PHOSPHOLIPID-BINDING ANTIBODIES AND RECURRENT PREGNANCY LOSS
P. de Moerloose
In the last few years, numerous publications have described a possible new cause for recurrent pregnancy loss, the so-called phospholipid-binding antibodies (Plb-Ab). In this chapter, we will first review the antibodies and the antigens towards which they are directed, and describe the tests used to detect Plb-Ab; next we will focus on their association with repetitive pregnancy loss and the possible mechanisms involved; lastly we will deal with some therapeutic implications.
Antibodies and antigens
Plb-Ab are heterogeneous immunoglobulins, the most common of which are IgG (all subclasses) and IgM (12,21). IgA was thought to be of less importance, but some recent publications have renewed interest in this isotype (16).
Until recently, Plb-Ab were thought to be directed towards negatively charged (anionic) phospholipids, the main ones being phosphatidylserine and phosphatidyldiglycerol (cardiolipin). But a non-anionic phospholipid, phosphatidylethanolamine, when present in an hexagonal phase, can also be an immunological target for Plb-Ab (14).
The concept that these antibodies are specifically directed to phospholipid epitopes has been recently challenged. Proteins, namely prothrombin and ß2-glycoprotein 1, are now considered as essential cofactors for the generation of Plb-Ab (1,9,19,20), and other new cofactors such as protein S have been identified. These proteins have phospholipid-binding capacities and when they form complexes with phospholipids, cryptic epitopes may appear, which could lead to the formation of autoantibodies such as Plb-Ab. Some authors have proposed to rename these Plb-Ab " anti lipid-bound protein antibodies ".
In summary Plb-Ab include heterogeneous populations of antibodies directed towards different combinations of phospholipids and phospholipid-binding proteins. In a given patient more than one type of Plb-Ab can be detected.
Assays for phospholipid-binding antibodies
Three different types of tests can detect Plb-Ab: flocculation (VDRL), coagulation and immunological assays. We will focus on the last two types of tests, the first one (false VDRL positive reaction) being less sensitive.
Bearing in mind a) that phospholipids are essential for the coagulation cascade and b) that prothrombin and phospholipid which form complexes are potential targets for Plb-Ab, interference of Plb-Ab in the coagulation cascade may lead to lengthening in clotting times. The usual test which detects these antibodies is activated partial thromboplastin time (aPTT), but as various causes (heparin, factor deficiencies and antifactor antibodies) can induce a prolongation of aPTT, other tests are necessary to confirm that this lengthening is due to Plb-Ab. Moreover aPTT is not a reliable test during pregnancy since it can be shortened because of increases in coagulation factors. A three step-procedure is usually performed: first a screening test (for example aPTT), then mixing studies to demonstrate that the abnormality in the first step is due to an inhibitor, and confirmatory tests to prove that the inhibitor is non factor-specific and phospholipid dependent (7,31).
Antibodies detected by coagulation tests are called Lupus-Like Anticoagulant (LLA) since a) they are frequently observed in lupus and b) they induce an anticoagulant effect in vitro. However this is a misnomer since a) diseases other than lupus can be associated with LLA and b) in vivo LLA usually does not induce bleeding, but thrombosis.
The most commonly performed assay is an ELISA in which the coated antigen is cardiolipin, hence the name of anticardiolipin antibodies (aCL). This assay is relatively simple to carry out, but there are still several unresolved problems which preclude a standardization of the assay. The main issues are the coated antigens, the choices of the protein’s buffered solution, the subtraction of blanks and the available standards (6,25,26). This makes the interpretation of the data sometimes difficult. Many authors agree that the presence of high titers of IgG is of greater clinical significance in identifying women at risk of pregnancy loss than IgM aCL.
Which assay to perform?
Both tests can detect different Plb-Ab in the same patient, so even if there is an overlap, both assays (coagulation and immunological) need to be performed. It is not clear whether LLA or aCL is more specific for fetal loss. It is well to bear in mind that various antibodies and various antigens are involved in this heterogeneous syndrome, and that with the usual assays some patients having Plb-Ab will be missed. On the other hand positive tests should be confirmed at least twice, more than three months apart, as transient Plb-Ab have less clinical significance (11).
Recurrent pregnancy loss
The two main clinical conditions associated with Plb-Ab are thrombosis and recurrent pregnancy loss (17). It is possible that both are linked, as pregnancy loss could be due to thrombosis in the placental vessels. Concerning pregnancy loss two situations will now be considered: a) the importance of Plb-Ab in systemic lupus erythematosus (SLE) or related disorders and b) the importance of Plb-Ab in non-lupus women.
Plb-Ab and SLE
There is general agreement that LLA or aCL in SLE women are linked with recurrent pregnancy loss. In a review of the literature (17) it was found that the incidence of pregnancy loss was 60% and 59% for SLE patients with LLA or aCL, respectively, and 13% and 5% for SLE patients without LLA and aCL, respectively. Pregnancy loss can occur in the first, second or third trimester of pregnancy. Even in the case of a successful pregnancy Plb-Ab have been associated with other unfavourable outcomes such as preeclampsia, HELLP syndrome, chorea gravidarum, intrauterine growth retardation, stillbirth, preterm births, and thrombosis (30).
