Chapter 131: The Antiphospholipid Syndrome

The antiphospholipid (aPL) antibody syndrome (APS) is a disorder in which vascular thrombosis or pregnancy complications attributable to placental insufficiency occur in patients with laboratory evidence for antibodies directed against proteins that bind to phospholipids. The syndrome was first proposed to be a distinct entity, “the anticardiolipin (aCL) syndrome,” in 1985¹ and soon was renamed APS.² While precise data are not available, the syndrome is thought to affect approximately 10 percent of patients with venous thrombosis,^3,4 and approximately 20 percent of women with three unexplained fetal losses before 12 weeks of gestation, or at least one intrauterine fetal death after 12 weeks of gestation.⁵

The term “aPL antibodies” can refer to (1) antibodies that recognize protein-phospholipid complexes as in cofactor-dependent aCL assays, (2) antibodies that recognize the proteins directly as in anti-β₂glycoprotein I assays (anti-β₂GPI), (3) an abnormal coagulation test in several assays that report inhibition of phospholipid-dependent coagulation reactions, collectively as termed lupus anticoagulant (LA) tests, and (4) antibodies that recognize phospholipid directly as in syphilis and are not associated with the APS disease entity.

A brief review of the history of APS^6,7,8 helps explain the confusing terminology (Table 131–1); the reader is referred to references 6,7,8 for more detailed accounts. In retrospect, the first assay for aPL autoantibodies was Moore and Mohr’s report of the “biologic false-positive” serologic tests for syphilis (BFP syphilis test) in 1952⁹; this abnormality came to be associated with systemic lupus erythematosus (SLE).¹⁰ The contemporaneous introduction, of the activated partial thromboplastin time (aPTT), which used cephalin, a phospholipid extract of animal brains, as the “partial thromboplastin” (distinct from the “complete thromboplastin,” tissue factor, and phospholipid),¹¹ led to the recognition of a unique type of anticoagulant in patients with SLE, that was frequently associated with BFP syphilis tests.¹² Because the of its initial association with SLE this phenomenon was misnamed LA.¹³ It became recognized that the LA was purely an in vitro phenomenon that was not limited to SLE and that was not associated with bleeding complications unless another hemostatic defect was present.⁶ Furthermore, the LA came to be associated with recurrent pregnancy losses^14,15 and with thrombotic and embolic manifestations.¹⁶ The development, in 1983, of the aCL antibody assay, that measured antibodies against the anionic phospholipid, cardiolipin (diphosphatidylglycerol), the primary antigen in the syphilis test reagent,¹⁷ was the advance that led to the identification of a new syndrome. Within a few years, it became recognized that aPL antibodies did not bind phospholipids directly but instead were directed against proteins that bound to the phospholipid, primarily β₂GPI (see “Pathogenesis” below). This information became important in helping to unravel the mechanisms for APS and in advancing diagnostic testing toward the goal of distinguishing between the syndrome and incidental false-positive tests. Table 131–2 describes the current consensus investigational criteria for diagnosing APS.¹⁸

Most patients with elevated aPL antibodies do not have the syndrome; elevated aPL antibody levels can occur in patients with several types of infections that induce formation of antibodies recognizing anionic phospholipids directly, patients taking medications such as chlorpromazine or procainamide, and even in normal healthy individuals. Testing of patients who have neither clinical manifestations of the disorder or SLE for aPL antibodies should be discouraged because it incurs the risk of inappropriate diagnostic and treatment decisions.

ETIOLOGY

As with most autoimmune conditions, the etiology of APS is not understood. It has been demonstrated that even normal healthy individuals have memory B cells that produce aPL antibodies. In a study of patients with infectious mononucleosis, 10 to 60 percent of immunoglobulin (Ig) M aPL-producing cells expressed CD27, the marker of memory B cells.¹⁹ The affinity of aPL antibodies for their target becomes increased by the inclusion of amino acids lysine, arginine, and asparagine within the complementary determining regions of the heavy and light chains.²⁰

Although antibodies against anionic phospholipid moieties arise during the course of infections such as syphilis and Lyme disease, those are distinct from antibodies generated by patients with the syndrome because they generally recognize phospholipid epitopes directly (also referred to as “cofactor independent”) and are not associated with the clinical manifestations of the syndrome. There are intriguing hints for molecular mimicry mechanisms and that infection and vaccination-induced APS could be related to autoimmune/autoinflammatory syndrome induced by adjuvants (ASIA).²¹ aPL antibodies have been reported in patients who developed thrombosis after varicella infection,^8,22,23 and in patients with hepatitis C.^24,25 aPL antibodies were reported in a patient with cytomegalovirus (CMV) infection, mesenteric and femoropopliteal thrombosis.^26,27 β₂GPI cofactor-dependent antibodies against cardiolipin, phosphatidyl serine, and phosphatidyl ethanolamine have been identified in sera from patients with parvovirus B19.²⁸ Bacterial infections are a predisposing risk factor for the catastrophic form of APS (CAPS).²⁹ A high proportion of HIV-1 patients have aPL antibodies; more than 40 percent in one study, in which 18 percent had aCL and 30 percent had anti-β₂GPI (mostly of the IgA isotype),³⁰; however, positivity for these antibodies was not associated with thrombosis. A link has been proposed between the cardiac valvular disease in acute rheumatic fever and the presence of aPL antibodies.³¹ aCL antibodies having β₂GPI dependence and LA activity have been generated in rabbits immunized with lipid A and lipoteichoic acid, suggesting that some bacteria can contribute to the production of pathogenic aPL antibodies.³² It has also been proposed that cellular apoptosis, with the resulting exposure of anionic phospholipids on cell surfaces, may trigger the generation of aPL antibodies.^33,34,35 Molecular mimicry between β₂GPI-related synthetic peptides and structures within bacteria, viruses, and tetanus toxoid³⁶ have been demonstrated in an experimental model for APS.³⁷ Mice immunized with a CMV-derived peptide developed aPL antibodies and thrombosis, and showed evidence for endothelial cell activation.³⁸

Reports of familial clustering of raised aPL antibody levels³⁹ indicate that genetic susceptibility can play a role in their development. In one study of 84 APS patients, more than 35 percent had at least one relative, and more than 20 percent had two or more relatives, with evidence of at least one clinical feature of APS, such as thrombosis or recurrent fetal loss.⁴⁰

PATHOGENESIS

Experimental Evidence That Antiphospholipid Antibodies Are Pathogenic

It has been clearly established in a number of experimental animal models for APS that aPL antibodies play a causal role in the development of thrombosis and pregnancy loss.^{41,42,43,44,45} Although it is reasonable to assume that the same holds for the human disorder, the epitopic specificities of the autoantibodies that cause disease and the mechanisms by which they produce clinical manifestations require further elucidation.

Antigenic Specificities

Antibodies against phospholipid that arise during the immunologic response to syphilis and other infections (with the notable exception of leprosy⁴⁶) recognize anionic phospholipid epitopes directly,⁴⁷ whereas pathogenic aPL antibodies recognize phospholipid-binding proteins, primarily β₂GPI.^48,49

β₂GPI (also named apolipoprotein H), a member of the complement control protein or short consensus repeat superfamily,⁵⁰ is a highly glycosylated single-chain plasma protein composed of 326 amino acids, with a molecular weight of approximately 50 kDa (Fig. 131–1). β₂GPI has five short consensus repeat (SCR) stretches of approximately 60 amino acids⁴⁵ (also referred to as complement control protein [CCP] repeats). Epitopic specificities for individual domains may have pathogenic and prognostic significance (see “Immunoassays” below).^51,52,53,54

The affinity of β₂GPI for anionic phospholipids derives from cationic residues from its aminoterminus that have affinity for anionic polar heads of phospholipids and a hydrophobic loop which inserts into the lipid bilayer. β₂GPI has five domains for which antibodies have been identified. IgG antibodies against an epitope comprising Gly40-Arg43 in the domain I of β₂GPI have been reported to have a stronger correlation with thrombosis than antibodies against other epitopes.⁵¹ Recent data support the concept that β₂GPI undergoes conformational changes that may be important for the APS disease process. By transmission electron microscopy, unbound β₂GPI appears to be in a closed conformation because of the affinity of a portion of carboxyterminal domain V for the protein’s amninoterminal domain I, where the phospholipid binding site is located near the carboxy-terminus of SCR domain V (see Fig. 131–1).⁵⁵ The binding of β₂GPI to anionic phospholipid membranes, requires a conformational change which exposes an epitope in domain I that had been cryptic in the unbound conformation (see Fig. 131–1).⁵⁵

Although β₂GPI bind to phospholipids, its role in aPL-mediated cell signaling (described in the section Proposed Pathogenic Mechanisms), is mediated via binding to toll-like receptors (TLRs) and not by direct binding to the lipid bilayer.

