Chapter 178. Hypercoagulable States

1. Which hospitalized patients are at risk for venous thromboembolism?

2. Who is at increased risk for complications from thromboprophylaxis?

3. How do you make sure that thromboprophylaxis is prescribed to all patients who should get it?

Many patients have an increased baseline propensity for thrombosis due to underlying hypercoagulable states, the majority of which are acquired. An acute insult may then result in hospitalization, which usually further increases the risk due to immobilization, procedures including central venous catheter placement or pacemaker, and/or administration of certain drugs. The main question in most patient groups detailed below is whether they can expect a net benefit from pharmacologic thromboprophylaxis, mainly low-molecular-weight heparins (LMWHs). Ideally, one has to balance the risk and consequences of venous thromboembolism (VTE) with the risk of bleeding and other complications of anticoagulants in every patient, while also taking the values and preferences of that patient into consideration. To facilitate this balancing process, institutions and national and international organizations have developed guidelines for the use of thromboprophylaxis. The use of these guidelines and local protocols has greatly improved patient outcomes, mainly by increasing the use of pharmacologic thromboprophylaxis in patients at risk for VTE. The authors strongly advise the development and implementation of a local protocol.

Deficiencies of Natural Inhibitors of Coagulation

Less than 1% of the general population has a congenital deficiency of one of the natural inhibitors of coagulation: antithrombin, protein C, and protein S. These deficiencies are associated with a high lifetime risk of VTE (60–70% in a recent study) and with a high risk of recurrence after a first event (50% after 10 years). A limitation to these data is that they are derived from highly selected families, which might have led to an overestimation of risk. No population-based data are available.

Acquired protein C deficiency may be seen with disseminated intravascular coagulation (DIC), extensive deep vein thrombosis (DVT), severe liver disease, infection, malignancy, acute respiratory distress syndrome (ARDS), the hemolytic uremic syndrome, thrombotic thrombocytopenic purpura (TTP), and following L-asparaginase therapy. Although levels of protein C may decrease postoperatively, the significance is unknown. Acquired protein S deficiency may be seen with DIC and severe liver disease. Total and free protein S decrease steadily during pregnancy, with the lowest levels at term. Acquired antithrombin deficiency may be seen due to decreased synthesis (liver disease, following L-asparaginase therapy), malnutrition, or consumptive states (acute thrombosis, DIC, malignancy, pre-eclampsia). Most patients, however, will have normal antithrombin levels within 24 hours of initiation of heparin therapy. A clue to antithrombin deficiency may be inadequate or failure of activated partial thromboplastin time (PTT) to prolong during intravenous heparin therapy. Because protein C and protein S are vitamin K–dependent factors, warfarin-induced skin necrosis may occur with these deficiencies.

Patients with a known deficiency, primarily of antithrombin but to some extent also protein C and protein S, have a strong indication for pharmacologic thromboprophylaxis in any situation that increases the risk of thrombosis, including lower-limb immobilization, pregnancy, and bed rest as a medical patient. Given the high risk of recurrence, long-term anticoagulation should be considered in these patients after the first episode of VTE, especially if it was unprovoked (ie, not associated with a transient major risk factor for VTE, such as major surgery or trauma).

The utility of testing for a deficiency of a natural inhibitor is strongly debated, mainly because of its low prevalence. Testing in the context of thromboprophylaxis is not indicated. Even in patients who present with DVT, the prevalence of these deficiencies is so low that testing is rarely cost effective. Most centers limit thrombophilia testing to those patients who are young, who have recurrent VTE, or who have a strong family history.

Common Thrombophilic Mutations

Heterozygous factor V Leiden (causing resistance to degradation by activated protein C) and prothrombin mutation occur in 5% and 2% to 3%, respectively, of the Caucasian population and rarely in other races. The lifetime risk of VTE associated with these mutations is 5% to 10% in unselected populations. While the risk of a first VTE is increased, these mutations hardly increase the risk of recurrence after a first event. Testing for factor V Leiden and prothrombin mutation is not recommended in routine clinical practice, as their (heterozygous) presence does not influence the indication for, or intensity of, thromboprophylaxis or the duration of treatment for VTE.

