The antiphospholipid syndrome (APS) is an autoimmune multisystem disease characterized by thromboembolic events, pregnancy morbidity, hematological, dermatological, neurological, and other manifestations in the presence of elevated titers of antiphospholipid antibodies (aPLs). APS may occur as an isolated clinical entity (primary APS) or in association with other diseases, mainly systemic lupus erythematosus (SLE). It occasionally occurs with other autoimmune conditions, infections, and malignancies.
Definition and Classification
Preliminary classification criteria for APS were developed by consensus in 1998. It was proposed that the term APS should designate patients who suffered from vascular thrombosis or recurrent fetal losses associated with the presence of aPLs, namely the lupus anticoagulant (LA) or the anticardiolipin antibodies (aCLs) of immunoglobulin G (IgG) and/or IgM isotype in medium or high titers detected on two or more occasions at least 6 weeks apart. The classification criteria for definite APS were revised in 2006 and include the presence of antibodies against β 2 glycoprotein I (anti-β 2 GPIs) of IgG and/or IgM isotype as part of the updated laboratory criteria and require aPLs to be positive on more than one occasion at least 12 weeks apart ( Table 24-1 ). The revised 2006 criteria were evaluated only in studies in adult populations and require further validation; however, it is generally assumed that they limit the risk of misclassification of patients with transient aPLs and provide a more selective and risk-stratified framework for evaluating patients with persistently positive aPLs.
APS in children has been largely reported in patients with vascular thromboses and less frequently in association with isolated neurological or hematological manifestations. Pregnancy morbidity, which represents one of the two clinical criteria for definite APS in adults, is not applicable to the pediatric population, and it is possible that current consensus criteria may fail to recognize a subgroup of pediatric patients who do not have vascular thrombosis but demonstrate typical nonthrombotic clinical features and fulfill the laboratory criteria for APS. A classification of probable APS has been given to patients with aPLs who have clinical features associated with APS that do not meet the APS criteria, such as heart valve disease, livedo reticularis, thrombocytopenia, nephropathy, and neurological manifestations.
The term seronegative APS was proposed for patients with clinical manifestations highly suggestive of APS, but with persistently negative results in the commonly used assays to detect aCLs, anti-β 2 GPIs, and LA. Some of these patients may have so-called noncriteria aPLs including antibodies to phosphatidylethanolamine, phospholipid-binding plasma proteins (prothrombin, protein C, protein S, annexin V, and domains of β 2 GPI), phospholipid–protein complexes (vimentin/cardiolipin complex), and anionic phospholipids other than cardiolipin (phosphatidylserine, phosphatidylinositol, and phosphatidic acid). The clinical relevance of noncriteria aPLs is still controversial; however, it is possible that some of the new aPL assays will be added as a laboratory diagnostic marker in future revisions of classification criteria for APS.
Catastrophic antiphospholipid syndrome (CAPS) is another subset of APS that is characterized by acute microvascular occlusive disease with subsequent multiorgan failure and a high mortality rate. Preliminary classification criteria for CAPS were established in 2002 ( Table 24-2 ). This syndrome is defined as clinical involvement of at least three organ systems and/or tissues over a very short period of time (less than a week) with histopathological evidence of small-vessel occlusion and laboratory confirmation of the presence of aPLs. Diagnostic algorithms for CAPS were updated in 2010 to emphasize possible overlap with other thrombotic microangiopathies.
|Definite Catastrophic APS|
|Probable Catastrophic APS|
APS is considered to be the most common acquired prothrombotic state of autoimmune etiology, with an estimated incidence around 5 per 100,000 persons per year and the prevalence of 40 to 50 cases per 100,000 persons. The cumulative retrospective analysis indicates that approximately one third of patients with aPLs have a history of thrombosis.
There are no reliable data on the incidence or prevalence of APS in the pediatric population because there are no validated criteria, and the diagnosis rests on the application of adult guidelines and clinical judgment. In a large cohort study of 1000 consecutive patients with APS from 13 European countries, 85% of patients were diagnosed between ages 15 and 50 years. Those with disease onset before 15 years of age accounted for 2.8% of patients with APS. Although the incidence of thrombosis in children is significantly lower than in adults, the proportion of thrombosis that is attributable to aPLs in children appears to be higher than in the adult population, which has other common prothrombotic risk factors such as atherosclerosis, cigarette smoking, hypertension, and use of oral contraceptives. The prevalence of aPLs in unselected children with thrombosis was reported between 12% and 25%. Meta-analysis of 16 observational studies investigating the association of aPLs and first onset of thromboembolism in children showed persistent aPL positivity in 1% to 22% of arterial and 2% to 12% of venous thrombotic events.
The demographic characteristics of 121 pediatric patients with aPL-related thrombosis, included in an international registry of APS, revealed a mean age at disease onset of 10.7 years (range, 1.0 to 17.9 years). There was a moderate female predominance in pediatric APS studies with a female-to-male ratio ranging from 1.2 : 1 to 3 : 1, whereas in adult APS studies the female-to male ratio is over 5 : 1. This difference may reflect, in part, a sampling bias, because the adult APS studies included patients with thrombosis as well as women with pregnancy morbidity. Very little is known about the geographic and racial distribution of pediatric APS.
Primary Antiphospholipid Syndrome
Primary isolated APS without other underlying disease accounts for 40% to 50% of pediatric patients with APS. This percentage may be somewhat overestimated because a number of children initially have primary APS and later during follow-up develop overt SLE. During the 6.1-year mean follow-up period, 21% of children who were initially diagnosed with primary APS progressed to have either clear-cut SLE or lupuslike disease. Comparisons between pediatric patients with primary APS and those with APS associated with underlying autoimmune disease suggest that children with primary APS are significantly younger and have a higher frequency of arterial thrombotic events, especially cerebrovascular ischemic events. In contrast, children with APS associated with underlying autoimmune disease have a higher frequency of venous thrombotic events associated with hematological and skin manifestations.
From the pediatric APS registry data, it is estimated that 50% to 60% of all APS cases in pediatric populations are associated with underlying autoimmune disease. There are only limited data addressing how autoimmune disease can modify the clinical expression of APS, and sometimes the clinical distinction between primary APS and APS associated with autoimmune disease can be difficult to make. This is especially true in SLE, which has many overlapping features with APS, including thrombocytopenia, hemolytic anemia, seizures, and proteinuria.
SLE and lupuslike disease account for the majority (80% to 90%) of pediatric APS cases associated with underlying autoimmune disease. SLE is the autoimmune disease in which aPLs can be found most often, with the reported aPL frequencies ranging from 19% to 87% for aCL, 27% to 48% for anti-β 2 GPI, and 10% to 62% for LA, respectively. A meta-analysis of the published studies that investigated the prevalence and clinical significance of aPLs in childhood-onset SLE showed a global prevalence of 44% for aCLs, 40% for anti-β 2 GPIs, and 22% for LA.
Patients with juvenile idiopathic arthritis (JIA) very rarely develop aPL-related thrombotic events in spite of the high percentage of positive aPLs. Anticardiolipin antibodies were reported in 7% to 53% of JIA patients, but anti-β 2 GPIs and LA, felt to be more specific for thrombosis risk than aCL, were detected in less than 5% of patients. The aPL profile observed in JIA appears to have limited pathogenic potential and may explain the low incidence of thromboembolism in this disease.
