Part 14: Disorders of Hemostasis, Thrombosis, & Antithrombotic Therapy

PRIMARY HEMOSTASIS

Hemostasis, the physiologic process that maintains vascular integrity and minimizes bleeding in the face of injury, requires a complex and coordinated interplay of cellular elements (platelets) and humoral (clotting) factors working in concert. Platelets are exposed to collagen, fibronectin, and other molecules in the subendothelial matrix at the site of vascular injury, which leads to platelet adherence, activation, and aggregation to form an initial platelet plug, a process that can be viewed as primary hemostasis. von Willebrand factor (vWF) facilitates platelet adherence to subendothelial collagen at the site of endothelial injury, especially in high-shear areas such as arterioles. While vWF also helps cross-link platelets, soluble fibrinogen is the primary molecule that links platelets together to form the platelet plug.

SECONDARY HEMOSTASIS

The serial, cascading activation of clotting factors at the site of endothelial injury leads to the formation of a fibrin clot, which constitutes secondary hemostasis. Platelets present at the site of a vascular injury facilitate clot formation by providing a phospholipid surface on which coagulation factor complexes assemble in space. The tissue injury (extrinsic) pathway of coagulation involves the conversion of factor X to factor Xa by activated factor VII (VIIa) in the presence of tissue factor. Factor Xa binds factor Va in the presence of phospholipid and calcium to form the prothrombinase complex. The common pathway involves the cleavage of factor II (prothrombin) by the prothrombinase complex to yield factor IIa (thrombin), and thrombin then cleaves fibrinogen to form fibrin, which is stabilized through cross-linking performed by factor XIIIa. The contact (intrinsic) pathway of coagulation begins with the activation of contact factors (kallikrein, XII, others), leading sequentially to the formation of factors XIa and IXa; the intrinsic pathway tenase complex (factors IXa, VIIIa), converts factor X to factor Xa, which binds factor Va in the presence of phospholipid and calcium, forming prothrombinase complex, which leads to formation of thrombin and fibrin mesh via the common pathway as outlined above.

In vivo, the tissue injury (extrinsic) pathway usually provides the initial activation of coagulation, which then activates the contact pathway coagulation cascade through thrombin-mediated activation of factor XI. Excessive activation of coagulation is controlled by the activity of endogenous proteins such as proteins C and S (which inactivate factor Va and factor VIIIa), thrombomodulin (via a negative feedback mechanism), and the plasminogen/plasmin system, which leads to fibrinolysis.

To evaluate patients for defects of hemostasis, the clinical context must be considered carefully (Table 14–1). Heritable defects are suggested by bleeding that begins in infancy or childhood, is recurrent, and occurs at multiple anatomic sites, although other patterns of presentation are possible. Acquired disorders of hemostasis typically are associated with bleeding that begins later in life and may relate to introduction of medications (eg, agents that affect platelet activity) or to onset of underlying medical conditions (such as kidney disease, liver disease, myelodysplasia, aortic stenosis, prosthetic aortic valve, myeloproliferative neoplasms), or may be idiopathic (acquired hemophilia A, acquired von Willebrand disease). Importantly, however, a sufficient hemostatic challenge (such as major trauma) may produce excessive bleeding even in individuals with normal hemostasis. A personal history of hemostatic challenges (eg, circumcision, trauma, injury during youth sports, tooth extractions, motor vehicle accidents, prior surgery, and pregnancy and delivery) and a family history of bleeding are critical when evaluating someone for a possible bleeding disorder.

Selected causes of thrombocytopenia are shown in Table 14–2. The age of the patient and presence of comorbid conditions can help direct the diagnostic workup.

The risk of clinically relevant spontaneous bleeding (including petechial hemorrhage and bruising) does not typically increase appreciably until the platelet count falls below 10,000–20,000/mcL (10–20 × 10⁹/L), although patients with dysfunctional platelets or local vascular defects can bleed with higher platelet counts. Suggested platelet counts to prevent spontaneous bleeding or to provide adequate hemostasis around the time of invasive procedures are found in Table 14–3. However, most medical centers develop their own local guidelines to have a consistent approach to such complex situations.

1. BONE MARROW FAILURE

General Considerations

Congenital conditions that cause thrombocytopenia include amegakaryocytic thrombocytopenia, the thrombocytopenia-absent radius syndrome, and Wiskott-Aldrich syndrome; these disorders usually feature isolated thrombocytopenia, whereas patients with Fanconi anemia and dyskeratosis congenita typically include cytopenias in other blood cell lineages. Mutations in genes (FLI1, MYH9, GATA1, ETV6, among others) that cause thrombocytopenia are being identified.

Acquired causes of bone marrow failure (see Chapter 13) leading to thrombocytopenia include, but are not limited to, acquired aplastic anemia, myelodysplastic syndrome (MDS), acquired amegakaryocytic thrombocytopenia (albeit a rare disorder), alcohol, and drugs. Unlike aplastic anemia, MDS is more common among older patients.

Clinical Findings

See Chapter 13 for symptoms and signs of aplastic anemia. Acquired aplastic anemia typically presents with reductions in multiple blood cell lineages, and the CBC reveals pancytopenia (anemia, thrombocytopenia, and neutropenia). A bone marrow biopsy is required for diagnosis and reveals marked hypocellularity. MDS also presents as cytopenias and can have pancytopenia, but the marrow typically demonstrates hypercellularity and dysplastic features. The presence of macrocytosis, ringed sideroblasts on iron staining of the bone marrow aspirate, dysplasia of hematopoietic elements, or cytogenetic abnormalities (especially monosomy 5 or 7 and trisomy 8) is more suggestive of MDS.

Differential Diagnosis

Adult patients with acquired amegakaryocytic thrombocytopenia (rare) have isolated thrombocytopenia and reduced or absent megakaryocytes in the bone marrow, which along with failure to respond to immunomodulatory regimens typically administered in immune thrombocytopenia (ITP), distinguishes them from patients with ITP.

Treatment

A. Congenital Conditions

Treatment is varied but may include blood product support, blood cell growth factors, androgens and, in some cases, allogeneic hematopoietic stem cell transplantation.

B. Acquired Conditions

Patients with severe aplastic anemia are treated with immunosuppressive therapy or allogeneic hematopoietic stem cell transplantation (see Chapter 13).

Treatment of thrombocytopenia due to MDS, if clinically significant bleeding is present or if the risk of bleeding is high, is limited to chronic transfusion of platelets in most instances (Table 14–3). Additional treatment is discussed in Chapter 13.

2. BONE MARROW INFILTRATION

Replacement of the normal bone marrow elements by leukemic cells, plasma cell myeloma, lymphoma, or nonhematologic tumors or by infections (such as mycobacterial disease or ehrlichiosis) may cause thrombocytopenia; however, abnormalities in other blood cell lines are usually present. These entities are easily diagnosed after examining the bone marrow biopsy and aspirate or determining the infecting organism from an aspirate specimen, and they often lead to a leukoerythroblastic peripheral blood smear (left-shifted myeloid lineage cells, nucleated red blood cells, and teardrop-shaped red blood cells). Treatment of thrombocytopenia is directed at eradication of the underlying infiltrative disorder, but platelet transfusion may be required if clinically significant bleeding is present.

3. CHEMOTHERAPY & IRRADIATION

Chemotherapeutic agents and irradiation may lead to thrombocytopenia by direct toxicity to megakaryocytes, hematopoietic progenitor cells, or both. The severity and duration of chemotherapy-induced depressions in the platelet count are determined by the specific regimen used, although the platelet count typically resolves more slowly following a chemotherapeutic insult than does neutropenia or anemia, especially if multiple cycles of treatment have been given. Until recovery occurs, patients may be supported with transfused platelets if bleeding is present or the risk of bleeding is high (Table 14–3). Initial studies suggest that that platelet growth factors, such as eltrombopag and romiplostim, may help prevent chemotherapy-induced thrombocytopenia and allow patients to receive their full chemotherapy doses on schedule. Additional prospective trials are underway to better define their role and efficacy. Checkpoint inhibitors can also lead to thrombocytopenia that mimics immune thrombocytopenic purpura.

4. NUTRITIONAL DEFICIENCIES

Thrombocytopenia, typically in concert with anemia, may be observed with a deficiency of folate (that may accompany alcoholism) or vitamin B₁₂ (concomitant neurologic findings may be manifest). In addition, thrombocytopenia can occur in very severe iron deficiency, albeit rarely, whereas thrombocytosis is far more common. Replacing the deficient vitamin or mineral results in improvement in the platelet count.

5. CYCLIC THROMBOCYTOPENIA

Cyclic thrombocytopenia is a rare disorder that produces cyclic oscillations of the platelet count, usually with a periodicity of 3–6 weeks. The pathophysiologic mechanism responsible for the condition is unclear. Severe thrombocytopenia and bleeding typically occur at the platelet nadir. Oral contraceptive medications, androgens, azathioprine, and thrombopoietic growth factors have been used successfully in the management of cyclic thrombocytopenia.

1. IMMUNE THROMBOCYTOPENIA

General Considerations

ITP is an autoimmune condition in which pathogenic antibodies bind platelets, accelerating their clearance from the circulation. Growing evidence suggests additional pathophysiologic mechanisms, including a role for T cells. Many patients with ITP also lack appropriate compensatory platelet production, thought, at least in part, to reflect the antibody’s effect on megakaryocytopoiesis and thrombopoiesis. ITP is primary (idiopathic) in most adult patients, although it can be secondary (ie, associated with autoimmune disease, such as systemic lupus erythematosus [SLE]; lymphoproliferative disease, such as lymphoma; medications; and infections caused by hepatitis C virus, HIV, and H pylori). Antiplatelet antibody targets include glycoproteins IIb/IIIa and Ib/IX on the platelet membrane, although antibodies are demonstrable in only two-thirds of patients; testing for such antibodies is not standard of care given the significant false-positive and false-negative results. In addition to production of antiplatelet antibodies, HIV and hepatitis C virus may lead to thrombocytopenia through additional mechanisms (for instance, by direct suppression of platelet production [HIV] and cirrhosis-related decreased thrombopoietin [TPO] production and secondary splenomegaly [hepatitis C virus]).

Clinical Findings

A. Symptoms and Signs

Mucocutaneous bleeding may be present, depending on the platelet count. Clinically relevant spontaneous bruising, epistaxis, gingival bleeding, or other types of hemorrhage generally do not occur until the platelet count has fallen below 10,000–20,000/mcL (10–20 × 10⁹/L). Individuals with secondary ITP (such as due to autoimmune disease, HIV or HCV infection, SLE, or lymphoproliferative malignancy) may have additional disease-specific findings.

B. Laboratory Findings

Typically, patients have isolated thrombocytopenia. If substantial bleeding has occurred, anemia may also be present. Hepatitis B and C viruses and HIV infections should be excluded by serologic testing. H pylori infections can sometimes cause isolated thrombocytopenia.

