Common Acquired Causes of Thrombosis in Children

https://doi.org/10.1016/j.cppeds.2013.05.005Get rights and content

Compared to adults, venous thromboembolism in the pediatric population is a rare event. Cancer, cardiac disease, antiphospholipid antibodies, and indwelling catheters are established risk factors for thromboembolism in children. We examined the literature related to thrombophilia in children, childhood cancer and thrombosis, cardiac disease and thrombosis, and antiphospholipid antibody syndrome in children. Citations in identified articles yielded additional articles for review. We found that studies of acquired thrombophilia in children are limited. Current treatment for thromboembolism in children is based on adult data therefore optimal treatment in this population remains unclear.

Introduction

The annual incidence of venous thromboembolism (VTE) in children is estimated at 0.07–0.49 per 10,000 children1, 2, 3, 4, 5 with the highest frequency of events occurring in the neonatal period. More than 90% of thrombotic events in children are a result of underlying risk factors or disease processes. Here, we discuss some of the more common diseases associated with thrombosis in the pediatric population: congenital heart disease, malignancies, and the antiphospholipid antibody syndrome.

Congenital heart disease (CHD) affects approximately 1% of all live births. Over the past 20 years, medical and surgical advances have improved the survival rate of children born with these defects.6, 7 Despite advances in treatment, thromboembolic complications remain a major cause of morbidity in children with CHD.8, 9 Thromboembolism in these patients occurs as venous or arterial thromboembolism, intracardiac thrombi, pulmonary embolism, and embolism to the central nervous system. Other clinical factors that contribute to the formation of thrombosis include major surgery, angiography, the presence of a central venous catheter, and local or systemic infection, with central venous catheter placement and surgery being the most common precipitants.9, 10, 11

Cardiac catheterization is an important diagnostic and treatment procedure and the most common procedure performed in children with cardiac disease.12 In this procedure, access is obtained via the femoral artery, which can lead to thrombus formation in the iliofemoral system, causing loss of arterial blood flow, ischemia, and limb loss. Thrombosis associated with cardiac catheterization in children ranges from 0.8% to 40% for arterial thrombosis and 0–20% for venous thrombosis.13, 14, 15 The use of unfractionated heparin has been proven to decrease arterial thrombosis and is currently recommended in all children undergoing cardiac catheterization. In the absence of prophylactic unfractionated heparin, thrombosis following cardiac catheterization occurs in approximately 40% of children.16, 17, 18

Cardiac surgical procedures, particularly those designed to treat patients with pulmonary valve obstruction and/or single-ventricle physiology, are generally performed in a stepwise fashion and the risk of thrombosis differs with each procedure.19, 20, 21, 22 In most cases, the Blalock–Taussig shunt (BTS) is first placed during the neonatal period. This is followed by the Glenn shunt, and finally, the Fontan shunt, which is considered to be the definitive palliative procedure.19, 20, 21, 22

The BTS is also commonly used in pulmonary atresia, tricuspid atresia, and Tetralogy of Fallot. In this procedure, a graft is placed between the subclavian artery and the ipsilateral pulmonary artery in an attempt to increase pulmonary blood flow in the neonatal period. Thrombotic occlusion of a BTS ranges from 1% to 17%.23, 24 Risk factors for the development of occlusion include patient age, graft size, underlying heart defect, and weight at the time of the operation. Acute occlusion of the shunt is usually treated with thrombolytic therapy, stent insertion, or balloon dilation.25, 26, 27 The prophylactic use of aspirin at low doses (1–10 mg/kg/day) in these patients was shown to significantly decrease risk of shunt thrombosis and death.28

The Glenn shunt is an intermediate step that is typically performed after a BTS and before the definitive Fontan surgery in patients with single-ventricle physiology. It allows blood to flow directly from the SVC into the pulmonary arteries and the lungs, increasing pulmonary blood flow and oxygen saturation without increasing the already heavy workload of the single ventricle. Occlusion of the shunt via thrombosis leads to a dramatic decrease in pulmonary blood flow. Thrombosis after Glenn shunts has been inconsistently reported in the literature, and there are currently no recommendations for routine thromboprophylaxis in these patients.29