Isolated cases of APS were reported in a variety of other pediatric autoimmune diseases including Henoch–Schönlein purpura, Behçet disease, polyarteritis nodosa, immune thrombocytopenic purpura, hemolytic uremic syndrome, and rheumatic fever. The risk of aPL-associated thrombotic events seems to be particularly high in systemic vasculitides, which are generally regarded as hypercoagulable disorders. Patients with immune thrombocytopenic purpura associated with aPLs are at risk for developing both bleeding and thrombotic complications.
Many viral and bacterial infections in childhood can induce de novo production of aPL in previously negative patients. Infection-induced aPLs tend to be transient, present in low titer, and are generally not associated with clinical manifestations of APS. The majority of postinfectious aPLs differ immunochemically from those seen in patients with autoimmune diseases and do not require the presence of cofactor plasma proteins such as β 2 GPI for binding. Because common viral and bacterial infections occur so frequently in children, a high percentage of incidental aPL positivity might be expected in a pediatric population. The association between infections and aPLs is supported also by some indirect evidence, such as seasonal distribution of aPLs.
The distinction between nonpathogenic “postinfectious aPLs” and thrombogenic “autoimmune aPLs” is not absolute, and it was demonstrated that several infections may induce production of heterogeneous aPLs, including pathogenic antibodies against β 2 GPI and prothrombin, resembling those found in autoimmune diseases. Preceding or concomitant infections were found in approximately 10% of children with primary APS or APS associated with autoimmune disease and as high as 60% in children with CAPS. Pediatric APS was most frequently reported in association with varicella-zoster virus, parvovirus B19, human immunodeficiency virus (HIV), streptococcal and staphylococcal infections, Gram-negative bacteria, and Mycoplasma pneumonia. The main causes for infectious induction of APS could be molecular mimicry between infectious agents and β 2 GPI in certain predisposed subjects or unmasking of cryptic antigenic determinants of naturally occurring anti-β 2 GPIs.
There have been isolated case reports of the association of aPLs with thrombotic events in children with various malignancies, including solid tumors and lymphoproliferative and hematological malignancies. Malignancy is an important risk factor for the development of childhood thrombosis and the presence of aPLs may enhance the thrombophilic state of patients with neoplasms. However, aPLs were found as one of the acquired prothrombotic risk factors in less than 3% of thrombotic children with malignancy. Published data suggest that APS associated with malignancies accounts for less than 1% of all children with APS. It appears that aPL-related thrombotic events associated with malignancies are more common in elderly patients, particularly in association with solid tumors.
Low levels of aPLs can be found in up to 25% of apparently healthy children, which is higher than the rate seen in the normal adult population. Such naturally occurring aPLs are usually transient, present in low titer, and could be the result of previous infections or vaccinations. In apparently healthy children, the estimated frequency of aCLs ranges from 3% to 28%, and of anti-β 2 GPIs from 3% to 7%. LA has also been described in apparently healthy children, usually as an incidental finding of prolonged activated partial thromboplastin time (aPTT) in preoperative coagulation screening. The risk of future thrombosis is exceedingly low in otherwise healthy children who were incidentally found to have positive aPLs.
Increasing evidence suggests that alternative responses of the developing immune system to nutritional antigens can result in production of specific nonpathogenic anti-β 2 GPIs during childhood. It was shown that dietary bovine β 2 GPI from milk or meat products may act as an oral immunization agent and induce transitory production of IgG anti-β 2 GPIs in up to 55% of healthy infants whose intestinal mucosa is more permeable to large molecules than that of adolescents and healthy adults. During the prospective follow-up of infants born to mothers with APS or aPL-positive autoimmune disease, it was observed that aCL titers in the newborns’ sera progressively decrease; at 6 and 12 months of age, all infants were negative for aCLs. In contrast, anti-β 2 GPIs were found at 12 months of age in up to 64% of infants born to aPL-positive mothers and in 33% of infants born to mothers with aPL-negative autoimmune disease, further implying postnatal de novo synthesis of anti-β 2 GPIs in infants. Current evidence suggests that postnatally produced anti-β 2 GPIs found in infants have low thrombosis risk, display unusual epitope specificity directed against domain V of β 2 GPI, and are clearly different from pathogenic anti-β 2 GPIs found in patients with APS, which preferentially target domain I. For clinical practice it seems prudent to consider the detection of aCLs, but not anti-β 2 GPIs, to evaluate the disappearance of transplacentally acquired maternal aPLs.
Etiology and Pathogenesis
Although aPLs are found in healthy children, it is clear these antibodies can be associated with thrombosis (see the section “ Clinical Manifestations ”); pregnancy morbidity; hematological, skin and neurological conditions; and they can produce signs and symptoms of microangiopathy. Production of aPLs appears to be triggered by infections of all sorts, and in some instances, it can be familial or hereditary. The presence of any one of several prothrombotic risk factors, including underlying autoimmune disease such as childhood-onset SLE, dramatically increases the risk of thromboembolic disease when the child has aPL. This “two hit-theory” implies that although aPL may inhibit repair of endothelial and platelet surface defects, thereby promoting thrombosis, a perturbation of these membranes is in some way required as well. Animal models have contributed significantly to our understanding of the pathogenesis of APS in humans, as discussed in the sections that follow. Still, it is difficult to ascertain which autoantibodies and in what setting aPLs will cause thromboembolic disease in an individual child.
aPLs cause disease through a variety of effects on endothelial cells, platelets, monocytes, and neutrophils. These antibodies promote complement activation, inhibit physiological anticoagulants such as activated protein C, antithrombin, and the annexin A5 anticoagulant shield, and can impair fibrinolysis. Furthermore, aPLs increase the procoagulant function of cells such as platelets, endothelial cells, and leukocytes. Recent evidence implicates oxidative stress, including decreases in paraoxonase activity ; effects on endothelial nitric oxide synthase ; increased lipid peroxidation in plasma ; and direct antibody-mediated cross-linking and activation of apolipoprotein E receptor 2 in the pathogenesis of thrombosis as well.
As recently reviewed, aPLs have broad-ranging biological effects that not only perturb the interaction between cells and the plasma that bathes them, but also disrupt the orderly function of biological membranes. There is increasing evidence that aPLs promote atherosclerosis, cardiac valve disease, and directly contribute to trophoblast and neuronal dysfunction.
Characteristics of aPLs That Influence Pathogenicity
In humans, thrombotic disease related to aPLs is more likely to occur in individuals with abnormal functional assays such as LA, or prolonged phospholipid-dependent clotting assays; very high titers of aPLs confer higher risk of thrombotic disease than do lower antibody titers. IgG and IgM antibodies are generally associated with hypercoagulable complications. However, IgA antibodies to cardiolipin can be found as the only isotype of aPLs in Afro-Caribbean adults with thrombophilia. Furthermore, sera from patients with IgA aCLs can produce thrombosis in mice. IgA aCL in adults with SLE or other related autoimmune diseases is associated with thrombocytopenia. Likewise, in a large multiethnic cohort, IgA anti-β 2 GPI, even as the only detectible aPL, was associated with at least one APS-related clinical syndrome in the majority of patients.
