Antiphospholipid syndrome: pathogenesis and a window of treatment opportunities in the future

Authors


Imad W. Uthman, MD, MPH, Department of Internal Medicine, American University of Beirut Medical Center, PO Box: 11-0236, Riad El-Solh 1107, 2020 Beirut, Lebanon. Tel.: +961-3-379098; Fax: +961-1-744464;
e-mail: iuthman@aub.edu.lb

Abstract

Eur J Clin Invest 2010; 40 (5): 451–464

Abstract

Background  Antiphospholipid syndrome (APS) is a systemic autoimmune vascular disease characterized by recurrent thrombotic episodes and/or obstetric complications. Management of this disease has been restricted mainly to anticoagulation; however, in recent years, significant advancement has been made in elucidating the pathophysiology of the disease including antiphospholipid antibody (aPL)-induced activation of the platelets, endothelial cells, monocytes, complement and coagulation cascade. Stemming from these advances, potential targeted therapeutic approaches have been proposed.

Materials and methods  We utilized a computer-assisted search of the literature (MEDLINE, National Library of Medicine, Bethesda, MD, USA) up until September 2009 using the keywords: antiphospholipid syndrome, antiphospholipid antibodies, anticardiolipin antibodies, lupus anticoagulant, anti beta-2 glycoprotein antibodies, complement system, tissue factor, p38 mitogen-activated protein kinase (p38 MAPK), nuclear factor kappa B, toll-like receptors, annexin, Rituximab, statins and tumour necrosis factor.

Results  Several study groups have separately demonstrated the importance of inflammatory mediators in the pathogenesis of APS. It was also established that tissue factor, MAPK, nuclear factors kappa B, and the complement system are integral to the disease process. Toll-like receptors and annexin have as well been associated with the disease pathophysiology. Some study groups proposed new targeted therapeutic strategies some of which have shown promising results in preclinical studies. These include Rituximab, complement inhibition, anti-cytokine therapy, p38 MAPK inhibitors, nuclear factor inhibitors and tissue factor inhibitors.

Conclusion  As more insight is being gained into the pathophysiology of APS, newer therapeutic strategies are being proposed that might lead to safer and more efficacious treatment modalities in the future.

Introduction

Antiphospholipid syndrome (APS) is a systemic autoimmune disorder that is characterized by the presence of antiphospholipid antibodies (aPL) in the serum of patients and a generalized prothrombotic state [1]. This state is associated with significant morbidity and mortality form vascular thrombotic events and an array of obstetric complications (Table 1).

Table 1.   The Updated International Preliminary Classification Criteria for APS [1]. One clinical and one serological criteria needed to make the diagnosis
Clinical criteriaLaboratory criteria
  1. APS, antiphospholipid syndrome; ELISA, enzyme-linked immunosorbent assay.

Vascular thrombosis (arterial/venous/microvascular)Lupus anticoagulant present in plasma on two or more occasions at least 12 weeks apart, detected according to the guidelines of the International Society on Thrombosis and Haemostasis
Obstetric morbidity
 One or more premature birth of normal neonate before 34 weeks of gestationAnticardiolipin antibody (IgG and/or IgM isotype) in serum or plasma, present in medium or high titres on two or more occasions at least 12 weeks apart, measured by a standardized ELISA
 One or more unexplained death of a normal foetus at or beyond 10 weeks of gestationAnti-β2 glycoprotein I antibody (IgG and/or IgM isotype) in serum or plasma, present on two occasions at least 12 weeks apart, measured by a standardized ELISA
 Three or more unexplained spontaneous abortions before 10 weeks gestation 

As a result of the apparent prothrombotic state provoked by aPL, anticoagulation and anti-platelet medications have been logically the mainstay of the treatment of this disease. However, several dilemmas exist regarding the optimal management of the different categories of APS patients. Many authorities advocate indefinite high intensity anticoagulation with an international normalization rate (INR) higher than three for APS patients who develop an arterial thrombotic episode; this option, however, might be limited by bleeding complications. Moreover, recurrence despite this high intensity anticoagulation is not a rare event. As for APS patients with index venous events, most field experts recommend indefinite moderate intensity anticoagulation (INR:2–3). Again however, recurrence despite this seemingly optimal therapy does occur. Similarly, for APS patients with obstetric morbidity, despite the now widely accepted and literature supported treatment algorithm for pregnant APS patients that includes aspirin and heparin, a lot of patients still experience obstetric complications while on this regimen. All these dilemmas indicate that our currently available treatment modalities and management algorithms for APS are far from being optimal or ideal.

