Chapter 6Arterial, Venous, and Microvascular Hemostasis/Thrombosis

Publication Details

While hemostasis represents a physiological response to prevent bleeding, the term thrombosis typically refers to the pathologic formation of a thrombus (clot). This may result in severe consequences, due to reduction or cessation of blood flow to a tissue by the thrombus itself, or by rupture and release of thrombi (known as embolization). Thrombosis contributes to morbidity and mortality in common clinical diseases like myocardial infarction, stroke, deep venous thrombosis, and cancer, among many others. Since these disorders represent the most common causes of death [205], thrombosis is a major health and economic burden. Clinical and experimental thrombosis may occur throughout the vascular tree; while they share some fundamental features among the various vessel types, certain differences in pathogenesis and molecular mechanisms are evident [206209]. A number of clinical risk factors for development of thrombosis have been identified [210,211]; these are outlined in Table 6.1.

Table 6.1. Selected Risk Factors for Thrombosis.

Table 6.1

Selected Risk Factors for Thrombosis.

6.1. Venous Thrombosis

The pathophysiology of venous thrombosis is traditionally attributed to the experiments performed by Rudolph Virchow in the mid 19th century, in which he described that the consequences of thrombosis in dog pulmonary arteries could be grouped according to irritation of the blood vessel and its surroundings, to blood coagulation, and to interruption of blood flow [212]. Those observations formed the basis of “Virchow’s triad” of endothelial injury, hypercoagulability and stasis, as mechanisms predisposing to formation of venous thrombosis. Venous thrombosis is a common clinical entity, associated with a number of acquired and intrinsic risk factors, including cancer, surgery, indwelling catheter use, stroke, medications, inherited hypercoagulable states, and others [211]. The relative contribution of the factors of Virchow’s triad varies according to the type of venous thrombosis, and between venous and arterial thrombi. These will be discussed individually.

6.1.1. Endothelial Injury

The role of endothelial injury in most clinical cases of venous thrombosis is relatively unclear. Venous thrombosis occurs frequently in the veins of the lower extremities, particularly in the pockets of venous valves, which serve to direct flow against gravity. In a classic study, Sevitt [213] reported histologic findings of 50 thrombi of femoral vein valves obtained at autopsy. The thrombi consisted mainly of fibrin and varying degrees of platelets, but there was no significant evidence of damage to the vascular intima; the endothelium beneath the thrombi appeared normal. These findings suggest that overt damage to endothelium may not contribute significantly to the common entity of deep venous thrombosis. More subtle forms of endothelial injury, without overt denudation, have been proposed to contribute to thrombosis in this entity [214]. One potential stimulus for endothelial injury is hypoxia, which has been demonstrated to occur in the pockets of venous valves [215], and may result in increased expression of P-selectin on endothelial cells [216]. As outlined above in Chapter 3, P-selectin may contribute to thrombosis by various mechanisms including promoting platelet-endothelial and platelet-leukocyte-endothelial interactions. Endothelial P-selectin surface expression is associated with a variety of inflammatory stimuli, and local inflammation has also been postulated to contribute to the initiation of deep venous thrombosis [217]. The association between inflammation and venous thrombosis is illustrated by the clinical use of the term “thrombophlebitis”, or inflammation of a vein related to a thrombus [218]. Inflammation has also been suggested to mediate the contribution of stasis to venous thrombosis as outlined in the next section.

Although endothelial denudation was not detected in patients with deep vein thrombosis of the lower extremities [213], denudation may contribute to some cases of venous thrombosis, such as those associated with the use of indwelling central venous catheters [219]. Ultrastructural studies on animal models of catheter-associated thrombosis, using catheters analogous to those used clinically, demonstrate significant endothelial denudation [220]. Of interest, an association between infection and thrombosis has been well established in patients with catheter-related thrombosis [221], suggesting a role for inflammation in this entity.

6.1.2. Hypercoagulability

The term “hypercoagulable state” is used to denote conditions that favors a procoagulant state due to imbalances in the hemostatic mechanisms, resulting in an unusual tendency toward thrombosis [210]. The changes in the hemostatic factors can often be used to predict the risk of future vascular events. The activation of coagulation factors plays a major role in venous thrombosis by concentrating these activated factors in regions of reduced flow, such as the valve pockets. Consequently, inherited defects in the anticoagulant mechanisms such as antithrombin deficiency, the protein C pathways and fibrinolytic pathways play a major role in venous thrombosis [210]. Furthermore, activation of the coagulation cascade due to release of tissue factor from damaged tissues, tumors, or cytokine stimulation also increases the risk of venous thrombosis. Stasis accentuates the local concentration of the activated coagulation factors at areas of vascular injury as discussed below.