Antibodies to specific membrane phospholipids such as phosphatidylcholine, phosphatidylethanolamine, and phosphatidylserine are less reliable indicators of risk of thrombotic disease than protein or protein-phospholipid–directed antibodies. A study of 230 adults with SLE examined six tests for aPLs, including LAs, aCLs, anti-β 2 GPIs, solid-phase antiprothrombin, antiphosphatidylethanolamine and antiphosphatidylserine/prothrombin (aPS/PT) (23 possible combinations of test results). Antibodies to domain I of β 2 GPI are more closely associated with thrombosis and with abnormalities of functional coagulation tests in children. The combination of abnormal results for LA in combination with both the anti-β 2 GPI and aPS/PT antibody (“triple positive”) was most closely associated with APS (odds ratio [OR] 23.2) than any combination of two positive tests (OR 3.1-7.3). In children as in adults, aPLs that are persistent for at least 12 weeks are more likely to be associated with thrombosis than transient autoantibodies, hence the international consensus definition requiring detection at least 12 weeks apart.
Most authors recommend screening with functional tests such as the lupus-sensitive aPTT or dilute Russell viper venom test (dRVVT), then performing a mixing test to ensure that the prolonged clotting test is not due to deficiency of one or more clotting factor, and then confirming with a different phospholipid-dependent assay (kaolin clotting time, hexagonal phospholipid clotting test, etc.). To better assess the risk for thrombosis or obstetric risk in aPL-positive patients and those with a variety of autoimmune diseases, a score was developed by Otomo et al. that utilizes the combined strength of several assays to predict risk. Pathogenic autoantibodies to annexin A2, galectins, and a variety other antigens have been described in association with APS in adults; their diagnostic and prognostic utility has yet to be confirmed, especially in children.
Anti-β 2 GPI binds to the platelet von Willebrand factor (vWF) receptor, glycoprotein Ib alpha, and the apolipoprotein E receptor 2 (ApoE2R), promoting platelet adherence to endothelium. In vivo platelet activation in patients with APS has been demonstrated by the finding of high levels of urinary thromboxane metabolites; in vitro aPLs induce release of thromboxane from platelets. aPL-induced platelet aggregation in vitro requires partial preactivation of the platelets with thrombin, collagen, or ADP; aPL then completes that activation via p38 mitogen-activated protein (MAP)-kinase phosphorylation, leading to release of thromboxane and activation of phospholipase A2. For anti-β 2 GPI to induce this effect, expression of the platelet ApoE2R and the platelet glycoprotein Ib alpha subunit is required. Anti-β 2 GPI from sera of APS patients augments the expression of platelet P-selectin in response to thrombin receptor-activating peptide 6, and antibody derived from patients during the catastrophic phase of APS shows greater enhancement than that from quiescent stage donors. The same study showed APS subjects have higher plasma soluble P-selectin, CD40 ligand, monocyte chemoattractant protein 1, and soluble vascular cell adhesion molecule 1. In addition, platelet factor 4, released by activated platelets, promotes dimerization of β 2 GPI on the platelet surface. The interaction of aPLs with platelets is illustrated in simplified format in Fig. 24-1 . These platelet-activating effects of aPLs are blocked in vitro with hydroxychloroquine, suggesting an important mechanism for the effect of this drug in APS. The hydroxychloroquine effect can be seen in images of artificial membrane bilayers treated with β 2 GPI and then with monoclonal anti-β 2 GPI. This disrupts the membrane, making globular outcroppings that are dissipated following the addition of hydroxychloroquine ( Fig. 24-2 ).
Endothelial Cells in Antiphospholipid Syndrome
ALPs increase the adhesiveness of endothelial cells to leukocytes and platelets by increasing expression of intracellular adhesion molecule (ICAM)-1, vascular adhesion molecule (VCAM)-1, and E-selectin. Mice genetically deficient in these adhesion molecules have impaired aPL-induced thrombus formation, thereby confirming the importance of this effect in the pathology of APS. Endothelial cells treated with aPLs increase production of interleukin (IL)-6 and reactive oxygen species, promoting an endothelial inflammatory phenotype. Tissue factor, an important initiator of coagulation via the extrinsic pathway, is highly expressed by endothelium of vessels from patients with APS. Anti-β 2 GPIs activate p38-MAP kinase on endothelium as well as platelets, increasing expression of tissue factor by endothelium exposed to aPLs. This tissue factor overexpression contributes to the prothrombotic effects of aPLs. Activation of nuclear factor (NF)-κB is an essential step in the endothelial effects of aPLs, and MG132, an inhibitor of NF-κB, blocks the tissue factor upregulation induced by aPLs. Pretreatment with fluvastatin blocks the aPL endothelial cell effects on tissue factor and adhesion molecule expression. β 2 GPI blocks vWF-dependent platelet aggregation, and aPLs interfere with this inhibition of platelet aggregation. Plasma concentrations of von Willebrand protein are high in patients with primary APS, further promoting platelet aggregation and the prothrombotic effects of aPLs. IgG from subjects with APS can also stimulate endothelial release of microparticles, which themselves can promote thrombosis.
β 2 GPI–aPL complexes bind tightly to endothelial cells, and that binding is mediated by annexin A2. Antibodies to annexin A2 produce the same activation of endothelial cells as do anti-β 2 GPIs, with the same kinetics. This effect is only present with cross-linking antibodies and not present with F(ab′) monomers, suggesting that perturbation of annexin A2 in the membrane mediates the anti-β 2 GPI–induced endothelial activation through effects on annexin A2. β 2 GPI promotes the activation of plasminogen by tissue plasminogen activator; this suggests that β 2 GPI may be an endogenous regulator of fibrinolysis. Consequently, impairment of β 2 GPI-augmented fibrinolysis by anti-β 2 GPI may contribute to thrombosis in patients with APS.
aPLs activate endothelial cells via the Toll-like receptor 4 (TLR4), activating the adapter protein MyD88 and downstream effects; TLR4 knockout mice are resistant to aPL-induced thrombosis. In mouse fibroblasts a similar role for TLR2 has been demonstrated. This involvement of the innate immune system may be a critical link in comprehending the pathogenesis of this complex disease. The interaction of aPLs with endothelial cells is illustrated in Fig. 24-3 .
Leukocyte Activation by Antiphospholipid Antibodies
aPLs affect monocytes in ways similar to how they affect platelets. They upregulate tissue factor expression and adhesion molecule expression, the production of IL-1, IL-6 and IL-8 via phosphorylation of p38 MAPK, and the activation of NF-κB and MEK/ERK kinases. Recently, involvement of both NF-κB and c-Jun/activator protein 1 (AP1) pathways were demonstrated in this effect. Monocyte effects of aPLs are likely important in preactivating endothelial cells to a prothrombotic phenotype. Neutrophils are required for fetal loss related to aPLs, as neutrophil depletion in mice prevents the intrauterine growth retardation and fetal demise related to passive aPL antibody administration. Neutrophil accumulation in decidual tissues is mediated by complement activation products, particularly C5a. Tissue factor expression on neutrophils is increased by aPLs, and blockade of tissue factor by monoclonal antibody treatment prevents fetal injury.
