The Hypercoagulable Patient

Christine S. Rinder

What Is the Influence of Hypercoagulability in Thrombosis?

Thrombosis in either the venous or arterial circulation may have catastrophic consequences. In the venous circulation, three vascular conditions occurring singly or, more often, in concert, put patients at increased risk for thrombosis. These include (a) stasis, (b) injury to vascular endothelium, and (c) hypercoagulability. Hypercoagulability may be thought of as a state of exaggerated activation of coagulation. A comparable triad that predisposes to arterial thrombi includes flow limitation from atherosclerosis, plaque rupture as a uniquely arterial form of endothelial injury, and arterial hypercoagulability, an entity just beginning to be understood. Fundamental to the prevention and management of hypercoagulability is a better understanding of thrombosis and its endogenous regulators. Accordingly, this review will begin by describing the most current model for the initiation and propagation of clotting, together with the role of natural anticoagulants, the activity of which is critical to the prevention of thrombosis. Using that model, the etiology of different hypercoagulable states will be explored.

Sources of hypercoagulability can be divided into two major classes: (a) a congenital predisposition caused by one or more genetic abnormalities, often referred to as thrombophilia and (b) acquired, or environmental hypercoagulability.

Genetic mutations that reduce the functional levels of endogenous anticoagulants—antithrombin, protein C, and protein S—have long been known to predispose to venous thromboembolism (VTE). However, new genetic abnormalities are being discovered that contribute to hypercoagulability. Indeed, with appropriate testing, causes of thrombophilia are being detected in as many as 50% of VTE cases. However, genetic factors are only rarely the sole contributor to VTE. Genetic sources of thrombophilia create a lifelong hypercoagulable state that give rise to clinical VTE only episodically. In most cases of VTE, hypercoagulability—acquired or environmental—serves as a triggering factor. As anesthesiologists are often actively involved in clinical events that create acquired hypercoagulability, for example, pregnancy, surgery, and malignancy, it is important to stay abreast of advances in the understanding and treatment of thrombosis.

Historically, blood clotting has been viewed as a series of enzymatic reactions in which the participants, sequence, and regulation were independent of location. Increasingly, recognition of the unique conditions operating in the arterial circulation, including (a) the high velocity of blood flow, creating shear forces in vessels of smaller diameter, (b) its complex rheology at vessel branch points, and (c) its unique anticoagulant requirements (i.e., better efficacy of antiplatelet agents as opposed to antithrombin agents), has mandated the development of an arterial clotting model distinct from that operating in the venous circulation.¹ In concert with this improved understanding of the physiology of arterial hemostasis has come an appreciation for how tightly regulated the controls on this system must be maintained. Indeed, given the spectrum of clinical complications associated with defects in arterial hemostasis, including catastrophic blood loss at one end and infarction at the other, it is not surprising that multiple interdependent factors keep a very tight rein on this process. Accordingly, although there are conditions that give rise to both venous and arterial hypercoagulability, a subset of heritable and acquired conditions exist that
uniquely predispose to arterial thrombosis. The current understanding in the field of arterial hypercoagulability is not as advanced as that of venous hypercoagulability, but has the potential to significantly improve the medical treatment of heart disease and stroke.

How Is Venous Thromboembolism Defined and Assessed?

VTE has an annual incidence that exceeds 1 per 1,000, with approximately 2 million cases in the United States alone, and more than 150,000 deaths per year resulting from pulmonary embolism. These statistics put VTE in a class with clinical events such as stroke. As early as the late 1800s, Virchow identified the three vascular conditions predisposing to VTE: (a) stasis, (b) injury to vascular endothelium, and (c) hypercoagulability. The first two made intuitive sense, but very little was known at the time about the pathophysiology of hypercoagulability. A better understanding of the physiology of normal hemostasis, especially factors that regulate the velocity of clot growth and its composition, has paved the way for recognizing the genetic and acquired conditions that create a state of heightened coagulation activation, that is, hypercoagulability. Accordingly, this review will begin with an overview of the current model of normal coagulation.

How Does Normal Venous Coagulation Occur? The Two-Phase Model

Fifty years ago, two groups simultaneously described the “waterfall” or “cascade” model (see Fig. 38.1) of soluble coagulation.²,³ This model allowed great strides to be made in identifying the series of proteolytic reactions that culminate in a fibrin clot. The cascade model seemed to fit with the clotting assays that were developed to guide warfarin and heparin dosing. These tests became the gold standard for measuring soluble coagulation. Although this cascade model is practical in clinical scenarios associated with deficits of one or more factors, it fails to explain the bleeding diathesis associated with hemophilia, and offers little insight into thrombophilias. As will be emphasized later in the chapter, the critical determinants of whether a clot will form—be it an appropriate response to bleeding or a pathologic event—are the efficiency and velocity of each of the reaction steps (e.g., how much, how fast).