Elevated Levels of Factor VIII

Increased levels of coagulation factor VIII are a mild risk factor for VTE. As with heterozygous factor V Leiden and prothrombin mutation, routine testing for high factor VIII in patients who present with VTE is not useful. A previously documented high level of factor VIII does not influence the indication or intensity of thromboprophylaxis.

However, evidence is accumulating that high levels of factor VIII and D-dimers in patients one month after discontinuation of anticoagulation for VTE might predict a high risk of recurrence of VTE. It is not yet common clinical practice to tailor duration of anticoagulation according to these tests, but trials are ongoing.

Homocysteinemia and Homocystinuria

Deficiency of folic acid or vitamin B12 accounts for approximately two-thirds of patients with elevated plasma homocysteine levels. This condition becomes more prevalent with increasing age, end-stage renal disease, and in pregnancy, and has been associated with occlusive arterial disease and with VTE, often at unusual sites (thrombosis in the central, cerebral, or splanchnic veins). Certain mutations that result in a high homocysteine level have only been weakly associated with VTE, and lowering of the level with vitamin supplementation has not been shown to be effective in reducing the risk of thromboembolism.

Homocystinuria, an autosomal recessive disease resulting from cystathionine beta-synthase deficiency, is associated with a tenfold or greater increase in homocysteine levels and has a prevalence of four per million. Approximately 20% of affected patients suffer from arterial and venous thromboembolism.

Antiphospholipid Syndrome

The antiphospholipid syndrome (APS) is defined as a combination of laboratory evidence (ie, presence of lupus anticoagulant or antibodies against cardiolipin or β2-glycoprotein I) and clinical manifestations (ie, at least one thrombotic event or obstetric complications). Thromboembolic events are more often venous than arterial and often involve unusual locations. The strongest laboratory association with thromboembolism is for lupus anticoagulant, followed by β2-glycoprotein I antibodies, and then cardiolipin antibodies; combined serologic abnormalities result in even stronger associations. When the APS is secondary to systemic lupus erythematosus, the presence of lupus anticoagulant or antibodies against cardiolipin increases the risk of VTE sixfold and twofold, respectively.

Primary prophylaxis with anticoagulation in APS should be provided for high-risk situations (surgery, trauma, puerperium), whereas aspirin is suggested for prophylaxis during pregnancy.

After a VTE event, secondary prophylaxis should be with the same intensity (warfarin with a target international normalized ratio (INR) of 2.0–3.0) as for other VTE. The duration of anticoagulation needs to be assessed after three to six months, taking into account whether the event was unprovoked, the presence of multiple serologic abnormalities, the serologic titer, compliance, and patient preferences. The risk of recurrence, including fatal events, is increased and some patients may experience recurrence in spite of anticoagulant therapy. This may be due to falsely prolonged coagulation times, and a switch from warfarin to long-term LMWH, or to monitoring warfarin with a coagulation factor X method, is then helpful. Patients with APS and arterial events should receive aspirin since there is no evidence that warfarin is more effective. Pregnancy loss in APS is most typical in patients with systemic lupus or with previous pregnancy complications. For these patients, heparin (unfractionated or LMWH) in combination with aspirin provides the highest success rate regarding live births.

The catastrophic APS (cAPS, Asherson syndrome) is characterized by microthrombi, anemia with schistocytes, and thrombocytopenia and is often triggered by an infection. The clinical course is rapid with multiorgan failure (usually pulmonary, cardiac and renal) and about 50% mortality, but this occurs only in 1% of patients with APS. Treatment with a combination of intravenous heparin, high-dose corticosteroids, and plasma exchange results in the highest rate of recovery (78%). The addition of intravenous immunoglobulin does not seem to improve the prognosis. The survivors may sometimes develop additional APS manifestations, but relapse of cAPS is very unusual.

Medically ill patients have a moderately increased risk of VTE, depending on their conditions, age, and treatments. In addition to potentially preventable mortality due to pulmonary embolism, VTE in medically ill patients prolongs hospitalization and increases costs. A growing body of evidence shows that thromboprophylaxis is underused in this patient group.