The Effect of aPLs on Placental Trophoblasts
In addition to promoting inflammation and coagulation at the level of endothelial cells and leukocytes within vessels, there is mounting evidence that aPLs have direct toxic effects on the functions of a variety of other cells. Most prominent among these are the effects on placental trophoblasts. This has best been demonstrated in a murine model of APS, where serum IgG antibody and monoclonal human aPL can induce fetal loss in pregnant mice. In particular, aPL can bind directly to the trophoblast in the developing embryo. Following binding, these antibodies inhibit the normal invasion of the trophoblast into the decidual tissues, which results in defective placentation. Trophoblast production of chorionic gonadotropin and placental lactogen is impaired following aPL treatment. Because trophoblasts express anionic phospholipids on the surfaces during differentiation, they bind β 2 GPI in vivo , making them a target for aPLs. Matrix metalloproteinases, required for placental invasion, are underexpressed in APS, and heparin increases trophoblast invasiveness by increasing matrix metalloproteinases expression.
Complement Activation in the APS
Antigen–antibody complexes made of aPLs and the membrane phospholipids to which they are directed, like most immune complexes, can activate complement via the classical pathway. Complement activation generates a variety of biologically active cleavage fragments that have serine protease activity, anaphylatoxic effects, and promote neutrophil adhesion to activated endothelium and influx into tissues. In an animal model of aPL-induced fetal loss, complement activation appears to be necessary for fetal demise and resorption, and aPL-induced thrombus formation. This complement effect makes intuitive sense, as both complement and the thrombosis systems are cascades of interacting serine proteases. Serine protease inhibitors (serpins), several of which are common to both pathways, regulate both coagulation and thrombosis. Studies utilizing intravital microscopy following immunological “priming” with lipopolysaccharide then intraarterial injection of aPLs in rats demonstrated complement activation in the aPL-treated thrombotic vessels. In contrast, thrombus did not form as efficiently in complement-deficient C6-/- animals. In addition, an anti-C5 antibody that blocks C5 activation, C5a release, and C5b-9 assembly also had diminished thrombus formation. Thus complement plays a crucial role in promoting the placental inflammation that leads to fetal demise in APS, and in promoting thrombus formation.
aPLs can have antifibrinolytic effects and have been shown to inhibit the activities of thrombin, activated protein C, plasminogen, and plasmin. Anti-β 2 GPIs inhibit the anticoagulant activity of activated protein C. Some APS patients have demonstrable antibody to protein S and/or protein C. There is in vivo evidence that this inhibition of thrombolysis contributes to the pathogenicity of aPL. In fact, measuring inhibition of activated protein C activity may be more sensitive for pathogenic aPL than currently employed assays.
In summary (see Table 24-3 ), aPLs have a variety of procoagulant effects, including the activation of platelets to aggregate and release thromboxane, activation of monocytes (inflammatory cytokine production and increased tissue factor expression), and activation of endothelium (increased tissue factor expression and secretion, increased vWF, and expression of adhesion molecules). They also impede downregulation of thrombosis via impairment of activated protein C. Furthermore, through aPL actions on complement, endothelium and neutrophils are activated to promote tissue accumulation of neutrophils, thus promoting local inflammation.
|Impaired Function of Activated Protein C|
|Decreased Binding of C4 to C4BP|
|Promote Clot Formation|
That APS may be familial in a minority of patients has been recognized for decades. The findings of a high incidence of aCLs in first-degree relatives of people with APS or with SLE suggest that the capacity to develop antibodies to charged phospholipids, at the least, may be influenced by genetics, probably with important environmental influences. Allelic variations in β 2 GPI protein associated with APS have been identified. Inheritance in familial APS may be autosomal recessive, or dominant/codominant.
In interpreting the studies of familial APS, it is important to remember several factors: although primary and secondary APS share a number of features, it is not clear that their genetic basis is similar. Reports of aPL rates in pediatric family members of patients with APS must be interpreted in light an evidence of higher antibody positivity rates in healthy children than in adults.
There are several reports of an association with aPLs and familial Sneddon syndrome, a syndrome of livedo reticularis with hypertension and cerebrovascular disease, which is inherited as an autosomal-dominant trait. The role of aPLs in this disease is as yet uncertain.
Most reports of familial APS are anecdotal and do not provide information on the genetic transmission of the disease. The number of reports and small series published provide evidence that familial occurrence and possibly genetic influence are important, at least in a subset of human APS. Other issues that cloud reports of familial APS are variations in whether, and how completely, patients were evaluated for other prothrombotic traits such as inherited heterozygous deficiency of protein C or protein S, or resistance to activated protein C from such genetic disorders as factor V Leiden and methylenetetrahydrofolate reductase (MTHFR) deficiencies. Any of these defects can coexist with aPLs and increase the likelihood of thrombosis in the presence of aPLs. The number of patients reported with aPLs who have genetic defects predisposing to pathological thrombosis raises the possibility that disordered coagulation itself might indeed predispose to the formation of aPLs. This question has not been directly addressed.
It is not known how often APS is familial. Estimates based on family history usually do not include complete laboratory evaluation of family members for APS, or for concomitant thrombophilia disorders. Consequently, such studies likely overestimate the rate of “affected” relatives. Prospective studies of families utilizing current definitions of APS and standardized testing technology are not yet available.
A variety of human leukocyte antigen (HLA)-DR and HLA-DQ antigens have been associated with aPLs and both primary and secondary APS (summarized in eTable 24-4 ). Many of these associations lose significance when corrected for the number of variables tested, however. These studies are generally small series, and most do not include appropriate ethnic controls. Arnett and colleagues studied major histocompatibility complex (MHC) genotypes in 20 patients with LAs, of whom 8 had primary APS and 12 had other rheumatic diseases (7 with significant thrombotic complications). The strongest association with LAs was seen for DQB1*0301 (DQw7), and all who were negative for this allele had DQB1*0302 (DQw8). These two antigens share an identical 7-amino acid sequence in the third hypervariable region of the DQ molecule, suggesting this epitope might be important in mediating an immune response to phospholipid. Recently, a large genotype analysis of HLA associations with vascular disease and aPL positivity in Swedish patients with SLE showed DRB 1*04/*13 alleles were closely associated with aPLs and thrombotic disease.
|CLINICAL CONDITION||HLA-ALLELE ASSOCIATION|
|Primary APS||DR4, DR5|
|Lupus anticoagulant||DR5, DRw52b, DQB1*0301 (DQw7), or DQB1*0302 (DQw8)|
|Anticardiolipin antibody||DR7, DR4|
|Anti-β 2 glycoprotein-I antibody||DQB1*0604/5/6/7/9; DQA1*0102; DRB1*1302 and DQB1*0303|
The large number of associations reported between various measures of aPLs and related thrombotic disease and alleles in the human MHC support the fact that these genes may play a significant role in facilitating formation of antibody to phospholipid antigens, but other factors are likely to be important in producing familial APS. As of now, single gene associations with APS are not confirmed and remain an area of investigation.
Children with aPLs may present with any combination of vascular occlusive events or with a variety of nonthrombotic clinical manifestations. Most of the clinical features that occur in adults with aPLs have also been described in children. However, the clinical expression of aPLs in children is modified by several characteristics such as the immaturity of the immune and other organ systems, the absence of common prothrombotic risk factors often present in adults, the lack of pregnancy morbidity, and the presence of routine immunizations and frequent exposure to common viral and bacterial infections. Data from large registries of pediatric patients with APS have provided information on the spectrum of thrombotic and nonthrombotic clinical manifestations. At the time of the initial thrombotic event in children with definite APS, the estimated frequencies of associated nonthrombotic manifestations were 38% to 53% for hematological manifestations, 6% to 18% for dermatological manifestations, and 16% to 22% for nonthrombotic neurological manifestations.
