesion and activation of the contact phase of the coagulation system. Vascular integrity may be damaged with venipuncture, intravenous catheter placement, infectious and inflammatory vasculitis, and neoplastic invasion of vessels. Blood stasis favors thrombosis by retarding the removal of activated coagulation factors and by causing local hypoxia and vascular injury. Stasis can result from hypovolemia, shock, cardiac insufficiency, blood vessel compression, or immobility. Vascular obstruction by implants, foreign bodies, or neoplasms may induce local thrombosis as well. Cardiac chamber enlargement or vascular aneurysm leads to focal sludging of flow and thrombus formation. Hyperviscosity, as occurs with dehydration, polycythemia, leukemia, hyperglobulinemia, and hyperfibrinogenemia, also results in stasis.

Whereas vascular stasis and injury are prothrombotic, true hypercoagulability refers to a quantitative or qualitative change in the coagulation system. In the simplest sense, the coagulation system is composed of procoagulants (platelets and coagulation factors), anticoagulants (protein C, protein S, antithrombin, heparin cofactor II), and the fibrinolytic system. Any imbalance in these processes may result in excessive bleeding or hypercoagulation, depending on the nature of the imbalance. Hypercoagulation can result from platelet hyperaggregability, excessive activation or decreased removal of coagulation factors, deficiencies of natural anticoagulants, or defective fibrinolysis.

In regard to platelets, there appears to be no correlation between thrombocytosis and hypercoagulation. Hyperaggregability of platelets, however, can increase thromboembolic risk. Platelet aggregation is finely controlled by interactions between the platelets themselves and the vascular endothelium. Additionally, platelets produce and release various proaggregating substances, including thromboxane A2, adenosine diphosphate (ADP), and prostaglandins G2 and H2. The endothelium in turn, releases several inhibitors of platelet aggregation, namely prostacyclin (PGI2), ADPase, and nitric oxide. Disturbances in the balance between platelets and the endothelium can lead to thrombosis.

Increased activation of coagulation factors (by vascular injury or inflammatory mediators) or decreased removal of factors from areas of injury (owing to stasis or decreased activity of the reticuloendothelial system) may contribute to thrombosis. Whether increased levels of individual clotting factors alone heighten the trend toward a hypercoagulable state remains controversial, but elevated fibrinogen levels, and elevations of factors VIII and XII as seen with hyperadrenocorticism, have been incriminated in people [3-5].

The mechanisms for triggering coagulation are offset by equally powerful mechanisms for controlling it. Three principal mechanisms are involved in this regulation: antithrombin (AT), protein C, and the fibrinolytic system. Protein C is a small (56,000 daltons) vitamin K-dependent serine protease that is synthesized in the liver. It circulates in plasma as an inactive protein, but is activated by thrombin complexed with thrombomodulin. Activated protein C (APC) then exerts its anticoagulant effect by inactivating membrane-bound factors Va and VIIIa (the amplifiers of coagulation). APC also encourages fibrinolysis by neutralizing plasminogen activator activity. Both actions of protein C are greatly enhanced by protein S.

Antithrombin is a serine protease inhibitor of low molecular weight (58,000 daltons) that is synthesized in the liver and endothelial cells. It is one of the most important inhibitors and effective modulators of hemostasis. AT functions as an inactivator of coagulation factors IIa, IXa, Xa, XIa, and XIIa. This effect occurs too slowly to be physiologically relevant unless heparin binds to AT, causing AT to undergo a conformational change that increases its anticoagulant activity by 1000-fold. Natural heparin-like substances (e.g., sulfated aminoglycans and heparin sulphates) are present on endothelial surfaces and are the major contributors to the anticoagulant properties of the endothelium. Heparin cofactor II (HC II) is a glycoprotein with a molecular mass of 64,000 daltons. It acts as an inhibitor of thrombin, with its action also accelerated by heparin.

AT deficiency is well described in dogs [6,7]. Acquired deficiencies occur because of decreased production, excessive losses, or consumption. Decreased hepatic production of AT occurs with hepatopathies, but thrombosis generally does not result because concomitant clotting factor deficiencies favor hemorrhage. A similar situation occurs with generalized protein loss, such as protein-losing enteropathies. However, there have been cases of thrombosis with protein-losing enteropathies [8]. In contrast, the increased glomerular permeability that accompanies protein-losing nephropathies (e.g., glomerulonephritis and amyloidosis) permits selective loss of lower molecular weight proteins, and consistently results in hypercoagulability [6,7]. AT is smaller than albumin (69,000 daltons) and procoagulant proteins, creating sufficient imbalance to favor a hypercoagulable state. Disseminated intravascular coagulation (DIC) is a situation of increased AT consumption. Measurement of AT activity has been helpful in detecting patients with DIC, and some authors have used it to guide therapy [9].