Recent advances in cell-based research models have made significant strides in clarifying the dynamics of coagulation.⁴ In vivo coagulation follows exposure of the blood to a source of tissue factor (TF), typically on the surface of a fibroblast or other subendothelial cell in contact with blood elements through damage to the endothelial cells lining the vessel lumen. The intrinsic, or contact, pathway of coagulation plays no role in these earliest clotting events. Tissue factor-initiated coagulation has two phases: An initiation phase and a propagation phase⁵,⁶ (see Fig. 38.2).

FIGURE 38.1 The Cascade or Waterfall model of coagulation. The intrinsic pathway consists of high molecular weight kininogen (HK), prekallikrein (PK), and the zymogens, factors XII, XI, IX, and VIII. The extrinsic pathway consists of tissue factor (TF) and factor VII, and the common pathway, factors X, V, prothrombin, and fibrinogen, culminating in the generation of thrombin and fibrin. The activated form of these factors is indicated by adding the letter “a” as a suffix. Reactions accelerated by the presence of a phosphatidylserine surface are indicated by “PS^–”.

The initiation phase begins as exposed TF binds to factor VIIa, picomolar amounts of which are present in the circulation at all times. This VIIa-TF complex catalyzes the conversion of very small amounts of factor X to Xa, which in turn generate nanomolar amounts of thrombin. Collagen exposed in the subendothelium is capable of activating nearby circulating platelets, causing them to expose phosphatidylserine and release factor V, thereby increasing prothrombinase formation and thus the likelihood of thrombin generation.

The seemingly trivial amount of thrombin formed during initiation sparks the inception of the propagation phase, the successful completion of which culminates in explosive thrombin generation and, ultimately, fibrin deposition. More than 96% of the total thrombin generated during clotting occurs during propagation. The commonly used laboratory tests of soluble coagulation only measure the kinetics of the initiation phase.⁵ The prothrombin time (PT) and activated partial thromboplastin time (aPTT) both have as their endpoints the first appearance of fibrin gel, which occurs with <5% of the total reaction complete. Therefore, the fibrin clotting that signals completion of the PT/aPTT occurs when only minimal thrombin levels have been formed. These tests are sensitive at detecting complete deficiencies in clotting factors (e.g., hemophilia) and guiding warfarin/heparin
therapy; however, they do not model the entire sequence of events necessary for effective hemostasis. They fail to give information relevant to thrombin generation during the propagation phase, which determines whether a persistent clot will form, or whether endogenous anticoagulants and fibrinolytic regulators are able to constrain excess clot growth.

FIGURE 38.2 The two-phase model of coagulation. In the initiation phase of coagulation, tissue factor (TF) exposed on a subendothelial fibroblast after vessel injury complexes with small amounts of factor VIIa present in the circulation. This complex then activates a small amount of factor X to Xa in the presence of an activated platelet. The platelet-bound Xa converts a tiny amount of prothrombin to thrombin. This small amount of thrombin then sparks the propagation phase of coagulation. Thrombin activates factors XI, VIII, and V at or near the activated platelet. Factor IX is activated by either factor XIa or the TF-VIIa complex. Factor IXa complexes with the factor VIIIa activated by thrombin, and on the platelet surface generates factor Xa with remarkable kinetic efficiency. The platelet-bound factor Xa complexes with factor Va, which converts prothrombin to explosive amounts of thrombin. This thrombin in turn converts fibrinogen to fibrin, thereby sealing the vessel injury beneath.

Thrombin generated during the initiation phase is a potent platelet activator, thereby providing both an activated platelet surface membrane and platelet-released factor V (which thrombin promptly converts to Va). Factor VIII, conveniently brought to the bleeding site by its carrier, the von Willebrand factor (vWF), is also activated by thrombin—a step that causes its release by vWF. FVIIIa then complexes with the picomolar amounts of factor IXa, also generated by the TF-VIIa complex during the initiation phase, forming the FVIIIa-IXa complex, a pivotal point in the successful generation of a clot.