A high percentage of children with persistent aPLs apparently do not present with overt thrombotic events. Studies in children with aPLs demonstrated that thromboses occurred only in 16% to 36% of patients, whereas the nonthrombotic aPL-related clinical manifestations alone were observed in more than 40% of patients.
The classical clinical picture of APS in pediatric populations is characterized by venous, arterial, or small-vessel thrombosis. Vascular occlusion in APS may involve arteries and veins at any level of the vascular tree and in all organ systems, giving rise to a wide variety of clinical presentations (summarized in Table 24-5 ). Meta-analyses investigating the association of aPLs and thromboembolism in children revealed that the presence of persistent aPLs shows a significant association with a first thrombotic event during childhood with an overall summary OR of 5.9. The association of aPLs appears slightly stronger for arterial thrombosis (OR 6.6) than for venous thrombosis (OR 4.9). Thrombosis is more likely to occur with the existence of additional hereditary and acquired prothrombotic risk factors. It is now well established that APS may develop as an initial manifestation of childhood-onset SLE, and all children presenting with aPL-related thrombosis require thorough assessment for evidence of underlying systemic disease.
|VESSEL INVOLVED||CLINICAL MANIFESTATIONS|
|Limbs||Deep vein thrombosis|
|Skin||Livedo reticularis, chronic leg ulcers, superficial thrombophlebitis|
|Large veins||Superior or inferior vena cava thrombosis|
|Lungs||Pulmonary thromboembolism, pulmonary hypertension|
|Brain||Cerebral venous sinus thrombosis|
|Eyes||Retinal vein thrombosis|
|Liver||Budd–Chiari syndrome, enzyme elevations|
|Adrenal glands||Hypoadrenalism, Addison disease|
|Brain||Stroke, transient ischemic attack, acute ischemic encephalopathy|
|Eyes||Retinal artery thrombosis|
|Kidney||Renal artery thrombosis, renal thrombotic microangiopathy|
|Gut||Mesenteric artery thrombosis|
Venous thrombosis is the most common vascular occlusive event seen, occurring in up to 60% of pediatric APS patients. The most frequently reported site of venous thrombotic events is deep vein thrombosis in the lower extremities, followed by cerebral sinus vein thrombosis, portal vein thrombosis, deep vein thrombosis in the upper extremities, and superficial vein thrombosis. Venous thrombotic events are particularly common in APS associated with childhood-onset SLE. In a Canadian retrospective cohort study in childhood-onset SLE, 13 of the 149 patients (9%) had one or more thromboembolic events, and all 13 patients with thromboembolic events were LA positive. In total, venous thrombosis occurred in 76% of episodes (cerebral venous thrombosis in nine, deep vein thrombosis in four, pulmonary embolism in two, and retinal vein occlusion in one) and arterial thrombosis in 24% (arterial stroke in three, retinal artery occlusion in one, and splenic infarct in one). The overall incidence of thromboembolic events in childhood-onset SLE patients with positive LAs was 54% (13 of 24 patients) and with positive aCLs, 22% (12 of 54 patients). LAs were found to be the strongest predictor of thrombosis risk in another study in 58 children with SLE, and positivity for multiple aPL subtypes indicates stronger associations with thromboembolic events than for individual aCL, anti-β 2 GPI, or antiprothrombin antibodies. Overall, it has been estimated that childhood-onset SLE patients with persistent LA positivity have a twenty-eight-fold increased risk of thrombotic events compared with patients who are negative for LA.
Arterial thrombosis occurs in approximately 30% of pediatric APS patients. The most frequently reported site of arterial thrombotic event is ischemic stroke, followed by peripheral arterial thrombosis. Arterial thrombotic events are significantly more common in primary APS patients than in APS associated with underlying autoimmune disease. Several studies have demonstrated that the prevalence of aPL-related cerebral ischemia is particularly high in pediatric and young adult patients, ranging from 15% to 75 %. In an Israeli study evaluating the importance of various thrombophilia markers in 58 pediatric patients with stroke, positive aPLs were found in 15% of patients, and only factor V Leiden and aPLs were found to be significant risk factors for ischemic stroke in children. aPLs are also an independent risk factor for recurrent ischemic stroke in children, but this effect has not been confirmed for IgG aCLs. Sneddon syndrome is associated with aPLs in approximately 50% of all patients.
Small-vessel thrombosis occurs in less than 10% of pediatric APS patients and may present as an aggressive microvascular occlusive disease (CAPS) or localized small-vessel thrombosis. Localized small-vessel thrombosis has been described primarily in children with isolated thrombosis of digital vessels or renal thrombotic microangiopathy. Peripheral vascular disease leading to digital gangrene is a well-recognized complication of APS, particularly in patients with SLE, and it may be difficult to distinguish this from vasculitis, cryoglobulinemia, or disseminated intravascular coagulation. In a group of 32 Mexican children with APS, digital ischemia was reported as the most frequent thrombotic event present in 44% of patients at the onset of APS.
The most common hematological manifestations associated with APS in children are thrombocytopenia, autoimmune hemolytic anemia, and leucopenia. Thrombocytopenia was observed in 20% to 25% of children with APS, often in association with Coombs-positive hemolytic anemia (Evans syndrome). Thrombocytopenia associated with APS is usually mild or moderate, with platelet counts greater than 50 × 10 9 /L. Treatment is usually not required except in cases with a platelet count less than 30 × 10 9 /L and symptomatic with bleeding. The pathogenesis of aPL-related thrombocytopenia is probably heterogeneous, including direct binding of aPL to platelet phospholipids, immune-mediated and platelet activation. Occasionally, thrombocytopenia may be severe, causing major bleeding. Thrombotic events are unusual with severe thrombocytopenia but may occur if the platelet counts increase during the disease course. Leucopenia or lymphopenia were reported in 8% of children with APS.
A high percentage (greater than 25%) of patients with isolated immune thrombocytopenic purpura also have aPLs, particularly in the pediatric population. In a case control study of 42 children with immune thrombocytopenic purpura, IgG aCLs were found in 78%, and anti-β 2 GPIs were found in all chronic cases, whereas patients with acute immune thrombocytopenic purpura demonstrated IgG aCLs in just 27% and anti-β 2 GPIs in 13%, respectively. During a 4-year follow-up period, 17% of children who were initially diagnosed with immune thrombocytopenic purpura developed overt SLE, and closer follow-up has been suggested for these children. Several studies have reported an increased risk of thrombosis in aPL-positive patients who present with isolated hematological manifestations. Some patients with aPL-related hematological manifestations, however, continue to have thrombocytopenia or hemolytic anemia as isolated clinical manifestations and do not develop thrombosis or SLE during follow up.
The acquired LA-hypoprothrombinemia syndrome is a rare complication consisting of a severe bleeding diathesis associated with the presence of LA. It has been described in both primary APS and APS associated with SLE, and is often preceded by a viral infection. This complication has been attributed to the presence of antiprothrombin antibodies that cause rapid depletion of plasma prothrombin and consequent hemorrhagic diathesis.
A wide variety of dermatological manifestations have been reported in patients with APS, ranging from minor signs to life-threatening conditions such as widespread cutaneous necrosis. The most common dermatological manifestations present in children included in the pediatric APS registry were livedo reticularis (6%), Raynaud phenomenon (6%), and skin ulcers (3%) ( Fig. 24-4 ).