The fibrinolytic system is extremely complex, and remains the least completely understood. Evidence is growing that fibrinolysis is exceedingly important in hypercoagulation [10]. Persistence of thrombi is abnormal and implies defective fibrinolysis. More recently, various degrees of hypofibrinolysis have been confirmed in many types of hypercoagulable states [10,11].

In its simplest form, the fibrinolytic system consists of plasminogen and its activators, which convert plasminogen to plasmin. Plasmin is responsible for dissolution of the fibrin clot. The two physiologic plasminogen activators are tissue-type plasminogen activator (t-PA) and urokinase (UK). T-PA is released from intact endothelium in its active form, whereas UK is released as an inactive zymogen (pro-UK, which is converted to UK by kallikrein. Both t-PA and UK are inhibited by plasminogen activator inhibitor (PAI-1), the major inhibitor of the system. Other inhibitors include: histidine-rich glycoprotein, which interferes with the binding of plasminogen to fibrin; α2-antiplasmin, which inhibits plasmin; and C1-inhibitor, which inactivates kallikrein. Alpha2-macroglobulins are scavenging protease inhibitors that intervene in situations in which antiplasmin is markedly decreased (i.e., thrombolytic therapy). Hypofibrinolysis and resultant thrombosis can occur because of decreases in plasminogen, t-PA, or UK, or because of increases in circulating inhibitors. Increased levels of PAI-1 are by far the most frequent cause of ineffective fibrinolysis in people [10].

Thrombosis associated with the development of antiphospholipid antibodies ("lupus-type anticoagulants") is common in people but has only been described in one dog with hemolytic anemia [12]. The exact mechanism by which lupus-type anticoagulants cause thrombosis is unclear, but these antibodies may develop secondarily to autoimmune disease, neoplasia, infectious or inflammatory diseases, or drug reactions. Defects of platelet function, dysfibrinolysis, plasminogen deficiency, dysfibrinogenemia, and t-PA deficiency may also lead to thrombosis in people, but have not yet been described in veterinary patients.

Specific Prothrombotic Disorders

Nephrotic Syndrome

Glomerular damage leads to leakage of small molecular weight proteins from blood. Because all of the major inhibitors of coagulation are proteins smaller than albumin, significant protein loss will result in acquired coagulation inhibitor deficiency. Antithrombin deficiency is well documented in protein-losing glomerulonephropathies [6,7]. The hypercoagulable state results from retention of high molecular weight hemostatic factors and selective loss of hemostatic regulators. Underlying inflammation also contributes to the hypercoagulable state by increasing production of acute-phase proteins that play a role in hemostatic reactions, e.g., fibrinogen, factor VIII, and α2-macroglobulin [6,7]. Hypoalbuminemia contributes to hypercoagulability owing to the development of platelet hyperreactivity [7]. Pulmonary thromboembolism (PTE) is the most common manifestation of hypercoagulability in this disorder.6

Immune-Mediated Hemolytic Anemia (IMHA)

PTE is also a common complication of IMHA, occurring in up to 35% of IMHA patients [13-19]. The mechanism is incompletely understood but it has been suggested that endothelial exposure to anti-erythrocyte antibodies and the subsequent complement-mediated damage triggers thrombosis [13-16]. Other hypotheses include increased activity of clotting factors, decreased activity of anticoagulants, defective thrombolysis, increased platelet reactivity, and the presence of antiphospholipid antibodies [16]. In complement-mediated erythrolysis, a significant amount of thromboplastin may be released into the circulation. Other factors implicated in the development of PTE include the release of inflammatory cytokines, the use of corticosteroids, frequent venipuncture and catheterization, and patient inactivity, resulting in sluggish blood flow [19].

Cardiac Disease

Arterial thromboembolism (ATE) occurs commonly in cats with cardiac disease, especially cats with hypertrophic cardiomyopathy [20-24]. ATE has also been reported in dogs, but the incidence is much lower [25,26]. Unlike other hypercoagulable states where the lesions center within veins, the thrombosis associated with cardiac disease affects arteries. Moderate to severe left atrial enlargement and swirling echocardiographic densities in the left atrium ("smoke") are considered risk factors for ATE [22]. The pathophysiology of ATE is thought to involve severe dilatation of the left atrium (resulting in stasis of blood flow), endothelial damage leading to platelet activation, increased platelet reactivity, and a number of as-of-yet unidentified factors [21,22,24].

