Guidelines On The Investigation and Management Of The Antiphospholipid Syndrome

In 1991, the Haemostasis and Thrombosis Task Force of the British Society for Haematology published guidelines on testing for the lupus anticoagulant (LA) (Machin et al, 1991). Since then there have been major developments in our understanding of the nature of ‘antiphospholipid antibodies’ (aPLs) and improvements in our knowledge of the clinical course of the antiphospholipid syndrome. These revised guidelines have been prepared against this background. They cover both clinical and laboratory aspects. Although most early observations were on LA and anticardiolipin (aCL), it is now clear that so-called antiphospholipid antibodies comprise a family of antibodies reactive with epitopes on proteins which are themselves complexed with negatively charged phospholipid (Roubey, 1996). Thus, many antiphospholipid antibodies require β2-glycoprotein I (β2-GP-I), a phospholipid-binding plasma protein with weak anticoagulant activity, for binding to acidic phospholipids such as phosphatidylserine and cardiolipin (Galli et al, 1990; McNeil et al, 1990; Matsuura et al, 1990). The precise relationships among β2-GP-I, phospholipid and autoantibody are disputed. One possibility is that cryptic epitopes are exposed on β2-GP-I when it binds to phospholipid. Alternatively, binding of the glycoprotein to phospholipid may concentrate the antigenic sites and promote bivalent antibody binding. Whichever mechanism is active, it also applies when β2-GP-I interacts with other negatively charged surfaces, including the plastic of an enzyme-linked immunosorbent assay (ELISA) plate. This observation has allowed the development of new, possibly more specific, assays for antiphospholipid antibodies which use purified β2-GP-I. However, other proteins share this property of binding to phospholipid in a manner that promotes interaction with antiphospholipid antibodies. These include prothrombin, annexin V, protein C, protein S, thrombomodulin and high molecular weight kininogen. The in vitro phenomenon known as LA can be due to antibodies reactive to β2-GP-I/phospholipid or to prothrombin-phospholipid. The β2-GP-I-dependent antibodies also bind in traditional anti-cardiolipin assays as the glycoprotein is present in test serum and often in assay reagents. Despite this improved understanding of the true nature of aPL and because the clinical utility of the newer assays is incompletely evaluated, the laboratory diagnosis of antiphospholipid syndrome still relies predominantly on coagulation-based assays for LA and solid phase assays (ELISAs) using cardiolipin. Antiphospholipid syndrome (APS) may be diagnosed when arterial or venous thrombosis, or recurrent miscarriage, occurs in a subject in whom laboratory tests for antiphospholipid antibody (aCL, LA or both) are positive. Because thrombotic disease, miscarriage and transient antiphospholipid antibody positivity are all common events, persistence of the positive tests must be demonstrated and other causes and contributory factors considered. Additional clinical and laboratory features are variably present in APS, particularly thrombocytopenia and livedo reticularis. Where the condition exists against a background of chronic inflammatory disease, especially systemic lupus erythematosus (SLE), it is referred to as secondary antiphospholipid syndrome to distinguish it from the primary syndrome, in which there is no evidence for another relevant underlying disease. An international workshop recently reported on a consensus statement on classification criteria for antiphospholipid syndrome (Wilson et al, 1999). These stringent criteria are intended only for use in the context of clinical and scientific investigations of APS. The range of disease associations with antiphospholipid antibodies is extremely broad (Table I). The mechanisms underlying the prothrombotic state in APS have not been clarified Greaves, 1999). Whether antibodies to β2-GP-I/phospholipid are causal is unproven. A range of pathogenetic mechanisms has been reported in subjects with APS, including protein C resistance, vascular endothelial autoimmunity and activation and impaired fibrinolytic capacity. Furthermore, although thrombosis in the uteroplacental vasculature has been implicated in the pathogenesis of miscarriage, placental infarction is not a universal finding and non-thrombotic mechanisms may be involved, such as failure of implantation or autoantibody binding to the trophoblast. Because pathogenicity of the autoantibodies has not been conclusively demonstrated, it remains possible that, in at least some cases, aPLs are surrogate markers for a prothrombotic syndrome with a multifactorial, possibly autoimmune, pathogenesis. Nevertheless, their detection is frequently diagnostically and prognostically useful. Subjects with APS present to a wide range of clinical specialists and a multidisciplinary approach to investigation and management is often appropriate. This may involve, for example, clinical haematologists, neurologists, cardiologists, rheumatologists, dermatologists and obstetricians. Clinical presentations most commonly leading to the diagnosis of APS are venous thromboembolism, arterial occlusive events and recurrent miscarriage, but antiphospholipid antibodies may also be detected incidentally in healthy subjects (Table I). aPLs occur in relation to use of some drugs, particularly chlorpromazine, and transiently after certain infections. Persistent antiphospholipid antibody positivity may be a result of chronic infection, for example in cases of syphilis, hepatitis C and HIV infection. Familial APS has been reported but is uncommon. In a significant proportion of subjects, the detection of aPLs is incidental in apparently healthy individuals. The thrombotic risk associated with incidental aPL positivity appears to be relatively low (Finazzi et al, 1996), although moderate to high titre aCL predicted future venous thrombosis in one epidemiological study (Ginsburg et al, 1992). A clinical assessment is