In a single-center series of 200 consecutive pediatric and adult patients with APS, skin manifestations were observed in 49% of the patients and were the presenting symptom in 30%. The most frequent manifestation was livedo reticularis (25%), followed by digital necrosis (7%), subungual splinter hemorrhages (5%), superficial venous thrombosis (5%), postphlebitic skin ulcers (4%), circumscribed cutaneous necrosis (3%), thrombocytopenic purpura (3%), and other changes (7%). Livedo reticularis, and especially the coarser livedo racemosa variant, is considered a major clinical feature of APS, strongly associated with arterial and microangiopathic subtypes of APS ( Fig. 24-5 ). The mechanism of livedo reticularis is not well understood and may include both patchy thrombosis and/or the interaction of aPLs with endothelium and induction of vasoconstriction. The association of aPLs with Raynaud phenomenon was noted also in a retrospective study of 123 children, in which at least one aPL subtype was positive in up to 36% of patients with primary and 30% of patients with secondary Raynaud phenomenon.
Typical neurological manifestations of APS are ischemic stroke and cerebral sinus vein thrombosis, both caused by thrombotic occlusion of cerebral vessels. Several other neurological manifestations have been associated with the presence of aPLs, including various movement disorders, epilepsy, migraine, cognitive defects, psychiatric diseases, transverse myelitis, multiple sclerosis–like disorders, sensorineural hearing loss, and Guillain–Barré syndrome. These manifestations are not fully explained by the thrombogenic effects of aPLs and may result from both thrombotic and nonthrombotic immune-mediated mechanisms such as direct interaction between aPLs and neuronal tissue or immune complex deposition in the cerebral blood vessels wall.
The most common nonstroke neurological manifestations observed in children with APS were migraine headache (7%), chorea (4%), and seizures (3%). Chorea has been strongly linked to the presence of aPLs as an isolated clinical finding or in children with SLE. A large retrospective cohort study of 137 children with SLE demonstrated an association between LA and chorea over the disease course, but not between aPLs and other neuropsychiatric manifestations. A significant association was found between aPLs and childhood seizure disorder in two prospective studies, but not all studies have found such a link. The association between epilepsy and aPLs is clouded also by the role of anticonvulsant drugs that could trigger autoantibody production. There has been controversy concerning a possible association of aPLs and migraines, which has not been confirmed in a prospective study in an unselected group of children with migraines. Prospective studies of adult patients with SLE demonstrated an association between persistent aPLs and cognitive impairments, especially in areas of attention, psychomotor speed, and executive functioning. Neurocognitive defects were frequently observed also in childhood-onset SLE patients, but there is no clear evidence suggesting association with aPLs. In general, APS may constitute a potentially treatable cause of a variety of neurological diseases, and it is recommended to routinely measure aPLs in children with otherwise unexplained neurological disorders.
Cardiac manifestations are frequent in adult patients with APS but have not been extensively investigated in childhood. The most prominent cardiac manifestations include valvular disease, occlusive coronary artery disease, cardiomyopathy, and intracardiac thrombosis. Nonbacterial (Libman–Sacks) vegetations were disclosed by echocardiographic studies in 11% of adult patients with APS, but were only rarely observed in patients with pediatric APS. Several cases of pediatric patients with aPL-related myocardial infarction have been reported, often in association with underlying SLE or congenital heart disease. Multiple small vascular occlusions are responsible for APS cardiomyopathy, especially in CAPS, in which it is one of the most common causes of death.
Pulmonary embolism and infarction constitute the most frequent pulmonary manifestation of APS. aPLs were reported in 30% to 40% of children with pulmonary embolism who were referred for hematology evaluation. Rarely, recurrent pulmonary embolism may lead to pulmonary hypertension.
The kidney is a major target organ of pediatric APS with manifestations that include renal vascular occlusion, thrombotic glomerular microangiopathy, and hypertension. The term antiphospholipid syndrome-associated nephropathy (APSN) was proposed to describe thrombotic microangiopathy involving both arterioles and glomerular capillaries that cause hypertension, acute renal failure, proteinuria, and poor renal function with a tendency to develop end-stage renal disease. This entity was reported in 16% of Thai patients with childhood-onset SLE who underwent renal biopsy and was significantly more frequent in adult SLE patients (41%). aPLs are also associated with hepatic, digestive, and adrenal manifestations resulting from occlusive vascular disease of intraabdominal vessels.
Osteoarticular manifestations such as avascular necrosis of bone, nontraumatic fractures, and bone marrow necrosis are rarely seen in APS patients. Adult patients with primary APS and no prior glucocorticoid treatment appear to have an increased risk of avascular necrosis. aPLs were reported as one of the most common prothrombotic alterations in patients with multifocal osteonecrosis, which is present in 20% of cases. Perthes disease has been linked with aPLs in two pediatric studies, but the association does not appear to be strong.
Catastrophic Antiphospholipid Syndrome
CAPS is a rare, potentially life-threatening variant of APS, characterized by multiple small-vessel occlusions that can lead to multiorgan failure. Large-vessel occlusions may also occur in CAPS, but they do not dominate the clinical picture. The most commonly affected organ systems include the kidney, lung, central nervous system (CNS), heart, and skin.
In a large cohort of 446 cases collected in the CAPS international registry, 45 patients (10%) developed catastrophic events before 18 years of age, and 87% of pediatric patients had CAPS as initial presentation without any previous history of thrombosis. Primary APS occurred in 69% of pediatric cases with CAPS, and 31% of patients suffered from underlying childhood-onset SLE or lupuslike disease. The most frequently affected organs in pediatric patients with CAPS were kidneys and lung (both 63%), followed by heart (57%), peripheral vessel thrombosis (52%), brain (48%), liver (40%), skin (37%), and gastrointestinal tract (17%). Thrombocytopenia was observed in more than 70% of patients and was one of the hallmarks of pediatric CAPS. Precipitating factors were identified in more than 75% of cases, most frequently infections (61%), malignancy (17%), surgery (6%) and SLE flares (4%).
CAPS has been described also in several pediatric case reports, but overall this aggressive disease represents less than 5% of pediatric APS patients. It is not known why this disorder behaves in such an aggressive fashion in some patients. Frequently, patients with CAPS have other non-aPL risk factors that contribute to the acute microvascular thrombosis such as infection with or without sepsis; disseminated, intravascular coagulation; surgery; and underlying autoimmune or malignant disease. The reported mortality rate in pediatric patients with CAPS is significantly higher than in classic APS and ranges between 26% and 33%.
Microangiopathic Antiphospholipid-Associated Syndrome
aPLs have also been associated with a variety of microangiopathic syndromes that cannot be attributed simply to the microvascular thrombotic process. The term microangiopathic antiphospholipid-associated syndrome was introduced to refer to patients with aPLs and clinical features of thrombotic microangiopathy with hemolytic anemia, severe thrombocytopenia, and the presence of schistocytes. The clinical presentations of 46 adult patients with thrombotic microangiopathic hemolytic anemia associated with aPLs included hemolytic uremic syndrome (HUS) (25%), CAPS (23%), acute renal failure (15%), malignant hypertension (13%), thrombotic thrombocytopenic purpura (TTP) (13%), and other clinical presentations (11%).