Our understanding of the disease’s aetio-pathophysiology has been steadily widening. Reliable animal models for APS have been successfully produced in the early 90s via immunization with various human aPL [2]. These animal models have been available ever since and served as an important element in the quest to elucidate the disease’s pathophysiology. Over the last decade, there has been a remarkable advancement in the understanding of the aetio-pathogenetic basis of this syndrome. Several studies have been published trying to uncover the basis of aPL pathogenicity including the cellular components targeted, the systems affected, the receptors involved, the intracellular cascades utilized, as well as the effector molecules altered in the process (Fig. 2). With this newly gained insight, the impact of inflammation in the pathogenesis of the disease has been appreciated to the extent that APS is now widely described as an inflammatory pro-thrombotic disorder. Stemming from these advancements, our repertoire of therapeutic options could expand to encompass agents that block the inflammatory component of the disease. Based on the uncovered aetio-pathogenesis so far, new potential targeted therapeutic agents have already been proposed to represent a safer and more efficacious treatment modality for APS patients (Fig. 3). In fact some of these agents have already been tried in preclinical trials and attained promising preliminary results.

Figure 2.

 Schematic drawing of the patho-physiological processes involved in aPL-mediated thrombosis. aPL activate the complement system that ultimately generates C3a, C5a and MAC. The antibodies activate platelets through apoER2. This activates the p38MAPK/cPLA2 pathway that leads to TXB2 generation. TXB2 leads to platelet activation and release and leads to upregulation of GPIIbIIIa. aPL also activate endothelial cells and monocytes through TLR-4, annexin A2 and heparan sulphate moiety of the cell membrane. This activates the p38MAPK-NF-κB pathway ultimately leading to upregulation of adhesion molecules and tissue factor and release of pro-inflammatory cytokines. Abbreviations: MAC, membrane attack complex; C5a, complement protein 5 activated; C3a, complement protein 3 activated; C5aR, C5a receptor; C3aR, C3a receptor; p38MAPK, p38 mitogen-activated protein kinase; NF-κB, nuclear factor kappa B; TLR-4, toll-like receptor 4; aPL, antiphospholipid antibodies; CAM, cell adhesion molecule; TF, tissue factor; IL-1,6,8, interleukin 1,6,8; TNF, tumour necrosis factor; apoER2, apolipoprotein E receptor 2; cPLA2, cytosolic phospholipase A2; AA, arachidonic acid; TXB2, thromboxane B2; GPIIbIIIa, glycoprotein IIbIIIa. Solid black arrows represent activation of; dashed black arrows represent possible alternate activation pathway of.

Figure 3.

 Schematic drawing of the patho-physiological processes involved in aPL-mediated thrombosis and the possible sites of action of future targeted therapeutic agents. This includes complement inhibitors (Anti C5a antibodies), blockers of aPL binding to its target cell (anti-annexin A antibody, TLR-4 mutations, TIFI, HCQ), p38MAPK inhibitors (SB203580), NF-κB inhibitors (MG132), inhibitors of TF expression (ACEI, statins, dilazep and defibrotide), inhibitors of expression of adhesion molecules (statins), anti-cytokine agents (statins, anti-TNF agents and anti IL-6 agents), specific inhibitors of GPIIbIIIa (abciximab and HCQ), and at the level of production of aPL (Rituximab). Abbreviations: MAC, membrane attack complex; C5a, complement protein 5 activated; C3a, complement protein 3 activated; C5aR, C5a receptor; C3aR, C3a receptor; p38MAPK, p38 mitogen-activated protein kinase; NF-κB, nuclear factor kappa B; TLR-4, toll-like receptor 4; aPL, antiphospholipid antibodies; CAM, cell adhesion molecule; TF, tissue factor; IL-1,6,8, interleukin 1,6,8; TNF, tumour necrosis factor; apoER2, apolipoprotein E receptor 2; cPLA2, cytosolic phospholipase A2; AA, arachidonic acid; TXB2, thromboxane B2; GPIIbIIIa, glycoprotein IIbIIIa; TIFI, a 20 amino-acid synthetic peptide that shares similarity with domain V of β2GPI thus competes with it for binding to the membrane of target cells; HCQ, hydroxychloroquine; ACEI, angiotensin converting enzyme inhibitor. Solid black arrows represent activation of; dashed black arrows represent possible alternate activation pathway of; dashed red lines represent blockade/inhibition of.

In this review, we will summarize the latest advances in our understanding of the aetio-pathogenetic basis of APS and present the possible therapeutic implication that could result from these advances.

Antiphospholipid antibodies

In contrast to what the name indicates, aPL are a heterogeneous family of antibodies directed against phospholipid binding proteins and not phospholipids per se. The most important of these proteins is β-2 glycoprotein I (β2GPI) [3,4]; however, other targets for aPL are less frequently encountered; these include prothrombin, tissue plasminogen activator (tPA), annexin A2, thrombin and protein C [5,6].

Several theories have been proposed to explain aPL production in APS; however, the most accepted one suggests that anti-β2GPI antibodies may result from molecular mimicry between human β2GPI and similar molecules in pathogenic bacteria [7]. This notion was developed when mice immunized with certain bacteria developed anti-β2GPI which when extracted and transferred to pregnant mice resulted in foetal loss [8].