The formation of the FVIIIa-IXa complex on the platelet surface heralds the switch from FXa generation by the TF-VIIa complex to the intrinsic Xase pathway. This switch is of enormous kinetic advantage, with the intrinsic Xase complex exhibiting approximately a 50-fold greater efficiency at Xa generation. The bleeding diathesis associated with hemophilia is testament to the hemostatic importance of the exuberant thrombin generation formed during the propagation phase.⁷ Although congenital deficiencies in VIII and IX do prolong the aPTT, it is the thrombin generation in the propagation phase—a function not evaluated by the aPTT—that is most impaired in hemophilia.

The activated platelet, stimulated by thrombin formed in the initiation phase, expresses membrane receptors for VIIIa and IXa. When these active proteases are bound in combination with negatively charged phosphatidylserine, the resulting enzyme complex enhances the binding of its substrate, factor X. The speed of X activation by this complex is also increased by platelet binding; indeed, when compared to the reaction speed of free proteases in solution, assembly of the entire reaction on the platelet membrane increases the catalytic efficiency of X activation by approximately 13 million-fold.

Assembly of the prothrombinase complex is similarly dependent on the activated platelet surface for optimum activity. Much of factor V released by the activated platelet (subsequently thrombin-activated) stays bound to the platelet surface. Like the Xase complex, the platelet-bound prothrombinase complex (i.e., Xa-Va-phosphatidylserine) enhances prothrombin activation to thrombin 300,000-fold faster than free Xa and Va acting on prothrombin formed in solution. Platelet-bound Xa is the rate-limiting reagent in prothrombin cleavage. The substrate for this enzyme complex, prothrombin, binds to GPIIb/IIIa on both activated and inactivated platelets, potentially providing a source for thrombin generation in both the initiation and propagation phases.

Evidence that factor XI further amplifies the propagation phase is growing.⁷ As noted in the preceding text, Xa levels are rate limiting to the prothrombinase complex, particularly once the switch is made from extrinsic (FVIIa:TF complex) to intrinsic (FVIIIa:FIXa complex) Xase. Small amounts of IXa can be generated by the TF-VIIa complex, but its ability to sustain IXa generation is limited by its endogenous inhibitor, tissue factor pathway inhibitor (TFPI).⁷ To generate Xa in amounts sufficient to fuel the propagation phase, an alternative and kinetically superior source of IXa is needed. Factor XI is another zymogen activated by the minute amounts of thrombin generated during initiation, but this activation only occurs on the activated platelet surface.⁷ Platelet-bound FXIa is ideally located to activate FIX, which also binds to the platelet surface, helping to speed and localize this activity. Additionally, binding to the platelet surface protects FXIa from its inhibitor, protease nexin 2. Therefore, FXIa generation on the activated platelet is key to providing sufficient FIXa to maintain Xa generation through the catalytically more efficient intrinsic Xase complex.

In the venous circulation, the kinetic advantage of coagulation cascade assembly on the activated platelet surface is readily apparent; recent in vitro work has clarified the minimum platelet count (i.e., the dose of platelets) necessary for the reaction sequence to proceed.⁸ When all coagulation zymogens are present, thrombin is not generated unless platelets are present as a source of phospholipid. Once platelets are added, thrombin generation begins and grows with increasing numbers
of platelets, up to a threshold of 10,000 platelets per µL. Increases in platelets beyond this number have no effect on the efficiency of the reaction, suggesting that, as had been empirically observed in thrombocytopenic patients, the platelet level must decrease to very low levels (i.e., <10,000 per µL) to increase the risk of venous bleeding. This contrasts sharply with the arterial circulation where the minimum platelet count needed to ensure hemostasis for operative procedures is at least five times that number.

What Are Endogenous Anticoagulants?

In addition to providing a kinetically favorable orientation, the platelet surface membrane protects the active coagulation enzymes from inactivation by proteases circulating in plasma. During the initiation phase, binding of FXa to the platelet membrane protects it from inactivation by both TFPI and antithrombin. Normal plasma levels of both TFPI and antithrombin inhibit soluble FXa so efficiently that its plasma half-life is ≤1 minute.⁴ Preservation of small amounts of FXa that are generated during this “make or break” stage of coagulation is critical to formation of the nanomolar amounts of thrombin needed to begin the propagation phase. Similarly, the thrombin formed during the initiation phase must also be protected from inactivation by antithrombin. Antithrombin is present at over twice the concentration (3.2 µmol per L) of the highest concentration of thrombin reached during the propagation phase (1.4 µmol per L) and approximately 1,000 times the nanomolar amounts of thrombin generated during the initiation phase. Without the protection conferred by the platelet membrane, normal plasma levels of antithrombin inhibit soluble thrombin, with a half-life of <1 minute. Therefore, thrombin generated during the initiation phase is critically dependent on protection by the activated platelet membrane to have sufficient time to transition from the initiation phase to the propagation phase.