In a recent study (11), 42 consecutive SLE patients with 122 pregnancies, and with a history of at least one pregnancy were studied; 11 had a history of one previous pregnancy loss and seven had a history of two or more episodes of pregnancy loss. A history of pregnancy loss occurred in 24 of 55 pregnancies in patients with persistent LLA compared with 8 of 67 pregnancies in LLA-negative patients (odds ratio [OR] 5.7, 95% confidence intervals [CI] 2.3 to 13.9, p <.0001). All seven women with multiple pregnancy loss showed persistently positive coagulation tests. Women with transiently positive LLA did not show a significant association. Similar results were found for aCL: 10 of 14 pregnancies in women who were persistently positive resulted in pregnancy loss compared with 22 of 108 pregnancies in patients who were negative or transiently positive (OR 9.8, 95% CI, 3.0 to 32.4, p <.002). This study also showed that a combination of coagulation tests for LLA plus testing for aCL on more than one occasion is superior to a single test in attempting to correlate the laboratory results with clinical events. This study was retrospective however and we still need prospective studies to establish a temporal relationship between the diagnosis of SLE and the development of Plb-Ab with fetal wastage.
Plb-Ab in non SLE-women
In women without lupus, the possible association with pregnancy loss is less clear. Indeed the prevalence of Plb-Ab in non-lupus women with recurrent pregnancy loss varies between 3 and 50% (17). There are several explanations for these discordant results, the main ones being patient selection bias and lack of standardization of the tests. An important case-control study (23) has recently evaluated the prevalence of LLA and aCL in 220 women with two or more unexplained spontaneous abortions and without immunologic disorders as well as in 193 controls. LLA was detected in 7% of cases, but in none of the 193 controls (p <.001) and aCL in 19% of cases versus 3% in controls (OR 9.1, 95% CI 3.5-23.6). Another report (24) has described Plb-Ab at a statistically greater prevalence in 81 women with recurrent spontaneous abortion (16%) than in 88 women with successful pregnancies (7%) or in 64 women who had never been pregnant (3%). In another recent study (22) however, when women with lupus-like disease, or a history of thrombosis were excluded, no significant difference in the prevalence of anticardiolipin between patients and controls was found.
When is it necessary to search for antibodies?
In a woman with SLE or SLE-related disorder, Plb-Ab should be investigated, whichever her past gynecological history. Indeed, because of the close association of Plb-Ab with an unfavourable outcome, preventive measures should be taken from the first pregnancy.
In a woman without lupus but with recurrent pregnancy loss (three consecutive fetal losses), and without any other factor, these antibodies should also be looked for. Positive tests may indicate a subclinical autoimmune disease, a primary antiphospholipid syndrome (Pl-bAb linked with repeated pregnancy loss, thrombosis and thrombocytopenia), or a not yet developed clinical autoimmune disease. It is well to bear in mind the particular role of gynecologists and obstetricians who perform routine medical evaluations on the highest risk group for the onset of autoimmune diseases, i.e. women between the age of 20 and 40.
In a woman with a first pregnancy loss we would, as others (13), not consider performing this test routinely. However, a distinction should be made between a spontaneous abortion in the first trimester (a relatively common event) and a fetal death in the second or third trimester, which is unusual (<4% of recognized gestations). We believe that Plb-Ab should also be studied after a single episode of pregnancy loss in women a) having unexplained episodes of thrombosis or thrombocytopenia, b) with symptoms suggestive of an autoimmune disease and c) with a family history of autoimmune disease.
In women with two consecutive pregnancy losses the question is open; we think that these tests are worthwhile to be considered if no other explanations are found for repetitive pregnancy loss.
In spite of considerable interest in these autoantibodies, the mechanism(s) implicated in the association between Plb-Ab and fetal wastage remain unclear. There are still doubts whether Plb-Ab have a pathogenic potential on their own, or whether they are epiphenomena. The cause of fetal wastage is thought to be thrombosis (or eventually vasculopathy) in placental and decidual vessels. A large number of potential pathological mechanisms inducing thrombosis have been proposed for Plb-Ab such as increased tissue factor generation, inhibition of the thrombomodulin-protein C pathway (the placenta is rich in thrombomodulin), interference with the prostacyclin-thromboxane balance, inhibition of the fibrinolytic system and so on, as well as combinations of them. Based on the clinical and laboratory heterogeneity of Plb-Ab, it is logical to assume a similar diversity for the pathophysiology of pregnancy loss.
Some interesting animal models have been recently developed which may help us in the future to have a better understanding of the pathogenic mechanisms of Plb-Ab (2,3,10).
Due to the numerous uncertainties mentioned, it is not astonishing that there is little agreement among clinicians on the best treatment for women with Pl-binding antibodies. The first successes were reported with high-dose corticosteroids (40 mg daily) and low-dose aspirin (18,30). Not all reports using this regimen or aspirin alone have shown efficacy however (15) and long-term high-dose steroids have significant adverse effects. To decrease the side-effects, lower doses of prednisone (5 mg daily) can be administered; a recent prospective study on 11 patients reported a 100% success rate (29).
Other treatments have been evaluated. Rosove et al (28) have successfully prescribed adjusted full-dose heparin in 14 women with 28 previous fetal losses; of 15 pregnancies 14 resulted in live births. In some anecdotal reports intravenous IgG, plasmapheresis, immunosuppressive therapy, either alone or combined, were given with some success (4,8,27).
The only randomized study (5) has compared heparin versus steroids, both combined with aspirin (80 mg daily): results indicated that the association of aspirin-heparin was as effective as aspirin with steroids (75% live birth rate), but with significantly fewer side-effects.
Due to the maternal risk of thrombosis during pregnancy, treatment with aspirin or prophylactic heparin may be advised, regardless of concern for the fetus.
Phospholipid-binding antibodies are now recognized as a risk factor for pregnancy loss in women with SLE and SLE-related disorders, but their importance is less clear in non-SLE women. To improve our knowledge in this fascinating area we need to define the immunological targets of these antibodies. When this is achieved, more appropriate tests will probably be available, and the possible role of these antibodies as a cause of pregnancy loss will be clarified. Large-scale randomized prospective trials will then be necessary to offer the best therapy to pregnant women with phospholipid binding antibodies.
Edited by Aldo Campana,