Pharmacologic prophylaxis is recommended for all patients admitted with congestive heart failure, severe respiratory disease, and those in intensive care unless there is an increased risk of bleeding. Hospitalized patients who have no contraindications should be prescribed pharmacologic prophylaxis if they have additional risk factors for VTE, including previous VTE, known thrombophilic states pregnancy (discussed later), inflammatory diseases, or malignancy. In case of recent, current, or high risk of bleeding, compression therapy of the legs should be used.

The elderly have a higher risk of VTE but also a higher risk of bleeding. Clinicians tend to overestimate the risk of bleeding and to underestimate the risk (and consequences) of VTE, thus withholding thromboprophylaxis from patients who are likely to benefit. However, older patients are more likely than younger patients to have renal dysfunction, which increases the bleeding risk with any anticoagulant. In patients with a creatinine clearance less than 30 ml/min, unfractionated heparin or vitamin K antagonists might be preferred over LMWH.

Sickle cell disease is associated with hypercoagulability through activation of both plasma coagulation and platelets. Ischemic stroke, pulmonary hypertension, avascular joint necrosis, and leg ulcers are viewed as thrombotic complications. The hypothesis that antiplatelet agents, warfarin, and heparin can prevent vasoocclusive events and subsequent complications has been tested, but studies were small and results not conclusive. Therefore, maintenance antithrombotic therapy is not routinely given. Patients who are hospitalized with a sickle cell crisis should be given thromboprophylaxis to prevent VTE.

Patients with the nephrotic syndrome have an increased risk of renal vein thrombosis, deep vein thrombosis, and pulmonary embolism. This is probably mediated by imbalances between pro- and antithrombotic factors. Estimates of the absolute risk vary from 2% to 40% of patients during their course of disease, with higher incidences in older studies. Randomized studies to guide prophylactic anticoagulant treatment in these patients have not been performed. Clinical practice is to consider vitamin K antagonists in patients with serum albumin below 2.0 to 2.5 g/dL (20 to 25 g/L) who do not have an increased risk of bleeding, especially when other risk factors including previous VTE, positive family history, and high body mass index are present.

The risk of VTE and need for adequate thromboprophylaxis in postoperative patients is well established. Early ambulation is sufficient in low-risk patients who undergo minor or laparoscopic or arthroscopic procedures, and mechanical prophylaxis can be used in patients with increased risk of bleeding, but all others should receive pharmacologic prophylaxis. This includes patients who undergo major general, orthopedic, gynecologic, urologic, and thoracic surgery. A prophylactic dose of LMWH is adequate for most patients, with fondaparinux as an alternative, and subcutaneous unfractionated heparin for patients with renal dysfunction. In most patients, thromboprophylaxis is given until discharge. For very high-risk patients (for example, major surgery for malignancy in a patient with previous VTE), thromboprophylaxis should be continued for about a month after surgery.

Patients after hip fracture surgery or hip or knee arthroplasty are at very high risk of VTE. This is illustrated by the fact that the first phase 3 studies with new anticoagulants are almost always performed in these patient groups. They should receive pharmacologic prophylaxis for at least 10 but preferably 35 days after surgery. The preferred agents are fondaparinux and LMWH, with vitamin K antagonists recommended for patients with renal dysfunction.

The main challenge in postoperative patients, as in the medically ill patients, is to ensure that thromboprophylaxis is prescribed for all patients who do not have contraindication. Every hospital should use a protocol or guideline, and adherence should be monitored.

During pregnancy and the postpartum period, women have a four to five times increased risk of VTE. One to two per 1000 pregnancies are complicated by VTE, which is the cause of 20% to 30% of all maternal deaths in the developed world. The risk of VTE is equal in the three trimesters of pregnancy, but does increase postpartum. About one-third of all pregnancy-associated deep vein thrombosis and about half of all pregnancy-related pulmonary embolism occur during the six weeks postpartum, as compared to the nine months antepartum. The increased risk of VTE during and after pregnancy is caused by both altered hemostatic balance and compromised venous flow. The risk of VTE is higher in obese and older women and after caesarian section as compared with vaginal delivery. Black women seem to be at higher risk than Caucasian women, while the risk is probably lower in Asian women.

