CHAPTER 17: Thrombotic Disorders

As mentioned in the introduction to this section of the book, thrombosis plays a commanding role as a complication of highly prevalent disorders such as obesity, diabetes, and cancer. Myocardial infarction is the leading cause of mortality from arterial thrombosis, whereas pulmonary embolus accounts for most deaths due to venous thrombosis. In this chapter, we will stress the pathophysiologic principles underlying in vivo thrombus formation and the acquired and inherited factors that put patients at increased risk.

Insight into the pathogenesis of thrombotic disorders began with the 19th century pathologist Rudolf Virchow, whose clinical and postmortem observations led him to postulate that the formation of blot clots sufficiently large to occlude major blood vessels was dependent upon three independent processes:

• Abnormal blood flow—principally stasis.

• Irritation of the vessel wall due to trauma, or a degenerative/inflammatory process such as atherosclerosis.

• Enhanced blood coagulation, or what we now call hypercoagulability.

Virchow's triad provides a useful framework for evaluating any patient who has developed clinically significant thrombotic disease. The history will often reveal factors that predispose to stasis, such as immobilization of a limb, pregnancy, or prolonged inactivity. Further studies may identify a specific type of vascular pathology as well as abnormalities in blood coagulation. In this chapter, venous thrombosis will be covered in more depth than arterial thrombosis because the former is closely linked to well-defined abnormalities in coagulation, whereas the latter is much more dependent on vascular abnormalities, particularly atherosclerosis, which lie outside the scope of this book.

As outlined in Table 17-1, both nature and nurture contribute to the development of venous thrombosis. As listed in the left hand column, a wide spectrum of acquired or secondary factors puts individuals at increased risk. Several, such as increasing age, male sex, and obesity, are weak but statistically significant risk factors. In contrast, a history of venous thrombosis is a strong predictor of repeated episodes. Immobilization due to extended bed rest or prolonged air travel leads to blood stasis, particularly in the lower extremities. Immobilization is also an important contributor to the increased risk of venous thrombosis in patients recovering from major surgery. Pregnancy causes stasis in the legs, owing to impairment of venous return by the enlarging uterus. Malignancy and estrogen use will be discussed later in the chapter along with other less common acquired risk factors.

During the last 15 years, it has become apparent that inherited abnormalities of blood coagulation play an important role in determining which individuals are at a higher risk for development of venous thrombosis. The right-hand column of Table 17-1 lists monogenic defects linked to the development of venous thrombosis. All consist of mutations in clotting factors or in inhibitors of clotting. Figure 17-1 depicts where these proteins are situated in the coagulation cascade.

Two of these genetic variants are mutations in genes encoding clotting proteins—factor V and prothrombin:

• Factor V Leiden is a genetic variant created by a single base change that results in the replacement of arginine by glutamine at position 506. In normal factor Va, Arg506 is the first site cleaved by activated protein C (Fig. 17-2). Subsequent cleavage by this serine protease at Arg306 and Arg679 inactivates factor Va. The absence of the initial cleavage at position 506 in factor V Leiden markedly delays cleavage at the two other sites and thereby results in higher levels of factor Va, which acts as a cofactor for factor Xa.

• Prothrombin G20210A is a single base substitution in the 3' untranslated region of the prothrombin gene that increases the biosynthesis of prothrombin protein. As a result, individuals heterozygous for this mutation have approximately 30% higher plasma levels of structurally normal prothrombin compared with normal individuals. This seemingly modest increase in concentration significantly increases the risk of venous thrombosis.

Both factor V Leiden and prothrombin G20210A mutations are surprisingly common in White populations, with prevalences of about 5% and 2%, respectively. In fact, the high frequency of these two genetic variants suggests that they confer some type of selective advantage. However, their clinical importance lies in the striking finding that one, or sometimes both, of these mutations is encountered in a substantial percentage of individuals who develop deep venous thrombosis either spontaneously or in the presence of acquired risk factors.

Inherited defects that increase risk of deep venous thrombosis also include deficiencies of protein C, protein S, and antithrombin. The mutations in these proteins differ from factor V Leiden and prothrombin G20210A mutations in three major respects:

• They involve inhibitors of coagulation rather than procoagulant factors.

