Jpn J Thromb Hemost 9(3) :159~165, 1998

Special Lecturf

Antiphospholipid- -A Paradigm for -Mediated Thrombosis-

Douglas A. TRIPLETT, M.D."

Antiphospholipid-protein antibodies (APA) are a family of immunoglobulins which recognize a number of plasma that bind to activated membranes. Among these

plasmaproteinsare:β2 GlycoproteinI(β2 GPI),prothrombin,AnnexinV,andothervitaminK dependent proteins (e.g., ). APA have been associated with venous and arterial events as well as recurrent spontaneous abortion (RSA). It is estimated that 8 to 14% of venous thrombosis is secondary to the presence of APA. APA are the most common cause of acquired thrombophilia. The concept of antibody-mediated thrombosis is now widely accepted. The laboratory diagnosis of APA requires the use of -based tests to identify anticoagulants (LA) and ELISA assays to identify anticardiolipin antibodies (ACA). The importance of testing for antibodies to β2 GPI is now well recognized. Antibodies to β2 GPI are more specific in identification of patients at risk for thrombosis.

Key words : anticardiolipin, lupus anticoagulant, thrombosis

found lupus anticoagulants (LA) recognize Introduction human prothrombin rather than Antiphospholipid-protein antibodies phospholipidss'. Both ACA and LA recognize (APA) are a family of immunoglobulins specific plasma proteins when they are bound which have been defined by in vitro labora- to activated phospholipid (PL) membranes tory tests (Table 1) 1)2).The Immunoglobulins in vivo (e.g., perturbed endothelium, may be IgG, IgM, or IgA isotypes. Individual platelets, monocytes). Presumably, these patients often have a mixture of isotypes. PL-binding proteins when reacting with Recently, the concept of these antibodies having specificity for has been Table 1 Antiphospholipid Antibodies (APA) challenged. The work of McNeil et al., Galli et al., and Matsuura et al. demonstrated that anticardiolipin antibodies (ACA) are not directed against but instead apPear to recognize a plasma protein:β2

Glycoprotein I(β2 GPI)3)4)5)Following these initial observations, Bevers and colleagues

* Director , Midwest Thrombosis and Hemostasis Laboratories, Muncie, Indiana. Professor of Pathology and Asistant Dean, Indiana University School of Medicine. The 20th Congress of the Japanese Society of Thrombosis and Hemostasis 160 Jpn J Thromb Hemost Vol. 9 No. 3

Table 2 APA Classification

activated PL membranes expose neoepitopes. the presence of one or more of the clinical The subsequent immunologic response to findings noted above together with positive exposure of the neoepitopes results in LA and laboratory tests for ACA or LA. However, ACA formation. In vitro test systems using the criteria required to diagnose APS need to "high sensitivity" microtiter plates relay on be revised. With the realization that APA the same phenomenon when testing for anti- target PL-binding proteins, new test systems bodies toβ2 GPI or human prothrombin. have been introduced including those which With the realization that APA, in fact, recog- identify antibodies to specific proteins (e.g.,