indicated to exclude evidence of SLE, infection and the use of relevant medications (especially chlorpromazine). Objective confirmation of thrombosis is essential as immediate and long-term management decisions may rest on the diagnosis. This is particularly important in women of child-bearing age. In APS, limb deep vein thrombosis, with or without pulmonary embolism, is most common. Thrombosis in unusual sites, including the cerebral venous sinuses and intra-abdominal visceral veins, is an occasional feature. Clinical assessment should include a search for additional risk factors, present in over 50% of instances, and a detailed and for features of APS or and aPLs may not a thrombotic the is relevant in the assessment of positive tests and clinical the diagnosis of APS may the of anticoagulant because of the high risk of testing for antiphospholipid antibodies at with venous and the of anticoagulant may be However, this is frequently not in the and anticoagulant should not be for the of tests for antiphospholipid antibodies. may not be possible to a diagnosis of antiphospholipid syndrome anticoagulant has been This is because of persistence of positive tests is and this may not be in an only laboratory of APS is lupus is most often at a age. cerebral vascular may be and transient are a feature. A high of recurrent and events is arterial with is common. Although an of aPL positivity has been in of infarction and to thrombosis as a of antiphospholipid syndrome appears to be common Because of the high risk of recurrent events (Finazzi et al, and the of anticoagulant early diagnosis is for Clinical assessment should include a for evidence of and and for thrombotic events of livedo in subjects with is frequently associated with may be a feature. including also occur in APS, but may only be on Additional risk factors for arterial thrombosis should be and A major of APS is recurrent miscarriage or in or more The of antiphospholipid antibodies among women with recurrent miscarriage has been reported to be and This may be for the of of laboratory to antiphospholipid the of women with transiently positive test and a approach with testing for both LA and and the of positive tests for aPLs is et al, In the women with recurrent miscarriage due to APS, the may be as high as et al, In the of positive tests in women of child-bearing is and positive tests are not of in women with no of et al, 1991). women with recurrent antiphospholipid antibodies only early have both early and or for antiphospholipid antibodies is to all women with recurrent miscarriage, especially as is may also be an aPLs and other especially placental and early of are in Although thrombocytopenia is frequently present in APS, is uncommon. The low is due to an to that in thrombocytopenia as in the antibodies are epitopes on and are from antibodies et al, Furthermore, aPLs have been reported in of subjects with In the condition of antiphospholipid there is failure due to The is Because of the high risk of thrombosis and miscarriage and the for positive aPL tests to can be in a wide range of In relation to venous thromboembolism, all subjects with apparently events should be for The of positive tests is to be in with and venous thromboembolism, in the of other risk factors, may be an for testing for antiphospholipid antibodies. Subjects with and with arterial occlusive events at a example should be for especially when risk factors for arterial disease are not The can be for subjects are and are not other risk factors such as or Where recurrent arterial occlusive events occur APS should be In subjects with SLE, aPLs should be as of the assessment of the autoantibody as risk of thrombosis is in with aPLs and the finding may the use of at of Because miscarriage is a common for aPLs is not after a In women with or more testing for aPLs should of the including and is possible that for aPLs be to include women not the criteria for APS but have miscarriage, as or or more of in the or may be an for testing for should also be to the possible diagnosis of APS in women with early or placental in Because antiphospholipid antibodies may be et al, tests are or early in when A proportion of women with aPLs also have antibodies. detection is important as it is associated with a risk of The diagnosis of APS relies on the of the of LA tests or aPLs solid phase The use cardiolipin as the but some use using purified β2-GP-I in the of phospholipid have been but the of assays in diagnosis and management has not been some evidence that the of β2-GP-I antibodies more aCL with thrombotic clinical events in APS et al, et al, et al, et al, the traditional tests the of laboratory investigation of APS and it is clear that both LA and solid assays must be for the detection of aPLs in certain as cases with LA but no aCL and are on one of assay may to aPL In both (Machin et al, and international et al, guidelines for the detection of LA have been A of tests and have been and a of and have been Despite criteria for the of LA of a of an demonstrated of the nature of the additional criteria are the failure to against a and of with LA These the of most clinical LA may also result in and low in In the laboratory tests should use a detection or and a confirmation (Table the often using in the of The must be as as possible to the confirmation to positive both of which may have important clinical LA test and because of the nature of more one test should be for detection of of the and tests is important as the in the of LA some in assays and should be or are is to of in subjects with of the due to In common with the and of for other venous and immediate are should be prepared of at at for with and other must be as the of particularly after plasma This may be in the plasma a and the for at in a which or a A of must be These also to the of and must be for LA and of This should be the LA test on at least plasma from healthy Where are the of plasma is the and the range plasma for the of LA must include from at least healthy The is frequently as the test for is in the