Various in vitro and in vivo animal studies demonstrated that aPL are pathogenic [9–11]. However, the exact mechanism through which these antibodies mediate the prothrombotic state is yet to be clearly delineated. However, it has become clear that the heterogeneity of this family of antibodies correlates with multiple mechanisms of action. These include activation of cellular elements (platelets, endothelial cells and monocytes), inhibition of the fibrinolytic system, activation of the coagulation cascade and activation of the complement system (Fig. 1).

Figure 1.

 Main pathogenetic mechanisms of aPL induced thrombosis. Abbreviations: aPL, antiphospholipid antibodies; EC, endothelial cell; TXA2, thromboxane A2; TF, tissue factor; TLR4, toll-like receptor 4; tPA, tissue plasminogen activator; PAI, plasminogen activator inhibitor; TNF, tumour necrosis factor; MAC, membrane attack complex.

The updated international preliminary (Sapporo) classification criteria for APS require one clinical criterion and one laboratory criterion for the diagnosis of APS [1] (Table 1). However, despite the great diversity of aPL, only three tests are accepted to document the presence of these antibodies in the patients’ blood. Additional antibodies can be demonstrated via different laboratory testing modalities; none, however, has demonstrated the required specificity and standardization and thus none is incorporated in the Sapporo classification criteria. These antibodies include IgA aCL, IgA anti-β-2GPI, antiphosphatidylserine, antiphosphatidylethanolamine, antiprothrombin and antiphosphatidylserine-prothrombin complex. A lot of work is currently being performed to establish the importance of these new antibodies and their pathogenic significance. Researchers are relentlessly trying to establish a universal standardized testing modality to detect aPL.

Pathogenetic mechanisms in APS

Cellular targets for aPL

Platelets.  In the context of the thrombotic tendency of the disease, and with the observation that thrombocytopenia is a frequent manifestation in APS, we can speculate that platelet activation is integral to pathogenesis of the disease (Fig. 1). In fact, studies have yielded strong evidence that aPL enhance platelet activation and aggregation in vitro as well as in vivo [12–16]. One study using affinity-purified aPL from APS patients demonstrated increased platelet activation and aggregation with these antibodies when compared with normal IgG [12]. Moreover, aPL have been shown to increase the expression of glycoproteins on the membrane of the platelet particularly GPIIb-IIIa (fibrinogen receptor important in platelet aggregation) and GPIIIa [17] (Fig. 2). Moreover, Pierangeli et al. [18] did show that in mice that are GPIIb-IIIa deficient, thrombus formation was not increased when treated with aPL. This group also demonstrated that aPL-mediated thrombosis in vivo was reduced upon pretreatment of the mice with monoclonal anti-GPIIb-IIIa antibodies.

A recent hypothesis regarding platelet activation stemmed from the observation that SLE patients with anti β2GPI antibodies had increased amounts of active von Willebrand factor (vWF) in their plasma thus leading to an increased platelet activation and aggregation. It is hypothesized that β2GPI usually binds to vWF inhibiting its ability to promote platelet aggregation and adhesion. However, in the presence of anti β2GPI antibodies, this process is blocked leading to an increase in active vWF and thus increased platelet adhesion and aggregation [19].

Endothelial cells and monocytes. In vitro studies by Simantov et al. demonstrated that endothelial cells had a significantly higher expression of cellular adhesion molecules when treated with aPL in the presence of β2GPI [20]. Higher expression of intracellular adhesion molecule-1 (ICAM-1), vascular cell adhesion molecule-1 (VCAM-1) and E-selectin was demonstrated in this study. Similarly, Pierangeli et al. showed that aPL were capable of activating endothelial cells both in vitro as well as in mouse models. More importantly, this group correlated this activation with accelerated thrombus formation in vivo [10]. In a more recent work by Cugno et al. [21], endothelial perturbation in patients with APS was illustrated by utilizing a large panel of biological and functional parameters. This group hypothesizes that endothelial cell perturbation might represent the first hit in the pathogenesis of APS. Utilizing cellular adhesion molecules (CAMs) knockout mice and anti-VCAM-1 monoclonal antibodies, two separate groups further illustrated that the pathogenesis of the aPL-mediated endothelial cell activation was mediated via these CAMs [22,23]. Moreover, trying to extrapolate these results into human patients with APS, Kaplanski et al. illustrated higher levels of soluble adhesion molecules (VCAM-1 and P-selectin) in patients with aPL related thrombosis [24].

Tissue factor (TF) is a trans-membrane protein that is a member of the class II cytokine and haematopoietic growth factor receptor family. This protein is up-regulated upon activation of certain cell types mainly vascular endothelial cells, monocytes and smooth muscle cells [25]. Pro-inflammatory cytokines [tumour necrosis factor alpha (TNF-α), bacterial endotoxin] activate the endothelial cell and lead eventually to increased expression of the TF on the surface membrane. This is mainly accomplished via translocation of nuclear factor (NF-κB) to the nucleus of the cell (normally NF-κB is sequestered actively in the cell cytoplasm) [25]. Several studies have shown that TF expression and action was increased in monocytes and endothelial cells treated with aPL [26–28]. Moreover, serum TF in APS patients compared with control was found to be elevated in a study performed by Forastiero et al. [29].