nize proteins reconfigured on PL surfaces, a ,β2 GPI). Recently, patients with APS have more accurate name for this family of anti- been described who have negative test results bodies is antiphospholipidprotein antibodies for ACA and LA and only demonstrate anti- (APA). bodies to β2 GPI7). APA may be found in a variety of different In addition to autoimmune APA, alloim- clinical settings. In patients with underlying mune APA have been described following autoimmune disease (e.g., systemic lupus various infections including bacterial, viral, erythematosus„ rheumatoid arthritis, etc.) and protozoal agents. Often, alloimmune APA are often detected. The presence of APA are transient in contrast to persistence these antibodies in autoimmune disease has typically seen with autoimmune APA. Table been linked to a variety of clinical complica- 2 summarizes clinical setting of allo-and tions including recurrent venous and arterial autoimmune APA. An important charac- thromboembolic events, recurrent abortion/ teristic of alloimmune APA is its transient fetal loss, and thrombocytopenia. The anti- nature. This is in contrast to persistence of phospholipid syndrome (APS) is defined by APA in autoimmune disease(i.e.,persistent DA TRIPLETT: Antiphospholipid-Protein Antibodies and Thrombosis 161 positivity of laboratory results demonstrated also do parallel testing for ACA. The pres- 6-8 weeks after the initial positive test). ence of persistently positive LA and/or ACA identifies patients at increased risk for thromboembolic events. Patients at risk may Laboratory Detection of APA be positive for one or both of these tests Lupus Anticoagulants (more recently, it has been appreciated The term lupus anticoagulants was first patients may be negative for both tests and proposed by Feinstein and Rapaport in yet still have an increased risk of thrombosis 19728). This term is a misnomer since the due to anti-β2 GPI). Thus, when clinicians vast majority of patients with LA do not have request a LA evaluation, they are asking for underlying systemic lupus erythematosus a comprehensive evaluation for the presence (SLE). Many patients with LA are identified of APA. The laboratory is responsible for immediately following an intercurrent upper performing the appropriate range of tests respiratory tract infection. These patients do necessary to identify members of this family not manifest any of the findings typically of antibodies1)2). seen with APS. Subsequent testing will, in The Lupus Anticoagulant/Phospholipid- most cases, show disappearance of LA. In dependent Antibody Subcommittee of the many cases, patients positive for LA will also Scientific and Standardization Committee/ have ACA positivity (~40%). Conversely, International Society of Thrombosis and ACA positive patients, in many cases, do not Haemostasis (ISTH) has published specific have an associated LA. These differences guidelines required to identify LA. Four may reflect the underlying quantity of sequential steps are necessary15). antibody present in the patient plasma as (1) Demonstration of an abnormal well as differences in protein targets and phospholipid-dependent coagulation screen- antibody affinity. ing test (e.g., Kaolin Clotting Time (KCT), Demonstration of LA identifies a patient at dilute Russell Viper Venom Time (dRVVT), increased risk for thromboembolic events as Activated Partial Thromboplastin Time well as recurrent spontaneous abortions (APTT), Textarin Time (TT)) ; (RSA) 1)2)9)10)11)The presence of LA in the (2) Demonstration the abnormal screen- setting of autoimmune disease indicates a ing test is due to the presence of an inhibitor relative risk for thrombosis of about seven-to (synonym ; circulating anticoagulant ; eight-fold increase when compared to con- (3) Proof the inhibitor is phospholipid- trols1)12). Approximately two-thirds of throm- dependent; boembolic events are venous and the remain- (4) Rule out other coagulopathies (e.g., ing one-third arterial. Arterial events are presence of heparin in the sample, circulating typically found in the cerebral circulation13). inhibitors such as factor VIII inhibitor). The laboratory diagnosis of LA requires a With increasing recognition of the com- careful sequential series of steps14). When a plexity of the diagnosis of LA, a variety of clinician requests the laboratory to rule out new test systems have been introduced. In LA, it is very important for the laboratory to some cases, these tests are complex involving 162 Jpn J Thromb Hemost Vol. 9 No. 3

incorporation of Steps 1 through 3 into a eliciting Procoagulant reactions23). As noted

singlesystem.TheStaclotLAｮisanexam一 above, antibodies to,β2 GPI are now being

pie of such an integrated approach16'. Also, evaluated as a means of identifying patients dRVVT-based systems have been at increased risk for thrombosis. Anti-β2 GPI marketed17). appears to be more specific in identifying