Thromboprophylaxis is not indicated in healthy women during uncomplicated pregnancy or after uncomplicated vaginal delivery. There are no high-quality studies on routine thromboprophylaxis after caesarean section. The American College of Chest Physicians guidelines advise against it, while it is commonly used in European practice. Thromboprophylaxis should be used after caesarean section in patients with additional risk factors, such as emergency indication, delayed immobilization, and obesity (Table 178-1).

Antepartum thromboprophylaxis is indicated in women with a markedly increased risk of VTE (see Table 178-1). This includes women with a history of spontaneous or recurrent VTE and those with high-risk thrombophilia (deficiency of antithrombin, antiphospholid syndrome, and homozygous/compound heterozygous factor V Leiden or prothrombin mutation). Thromboprophylaxis should also be considered in hospitalized pregnant patients if they are ill or bedridden. Thromboprophylaxis for six weeks postpartum is given at a lower threshold of risk than prophylaxis antepartum, as the daily risk of VTE is much higher. In addition to women with an antepartum indication, this includes women with any history of VTE and lower-risk thrombophilia (deficiencies of protein C or protein S). For asymptomatic patients with low-risk thrombophilia (heterozygous factor V Leiden or prothrombin mutation), thromboprophylaxis is not usually prescribed. In many cases of increased risk of VTE, the decision to give or withhold thromboprophylaxis is not clear-cut, and the pros and cons should be discussed with the patient.

In the setting of thromboprophylaxis in pregnant women, LMWHs are preferentially used. They do not pass the placenta and do not have the potential for teratogenicity or fetal bleeding. However, they are associated with hypersensitivity skin reactions. If these appear, patients can be switched to another brand of LMWH, although the skin reactions tend to recur after a number of weeks. Fondaparinux or danaparoid can then be used as an alternative. Unfractionated heparin has the disadvantages of increased risk of osteoporosis when used for long (greater than one month) periods of time, heparin-induced thrombocytopenia, and inferior efficacy when used in nonpregnant populations. Warfarin is associated with teratogenicity when used between 6 and 12 weeks of gestation and with risk of fetal bleeding. If long-term use of LMWH is not feasible, warfarin might be an acceptable alternative between 12 and 36 weeks of gestation. For postpartum prophylaxis, both LMWH and warfarin are good options. Warfarin can be safely used in breastfeeding women.

One fifth of patients with a first event of VTE have undiagnosed cancer. The neoplasm causes hypercoagulability via production and secretion of procoagulant substances, by endothelial injury, platelet and monocyte activation, and by vascular occlusion. Solid tumors with the greatest risk of associated venous thromboembolism include glioma and adenocarcinoma of the pancreas, stomach, biliary tree, breast, and lung. Many chemotherapeutic agents and growth factors increase the risk of thrombosis further, and the risk also increases with the extent of the disease. In myeloproliferative disease, presence of the JAK2 mutation has been associated with increased risk for thrombosis. Strategies for screening for cancer in patients with VTE are shown in Table 178-2.

All hospitalized patients with active cancer should receive primary prophylaxis. If unfractionated heparin is used, the dose should be 5000 units three times a day. Prophylaxis should be given for at least a week after surgery and extension to one month is suggested. When a patient with cancer has developed VTE, anticoagulant treatment is more challenging than in other patients. The risk of recurrence despite adequate treatment is paralleled with an increased risk of bleeding complications. LMWH is a better choice than warfarin for secondary prophylaxis and should be given at full therapeutic dose for at least one month and may then be reduced to three-quarters of that dose. Anticoagulation should continue as long as there is active cancer but may be switched to warfarin (INR, 2 to 3) after three to six months.

Disseminated intravascular coagulation (DIC) is an acquired syndrome, caused typically by sepsis, trauma, organ destruction, cancer, obstetric complications, vascular malformations, severe liver failure, or severe toxic or immunologic reactions. Any of these may trigger excessive activation of the blood coagulation, with or without concomitant inflammation. The result is widespread fibrin formation in the microvasculature, rather than the normal concentration of a clot to an injured locus. In severe, uncompensated DIC organ dysfunction will ensue. In the early phase microthrombi will cause organ hypoxia, but if the DIC is not controlled and the hemostatic components are consumed faster than they can be replaced the patient will develop bleeding tendency, which eventually becomes generalized. The International Society on Thrombosis and Hemostasis has proposed an algorithm for diagnosis of overt DIC (Table 178-3).