The Fontan procedure is the definitive palliative surgical procedure for patients with single-ventricle physiology such as tricuspid atresia, pulmonary atresia with intact ventricular septum, double-inlet ventricle, and hypoplastic left heart syndrome.22 It is usually performed after a Glenn procedure and involves the routing of the systemic venous return to the pulmonary arteries while bypassing the ventricle. This allows the single ventricle to function as the left ventricle by generating systemic blood flow. Although considered the definitive palliative procedure in this population, Fontan shunts are associated with significant morbidity and mortality due to thromboembolism, which occurs at an incidence of 3–20%.30 Clots may occur at any time after the procedure but usually present months after. The reason behind thrombus formation after the procedure remains unclear but is believed to be multifactorial and includes factors such as underlying physiology, presence of arrhythmias, and presence of a CVL. There is currently no consensus regarding routine thromboprophylaxis for patients after Fontan, although most centers use either low-dose aspirin or a vitamin K antagonist such as warfarin.31

Mechanical prosthetic valves are placed in children with a variety of congenital or acquired cardiac valve disease. As a result of this introduction of foreign material into the body, the placement of these valves poses a great risk for thrombosis. The American College of Chest Physicians (ACCP) published guidelines in 2012 regarding anticoagulation for specific clinical situations in children with cardiac disease.29 Current recommendations for adults and children with artificial valves involve the use of vitamin K antagonists (VKA) or aspirin for those who developed thrombosis while taking VKA therapy.

Children with CHD represent a significant portion of pediatric patients with thrombosis. Optimal prophylaxis and treatment for these patients remains unclear and strongly indicates the need for more research.

Thromboembolism has been reported at the time of cancer diagnosis and during treatment.32 The pathogenesis of thrombosis in childhood cancer is related to the disease itself and its effects on the coagulation and fibrinolytic systems, the presence of central venous lines (CVL), and the effects of chemotherapeutic agents used.33, 34, 35 Children with cancer comprise approximately 8–22% of pediatric patients with thrombosis.1

The presence of a central venous catheter is the most common risk factor for thrombosis in all pediatric patients.1, 36 Thrombosis is more commonly associated with external CVLs than internal CVLs and usually involves the upper venous system. Compared to children with other malignancies, children with acute lymphoblastic leukemia (ALL) are more prone to develop CVL-related thrombosis. In a study by Glaser et al.37 in 2001, 50% of children with cancer and implantable ports had evidence of thrombosis at diagnosis. CVL-related thrombosis may present as line malfunction, DVT, or prominent chest wall veins depending on the location of the thrombus.

Malignant cells have procoagulant, pro-inflammatory, and antifibrinolytic properties. These cells express procoagulants such as tissue factor on their cell surfaces, which results in the initiation of coagulation through the activation of thrombin.38, 39 In one study, children with acute myeloblastic leukemia (AML) were found to have increased tissue factor activity in the bone marrow whereas children with ALL did not. Children with ALL were reported to have decreased amounts of circulating protein C and S activity.40 They also display increased expression of plasminogen activator inhibitor-1 (PAI-1). PAI-1 inhibits fibrinolysis, thereby promoting coagulation.41 Children with leukemia and lymphoma have been shown to have increased generation and activity of thrombin at the time of diagnosis which persists throughout treatment.42

ALL is the most common childhood malignancy and therefore has the highest incidence of reported thromboembolic events. The majority of these events are related to the presence of central venous lines.1, 36, 37, 43 Thrombosis related to l-asparaginase, a mainstay of ALL therapy, has also been reported.44, 45, 46 The incidence of thromboembolism in childhood ALL is reported to range from 1.1% to 36.7%.47 Caruso et al. conducted a meta-analysis of 17 prospective studies of thromboembolic events in pediatric patients with ALL. The study involved a total of 1752 patients in which the overall risk of symptomatic thromboembolic events was reported to be 5.2% over the entire course of therapy, a rate at least 100 times greater than that observed in the general pediatric population.48

Thromboembolic events have also occurred in solid tumors such as Ewing's sarcoma, rhabdomyosarcoma, nephroblastoma, and neuroblastoma. The incidence of thromboembolic events in these and other childhood cancers remains unclear as less data is available.49 In a retrospective study of 122 pediatric patients with sarcoma, Paz-Priel et al. found that approximately 16% of patients had thrombosis with half of those identified at the time of diagnosis. They also reported that 80% of those patients with thrombosis at the time of diagnosis also had metastatic disease.50