While the in vivo biologic function(s) of the protein has (have) not been defined, several interesting properties have been demonstrated. The molecule binds to apoptotic cells,⁵⁶ and may play a role in their phagocytosis and clearance.⁵⁷ β₂GPI binds to oxidized low-density lipoprotein (LDL) and may play a role in its clearance.⁵⁸ β₂GPI binds to lipopolysaccharide and the scavenged complex is taken up by monocytes/macrophages.⁵⁹ β₂GPI reduces platelet adhesion to collagen in flow chambers by interfering with the platelet–von Willebrand factor (VWF) interaction by binding to its A2 domain, thereby interfering with its binding to the platelet glycoprotein Ib complex.⁶⁰ β₂GPI may also promote fibrinolysis as a cofactor for tissue-type plasminogen activator (t-PA) via its SCR domain V, which increases fibrinolytic activity.⁶¹ The protein may have a further effect on fibrinolysis by binding to endothelial cells via annexin A2, a protein that also serves as a receptor for plasminogen and t-PA.⁶² Homozygous β₂GPI-null mice have not been demonstrated to display a thrombotic phenotype.⁶³ However, the protein may play a role—though not a critical one—in the reproductive process, as there was a reduction in the number of viable implantation sites in β₂GPI-null mice and reduced fetal weight and fetal-to-placental weight ratio in late gestation, suggesting compromised placental function.⁶⁴

In addition to β₂GPI, a number of other antigenic targets have been identified for aPL, including, but not limited to, prothrombin, coagulation factor V, protein C, protein S, annexin A2, annexin A5, high- and low-molecular-weight kininogens, and factors VII/VIIa and vimentin–cardiolipin complex.^65,66,67,68

Proposed Pathogenic Mechanisms

Table 131–3 and Fig. 131–2 summarize several of the main current hypotheses for pathogenic mechanisms in APS. The mechanisms of the human APS disease process have been difficult to elucidation, mainly for two reasons: (1) The phenotypes of vascular thrombosis and pregnancy morbidity are not unique to APS, so it is difficult to ascertain whether the candidate mechanism is playing a causal role or is incidental. (2) Antibodies isolated from APS patients recognize a multiplicity of antigenic determinants^69,70 that can have a broad range of effects, so it is difficult to determine which specificities and effects are responsible for disease manifestations in humans.

Disruption of the Endothelial Surface and Annexin A5 Anticoagulant Shield Annexin A5 is a potent anticoagulant protein with high affinity for phospholipid membranes that contain anionic phospholipids, specifically phosphatidyl serine.⁷¹ Annexin A5 forms two-dimensional crystals over the phospholipid bilayers that shield them from binding coagulation factors⁷² and it has been proposed that the protein may play a thrombomodulatory role on the surfaces of cells lining the placental and systemic vasculatures. Annexin A5 is highly expressed on the apical membranes of placental syncytiotrophoblasts, the location where maternal blood interfaces with fetal cells.⁷³ Pregnant annexin A5-null mice develop placental infarctions of fetuses and yield reduced litter sizes.⁷⁴ Pregnant mice treated with anti–annexin A5 antibodies developed placental necrosis, fibrosis, and pregnancy loss.⁷⁵ Dissociation of annexin A5 from the surface of human placental trophoblasts and human umbilical vein endothelial cells accelerates the coagulation of plasma exposed to those cells.⁷⁶ Annexin A5 binds to the surfaces of endothelial cells and inhibits thrombin formation.⁷⁷

aPL antibody–antigen complexes disrupt the crystallization of annexin A5 and displace the protein from phospholipid membrane surfaces (Fig. 131–3).^78,79,80,81 In contrast to the LA phenomenon, aPL antibodies accelerate coagulation reactions in systems that include annexin A5.^{78,82,83,84,85} IgG fractions from APS patients reduce the quantity of annexin A5 on cultured placental trophoblasts^76,86 and endothelial cells^76,87 and accelerate the coagulation of plasma exposed to these cells.⁷⁶ This effect of aPL antibodies on annexin A5 binding has been correlated with IgG antibodies that recognize a specific epitope—domain I of β₂GPI in patients with APS who have thrombosis⁵² and spontaneous pregnancy losses. Figure 131–2 includes a model for this mechanism.

Binding to Endothelial Surface Receptors Enhances Cell Signaling aPL antibodies can bind, injure, and activate cultured vascular endothelial cells.^88,89,90,91 Cultured endothelial cells incubated with aPL antibodies with specificity for cell surface β₂GPI express increased levels of cell adhesion molecules⁹² triggered by their binding to cell surface β₂GPI.⁹³ Annexin A2 serves as a receptor for β₂GPI,⁶² and anti-β₂GPI antibodies may thereby stimulate expression of tissue factor on endothelial cells.⁹⁴ In animal models, the signaling effects of aPL antibodies were significantly reduced in mice treated with an anti-annexin A2 monoclonal antibody and also in annexin A2-null transgenic mice.⁹⁵ These effects of the aPL antibodies may be mediated by TLR-4 of the innate immunity system,^96,97 although there are data indicating participation of other TLRs, particularly TLR-2.⁹⁸ This binding results in downstream signaling that involves TRAF6 (tumor necrosis factor receptor-associated factor 6) and MyD88 (myeloid differentiation factor 88).⁹⁹ Increased expression of tissue factor is mediated by p38 mitogen-activated protein kinase (MAPK).¹⁰⁰

Binding of autoantibodies to annexin A2 may also promote thrombosis by inhibiting fibrinolysis. APS patients have increased titers of antibodies against annexin A2, an endothelial surface receptor for t-PA, and plasminogen.¹⁰¹ The blocking of annexin A2 by aPL antibodies impedes plasmin generation in a t-PA–dependent generation assay, and inhibits cell surface plasmin generation on human umbilical vein endothelial cells.⁹⁴ Several additional mechanisms have been identified by which aPL antibodies can interfere with fibrinolysis. β₂GPI is a cofactor for t-PA–mediated activation of plasminogen, and aPL antibodies against β₂GPI interfere with its binding to t-PA, thereby downregulating plasminogen activation.⁶¹ Finally, fibrinolysis may also be impaired by autoantibodies directed against the catalytic site of plasmin or t-PA,^102,103 by an increased level of plasminogen activator inhibitor-1,¹⁰⁴ and by inhibition of autoactivation of factor XII with ensuing reductions of kallikrein and urokinase.¹⁰⁵

Apolipoprotein E receptor 2 (apoER2), a member of the LDL-receptor family is found on endothelial surfaces,¹⁰⁶ monocytes,¹⁰⁷ and platelets, and may also serve as receptor for anti-β₂GPI–β₂GPI complexes where it can also trigger the phosphatidylinositol 3′-kinase (PI3K)/Akt pathway¹⁰⁸ and increase tissue factor and cell adhesion molecule expression. IgG-mediated dimerization of β₂GPI and binding to ApoER2′ increases the sensitivity of platelets to agonists of aggregation.¹⁰⁹

Complement-Mediated Injury Complement activation may play a role in the APS disease process. The IgG₂ subtype of aPL most closely correlates with thrombosis.^110,111 Blockade of complement activation using a C3 convertase inhibitor or genetic deletion of C3 protected mice from pregnancy complications induced by aPL antibodies.^112,113,114 These effects involve the aPL-stimulated expression of tissue factor by myeloid cells,¹¹⁵ and to involve proteinase-activated G-protein–coupled receptor (PAR)-2 signaling,¹¹⁶ indicating that complement activation can be pathogenic via both direct injury and downstream signaling.

Induction of Tissue Factor Activity in Leukocytes aPL antibodies can promote tissue factor expression by leukocytes.^{115,117,118,119} The specific binding site on these cells has not been elucidated.