The other critical physiologic regulator of thrombin generation is protein C. Unlike antithrombin, which circulates in an active form (albeit capable of enhanced kinetics when bound to heparin or endogenous heparan sulfate), protein C requires activation by thrombinstimulated endothelial cell thrombomodulin. Activated protein C (APC) then acts in concert with its cofactor, protein S, to limit the rate of thrombin generation by inactivating the essential procoagulant cofactors, FVa and FVIIIa.⁹ Again, the activated platelet membrane promotes clot success by protecting FVIIIa and FVa from inactivation by APC.¹⁰

TABLE 38.1 Major Hereditary Conditions Linked to Hypercoagulability^a

Conditions	Prevalence in Healthy Controls	Prevalence in Patients with First DVT	Likelihood of DVT by Age 60
Antithrombin deficiency	0.2%	1.1%	62%
Protein C deficiency	0.8%	3%	48%
Protein S deficiency	1.3%	1.1%	33%
Factor V_Leiden	3.5%	20%	6%
Prothrombin 20210A	2.3%	18%	<5%
^aAll numbers pertain to heterozygous state.
DVT, deep venous thrombosis.
Adapted with permission from: van der Meer FJ, Koster T, Vandenbroucke JP, et al. The leiden thrombophilia study (LETS). Thromb Haemost. 1997; 78:631.

What Contributory Role Does Genetics Play in Heritable Hypercoagulability?

The model of the soluble coagulation system and its opposing anticoagulant system highlights the importance of both the speed and amounts of formed thrombin in determining whether a clot will persist. Accordingly, the genetic abnormalities contributing to hypercoagulability can be divided into two groups: One consists of genes that cause a surplus in prothrombotic proteins and the other genes that cause deficiencies in antithrombotic proteins. The net effect of either of these thrombophilias is enhancement in either the generation of thrombin or its persistence in the circulation.

▪ THROMBOPHILIA DUE TO DECREASED ANTITHROMBOTIC PROTEINS

Antithrombin (aka ATIII) is the most important of the body’s defenses against clot formation in healthy vessels or at the perimeter of a site of active bleeding.¹¹ Homozygous AT deficiency is generally not compatible with life or even fetal survival to term. Individuals heterozygous for AT deficiency are approximately 20 times more likely than nondeficient individuals to develop VTE at some point in their lives, usually in association with a triggering event that further increases their hypercoagulability (see Table 38.1). In one study of 18 AT-deficient individuals,
more than 40% of the VTE that developed did so in association with either surgery or pregnancy.¹² Only 11% of VTE were spontaneous, that is, had no known precipitating factors.

Hereditary deficiencies in protein C (PC) and protein S (PS) also adversely impact thrombin regulation. However, instead of limiting the activity of thrombin already formed, congenital deficiencies in PC and PS hamper the affected individual’s ability to limit rates of thrombin generation. Similar to homozygous AT deficiency, many infants homozygous for PC or PS deficiency do not survive long after birth. With heterozygous deficiencies, the relative surplus of Va and VIIIa that results from defective inactivation ensures that both the tenase and prothrombinase complexes are able to act with enhanced kinetics, generating an overabundance of thrombin and setting the stage for VTE risk of the same order of magnitude as AT deficiency (Table 38.1). Moreover, the synthesis of PC and PS are both vitamin K dependent, with PC having the shorter half-life. Accordingly, individuals who are PC-deficient are at particular risk for thrombosis if warfarin therapy is initiated without the protection of initial anticoagulation with heparin. Specifically, during the first days of warfarin treatment, before inhibition of vitamin K has decreased factors VII, IX, and XI sufficiently to provide the intended anticoagulation, modest suppression of PC synthesis may compound the already subnormal PC levels, resulting in paradoxically heightened hypercoagulability.

Considerable evidence suggests that, in addition to serving as a cofactor for APC, PS may generate additional, direct anticoagulant activity. Protein S acts to inhibit the development of the prothrombinase complex by inhibiting Xa in the absence of Va, and by inhibiting the binding of the substrate, prothrombin, to factor Va.¹³ Therefore, decreases in PS, in addition to limiting the effectiveness of APC, may also allow the generation of an overabundance of the rate-limiting FXa.