of in of the and phospholipid The of the phospholipid of the to be in LA and in both the of phospholipid present and in their et al, This in of the test et al, 1991). The phase and are associated with of and which to the and a weak lupus a is to exclude LA and additional tests must be A of of the have been for use as tests for including the use of et al, of the with and et al, and phase which the et al, In the et al, which in in the of and leading to the of a the is to a and the phospholipid is that of the phospholipid LA in a A of are now but in and and is the of et al, 1999). is essential to use a range for and and use some of or confirmation to the the of for the with phospholipid is and A is in the guidelines (Machin et al, 1991). are the must be to a for the in use The of phospholipid must then be to and of the the is the test may be to weak LA and may be with in an on the The of the phospholipid is also important and The can be the with a plasma and a known The of phospholipid that to a with plasma and a significant with LA plasma should be The may then be in at or the may be for future tests with the of must be and against the The of the also and some at a for LA that have been for LA in the of phospholipid and that a phospholipid is the test to LA et al, The can be improved the use of tests a confirmation using a in a but also the of The is improved although the of an additional test using a has been et al, is an purified from the of which but with no for phospholipid. Thus, the test is not the of LA and the be with but with a high These tests may be additional in cases but are not as In the no additional phospholipid is The for this test are that the of the et al, to LA is to on the of and plasma on this the is particularly to of the plasma which the of the especially after and LA is when the to after relatively of plasma are in the is with of The use of of plasma is and to is that and are as as one A an and a of should be as of LA a and is However, for and should be In a of of the plasma with and the The is not for all of particularly some to the nature of the which to the and to A of the using of has been et al, and this has with a that it is for The and for LA to be to tests using One common to both the and is the for most factors to be present in which that is or the of and The test The is in with LA and are or the is low because of the of a high of which to the of However, when the is the phospholipid a and of the LA causes a of the This has been in the test et al, for this test are that the is using or more and the of test to with the is A in the with is of The and of the phospholipid of Thus, the and range and Although the test not in a (Machin et al, it has recently been demonstrated that a with phospholipids is particularly to LA et al, et al, of The of an can be the of tests with In the is the of of but in the of LA relatively of plasma must often be to the The of plasma may the a phenomenon known as the lupus A 50% with plasma is However, tests result in of the LA antibody and this can to Nevertheless, the approach is with moderate and the of LA tests is of phospholipid The is commonly are with or and These are to plasma and their is with a to LA the mechanism of the is not and the in the remains This has been for confirmation of LA in the the and the for the of the are A test on has also been et al, and may be useful. An for the confirmation of LA a phase which bind LA and on the In the an to the is the use of a which a high of phospholipid or an are can the of and it is to for of the are in it is essential to to in relation to the and phospholipid reagents. for the of in the have been reported and there is no consensus on the of The of plasma for the phospholipid and for the or are is to A result the range with which to the or with the high phospholipid or is of of The is from the test and the is the to a for phospholipid and for the The is then as In of are of The with phospholipid is the with the A the the of is to as of and the are to the range is evidence for the of The use of the not in the with the in or in and be for The of and the of the most although the is more 1999). for LA in subjects with detection of LA may not be possible in a subject with One approach is to the on of and test In this the should also be on a plasma to because of the of LA A clear positive result may be but an or result is and may be particularly diagnostically in plasma from as are relatively to et al, et al, no other is in the of the in anticoagulant remains However, a test or confirmation is their is is and to test for LA on plasma from a subject of although some include which The of an is important in all LA of the and the of reagents. A is the of known and in of and in at may be are but many are and LA may which their and have been prepared in the British for in Haematology and the for and British for LA from The assessment of LA tests the of plasma to all have in the of LA et al, phase assays for such as the aCL have been et al, The testing and the are not or the use of are the of the cardiolipin and the for the The binding of most aCL the of the protein β2-GP-I, which is essential for and aCL This is serum or serum in the to binding to the The use of as is essential to the of binding due to in the test assays on other phospholipids have been but to have to aCL assays and is that the aCL test is the use of et al, or secondary from These the of aCL in or antiphospholipid and to a of aCL and should also be in of An is in the from The detection of aCL the diagnosis of APS in a subject with an clinical when LA is However, the aCL assay is not a for the LA test it that LA is present because antibodies to be for the Furthermore, the clinical of aCL is Thus, in cases the aCL titre is and tests for LA are a diagnosis of APS may not be it is particularly important to other causes of thrombosis or is also over the of the aCL antibodies may be more although aCL appears to be associated with thrombotic events and miscarriage in some for antibodies may not be et al, and their use