Inhibition of Fibrinolysis and Endogenous Anticoagulation aPL antibodies can interfere with fibrinolysis in several ways. Antibodies against annexin A2, an endothelial surface receptor for t-PA and plasminogen, can interfere with binding of plasminogen and t-PA and thereby reduce plasmin formation and fibrinolysis.^94,101,103 Monoclonal aPL antibodies derived from APS patients can directly inhibit plasmin’s enzymatic activity.¹⁰² β₂GPI is a cofactor for t-PA–mediated activation of plasminogen,¹²⁰ and anti-β₂GPI can interfere with this activity. Also, it has been reported that women with APS have significantly increased levels of circulating plasminogen activator inhibitor-1 (PAI-1), implying impaired fibrinolysis.¹⁰⁴

Interference with Components of the Protein C Activation Pathway The protein C pathway (Chaps. 114 and 139) is initiated by thrombin binding to thrombomodulin (TM), which activates protein C bound to the endothelial protein C receptor (EPCR). Activated protein C (APC), together with free protein S, then proteolyses coagulation factors Va and VIIIa. APC also modulates signaling events by interfering with PAR-1 signalling.^121,122 aPL antibodies can interfere with the activation of protein C by TM–thrombin and with the activity of APC, as well as protect factors Va and VIIIa from proteolysis by APC.⁶⁵ Acquired APC resistance has been described in APS plasmas¹²³ and has been correlated with anti-β₂GPI domain I antibodies,¹²⁴ a risk factor for thrombosis. The presence of antibodies against EPCR in APS patients was proposed to be a risk factor for fetal death.¹²⁵

Antiphospholipid Antibodies Activate Platelets An experimental animal model that includes in vivo imaging has provided data indicating that aPL-induced thrombosis is a consequence of platelet activation that then promotes endothelial activation and fibrin formation.¹²⁶ aPL antibodies can induce platelet aggregation,¹²⁷ an effect that might be promoted via signaling through apoER2 receptors; the β₂GPI binding site for apoER2 on platelets was localized to its domain V.¹²⁸ As described above (see “Antigenic Specificities”), β₂GPI also has a dampening effect on platelet adhesion by interfering with the platelet–VWF interaction, and consequently aPL antibodies, by interfering with this β₂GPI-mediated dampening, can increase platelet adhesion in flow systems.⁶⁰

Other Mechanisms APS patients have been shown to have autoantibodies against tissue factor pathway inhibitor.¹²⁹ Some aPL antibodies cross-react with heparin and heparinoid molecules, which are highly polyanionic, and hence, inhibit their contribution to antithrombin activity.⁶⁹ aPL antibodies show cross-reactivity against oxidized LDL¹³⁰ and are associated with an increased risk of atherosclerosis.¹³¹ Antibodies against β₂GPI-oxidized LDL complexes have been proposed to be atherogenic by reducing their clearance.¹³² Finally, in addition to promoting thrombosis, aPL antibodies may contribute to other vascular lesions by stimulating the mammalian target of rapamycin complex (mTORC) pathway (Fig. 131–4).¹³³

Table 131–4 summarizes the clinical features of APS. Patients generally present with thrombotic manifestations, that is, evidence for vasoocclusion or end-organ ischemia or infarction, and/or pregnancy losses and complications attributable to placental insufficiency. The usual age at presentation with thrombosis is approximately 35 to 45 years. Except for patients with SLE, men and women are equally susceptible to thrombotic manifestations. No differences have been observed between the arterial and venous distributions of thromboses of primary and secondary APS patients.¹³⁴

SYSTEMIC VASCULAR THROMBOSIS

Patients can present with spontaneous venous and/or arterial thrombosis or embolism in any site; however, about half of all patients have deep vein thrombosis of the lower extremities.^135,136 Other sites of venous thromboembolic events include pulmonary embolism, thoracic veins (superior vena cava, subclavian vein, or jugular vein), and abdominal or pelvic veins.¹³⁶ Approximately one-fourth of patients present with arterial thromboses; the remainder present with concurrent arterial and venous thrombosis.¹³⁶ Patients may also present with stroke, cerebral venous thrombosis, upper-extremity venous thrombosis,¹³⁵ myocardial infarction, adrenal infarction, acalculous gallbladder infarction, aortic thrombosis with renal infarction,¹²⁰ and mesenteric artery thrombosis.^137,138 Thrombosis may occur spontaneously or in the presence of some other risk factor such as estrogen replacement therapy, oral contraceptives,^134,139 vascular stasis, surgery, or trauma. Women are at particularly high risk for venous thrombosis during pregnancy and in the postpartum period.¹³⁴ Some APS patients with venous thrombosis have concurrent genetic thrombophilic conditions such as the factor V Leiden variant, and it has been postulated that this may increase the risk of thrombosis.^{140,141,142,143} Concurrent positivity for all three aPL antibody assays—that is, the aCL antibody assay, anti-β₂GPI antibody assay, and the LA assay—appears to be a highly significant risk factor for having an initial thrombotic event (Fig. 131–5).¹⁴⁴ If this finding is confirmed, it could identify a group of patients who might warrant consideration for prophylactic treatment. A history of having had a thromboembolic event is probably the most significant risk factor for recurrence of venous thromboembolism in APS, with a reported frequency that reached approximately 30 percent in patients followed for 4 years after the first episode.¹⁴⁵ The risk of recurrent thromboembolism correlates with the titer of antibodies,^145,146 and with the presence of LA. In addition, it appears that the presence of anti-β₂GPI domain I antibodies significantly increase the risk of thrombosis, compared to patients who have antibodies that are not domain I dependent and LA that is not domain I dependent.⁵¹

SYSTEMIC LUPUS ERYTHEMATOSUS AND OTHER AUTOIMMUNE CONDITIONS

APS patients frequently present with other autoimmune conditions. A significant proportion of SLE patients have elevated aPL antibodies, with estimates ranging between 12 and 30 percent for aCL antibodies and 15 and 34 percent for LA antibodies.¹⁴⁷ APS has been associated with the concurrence of other autoimmune conditions including, but not limited to, rheumatoid arthritis,¹⁴⁸ Sjögren syndrome,¹⁴⁹ myasthenia gravis,¹⁵⁰ Budd-Chiari syndrome in the setting of SLE,¹⁵¹ Graves disease,¹⁵² autoimmune hemolytic anemia, progressive systemic sclerosis,¹⁵³ Evans syndrome,¹⁵⁴ Takayasu arteritis,¹⁵⁵ polyarteritis nodosa,¹⁵⁶ and immune thrombocytopenia (see section Thrombocytopenia below and also Chap. 117).

STROKE AND OTHER NEUROLOGIC CONDITIONS

The most common neurologic manifestations of APS are stroke or transient ischemic attacks (TIAs), which were the initial presentation of up to 30 percent of adults with APS in a large European cohort.¹⁵⁷ In the European Catastrophic Antiphospholipid Antibody Syndrome (CAPS) Registry, cerebral manifestations occurred in 62 percent of patients and caused 13 percent of deaths.¹⁵⁸ Recurrent strokes are more likely in patients with APS and other risk factors for cerebrovascular disease, such as cigarette smoking, hypertension hyperlipidemia, oral contraceptive use, and SLE.^159,160

Most APS patients with stroke have arterial thromboembolic occlusive events that are clinically indistinguishable from the more common arteriosclerotic strokes. APS should be suspected in young patients with TIAs or stroke, particularly when the more typical risk factors for cerebrovascular disease are absent.¹⁶¹ Cerebral venous thrombosis is less common in APS patients and presents at a younger age, and is more extensive than in non-APS patients with the disorder.¹⁶² In one series of 40 cases of cerebral venous thrombosis, three patients (8 percent) had elevated aPL antibody levels.¹⁶³ Superior sagittal sinus thrombosis has been reported with primary APS.¹⁶⁴

There is controversy about whether migraines in patients with aPL antibodies should be regarded as thromboocclusive events.¹⁶⁵ Other neurologic abnormalities reported to be associated with aPL antibodies include seizures/epilepsy,¹⁶⁶ chorea, Guillain-Barré syndrome, transient global amnesia, dementia, diabetic peripheral neuropathy, and orthostatic hypotension.¹⁶⁷ Recurrent acute transverse myelopathy has been described with APS.^{168,169,170,171,172} However, in one study of 315 SLE patients, including 10 with a history of transverse myelopathy, that disorder was not associated with aPL antibodies.¹⁷³ Multiple sclerosis patients have a high incidence of elevated aCL antibody levels (in one series, 9 percent had IgG antibodies and 44 percent had IgM antibodies),¹⁷⁴ however, no clinical distinctions were found between aPL-positive and aPL-negative patients and the antibodies do not appear to be associated with thrombosis. Patients with psychotic disorders have an increased prevalence of LA and aCL antibodies, even in the absence of treatment with antipsychotic drugs.¹⁷⁵