With respect to the regulatory protein, TFPI, no congenital or acquired deficiencies have been described that appear to predispose to VTE at this time. One investigation,¹⁴ however, suggests that among patients with VTE of unknown etiology, the average response to recombinant TFPI is reduced compared to controls, and individuals whose TFPI sensitivity was below the 10th percentile of controls had 13 times greater likelihood of a VTE. Further investigation is required to substantiate this observation and determine whether a reduced response to TFPI is attributable to alterations in FVII, FX, or another pathophysiology.

▪ THROMBOPHILIA DUE TO INCREASED PROTHROMBOTIC PROTEINS

A number of thrombophilias have been described resulting from increased levels of prothrombotic proteins. Two thrombophilias deserve particular notice because of their relatively high prevalence. Dahlback first described APC resistance in a single family in 1993, and subsequently found that among other VTE patients, their plasma often exhibited resistance to the normal anticoagulant effect of APC.¹⁵ Specifically, the addition of exogenous APC to their plasma did not prolong the aPTT of these VTE patients when compared with the prolongation found by APC treatment of plasma from non-VTE controls. Several groups subsequently demonstrated that approximately 90% of patients with APC resistance have an activated form of FV that is resistant to proteolysis by APC.¹⁶ The gene responsible for this effect, factor V_Leiden, differs from the normal gene by a single nucleotide, producing an amino acid substitution at one of the sites where APC normally cleaves FVa, thereby rendering it refractory to inactivation. Accordingly, FVa_Leiden stays active in the circulation longer than normal, fostering increased thrombin generation.

As the sole source of hypercoagulability, FV_Leiden is viewed as having low to intermediate procoagulant risk. Patients heterozygous for FV_Leiden have a fivefold to sevenfold increased risk of VTE (Table 38.1), whereas the risk for homozygous carriers is increased up to 80-fold. FV_Leiden may be present at a high frequency in the general population. Its prevalence varies considerably in different ethnic populations. It is present in approximately 5% of people of northern European descent but rarely in patients of African or Asian descent.¹⁷ Accordingly, depending on the ethnic makeup of the community, up to 1 in 20 patients presenting for routine surgery can be expected to have a degree of heightened risk attributable to this gene. A very small minority of patients who demonstrate APC resistance do not carry the gene for FV_Leiden, but their VTE risk is similar to that for heterozygous FV_Leiden carriers, and therefore, for the purposes of perioperative risk, they may be treated identically. Currently, however, testing for APC resistance is not indicated for routine preoperative screening in the absence of a VTE history.

Another thrombophilia that operates through an increase in prothrombotic proteins is known as the prothrombin gene mutation (G20210A). This gene was described by Poort et al. in 1996,¹⁸ noting that 18% of VTE patients and approximately 1% of healthy controls had a mutation in the gene for prothrombin at base 20210. This particular location is in the 3-prime region of the gene that is not translated but contains the “stop” condon. The mutation renders the “end” cleavage signal of the gene inefficient, causing additional amounts of mRNA to be transcribed. Accordingly, the levels of the inactive zymogen, prothrombin, are considerably higher in affected individuals than in the general population. Unlike FV_Leiden, which enhances levels of the active enzyme, the prothrombin gene mutation increases the levels of substrate for the enzyme complex. When this mutation is the only thrombophilic risk factor, the VTE risk is relatively low¹⁹ (Table 38.1); most carriers of this gene will not have had an episode of VTE before age 50. The importance of this thrombophilia, as for FV_Leiden, resides in the frequency of the gene, rather than its potency. Also similar to FV_Leiden, ethnicity plays a significant role in the prevalence of this gene, occurring in approximately 4% of
individuals of European descent, but rarely in patients of African or Asian descent.²⁰

After excluding the thrombophilias described in the preceding text, patients with VTE are more likely to have increased concentrations of zymogens for factors VIII, IX, and XI compared with healthy controls.²¹,²²,²³ Other than theoretically providing a surplus substrate to optimize enzyme kinetics, it is not known exactly why elevated levels predispose to thrombosis or what causes their plasma factor levels to be higher. Genetic mutations that have yet to be elucidated may be present, possibly resulting in either elevated zymogen production or prolonged plasma survival. The degree of risk enhancement for these increased levels is low to moderate. One study²³ estimated that VTE risk increased 10% for each 100% FVIII increment, and another study²¹ demonstrated that patients with FXI levels above the 90th percentile had more than a twofold increased VTE risk. Van Hylckama et al.²² demonstrated a twofold to threefold increased risk for patients with higher FIX levels. These risks are comparable to the risks for the heterozygous FV_Leiden or prothrombin 20210A thrombophilias.