Bone marrow should be examined in patients with unexplained cytopenias in two or more lineages, in patients older than 40 years with isolated thrombocytopenia, or in those who do not respond to primary ITP-specific therapy. A bone marrow biopsy is not necessary in all cases to make an ITP diagnosis in younger patients. Megakaryocyte morphologic abnormalities and hypocellularity or hypercellularity are not characteristic of ITP. ITP patients often have increased numbers of bone marrow megakaryocytes. If there are clinical findings suggestive of a lymphoproliferative malignancy, a CT scan should be performed. In the absence of such findings, otherwise asymptomatic patients younger than 40 years lacking the above infections and with unexplained isolated thrombocytopenia of recent onset may be considered to have ITP.

Treatment

Individuals with platelet counts less than 25,000–30,000/mcL (25–30 × 10⁹/L) or those with significant bleeding should be treated; the remainder may be monitored serially for progression, but that is a patient-specific decision. The mainstay of initial treatment of new-onset primary ITP is a short course of prednisone with or without intravenous immunoglobulin (IVIG) or anti-D (WinRho) (Figure 14–1). A short course of high-dose dexamethasone is also an option for initial treatment. Response to corticosteroids is generally seen within 3–7 days of initiating treatment, with responses to IVIG typically seen in 24–36 hours. Platelet transfusions may be given concomitantly if active bleeding is present. Adding the anti-B cell monoclonal antibody rituximab to corticosteroids as first-line treatment may improve the initial response rate, but it is associated with increased toxicity and is not regarded as standard first-line therapy in most centers.

Although over two-thirds of patients with ITP respond to initial treatment with oral corticosteroids, most relapse following reduction of the corticosteroid dose. Patients with a persistent platelet count less than 30,000/mcL (30 × 10⁹/L) or clinically significant bleeding are appropriate candidates for second-line treatments (Figure 14–1). These treatments are chosen empirically, bearing in mind potential toxicities and patient preference. IVIG or anti-D (WinRho) temporarily increases platelet counts (duration, up to 3 weeks, rarely longer). Serial IVIG or anti-D treatment is an option for some adult patients while alternate safe treatment is pursued. Rituximab leads to clinical responses in about 50% of adults with corticosteroid-refractory chronic ITP, which decreases to about 20% at 5 years. The TPO-mimetics romiplostim (administered subcutaneously weekly), eltrombopag (taken orally daily), and avatrombopag (taken orally daily) are approved for use in adult patients with chronic ITP who have not responded durably to corticosteroids, IVIG, or splenectomy. Romiplostim, eltrombopag, or avatrombopag can be taken indefinitely to maintain the platelet response and can be used as second-line therapy. The Syk inhibitor fostamatinib represents a novel mechanism of action to treat ITP patients who do not respond to corticosteroids, TPO-mimetics, or rituximab. Splenectomy has a durable response rate of over 50% and may be considered for cases of severe ITP that fail to respond durably to initial treatment or are refractory to second-line agents; patients should receive pneumococcal, Haemophilus influenzae type b, and meningococcal vaccination at least 2 weeks before therapeutic splenectomy. If available, laparoscopic splenectomy is preferred. Additional treatments for ITP are found in Figure 14–1.

Management goals for pregnancy-associated ITP are a platelet count of 10,000–30,000/mcL (10–30 × 10⁹/L) in the first trimester, greater than or equal to 30,000/mcL (30 × 10⁹/L) during the second or third trimester, and greater than 50,000/mcL (50 × 10⁹/L) prior to cesarean section or vaginal delivery. Moderate-dose oral prednisone or intermittent IVIG infusions are standard treatment options. Splenectomy is reserved for failure to respond to these therapies and may be performed in the first or second trimester. Management requires close interaction between obstetrician and hematologist. TPO-mimetics are not approved for use during pregnancy.

For thrombocytopenia associated with HIV or hepatitis C virus, effective treatment of either infection leads to an amelioration of thrombocytopenia in most cases; refractory thrombocytopenia may require the use of IVIG, splenectomy, TPO-mimetic, or anti-CD20 therapy. Occasionally, ITP treatment response is impaired due to H pylori infection, which should be ruled out in the appropriate situation.

When to Refer

All patients with ITP need to be managed by a hematologist because of the complexity of the decision making.

When to Admit

Patients with major hemorrhage or very severe thrombocytopenia associated with bleeding should be admitted and monitored in-hospital until the platelet count has consistently risen to more than 20,000–30,000/mcL (20–30 × 10⁹/L) and hemodynamic stability has been achieved.

2. THROMBOTIC MICROANGIOPATHY

General Considerations

The TMAs include, but are not limited to, thrombotic thrombocytopenic purpura (TTP) and HUS. These disorders are characterized by thrombocytopenia due to the incorporation of platelets into fibrin thrombi in the microvasculature, and microangiopathic hemolytic anemia, which results from shearing of erythrocytes in fibrin networks in the microcirculation.

In idiopathic TTP, autoantibodies against ADAMTS-13 (a disintegrin and metalloproteinase with thrombospondin type 1 repeat, member 13), also known as the von Willebrand factor (vWF) cleaving protease (vWFCP), lead to accumulation of ultra-large vWF multimers. The ultra-large multimers bridge and aggregate platelets in the absence of hemostatic triggers, which in turn leads to the vessel obstruction and various organ dysfunctions seen in TTP. In some cases of pregnancy-associated TMA, an antibody to ADAMTS-13 is present. In contrast, the activity of the ADAMTS-13 in congenital TTP is decreased due to a mutation in the gene encoding the molecule. Classic HUS, called Shiga toxin–mediated HUS, is thought to be secondary to toxin-mediated endothelial damage and is often contracted through the ingestion of undercooked ground beef contaminated with Escherichia coli (especially types O157:H7 or O145).

Complement-mediated HUS (formerly called atypical HUS) is not related to Shiga toxin. Patients with complement-mediated HUS often have genetic defects in proteins that regulate complement activity. Mutations in complement genes (such as factor H, a complement regulator) account for the uncontrolled activation of complement that characterizes the condition. Damage to endothelial cells—such as the damage that occurs in endemic HUS due to presence of toxins from E coli (especially type O157:H7 or O145) or in the setting of cancer, hematopoietic stem cell transplantation, or HIV infection—may also lead to TMA. Certain drugs (eg, cyclosporine, quinine, ticlopidine, clopidogrel, mitomycin C, and bleomycin) are associated with the development of TMA, possibly by promoting injury to endothelial cells, although inhibitory antibodies to ADAMTS-13 have been demonstrated in some cases.

Clinical Findings

A. Symptoms and Signs

Microangiopathic hemolytic anemia and thrombocytopenia are presenting signs in all patients with TTP and most patients with HUS; in a subset of patients with HUS, the platelet count remains in the normal range. Only about 25% of patients with TMA manifest all components of the original pentad of findings (microangiopathic hemolytic anemia, thrombocytopenia, fever, kidney disease, and neurologic abnormalities) (Table 14–4). Most patients (especially children) with HUS have a recent or current diarrheal illness, often bloody. Neurologic manifestations, including headache, somnolence, delirium, seizures, paresis, and coma, may result from deposition of microthrombi in the cerebral vasculature.

B. Laboratory Findings

Laboratory features of TMA include those associated with microangiopathic hemolytic anemia (anemia, elevated lactate dehydrogenase [LD], elevated indirect bilirubin, decreased haptoglobin, schistocytes on the blood smear, elevated reticulocyte count, and a negative direct antiglobulin test); thrombocytopenia; elevated creatinine; positive stool culture for E coli O157:H7 or stool assays for Shiga toxin; reductions in ADAMTS-13 activity with the presence (acquired TTP) or absence (inherited TTP) of ADAMTS-13 inhibitor; and mutations of genes encoding complement proteins (complement-mediated HUS; specialized laboratory assessment). Routine coagulation studies (prothrombin time [PT], activated partial thromboplastin time [aPTT], fibrinogen) are within the normal range in most patients with TTP or HUS.

Treatment

With the exception of children or adults with endemic diarrhea-associated HUS, who generally recover with supportive care only, plasma exchange must be initiated as soon as the diagnosis of TMA is suspected and in all cases of TTP. Immediate administration of plasma exchange is essential in most cases of TTP because the mortality rate without treatment is over 95%. Plasma exchange usually is administered once daily until the platelet count and LD have returned to normal for at least 2 days, after which the frequency of treatments may be tapered slowly while the platelet count and LD are monitored for relapse. In cases of insufficient response to once-daily plasma exchange, twice-daily treatments can be considered. Fresh frozen plasma (FFP) may be administered if immediate access to plasma exchange is not available or in cases of familial TMA. Platelet transfusions are contraindicated in the treatment of TMA due to reports of worsening TMA, possibly due to propagation of platelet-rich microthrombi. In cases of documented life-threatening bleeding, however, platelet transfusions may be given slowly and preferably after plasma exchange is underway. Red blood cell transfusions may be administered in cases of clinically significant anemia. Hemodialysis should be considered for patients with significant kidney injury. Caplacizumab, a bi-specific antibody that targets the A1 domain of vWF and prevents vWF interaction with the platelet glycoprotein Ib-IX-V receptor, can reduce the time to platelet count normalization and 30-day mortality. The role of caplacizumab in the treatment of TTP remains controversial given its high cost and limited benefit, despite its inclusion in 2020 guidelines.

In cases of TTP relapse following initial treatment, plasma exchange should be reinstituted. If ineffective, or in cases of primary refractoriness, second-line treatments including rituximab (which has shown efficacy when administered preemptively in selected cases of relapsing TTP), corticosteroids, IVIG, vincristine, cyclophosphamide, and splenectomy should be used. Idiopathic TTP is a relapsing autoimmune disorder (antibody inhibitor to ADAMTS-13) for most patients; careful monitoring of the ADAMTS-13 activity and inhibitor status and use of rituximab can prevent dangerous relapses.

Cases of complement-mediated HUS may respond to plasma infusion initially; however, once this diagnosis is strongly suspected, apheresis is typically stopped and serial infusions of the anti-complement C5 antibody eculizumab are given, which have produced sustained remissions in some patients. If irreversible kidney injury has occurred, hemodialysis or kidney transplantation may be necessary.

When to Refer

Consultation by a hematologist or transfusion medicine specialist familiar with plasma exchange is required at the time of presentation. Patients with TMA and TTP require ongoing care by a hematologist.

When to Admit

All patients with newly suspected or diagnosed TMA should be hospitalized immediately.

3. HEPARIN-INDUCED THROMBOCYTOPENIA

General Considerations

Heparin-induced thrombocytopenia (HIT) is an acquired disorder that affects approximately 3% of patients exposed to unfractionated heparin and ~0.6% of patients exposed to low-molecular-weight heparin (LMWH). The condition results from formation of IgG antibodies to heparin-platelet factor 4 (PF4) complexes; the antibody:heparin-PF4 complex binds to and activates platelets independent of physiologic hemostasis, which leads to thrombocytopenia and thromboses. von Willebrand factor has been postulated to play a role in the thrombotic events that take place long after heparin is cleared from the patient’s system.

Clinical Findings

A. Symptoms and Signs

Patients are often asymptomatic, and due to the prothrombotic nature of HIT, bleeding usually does not occur. Thrombosis (at any venous or arterial site), however, may be detected in up to 50% of patients, up to 30 days post diagnosis. If thrombosis has not already been detected, the use of duplex Doppler ultrasound of the lower extremities should be considered to rule out subclinical deep venous thrombosis (DVT).