More than 90% of thrombotic events in childhood leukemia occur during induction therapy. The remainder of events occurs during the consolidation or intensification phases of leukemia treatment.32 As a result, research has focused on the association between thrombosis and chemotherapy given during these treatment phases. l-asparaginase is an important component of induction and intensification therapy in childhood ALL. The use of l-asparaginase either alone or in combination therapy for the treatment of ALL has improved disease response and cure rates. It is well documented that l-asparaginase therapy is associated with severe thromboembolic complications.45, 46, 47, 48, 51 Asparaginase-related thrombosis usually involves the cerebral venous sinuses.52 The mechanism of action by which l-asparaginase leads to thrombosis is through decreased hepatic synthesis of plasminogen, fibrinogen, and antithrombin (AT), protein C, and protein S. Deficiencies of these proteins lead to impaired inhibition of thrombin and therefore promote coagulation. Vyzantiadis et al.53 hypothesized that l-asparaginase may also produce thrombosis via a platelet agonist effect in vitro. Some reports have shown that Erwinia asparaginase is associated with a lower risk of thrombosis compared with Escherichia coli asparaginase. However E. coli asparaginase has been shown to result in superior outcomes when used in the initial treatment of ALL in children.54

Steroids such as prednisolone are usually given with asparaginase in the treatment of childhood ALL and are also known to promote thrombosis. Steroids cause an increase in the levels of factor VIII, VWF, prothrombin, AT, and PAI-1, thereby promoting coagulation. The combination of prednisolone and asparaginase in pediatric patients suggests potential synergy between the two in regards to clot formation.55

Inherited thrombophilic risk factors have also been reported to play a role in the development of VTE in childhood cancer. These include the prothrombin G20210A variant, the factor V G1691A mutation, and deficiencies of protein C, protein S, or antithrombin. In a prospective study by Nowak-Gottl et al.56 children with ALL and at least one inherited prothrombotic risk factor had a higher incidence of thrombosis compared to those without any inherited risk factors. This study also found that protein C, protein S, and AT deficiencies were associated with the greatest risk of VTE and that patients with multiple inherited defects had a significantly higher risk.

Treatment of thromboembolic disease in children with cancer should be based on the specifics of the patient's clinical situation. As in other children with thromboembolism, treatment of thrombosis in children with cancer generally involves the use of unfractionated heparin, low-molecular-weight heparin, and warfarin. Factors that affect choice of pharmacologic therapy include clinical status, presence of other thrombotic risk factors, patient age as well as frequency of monitoring involved.82 Routine thromboprophylaxis is not recommended.29, 57

Antiphospholipid antibody syndrome (APS) is a multisystem autoimmune hypercoagulable state characterized by arterial and/or venous thrombosis, history of recurrent fetal loss, and by the presence of persistent circulating antiphospholipid antibodies.58, 59 APS is considered the most commonly acquired hypercoagulation state of autoimmune disorder in children.58, 59 Primary APS, or that occurring in the absence of another underlying disease, is very rare in the pediatric population. In the largest published APS cohort, disease onset prior to age 15 years occurred in only 2.8% of patients. Avcin et al. in their retrospective analysis of 121 pediatric patients with APS reported mean age at the onset of antiphospholipid syndrome was 10.7 years. In this study, patients with primary antiphospholipid syndrome were younger and had a higher frequency of arterial thrombotic events.60

The antiphospholipid antibodies (aPLs) are immunoglobulins (IgG, IgM, and IgA) directed against phospholipids and phospholipid-binding proteins expressed on, or bound to, the surface of vascular endothelial cells or platelets. In the healthy child, the aPLs are usually transient and present in low amounts. Their presence in otherwise healthy children is likely secondary to vaccinations, drug administration, or previous infections.61

The three major aPLs in APS most readily identified in the lab are anticardiolipin antibody, lupus anticoagulant, and β2-glycoprotein 1 antibody.