Because of the heterogeneity of the anti- patients at risk of thrombosis24). Consequent- body population seen in patients with LA, ly, many laboratories now perform both con-

more than one screening procedure is ventional ACA and an ELISA assay for anti-

required. Most often, laboratories choose bodies to β2 GPI. screening procedures which differ in the por- tion of the coagulation test system analyzed Antibody-Mediated Thrombosis (e.g., APTT : Intrinsic ; PT : Extrinsic, dRVVT, Final Common Pathway). For APA are perhaps the most common instance, the most common combination is a acquired cause of thrombophilia. With the dRVVT and APTT-based systems. Also, the recent discovery of hereditary abnormalities dilute Prothrombin Time (dPT) and Textar- leading to thrombophilia (e.g., in Time have been used18). Leiden, 3' untranslated portion of the proth- Anticardiolipin Antibodies (ACA) rombin gene (G20210A), etc.), laboratories ACA testing was originally described utiliz- are increasingly testing patients for both ing solid phase redioimmunoassay19). Because hereditary and acquired risk factors for this assay system was expensive and also thrombosis25). The concept of antibody- required utilizing radioisotopes, many labora- mediated thrombosis has been debated for a tories quickly moved to ELISA technology20). number of years. Perhaps the classic example As discussed earlier, ACA measures anti- is heparin-induced thrombocytopenia bodies toβ2 GPI5)21)22). Other plasma proteins (HIT) 26).HIT is associated with antibodies which bind to PLs may be identified in the which appear to recognize a complex of conventional ACA assay(antibodies to An- heparin- (or other plasma nexin V). β2 GPI is a member of the short proteins such as IL-6)27). This immune com- consensus repeat family of proteins. plex binds to FcγIIa receptors on the surface Physiologically it has anticoagulant prop- of platelets resulting in activation of platelets erties (inhibits contact activation of the and subsequent thrombosis28). The patho- intrinsic pathway, inhibits ADP-induced physiology of thrombosis seen in the APS is platelet aggregation, inhibits the prothrom- similar. In this case, however, β2 GPI binds to binase reaction). Despite its anticoagulant activated cell membranes (with exposed activities, hereditary deficiency of/β2 Glyco一 phosphatidylserine). Antibodies then react protein I is not associated with a predisposi- with neoepitopes on bound,β2 GPI. This tion to thrombosis. Presumably, the mecha- antigen antibody complex, in turn, activates nism of thrombosis associated with anti- FcγIIa receptors leading to platelet activa- bodies to β2 I requires trans一 tion and thrombotic events(Table 3). membrane signaling with cellular activation Also, APA may inhibit regulatory systems DA TRIPLETT : Antiphospholipid-Protein Antibodies and Thrombosis 163

Table 3 Similarities of APS & HIT*

Legend : APS= Antiphospholipid Syndrome HIT= Heparin Induced Thrombocytopenia PL-II : Phospnolipid-Prothrombin Complex PF 4-H: Platelet Factor 4-Heparin Complex VT : Venous Thrombosis AT : Arterial Thrombosis *Predominant immunoglobulin , often patients will have a mixture of immunoglobulin based on membrane localized reactions. The / system requires activa- tion of protein C and subsequent downregula- tion of factor Va and factor Villa by a com- plex of Activated Protein C (APC) and pro- tein S on cellular membranes29~30'. Antibodies to β2 GPI can inhibit this reaction. Again, predisposing to thrombosis. HIT and the antiphospholipid syndrome Fig. 1 Antibody Induced Thrombosis represents paradigms of antibody-mediated Legend : AECA=Antiendothelial thrombosis31). Other examples include the Cell Antibodies pathogenicity of antibodies to oxidized LDL Anti-Ox-LDL=Antibodies to Oxid- ized Low Density and their association with accelerated ather- Anti-β2 GPI=Antibodies to Beta2 osclerosis32). (Fig. 1) In addition, antibodies Glycoprotein I to components of the protein C system have Anti-Prothrombin = Antibodies to Prothrombin been described (anti-PC and anti-PS as well APA=Antiphospholipid-Protein as antithrombomodulin). Antiendothelial cell Antibodies membrane antibodies have also been linked to thrombosis. In the coming years, addi- References tional mechanisms for antibody-mediated thrombosis will be described. It now appears 1) Triplett DA : Antiphospholipid-protein anti- antibody-mediated thrombosis may be one of bodies : laboratory detection and clinical rele- the most important acquired mechanisms for vance. Thromb Res 78: 1-31, 1995. 2) Triplett DA : Antiphospholipid antibodies and thrombosis affecting the arterial and venous thrombosis. a consequence, coincidence, or circulation. cause? Arch Pathol Lab Med 117: 78-88, 1993. 3)McNeil HP, Simpson RJ, Chesterman CN, Krilis 164 Jpn J Thromb Hemost Vol. 9 No. 3