Hyperadrenocorticism

Patients with endogenous or exogenous glucocorticoid excess may develop venous, pulmonary, or arterial thrombosis. The mechanism of hypercoagulability associated with hyperadrenocorticism is incompletely understood. Contributing factors may include increased activity of coagulation factors (factors VIII, V, and prothrombin), which have been documented in dogs with hyperadrenocorticism, or the occurrence of secondary hypertension [3]. The presence of hyperadrenocorticism alone probably does not confer a high risk for thromboembolic complications necessitating prophylactic therapy. However, the presence of other concomitant risk factors may require anticoagulation.

Pancreatitis

The relationship between inflammatory conditions and the predisposition for developing thromboembolic events is well known. Proinflammatory cytokines such as TNF-α and IL-1 play a role in activating the coagulation system but are not the only contributors. In the case of severe pancreatitis, proteolytic enzymes that leak into the circulation are removed by the reticuloendothelial system once bound to α-macroglobulins [27]. Depletion of plasma α-macroglobulins or the inability of the reticuloendothelial system to remove bound proteases leads to activation of many plasma proteins, including those of the coagulation cascade and fibrinolytic system [27]. Although the depletion of α-macroglobulins in patients with pancreatitis has been used to justify the use of fresh frozen plasma (FFP) transfusions, no veterinary trial has demonstrated an improvement in AT activity following transfusions [15,28]. A benefit to the course or outcome in human patients with severe pancreatitis by the administration of FFP has similarly not been demonstrated [29,30].

Sepsis

In septic patients, several mechanisms exist by which thromboembolic events can occur. Generalized inflammation and circulation of cytokines are thought to activate the coagulation system. Furthermore, endotoxins and exotoxins released by microbes can contribute to thrombosis by causing direct endothelial damage, stimulating tissue factor (TF) expression, and activating clotting factors and platelets [31]. The progression of a hypercoagulable state to fulminant DIC occurs with concomitant dysfunction of the natural anticoagulant system. Depletion of AT, protein C, and protein S, and reduced plasmin activity, are key players in this progression. A more detailed discussion of DIC is found elsewhere. (See Chapter on Hemostasis and Disseminated Intravascular Coagulation). Necrotizing fasciitis is another important condition that exemplifies the relationship between infections and coagulation. Microorganisms can possess unique cell membrane components that elicit specific prothrombotic tendencies that result in the thrombosis of regional vessels and contribute to the extensive tissue necrosis and sloughing of muscle and skin seen with this condition [32]. In fact, the rapidity and extent of tissue destruction is more consistent with microvascular thrombosis and resultant ischemia than with direct microbial destruction of tissue [31].

Orthopedic Surgery

The occurrence of PTE stemming from deep venous thrombosis in patients requiring orthopedic surgery is well documented in people, but this has not been described in the veterinary literature. However, the occurrence of thromboembolic events during or after total hip replacement in dogs has been described [33-36]. In the unique context of cemented total hip replacement, a marked increase occurs in proximal intramedullary pressure within the femur during femoral stem implantation. As cement intrudes into the cancellous bone, it dislodges particles of fat and bone, which enter the circulation and migrate to the right heart and pulmonary circulation [35,36]. This has been documented in studies using transesophageal echocardiography, capnography, pulmonary scintigraphy, and histologic evaluation [33-36]. Despite the fact that some form of thromboembolism occurs frequently in dogs undergoing total hip replacement, clinically significant abnormalities or fatalities remain rare. The pathophysiology of thromboembolism associated with total hip replacement differs from other diseases in that there is no direct perturbation of the coagulation system; therefore, preventative and therapeutic interventions should not involve the use of anticoagulants or antithrombotics. Nevertheless, a discussion of this phenomenon is important because the clinical manifestations (e.g., respiratory distress and sudden death) are identical to PTE caused by hypercoagulability. Techniques to reduce the risk of thromboembolic complications include controlling the rise in intramedullary pressure during implant insertion, use of vacuum drainage in the proximal femur, and meticulous, copious, pulsatile lavage of the bone-cement interface to remove embolic substrate [35,36].

Laboratory Diagnosis of Hypercoagulation

The detection of hypercoagulability is extremely difficult in clinical practice. Little correlation exists between thrombocytosis, shortened coagulation times (decreased prothrombin time (PT), and activated partial thromboplastin time (aPTT)) and a prothrombotic tendency. Fibrin degradation products (FDP) are generated by the dissolution of fibrin by plasmin, and therefore, increased concentrations only indirectly imply thrombus formation.