CATASTROPHIC ANTIPHOSPHOLIPID SYNDROME

CAPS is a relatively infrequent but devastating presentation of APS, is characterized by severe widespread vascular occlusions.¹⁷⁶ Diagnostic criteria for CAPS include evidence of involvement of at least three organs, systems, and/or tissues; development of manifestations simultaneously or in less than 1 week; histopathologic confirmation of small-vessel occlusion; and laboratory confirmation of the presence of aPL antibodies.¹⁷⁷ According to the CAPS Registry, a web-based database of 433 patients with CAPS (https://ontocrf.costaisa.com/en/web/caps/), the majority of CAPS patients are female (69 percent), in their late thirties (mean age of 38.5 years), but the condition can present at any age (range: 0 to 85 years). In half of the CAPS cases, the patients’ catastrophic event was their first APS manifestation. Precipitating factors of CAPS include infections, drugs (sulfur-containing diuretics, captopril, and oral contraceptives), surgical procedures, and cessation of prior anticoagulant therapy. In 26.9 percent of cases, the patients also had SLE. The most frequently affected organ was the kidney (73 percent of episodes), followed by lungs (58.9 percent), the brain (55.9 percent), the heart (49.7 percent), and the skin (45.4 percent). Other organs were also affected, including the peripheral vessels, intestines, spleen, adrenal glands, pancreas, retina, and marrow. Patients present with evidence for severe multiorgan ischemia/infarction, often with concurrent disseminated microvascular thrombosis. Patients with CAPS can present with renal insufficiency, respiratory failure resulting from acute respiratory distress syndrome (ARDS) and pulmonary emboli, cerebral manifestations such as encephalopathy, stroke and seizures, cardiac problems such as heart failure, myocardial infarction and valvular defects, and skin complications such as livedo reticularis and skin necrosis.¹⁷⁸ Most patients show histologic evidence of microangiopathy mainly affecting small vessels of the kidneys, lungs, brain, heart, and liver. Only a minority of patients experience large-vessel occlusions. Laboratory evidence for disseminated intravascular coagulation is frequently present. An LA is present in 81.7 percent of patients, and the aCL IgG is the most common positive aPL antibody.¹⁷⁸ Improved treatment has reduced mortality from approximately 50 percent to approximately 20 percent.¹⁷⁶ Relapse is rare in survivors. The only identified predictive factor for adverse outcome is underlying SLE.¹⁷⁹

PREGNANCY LOSSES, OBSTETRIC COMPLICATIONS, AND INFERTILITY

At this time, aPL screening of otherwise asymptomatic with no history of prior complications obstetrical patients is not warranted because of the high frequency of false-positive tests; most studies have estimated the prevalence of aPL antibodies among general obstetric populations to be approximately 5 percent or less and most of these aPL-positive patients are not clinically affected.¹⁸⁰

Among obstetric patients with recurrent fetal losses, approximately 16 to 38 percent have aPL antibodies. In addition, pregnant women with elevated aPL antibodies had significantly more obstetric complications, including preeclampsia, abruption placentae, miscarriage, prematurity, intrauterine fetal demise, intrauterine growth restriction, and oligohydraminos, than aPL antibody-negative pregnant women.^181,182,183 In approximately half of patients, the pregnancy losses occur in the first trimester. Other patients present with later losses, most in the second trimester, but some even later, including stillbirth. Pregnancy complications attributable to APS include three or more recurrent spontaneous first trimester miscarriages, one or more fetal losses during the second trimester, stillbirth, episode of preeclampsia, preterm labor, placental abruption, intrauterine growth restriction, and oligohydramnion.^184,185,186 Pregnant patients with APS are also more prone to developing deep vein thrombosis during pregnancy or the puerperium. Rarely, pregnant patients develop CAPS.^187,188 The best predictor for pregnancy loss in a patient who tests positive for aPL antibodies has been demonstrated to simply be a previous history of pregnancy loss, complications, or thrombosis.^146,189 A recent study reported that any aPL antibody-positive women with an unexplained early loss prior to 10 weeks have a higher risk of complications in their second pregnancy¹⁹⁰ but the degree of laboratory abnormalities were not associated with increased risk. Positivity by more than one assay appears to correlate with increased pregnancy morbidities.^182,191 Histologic abnormalities were found in many, but not all, placentas of aPL patients.¹⁹² Studies of placental pathology in patients with aPL antibodies, but without a prior history of fetal loss, showed that approximately half had evidence of uteroplacental vascular pathology, approximately half had evidence of thrombotic occlusion, and approximately one-third had chronic villitis and/or decidual plasma cell infiltrates.^193,194

Overall, it appears that the presence of aPL antibodies does not affect the success rate of implantation. Although one group has reported data that indicated that women with recurrent implantation failure were more likely to have positive assays for aPL antibodies compared to fertile negative controls,¹⁹⁵ a review of 29 studies showed mixed results.¹⁹⁶ Many of the studies were noted to have limitations, including problems with study design and statistical power. The current consensus is that aPL antibodies are not a cause of infertility.¹⁹⁷ Recently, the14th International Congress on Antiphospholipid Antibodies Task Force also concluded that “there are no data to support the inclusion of infertility as criteria for APS and investigation of APS in patients with infertility should not be done in routine clinical practice, being reserved only for research purposes.”¹⁹⁶

CUTANEOUS MANIFESTATIONS

The cutaneous manifestations of APS may comprise the first signs of APS in some patients.^198,199 Livedo reticularis is relatively common, occurring in 24 percent of a series of 1000 aPL patients,²⁰⁰ and occasionally presents in a necrosing form.²⁰¹ Noninflammatory vascular thrombosis is the most frequent histopathologic feature. Necrotizing vasculitis, livedoid vasculitis, thrombophlebitis, cutaneous ulceration and necrosis, erythematous macules, purpura, ecchymoses, painful skin nodules, and subungual splinter hemorrhages, anetoderma (macular atrophy), discoid lupus erythematosus, and cutaneous T-cell lymphoma have all been reported.

CORONARY ARTERY DISEASE

aPL antibodies are associated with increased susceptibility to coronary artery disease,²⁰² particularly premature atherosclerosis.^203,204 APS should be considered in patients who lack the usual risk factors for coronary artery disease and in patients with evidence for thrombotic or embolic coronary artery occlusion that lack angiographic evidence of atherosclerotic disease. aPL antibodies appear to be a risk factor for adverse outcomes following all coronary revascularization procedures,²⁰² and for restenosis after percutaneous transluminal coronary angioplasty.^205,206 An ultrasound study of carotid arteries provided evidence supporting an association of aPL antibodies with premature atherosclerosis; relatively young primary APS patients (mean age: 37 ± 11 years) had significantly increased intimal medial thickness compared to control non-APS groups.²⁰⁷

VALVULAR HEART DISEASE

Approximately 35 percent of patients with primary APS have cardiac valvular abnormalities detected by echocardiography.²⁰⁸ In one study, approximately 20 percent of cardiac patients with valvular heart disease had evidence for aPL antibodies compared with approximately 10 percent of matched control subjects.²⁰⁹ Valvulopathy includes leaflet thickening, vegetations, regurgitation, and stenosis.²¹⁰ The mitral valve is mainly affected, followed by the aortic valve.²¹¹ Histologically, APS valvular lesions consist mainly of superficial or intravalvular fibrin deposits in association with variable degrees of vascular proliferation, fibroblast influx, fibrosis, and calcification. These can result in valve thickening, fusion, and rigidity, and lead to functional abnormalities.²¹² Deposits of immunoglobulins, including aCL antibodies, and of complement components are commonly found in the affected valves.²¹³

PERIPHERAL VASCULAR DISEASE

Approximately one-third of patients with peripheral arterial disease undergoing bypass grafting procedures had elevated aPL antibody levels (mostly aCL antibodies).²¹⁴ Intraarterial thromboembolic events are common at presentation of these patients and may complicate surgical management. However, these patients did not have an increased risk for reocclusion, a finding that was attributable to the use of anticoagulant therapy.