• They are much less common.

• They are located at multiple sites in the respective genes.

The activities of protein C, protein S, and antithrombin also can be decreased in genetically normal individuals by certain conditions or medications.

Because of the low frequency of these mutations, homozygous individuals are very rarely encountered. A functional antithrombin level of less than 10% of normal is incompatible with life. Homozygous individuals with severe protein C or protein S deficiency have a very severe phenotype marked by thrombosis in childhood, adolescence, or even infancy (neonatal purpura fulminans). In contrast, the more frequently encountered heterozygotes are at increased risk of developing thrombosis in adulthood, albeit at low penetrance. The responsible mutations may either suppress expression of the protein (type I) or impair the functional activity of the protein (type II). In the former, both the antigen and activity levels are low, whereas in the latter, the antigen level is normal but the activity level is low.

In contrast to factor V Leiden and prothrombin G20210A mutations, which can be identified unequivocally by DNA analysis, inherited deficiencies of antithrombin, protein S, and protein C are diagnosed by measurements of plasma levels. Interpretation of these levels is often confounded by the fact that they can be lowered by a number of acquired factors (Table 17-2). Of most concern is the impact of acute thrombosis and anticoagulant therapy. Thus, in order to make a convincing diagnosis of an inherited deficiency of antithrombin, protein S, or protein C, measurements should be made after the patient has recovered from the thrombotic episode and is no longer receiving anticoagulants.

An appreciation of the clinical significance of deficiencies in antithrombin, protein C, and protein S requires an understanding of the role of these proteins in pathways that inhibit clot formation. Figure 17-3 depicts how protein C and protein S limit the extent of clot formation. Thrombomodulin is a transmembrane protein expressed on endothelial cells that has on its extracellular domain a binding site for thrombin. Upon docking onto thrombomodulin, thrombin's substrate specificity becomes restricted to protein C. Protein C, like factors VII, IX, X, and prothrombin, is a vitamin K–dependent zymogen. Protein C binds to a receptor termed the endothelial protein C receptor on the endothelial cell surface, where protein C is activated by the thrombin-thrombomodulin complex. The rate of protein C activation is accelerated about 20,000-fold by binding to this complex. Activated protein C cleaves and degrades factors Va and VIIIa, as depicted in Figure 17-3. Importantly, the inactivation rate is enhanced by protein S, which is also vitamin K–dependent but not a zymogen.

Individuals with protein C deficiency are at risk for a rare but serious complication termed warfarin-induced skin necrosis. When treatment with the anticoagulant warfarin (Chapter 13) is initiated, these patients occasionally develop painful ecchymoses in fat-containing areas of the body such as the breasts or buttocks, which can be quite extensive and debilitating (Fig. 17-4). Microscopically, these lesions contain fibrin thrombi within venules, accompanied by hemorrhagic necrosis. The explanation for warfarin-induced skin necrosis lies in wide differences in the stability of vitamin K–dependent coagulation proteins in the plasma. Factor VII and protein C have the shortest half-lives (6 and 14 hours, respectively), whereas the other vitamin K–dependent proteins have half-lives of 24 to 60 hours. In normal individuals, the fall in protein C activity following the start of warfarin is generally not of clinical significance, but in heterozygotes with levels of protein C that are about half of normal, a further drop in protein C activity during the first day of therapy can trigger thrombosis in the skin.

Antithrombin is a member of a large family of serine protease inhibitors. Although initially described as the primary physiologic inhibitor of thrombin, antithrombin was subsequently shown to also neutralize factors IXa and Xa and, to a lesser extent, factors XIa and XIIa. Antithrombin is also known as heparin cofactor I. In the absence of heparin, antithrombin is a relatively slow inactivator of these activated clotting factors. In contrast, as shown in Figure 17-5, the binding of the heparin polysaccharide to antithrombin changes its conformation in a way that markedly enhances its affinity for thrombin as well as for factors IXa, Xa, XIa, and XIIa. Heparin's binding site to antithrombin has been localized to a sulfated pentasaccharide (Fig. 17-5). This small domain is sufficient to inactivate factor Xa but not thrombin, because thrombin inhibition by heparin requires longer saccharides. This insight has permitted the development of heparin-like drugs that target factor Xa only (such as the synthetic pentasaccharide fondaparinux). Endogenous heparan sulfate polysaccharides on the surface of vascular endothelium also have domains with structures very similar to this pentasaccharide and appear to activate antithrombin, thereby serving as natural anticoagulants.