The underlying cause has to be identified and treated or removed. The hypercoagulability may be controlled with active protein C concentrate in severe sepsis. Other treatment modalities include replacement fresh frozen plasma or more specific components (fibrinogen or cryoprecipitate, prothrombin complex concentrate, antithrombin). Unfractionated or low-molecular-weight heparin in subtherapeutic doses is sometimes used but should not be combined with antithrombin.

Thrombotic thrombocytopenic purpura (TTP) is a disease defined by microangiopathic hemolytic anemia (Figure 178-1) and thrombocytopenia, without an alternative explanation (for example, DIC, malignancy, or obstetric disorders).

Patients often present with fever and renal and neurologic abnormalities, but these are not required for the diagnosis. The annual incidence is low at about four per million. Primary TTP is caused by a deficiency or inhibition of ADAMTS13. This protein cleaves and inactivates the large, very active von Willebrand multimers that interact with platelets at the site of vascular injury. The absence of ADAMTS13 leads to unchecked accumulation of platelets and growth of thrombi, particularly microvascular thrombosis. ADAMTS13 levels can be measured, but turnaround times are commonly too long to help in making the diagnosis. A test result can be helpful to confirm the diagnosis though, especially when the response to therapy is not optimal. The natural course of TTP is fatal in over 90% of patients. Observations in the 1960s and 1970 led to the conclusion that normal plasma contained a factor that could reverse TTP (ie, what is now known to be ADAMTS13). A subsequent randomized trial showed that plasma exchange has even better results than plasma infusion: 78% versus 63% survival at six months. The current standard of care is that plasma exchange should be started the same day a clinical diagnosis of TTP is made. When plasma exchange is not immediately available, plasma infusion should be started first. A useful strategy is to continue daily exchanges of the patient's plasma volume until the neurologic status, platelet count, and lactate dehydrogenase (LDH) level have normalized. After this, plasma exchange on alternate days is continued for about a week. Some centers start low-dose aspirin (80 mg) as soon as the platelet count rises above 50 × 109/L. Although most patients will respond to this therapy, relapse is expected in about 30%. Nonresponding or relapsing patients are treated with immunosuppression, either steroids or rituximab.

The vasculitides have inflammation and fibroid necrosis in vessel walls as common denominators, but they differ regarding the size and type of blood vessels affected, cellular reactions, and serologic manifestations. In systemic lupus erythematosus both arterial disease with accelerated atherosclerosis, typically affecting the heart and the kidneys, and venous thrombosis are seen. Behçet disease (oral ulcers, genital ulcers, uveitis, retinitis, cutaneous vasculitis, synovitis, meningoencephalitis) is also associated with arterial and venous thrombosis in up to 40% of the patients, but pulmonary embolism is rare due to adherence to the vessel wall. Thromboembolic events may be the initial presentation of disease. In Buerger disease small arteries and veins distally in the extremities manifest occlusions. Conversely, in Kawasaki disease the systemic vasculitis has a predilection for the coronary arteries. In patients with antineutrophil cytoplasmic antibody–associated vasculitis the incidence of venous thrombosis is high. The interaction between the inflammatory process and the hemostatic mechanisms appears in both directions, and, therefore, the risk of thrombotic complications relates to the activity of the vasculitis.

Treatment with antiplatelet agents is recommended in giant cell arteritis and in Kawasaki disease, and for the latter, anticoagulation and thrombolytic therapy are indicated in certain phases. Long-term anticoagulation is also required in association with the antiphospholipid syndrome (discussed earlier). In general, there is a lack of evidence-based guidelines on antithrombotic prevention and therapy in the vasculitides.

Drug-induced hypercoagulapathy is associated with a number of drugs (Table 178-4).