B. Laboratory Findings

A presumptive diagnosis of HIT is made when new-onset thrombocytopenia is detected in a patient (typically a hospitalized patient) within 5–14 days of initial exposure to heparin; other presentations (eg, rapid-onset HIT) are less common and reflect recent prior heparin exposure. A decline of 50% or more from the baseline platelet count is typical. The 4T score (http://www.qxmd.com/calculate-online/hematology/hit-heparin-induced-thrombocytopenia-probability) is a clinical prediction rule for assessing pretest probability for HIT. Low 4T scores have been shown to be more predictive of excluding HIT than are intermediate or high scores of predicting its presence. Once HIT is clinically suspected, the clinician must establish the diagnosis by performing a screening PF4-heparin antibody enzyme-linked immunosorbent assay (ELISA). If the PF4-heparin antibody ELISA is positive, the diagnosis must be confirmed using a functional assay (such as serotonin release assay). The magnitude of a positive ELISA result correlates with the clinical probability of HIT, but even high ELISA optical density values may be falsely positive. The confirmatory functional assay is essential.

Treatment

Treatment should be initiated as soon as the diagnosis of HIT is suspected, before results of laboratory testing are available.

Management of HIT (Table 14–5) involves the immediate discontinuation of all forms of heparin. Despite thrombocytopenia, platelet transfusions are rarely necessary and should be avoided. Due to the substantial frequency of thrombosis among HIT patients, an alternative anticoagulant should be administered immediately while awaiting confirmatory testing. A direct thrombin inhibitor (DTI), such as argatroban or bivalirudin, is preferred in critical illness because of the shorter duration of action. The use of the subcutaneous indirect anti-Xa inhibitor fondaparinux for initial treatment of HIT is a reasonable option in clinically stable patients. For confirmed HIT, the DTI should be continued until the platelet count has recovered to at least 100,000/mcL (100 × 10⁹/L), at which point treatment with a vitamin K antagonist (warfarin) may be initiated. The DTI should be continued until therapeutic anticoagulation with the vitamin K antagonist warfarin has been achieved (ie, international normalized ratio [INR] of 2.0–3.0); the infusion of argatroban must be temporarily discontinued before the INR is obtained so that it reflects the anticoagulant effect of warfarin alone. There is a growing acceptance for using oral anti-Xa agents instead of vitamin K antagonists in selected patients. In all patients with HIT, some form of anticoagulation (warfarin or other) should be continued for at least 30 days, due to a persistent risk of thrombosis even after the platelet count has recovered, but in patients in whom thrombosis has been documented, anticoagulation should continue for 3–6 months.

Subsequent exposure to heparin should be avoided in all patients with a prior history of HIT, if possible. If its use is regarded as necessary for a procedure, it should be withheld until PF4-heparin antibodies are no longer detectable by ELISA (usually as of 100 days following an episode of HIT), and exposure should be limited to the shortest time period possible. A common example is a cardiac catheterization. The heparin is gone before the antibody returns, so HIT is avoided.

When to Refer

Due to the tremendous thrombotic potential of the disorder and the complexity of use of the DTI, all patients with HIT should be evaluated by a hematologist.

When to Admit

Most patients with HIT are hospitalized at the time of detection of thrombocytopenia. Admission is a clinical decision for an outpatient in whom HIT is suspected and who is a candidate for subcutaneous fondaparinux. Other outpatients should be admitted because the DTIs must be administered by continuous intravenous infusion. Regardless, a hematologist needs to be involved as soon as the diagnosis is suspected or treatment is indicated.

4. DISSEMINATED INTRAVASCULAR COAGULATION

General Considerations

Disseminated intravascular coagulation (DIC) is caused by uncontrolled local or systemic activation of coagulation, which leads to depletion of coagulation factors and fibrinogen, and often results in thrombocytopenia as platelets are activated and consumed.

Numerous disorders are associated with DIC, including sepsis (in which coagulation is activated by presence of lipopolysaccharide), cancer, trauma, burns, and pregnancy-associated complications (in which tissue factor is released). Aortic aneurysm and cavernous hemangiomas may promote localized intravascular coagulation, and snake bites may result in DIC due to the introduction of exogenous toxins.

Clinical Findings

A. Symptoms and Signs

Bleeding in DIC usually occurs at multiple sites, such as intravenous catheters or incisions, and may be widespread (purpura fulminans). Malignancy-related DIC may manifest principally as thrombosis (Trousseau syndrome).

B. Laboratory Findings

In early DIC, the platelet count and fibrinogen levels often remain within the normal range, albeit reduced from baseline levels. There is progressive thrombocytopenia (rarely severe), prolongation of the PT, decrease in fibrinogen levels, and eventually elevation in the aPTT. D-dimer levels typically are elevated due to the activation of coagulation and diffuse cross-linking of fibrin followed by fibrinolysis. Schistocytes on the blood smear, due to shearing of red cells through the microvasculature, are present in 10–20% of patients. Laboratory abnormalities in the HELLP syndrome (hemolysis, elevated liver enzymes, low platelets), a severe form of DIC with a particularly high mortality rate that occurs in peripartum women, include elevated liver transaminases and kidney injury due to gross hemoglobinuria and pigment nephropathy. Malignancy-related DIC may feature normal platelet counts and coagulation studies, but clinicians often see a dropping platelet count and fibrinogen, with a rising INR, highlighting the importance of serial laboratory values to help make the diagnosis.

Treatment

The underlying causative disorder must be treated (eg, antimicrobials, chemotherapy, surgery, or delivery of conceptus). If clinically significant bleeding is present, hemostasis must be achieved (Table 14–6).

Blood products are administered if clinically significant hemorrhage has occurred or is thought likely to occur without intervention based on progressively increasing PT and PTT and decreasing fibrinogen and platelets levels (Table 14–6). The goal of platelet therapy for most cases is greater than 20,000/mcL (20 × 10⁹/L) or greater than 50,000/mcL (50 × 10⁹/L) for serious bleeding, such as intracranial bleeding. FFP is typically given only to patients with a prolonged aPTT and PT and significant bleeding. Cryoprecipitate may be given for bleeding or for fibrinogen levels less than 80–100 mg/dL. The clinician should correct the fibrinogen level with cryoprecipitate prior to giving FFP for prolonged PT and aPTT to see if the fibrinogen replacement alone corrects the PT and aPTT. The PT, aPTT, fibrinogen, and platelet count should be monitored at least every 6–8 hours in acutely ill patients with DIC.

In some cases of refractory bleeding despite replacement of blood products, administration of low doses of heparin can be considered. The clinician must remember that DIC is primarily a disorder of excessive clotting with secondary fibrinolysis, and that heparin can interfere with thrombin generation, which leads to less consumption of coagulation proteins and platelets. An infusion of 5 units/kg/h (no bolus) may be used with appropriate clinical judgement, uptitrated as clinically indicated. Heparin, however, can be contraindicated if the platelet count cannot be maintained above 50,000/mcL (50 × 10⁹/L) and in cases of central nervous system hemorrhage, gastrointestinal bleeding, placental abruption, and any other condition that is likely to require imminent surgery. Fibrinolysis inhibitors may be considered in select DIC patients with bleeding, but this can promote dangerous clotting and should be undertaken with great caution and only in consultation with a hematologist.

The treatment of HELLP syndrome must include evacuation of the uterus (eg, delivery of a term or near-term infant or removal of retained placental or fetal fragments). Patients with Trousseau syndrome require treatment of the underlying malignancy and administration of unfractionated heparin or subcutaneous therapeutic-dose LMWH as treatment of thrombosis, since warfarin typically is ineffective at secondary prevention of thromboembolism in the disorder. Typically, the heparin or LMWH treatment will gradually return the fibrinogen, PT (INR), aPTT, and platelet count back to normal, but it can take many days. Oral anti-Xa agents or oral DTIs can be considered once stabilized with parenteral heparin or LMWH, but extended LMWH is often used in this setting.

Immediate initiation of medical treatment (usually within 24 hours of diagnosis) is required for patients with acute promyelocytic leukemia (APL)–associated DIC, along with administration of blood products as clinically indicated.

When to Refer

• Diffuse bleeding unresponsive to administration of blood products should be evaluated by a hematologist.

• All patients with DIC should be cared for by a hematologist prior to starting treatment with heparin or LMWH.

When to Admit

Most patients with DIC are hospitalized when DIC is detected.

1. DRUG-INDUCED THROMBOCYTOPENIA

Drug-induced thrombocytopenia is often immune-mediated but can also be due to marrow suppression. Table 14–7 lists medications associated with thrombocytopenia. The typical presentation of drug-induced, antibody-mediated thrombocytopenia is severe thrombocytopenia and mucocutaneous bleeding 5–14 days after exposure to a new drug, although a range of presentations is possible. Discontinuation of the offending agent leads to resolution of thrombocytopenia within 3–7 days in most cases, but recovery kinetics depend on rate of drug clearance, which can be affected by liver and kidney function. Patients with severe thrombocytopenia should be given platelet transfusions with or without IVIG. The University of Oklahoma Health Sciences center maintains a useful website for drug-induced thrombocytopenia (https://www.ouhsc.edu/platelets/).

2. POSTTRANSFUSION PURPURA

Posttransfusion purpura (PTP) is a rare disorder of sudden-onset thrombocytopenia that occurs within 1 week after transfusion of red cells, platelets, or plasma. Antibodies against the human platelet antigen PL^A1 are detected in most individuals with PTP. Patients with PTP often are either multiparous women or persons who have received transfusions previously. Severe thrombocytopenia and bleeding are typical. Initial treatment consists of administration of IVIG (1 g/kg/day for 2 days), which should be administered as soon as the diagnosis is suspected. Platelets are not indicated unless severe bleeding is present, but if they are to be administered, HLA-matched PL^A1-negative platelets are preferred. A second course or IVIG, plasma exchange, corticosteroids, TPO-mimetics, or splenectomy may be required in case of refractoriness. PL^A1-negative or washed blood products are preferred for subsequent transfusions, but data supporting various treatment options are limited.

3. VON WILLEBRAND DISEASE TYPE 2B

von Willebrand disease (vWD) type 2B leads to chronic, characteristically mild to moderate thrombocytopenia via an abnormal vWF molecule that binds platelets with increased affinity, resulting in aggregation and clearance.

4. PLATELET SEQUESTRATION

One-third of the platelet mass is typically sequestered in the spleen. Splenomegaly, due to a variety of conditions (eTable 14–1), may lead to thrombocytopenia of variable severity. When possible, treatment of the underlying disorder should be pursued, but splenectomy, splenic embolization, or splenic irradiation may be considered in selected cases.

5. PREGNANCY

Gestational thrombocytopenia is thought to result from progressive expansion of the blood volume that typically occurs during pregnancy, leading to hemodilution. Cytopenias result even though blood cell production is normal or increased. Platelet counts less than 100,000/mcL (100 × 10⁹/L), however, are observed in less than 10% of pregnant women in the third trimester; decreases to less than 70,000/mcL (70 × 10⁹/L) should prompt consideration of pregnancy-related ITP as well as preeclampsia or a pregnancy-related thrombotic microangiopathy.