6.1.3. Stasis

An association between stasis and venous thrombosis has long been recognized based on a number of clinical observations. In a classic study of patients with stroke and paralysis or weakness of one half of the body (hemiplegia or hemiparesis, respectively), 53% developed venous thrombosis in the paralyzed limb, whereas only 7% developed thrombosis in the non-paralyzed limb. A number of other conditions resulting in hospitalization, such as the postoperative state, are well known risk factors of venous thrombosis, particularly with prolonged bed rest [222]. Similarly, patients with factures, immobilization and orthopedic surgery are at especially high risk [223], and prolonged immobility during long-haul air flights is increasingly associated with development of venous thrombosis [224]. Although stasis clearly predisposes to development of venous thrombi, it is unclear whether stasis alone, in the absent of other components of Virchow’s triad, is sufficient to initiate venous thrombosis [209]. Animal models of venous thrombosis associated with stasis typically require concomitant forms of endothelial injury such as vessel ligation or vascular injury [225,226]. Similarly, some clinical data failed to demonstrate independent effects of bed rest per se, on venous thrombosis after controlling for other clinical conditions associated with thrombosis [227]. A potential mechanism by which stasis predisposes to venous thrombosis is by inducing local hypoxia and stimulating endothelial release of P-selectin, as discussed above. In addition to P-selectin, hypoxia-induced exocytosis of Weibel-Palade bodies results in increased release of vWF [228], which may promote platelet recruitment by the mechanisms outlined previously. Hypoxia-induced stimulation of tissue factor synthesis by monocytes has also been postulated as another mechanism linking hypoxia and thrombosis [229].

6.1.4. Coexistence of Virchow’s Triad Components: Cancer and Thrombosis

In a number of clinical entities, more than one component of Virchow’s triad may contribute to the pathogenesis of thrombosis. A prominent example of this is evident in thrombosis associated with cancer [230]. Patients with cancer have a nearly 7-fold increased risk of venous thrombosis as compared to patients without malignancy, and approximately 20% of new cases of venous thrombosis are associated with cancer [231233]. Cancer may contribute to a hypercoagulable state; for example, cancer cells express tissue factor and can release tissue factor-bearing microparticles [234]. Of interest, the degree of tissue factor expression on some tumors has been correlated with greater risk of development of venous thrombosis [234,235]. Other proposed factors derived from cancer leading to a hypercoagulable state include cancer procoagulant, cytokines, and fibrinolytic substances [236]. In addition to hypercoagulability, endothelial injury has been suggested to contribute to venous thrombosis in cancer. In certain tumors, direct extension of cancer cells into blood vessels may result in formation of thrombi [237]. In addition, indirect endothelial activation via cytokines may promote thrombosis [236], by similar mechanisms as those outlined under platelet-endothelial adhesion in Section 3.3. Further, cancer-derived microparticles may mediate endothelial injury; recent data demonstrate incorporation of these microparticles at the sites of endothelial injury in a mouse model of thrombosis [238]. Cancer may also promote thrombosis by contributing to stasis; for example as a result of vascular compression by the tumor or enlarged lymph nodes [239], or by the associated reduced mobility [240]. Finally, catheter-associated thrombosis are common in patients with cancer, due to the frequent use of these devices in this patient population [241].

6.2. Arterial Thrombosis

Consequences of arterial thrombosis such as myocardial infarction and stroke are the most common causes of morbidity and mortality in middle-aged Americans. Though the clinical manifestation of myocardial infarction and strokes are different, they are the result of the same pathogenic process, formation of a thrombus over an underlying atherosclerotic plaque in the setting of high flow and high shear arterial circulation [242]. In most cases, the thrombus overlies a ruptured plaque or an intact plaque with superficial endothelial erosion. In recent years, it has been recognized that plaque composition rather than plaque size or severity of stenosis is important for plaque rupture and subsequent thrombosis [243] . Exposure of blood to the procoagulant materials in the ruptured plaque promotes thrombosis. Catalytically active tissue factor is present in the atherosclerotic plaque and seems to play a major role in the initialization of thrombosis following rupture [244,245]. Under arterial shear stresses, only platelets are capable of adhesion to the damaged vessel wall. Several adhesion molecules and platelet agonists are present in the plaque such as collagen, and oxidized lipids [246]. In contrast to venous thrombosis, activation of coagulation factors does not appear to play a major role in arterial thrombosis as these factors are likely to be removed by the high flow in the arterial system. Vessel wall damage due to atherosclerosis, hypertension, or vascular anomalies is a major risk factor for arterial thrombosis, by inducing turbulence and altered blood flow, which allows platelet adhesion. Consequently, hyperactivity of platelets also plays a role in the pathogenesis. Antiplatelet therapy such as aspirin is used very successfully to prevent vascular events in arterial thrombosis while they are of limited value in preventing venous thrombosis.