Two pediatric series reported a high frequency of aCLs in children with diarrhea-associated HUS, but without a clear role in the pathogenesis of microangiopathy. Microangiopathic antiphospholipid-associated syndrome was reported in a 4-year-old child who presented with atypical HUS and later developed rapidly progressive thrombotic microangiopathy associated with the transient presence of aPLs, decreased serum factor H, and positive anti–ADAMTS13 antibodies ( Fig. 24-6 ). A study of eight children initially diagnosed with acquired TTP and reduced ADAMTS13 activity showed that SLE was concurrently or subsequently diagnosed in seven of eight patients during 42 months follow-up; all six patients tested for aPLs eventually developed positive aPLs, suggesting a potential association.
In general, published data demonstrate that clinical conditions with thrombotic microangiopathic hemolytic anemia are characterized by major endothelial dysfunction; the pathogenic role of aPLs remains controversial. It is assumed that the majority of these conditions do not form part of APS (with the exception of CAPS), but aPLs can be produced as an immune response to the exposure of phospholipids that occur with endothelial injury. Apparently aPLs represent only one of the pathogenic factors in these conditions, and various infectious causes, inherited defects in complement genes (e.g., complement factors H, I, B, and membrane cofactor protein), acquired autoantibodies against complement regulatory proteins, deficiency of vWF-cleaving protease (ADAMTS13), and inherited prothrombotic disorders must be taken into consideration, particularly in the pediatric population.
Perinatal Complications Associated with aPLs
The presence of maternal aPLs during pregnancy is associated with a number of serious obstetric and fetal complications, including preeclampsia, uteroplacental insufficiency, intrauterine growth restriction, fetal distress, premature birth, and fetal loss. With recent therapeutic approaches including low-dose aspirin alone or together with either unfractionated heparin or low molecular weight heparin (LMWH) throughout pregnancy, the percentage of pregnancies in women with aPLs that end in live births ranges from 75% to 80%. The prematurity rate in babies born to mothers with APS is 10% to 15%, and the incidence of growth restriction is 15% to 20%.
Perinatal thrombosis and other aPL-related clinical manifestations are rare complications of transplacental passage of aPLs in neonates. It is well established that maternal aCLs and anti-β 2 GPIs can cross the placenta and be detected in cord blood. The estimated rate of transplacental passage of maternal aPLs is 30% to 50%, significantly lower than that observed for antinuclear antibodies (approximately 80%). Several cohort studies examining the outcome of infants born to mothers with APS have consistently shown that, except for prematurity and its potential associated complications, these neonates rarely had other clinical manifestations. A recent report from the European multicenter registry of 134 children born to mothers with APS showed that only 2 neonates (1%) developed thrombocytopenia, but there were no cases of neonatal thrombosis.
In contrast to the encouraging results from the cohort studies, there are a growing number of isolated case reports of perinatal thrombotic events associated with transplacentally transferred aPLs. Analysis of 16 infants with thrombosis born to mothers with aPLs revealed arterial thrombotic events in 80%, and ischemic stroke occurred in approximately half of the cases ( Fig. 24-7 ). Moreover, in an Israeli study of 47 infants with perinatal arterial ischemic stroke, aPLs were detected in more than 20% of patients and were identified as a potential risk factor for stroke, together with factor V Leiden. Thromboses in neonatal APS were also reported in other vessels, including the aorta, peripheral arteries, mesenteric arteries, cerebral sinus veins, renal veins, and subclavian veins. More than 60% of infants with aPL-related thrombosis had at least one additional thrombophilic risk factor identified, most commonly arterial or venous catheters, sepsis, asphyxia, and/or congenital thrombophilia. In general, transplacentally transferred aPLs alone seem to be insufficient for induction of fetal or neonatal thrombosis, and other inherited and acquired thrombophilic risk factors should be systematically evaluated in case of perinatal thrombotic event in a child born to mother with aPLs.
Evidence of neurodevelopmental abnormalities has been described in the prospective long-term studies of children born to mothers with APS. An Italian study systematically evaluated neuropsychological development in 17 children born to mothers with APS and reported normal cognitive capacity, but with presence of learning difficulties such as dyslexia and dyscalculia in 4 children (24%) upon beginning school. In a recent French study, autism spectrum disorders were observed in 3 of 36 children(8%) born to mothers with primary APS, but in no child born to a mother with SLE. Among the group of 134 children included in the European registry of children born to mothers with APS, 4 children (3%) displayed behavioral abnormalities between 3 months and 3 years of age including autism, hyperactivity disorder, language delay, and learning disabilities. The mechanism of neurodevelopmental abnormalities in children born to mothers with APS is not clear and could be related to prematurity or in utero exposure to aPLs, just as prolonged exposure to aPLs induced behavioral and cognitive deficits in an experimental mice model. Regular psychomotor and cognitive assessments are recommended for long-term follow-up of these high-risk children.
Histopathological changes in APS may be grouped under several categories, including thrombotic, microangiopathic, ischemic, or coincidental with underlying disease–related pathology.
Classic vasoocclusive lesions in APS are thrombotic, recent, or organized, and can either be systemic or involve single organs. Thrombotic lesions are characterized by predominant noninflammatory occlusive or mural thrombosis and its consequences. Acute thrombotic lesions may be seen as occlusive thrombi with or without features of thrombolysis and limited reactive changes. Microvascular thrombi can undergo rapid dissolution and disappear. Otherwise, thrombus organization starts early by proliferation of neighboring endothelial cells and is later dominated by intraluminal proliferation of myofibroblasts and recanalization. Occlusive vascular thrombosis causes secondary acute and chronic ischemic changes. True vasculitis may occur as a coincidental finding that is causally not related to APS but to the coexisting underlying disease, most commonly SLE.
Microangiopathic lesions are characterized predominantly by endothelial cell injury–subendothelial plasma insudation often associated with thrombotic necrotizing lesions and its chronic consequences. Endothelial cells of the small blood vessels that are activated and injured by aPLs play a major role in the initiation and development of acute microangiopathic lesions. The acute phase is characterized by endothelial cell swelling, detachment, and necrosis, as well as prominent insudation into the subendothelium of blood constituents through a leaky endothelium, which may be seen by immunofluorescence microscopy as intensely positive staining for IgM with a lumpy pattern along the small-vessel walls. Chronic microangiopathic glomerular lesions are characterized by a double-contoured basement membrane in the thickened glomerular capillary walls without thrombosis and no association with inflammatory cell proliferation and exudation. In addition, chronic microangiopathic vascular lesions with obliterating fibrous intimal hyperplasia of arterioles and interlobular arteries may become evident.
Diagnosis and Differential Diagnosis
Children with APS will be treated by a wide range of clinical specialists, including pediatric rheumatologists, hematologists, neurologists, and others. A multidisciplinary approach to investigation and management is often appropriate. Testing for aPLs is usually not a first-line investigation and should be limited to those children with an indication as listed in Table 24-6 .
Given the spectrum of clinical manifestations, the differential diagnosis of APS is very broad and depends on target organ involvement. Characteristically, pathological thrombosis in children requires the presence of multiple risk factors to produce abnormal clotting ; therefore, clinical assessment of all children who have an aPL-related thrombotic event should include a search for additional congenital and acquired prothrombotic risk factors, and a thorough evaluation for evidence of underlying SLE or other systemic disease.