PULMONARY MANIFESTATIONS

Patients with APS may present with in situ thrombosis in pulmonary vessels. aPL antibodies are associated with pulmonary hypertension.²¹⁵ In one prospective trial of 38 consecutive patients with precapillary pulmonary hypertension, approximately 30 percent had aPL antibodies with various phospholipid specificities.²¹⁶ An interinstitutional study of 687 patients with chronic thromboembolic pulmonary hypertension reported that aPL antibodies were a significant risk factor.²¹⁷ The majority of patients with CAPS (see “Catastrophic Antiphospholipid Syndrome” above) have dyspnea, and most of these individuals have ARDS.²¹⁸

ABDOMINAL MANIFESTATIONS

The liver is the most frequently affected abdominal organ in APS, with occlusion of hepatic vessels, including those supplying the biliary tree.²¹⁹ aPL antibody levels frequently are elevated in patients with chronic liver disease of various causes. In one prospective study of patients with liver disease, approximately half of patients with alcoholic liver disease and one-third of patients with chronic hepatitis C virus had elevated aPL antibody levels. The frequency was even higher in patients with more severe cirrhosis.²²⁰ A review reported that approximately 20 percent of patients with chronic hepatitis B and hepatitis C had aPL antibodies, most of which were cofactor independent.²²¹ Some patients with hepatitis C present with true autoimmune aPL antibodies, the most common features reported being intraabdominal thrombosis and myocardial infarction.²²²

Reported gastrointestinal manifestations of APS also include esophageal necrosis with perforation, intestinal ischemia and infarction, pancreatitis, and colonic ulceration. Primary biliary cirrhosis,²²³ acute acalculous cholecystitis with gallbladder necrosis,^224,225 and giant gastric ulceration are associated with APS.²²⁶ APS has been reported in patients with mesenteric inflammatory venoocclusive disease²²⁷ and in patients with mesenteric and portal venous obstruction.²²⁸

THROMBOCYTOPENIA

Approximately 20 to 40 percent of patients with APS have varying degrees of thrombocytopenia. The decrease in platelet count is generally mild or moderate and is rarely significant enough to cause bleeding complications or affect anticoagulant therapy.^229,230 The majority of patients with APS and thrombocytopenia have antibodies against α_IIbβ₃ integrin and/or glycoprotein Ib-IX complex.²³¹ Patients presenting with immune thrombocytopenic purpura frequently have elevated aPL antibodies, and these patients are more prone to thrombosis.²³² aPL antibodies and antibodies against platelet membrane glycoprotein were present simultaneously in approximately 70 percent of patients with immune-mediated thrombocytopenia.²³³ Thrombocytopenia itself is not protective against thrombosis in these patients. In a prospective cohort study, 5-year thrombosis-free survival of aPL-positive and aPL-negative immune thrombocytopenic purpura patients were 39 percent and 98 percent, respectively.²³⁴

BLEEDING

The presence of a concurrent hemostasis defect needs to be considered when patients with APS exhibit a bleeding tendency (Table 131–5). Acquired hypoprothrombinemia with severe bleeding has been reported.^235,236 This diagnosis may be missed when coagulation abnormalities are attributed only to the LA effect, so a specific assay for prothrombin should be performed when the prothrombin time is prolonged. Other causes of bleeding in APS include acquired thrombocytopathies, thrombocytopenia (see “Thrombocytopenia,” above), acquired inhibitors against specific coagulation factors, such as factor VIII, and the acquired von Willebrand syndrome (AVWS).

RETINAL ABNORMALITIES

The diagnosis of aPL antibody retinopathy should be suspected in patients with diffuse retinal vasoocclusion, particularly when characterized by involvement of arteries and veins, neovascularization at presentation, and symptoms of systemic rheumatologic disease.²³⁷ aPL antibodies were present in 5 to 33 percent of patients with retinal vein occlusion.^238,239 Cilioretinal artery occlusion,²⁴⁰ optic neuropathy,²⁴¹ and severe vasoocclusive retinopathy²⁴² have been described with APS.

KIDNEY DISEASES

APS may affect the renal system. Patients may present with renal artery stenosis and/or thrombosis, renal infarction, renal vein thrombosis, and glomerulonephritis that is distinct from vasoocclusive disease.^200,243 An entity named “APS nephropathy” has been described, which consists of a vasoocclusive disease of small-size intrarenal vessels.²⁴⁴ This nephropathy features fibrous intimal hyperplasia, focal cortical atrophy, and thrombotic microangiopathy. A review of 29 consecutive renal biopsies from patients with primary APS, performed at two institutions over 22 years, described 20 cases of APS nephropathy and nine cases with other distinct pathologic features.²⁴³ These features included membranous nephropathy, minimal change disease/focal segmental glomerulonephritis, mesangial C3 nephropathy, and pauci-immune crescentic glomerulonephritis.

ANTIPHOSPHOLIPID SYNDROME AND AIDS

Although patients with HIV-1 infection frequently have elevated aPL antibody levels, they do not often have thrombotic manifestations. A review indicated that approximately 50 percent of HIV-1 patients test positive for aPL antibodies, most of which are not cofactor dependent.²²¹ HIV-infected patients with manifestations of APS have also presented with avascular bone and cutaneous necrosis.²²²

ANTIPHOSPHOLIPID SYNDROME IN CHILDREN

Increasingly, APS has become recognized in children,²⁴⁵ in whom diverse clinical features are common. The results of a European registry have been reported.²⁴⁶ Review of 121 cases indicated that although the thrombotic manifestations were similar to adults with APS, there was a difference between children with primary APS and the secondary APS; the children with primary APS were younger and had a higher frequency of arterial thrombotic events, whereas the children with secondary APS patients had a higher frequency of venous thrombotic events associated with hematologic and skin manifestations. CAPS has been reported in children, but is rare.^247,248

Thrombosis is rare in newborns delivered from mothers with APS, and only a few cases are reported, mostly associated with other prothrombotic factors.²⁴⁹ aPL antibodies have been found in up to 30 percent of offspring of mothers with APS.²⁵⁰

In a recent European prospective study, 17 percent of neonates born to APS mothers were premature; however, no specific complications were found during the 5-year followup.²⁵¹ The study did show a higher rate of neurodevelopmental abnormalities with learning disabilities similar to two retrospective reports where learning disabilities without other neurodevelopmental abnormalities were present in 15 to 20 percent of cases.^252,253

OTHER MANIFESTATIONS

Acute adrenal failure secondary to bilateral infarction of the adrenal glands has been reported as the first manifestation of primary APS.²⁵⁴ Adrenal hemorrhage has been reported.²⁵⁵ aPL antibodies have been associated with marrow necrosis.²⁵⁶

Diagnosis of APS requires the demonstration of antibodies against phospholipids and/or relevant protein cofactors (Table 131–6). The current tests for APS recommended by the most recent consensus on investigational criteria¹⁸ and the Scientific Standardization Committee of the International Society of Thrombosis and Hemostasis are aCL, anti-β₂GPI (IgG and IgM), and LA.²⁵⁷ The laboratory diagnosis of APS frequently is problematic, with limitations that have been detailed.^258,259 aCL IgG and IgM assays are the most sensitive, but the least specific. Anti-β₂GPI IgG and IgM assays are more specific but less sensitive. LA assays, of which the dilute Russell viper venom time (dRVVT) is the most common, generally tend to be the least sensitive but the most specific. The current recommended tests are less than ideal because they are empirically derived tests and do not yet measure antibodies directed against disease specific epitopes or functional parameters that correlate with disease mechanisms. Despite these limitations, positivity for all three criteria assays (“triple positivity”) has been correlated with an increased risk for a future thrombotic event (see Fig. 131–5).¹⁴⁴ Investigational criteria have been developed to identify patients with the “definite” autoimmune APS (see Table 131–3).¹⁸ Because no single test is sufficient for diagnosing the disorder, a panel of tests, including antibodies against cardiolipin, and β₂GPI, and coagulation tests for LA should be performed when APS is suspected.²⁵⁷ The investigational criteria require that positive results have to be obtained on two or more occasions at least 12 weeks apart to qualify for diagnosis.