6.3. Microvascular Thrombosis

Thrombosis of large arteries and veins typically receive the greatest attention clinically and experimentally, likely due to the health and economic impact of those disorders. Thrombosis may also occur within the microcirculation, which represents the largest proportion of the surface area of the vasculature. Microvascular thrombi have been demonstrated in a number of severe clinical diseases, including thrombotic thrombocytopenic purpura, sepsis, disseminated intravascular coagulation, and antiphospholipid syndrome among others [247250]. The ultrastructure of thrombi in these diseases varies widely, ranging from platelet-rich “white” thrombi to fibrin- and erythrocyte-rich “red” thrombi. Microvascular thrombosis has been demonstrated in various organs and various vessel types in those disorders, though their tissue and vessel distribution vary among those entities. Microvessels consist of arterioles, capillaries and venules and demonstrate considerable heterogeneity in vascular wall structure, function, and endothelial interactions with blood constituents. Figure 6.1 illustrates a few of the prominent differences between arterioles and venules: endothelial cell morphology, leukocyte-endothelial interactions, vascular smooth muscle content, and endothelial von Willebrand factor expression.

Figure 6.1. Example of several structural differences between arterioles and venules.

Figure 6.1

Example of several structural differences between arterioles and venules. A) Silver-stained microvessels of the rat mesentery. The endothelial cell borders (arrows) reveal long spindle-shaped cells in arterioles (Art.) and polygonal cells in venules (Ven.). (more...)

Of interest, in addition to the marked structural differences between arterioles and venules, these vessels demonstrate significant differences in their responses to thrombosis induced by various forms of endothelial injury, as determined by intravital microscopy-based models. These include a predisposition of venules to microvascular thrombosis as well as significant differences in kinetics of thrombus formation and embolization between venules and arterioles [86,134,167]. These studies also suggest differences in endogenous mechanisms preventing thrombosis and embolization among venules and arterioles. For example, data derived from micropuncture experiments demonstrate that prostacyclin-induced inhibition of thromboembolism predominated in arterioles, while nitric oxide-induced inhibition predominated in venules [109,110,251,252]. Similarly, the influence of pathologic stimuli on experimental microvascular thrombosis differs among these vessel types, while endotoxemia promotes photochemically-induced microvascular thrombosis predominantly in venules, experimental colitis promotes thrombosis mainly on arterioles [92,138,253].

The mechanisms accounting for the differences in microvascular thrombotic responses between venules and arterioles are not fully understood. Differences in wall shear rate (much higher in arterioles) and leukocyte-endothelial interactions (which occur mainly in venules) have been suggested [134,254], though experimental evidence has failed to support them consistently [86,255]. Vascular-specific differences in nitric oxide, prostacyclin, or vWF may represent additional mechanisms contributing to the differential responses [92,109,110], and synergistic effects between these various candidates are yet to be defined.

While the mechanisms for microvascular thrombosis in each of the above-mentioned clinical disorders remain to be elucidated, a role for vWF (particularly ULVWF) is evident in certain conditions. For example, patients with a familial form of thrombotic thrombocytopenic purpura have a deficiency in ADAMTS-13, reduced cleavage of ULVWF and evidence of increased plasma levels of the large adhesive ULVWF multimers [256]. Platelet adhesion to ULVWF in the absence of ADAMTS-13 has been demonstrated both in vitro (see Figure 3.5) and in vivo [145,146]. Similarly, some patients with severe forms of sepsis, a systemic inflammatory response to an infection, have been shown to have a secondary form of ADAMTS-13 deficiency and evidence of circulating ULVWF [148,149]. The reduced ADAMTS-13 may contribute to microvascular thrombosis accompanying severe sepsis and disseminated intravascular coagulation [148,149]. The increasing interest in intravital microscopy-based studies is expected to broaden our understanding of the mechanisms of microvascular thrombosis in relevant models of human diseases.