6. INFECTION OR SEPSIS

Both immune- and platelet production–mediated defects are possible, and there may be significant overlap with concomitant DIC. Regardless, the platelet count typically improves with effective antimicrobial treatment or after the infection has resolved. Hemophagocytosis may occur in some critically ill patients; a defect in immunomodulation may lead to bone marrow macrophages (histiocytes) engulfing cellular components of the marrow. The phenomenon typically resolves with resolution of the infection, but with certain infections (Epstein-Barr virus) immunosuppression may be required. Hemophagocytosis also may occur with malignancy, in which case the disorder is usually unresponsive to treatment with immunosuppression and requires treatment of the malignancy. Sepsis-related thrombocytopenia may be at least in part due to increased hepatic clearance of platelets caused by loss of asialoglycoprotein moieties on the platelet surface.

7. PSEUDOTHROMBOCYTOPENIA

Pseudothrombocytopenia results from ethylenediaminetetraacetic acid (EDTA) anticoagulant-induced platelet clumping; the phenomenon typically disappears when blood is collected in a tube containing citrate anticoagulant. Pseudothrombocytopenia diagnosis requires review of the peripheral blood smear and is not associated with bleeding.

GENERAL CONSIDERATIONS

Heritable qualitative platelet disorders are far less common than acquired platelet function disorders and lead to variably severe bleeding, often beginning in childhood. Occasionally, however, disorders of platelet function may go undetected until later in life when excessive bleeding occurs following a sufficient hemostatic challenge. Thus, the true incidence of hereditary qualitative platelet disorders is unknown.

Bernard-Soulier syndrome (BSS) is a rare, autosomal recessive bleeding disorder due to reduced or abnormal platelet membrane expression of glycoprotein Ib/IX (vWF receptor).

Glanzmann thrombasthenia results from an abnormality in the platelet glycoprotein IIb/IIIa receptor on the platelet membrane. Glycoprotein IIb/IIIa is the fibrinogen receptor critical for linking platelets during initial platelet aggregation/platelet plug formation. Inheritance is autosomal recessive.

Under normal circumstances, activated platelets release the contents of platelet granules to reinforce the aggregatory response. Storage pool disease includes a spectrum of defects in release of alpha or dense (delta) platelet granules, or both (alpha-delta storage pool disease).

CLINICAL FINDINGS

A. Symptoms and Signs

Bleeding due to defective platelets is usually mucocutaneous, but it is not limited to mucocutaneous surfaces. The onset of bleeding with Glanzmann thrombasthenia is usually in infancy or childhood, but some forms are milder and present later in life. The degree of deficiency in IIb/IIIa may not correlate well with bleeding symptoms. Patients with storage pool disease are affected by variable bleeding, ranging from mild and trauma-related to spontaneous.

B. Laboratory Findings

In Bernard-Soulier syndrome, there are abnormally large platelets (approaching the size of red cells), moderate thrombocytopenia, and a prolonged bleeding time. Platelet aggregation studies show a marked defect in response to ristocetin, whereas aggregation in response to other agonists is normal; the addition of normal platelets corrects the abnormal aggregation. The diagnosis can be confirmed by platelet flow cytometry.

In Glanzmann thrombasthenia, platelet aggregation studies show marked impairment of aggregation in response to stimulation with various agonists, which reflects the critical role of the fibrinogen receptor in platelet plug formation.

Storage pool disease describes defects in the number, content, or function of platelet alpha or dense granules, or both. The gray platelet syndrome comprises abnormalities of platelet alpha granules, thrombocytopenia, and marrow fibrosis. The blood smear shows agranular platelets, and the diagnosis is confirmed with electron microscopy.

Albinism-associated storage pool disease involves defective dense granules in disorders of oculocutaneous albinism, such as the Hermansky-Pudlak and Chediak-Higashi syndromes. Electron microscopy confirms the diagnosis.

Non–albinism-associated storage pool disease results from quantitative or qualitative defects in dense granules and is seen in Ehlers-Danlos and Wiskott-Aldrich syndromes, among others.

The Quebec platelet disorder comprises mild thrombocytopenia, an abnormal platelet factor V molecule, and a prolonged bleeding time. Patients typically experience moderate bleeding. Interestingly, platelet transfusion does not ameliorate the bleeding. Platelet aggregation studies characteristically show platelet dissociation following an initial aggregatory response, and electron microscopy confirms the diagnosis.

TREATMENT

The mainstay of treatment (including periprocedural prophylaxis) is transfusion of normal platelets, although desmopressin acetate (DDAVP), antifibrinolytic agents, and recombinant human activated factor VII each have a role in selected clinical situations.

Platelet dysfunction is more commonly acquired than inherited; the widespread use of platelet-altering medications accounts for most of the cases of qualitative defects (eTable 14–2). In cases where platelet function is irreversibly altered, platelet inhibition typically recovers within 7–9 days following discontinuation of the drug, which is the time it takes to replace all of the impaired platelets with newly produced platelets. In cases where platelet function is non-irreversibly affected, platelet inhibition recovers with clearance of the drug from the system. Transfusion of platelets may be required if clinically significant bleeding is present.

Laboratory manifestations of aspirin toxicity include a prolonged epinephrine cartridge closure time in the platelet function analyzer (PFA)-100 system, or a decreased aggregation to low-dose collagen and thrombin (and preserved aggregation in response to high-dose collagen and thrombin) on platelet aggregation studies. Abnormal platelet aggregation studies may be observed in the setting of qualitative platelet defects induced by other conditions, but specific defects vary considerably.

1. HEMOPHILIA A & B

General Considerations

The frequency of hemophilia A is ~1 per 5000 live male births, whereas hemophilia B occurs in ~1 in 25,000 live male births. Inheritance is X-linked recessive, leading to affected males and carrier (affected) females with variable bleeding tendencies. Daughters of all affected males are obligate carriers. There is no race predilection. Factor activity testing is indicated for male infants with a hemophilic maternal pedigree who are asymptomatic or who experience excessive bleeding, for all daughters of affected males (100% chance of being affected) and carrier mothers (50% chance of being affected), and for otherwise asymptomatic adolescents or adults who experience unexpected excessive bleeding with trauma or invasive procedures.

Inhibitors to factor VIII will develop in approximately 20–25% of patients with severe hemophilia A; inhibitors to factor IX will develop in less than 5% of patients with severe hemophilia B. The risk of development of an inhibitor to factor VIII or factor IX is highest in patients with severe hemophilia who have a sibling in whom inhibitor formation occurred; the characteristics of initiation of administration of factor replacement in childhood, as well as mutation type, also may be influential. Recent data suggest that recombinant products may have an increased risk of inhibitor formation than products made from pooled plasma.

A substantial proportion of older patients with hemophilia acquired infection with HIV or HCV or both in the 1980s due to exposure to contaminated factor concentrates and blood products.

Clinical Findings

A. Symptoms and Signs

Severe hemophilia (factor VIII activity less than 1%) presents in infant males or in early childhood with spontaneous bleeding into joints, soft tissues, or other locations. Spontaneous bleeding is much less common in patients with mild hemophilia (factor VIII activity greater than 5%), but bleeding is common with provoked bleeding (eg, surgery, trauma). Intermediate clinical symptoms are seen in patients with moderate hemophilia (factor VIII activity 1–5%). Female carriers of hemophilia can have a wide range of factor VIII activity and therefore have variable bleeding tendencies.

Significant hemophilic arthropathy is usually avoided in patients who have received long-term prophylaxis with factor concentrate starting in early childhood, whereas destructive joint disease is common in adults who have experienced recurrent hemarthroses. Patients tend to have one or two “target” joints into which they bleed most often.

Inhibitor development to factor VIII or factor IX is characterized by bleeding episodes that are resistant to treatment with clotting factor VIII or IX concentrate, and by new or atypical bleeding.

B. Laboratory Findings

Hemophilia A or B is diagnosed by an isolated reproducibly low factor VIII or factor IX activity level, in the absence of other conditions. If the aPTT is prolonged, it typically corrects upon mixing with normal plasma. A variety of mutations, including inversions, large and small deletions, insertions, missense mutations, and nonsense mutations may be causative. Depending on the level of residual factor VIII or factor IX activity, and the sensitivity of the thromboplastin used in the aPTT coagulation reaction, the aPTT may or may not be prolonged, although it typically is markedly prolonged in severe hemophilia. Hemophilia is classified according to the level of factor activity in the plasma. Mild hemophilia has greater than 5% factor activity; moderate hemophilia has 1–5% factor activity; and severe hemophilia has less than 1% factor activity. Female carriers may become symptomatic if significant lyonization has occurred favoring the defective factor VIII or factor IX gene, leading to factor VIII or factor IX activity level markedly less than 50%. Typically, a clinical bleeding diathesis occurs once the factor activity is less than 20%, but this appears to be patient-specific, and bleeding can occur in trauma, surgery, and delivery if the factor activity is less than 50%.

In the presence of an inhibitor to factor VIII or factor IX, there is accelerated clearance of and suboptimal or absent rise in measured activity of infused factor, and the aPTT does not correct on mixing. The Bethesda assay measures the potency of the inhibitor.

Treatment

A. Factor VIII or IX Products

Plasma-derived or recombinant factor VIII or IX products are the mainstay of treatment. The standard of care for most individuals with severe hemophilia is primary prophylaxis: by the age of 4 years, most children with severe hemophilia have begun twice- or thrice-weekly infusions of factor to prevent the recurrent joint bleeding that otherwise would characterize the disorder and lead to severe musculoskeletal morbidity. In selected cases of less severe hemophilia, or as an adjunct to prophylaxis in severe hemophilia, treatment with factor products is given peri-procedurally, prior to high-risk activities (such as sports), or as needed for bleeding episodes (Table 14–8). Recombinant factor VIII and factor IX molecules that are bioengineered to have an extended half-life may allow for extended dosing intervals in patients who are treated prophylactically. The decision to switch to a long-acting product is patient specific. The long-acting factor IX products have clear added value in reducing frequency of factor injections often to weekly or less. Long-acting factor VIII products have not achieved a similar degree of extended half-life. Patients with mild hemophilia A may respond to as-needed (on demand) intravenous or intranasal treatment with DDAVP. Antifibrinolytic agents may be useful in cases of mucosal bleeding and are commonly used adjunctively, such as following dental procedures.

B. Factor VIII or IX Inhibitors

Factor inhibitors (antibodies that interfere with activity or half-life) are a major clinical problem for patients with hemophilia. It may be possible to overcome low-titer inhibitors (less than 5 Bethesda units [BU]) by giving larger doses of factor, whereas treatment of bleeding in the presence of a high-titer inhibitor (more than 5 BU) requires infusion of an activated prothrombin complex concentrate (such as FEIBA [factor eight inhibitor bypassing activity]) or recombinant activated factor VII. Recombinant porcine factor VIII is also an option but is reserved for selective circumstances because of its cost. Inhibitor tolerance induction, achieved by giving large doses (50–300 units/kg intravenously of factor VIII daily) for 6–18 months, succeeds in eradicating the inhibitor in 70% of patients with hemophilia A and in 30% of patients with hemophilia B. Patients with hemophilia B who receive inhibitor tolerance induction, however, are at risk for development of nephrotic syndrome and anaphylactic reactions, making eradication of their inhibitors more problematic. Additional immunomodulation may allow for eradication in selected inhibitor tolerance induction–refractory patients. Emicizumab is a novel bi-specific antibody that brings activated factor IX and factor X together, effectively replacing the cofactor function of factor VIII in the clotting cascade, providing a major therapeutic advance for patients with inhibitors. Emicizumab has also been demonstrated to be an effective option for patients without inhibitors.