IMMUNOASSAYS

Anticardiolipin Antibody Assays

Most patients with APS are identified by elevated levels of aCL antibodies, a test with high sensitivity but poor specificity. The prevalence of positive tests in the asymptomatic healthy population has generally ranged from approximately 3 to 10 percent. In a prospective study of 2132 consecutive Spanish patients with venous thromboembolism, 4.1 percent had elevated levels of aCL antibodies (i.e., about the same prevalence as in the asymptomatic healthy population),²⁶⁰ but in a group of healthy young women in another study, the prevalence of elevated levels of aCL was 18.2 percent.²⁶¹ Many individuals have antibody levels that are elevated in response to infections that are not associated with thrombotic complications. Patients with syphilis, Lyme disease, and other infections may be misdiagnosed with APS based on elevated aCL antibody levels when concurrent stroke or arterial thrombosis is present, so these conditions must always be ruled out in susceptible patients. In a systematic literature review, 15 of 28 studies showed significant associations between aCL antibodies and thrombosis.²⁶² In all cases, a correlation existed between high antibody titers and a high risk of thrombosis. Elevated levels of aCL antibodies, whether high or low titer, were significantly associated with both myocardial infarction and cerebral stroke. Only high-titer aCL antibodies significantly increased the risk of deep vein thrombosis. During a 10-year followup of patients with elevated levels of aCL antibodies, approximately 50 percent of patients who presented with the antibodies but without clinical manifestations of the syndrome subsequently developed the APS.²⁶³ The presence of elevated titers of aCL antibodies 6 months after an episode of venous thromboembolism is a predictor for increased risk of recurrence and of death.¹⁴⁵ Women with aCL IgM antibodies, or with an aCL IgG antibody titer less than 20 IgG-binding units, and without a positive LA do not appear to be at risk for APS.²⁶⁴ In contrast, women with an aCL IgG titer greater than 20 binding units or a positive LA were more likely to develop complications.²⁶⁴ With respect to pregnancy losses, a meta-analysis of 25 studies on aPL antibodies in women with recurrent fetal losses showed a significant correlation with increased aCL IgG, and particularly with the LA.²⁶²

With respect to stroke, elevated levels of aCL antibodies of IgG or IgM isotype were reported to be significant risk factors.²⁶⁵ aPL antibodies also are an independent risk factor for stroke in young women.²⁶⁶

Approximately 20 percent of patients taking procainamide have moderate to high levels of aCL antibodies.²⁶⁷ In these patients, the antibodies are associated with anti-β₂GPI specificity. There have been case reports of associated thrombosis.²⁶⁸ Treatment with chlorpromazine has been associated with the development of aCL antibodies,²⁶⁹ although more recent data indicate that aPL antibodies are frequently elevated in patients with severe psychiatric disorders whether or not they are treated with antipsychotic medications.²⁷⁰

Anti-β₂ Glycoprotein I Antibody Assays

β₂GPI is believed to be the major protein cofactor for aPL antibodies. Enzyme-linked immunosorbent assays (ELISAs) for anti-β₂GPI antibodies are considered to be more specific but less sensitive to APS than to aCL assays.²⁷¹ In a systematic literature review, 34 of 60 studies showed significant associations between anti-β₂GPI antibodies and thrombosis.²⁶² None of the studies were prospective. Of the 10 studies that included multivariate analysis, only two confirmed that anti-β₂GPI IgG antibodies were independent risk factors for venous thrombosis. Anti-β₂GPI antibodies were more often associated with venous than arterial thrombosis. Anti-β₂GPI IgA antibodies were significantly associated with thrombosis.

Although these antibodies usually are present in conjunction with abnormal aCL and antiphosphatidyl serine antibodies, some patients with APS present solely with antibodies to β₂GPI.^272,273 Despite their higher specificity for APS (98 percent), β₂GPI antibodies alone cannot be relied upon for the diagnosis because of their low sensitivity (40 to 50 percent).^274,275 Also, interlaboratory variability is a significant problem with anti-β₂GPI antibody assays.²⁷⁶

Epitope-specific anti-β₂GPI antibodies, not yet in general use, may offer a better predictive value for diagnosis and prognosis of APS. A recent analysis of 198 samples from patients with a variety of autoimmune conditions revealed that the 52 patients with anti-β₂GPI IgG antibodies could be divided into those that recognize domain I alone and those with reactivity for all domains⁵¹; the former were positive for LA and were associated with an increased risk for thrombosis. As mentioned earlier, positivity for this assay has been correlated with positivity for a functional coagulation assay that measures resistance to annexin A5 anticoagulant activity.⁵²

Assays for Antibodies Against Prothrombin and Phosphatidyl Serine

Prothrombin is considered the second major cofactor for aPL antibodies. In a systematic literature review, 17 of 46 studies showed significant associations between antiprothrombin antibodies and thrombosis.²⁶² Of the eight studies that included multivariate analysis, two confirmed that antiprothrombin antibodies were independent risk factors for thrombosis, and three other studies showed that antiprothrombin antibodies added to the risk borne by LA or aCL antibodies. A recent study indicated that the sensitivity of antiprothrombin immunoassays in the primary aPL syndrome is too low to warrant inclusion in recommendations for APS testing.²⁷⁷

Tests for antibodies against phosphatidyl serine have been hypothesized to be more relevant than antibodies against cardiolipin, because the latter are present in intracellular membranes, whereas phosphatidyl serine is exposed on syncytialized cells, on apoptotic cells, and on activated platelets. Antibodies against phosphatidyl serine were reported to correlate more specifically with APS than aCL antibodies, particularly in arterial thrombosis.^278,279 However, antiphosphatidyl serine antibody tests are not included as an international consensus criterion.²⁵⁷

Newer studies indicate that assays for detecting autoantibodies that recognize prothrombin complexed to the anionic phospholipid, phosphatidyl serine, may have improved utility for diagnosing APS by identifying patients suspected for the disorder but with negative results in the conventional assays; in one series of 728 patients who were suspected of having APS but tested negative by conventional assays, 41 were found to be positive for anti–prothrombin/phosphatidyl serine.²⁸⁰ It has also been suggested that positivity for these assays may correlate with positivity for LA assays.²⁸¹

Assays for Antibodies Against Other Phospholipids

Some investigators have advocated testing for antibodies against a panel of phospholipids other than cardiolipin,^{282,283,284,285} but others have disagreed.²⁸⁶ Although one review has indicated that measurement for antiphosphatidyl ethanolamine antibodies may identify some patients whose conventional tests are negative for aPL antibodies,²⁸⁷ the current consensus holds that no benefit has been demonstrated for tests for antibodies against panels of different phospholipids.²⁵⁷

COAGULATION TESTS

Lupus Anticoagulants

One of the most intriguing aspects of APS is the LA phenomenon.^288,289 The various LA tests all report the inhibition of phospholipid-dependent blood coagulation reactions,⁶ but by different detection methods. These include modifications of the aPTT test with LA-sensitive and LA-insensitive reagents, the kaolin clotting time, the dRVVT, the tissue thromboplastin inhibition time, the hexagonal phase array test, and the platelet neutralization procedure. Readers who are interested in the latest consensus recommendations on details of the procedures are directed to reference 257.

The results of LA tests can be so variable that even specialized laboratories will disagree as to the results of LA tests. For example, three surveys in the United Kingdom have shown that although most laboratories agreed on identification of plasmas containing strong positive LA activity, they frequently have disagreed about samples with a weak LA activity.²⁸⁶

Despite these limitations, the presence of a positive LA appears to be the strongest predictive diagnostic test for future thrombosis. In a meta-analysis of the risk for aPL-associated venous thromboembolism in individuals with aPL antibodies without an underlying autoimmune disease or previous thrombosis, the mean odds ratios were 1.6 for aCL antibodies, 3.2 for high titers of aCL, and 11.0 for LA.²⁹⁰ In a systematic literature review, 12 of 12 studies showed significant associations between LA and thrombosis, with odds ratios from 5.7 to 9.4.²⁶² LA increased the risks of arterial and venous events to the same extent. Positivity for both LA and aCL, but not for aCL alone, predicted a higher risk of recurrent thromboocclusive events in patients with first ischemic stroke.¹³⁵ In a prospective study of pregnant women, the PROMISSE study (Predictors of Pregnancy Outcome: Biomarkers in Antiphospholipid Antibody Syndrome and Systemic Lupus Erythematosus study), LA was the primary predictor of adverse pregnancy outcome after 12 weeks’ gestation in aPL-associated pregnancies.; aCL antibody and anti-β₂GPI, did not predict adverse pregnancy outcome if LA was not also present.²⁹¹

In patients with SLE as well, the presence of LA activity is more predictive and more specific for the occurrence of thrombosis or pregnancy loss than aCL assays.²⁹² This was also found in a meta-analysis of women without autoimmune conditions who had recurrent pregnancy losses.²⁹³

Dilute Russell Viper Venom Time

dRVVT is considered to be one of the most sensitive of the LA tests. The assay uses Russell viper venom (RVV) in a system containing limiting quantities of diluted rabbit brain phospholipid. RVV directly activates coagulation factor X, leading to formation of fibrin clot. LA prolongs dRVVT by interfering with assembly of the prothrombinase complex; however, the prolongation is reversed by adding excess phospholipid to the reaction (sometimes referred to a “confirmatory test”). To ensure that prolongation of the clotting time is not a result of a factor deficiency, the procedure includes mixture of patient and control plasmas. Anticoagulant therapy with heparin, warfarin, or direct thrombin inhibitors can yield falsely abnormal test results.