Therefore, there is convincing evidence that aPL activate endothelial cells in vivo and in vitro and that this activation is illustrated mainly through an increased expression of different cellular adhesion molecules and of TF (Figs 1 and 2).

Receptors for aPL

Platelets.  Platelet activation by aPL is thought to be mediated by two receptors, the apolipoprotein E receptor 2 (apoER2) (Fig. 2) and the platelet adhesive receptor glycoprotein GPIbα (subunit of the GPIb-IX-V platelet receptor that binds multiple ligands including von Willebrand factor) [30]. These conclusions were drawn after Lutters et al. [31] demonstrated that blocking apoER2 resulted in loss of the increased adhesion by platelets to collagen induced by the β2GPI-antibody complex. They also demonstrated that dimerized β2GPI coprecipitated with apoER2 illustrating the direct and close interaction between apoER2 and β2GPI. On the other hand, Shi et al. [32] demonstrated that β2GPI is able to bind directly to GPIbα which would provide a route for platelet activation by anti-β2GPI.

Endothelial cells and monocytes.  As with the aPL platelets interaction, the exact receptors to which aPL bind on endothelial cells have not been fully uncovered. β2GPI domain V is unique in its high ‘Lysine’ content rendering it a highly positive charged phospholipid binding domain. Studies showed that β2GPI bind to the surface of the endothelial cell through this domain, probably via interacting with negatively charged heparan sulphate residues within the endothelial cell membrane [33] (Fig. 2). It is thus hypothesized that this interaction will facilitate the binding of β2GPI to the endothelial cell/monocyte cell membrane which will in turn increase the affinity of aPL to the receptors. Interestingly, Vega-Ostertag et al. employed a synthetic peptide similar to domain V of β2GPI. It was noted that this peptide was able to abrogate the thrombogenic properties of human aPL in mouse models through competing with β2GPI for the negatively charged binding site on the endothelial cell/monocyte membrane [34].

Other than heparan sulphate of the cell membrane, other structures may act as receptors for the β2GPI. Two recent studies have demonstrated the importance of Annexin A2 of the endothelial cell membrane in binding aPL (Fig. 2). These studies hypothesized that aPL would bind β2GPI and the complex formed would in turn bind Annexin A2 [35,36]. A more recent in vivo study by Romay-Penabad et al. utilizing Annexin A2 ‘knock out’ mice demonstrated lower TF activity and VCAM expression in these mice compared with their normal controls when both groups were exposed to monoclonal anti β2GPI antibodies or an IgG isolate of APS patients. Another group utilized anti-A2 monoclonal antibodies and showed a resulting decrease in the aPL-induced CAMs expression and TF activity on cultured endothelial cells [37].

Recently Raschi et al. showed that the myeloid differentiation factor 88 (MyD88) transduction cascade is involved in the endothelial activation by aPL [38]. MyD88 is a common adaptor protein for toll-like receptors (TLR) that act as receptors for bacterial endotoxin (LPS) and lead to up-regulation of pro-inflammatory cytokines [39]. Based on Raschi’s work, a possible role for TLR-4 in the aPL-induced activation of endothelial cells/monocytes was speculated (Fig. 2). This hypothesis was investigated by Pierangeli et al. who utilized LPS responsive and nonresponsive mice (mutated TLR-4). The mutated LPS unresponsive mice showed no signs of aPL-induced thrombosis in vivo [40]. This shows that TLR-4 is involved in the pathogenesis of aPL-induced thrombosis.

Along the same line of thought, Doring et al. [41] carried on a well designed experiment to evaluate the importance of TLR in the aPL-mediated activation of endothelial cells and monocytes. Based on previous data that TNF-α, a well known factor to trigger for TF induction, is induced in aPL-treated monocytes [42]; the study demonstrated that the aPL-induced up-regulation of TNF-α is caused by specific up-regulation of TLR8 mRNA and protein expression. The TNF-α over-expression was abrogated upon utilizing a TLR8-specific inhibitory oligonucleotide and was induced by adding specific TLR8 ligands. Based on this observation, the authors hypothesize that aPL bind to monocytes through TLR8 and induce TNF-α over-expression that in turn will upregulate TF expression (Figs 1 and 2).

Triggered intracellular mechanisms

Platelets.  Thromboxane A2 (TXA2) is one of the metabolites of arachidonic acid. It is the major eicosanoid produced in the platelet that results in platelet aggregation. Studies have shown that patients with APS have a higher level of TX metabolic breakdown products than normal controls [43]. Moreover, a significant correlation was found between the level of aPL (IgG anti-β2GPI) and the level of urinary 11-DihydroxyThromboxane B2 (11DHTXB2) [15]. In separate work by Forastiero et al., it was shown that TXB2 production was enhanced upon treating normal platelets with sub-activating doses of thrombin and the F(ab)2 fragments of aPL from APS patients [14]. All this showed that thromboxanes are definitely up regulated in patients with APS (Figs 1 and 2).

