Modulating thrombotic diathesis in hereditary thrombophilia and antiphospholipid antibody syndrome: a role for circulating microparticles?

Introduction

Venous thromboembolism (VTE), the collective term for deep vein thrombosis (DVT) and pulmonary embolism (PE), is a major medical problem, affecting one to three individuals per 1000 per year in industrialized countries [1], [2], [3]. VTE is a multifactorial disorder resulting from the interaction between an array of acquired and genetic factors. Prevention of a first or second event is key to reducing the morbidity and mortality rates associated with VTE [2], [4], [5], [6]. In general, the term thrombophilia refers to the tendency to develop VTE on the basis of a hypercoagulable state, owing to inherited or acquired disorders of blood coagulation or fibrinolysis. Indeed, inherited thrombophilia is associated with a high risk of VTE particularly in the presence of triggering conditions such as surgery, trauma, hormonal therapy, pregnancy, and puerperium [2], [7], [8]. In large case-control studies, the overall risk of clinical manifestations of VTE was two- to seven-fold higher in carriers of any thrombophilic disorder than in non-carriers [9], [10], [11], [12], [13]. Over the past decades, there have been great advances in the understanding of the pathogenesis of VTE in patients with inherited risk factors or antiphospholipid syndrome. However, a number of questions remain unanswered. In particular, we already know that some subjects carrying inherited thrombophilia will never experience a thrombotic episode while others with the same defects will develop recurrent VTE with no known additional acquired risk factors [14]. Moreover, it is also apparent that carriers of antiphospholipid antibodies (aPL), at increased risk of thrombosis compared to the general population [15], may remain asymptomatic over their lifetime. The reason for these discrepancies remains unknown. The development of biomarkers able to accurately stratify patients according to their VTE risk is of high clinical importance, since currently available risk models are insufficient [16]. Microparticles (MPs) or microvesicles (also referred to as shedding vesicles, ectosomes) include all heterogeneous structures created through budding and fission directly from the plasma membrane that range in size from 0.1 to 1.0 μm [17], [18], [19], [20]. MPs are constitutively released from the surface of blood and endothelial cells after structural changes involving local cytoskeletal rearrangements and membrane budding, and their formation can be up-regulated by cellular activation and apoptosis [21]. MPs circulate in blood under physiological conditions and are detected and characterized based on their phospholipids and antigens compositions, which reflect their cellular origins. A wide spectrum of pathologies have been associated with increased MPs levels, namely coagulative disorders, inflammatory states, and vascular dysfunction [19], [21], [22], [23], [24], [25], [26]. Most of these conditions are characterized by a hypercoagulable state, which may lead to venous thrombosis. Indeed, the direct link between MPs and VTE is supported by a considerable amount of new evidence [24], [25], [27]. MPs may become in the future one of the novel biomarkers for early primary and/or secondary VTE prevention protocols [28]. Therefore, the aim of this review is to offer a new perspective on the current knowledge and research trends regarding the possible role of MPs in hereditary thrombophilia and antiphospholipid syndrome.

Prothrombotic role of MPs

Perhaps the best established property of MPs is their ability to promote coagulation, which is largely linked to their physical characteristics with two specific surface features [25]. First, the externalization of anionic phospholipids [predominantly phosphatidylserine (PS)] promotes the interaction with clotting proteins cationic domains, the subsequent assembly of coagulation factors (Va, VIII, IXa) and ultimately thrombin formation. The externalization of PS is believed to be a property of all types of MPs and is a strong promoter of coagulation, thus of optimal thrombin generation and efficient haemostasis. Interestingly, it has been estimated that a platelet-derived MP generated ex vivo has a 50- to 100-fold higher procoagulant activity than the same blubbing area on an activated platelet membrane [29], which may explain the potential thrombogenicity of MPs. Second, some populations of MPs have been shown to display tissue factor (TF) on their surface. The presence of TF on MPs dramatically increases their procoagulant activity, especially since PS boosts the procoagulant activity of TF and contributes to the propagation of the coagulation cascade. TF has a high affinity for FVII/FVIIa, and therefore TF-MPs in blood will readily bind FVII/FVIIa, leading to the initiation of coagulation [25]. It is interesting to note that the density of active TF on microvesicles is higher than that on their parental cells [30], which supports the hypothesis that MP formation is not an entirely random process and that MPs are effective products made in response to a changing environment. More recent data showed that MPs not only propagate coagulation by exposing PS but also initiate thrombin generation independently of TF and the extrinsic pathway [24]. These findings shed new light on the procoagulant properties of MPs and their possible hypercoagulable impact.