Activated Partial Thromboplastin Time Tests

Prolongation of the aPTT detects some LAs, and prolonged aPTTs in otherwise healthy individuals are most frequently caused by LAs.²⁹⁴ The various commercial aPTT reagents vary widely with regard to sensitivity to LA, so it is important to know the characteristics of the particular reagent(s) that is (are) being used. When the aPTT of a particular plasma sample is prolonged and not correctable by immediate mixture with normal plasma, the presence of an LA should be suspected, especially if the patient does not have bleeding symptoms. The LA needs to be differentiated from inhibitors of specific coagulation factors and from anticoagulants such as heparin. Besides specific assays to exclude the latter two possibilities, the clinician should check whether the aPTT normalizes when an LA-insensitive aPTT reagent is used or when the assay is performed using frozen washed platelets as the source of phospholipid, a procedure referred to as the platelet neutralization procedure. The effects of incubation with normal plasma may be helpful in differentiating LAs from coagulation factor inhibitors. aPTTs performed on mixtures of normal plasma and plasma containing a factor VIII inhibitor usually show no prolongation immediately after mixing but marked prolongation following incubation for 1 to 2 hours at 37°C, whereas LA-containing plasmas usually prolong the aPTT immediately after mixing with normal plasma and show no further prolongation with incubation. The clinician should be aware that both types of anticoagulants, LA and specific coagulation factor inhibitors, may coexist in rare patients and yield confusing laboratory results. Specific coagulation factor inhibitor assays and using an aPTT reagent that is insensitive to LA are helpful for clarifying most of these cases. LAs may result in artifactual decreases in contact activation pathway coagulation factor assays, because these assays are based on aPTT. Consequently, these patients are sometimes misdiagnosed as having multiple coagulation factor deficiencies. This problem can be handled by repeating the coagulation factor assays following dilution of the plasma samples; this usually results in complete or partial normalization of coagulation factor levels with progressive dilution. The use of an aPTT reagent that is insensitive to LA for specific factor assays is another way to solve this problem.

Other Methods for Detecting Lupus Anticoagulant

The dilute prothrombin time (dPT) (also known as the tissue thromboplastin inhibition test [TTIT]) is essentially a prothrombin time assay done with diluted tissue factor–phospholipid complex. It can be performed with standard or recombinant tissue factor.^295,296 The results are expressed as a ratio of the patient-to-control clotting times.

The kaolin clotting time (KCT) depends on the ability of aPL antibodies to block the coagulant activity of trace amounts of phospholipid present in centrifuged plasma. Some authors maintain that the KCT–LA test reflects dependence on prothrombin as a cofactor and is less likely to be associated with thrombosis than the dRVVT, which appears to be more dependent on β₂GPI.^297,298 The hexagonal phase array test is based on a prior idea that aPL antibodies recognize phosphatidyl ethanolamine directly in the hexagonal phase array configuration but not in the lamellar phase. Although this assay remains in use, the correction of the prolonged clotting time with hexagonal phase phosphatidyl ethanolamine is probably similar to the confirmatory step used in the other LA assays, that is, a result of the excess of phospholipid in the reaction.

The textarin/ecarin test depends on the difference in phospholipid dependence of coagulation mechanisms triggered by two snake venoms: textarin, which activates prothrombin via a phospholipid-dependent pathway, and ecarin, which activates prothrombin directly without phospholipid.²⁹⁶

Annexin A5 Resistance Assay

In addition to the various LA tests, there is a coagulation test that reports on a thrombogenic mechanism—resistance to annexin A5 anticoagulant activity.⁸⁰ This assay has been correlated with an immunoassay for IgG antibodies against domain I of β₂GPI.⁵² The assay has two stages: a first, in which a tissue factor-phospholipid suspension is exposed to test plasma, and a second, in which the washed suspension is used to coagulate pooled normal plasma in the presence and absence of annexin A5. Patients with annexin A5 resistance show a less-than-expected annexin A5 anticoagulant effect, reported as a reduction in the annexin A5 anticoagulant ratio. In contrast to the lupus “anticoagulant” effect, this assay measures and reports a procoagulant effect for the antibodies.⁷⁰ (Disclosure: One of the authors (J.H.R.) is inventor of this assay, U.S. Patents #6284475 and #7252959.)

Chapters 133 and 134 address the general subject of vascular thrombosis and its differential diagnosis. When vascular occlusion occurs in the setting of a known autoimmune disorder such as SLE, the possibility of a vasculitis, rather than a thrombotic condition, should be considered. Patients with CAPS may, at first, appear to have other multisystem vasoocclusive disorders, such as thrombotic thrombocytopenic purpura or disseminated vasculitis, and may also manifest laboratory findings of disseminated intravascular coagulation.

The differential diagnosis of a prolonged aPTT includes hereditary and acquired coagulation factor deficiencies, inhibitors to coagulation proteins (e.g., acquired hemophilia A; Chap. 128), and the presence, or use, of anticoagulants. The diagnosis of LA is substantiated through plasma mixing studies and specific factor assays. A positive immunoassay for aPL—that is, aCL and/or anti-β₂GPI antibodies—helps confirm the diagnosis.

When an elevated level aPL antibody level is detected, the clinician must exclude the possibility of an infectious etiology for the antibodies; these occur frequently in syphilis, Lyme disease, HIV-1, and hepatitis C. Occasional patients may have artifactually elevated antibodies from increased polyclonal immunoglobulin levels.²⁹⁹ In such cases, diagnosis is aided by specific tests for suspected infection, quantitative measurement of serum immunoglobulins, and subtraction of background controls using uncoated microtiter wells. Antipsychotic or other medications should be excluded as causative agents.

PREGNANCY COMPLICATIONS

A systematic review of treatments given to maintain pregnancy in women with prior miscarriages and APS concluded that combined unfractionated heparin and aspirin may reduce pregnancy loss by 54 percent compared to aspirin alone.³⁰⁰ Three trials of aspirin alone showed no significant reduction in pregnancy loss³⁰⁰; intravenous immunoglobulin, whether in combination with or without unfractionated heparin (UFH) and aspirin, was associated with an increased risk of pregnancy loss or premature birth when compared to UFH or low-molecular-weight heparin (LMWH) combined with aspirin.

Taking together the available data, women with a history of three or more spontaneous pregnancy losses and evidence of aPL antibodies should be treated with a combination of low-dose aspirin (75 to 81 mg/day) and prophylactic doses of UFH (i.e., 5000 U every 12 hours subcutaneously). Treatment should be started as soon as pregnancy is documented and continued until delivery so as to reduce the rate of late complications.^183,301 In especially high-risk situations, induction of early delivery may be necessary. Unfractionated heparin at the prophylactic dosage of 5000 U q12h subcutaneously should be started approximately 4 to 6 hours after delivery, if significant bleeding has ceased, and continued at least until the patient is fully ambulatory. Many physicians recommend continuing prophylactic therapy for 6 weeks after delivery even if the patients have not experienced thrombosis. For patients who experienced thromboembolism, prophylaxis by heparin or oral anticoagulant therapy is warranted for at least 6 weeks after delivery.

Although treatment with LMWH has become widely used for recurrent fetal loss as a replacement for prophylactic dose UFH, a prospective randomized controlled trial has questioned the benefit of LMWH treatment versus aspirin therapy. In the LMWH/aspirin group 35/47 (77.8%) had a livebirth, and in the aspirin group 34/43 (79.1 percent) had a livebirth (p = 0.7).³⁰²

The presence of positive aPL laboratory assays alone is not an indication for treatment in pregnant women without a history of spontaneous pregnancy losses, other attributable pregnancy complications, thrombosis, or SLE. Therefore, the inclusion of aPL tests in routine prenatal testing panels should be discouraged.

Although prednisone was reported to possibly improve the outcome of pregnancy in women with APS, those benefits are associated with significant toxicity.³⁰³ Glucocorticoids or intravenous IgG should be considered only for patients who are refractory to anticoagulant therapy, who have a severe immune thrombocytopenia or a significant contraindication to heparin therapy. Treatment with the combination of prednisone and heparin is associated with an increased risk of osteopenia and vertebral fractures.³⁰⁴

There is general agreement that APS patients with recurrent spontaneous thrombosis require long-term, and perhaps lifelong, anticoagulant therapy and APS patients with recurrent spontaneous pregnancy losses require antithrombotic therapy for most of the gestational period. There are differences of opinion among experts regarding the approaches to treatment of patients with a single thrombotic event, patients with a history of thrombotic events in the distant past (>5 years), patients with stroke, and patients with thrombotic events that were associated with a provocative factor such as trauma, surgery, stasis, pregnancy, and estrogens.