It was previously known that p38 mitogen activated protein kinase (p38MAPK) is part of the signal transduction cascade triggered in platelets by thrombin and collagen. It was also known that this enzyme was part of the cascade ultimately ending in phosphorylation of cytosolic phospholipaseA2 (cPLA2) with production of TXB2 as an end result. Vega-Ostertag et al. [44] evaluated whether this pathway is also utilized in the aPL-mediated platelet activation. They demonstrated that upon treatment of normal platelets with IgG aPL or their F(ab)2 fragments along with sub-activating doses of thrombin, phosphorylation of p38MAPK increased significantly. They also showed that pretreatment with p38MAPK specific inhibitor (SB203580) resulted in loss of the aPL-mediated platelet aggregation. In addition, the increase in TXB2 seen after treating the normal platelets with aPL was as well abrogated upon pretreatment with the p38MAPK inhibitor. Phosphorylation of cPLA2 was noted to increase upon treating the platelets with the F(ab)2 fragments of aPL. This clearly shows that the aPL-mediated platelet activation and aggregation utilizes the p38MAPK-cPLA2 pathway ending in TXB2 production (Fig. 2).

Endothelial cells and monocytes.  Several research groups working independently have shown the integral role for p38MAPK and NF-κB in the transduction cascade of aPL-induced endothelial cell activation. NF-κB was the first molecule to be proposed as part of the intracellular pathway of aPL pathogenesis [23,45]. Furthermore, Vega-Ostertag et al. demonstrated that endothelial cells treated with aPL demonstrated a significant increase in phosphorylation of p38MAPK in addition to the increase in the expression and function of TF and pro-inflammatory cytokines [interleukin (IL)-6 and IL-8] [46]. This group also showed that treating the endothelial cells with specific inhibitors of p38MAPK (SB203580) and specific NF-κB inhibitors (MG132) resulted in loss of the TF up-regulation. This result was demonstrated separately by Bohgaki et al. in monocytes as well [47]. More recently, two separate in vivo studies showed that aPL resulted in up regulation of TF in the carotid artery as well as in monocytes. VCAM-1 expression was also significantly increased in aorta of mice treated with aPL. Interestingly, all these findings were abrogated by the use of the specific NF-κB and p38MAPK inhibitors [48,49]. More recently, Doring et al. [41] showed that aPL-mediated upregulation of TNF-α was caused by aPL interaction with TLR8 on the surface of monocytes. The downstream signalling of all known TLRs involves the activation of MyD88 that ultimately leads to the activation of the transcription factor NF-κB [50].

This data clearly demonstrate that NF-κB and p38MAPK are at the base of the intracellular cascade triggered by aPL in endothelial cells and monocytes (Fig. 2).

Effect of aPL on the haemostatic reactions

Several aspects of the haemostatic system are disturbed in APS patients leading to the generalized prothrombotic state. This state originates from a combination of activation of the coagulant system in addition to a hypofibrinolyitic state (Fig. 1). On one side, the aPL-induced overproduction of tissue factor will activate the extrinsic pathway of the coagulation cascade. Moreover, aPL-induced complement activation promotes a hypercoagulable state [51,52] directly through the C5b-9 Membrane Attack Complex (MAC) and indirectly through C5a receptor-mediated effects [53] (Figs 1 and 2). On the other side, Ames et al. [54] showed that in aPL positive patients plasminogen activator inhibitor 1 (PAI1) was upregulated when compared with normal controls while tPA was reduced. The study hypothesized this imbalance in the PAI/tPA ratio leads to the hypofinbrinolytic state in APS. Moreover, it was observed that anti β2GPI antibodies are capable of cross reacting with enzymatic domains of several serine proteases involved in haemostasis including protein C, protein S, thrombin and tissue plasminogen [55]. This might be a mechanism for the impaired fibrinolysis observed in APS. In fact, protein C and protein S inhibition has been reported in some studies to be associated with the aPL-induced hypofibrinolytic/prothrombotic state of APS [56,57] (Fig. 1).

Effect of aPL on the complement system

Complement activation and the prothrombotic state.  The complement cascade has attracted much attention in the past years in the context of APS in general and in foetal loss in particular. Inflammation is now believed to be at the core of the pathogenesis of this syndrome and tissue injury is speculated to be the result of complement-mediated inflammatory reaction in addition to thrombosis. It was previously realized that the complement system is capable of producing the thrombotic manifestations of APS through direct activation of the coagulation cascade via MAC or indirectly via C5a-C5a receptor interaction leading to upregulation of TF and thus activating the coagulation cascade [58] (Fig. 2). This theory was tested and validated when Pierangeli et al. demonstrated reversal of aPL-induced pro-thrombotic characteristics in vivo using monoclonal antibodies targeting C5 [51]. Along the same line of thought, Romay-Penabad et al. showed that C5a receptor deficient mice were protected from aPL-induced thrombophilia [52]. Then Fischetti et al. presented their well designed study and demonstrated that complement activation is required for aPL-mediated thrombosis [59]. More recently, Oku et al. has shown that hypocomplementaemia is actually frequent in patients with primary APS and attributed this to the state of complement activation and consumption [60].