C. Gene Therapy

Gene therapy clinical trials for hemophilia A and B have shown great promise for patients with severe hemophilia A and B. For most patients, gene therapy has eliminated spontaneous bleeding as well as the need for factor replacement. While phase III clinical trials have been restricted to patients 18 years of age and older, the results look extremely promising. It is hoped that this potentially life-changing therapy will become an approved treatment outside of clinical trials in 2022.

D. Antiretroviral Therapy

Antiretroviral treatment should be administered to hemophilia patients with HIV infection. Patients with hepatitis C infection should be referred for treatment to eradicate the virus.

E. NSAIDs

The cyclooxygenase (COX)-2 selective nonsteroidal anti-inflammatory drug, celecoxib, may be used to treat arthritis symptoms; generally, other NSAIDs and aspirin should be avoided due to the increased risk of bleeding from inhibition of platelet function. Oral opioid medications are commonly used to control pain, including joint pain and surgical pain from the often-needed total joint replacements.

When to Refer

All patients with hemophilia should be seen regularly in a comprehensive hemophilia treatment center.

When to Admit

• Major invasive procedures because of the need for serial infusions of clotting factor concentrate.

• Bleeding that is unresponsive to outpatient treatment.

2. VON WILLEBRAND DISEASE

General Considerations

vWF is an unusually large multimeric glycoprotein that binds to subendothelial collagen and its platelet receptor, glycoprotein Ib, bridging platelets to the subendothelial matrix at the site of vascular injury and contributing to linking them together in the platelet plug. vWF also has a binding site for factor VIII, prolonging factor VIII half-life in the circulation.

Between 75% and 80% of patients with vWD have type 1, a quantitative abnormality of the vWF molecule that usually does not feature an identifiable causal mutation in the vWF gene.

Type 2 vWD is seen in 15–20% of patients with vWD. In type 2A or 2B vWD, a qualitative defect in the vWF molecule is causative. Type 2N and 2M vWD are due to defects in vWF that decrease binding to factor VIII or to platelets, respectively. Importantly, type 2N vWD can clinically resemble hemophilia A because factor VIII activity levels are decreased, and vWF activity and antigen (Ag) are normal. Type 2M vWD features a normal multimer pattern. Type 3 vWD is rare, and like type 1, is a quantitative defect, with mutational homozygosity or compound heterozygosity yielding very low levels of vWF and severe bleeding in infancy or childhood. Due to its factor VIII carrier function, a severely low vWF level leads to low factor VIII activity and prolonged aPTT.

Clinical Findings

A. Symptoms and Signs

Patients with type 1 vWD usually have mild or moderate platelet-type bleeding (mucocutaneous) that may be evident in childhood. Heavier bleeding may occur with menses, surgery, or delivery. Patients with type 2 vWD usually have moderate to severe bleeding that presents in childhood or adolescence. Patient with type 3 vWD demonstrate a severe bleeding phenotype that typically manifests in childhood or infancy.

B. Laboratory Findings

In type 1 vWD, the vWF activity (ristocetin co-factor assay) and the vWF Ag are mildly depressed, whereas the vWF multimer pattern is normal (Table 14–9). Laboratory testing of type 2A or 2B vWD typically shows a ratio of vWF Ag:vWF activity of approximately 2:1 and a multimer pattern that lacks the highest molecular weight multimers. Thrombocytopenia is common in type 2B vWD due to a gain-of-function mutation of the vWF molecule, which leads to increased vWF binding to its receptor on platelets, resulting in platelet clearance; a ristocetin-induced platelet aggregation (RIPA) study shows an increase in platelet aggregation in response to low concentrations of ristocetin. Except in the more severe forms of vWD that feature a significantly decreased factor VIII activity, aPTT is most commonly normal in patients with vWD. The PT is not affected by vWD.

Treatment

The treatment of vWD is outlined in Table 14–8. DDAVP is useful in the treatment of mild bleeding in most cases of type 1 and some cases of type 2 vWD. DDAVP causes release of vWF and factor VIII from storage sites (endothelial cells), leading to a two- to sevenfold increase in vWF and factor VIII. A therapeutic DDAVP trial to document sufficient rise in vWF level is critical prior to relying on DDAVP as a treatment option. Due to tachyphylaxis and the risk of significant hyponatremia secondary to fluid retention, DDAVP treatment is limited to one dose per 24 hours and no more than three doses over 5 days. vWF-containing factor VIII concentrates or recombinant VWF products are used in all other clinical scenarios, and when bleeding is not controlled with DDAVP. Cryoprecipitate is no longer used as a source of vWD in clinical practice. Antifibrinolytic agents (eg, aminocaproic acid or tranexamic acid) may be used adjunctively for mucosal bleeding or procedures. Pregnant patients with type 1 vWD usually do not require treatment at the time of delivery because of the natural physiologic increase in vWF levels (up to threefold that of baseline) that are observed by parturition. However, levels need to be confirmed in late pregnancy, and if they are low or if excessive bleeding is encountered, vWF products may be given. Moreover, patients are at risk for significant bleeding 1–2 weeks postpartum when vWF levels fall secondary to the fall in estrogen levels and related return to baseline vWF levels.

3. FACTOR XI DEFICIENCY

Factor XI deficiency (also called hemophilia C) is inherited in an autosomal recessive manner(in contrast to X-linked inheritance for hemophilia A and B), leading to heterozygous, compound heterozygous, or homozygous defects. It is most prevalent among individuals of Ashkenazi Jewish descent, yet it is in the differential diagnosis of anyone with an unexplained prolonged aPTT. Levels of factor XI, while variably reduced, do not correlate well with bleeding symptoms. Mild bleeding is most common, and diagnosis is often made after unexpected, excessive bleeding following surgery or trauma. Importantly, factor XI deficiency that can lead to provoked excessive bleeding does not always prolong the aPTT. FFP is the mainstay of treatment when plasma-derived factor XI concentrate is not available. Administration of adjunctive aminocaproic acid or tranexamic acid is regarded as mandatory for procedures or bleeding episodes involving the mucosa (Table 14–8).

4. LESS COMMON HERITABLE DISORDERS OF COAGULATION

Congenital deficiencies of clotting factors II, V, VII, and X are rare and typically are inherited in an autosomal recessive pattern. A prolongation in the PT (and aPTT for factor X, factor V, and factor II deficiency) that corrects upon mixing with normal plasma is typical. Definitive diagnosis requires testing for specific factor activity. The treatment of factor II deficiency is with a prothrombin complex concentrate; factor V deficiency is treated with infusions of FFP or platelets (which contain factor V in alpha granules); factor VII deficiency is treated with recombinant human activated factor VII. Factor X deficiency is treated with an FDA-approved plasma-derived factor X product (Coagadex).

Deficiency of factor XIII characteristically leads to delayed bleeding that occurs hours to days after a hemostatic challenge (such as surgery or trauma). The condition is usually life-long, and spontaneous intracranial hemorrhages as well as recurrent pregnancy loss appear to occur with increased frequency in these patients compared with other congenital deficiencies. Cryoprecipitate can be used to provide factor XIII, but if available, plasma-derived factor XIII concentrate (Corifact) is preferred to treat bleeding or for surgical prophylaxis. Regular prophylactic factor XIII replacement is indicated for patients with severe factor XIII deficiency. Factor XIII has an A and B subunit. Recombinant factor XIII A-subunit (Tretten) is an option for patients deficient in the factor XIII A subunit. Factor XIII deficiency does not cause a prolongation of the PT or aPTT.

Alpha-2-antiplasmin deficiency is a rare disorder that leads to accelerated fibrinolysis via insufficient inhibition of plasmin. Heterozygosity for the condition usually produces a mild bleeding tendency, while bleeding symptoms in homozygotes may be severe. The diagnosis is made by a documented antiplasmin level below the reference range; the aPTT and PT are normal. In some cases, treatment of bleeding or surgical prophylaxis is with aminocaproic acid. A congenital deficiency of plasminogen activator I (PAI-1) is extremely rare and can lead to mild to moderate bleeding; testing for the disorder can be difficult due to the extremely low extension of the normal reference range for PAI-1.

Congenital afibrinogenemia is exceedingly rare and produces mild to severe bleeding; the frequency of first trimester miscarriage is increased among women with the disorder. The PT is more typically prolonged than the aPTT, and a functional fibrinogen assay shows reduced activity. Treatment is with a fibrinogen concentrate (RiaSTAP) (preferred and now FDA approved), cryoprecipitate, or FFP and is aimed at increasing the plasma fibrinogen concentration to greater than 80 mg/dL.

Congenital deficiencies of factor XII, prekallikrein, high molecular-weight kininogen may lead to a prolonged aPTT that corrects with extended incubation but do not lead to bleeding.

1. ACQUIRED ANTIBODIES TO FACTOR II

Patients with antiphospholipid antibodies occasionally have antibody specificity to coagulation factor II (prothrombin) that accelerates factor II clearances and can lead to severe hypoprothrombinemia and bleeding. Mixing studies may or may not reveal presence of an inhibitor, as the antibody typically binds a non-enzymatically active portion of the molecule leading to accelerated clearance, but characteristically the PT is prolonged and levels of factor II are low. FFP should be administered to treat bleeding. Treatment is immunosuppressive.

2. ACQUIRED ANTIBODIES TO FACTOR V

Products containing bovine factor V (such as topical thrombin or fibrin glue, frequently used in surgical procedures) can lead to formation of an anti–factor V antibody that cross-reacts with human factor V. Clinicopathologic manifestations range from a prolonged PT in an otherwise asymptomatic individual to severe bleeding. Mixing studies suggest the presence of an inhibitor, and the factor V activity level is low. In cases of serious or life-threatening bleeding, IVIG or platelet transfusions, or both, should be administered, and immunosuppression (as for acquired inhibitors to factor VIII) may be offered.

3. ACQUIRED ANTIBODIES TO FACTOR VIII

Acquired hemophilia A due to factor VIII inhibitors is the most common acquired factor-based bleeding disorder. Spontaneous antibodies to factor VIII (acquired hemophilia A) can occur in adults without a prior history of hemophilia; older adults and patients with lymphoproliferative malignancy or autoimmune disease and those who are postpartum or postsurgical are at highest risk. The clinical presentation, which should be viewed as a medical emergency, typically includes extensive soft tissue ecchymoses, hematomas, and mucosal bleeding, as opposed to hemarthrosis characteristic of congenital hemophilia A. The aPTT is typically prolonged and does not correct upon mixing; factor VIII activity is low and a Bethesda assay reveals the titer of the inhibitor. Inhibitors of low titer (less than 5 BU) may often be overcome by infusion of high doses of factor VIII concentrates, whereas high-titer inhibitors (greater than 5 BU) must be treated with serial infusions of activated prothrombin complex concentrates, recombinant human activated factor VII, or recombinant porcine factor VIII. Along with establishment of hemostasis by one of these measures, immunosuppressive treatment with corticosteroids with or without oral cyclophosphamide or rituximab should be instituted. Treatment with IVIG and plasmapheresis can be considered in refractory cases. Unlike in congenital factor VIII deficiency, the patient’s bleeding does not correlate well with the factor VIII activity level, so the clinician must be concerned with any elevation of aPTT secondary to acquired factor VIII inhibitor. All such patients require immediate referral to a hematologist.