Assessing risk for thromboembolic complications involves the recognition of predisposing conditions and populations rather than any single clinicopathologic abnormality. A list of predisposing conditions can be found in Table 21.1. Some authors have attempted to devise a predictive scheme based on the patient's plasma AT concentrations, but this algorithm has neither been validated nor is the AT assay widely available [9]. Such a scheme recommends AT supplementation and heparin therapy when AT activity levels fall below 60%. Other tests that have been developed include screening for elevations in the by-product of prothrombin activation (fragments F1 and F2) or fibrinogen cleavage (fibropeptides A and B), and increased concentrations of thrombin: AT complex [36-39]. Again, these markers indicate that thrombin has been generated rather than the likelihood of thromboembolism. Newer assays for D-dimers (a more specific indicator of cross-linked fibrin generation) are currently considered the best clinical test for diagnosing the presence of active coagulation [40-42]. Because of its high specificity, this semi-quantitative test may be best used to rule out cases of suspected PTE [40]. D-dimer concentrations > 2000 ng/dL had a 98.5% specificity in a heterogeneous population of dogs, and no dog with a confirmed embolus had a negative d-dimer test [41]. Thromboelastography, a global in vitro test of coagulation, is perhaps the best tool for assessing hypercoagulability, but remains a research tool optimally suited to characterize the degree of hypercoagulability seen with specific diseases [43,44].

Anticoagulant and Antithrombotic Therapy

Unfractionated heparin is probably the most commonly employed form of anticoagulant therapy used in animals. The mechanism of action centers around the potentiation of AT, leading to the inactivation of factors IIa, Xa, IXa, XIa, and XIIa. Of these, the inactivation of IIa (thrombin) and Xa are considered the most important. Whereas heparin appears to increase tissue-type plasminogen activator (t-PA) activity, it does not appear to enhance fibrinolysis; however this subject remains highly controversial [45]. Various dosing schemes have been proposed and range from intermittent subcutaneous injections of 50 to 300 units/kg every 8 hours to continuous intravenous infusions of 5 to 20 units/kg/hour. It is important to realize that these dosing schemes are largely anecdotally derived, and studies to determine optimal dosing in patients with hypercoagulable tendencies have not been performed. Regardless of the dosing scheme, a common recommendation is to prolong the APTT by 1.5 to 2.5 times the baseline. Some limitations of monitoring heparin therapy via clotting times do exist, however, in that manipulation of the APTT does not necessarily decrease the risk of a thromboembolic event. Furthermore, considerable variation occurs in APTT sensitivity to heparin using different coagulometers.

One of the major concerns with heparin therapy is that it can lead to overt coagulopathy and major hemorrhage. Individual variation in the ability to metabolize heparin and changes in coagulation status in the course of many diseases explain the development of bleeding tendencies in previously hypercoagulable individuals. Although protamine sulfate is considered a reversal agent for heparin, few indications exist for its use given its side effects and the short half-life of heparin, particularly in hypercoagulable individuals. Another concern with the use of unfractionated heparin is that it can eliminate the important anti-inflammatory effects of AT, potentially limiting its use in critically ill patients [46-49].

Given concerns over the possible complications of using unfractionated heparin, interest has been increased in the use of low molecular weight heparin (LMWH). Major advantages of LMWH include improved bioavailability, predictable renal clearance, predictable antithrombotic responses, more tempered effects on coagulation factors (mainly Xa), and supposedly prolonged half-life. However, several studies have shown that maintenance of a desirable anti-Xa activity level in both dogs and cats requires more frequent administration than the once daily dosing recommended for human patients [50,51]. A dose of 150 to 200 units/kg subcutaneously every 8 to 12 hours, in cats and dogs, respectively has been anecdotally reported with dalteparin [20,50,51]. Monitoring of APTT during LMWH therapy is not effective and may require assaying anti-Xa activity, which is not readily available. Because of the reduced risk of hemorrhage with LMWH, some authors have used it empirically without specific monitoring. However, further studies are needed to determine the optimal use of LMWH in dogs and cats, especially given the high cost of LMWH.