THROMBOSIS

The accumulated evidence from randomized controlled trials indicates that patients with APS and thrombosis should be treated with warfarin for the long-term, and maintained at a therapeutic international normalized ratio (INR) of 2.0 to 3.0.³⁰⁵ Patients with arterial thrombosis may require a higher anticoagulant intensity as a retrospective study showed that a higher intensity (INR >3.0) was necessary for preventing recurrences in this group of patients, but this issue is controversial.³⁰⁶ Two other studies reported no benefit for high-intensity warfarin but the number of patients with arterial thrombosis was not high.^299,305 The issue of appropriate antithrombotic treatment of aPL-associated stroke is even more controversial. One major study concluded that there was no benefit for warfarin anticoagulation compared to aspirin therapy.³⁰⁷

For patients treated acutely with intravenous UFH, care must be taken to determine whether the patient has a preexisting LA that can interfere with aPTT monitoring of heparin levels. This problem can be circumvented by using an LA-insensitive aPTT reagent or be avoided by treatment, where appropriate, with a LMWH.

An important practical consequence of the LA effect is that prothrombin time and INR results can be artifactually elevated in some patients with APS and LAs treated with warfarin anticoagulant therapy.³⁰⁸ A multicenter study reported that all but one of the commercial thromboplastins in use at nine centers provided acceptable INR values for APS patients with LA.³⁰⁹ New thromboplastins should be checked for their responsiveness to LA prior to their use in monitoring oral anticoagulant treatment in patients with APS. Chromogenic factor X (CFX) assays can be used as an alternative to INR for APS patients, especially in patients with a prolonged baseline prothrombin time prior to initiating warfarin therapy, those who are persistently positive for LA and those patients who continue to have recurrent venous thromboembolism (VTE).³¹⁰ Therapeutic CFX values range from 20 to 40 percent; thus, a CFX of 40 percent would approximate an INR of 2.0, and a CFX of 20 percent would approximate an INR of 3.0.

New oral anticoagulants (NOACs), either direct factor Xa or thrombin inhibitors, are effective for treatment of VTE.³¹¹ However, their use specifically in APS patients has not been thoroughly evaluated. In two recent case studies, NOACs have failed to prevent thrombosis in APS patients. Of six APS patients studied, five suffered recurrent VTE and one suffered a recurrent TIA after transitioning to NOACs.^312,313 At the time of writing, a prospective randomized controlled trial of warfarin versus rivaroxaban named “RAPS” (Rivaroxaban in Antiphospholipid Syndrome) is ongoing in patients with thrombotic APS (study ISRCTN68222801). Until the trial is concluded (expected completion in 2015), caution should be applied in using NOACs for APS patients.

Fibrinolytic treatment has been reported for patients with primary APS and extensive thrombosis of the common femoral and iliac veins extending to the lower vena cava,³¹⁴ acute ischemic stroke,³¹⁵ and acute myocardial infarction.³¹⁶

The antimalarial drug hydroxychloroquine (HCQ) is associated with reduced risk of thrombosis in patients with APS^{271,272,273,317,318,319} and SLE.^319,320,321 The potential effectiveness of this treatment has been supported by an animal model for aPL thrombosis³²² and by a recent report that HCQ directly disrupts aPL IgG–β₂GPI complexes,³²³ and also reverses the aPL antibody-mediated disruption of annexin A5 binding on phospholipid bilayers³²⁴ and on human placental syncitiotrophoblasts.³²⁵ In a longitudinal cohort study consisting of 272 patients with the APS and 152 taking HCQ (17 of 272 patients on warfarin, 203 were on prednisolone, 112 on azathioprine, 38 on aspirin) investigators found fewer thrombotic complications for patients on HCQ (odds ratio [OR] 0.17, 95% confidence interval [CI] 0.07 to 0.44; p <0.0001).³²⁶ In asymptomatic aPL antibody-positive patients with SLE, primary prophylaxis with aspirin and HCQ appeared to reduce the frequency of thrombotic events.³²⁷ A published prospective, nonrandomized study compared oral anticoagulant plus HCQ versus oral anticoagulant alone. In this study, 30 percent (6/20) of patients had a thrombotic event if they were on oral anticoagulant alone, despite therapeutic range INR, versus no thrombotic events in the oral anticoagulant plus HCQ group (0/20). This study, however, was limited given the small number of patients studied and short followup.³²⁸ Recently, the natural 4-aminoquinolone, quinine, was shown in vitro to disrupt the immune complexes bound to phospholipid layers.³²⁹ However, both HCQ and quinine will require clinical testing in appropriately designed clinical trials. A prospective randomized controlled trial comparing HCQ to placebo in aPL-positive patients without a prior history for thrombosis is currently in progress.

Other proposed treatments for APS include statins, rituximab and vitamin D. Statins have immunomodulatory, antiinflammatory and antithrombotic properties that may benefit APS patients. In recent studies, APS patients treated with statins demonstrated downregulation of tissue factor and reduced proinflammatory/prothrombotic markers such as interleukin-1β, vascular endothelial growth factor, tumor necrosis factor α, interferon-inducible protein 10, and soluble CD40L.^330,331 It has been suggested that a B-cell inhibitor such rituximab may be useful in reducing aPL antibody titers in APS patients. The Rituximab in Antiphospholipid Syndrome (RITAPS) trial did not show reduction in aPL antibody profiles by rituximab; however, rituximab may be effective in controlling noncriteria manifestations of APS.³³² Because low vitamin D levels correlate with arterial and venous thrombosis as well as noncriteria APS manifestations,^333,334,335 it is recommended that vitamin D deficiency (<10 to 20 ng/mL) and insufficiency (<30 ng/mL) be corrected in aPL antibody-positive patients.³³⁶

Conventional anticoagulant therapy is usually not sufficient for treatment of CAPS; these patients require aggressive treatment because of the high mortality.¹⁷⁶ Treatment for CAPS is directed toward the thrombotic events and suppression of the cytokine cascade. This includes anticoagulation with heparin and immunosuppressive therapy in the form of high-dose glucocorticoids. A triple therapy strategy of anticoagulation, glucocorticoids, and either intravenous immunoglobulin or plasma exchange or both has improved outcomes. Cyclophosphamide is recommended for patients with CAPS and inflammatory features of SLE or high-titer aPL antibodies. Rituximab may be useful in refractory or relapsing cases of CAPS.¹⁷⁸

PREGNANCY COMPLICATIONS

The current approach to treating pregnant women with APS and recurrent pregnancy losses or the other aPL antibody–associated complications of pregnancy includes daily low-dose aspirin (75 to 81 mg/day) and either UFH or LMWH.^300,337,338 Although clinical studies have shown efficacy with UFH, most clinicians treat with LMWH because it has a better pharmacokinetic profile and a lower risk of heparin-induced thrombocytopenia and osteopenia. Heparin is then withheld when labor begins or 24 hours prior to a cesarean section. Anticoagulation is resumed 6 weeks postpartum because of the increased risk of VTE in this time period.³⁰¹ Interestingly, the current standard of care for pregnant APS patients is based on two randomized controlled trials conducted prior to 2000 that included only 150 patients. Newer trials show conflicting results, with some showing no difference in prevention of pregnancy loss in APS patients receiving aspirin alone versus aspirin and LMWH,³⁰² and others showing a small benefit.³³⁹ With this management, the likelihood of a good pregnancy outcome in women with APS has been estimated to be approximately 75 to 80 percent.

Other treatment modalities such as glucocorticoids or intravenous IgG (IVIG) should be considered only for patients who are refractory to anticoagulant therapy, who have a severe immune thrombocytopenia, or who have a significant contraindication to heparin therapy. The addition of glucocorticoids has shown no clear benefits and has been associated with premature rupture of membranes or preeclampsia; however, a newer study showed that for patients with refractory APS, the addition of low-dose prednisolone (10 mg) from time of a positive pregnancy test up to 14 weeks’ gestation may help to increase livebirth rates.³⁴⁰ Although the addition of IVIG has not been shown to be superior to heparin and aspirin in large multicenter clinical trials,³⁴¹ it has shown some efficacy in patients with refractory APS in small case studies.^342,343