4. VITAMIN K DEFICIENCY

Vitamin K deficiency may occur as a result of deficient dietary intake of vitamin K (from green leafy vegetables, soybeans, and other sources), malabsorption, or decreased production by intestinal bacteria (due to treatment with chemotherapy or antibiotics). Vitamin K is required for normal function of vitamin K epoxide reductase that assists in posttranslational gamma-carboxylation of the coagulation factors II, VII, IX, and X, which is necessary for their activity. Thus, mild to moderate vitamin K deficiency typically features a prolonged PT (activity of the vitamin K–dependent factors is more reflected than in the aPTT; aPTT is prolonged if the deficiency is more severe) that corrects upon mixing; activity levels of individual clotting factors II, VII, IX, and X typically are low. Importantly, a concomitantly low factor V activity level is not indicative of isolated vitamin K deficiency and may indicate an underlying defect in liver synthetic function. Hospitalized patients on broad-spectrum antibiotics and with poor or no oral intake are at high risk for vitamin K deficiency.

For treatment, vitamin K₁ (phytonadione) may be administered via intravenous or oral routes; the subcutaneous route is not recommended due to erratic absorption. The oral dose is 5–10 mg/day and absorption is typically excellent; at least partial improvement in the PT should be observed within 18–24 hours of administration. Intravenous administration results in faster normalization of a prolonged PT than oral administration; due to descriptions of anaphylaxis, parenteral doses should be administered at lower doses (1–5 mg/day) and slowly (eg, over 30 minutes) with concomitant monitoring. Overreplacement can make it difficult to resume warfarin when necessary.

5. COAGULOPATHY OF LIVER DISEASE

Impaired liver function due to cirrhosis or other causes leads to decreased synthesis of clotting factors, including factors II, V, VII, IX, X, and fibrinogen; whereas factor VIII levels, largely made in endothelial cells, may be elevated despite depressed levels of other coagulation factors. The PT (and with advanced disease, the aPTT) is typically prolonged and usually corrects on mixing with normal plasma. A normal factor V level, in spite of decreases in the activity of factors II, VII, IX, and X, however, suggests vitamin K deficiency rather than liver disease. Qualitative and quantitative deficiencies of fibrinogen also are prevalent among patients with advanced liver disease, typically leading to a prolonged PT, thrombin time, and reptilase time.

The coagulopathy of liver disease usually does not require hemostatic treatment unless bleeding occurs. Infusion of FFP may be considered if active bleeding is present and the aPTT and PT are prolonged; however, the effect is transient and concern for volume overload may limit infusions. Patients with bleeding and a fibrinogen level consistently below 80–100 mg/dL should receive cryoprecipitate. Liver transplantation, if feasible, results in production of coagulation factors at normal levels. The use of recombinant human activated factor VII in patients with bleeding varices is controversial, although some patient subgroups may experience benefit. The coagulopathy of liver disease can predispose to bleeding or thrombosis, so caution and experience are needed for optimal management.

6. WARFARIN INGESTION

See Antithrombotic Therapy section, below.

7. DISSEMINATED INTRAVASCULAR COAGULATION

See above.

8. HEPARIN/FONDAPARINUX/DIRECT-ACTING ORAL ANTICOAGULANT USE

See Classes of Anticoagulants, below.

9. LUPUS ANTICOAGULANTS

Lupus anticoagulants prolong the aPTT by interfering with interactions between the clotting cascade and the phospholipid surface on which they function, but they do not lead to bleeding. Rather, they predispose to thrombosis. Lupus anticoagulants were so named because of their early identification in patients with autoimmune disease, although they also occur with increased frequency in individuals with underlying infection, inflammation, or malignancy, and they also can occur in asymptomatic individuals in the general population. A prolongation in the aPTT is observed that does not correct completely on mixing but that normalizes with excessive phospholipid. Specialized testing such as a positive hexagonal phase phospholipid neutralization assay, a prolonged dilute Russell viper venom time, and positive platelet neutralization assays can confirm the presence of a lupus anticoagulant. Rarely, the antibodies also interfere with factor II activity, and that tiny subset of lupus anticoagulant patients are at risk for bleeding.

Occasionally, abnormalities of the vasculature and integument may lead to bleeding despite normal hemostasis; congenital or acquired disorders may be causative. These abnormalities include Ehlers-Danlos syndrome, osteogenesis imperfecta, Osler-Weber-Rendu disease (hereditary hemorrhagic telangiectasia) (see Chapter 40), and Marfan syndrome (heritable defects) and integumentary thinning due to prolonged corticosteroid administration or normal aging, amyloidosis, vasculitis, and scurvy (acquired defects). The bleeding time often is prolonged. The bleeding time reflects the integrity of the vasculature (which is abnormal in collagen synthesis disorders) in addition to activity of platelets and coagulation factors. If possible, treatment of the underlying condition should be pursued, but if this is not possible or feasible (ie, congenital syndromes), globally hemostatic agents such as DDAVP can be considered for treatment of bleeding.

The currently available anticoagulants include unfractionated heparin, LMWHs, fondaparinux, vitamin K antagonists (ie, warfarin), and direct-acting oral anticoagulants (DOACs) (ie, dabigatran, rivaroxaban, apixaban, edoxaban, betrixaban). (For a discussion of the injectable DTIs, see section Heparin-Induced Thrombocytopenia above.)

CLASSES OF ANTICOAGULANTS

A. Unfractionated Heparin and LMWHs

Unfractionated heparin is a biologic product most commonly derived from porcine intestinal tissue rich in heparin-bearing mast cells. It is heterogeneous with respect to sulfation and polymer length; individual molecules may range from 3000 to 30,000. Only about one-third of the molecules in a given preparation of unfractionated heparin contain the crucial pentasaccharide sequence necessary for binding of antithrombin and exerting its anticoagulant effect upon thrombin. Heparin is highly negatively charged, and upon intravenous infusion, it binds to a large array of blood components, such as endothelial cells, platelets, mast cells, and plasma proteins. The degree of anticoagulation with unfractionated heparin is typically monitored by aPTT or anti-Xa level in patients who are receiving the drug in therapeutic doses, although the pharmacokinetics of unfractionated heparin are poorly predictable. Only a fraction of an infused dose of heparin is metabolized by the kidneys, making it safe to use in most patients with significant kidney disease.

The LMWHs are produced from chemical depolymerization of unfractionated heparin, resulting in products of lower molecular weight (mean molecular weight, 4500–6500d, depending on the LMWH). Due to less protein and cellular binding, the pharmacokinetics of the LMWHs are much more predictable than those of unfractionated heparin, allowing for fixed weight-based dosing. All LMWHs are principally renally cleared and must be avoided or used with extreme caution in individuals with creatinine clearance less than 30 mL/min. A longer half-life permits once- or twice-daily subcutaneous dosing, allowing for greater convenience and outpatient therapy in selected cases. Most patients do not require monitoring, although monitoring using the anti-Xa activity level is appropriate for patients with moderate kidney disease, those with elevated body mass index or low weight, and selected pregnant patients. LMWHs are associated with a lower frequency of heparin-induced thrombocytopenia and thrombosis (approximately 0.6%) than unfractionated heparin (3%).

B. Fondaparinux

Fondaparinux is a synthetic molecule consisting of the highly active pentasaccharide sequence found in LMWHs. As such, it exerts almost no thrombin inhibition and works to indirectly inhibit factor Xa through binding to antithrombin. Fondaparinux, like the LMWHs, is almost exclusively metabolized by the kidneys, and should be avoided in patients with creatinine clearance less than 30 mL/min. Predictable pharmacokinetics allow for weight-based dosing.

C. Vitamin K Antagonist (Warfarin)

The vitamin K antagonist warfarin inhibits the activity of the vitamin K–dependent carboxylase that is important for the posttranslational modification of coagulation factors II, VII, IX, and X. Although warfarin is taken orally, leading to a significant advantage over the heparins and heparin derivatives, interindividual differences in nutritional status, comorbid diseases, concomitant medications, and genetic polymorphisms lead to a poorly predictable anticoagulant response. Individuals taking warfarin must undergo periodic monitoring to verify the intensity of the anticoagulant effect, reported as the INR, which corrects for differences in potency of commercially available thromboplastin used to perform the PT.¹

D. Direct-Acting Oral Anticoagulants

Unlike warfarin, the DOACs (1) have a predictable dose effect and therefore do not require laboratory monitoring, (2) have anticoagulant activity independent of vitamin K with no need for dietary stasis, and (3) are renally metabolized to varying degrees so there are restrictions or dose reductions related to reduced kidney function (Table 14–10). While the DOACs have fewer drug interactions than warfarin, if DOACs are given with potentially interacting medications, there is no reliable way to measure the impact on anticoagulant activity of the concomitant administration. There is also no reliable way to measure adherence. Data remain limited on use of DOACs in morbidly obese patients (more than 120 kg or BMI greater than or equal to 40) in VTE treatment. The clinician must carefully consider kidney function, concomitant medications, indication for use, candidacy for lead-in parenteral therapy (as required for acute VTE treatment with edoxaban and dabigatran only) and anticipated patient adherence. Providers must be careful to dose each DOAC properly for the indication, kidney function, and weight of patient, and to check for drug interactions. (See Table 14–10 for details.) There is a reversal agent available for dabigatran and for the anti-Xa inhibitors apixaban and rivaroxaban (Table 14–11).

Routine monitoring is not recommended for patients taking DOACs. However, there are clinical scenarios where assessing anticoagulant activity may be helpful, including active bleeding, pending urgent surgery, suspected therapeutic failure, or concern for accumulation. Drug-specific anti-Xa levels are not widely available, and guidance is lacking regarding clinical approach to the results. DOACs have varying effects on the PT and aPTT. In the absence of drug-specific levels, a normal dilute thrombin time excludes the presence of clinically relevant dabigatran levels; an elevated aPTT suggests clinically relevant levels of dabigatran. An elevated PT suggests clinically relevant levels of rivaroxaban. However, a normal aPTT or normal PT does not rule out clinically significant amounts of dabigatran or rivaroxaban, respectively.

PREVENTION OF VENOUS THROMBOEMBOLIC DISEASE

The frequency of venous thromboembolic disease (VTE) among hospitalized patients ranges widely. Up to 60% of VTE cases occur during or after hospitalization, with especially high incidence among critical care patients and high-risk surgical patients.

Avoidance of fatal PE, which occurs in up to 5% of high-risk inpatients as a consequence of hospitalization or surgery, is a major goal of pharmacologic prophylaxis. Tables 14–12 and 14–13 provide risk stratification for DVT/VTE among hospitalized surgical and medical inpatients. Standard pharmacologic prophylactic regimens are listed in Table 14–14; prophylactic anticoagulation regimens differ in their recommended duration of use. Prophylactic strategies should be guided by individual risk stratification, with all moderate- and high-risk patients receiving pharmacologic prophylaxis, unless contraindicated. Contraindications to VTE prophylaxis for hospital inpatients at high risk for VTE are listed in Table 14–15. In patients at high risk for VTE with absolute contraindications to pharmacologic prophylaxis, mechanical devices such as intermittent pneumatic compression devices should be used, ideally in portable form with at least an 18-hour daily wear time.