Complement activation and foetal loss.  Obstetric complications are a major aspect of the morbidity associated with APS (Table 1). The trophoblasts of the placenta express anionic phospholipids on their cell membrane enabling them to bind exogenous β2GPI [61]. Moreover, it was also noted that these trophoblasts are capable of synthesizing their own β2GPI [62]. All this explains why the placenta is a major target for aPL, namely anti-β2GPI antibodies.

It was previously believed that thrombosis was at the basis of the obstetric complications of APS and hence the current treatment protocol that includes aspirin and heparin. However, data from histological studies demonstrated the absence of blood clots from the miscarriage samples of APS patients [63]. The sample pathological description included reduced vascular trophoblast invasion and an inflammatory infiltrate of neutrophils and monocytes [64]. Thus it was postulated that inflammation may be the culprit in aPL-induced pregnancy loss.

Holers et al. performed a study showing that in vivo inhibition of complement system using the C3 convertase inhibitor complement receptor-1-related gene protein y (Crry) I-g prevented foetal loss and growth retardation mediated by aPL [65]. Salmon, Girardi et al. utilized a mouse model of APS. They found elevated TNFα levels, neutrophilic infiltration in the placental specimens, along with C3 deposition within the decidua. No evidence for blood clots was noted [66,67]. In addition, this group demonstrated that blocking the complement system or tissue factor abrogated aPL-induced foetal loss and placental inflammatory process [58]. More recently, these results were further solidified by Pierangeli et al. when C3 and C5 deficient mice were also protected against aPL-induced thrombosis and foetal loss [51]. Shamonki et al. extrapolated these findings to human placentas of APS patients by demonstrating increased C4d and C3b deposition in the trophoblasts [68]. These findings presented the evidence that complement induced inflammation might be the aetiology of aPL-mediated foetal loss.

However, it should be noted that treatment of pregnant APS patients with heparin and aspirin has significantly resulted in an increase in live birth rates in this population [69]. This increase is now thought to be related to the anti-inflammatory properties of heparin and resulting inhibition of complement activation [70].

Thus, it can be postulated that although aPL activate platelets, endothelial cells and monocytes as illustrated above; this step seems not to be enough to generate thrombosis. It can be proposed that aPL first bind to their target cells priming and setting them in a pro-thrombotic state. However, activation of the complement system is the key step that ultimately amplifies the stimulatory effects of aPL and leads to thrombosis. We can therefore foresee that complement inhibition will be an interesting research area for targeted therapeutic implications.

The future in the management of APS

As more insight is being gained about the pathophysiology of the disease and the involved receptors and intracellular pathways, targeted treatment modalities have been proposed as possible alternatives to the current treatment options (Fig. 3). Anti-inflammatory and immunomodulatory approaches have been increasingly investigated by different research groups. We present here a group of different agents that appear to be promising based on experimental preclinical and clinical data.

Statins

As the inflammatory component of APS is surfacing, it was hypothesized that statins might prove beneficial in APS patients. First Meroni et al. showed that statins prevented aPL-mediated endothelial cell activation [71]. Then Ferrara et al. demonstrated in vivo that Fluvastatin inhibited the thrombogenic and inflammatory properties of aPL [72] and inhibited TF up-regulation in aPL-treated endothelial cells [73]. Moreover, in a pilot proteomics study carried out by Lopez-Pedrera et al., it was shown that after 1 month fluvastatin therapy, the aPL-mediated over expression of inflammatory mediators could be reversed [74]. More recently, Jajoria et al. demonstrated a significant reduction in VEGF (vascular endothelial growth factor), serum TF and TNF-α in the serum of APS patients treated with fluvastatin for 30 days compared with the control group [75]. Given all this data, we believe that it may be justifiable to use statins as a co-therapy for thrombosis prevention in APS patients; however, randomized controlled trials (RCTs) are needed to further solidify this assumption.