is not A of assays for β2-GP-I antibodies have been et al, and are β2-GP-I antibody assays and with the in APS and assays for aCL and are to transient positive in with infection. The of or other with a high of must be to β2-GP-I binding. β2-GP-I which are or have should not be are aCL but are β2-GP-I antibody positive. This is due to antibodies which to bind to β2-GP-I in aCL These antibodies are significant because may be associated with APS. β2-GP-I may other including reactive with annexin V, protein C, protein or These may be the or the test antibodies for venous thrombosis and recurrent and may be in with infection. precise clinical is not One has an with infarction et al, but more is to the clinical of this A for LA a which to is for confirmation of a positive test may with to the as the test are other the or the must be to A should be to or anticoagulant The may be as a test for A must be the is of tests with plasma a is frequently A test should also be the or the with a in tests are tests using may be of other tests or the for and aCL possibly for β2-GP-I must also be The of aPL must be in at least to should be that the finding of a positive test for aCL in subjects with deep venous thrombosis of an risk of recurrent thrombosis in the study of et in in whom tests were In subjects of LA may not be The and on of and or the and may be useful. are to that management of thrombosis and miscarriage in APS. a low risk of thrombosis in this although some an of risk over background (Ginsburg et al, 1992). is not indicated when there has been no thrombotic it is that this subjects with APS have not a a low for use of at of high risk is of is anticoagulant for example in and after in the with positive tests is not at present as the associated risk of may that of is not known the risk of thrombosis associated with use of the or is in the of aPL in an In the of decisions the use of and in women with incidental aPL can only be on other such as the and of and the of additional thrombotic risk The management of the with or low molecular weight is not the diagnosis of APS should be in the with a international of range is over the for more anticoagulant the of and the of an risk of thrombosis at of and a of has been et al, are no and of should be on an the of additional risk factors, the of the and the risk of on C should be in that of of venous are although significant from syndrome in of with limb deep vein thrombosis, a significant proportion of et al, 1999). the of in subjects on a is at least et al, and this when the is more common in the Despite over there is evidence that long-term with may be in some subjects with venous and also that the thrombosis may be particularly high in APS et al, et al, 1999). The of long-term has not been for many subjects with deep vein thrombosis in antiphospholipid syndrome, for to at of and management of additional risk factors is venous thrombosis should be long-term the is and should to more range but this is uncommon. of the and of are in women with aPL and thrombosis, at least in not on but risk of thrombotic Although have been in women at risk of venous thrombosis, are is not of to on with should be C Because of the high risk of and of or due to cerebral infarction in APS should be with long-term anticoagulant range has been and may be in some cases Haemostasis and Thrombosis Task but of are the use of a of can be on the use of a Whether additional with is in this is not but the risk of is when is anticoagulant arterial of APS also of of long-term with in many in the of venous thrombosis, should be to the and of additional thrombotic risk The use of the to may be because of of the due to of an assay this with low molecular weight the for in most cases, is also an has been that the may be in subjects with LA on as the of may be because of an of the LA on the et al, Although this phenomenon appears to be more when certain are in the of other have also been The clinical of this phenomenon is not but evidence that the is when with a low international and for the and is in et al, thrombocytopenia is the only and especially the clinical is one of a the management should be to that for has been and and is indicated et al, Although the of thrombocytopenia is in APS, of are with may an risk in this but should be thrombosis is the clinical In women with APS and a of there is a for A of including and have been as or in in an to the of in women with antiphospholipid antibodies et al, et al, 1996). are the of in which have also criteria and and the of of laboratory assays to antiphospholipid antibodies. The use of in is associated with significant and et al, et al, appears to be and should be on the of the only clinical reported to et al, with and is to women with a of recurrent miscarriage associated with antiphospholipid antibodies should be as as the test positive. Because the of occur of and in of may the of a should be when is on In the study of et of low molecular weight in is for but the in use is not no low molecular weight is for use in but low molecular weight are to et al, 1999). The of has not been but should not be because of the risk of at is a in women with no of thrombosis in whom early has been a feature. have of to is and also be indicated in women with a of an additional thrombotic risk and is of the should be A for the and is including early detection of In women not the criteria for APS but have aPL and a or the of with and in has not been Although there is often to the for must be especially in of the of at this Although use of and appears to in APS, are frequently and et al, 1999). management the and the and to clinical and laboratory with antiphospholipid antibodies have a high of with of et al, et al, the for and to a understanding of the pathogenesis and clinical features of thrombotic disease is APS is a and condition which is with to both diagnosis and is a clear for the pathogenetic mechanisms involved, for the development of more laboratory to at risk of thrombosis and miscarriage and for the of in of The are to the the of the