It is recommended that VTE prophylaxis be used judiciously in hospitalized medical patients who are not critically ill since a comprehensive review of evidence suggested harm from bleeding in low-risk patients given low-dose heparin and skin necrosis in stroke patients given compression stockings. Risk assessment models like the Padua Risk Score (Table 14–13) and the IMPROVE risk score can help clinicians identify patients who may benefit from DVT prophylaxis. The IMPROVE investigators also developed a bleeding risk model that may aid in identifying acutely ill medical inpatients at increased risk for bleeding: https://www.outcomes-umassmed.org/IMPROVE/risk_score/index.html. While two of the anti-Xa oral anticoagulants (betrixaban and rivaroxaban) have been approved for extended duration prophylaxis after discharge for medically ill patients, how to identify those who will have clinical benefit from this practice is still unclear.

The Caprini score may help guide decisions in surgical patients about VTE prophylaxis (https://www.mdcalc.com/caprini-score-venous-thromboembolism-2005). In addition, certain high-risk surgical patients should be considered for extended-duration prophylaxis of up to 1 month, including those undergoing total hip replacement, hip fracture repair, and abdominal and pelvic cancer surgery. If bleeding is present, if the risk of bleeding is high, or if the risk of VTE is high for the inpatient (Table 14–12) and therefore combined prophylactic strategies are needed, some measure of thromboprophylaxis may be provided through mechanical devices such as intermittent pneumatic compression devices and graduated compression stockings.

A. Primary VTE Prevention in Patients with Active Cancer

Some ambulatory cancer patients undergoing chemotherapy who are at moderate to high risk of VTE (Khorana risk score ≥ 2) (https://www.mdcalc.com/khorana-risk-score-venous-thromboembolism-cancer-patients) may benefit from pharmacologic DVT prophylaxis, although bleeding risk is increased and caution should be taken, particularly in patients with gastrointestinal or intracranial malignancy, and other risk factors for anticoagulant-related bleeding (such as thrombocytopenia and kidney dysfunction). DOACs should be avoided when there are possible interactions with chemotherapeutic agents.

B. Primary VTE Prevention, Diagnosis, and Treatment in Patients with Severe COVID-19

Patients with severe COVID-19 appear to have an increased incidence of thrombotic complications, including venous (DVT, PE) and arterial (stroke, limb occlusion) events. Risk is especially high in the critical care setting. Although the reasons for this hypercoagulability are not yet well understood, the profound systemic inflammatory response associated with severe COVID-19 is thought to play a role. While the hypercoagulability in COVID-19 resembles DIC, laboratory and clinical findings are somewhat different. Laboratory findings in patients with severe COVID-19 may include markedly elevated D-dimer and modestly prolonged prothrombin time. However, patients with COVID-19 tend to have elevated fibrinogen levels; thrombocytopenia is rare and nonsevere; and bleeding complications are unusual. Thrombosis in patients with COVID-19 is associated with a poor prognosis and often occurs despite standard pharmacologic prophylaxis.

1. Risk stratification and initial prognostication of patients with severe COVID-19

Given the prevalence and prognostic value of abnormal laboratory findings at presentation, patients with COVID-19 should have RP/INR, PTT, D-dimers, and fibrinogen measured. When results are abnormal, especially significantly elevated D-dimers or decreased fibrinogen, admission for monitoring should be considered even in patients who are otherwise clinically stable. Worsening laboratory parameters during hospitalization should prompt consideration of transfer to a higher level of care and heightened clinical suspicion for thrombosis.

2. VTE prophylaxis for patients with severe COVID-19

In the absence of strong contraindications, all patients hospitalized with COVID-19 should receive pharmacologic VTE prophylaxis. LMWH is preferred over unfractionated heparin to minimize staff exposure and the chance of heparin-induced thrombocytopenia.

For patients with a prior history of VTE who take an oral anticoagulant for secondary prevention at the time of admission, transition to LMWH should be considered due to its shorter half-life and potential anti-inflammatory properties.

For updated recommendations regarding pharmacologic dosing and post-discharge prophylaxis, refer to professional society guidance (links at end of this section) since guidance in this area is evolving rapidly.

3. Diagnosis and management of thromboembolic disease in patients with severe COVID-19

Logistical challenges complicate the diagnosis of thromboembolism in patients with COVID-19 due to patient instability and risks of staff exposures. D-dimers are generally elevated in hospitalized patients who have COVID-19. A substantial increase in D-dimers may suggest COVID-19–associated coagulopathy with or without thrombotic events. Clinicians should remain vigilant for signs and symptoms of thrombosis and consider obtaining surveillance laboratory testing at least every 3–4 days with low threshold for imaging. Ideally, thrombosis should be confirmed radiographically, but in situations where these studies cannot safely be obtained and clinical suspicion is very high, empiric treatment may be considered.

Guidance from the Anticoagulation Forum (https://acforum.org/web/), the International Society for Thrombosis and Haemostasis (https://academy.isth.org/isth/#!*menu=8*browseby=2*sortby=1*label=19794), and the American Society for Hematology (https://www.hematology.org/covid-19) is evolving and should be frequently consulted.

TREATMENT OF VENOUS THROMBOEMBOLIC DISEASE

A. Anticoagulant Therapy

Treatment for VTE should be offered to patients with objectively confirmed DVT or PE, or to those in whom the clinical suspicion is high for the disorder but who have not yet undergone diagnostic testing (see Chapter 9). The management of VTE primarily involves administration of anticoagulants; the goal is to prevent recurrence, extension and embolization of thrombosis and to reduce the risk of post-thrombotic syndrome. Suggested anticoagulation regimens are found in Table 14–16.

B. Selecting Appropriate Initial Anticoagulant Therapy

Most patients with DVT alone may be treated as outpatients, provided that their risk of bleeding is low and they have good follow-up. Table 14–17 outlines proposed selection criteria for outpatient treatment of DVT.

Among patients with PE, risk stratification at time of diagnosis should direct treatment and triage. Patients with persistent hemodynamic instability are classified as high-risk patients (previously referred to as having “massive PE”) and have an early PE-related mortality of more than 15%. These patients should be admitted to an intensive care unit and generally receive thrombolysis and anticoagulation with intravenous heparin. Intermediate-risk patients (previously, “submassive PE”) have a mortality rate of up to 15% and should be admitted to a higher level of inpatient care, with consideration of thrombolysis on a case-by-case basis. Catheter-directed techniques, if available, may be an option for patients who are poor candidates for systemic thrombolysis and/or in centers with expertise. Low-risk patients have a mortality rate less than 3% and are candidates for expedited discharge or outpatient therapy.

For hemodynamically stable patients, additional assessment focusing on right ventricular dysfunction is warranted to differentiate between low-risk, low-intermediate risk, and high-intermediate risk PE. The Bova score (https://www.mdcalc.com/bova-score-pulmonary-embolism-complications) and the simplified PE severity index accurately identify patients at low risk for 30-day PE-related mortality (Table 14–18) (eTable 14–3) who are potential candidates for expedited discharge or outpatient treatment. Because the Bova score includes serum troponin and evidence of right ventricular dysfunction (by CT or echocardiography), it also identifies patients with high-intermediate risk PE who warrant close monitoring and may require escalation of therapy. An RV/LV ratio less than 1.0 on chest CT angiogram has been shown to have good negative predictive value for adverse outcome but suffers from inter observer variability. Echocardiography may provide better assessment of right ventricular dysfunction when there is concern. Serum biomarkers such as B-type natriuretic peptide and troponin are most useful for their negative predictive value, and mainly in combination with other predictors.

Selection of an initial anticoagulant should be determined by patient characteristics (kidney function, immediate bleeding risk, weight) and the clinical scenario (eg, whether thrombolysis is being considered, active cancer, thrombosis location) as described in Table 14–16.

1. Parenteral anticoagulants

HEPARINS

In patients in whom parenteral anticoagulation is being considered, LMWHs are more effective than unfractionated heparin in the immediate treatment of DVT and PE; they are preferred as initial treatment because of predictable pharmacokinetics, which allow for subcutaneous, once- or twice-daily dosing with no requirement for monitoring in most patients. Accumulation of LMWH and increased rates of bleeding have been observed among patients with severe kidney disease (creatinine clearance less than 30 mL/min), leading to a recommendation to use intravenous unfractionated heparin preferentially in these patients. If concomitant thrombolysis is being considered, unfractionated heparin is indicated. Patients with VTE and a perceived higher risk of bleeding (ie, post-surgery) may be better candidates for treatment with unfractionated heparin than LMWH given its shorter half-life and reversibility. Unfractionated heparin can be effectively neutralized with the positively charged protamine sulfate while protamine may only have partial reversal effect on LMWH. Use of unfractionated heparin leads to heparin-induced thrombocytopenia and thrombosis in approximately 3% of patients, so daily complete blood counts are recommended during the initial 10–14 days of exposure.

Weight-based, fixed-dose daily subcutaneous fondaparinux (a synthetic factor Xa inhibitor) may also be used for the initial treatment of DVT and PE, with no increase in bleeding over that observed with LMWH. Its lack of reversibility, long half-life, and renal clearance limit its use in patients with an increased risk of bleeding or kidney disease.

2. Oral anticoagulants

A. DIRECT-ACTING ORAL ANTICOAGULANTS

DOACs have a predictable dose effect, few drug-drug interactions, rapid onset of action, and freedom from laboratory monitoring (Table 14–10). Dabigatran, rivaroxaban, apixaban, and edoxaban are approved for treatment of acute DVT and PE. While rivaroxaban and apixaban can be used as monotherapy eliminating the need for parenteral therapy, patients treated with dabigatran or edoxaban must first receive 5–10 days of parenteral anticoagulation and then be transitioned to the oral agent per prescribing information. Unlike warfarin, DOACs do not require an overlap since these agents are immediately active; the DOAC is started when the parenteral agent is stopped. Compared to warfarin and LMWH, the DOACs are all noninferior with respect to prevention of recurrent VTE; both rivaroxaban and apixaban have a lower bleeding risk than warfarin with LMWH bridge. While DOACs are recommended as first-line therapy for acute VTE according to the CHEST 2016 VTE guidelines, agent selection should be individualized with consideration of kidney function, concomitant medication use, indication, ability to use LMWH bridge therapy, cost, and adherence.