Rituximab

Rituximab is a chimeric monoclonal antibody that targets the cluster of differentiation 20 (CD20) of B-lymphocytes. It has been proposed that B cells are involved in the aPL-induced clinical manifestations of the disease [76]. Kahn et al. demonstrated that blockade of the B-cell activating factor (BAFF) can prevent the manifestations of the disease indicating a role for B cells in the aPL pathogenesis [77]. It thus becomes reasonable to speculate a possible role for Rituximab in the management of APS. In fact, a number of case reports reported successful results with Rituximab therapy in APS patients with thrombocytopenia, autoimmune haemolytic anaemia, or skin ulcers [78]. The RITAPS study, an open label phase 2a descriptive pilot study, is currently underway to assess the effectiveness and safety of Rituximab in persistently aPL positive patients that are resistant to the standard anticoagulation therapy [79]. In addition, the BIOGEAS study group has recently published the results of their systematic review on the off label use of biological therapy in systemic autoimmune diseases [80]. The study group reported a 92% response rate with the use of Rituximab in resistant APS patients [80]. More recently, Ramos-Casals et al. presented a review of the literature regarding the off-label use of Rituximab in resistant SLE patients revealing a 91% positive response [81]. The extrapolation of this result to APS patients is yet to be validated. However, based on these results we speculate that it is only a matter of time before Rituximab takes its place in the management algorithm of APS.

Hydroxychloroquine

Hydroxychloroquine (HCQ) is an anti-inflammatory and anti-thrombotic agent that is used in the management of some systemic autoimmune diseases. It was also used historically as a DVT prophylactic agent after hip surgeries. Several studies demonstrated that HCQ might decrease the thrombotic episodes in patients with SLE. However, it is still not clear whether this data can be extrapolated to patients with APS. Nevertheless, HCQ has been shown to inhibit platelet aggregation by preventing the over-expression of GPIIbIIIa on the membrane of aPL activated platelets [17,18]. It was also shown to decrease thrombus size in aPL-injected mice in a dose-dependent manner [82]. More recently, it has been shown that HCQ might act at a more proximal step in the aPL pathogenetic cascade by reducing the binding of the β2GPI to the phospholipid bilayer of the target cell [83]. Noteworthy of mentioning is a recent systematic review by Ruiz-Irastorza et al. on the efficacy and safety of antimalarials in patients with SLE. The review concluded that this medication is both safe and effective and should be used in every stage of the disease [84]. However, the applicability of this data to the APS population is questionable yet we believe that despite insufficient evidence, it is reasonable to recommend HCQ for thrombosis prevention in APS patients as an add-on or alternative treatment option in cases resistant to the traditional modalities.

Specific GPIIbIIIa inhibitors

As shown earlier, aPL-mediated platelet activation leads to up-regulation of GPIIbIIIa [18]. Specific GPIIbIIIa inhibitors (such as abciximb) are currently part of the management of the acute coronary syndrome as well as strokes. Considering the thrombotic nature of these diseases, and comparing them to the pro-thrombotic state of APS, it seems logical that these inhibitors might play a role in the management of this syndrome. However, no data currently exist regarding the use of these agents in the treatment of APS. The only current evidence that targeting this receptor might actually be beneficial comes from the work of Espinola et al. who demonstrated a reduction in the aPL-mediated expression of GPIIbIIIa with HCQ [17]. This concept thus needs a lot of further investigation before it can reach clinical utility.

Inhibition of TF up-regulation

As detailed previously, aPL-mediated activation of endothelial cells and monocytes is thought to lead to up-regulation of TF [26–29,85,86]. Thus, it is reasonable to assume that agents leading to down-regulation of TF might be potential therapeutic options for thrombosis prevention in APS patients. Angiotensin converting enzyme inhibitors (ACEI) have been found to decrease TF expression in monocytes [87]. Similarly, defibrotide (an adenosine receptor agonist) and dilazep (an adenosine uptake inhibitor and anti platelet agent), were found to inhibit TF expression [88,89]. However, there are currently no data to recommend the use of these agents in the management of thrombosis in APS but it seems like a promising agent for future animal studies.

Complement inhibition

Complement activation, as demonstrated previously, seems to be an integral prerequisite for the activation of the cellular elements targeted by aPL and thus thrombus formation [51,59]. It has been shown in pinch-induced thrombosis models in mice that anti-C5 antibodies could reverse the aPL-induced thrombogenic properties decreasing thrombus size and inhibiting new thrombus formation [51]. This shows that complement inhibition as a treatment option for APS is an interesting field for future investigation. It is our belief that given the pivotal role established for complement activation in the pathogenesis of APS in general and in aPL-mediated foetal loss in particular; complement inhibition will represent a key therapeutic element for APS once it attains clinical applicability.

Tumour necrosis factor

It is now widely accepted that an element of inflammation exists in the APS. In fact, higher levels of TNF, IL-1 and IL-6 have been demonstrated in APS patients [79]. In this context, it is speculated that targeting these pro-inflammatory and pro-thrombotic cytokines might be of benefit in managing APS patients. First, Blank et al. suggested that TNF DNA vaccination might prevent the clinical manifestations of APS [90]. Then, Berman et al. demonstrated that TNF deficiency or blockade prevented foetal loss in aPL-treated mice [66]. However, although it seems that anti-TNF therapy might have a role in APS as it does in other autoimmune diseases, there are no reports of its use in this syndrome. This in fact might stem from the fear of the possible elevation of the aPL (particularly aCL) titres after anti-TNF therapy. This point, however, requires further illustration.