▪ HYPERCOAGULABILITY OF UNCLEAR ETIOLOGY

Hyperhomocysteinemia has been identified as a risk factor for both accelerated atherosclerosis and VTE, although the magnitude of the latter is unclear at present.²⁴ Homocysteine is a thiol-containing amino acid that readily undergoes auto-oxidation in the plasma, forming the oxidized dimer, homocysteine, which participates in redox reactions that produce oxygen radicals. At physiologic homocysteine levels, endothelial cell-released nitrous oxide binds to homocysteine, forming S-nitrosothiol, thereby inactivating it. It is hypothesized that at supraphysiologic homocysteine levels, nitrous oxide-mediated regulation is overwhelmed, and homocysteine-derived oxygen radicals can cause local tissue injury. Hyperhomocysteinemia can be found in patients with dietary deficiencies of folate and B vitamins, and can also be hereditary, most commonly due to mutations affecting the cysteine β-synthase gene or the methyltetrahydrofolate reductase gene. The homocysteine levels associated with these mutations are highly variable, and may only become apparent with coexisting dietary deficiencies of folate, vitamins B₁₂, and B₆. The exact mechanism by which hyperhomocysteinemia predisposes to both atherosclerosis and VTE is not certain, but evidence points toward endothelial cell injury, perhaps mediated by oxygen radicals and predisposing to plaque buildup in arterial vessels and impaired antithrombotic activity in their venous counterparts. Such risks for vascular injury are certainly magnified in patients homozygous for the cysteine β-synthase mutation, but it is currently much less clear that mild hyperhomocysteinemia represents another hereditary thrombophilia.

TABLE 38.2 Enhanced Venous Thromboembolism Risk during Pregnancy in the Thrombophilic Parturient

Criteria	Antithrombin Deficiency (%)	Protein C Deficiency (%)	Protein S Deficiency (%)	Factor V_Leiden or Prothrombin Mutation (Homozygous) (%)	Factor V_Leiden or Prothrombin Mutation (Heterozygous) (%)
No VTE history	10-20	10-20	<3	<3	<3
VTE in one or more first-degree relative	10-20	10-20	10-20	10-20	3-10
Personal VTE history	>20	10-20	10-20	10-20	3-10
VTE, venous thromboembolism. Adapted from: McLintock C, North R, Dekker G. Inherited thrombophilias: Implications for pregnancy-associated venous thromboembolism and obstetric complications. Curr Probl Obstet, Gynecol Fertil. 2001;24:109.

▪ HYPERCOAGULABILITY FROM THROMBOPHILIC COMBINATIONS

Many VTE patients are found to have more than one risk factor, whether carrying the genes for two congenital thrombophilias or a combination of a congenital and acquired thrombophilia. The coinheritance of FV_Leiden and prothrombin G20210A gene mutations doubles the risk of recurrent VTE, compared with carriers of the FV_Leiden gene mutation alone.²⁵ The risk of symptomatic VTE among FV_Leiden patients undergoing total hip replacement is increased fivefold compared to noncarriers.²⁶ Wahlander et al. demonstrated that the prothrombin G20210A mutation had a ninefold increase in symptomatic VTE after hip/knee replacement surgery despite 8 to 11 days of antithrombotic prophylaxis.²⁷ For some thrombophilias, the risk of surgery appears to synergize with a baseline thrombophilic risk. Antithrombin deficient patients exposed to surgery have a VTE risk of approximately 12%, compared to only 0.8% per year for spontaneous VTE in AT-deficient patients and 1.2% per year for VTE risk attributable to the surgery alone.¹¹ The risk of pregnancy-related VTE is also increased by the presence of a thrombophilia (see Table 38.2). Most studies have found the VTE risk to be higher in the postpartum period, compared to antepartum, and the risk is higher still after cesarean section than after a vaginal delivery, including a 10-fold higher risk of fatal pulmonary embolism.²⁸

TABLE 38.3 Epidemiology of Venous Thromboembolism (VTE)

Independent Risk Factors^a	Odds Ratio^b
No institutionalization or recent surgery	1.00
Pregnancy and first 6-8 weeks postpartum	5-6
Malignancy with chemotherapy	6-7
Institutionalization without surgery	7-8
Institutionalization with recent surgery	20-22
^aWithout antithrombotic prophylaxis. ^bAdapted from: Heit JA, Silverstein MD, Mohr DN, et al. The epidemiology of venous thromboembolism in the community. Thromb Haemost. 2001; 86:452.