  • Anticardiolipin antibodies (aCL) are directed against cardiolipin, an important component of the inner mitochondrial membrane. It is essential for optimal functioning of multiple enzymes involved in energy metabolism as well as maintaining membrane potential. aCLs are classified as IgM, IgG, or IgA or as either β2-glycoprotein dependent or independent. There is a subset of aCLs that also bind to apolipoprotein H. This leads to inhibition of protein C and subsequent dysregulation of the common pathway of coagulation. aCLs have been noted in multiple diseases including APS, syphilis (β2-glycoprotein independent), systemic lupus erythematosus (β2-glycoprotein dependent), and Behcet's syndrome.62, 63

  • Lupus anticoagulant (LA), despite its name, functions as both a procoagulant and anticoagulant. In vivo, it is believed that LA antibodies interact with phospholipids on the platelet membrane, increasing adhesion and aggregation of platelets and thereby precipitating the formation of thrombi. In vitro, the antibodies interfere with phospholipids that are used in coagulation and inhibit agglutination leading to a prolonged activated partial thromboplastin time (aPTT).

  • Anti-β2-glycoprotein 1 antibodies (also called anti-apolipoprotein H antibodies) are often found associated with both LA and aCL. They are mainly found in autoimmune diseases such as lupus and have a strong association with deep vein thrombosis in these conditions.64, 65, 66 Antiβ2-glycoprotein 1 antibodies increase activated factor X (factor Xa) production. In the coagulation cascade Factor Xa cleaves prothrombin to thrombin. Thrombin in turn cleaves fibrinogen to fibrin, leading to clot formation. β2-Glycoprotein 1 is required for the recognition of some aCLs. Those aCLs that bind β2-glycoprotein 1 have been associated with an increased risk of thrombosis.67

Antibodies to prothrombin, phosphatidylserine, phosphatidylinositol, phosphatidylethanolamine, and annexin V have also been shown to have an association with APS but their interpretation in the clinical setting has not yet been established.68, 69, 70

Persistent antiphospholipid antibodies in children are associated with an increased risk of thrombotic events and contribute to a higher risk of cerebrovascular accident. In spite of this association, the actual role these antibodies play in the development of thrombosis remains unclear.71, 72 Some proposed mechanisms include alterations in the protein C–protein S pathway in which antibodies bind to protein S and decrease protein C efficiency. Other proposed mechanisms in APS are a direct procoagulant effect on platelets, inhibition of endothelial cell release of prostacyclin, and impairment of fibrinolysis. The current school of thought regarding the pathogenesis of APS holds that once aPLs are present, a “second-hit” is required for the development of the full-blown syndrome.73, 74 This explanatory mechanism is based on the fact that many individuals who have elevated levels of aPL do not develop the APS.

Because of its tendency to produce thrombosis in any part of the vasculature, APS can present with a wide variety of clinical manifestations in multiple organs. In a study by Avcin et al.60 venous thrombosis occurred in 60%, arterial thrombosis in 32%, small-vessel thrombosis in 6%, and mixed arterial and venous thrombosis in 2% of children with APS. Although thrombosis is one of the main complications of APS, children with antiphospholipid antibodies generally experience a lower rate of thrombotic events compared to adults with APS. This is partially related to the fact that children have higher levels of physiologic anticoagulants and lower levels of coagulation factors than adults. The lower rate of thrombosis in children may also be related to the absence of other thrombophilic risk factors such as smoking, atherosclerosis, oral contraceptive use, and pregnancy in this age range. The healthier vascular endothelium in children greatly enhances the antithrombotic potential of the vessel wall.75

The thrombotic events most frequently reported in children with APS include lower extremity deep venous thrombosis (DVT), pulmonary embolism and thrombosis in the central nervous system. Of these, DVT is the most common, occurring in approximately 30% of all patients with APS. It is predominantly seen in patients with secondary APS. In patients with primary APS, arterial thromboembolism and stroke are the most common events seen, with stroke being the most common arterial thrombotic as well as neurologic event in childhood APS.76 LA antibodies, specifically those that target β2-glycoprotein 1, have been shown to have the closest association with thrombosis in pediatric APS.64, 65, 66, 67, 68, 69

APS in children more frequently occurs in association with SLE or lupus-like disease than in the primary form. Clinical manifestations in this group of patients include both arterial and venous thrombosis, leukopenia, thrombocytopenia, livedo reticularis, and neuropsychiatric manifestations. Thrombosis in patients with SLE and APS presents as pulmonary embolism, DVT, cerebral ischemic stroke, renal artery thrombus, and retinal artery occlusion.64, 77 These thrombotic events are particularly common in those patients with positive LA. Persistently positive aCL in patients with SLE and APS has a reported association with neuropsychiatric manifestations such as migraines, stroke, seizures, chorea, cognitive decline, and psychosis. Development of chorea is more common in children with APS and SLE than adults and has been shown to be associated with persistently positive LA.