B. WARFARIN

If warfarin is chosen as the oral anticoagulant it will be initiated along with the parenteral anticoagulant, which is continued until INR is in therapeutic range. Most patients require 5 mg of warfarin daily for initial treatment, but lower doses (2.5 mg daily) should be considered for patients of Asian descent, older adults, and those with hyperthyroidism, heart failure, liver disease, recent major surgery, malnutrition, certain polymorphisms for the CYP2C9 or the VKORC1 genes or who are receiving concurrent medications that increase sensitivity to warfarin (eTable 14–4). Conversely, individuals of African descent, those with larger body mass index or hypothyroidism, and those who are receiving medications that increase warfarin metabolism (eg, rifampin) may require higher initial doses (7.5 mg daily). Daily INR results should guide dosing adjustments in the hospitalized patient while at least biweekly INR results guide dosing in the outpatient during the initial period of therapy (Table 14–19). Web-based warfarin dosing calculators incorporating clinical and genetic factors are available to help clinicians choose appropriate starting doses (eg, see www.warfarindosing.org). Because an average of 5 days is required to achieve a steady-state reduction in the activity of vitamin K–dependent coagulation factors, the parenteral anticoagulant should be continued for at least 5 days and until the INR is more than 2.0. Meticulous follow-up should be arranged for all patients taking warfarin because of the bleeding risk associated with initiation of therapy. Once stabilized, the INR should be checked at an interval no longer than every 6 weeks and warfarin dosing should be adjusted by guidelines (Table 14–20) since this strategy has been shown to improve the time patients spend in the therapeutic range and their clinical outcomes. Supratherapeutic INRs should be managed according to evidence-based guidelines (Table 14–21).

C. Duration of Anticoagulation Therapy

Recurrence rates of VTE after discontinuation of therapy—The clinical scenario in which the thrombosis occurred is the strongest predictor of recurrence and, in most cases, guides duration of anticoagulation (Table 14–22). In the first year after discontinuation of anticoagulation therapy, the frequency of recurrent VTE among individuals whose thrombosis occurred in the setting of a transient, major, reversible risk factor (such as surgery) is approximately 3% after completing 3 months of anticoagulation, compared with at least 8% for individuals whose thrombosis was unprovoked, and greater than 20% in patients with cancer. Men have a greater than twofold higher risk of recurrent VTE compared to women; recurrent PE is more likely to develop in patients with clinically apparent PE than in those with DVT alone and has a case fatality rate of nearly 10%; and proximal DVT has a higher recurrence risk than distal DVT.

1. Provoked versus unprovoked VTE

Patients with provoked VTE are generally treated with a minimum of 3 months of anticoagulation, whereas unprovoked VTE should prompt consideration of indefinite anticoagulation provided the patient is not at high risk for bleeding. Merely extending duration of anticoagulation beyond 3 months for unprovoked PE will not reduce risk of recurrence once anticoagulation is stopped; if anticoagulants are stopped after 3, 6, 12, or 18 months in such a patient, the risk of recurrence after cessation of therapy is similar. Individual risk stratification may help identify patients most likely to suffer recurrent disease and thus most likely to benefit from ongoing anticoagulation therapy. Normal D-dimer levels 1 month after cessation of anticoagulation are associated with lower recurrence risk, although some would argue not low enough to consider stopping anticoagulant therapy, particularly in men.

2. Risk scoring systems to guide therapy duration

The HERDOO2 risk scoring system uses body mass index, age, D-dimer, and post-phlebitic symptoms to identify women at lower risk for recurrence after unprovoked VTE (https://www.mdcalc.com/herdoo2-rule-discontinuing-anticoagulation-unprovoked-vte). The Vienna Prediction Model, a simple scoring system based on age, sex, D-dimer, and location of thrombosis, can help estimate an individual’s recurrence risk to guide duration of therapy decisions.

3. Cancer-related VTE

LMWH has been the mainstay of treatment for cancer-related VTE based on lower VTE recurrence in cancer patients treated with dalteparin compared with warfarin. Studies have also shown that DOACs (edoxaban, rivaroxaban, and apixaban) are at least as effective as LMWH for VTE treatment. The use of edoxaban and rivaroxaban is at the expense of increased bleeding, particularly for patients with gastrointestinal cancer. The International Society for Thrombosis and Haemostasis suggests use of specific DOACs for cancer patients with a diagnosis of acute VTE, no drug-drug interactions, and a low risk of bleeding but suggests use of LMWH for those with a high risk of bleeding, including patients with luminal gastrointestinal cancers with an intact primary tumor, and those at risk for bleeding from the genitourinary or gastrointestinal tract. For patients with intracranial malignancy and VTE, bleeding risk depends on tumor type (primary versus metastatic) and other characteristics; whenever possible, interdisciplinary consultation is recommended to help determine risk of initiating anticoagulation. DOACs do not appear to confer higher bleeding risk compared to LMWH in patients with brain tumors. Clinicians must be aware that chemotherapeutic agents may interact with DOACs and their use should be avoided in cases of potential interactions because there is no easily accessible and reliable way to measure the anticoagulant effect of DOACs.

4. Thrombophilia workup in determining duration

Laboratory workup for thrombophilia is not recommended routinely for determining duration of therapy because clinical presentation is a much stronger predictor of recurrence risk. The workup may be pursued in patients younger than 50 years, with a strong family history, with a clot in unusual locations, or with recurrent thromboses (Table 14–23). In addition, a workup for thrombophilia may be considered in women of childbearing age in whom results may influence fertility and pregnancy outcomes and management or in those patients in whom results will influence duration of therapy. An important hypercoagulable state to identify is antiphospholipid syndrome because these patients have a marked increase in recurrence rates, are at risk for both arterial and venous disease, in general receive bridge therapy during any interruption of anticoagulation, and should not receive DOACs as first-line antithrombotic therapy due to increased arterial events compared to warfarin. Due to effects of anticoagulants and acute thrombosis on many of the tests, the thrombophilia workup should be delayed in most cases until at least 3 months after the acute event, if indicated at all (Table 14–24). The benefit of anticoagulation must be weighed against the bleeding risks posed, and the benefit-risk ratio should be assessed at the initiation of therapy, at 3 months, and then at least annually in any patient receiving prolonged anticoagulant therapy. Bleeding risk scores, such as the Riete score (https://www.mdcalc.com/riete-score-risk-hemorrhage-pulmonary-embolism-treatment) have been developed to estimate risk of these complications. Their performance, however, may not offer any advantage over a clinician’s subjective assessment, particularly in older individuals. Consideration of bleeding risk is of particular importance when identifying candidates for extended duration therapy for treatment of unprovoked VTE; it is recommended that patients with a high risk of bleeding receive a defined course of anticoagulation, rather than indefinite therapy, even if the VTE was unprovoked.

D. Secondary Prevention

Antithrombotic therapy offered after the initial 3–6 months of treatment should be considered in patients with VTE that is not majorly provoked; it is most compelling for those with unprovoked VTE. For most patients who continue to take a DOAC to prevent recurrence, the dose can be reduced to prophylactic intensity after the initial 6–12 months of therapy. In patients deemed poor candidates for ongoing DOAC or warfarin use but who warrant some secondary prevention, low-dose (81–100 mg) aspirin may be used; however, this will provide far less reduction in risk of recurrent VTE with similar bleeding risk.

E. Thrombolytic Therapy

Anticoagulation alone is appropriate treatment for most patients with PE; however, those with high-risk, massive PE, defined as PE with persistent hemodynamic instability, have an in-hospital mortality rate that approaches 30% and, absent contraindications (Table 14–25), require immediate thrombolysis in combination with anticoagulation (Table 14–26). Systemic thrombolytic therapy has been used in carefully selected patients with intermediate-risk, submassive PE, defined as PE without hemodynamic instability but with evidence of right ventricular compromise and myocardial injury. Thrombolysis in this cohort decreases risk of hemodynamic compromise but increases the risk of major hemorrhage and stroke. A lower dose of tPA commonly used for PE treatment has been evaluated in small trials but additional data are needed to recommend its use. Catheter-directed therapy for acute PE may be considered for high-risk or intermediate-risk PE when systemic thrombolysis has failed or as an alternative to systemic thrombolytic therapy.

In patients with large proximal iliofemoral DVT, data from randomized controlled trials are conflicting on the benefit of catheter-directed thrombolysis in addition to treatment with anticoagulation; the CaVenT trial showed some reduction in risk of postthrombotic syndrome, but the larger ATTRACT trial failed to show reduction in postthrombotic syndrome but did find an increased risk of major bleeding.

F. Nonpharmacologic Therapy

1. Graduated compression stockings

Graduated compressions stockings may provide symptomatic relief to selected patients with ongoing swelling but do not reduce risk of postthrombotic syndrome at 6 months. They are contraindicated in patients with peripheral vascular disease.

2. Inferior vena caval (IVC) filters

There is a paucity of data to support the use of IVC filters for the prevention of PE in any clinical scenario. There are two randomized, controlled trials of IVC filters for prevention of PE. In the first study, patients with documented DVT received full- intensity, time-limited anticoagulation with or without placement of a permanent IVC filter. Patients with IVC filters had a lower rate of nonfatal asymptomatic PE at 12 days but an increased rate of DVT at 2 years. In the second study, patients with symptomatic PE and residual proximal DVT plus at least one additional risk factor for severity received full intensity anticoagulation with or without a retrievable IVC filter. IVC filter use did not reduce the risk of symptomatic recurrent PE at 3 months. Most experts agree with placement of an IVC filter in patients with acute proximal DVT and an absolute contraindication to anticoagulation despite lack of evidence to support this practice. While IVC filters were once commonly used to prevent VTE recurrence in the setting of anticoagulation failure, many experts now recommend switching to an alternative agent or increasing the intensity of the current anticoagulant regimen instead. The remainder of the indications (submassive/intermediate-risk PE, free-floating iliofemoral DVT, perioperative risk reduction) are controversial. If the contraindication to anticoagulation is temporary (active bleeding with subsequent resolution), placement of a retrievable IVC filter may be considered so that the device can be removed once anticoagulation has been started and has been shown to be tolerated. Rates of IVC filter retrieval are very low, often due to failure to arrange for its removal. Thus, if a device is placed, removal should be arranged at the time of device placement.

Complications of IVC filters include local thrombosis, tilting, migration, fracture, and inability to retrieve the device. When considering placement of an IVC filter, it is best to consider both short- and long-term complications, since devices intended for removal may become permanent. To improve patient safety, institutions should develop systems that guide appropriate patient selection for IVC filter placement, tracking, and removal.

WHEN TO REFER

• Presence of large iliofemoral VTE, unprovoked upper extremity DVT, IVC thrombosis, portal vein thrombosis, or Budd-Chiari syndrome for consideration of catheter-directed thrombolysis.

• High-risk PE for urgent embolectomy or catheter-directed therapies.

• Intermediate-risk PE if considering thrombolysis.

• History of HIT or prolonged PTT plus renal failure for alternative anticoagulation regimens.

• Consideration of IVC filter placement.

• Clots in unusual locations (eg, renal, hepatic, or cerebral vein), or simultaneous arterial and venous thrombosis, to assess possibility of a hypercoagulable state.

• Recurrent VTE while receiving therapeutic anticoagulation.

WHEN TO ADMIT

• Documented or suspected intermediate- or high-risk PE, low-risk PE at high risk for bleeding, poor candidate for outpatient treatment.

• DVT with poorly controlled pain, high bleeding risk, or concerns about follow-up.

• Large iliofemoral DVT for consideration of thrombolysis.

• Acute DVT and absolute contraindication to anticoagulation for IVC filter placement.

• Venous thrombosis despite therapeutic anticoagulation.

• Suspected Paget-Schroetter syndrome (spontaneous upper extremity thrombosis related to thoracic outlet syndrome).