The main congenital prothrombotic states to be considered include factor V Leiden, prothrombin G20210A mutation, and deficiencies of antithrombin, protein C, and protein S. Factor V Leiden is the most common hereditary risk factor for venous thrombosis, and up to 5% of the Caucasian population carry this polymorphism. The G20210A mutation of the prothrombin gene is a common polymorphism associated with venous thrombosis, and its prevalence in Caucasians is 2% to 4%. Congenital deficiencies of antithrombin, protein C, and protein S are very uncommon, but acquired deficiency of protein S has been reported in children with APS, particularly in association with varicella infection. The evaluation for congenital prothrombotic disorders in children should also include testing of total cholesterol, triglycerides, lipoprotein(a), coagulation factor VIII, and fasting homocysteine concentration. It is noteworthy that the levels of antithrombin, protein C, and protein S may be transiently decreased because of consumption in the setting of acute thrombosis, whereas factor VIII and lipoprotein(a) can be elevated in inflammatory conditions. The most common acquired factors that may contribute to the risk of thrombosis are infection, malignancy, congenital heart disease, nephrotic syndrome, systemic vasculitis, central venous lines, surgery, and immobilization.
The differential diagnosis of nonthrombotic aPL-related clinical manifestations encompasses a variety of hematological, dermatological, neurological, and other diseases. The differentiation of isolated aPL-related thrombocytopenia from classic idiopathic thrombocytopenic purpura is important and indicates the need for closer follow-up regarding the increased risk of future thrombosis or progression to SLE. Neurological manifestations of aPLs should be distinguished from idiopathic neurological conditions, neuropsychiatric involvement in systemic autoimmune diseases, Sydenham chorea, multiple sclerosis, infections, intoxications, and other causes.
CAPS should be distinguished from severe SLE vasculitis, sepsis, HUS, TTP, heparin-induced thrombocytopenia, macrophage activation syndrome, and disseminated intravascular coagulation.
Antiphospholipid antibodies is an umbrella term used to describe a heterogeneous group of autoantibodies directed against negatively charged phospholipids or plasma phospholipid-binding proteins. In clinical practice, the most relevant aPLs for identifying patients at risk for immune-mediated thrombosis are aCLs, anti-β 2 GPIs, and LA. In a cohort of 121 children included in the pediatric APS registry, the presence of aCL was detected in 81% of cases, anti-β 2 GPI in 67%, and LA in 72%.
Persistent positivity of aPLs is of major importance for diagnosing APS, and all abnormal aPL values should be verified on at least two occasions at least 12 weeks apart, preferably at a time when the child has not had a recent infection.
aPLs can be detected by a variety of laboratory tests. The most sensitive test for aPLs is the aCL test, which uses enzyme-linked immunosorbent assay to determine antibody binding to solid plates coated either with cardiolipin or other phospholipids. The specificity of aCLs for APS increases with titer and is higher for the IgG isotype than for the IgM isotype. There have been numerous efforts to standardize the aCL test, but precise reproducible measurement of aCL levels is difficult; for clinical practice the use of semiquantitative measurements (low, medium, and high) is recommended. Anticardiolipin antibodies must be persistently present in medium or high titer to meet the definition of APS. The observation that many aCLs are directed at an epitope on β 2 GPI led to the development of anti-β 2 GPI immunoassays, which have improved specificities over the aCL test. Anti-β 2 GPIs have been reported to be associated primarily with thrombosis in patients with APS, particularly in patients with underlying SLE. The issue of the standardization of the anti-β 2 GPI immunoassay has also been the subject of considerable debate; however, despite these efforts, a considerable degree of interlaboratory variation has been reported. Furthermore, recent evidence suggests that IgG antibodies directed against domain I of β 2 GPI are more strongly associated with thrombosis than those detected using the standard anti-β 2 GPI assay. Inadequate data exist as to the clinical utility of other aPL assays for antibodies to prothrombin, phosphatidylethanolamine, anionic phospholipids other than cardiolipin (phosphatidylserine, phosphatidylinositol, and phosphatidic acid), IgA isotype of aCL and anti-β 2 GPI, and the annexin A5 resistance test.
The LA test is a functional assay measuring the ability of patient plasma containing aPLs to prolong in vitro phospholipid-dependent clotting reactions such as the aPTT and the RVVT. In vivo , however, the presence of LA is paradoxically associated with thrombotic events rather than with bleeding. LA are a mixture of different antibodies with various target antigens and also include aCL and anti-β 2 GPI, among others. According to the updated guidelines of the Scientific Standardization Committee of the International Society of Thrombosis and Haemostasis, the presence of LA should be confirmed with mixing tests with normal plasma and demonstration of the phospholipid-dependent nature of the inhibitor. LA is less frequently positive in APS and is thus regarded as a less sensitive but more specific test for detection of aPLs. The LA assay has been shown to correlate much better with the occurrence of thromboembolic events than the aCL or the anti-β 2 GPI assay and is considered to be the most important acquired risk factor for thrombosis.
Several assays may be required to confirm aPLs in some patients because they may be negative according to one test but positive according to another. Reliance on just one type of assay may lead to false negative aPL assessments. In general, only one third of pediatric APS patients are concurrently positive for three aPL subtypes (aCL, anti-β 2 GPI, and LA), whereas other patients are negative for at least one of the aPL subtypes. A full panel of current aPL tests, including aCL, anti-β 2 GPI, and LA, should be performed, and if possible, newer assays such as antiprothrombin, anti-β 2 GPI domain I, and annexin A5 resistance should be tested. Determination of complete aPL profile is also important for stratification of thrombosis risk and identification of high-risk patients with multiple (particularly triple) aPL positivity.
Measurement of the activation products of coagulation and fibrinolysis, such as d -dimer, prothrombin fragment 1 and 2, soluble fibrin, and thrombin-antithrombin complexes provides additional information on the hypercoagulability state in patients with inherited and acquired prothrombotic disorders such as APS. d -dimer was most extensively studied, and there is substantial evidence that it is a sensitive but nonspecific indicator of deep-vein thrombosis. Because of its high negative predictive value, it is particularly useful for the exclusion of deep-vein thrombosis and pulmonary embolism. Persistently elevated concentrations of d -dimer above 500 ng/mL and of coagulation factor VIII above 150 IU are associated with an increased risk of recurrent thromboembolism in children and adults.
Thrombosis in patients with APS must be confirmed by objective validated criteria. The ability to detect thrombosis in target organs in infants and children was markedly improved by the development of noninvasive imaging techniques using color-flow and pulsed Doppler ultrasound, echocardiography, computed tomography plus angiography (CT/CTA), and magnetic resonance imaging with or without angiography (MRI/MRA). The diagnosis of deep-vein thrombosis in the lower or upper extremity is usually established by compression and Doppler ultrasound, which can be easily performed in children ( Fig. 24-8 ). Echocardiography, CTA, or MRA can be used for thrombus imaging in the superior vena cava and proximal subclavian veins. MRA, CT pulmonary angiogram, or ventilation-perfusion scans are recommended for diagnosing suspected pulmonary embolism in children. Except during interventional procedures, venography and angiography are rarely used in children because of technical difficulties (peripheral venous access or arterial catheterization), the requirement for iodinated contrast, and the possibility of extending thrombus.