The pediatric classification of APS is adapted from adult APS criteria. As per the most recent revised Sapporro Criteria (Sydney Criteria) in 2006, APS is diagnosed if a patient fulfills at least one of the clinical criteria and at least one of the laboratory criteria (Table 1).59 Currently, there are several tests available for aPL. Overlap of the antibodies on testing is not uncommon, so it is important to use multiple tests for detection. The presence of aPL antibodies results in prolongation of the PT and aPTT which does not correct in the presence of normal plasma. Detection of LA is performed using phospholipid-based clotting assays such as aPTT, kaolin clotting time, or diluted Russell's viper venom time. The diluted Russell's viper venom time is more sensitive and is currently the preferred test.78 This testing is confirmed using exogenous phospholipid which adsorbs the aPL antibodies allowing normal coagulation to occur. The aCL assay is performed using enzyme-linked immunosorbent assay (ELISA) which measures both IgG and IgM aCL isotypes (IgG is more specific than IgM). Anti-β2-glycoprotein 1 antibodies are also detected using ELISA.79

Asymptomatic aPL positive children have a very low risk of thrombosis and rarely develop complications. Therefore, thromboprophylaxis is not recommended in these children. Treatment in this group of patients should focus on identification and removal of thrombophilic risk factors.

In the presence of an acute thrombotic event, anticoagulation is usually initiated with heparin or low-molecular-weight heparin followed by oral anticoagulants for long-term management. Long-term anticoagulation is recommended for these children to prevent recurrence of thrombosis, but a consensus on the duration or intensity of therapy does not exist. Low-molecular-weight heparin has become the anticoagulant of choice in most children, as it requires less frequent dosing and monitoring than vitamin K antagonists (VKA) and has fewer medication interactions.80 VKAs such as warfarin are generally used for medium- to long-term anticoagulation. An INR of 2–3 is the recommended goal for all children with aPL and thrombosis on anticoagulation therapy.29, 65 In children with APS, duration and intensity of therapy is usually decided by the physician based on the child's clinical status, risk factors, and recurrence. Anticoagulant therapy is also indicated for APS-related ischemic strokes.29

In secondary APS, treatment should be directed at aggressive treatment of the underlying disease. Studies suggest that children with aPL in the presence of underlying autoimmune diseases such as SLE have a higher risk of thrombosis and a higher risk of recurrence. It is suggested that therapy in these children should continue until the underlying disease is in a more dormant phase and that thromboprophylaxis should be considered.81, 83

Section snippets

Conclusion

Both inherited and acquired risk factors play a role in the development of VTE in children, with acquired risk factors being more common. These include cardiac disease, cancer, and the presence of antiphospholipid antibodies. The risk of thrombosis with or without one of these conditions is greatly increased in the presence of an indwelling catheter. Treatment recommendations for acute thrombotic events in the pediatric population are based largely on evidence from adult studies. The large

References (83)

  • B. Olas et al.

    Cancer procoagulant and blood platelet activation

    Cancer Lett

    (2001)
  • T.C. Haddad et al.

    Chemotherapy-induced thrombosis

    Thromb Res.

    (2006)
  • J.R. Priest et al.

    The effect of l-asparaginase onantithrombin, plasminogen, and plasma coagulation during therapy for acute lymphoblastic leukemia

    J Pediatr

    (1982)
  • V. Caruso et al.

    Thrombotic complications in childhood acute lymphoblastic leukemia: a meta-analysis of 17 prospective studies comprising 1752 pediatric patients

    Blood

    (2006)
  • M. Duval et al.

    Comparison of Escherichia coli-asparaginase with Erwinia-asparaginase in the treatment of childhood lymphoid malignancies: results of a randomized European Organisation for Research and Treatment of Cancer-Children's Leukemia Group phase 3 trial

    Blood

    (2002)
  • U. Nowak-Gottl et al.