Heparin-Induced Thrombocytopenia

Because millions of people are exposed to heparin each year, a high index of suspicion is required to prevent the sequelae of HIT which is associated with a increased thrombosis risk greater than 30 times that of a control population. HIT risk factors depend on:

• duration of heparin exposure (11–14 days > 5–10 days > 1–4 days)
• type of heparin (UFH (2.6%) > LMWH (0.2%)
• type of patient (surgery>medical>pregnancy)

Patients receiving unfractionated heparin for more than five days after surgery constitute a typical risk group for heparin-induced thrombocytopenia (HIT, see Chapter 175), which is an antibody reaction with new epitopes exposed on platelet factor 4 when it is bound to heparin. HIT may, however, also occur in medical patients and occasionally with LMWH. Only a small proportion of patients with this antibody reaction develop symptoms of venous or arterial thromboembolism, but in those few patients the hypercoagulability is pronounced and can result in loss of limb or life. Other symptoms indicative of HIT are necrotic skin reactions at the site of heparin injection, adrenal necrosis, and in the case of intravenous injection of heparin, anaphylactoid reactions. Patients at risk for HIT should be monitored with platelet count every two to three days for the first two weeks while on unfractionated heparin. Diagnosis using the 4T model is described in Table 178-5. When the platelet count falls by at least 50% or if a thrombotic event is observed within two weeks from starting treatment with heparin, even if already discontinued, screening for HIT antibodies should be done, heparin should immediately be replaced by a nonheparin parenteral anticoagulant, and warfarin should be reversed with vitamin K and should only be resumed when the platelet count has normalized or HIT is excluded. Alternative anticoagulants are lepirudin, argatroban, danaparoid, bivalirudin, and fondaparinux. Reexposure to heparin within 100 days from HIT is likely to cause recurrent symptoms without any delay and must be avoided.

Warfarin (Coumadin) Induced Skin Necrosis

Vitamin K antagonists may cause skin necrosis in approximately 1 in 5000 patients with 85% of the affected cases being women. The mechanism is an imbalance between the rapidly occurring profound depression of the natural anticoagulants protein C and protein S, and on the other hand a slower and mild reduction of the procoagulant vitamin K–dependent coagulation factors. Therefore, preexisting deficiency of protein C, protein S, or antithrombin is also a risk factor. Fibrin is then deposited in subcutaneous small veins and venules, typically in the breasts, thighs, or buttocks. The syndrome starts with localized pain and a maculopapular rash 3–10 days after initiation of warfarin therapy, and then progresses within 24–48 hour to hemorrhagic lesions, hemorrhagic bullae, and finally necrosis. This develops into an eschar with slow healing but often requiring plastic surgery or mastectomy. Immediately upon recognition of the syndrome, vitamin K should be given, warfarin should be stopped, the natural anticoagulants should be replaced with plasma or a specific concentrate (protein C or antithrombin), and anticoagulation with heparin should be continued. After healing of the lesions, the vitamin K antagonist may be resumed extremely carefully (start dose 1–2 mg/day) or more appropriately use of an alternative agent.

Paroxysmal Nocturnal Hematuria

Venous thrombosis occurs in about 40% of patients and may include unusual sites (eg, intraabdominal veins, cerebral veins and sinuses, small veins of the skin, and hepatic vein (Budd-Chiari) thrombosis).

Inflammatory Bowel Disease

The risk of thromboembolism in inflammatory bowel disease is moderately increased. It is often associated with increased levels of coagulation factors and a decrease in antithrombin.

Most hospitalized patients have multiple risk factors for VTE due to underlying hypercoagulability that they have acquired during their lives. The key questions to ask is whether the patient is at low risk of developing VTE (only a small minority of patients), is there a contraindication to pharmacologic prophylaxis (due to bleeding or allergy), and whether the patient is undergoing a procedure that makes pharmacologic prophylaxis problematic. All other patients should receive pharmacologic prophylaxis. The risk of VTE also extends beyond hospitalization and clinicians should include information about what the patient received and recommendations about prophylaxis when they transition patients to another care setting. Patients' first lifetime VTE may present in the community but have been acquired during a recent hospitalization.