SA ; Antiphospholipid antibodies are directed phospholipid antibodies to thromboembolic dis- against a complex antigen that includes a ease in systemic lupus erythematosus : a cross binding inhibitor of coagulation:β2 Glycoprotein sectional study. Thromb Haemost 66: 520, 1991. I ( H) . Proc Natl Acad Sci USA 12) Ginsburg KS, Liang MH, Newcomer L, 87: 4120-4124, 1990. Goldhaber SZ, Schur PH, Hennekens CH, Stamp- 4) Galli M, Comfurius P, Maassen C, Hemker HC, fer MJ : Anticardiolipin antibodies and the risk de Baets MB, van Breda•Vriesman PJC, Barbui for ischemic stroke and venous thrombosis. Ann T, Zwaal RFA, Bevers EM : Anticardiolipin anti- Int Med 117 : 997-1002, 1992. bodies (ACA) directed not to cardiolipin but to a 13) Ford PM, Brunet D, Lillicrap DP, Ford SE : plasma cofactor. Lancet 335: 1544-1547, 1990. Premature stroke in a family with lupus 5) Matsuura E, Igarashi M, Igarashi Y, Katahira T, anticoagulant and antiphospholipid antibodies. Nague H, Ichikawa K, Triplett DA, Koike T : Stroke 21: 66-71, 1990. Molecular studies on phospholipid binding sites 14) Triplett DA : New diagnostic strategies for lupus

and cryptic epitopes appearing on β2 Glyco- anticoagulants and antiphospholipid antibodies. protein I structure recognized by anticardiolipin Haemost 24:155-164, 1994. antibodies. Lupus 4 : S13-S17, 1995. 15) Brandt JT, Triplett DA, Alving B, Scharrer I: 6) Bevers EM, Galli M : Cofactors involved in the Criteria for the diagnosis of lupus anti- antiphospholipid syndrome. Lupus 1: 51-53, 1992 coagulants : an update. Thromb Haemost 74: (editorial). 1185-1190, 1995. 7) Alarcon-Segovia D, Mestanza M, Cabiedes J, 16) Triplett DA, Barna LK, Unger GA : A hexagonal Cabral AR : The antiphospholipid/cofactor syn- (II) phase phospholipid neutralization assay for drome. II. A variant in patients with systemic lupus anticoagulant identification. Thromb

