Antiphospholipid antibodies: a mosaic of pathogenic effects?

Comment on Rand et al, page 2783

The elucidation of the pathophysiology of aPL antibodies has attracted considerable interest in the last few years. Several mechanisms have been proposed. There is now evidence that aPL may induce thrombosis by interfering with the anticoagulant properties of annexin A5.

There is convincing evidence from studies that used animal models of thrombosis, endothelial cell activation, and pregnancy loss that antiphospholipid (aPL) antibodies are pathogenic in vivo. However, the mechanisms by which these antibodies mediate disease are only partially known and our understanding is limited by the apparent polyspecificity of the antibodies, the multiple potential end-organ targets, and variability of clinical context that disease may present. In fact, in vitro studies have reported that aPL may cause thrombosis by interfering with activation of protein C or inactivation of factor V by activated protein C, inhibiting endothelial prostacyclin production, or exerting stimulatory effect on platelet function. Reports have also shown that aPL antibodies cause up-regulation of adhesion molecules (intercellular adhesion molecule-1 [ICAM-1], vascular cell adhesion molecule 1 [VCAM-1], endothelial selectin [E-SEL]) and increased expression of tissue factor and induce a procoagulant/proinflammatory state in endothelial cells that precede thrombosis.^1,2  In platelets, aPLs induce phosphorylation of p38 mitogen-activated protein kinase (p38MAPK) and phospholipase A2 (cPLA2) and thromboxane production with subsequent platelet activation in the presence of low doses of thrombin.³ FIG1 

One new mechanism of action for aPL has recently been proposed. aPL antibodies may cause thrombosis by disrupting the binding of the natural anticoagulant protein annexin A5 (also called placental anticoagulant protein or annexin V) to phospholipids and hence hamper its anticoagulant properties. In an elegant and well-designed study published in this issue of Blood, Rand and colleagues conclusively show that aPL antibodies, present in the plasma of patients with antiphospholipid syndrome (APS) and thrombotic events, reduce both annexin A5 binding to phospholipids and annexin A5 anticoagulant activity. The authors used purified polyclonal and monoclonal aPLs and plasma from patients with APS and controls to confirm these important observations. The investigators concluded that these effects explain, at least in part, how aPLs not only cause thrombosis in any artery or vein of the body but also may be a mechanism of clotting in the placenta of patients with APS, leading to fetal distress and death, since annexin A5 is expressed not only in endothelial cells but also in trophoblast surfaces.

Interestingly, although the study was carried out with a small but well-characterized group of patients, the methods used by Rand and colleagues enable one to distinguish between APS patients with previous history of thrombosis and those with APS and no thrombosis and also controls (patients with thrombosis but without APS). Diagnosis of the APS is currently based on a persistently positive anticardiolipin (aCL) and/or lupus anticoagulant (LA) test and 1 of the 2 major clinical manifestations of the syndrome (thrombosis or fetal loss). Correct identification of patients with APS is important, since prophylactic anticoagulant therapy can prevent thrombosis from recurring and treatment of affected women during pregnancy can result in live births. Since there are many causes of thrombosis and of pregnancy loss, the confirmation of diagnosis of APS is often dependent on a positive aCL or LA test. The aCL is a sensitive test but lacks specificity and is frequently found positive in patients with infectious diseases and other nonrelated autoimmune disorders. LA, on the other hand, is a multi-step assay that requires special care in the processing of samples, the selection of reagents, and the interpretation of the results. It is known to be a more specific test for diagnosis of APS but lacks sensitivity. Furthermore, neither the aCL nor the LA test distinguishes between patients with APS and thrombosis from those with APS without thrombosis. If confirmed in prospective studies using a larger number of patients, the study of Rand and colleagues opens the possibility for evaluation of a new simple 1-step coagulation assay for diagnosis of APS that will also enable the distinction of these 2 subgroups of patients. This may have significant implications in the prognosis and possibly in the treatment of patients with APS.

The nature of the target antigen(s) recognized by aPL antibodies has been the subject of considerable debate over the last 10 years. Reports have challenged the initial belief that aPLs bind negatively charged phospholipids. While the majority of the investigators now agree that aPL antibodies bind to β₂glycoprotein I (β₂GPI) or to new cryptic epitopes exposed on β₂GPI, some others indicate that aPL may bind to prothrombin, protein C, protein S alone, and/or complexes of some of those proteins with phospholipids and hence interfere with the phospholipid-dependent coagulation pathways. In addition, some studies have shown that aPL antibodies may bind to β₂GPI directly on endothelial cells to induce their activation, while others have indicated involvement of annexin II.⁴  A recent publication has demonstrated that aPL antibodies bind toll-like receptor (TLR), a receptor for bacterial endotoxin, present in endothelial cells and induce intracellular signal transduction.⁵  This study provides convincing evidence that negatively charged phospholipids are part of the target antigen recognized by aPL antibodies, since they displace the binding of annexin A5 to negatively charged phospholipids (particularly phosphatidylserine) and hamper its natural anticoagulant effect.

In conclusion, the study by Rand and colleagues brings new light into 2 major areas of interest in the field of APS. It clearly defines a new pathophysiologic mechanism of action of aPL antibodies and opens the possibility to use these findings in designing a new, and possibly better, test for diagnosis of APS.