Individuals with inherited antithrombin deficiency are at an increased risk of developing venous thrombosis, similar to those with protein C and protein S deficiency. Among the inherited thrombotic disorders, the clinical penetrance of venous thromboembolism appears to be highest in families with type I antithrombin deficiency.

If an individual, particularly a child or young adult, develops venous thrombosis in the absence of any of the acquired factors listed in Table 17-1, he or she may have an inherited defect in one of the proteins discussed earlier. As shown in Figure 17-6, in the presence of one or more of these acquired risk factors, the chance of thrombosis is markedly enhanced by the presence of an inherited defect. The secondary risk factors can be divided into two groups: conditions such as older age, obesity, and cancer and transient precipitating factors such as estrogens, pregnancy, surgery, and immobilization.

The combinatorial impact of inherited and acquired risk factors has been demonstrated by a large, case-control study from Leiden, Holland on the risk and incidence of a first episode of deep venous thrombosis in a previously healthy population (summarized in Table 17-3). It is clear from this table that the risk of venous thrombosis associated with the factor V Leiden mutation and the secondary risk factor, estrogens in the form of oral contraceptives, are synergistic. The combination imposes significantly more risk and a higher incidence than the sum of the two factors alone.

Antiphospholipid antibody syndrome is characterized by thrombosis in association with an acquired inhibitor directed against proteins bound to membrane phospholipids and/or persistently elevated titers of antibodies against cardiolipin or β₂-glycoprotein I. Patients will often have a prolonged partial thromboplastin time and, less commonly, a prolonged prothrombin time. Mixing studies described at the end of Chapter 13 reveal that the long partial thromboplastin time is not corrected by the addition of normal plasma. Despite this laboratory abnormality, affected individuals almost never have increased bleeding but are at an increased risk for venous as well as arterial thrombosis. In this important respect, antiphospholipid syndrome along with heparin-induced thrombocytopenia and the hypercoagulable state associated with malignancy, described in the following section, differ from the inherited disorders described earlier in this chapter. Other clinical manifestations of antiphospholipid antibody syndrome include immune thrombocytopenia and recurrent fetal loss. Some patients with this syndrome have systemic lupus erythematosus. Indeed, the coagulation abnormality is often called lupus anticoagulant. These patients are at a high risk for recurrent thrombosis and should be treated with long-term anticoagulation after an initial, unprovoked thrombotic event.

As mentioned in Chapter 13, a small percentage of patients (1%-5%) treated with unfractionated heparin develop a potentially serious complication, heparin-induced thrombocytopenia, owing to development of an autoantibody against a complex of the drug and platelet factor 4 released from platelet alpha granules. As shown in Figure 17-7, the antigen-antibody complex binds to Fc receptors on platelets and triggers their activation. This commonly results in mild or moderate thrombocytopenia, but more importantly, in about 20% to 50% of patients causes both venous and arterial thrombosis. This serious adverse effect is encountered much less often in patients treated with fractionated low-molecular-weight heparin and rarely with the synthetic pentasaccharide fondaparinux.

There is a high incidence of venous thrombosis in cancer patients. Some evince migratory superficial thrombophlebitis (Trousseau syndrome), often in unusual sites such as the arms. In addition, thrombosis associated with chronic disseminated intravascular coagulation is sometimes encountered in cancer patients (Chapter 15).

Those who are terminally ill may develop nonbacterial thrombotic (marantic) endocarditis. The pathogenesis of these thrombotic complications is unclear but probably involves release of thrombogenic agonists from malignant cells, such as microparticles containing tissue factor. Thrombosis also may occur during treatment of cancer with various chemotherapeutic agents, such as L-asparaginase, or estrogen-like drugs such as tamoxifen.

Patients with two types of clonal hematologic disorders, the myeloproliferative disorders and paroxysmal nocturnal hemoglobinuria, are also at a high risk for developing venous and arterial thrombosis. These conditions are discussed in detail in Chapters 20 and 11, respectively.