lupus erythematosus with antibodies to β2 Glyco- Hemost 70: 787-793, 1993. protein I but no antibodies detectable in standard 17) Brandt JT, Triplett DA : The effect of phos- anticardiolipin assays. J Rheumatol 24:1545- pholipid on detection of lupus anticoagulants by 1551, 1997. the dilute Russell viper venom time. Arch Pathol 8) Feinstein DI, Rapaport SI : Acquired inhibitors Lab Med 113:1376-1378, 1989. of blood coagulation. Prog Hemost Thromb 1: 18) Triplett DA, Stocker K, Unger G, Barna L : The 75-95, 1972. textarin/ecarin ratio : a confirmatory test for 9) Ginsberg JS, Demers C, Brill-Edwards P, lupus anticoagulants. Thromb Hemost 70:925- Johnston M, Bona R, Burrows RF, Weitz J, 931, 1993. Denburg JA : Increased generation and 19) Harris EN, Gharavi AE, Boey ML, Patel BM, activity in patients with systemic lupus eryth- MacWorth-Young CG, Loizou S, Hughes GRV : ematosus and anticardiolipin antibodies : evi- Anticardiolipin antibodies : detection by radioim- dence for a prothrombotic state. Blood 81: 2958- munoassay and association with thrombosis in 2963, 1993. systemic lupus erythematosus. Lancet 2:1211- 10) Ginsberg JS, Brill-Edwards P, Johnston M, et al.: 1214, 1983. Relationship of antiphospholipid antibodies to 20) Loizou S, McCrea JD, Rudge AC, Reynolds R, pregnancy loss in patients with systemic lupus Boyle CL, Harris EN : Measurement of anticar- erythematosus : a cross-sectional study. Blood diolipin antibodies by an linked im- 80: 975-980, 1992. munosorbent assay (ELISA) : standardization 11) Long AA, Ginsberg JS, Brill-Edwards P, and quantitation of results. Clinn Exp Immunol Johnston M, Turner C, Denburg JA, Cividino A, 62:738-745, 1985.

Andrew M, Hirsh J : The relationship of anti- 21)Roubey RAS:Autoantibodies to phospholipid一 DA TRIPLETT: Antiphospholipid-Protein Antibodies and Thrombosis 165

binding plasma proteins : a new view of lupus heparin-induced thrombocytopenia : a study of 44 anticoagulants and other "antiphospholipid" cases. Thromb Haemost 73: 21-28, 1995. autoantibodies. Blood 84: 2854-2867, 1994. 28) Denomme GA, Warkentin TE, Horsewood P, 22) Roubey RAS, Eisenberg RA, Harper MF, Win- Sheppard J-AI, Warner MN, Kelton JG : Activa- field JB : "Anticardiolipin" autoantibodies recog- tion of platelets by sera containing IgGI heparin- nize f2-glycoprotein I in the absence of phos- dependent antibodies : an explanation for the pholipid. J Immunol 154: 954-960, 1995. predominance of the FcyRIIa "low responder" 23) Roubey RAS, Hoffman M : From antiphos- (his131) gene in patients with heparin-induced pholipid syndrome to antibody-mediated throm- thrombocytopenia. J Lab Clin Med 130: 278-284, bosis. Lancet 350: 1491-1493, 1997. 1997. 24) Sanmarco M, Soler C : Heterogeneity of anti-Q2 29) Potzsch B, Kawamura H, Priessner KT, Schmidt Glycoprotein I antibodies. Nouv Rev Fr Hematol M, Seelig C, Muller-Berghaus G : Acquired pro- 37: S57-S60, 1995. tein C dysfunction but not decreased activity of 25) Rosendaal FR : Risk factors for venous thrombo- thrombomodulin is a possible marker of throm- sis : prevalence, risk and interaction. Sem bophilia in patients with lupus anticoagulant. J Hematol 34:171-187, 1997. Lab Clin Mid 125: 56-65, 1995. 26) Chong BH, Castaldi PA : Platelet proaggregating 30) Carreras L0, Forastiero RR : Pathogenic role of effect of heparin : possible mechanism for non- antiprotein-phospholipid antibodies. Haemos- immune heparin-associated thrombocytopenia. tasis 26 : 340-357, 1996. Aust N Z J Med 16: 715-716, 1986. 31) Triplett DA : Antiphospholipid antibodies, lupus 27) Amiral J, Bridey F, Wolf M, Boyer-Neumann C, anticoagulants and thromboembolic disease. Fressinaud E, Vissac AM, Peynaud-Debayle E, Haematologica 80: 122-126, 1995. Dreyfus M, Meyer D : Antibodies to macro- 32) Vaarala 0 : Atherosclerosis in SLE and Hughes molecular platelet factor 4-heparin complexes in syndrome. Lupus 6 : 489-490, 1997.