    Thromboembolic events in children with acute lymphoblastic leukemia (BFM protocols): prednisone versus dexamethasone administration

    Blood

    (2003)
  • U. Nowak-Gottl et al.

    Prospective evaluation of the thrombotic risk in children with acute lymphoblastic leukemia carrying the methylenetetrahydrofolate reductase TT 677 genotype, the prothrombin G20210A variant, and further prothrombotic risk factors

    Blood

    (1999)
  • O. Vaarala et al.

    Anticardiolipin response in acute infections

    Clin Immunol Immunopathol

    (1986)
  • M. Galli et al.

    Anticardiolipin antibodies directed not to cardiolipin but to a plasma protein cofactor

    Lancet

    (1990)
  • T. Koike et al.

    Antiphospholipid antibodies: lessons from the bench

    J Autoimmun

    (2007)
  • E. Von Scheven et al.

    Clinical characteristics of antiphospholipid antibody syndrome in children

    J Pediatr

    (1996)
  • F. Fischetti et al.

    Thrombus formation induced by antibodies to β2-glycoprotein I is complement dependent and requires a priming factor

    Blood

    (2005)
  • S.S. Pierangeli et al.

    Experimental thrombosis and antiphospholipid antibodies: new insights

    J Autoimmun

    (2000)
  • J.I. Malowany et al.

    Canadian Paediatric Thrombosis and Hemostasis Network. Enoxaparin for neonatal thrombosis: a call for a higher dose for neonates

    Thromb Res

    (2008)
  • B. Schmidt et al.

    Neonatal thrombosis: report of a prospective Canadian and international registry

    Pediatrics

    (1995)
  • B.E.S. Gibson et al.

    Thromboembolism in childhood: a prospective two year BPSU study in the United Kingdom. February 2001–February 2003

    Thromb Haemost

    (2003)
  • J.R. Boris et al.

    The use of anticoagulation in pediatric cardiac disease

    Images Pediatr Cardiol

    (2003)
  • L.C. Jordan

    Stroke in childhood

    Neurologist

    (2006)
  • P. Monagle et al.

    Outcome of paediatric thromboembolic disease: a report from the Canadian childhood thrombosis registry

    Pediatr Res

    (2000)
  • M. Gillespie et al.

    Determinants of intensive care unit length of stay for infants undergoing cardiac surgery

    Congenit Heart Dis

    (2006)
  • D. Bennet et al.

    Incidents and complications during pediatric cardiac catheterization

    Paediatr Anaesth

    (2005 Dec)
  • D.E. Miga et al.

    Incidence of femoral vein occlusion after catheter ablation in children: evaluation with magnetic resonance angiography

    Pediatr Cardiol

    (1997)
  • Z.R. Bulbul et al.

    Arterial complications following cardiac catheterization in children less than 10 kg

    Asian Cardiovasc Thorac Ann

    (2002)
  • S. Laurin et al.

    Venous thrombosis after cardiac catheterization in infants

    Acta Radiol

    (1987)
  • C. Smith et al.

    Pediatric vascular injuries

    Surgery

    (1981)
  • W. Mortensson et al.

    Percutaneous catheterization of the femoral vessels in children II. Thrombotic occlusion of the catheterized artery: frequency and causes

    Pediatr Radiol

    (1975)
  • M.D. Freed et al.

    The use of heparinization to prevent arterial thrombosis after percutaneous cardiac catheterization in children

    Circulation

    (1974)
  • A. Blalock

    Physiopathology and surgical treatment of congenital cardiovascular defects

    Harvey Lect

    (1945–1946)

    41

    (1945–1946)
  • W. Glenn

    Circulatory bypass of the right side of the heart. IV. Shunt between superior vena cava and distal right pulmonary artery; report of clinical application

    N Engl J Med

    (1958)
  • F. Fontan et al.

    Surgical repair of tricuspid atresia

    Thorax

    (1971)
  • K.T. Tsai et al.

    Modified Blalock–Taussig shunt: statistical analysis of potential factors influencing shunt outcome

    J Cardiovasc Surg (Torino)

    (1996)
  • View full text