P38 mitogen-activated protein kinase and nuclear factor-κB inhibitors

P38MAPK is a key player in the intracellular pathway of aPL-induced activation of platelets, endothelial cells as well as monocytes [44,46,47]. It was illustrated previously how specific p38MAPK inhibitors (SB203580) resulted in inhibition of platelet activation and thromboxane production as well as loss of TF up-regulation in monocytes and endothelial cells [48,49]. No animal studies have been performed to evaluate the efficacy of p38MAPK inhibitors in APS patients. However, it is noteworthy that p38MAPK inhibitors have been shown to be efficacious in several disease models including arthritis, septic shock and myocardial injury [91,92]. Moreover, various structures of p38MAPK inhibitors have been tested in preclinical trials and two particular structures, BIRB796 and RWJ67657, are under clinical trials and showing promising results in terms of safety and efficacy [92,93].

On the other hand, NF-κB has been shown to be part of the intracellular transduction cascade in the aPL-mediated up-regulation of TF in endothelial cells and monocytes. The use of specific NF-κB inhibitors (MG132) has been shown to reverse this TF up-regulation. More recently, Kubota et al. demonstrated in an in vitro study a reduction of expression of TF, IL-1 and TNF-α by aPL stimulated monocytes upon treatment with a specific NF-κB inhibitor. Chemokines production [CX3CL1(fractalkine) and CCL5(RANTES)] by aPL-pretreated vascular endothelial cells were also found to be suppressed via the utilization of the specific NF-κB inhibitor. Importantly, these researchers demonstrated that these chemokines are integral in the process of platelet aggregation and adhesion thus illustrating that the inhibition of the NF-κB will ultimately abrogate the aPL-mediated activation of platelets as well as monocytes and endothelial cells [94]. Despite these interesting results, no data exists concerning the safety and efficacy of this agent in human APS patients.

We speculate that it is through targeting p38MAPK and NF-κB that we will be hitting the basis of the disease’s pathophysiological process. As illustrated above, these agents, and in particular NF-κB, are in direct relation to the activation and up-regulation of many effector molecules that are thought to precipitate the APS clinical picture. Thus, it is only by blocking these agents that the entire aPL-induced cascade could be brought to an end. However, validation of this theory will await the results of clinical trials in the future.

Blockers of receptors on target cells

By blocking the receptors for β2GPI or aPL on the target cells, a proximal step in the thrombosis cascade would be halted and thus the whole series of events leading to the disease manifestations would be put to an end. Studies have shown that the toll-like receptor-4 (TLR-4) (and more recently TLR-8) and the annexin A2 protein are receptors for aPL [35,36,40] and thus could be the target for future therapeutic interventions (anti-annexin A2 antibodies, TLR monoclonal antibodies). Moreover, TIFI, which is a 20 amino acid synthetic peptide that shares similarity with domain V of the β2GPI protein, has been shown to inhibit the aPL-induced pro-thrombotic state in mice via competing with β2GPI and preventing it from direct binding to the phospholipid membrane of its target cell [34]. Thus, the concept of inhibiting aPL binding to their target cells seems to be an interesting area of investigation from which possible therapeutic implications might ensue.

Conclusion and future perspective

APS is a systemic autoimmune disease with thrombotic tendency manifesting as vascular (arterial and venous) thrombotic events and obstetric morbidity within the context of persistent aPL positivity Treatment of this syndrome focuses on anti-platelet agents and anticoagulation; however, in some cases, and despite documented adequate anticoagulation, recurrent thrombotic events did occur. Moreover, the need for higher intensity anticoagulation in some cases has led to an increase in the haemorrhagic complications. This clearly shows that we are in need for new, safer and more efficacious treatment modalities.

Over the last years, more insight has been gained regarding the pathophysiology of the disease process including receptors on target cells, intracellular pathways employed, as well as effector proteins up-regulated.

These achievements in our understanding of the disease have opened the door to the possibility of new more targeted therapeutic options that might be safer and more efficacious than the standard treatment modalities. In the near future, we might find statins, anti-platelets (GPIIbIIIa inhibitors), Rituximab, HCQ, ACEI or even anti-TNF medications as part of the standard treatment for APS patients. Moreover, p38MAPK and NF-κB specific inhibitors, complement inhibitors, as well as receptor blocking agents (TIFI) are all very exciting and promising areas for future investigations. Ultimately, well-designed clinical trials will be in need to validate the utility of these new therapeutic agents in the management of APS.

Acknowledgements

The authors declare no conflicts of interest.

Address

Division of Rheumatology, Faculty of Medicine, American University of Beirut, Beirut, Lebanon (A. A. Mehdi, I. Uthman); Lupus Research Unit, King's College, London School of Medicine, St. Thomas’ Hospital, London, UK (M. Khamashta).

Ancillary