Coumarin-based anticoagulants (e.g., warfarin) block the g-carboxylation of several glutamate residues in factors II, VII, IX, and X, as well as the endogenous anticoagulant proteins C and S. The blockade results in incomplete molecules that are biologically inactive (unable to undergo calcium-mediated binding of prothrombin to platelet phospholipids), and are collectively known as PIVKA, or proteins induced by vitamin K antagonism or absence. The anticoagulant prevents the conversion of inactive vitamin K epoxide back to its active hydroquinone form that is required for proper clotting-factor synthesis. Therapy using coumarin should be carefully titrated against the PT, with the optimal goal of prolonging it to 1.5 to 2 times the patient's baseline PT. Recommended dose ranges are from 0.05 to 0.1 mg/kg PO SID for 3 days, then every other day to every third day, depending on individual response. Therapy should be recommended only in patients in which monitoring will be consistently and reliably performed.

Aspirin antagonizes thromboxane A2, which is responsible for allowing platelets to alter their shape, release their granules, and aggregate. Aspirin is a mild platelet inhibitor that slightly attenuates aggregation responses to ADP and collagen but does not inhibit thrombin- or PAF-induced aggregation responses. Aspirin inhibits synthesis of thromboxane A2 by irreversible acetylation of cyclooxygenase. Because the anuclear platelet cannot synthesize new proteins, it cannot manufacture any new enzyme during its 10-day lifetime. The typical dose of aspirin used as an anti-inflammatory agent may actually be procoagulant and, therefore, the use of an ultra-low dose has been recommended, which has been reported as 0.5 mg/kg/day PO [52,53].

Fibrinolytic Agents

Although the practice is controversial, in the situation of a life-threatening thromboembolism, fibrinolytic drugs may be used to rapidly lyse thrombi and restore perfusion. This is accomplished by the ability of these agents to catalyze the formation of plasmin from its precursor plasminogen. These drugs create a generalized thrombolytic state when administered intravenously. Thus, both protective homeostatic thrombi and pathologic thromboemboli are broken down. In people, to reduce the nonselective systemic effects of fibrinolytic agents, intra-arterial administration has been used [54]. Given the high rate of complications, considerable cost, difficulty in obtaining these products, and high mortality of affected patients, experience with these treatment modalities is limited in veterinary medicine.

The use of streptokinase has been reported sporadically in the veterinary literature [55-58]. Streptokinase combines with a proactivator of inactive plasminogen, and this enzymatic complex catalyzes the conversion of inactive plasminogen to active plasmin, which degrades fibrin into fibrin degradation products. Since streptokinase does not have an affinity for fibrin-bound plasminogen, it can induce a state of systemic fibrinolysis and lead to significant hemorrhage. The dose in cats has been reported as a loading dose of 90,000 units per cat in 20 to 30 minutes, followed by a maintenance dose of 45,000 units per hour for 3 hours [23]. In dogs, dosage protocols include loading doses of 90,000 units intravenously administered slowly, followed by maintenance doses of 45,000 units per hour intravenous infusion over various periods of time, ranging from 30 minutes to 12 hours [56,57]. In one successfully treated case of arterial thromboembolism, a total of 8 streptokinase infusions were administered over a 50-hour period [57].

Another potent thrombolytic agent is tissue-type plasminogen activator (t-PA), which is relatively clot-specific owing to its relative low affinity for circulating plasminogen and high affinity for cross-linked fibrin. In people, t-PA has been used for lysis of thrombi involved with acute myocardial infarction, PTE, and peripheral vascular thrombosis [54]. The clinical use of t-PA for the treatment of PTE in dogs and cats has not been reported. Only a few reports exist of successful thrombolysis in dogs using t-PA [8,59]. The cost and difficulty in obtaining this drug has precluded further evaluation of this treatment modality as a practical option. In cats, a total dose of 1 to 10 mg/kg administered as a 0.25 to 1 mg/kg/hour infusion has been reported, although the mortality rate associated with this therapy was 50% [24].

The clinical use of urokinase in veterinary medicine is limited but has been reported in both cats and dogs with various types of thromboembolism [26,60]. The greatest experience is in cats with ATE, where the success rate was similar to that of other thrombolytic agents, but the complications were less severe and less common [60]. The one possible advantage suggested by the limited data available is that complications associated with UK therapy, as compared with other thrombolytics, may be less frequent and less severe; however, more studies are warranted to address this issue.

Conclusion

The occurrence of thromboembolic disease can have devastating consequences. Recognition of high-risk populations is important because timely intervention can be essential in the successful management of these cases, and preventative measures may be employed to decrease the likelihood of thromboembolic events. Knowledge of the pathophysiology involved in the development of thrombi in the context of different diseases can help in raising the index of suspicion in particular cases and in devising effective therapeutic strategies. Further research in the area of antithrombotics and anticoagulants is needed to determine optimal use of these agents in high-risk populations.