MPs detection and limitations of the current available methods

Different methods and combinations of methods have been developed to detect and analyze MPs [16], [20], [24], [25], [31], [32]. Appropriate sampling conditions, processing, and sample storage are essential [33]. Flow cytometry is the most widely used method because it enables to analyze both quantitative and qualitative characteristics of MPs [32], [34], [35], [36], [37]. Meanwhile, functional assays measure the procoagulant activity of isolated MPs with high sensitivity, simplicity, and use of well-defined reagents [25], [38], [39]. The clinical research on MPs is hampered by the limitations of the currently available methods. First, pre-analytical conditions highly influence the isolation of MPs. For instance, isolation of MPs from blood is affected by vein puncture, tube transportation, time between blood collection and handling, the anticoagulant, centrifugation and washing procedures, the presence of lipoprotein particles and small platelets within the size range of MPs, storage, freezing and thawing procedures [33], [40]. The only multicenter study performed by the International Society on Thrombosis and Haemostasis (ISTH) Vascular Biology Standardization Subcommittee showed that a common pre-analytical protocol can reduce the inter-laboratory variability of flow cytometric enumeration of platelet-derived MPs in healthy individuals. However, the significant variability even when a common protocol is correctly applied remains unacceptably high for the clinical use of MPs [41]. Second, it is well known that flow cytometry does have many intrinsic drawbacks that have been recently reviewed [32], [37]. Briefly, the major points are: (i) only a small fraction of MPs can be detected by commercial flow cytometers because of their detection limit (300–500 nm) and because they can resolve only particles that differ by approximately ≥280 nm in size [37]. If very small MPs cannot be accurately detected, false-positive signals can arise from other non-cell-derived particles and/or background noise [36]; (ii) quantitative size information are imperfect because they are obtained by comparing the scattering intensity of MPs with that of polystyrene beads of known size. The scattering intensity, however, depends not only on size but also on shape, refractive index, and absorption [37]. Thus, it has been postulated that flow cytometry underestimates the number of MPs in a sample, by up to 100–1000 times [42]; (iii) fluorescence threshold is also imperfect because no panspecific marker for total MPs exists [36]. As far as the limitations of functional assays are concerned, these assays do not provide any information on MPs size, cellular source or physical properties and does not take into account the purity of the sample. Optimized protocols including a combination of both quantitative and qualitative methods should be used to best characterize MPs in clinical studies. However, no consensus on the best method has been reached so far.

MPs and venous thrombosis

While there have been consistent reports of an association between elevated MPs levels and prothrombotic conditions [24], [25], [43], data on a direct causal relationship between circulating MPs levels and thrombotic events are inconclusive so far [44]. From a clinical standpoint, MPs as biomarkers for thrombotic risk could be helpful (i) in the acute VTE setting to help secondary prevention [45], [46], [47], [48], [49], and (ii) as prognostic biomarker of VTE for primary prevention. (i) In the acute VTE, while some studies showed an increase in different types of MPs others did not, resulting in an overall inconsistency most likely traceable to differences in methodologies and techniques. Therefore, the effects of MPs formation on the inception of VTE remain unclear. Studies on the long-term trend of MPs after the acute thrombotic event [50], [51], [52], [53], [54], despite the different assays used and the different aims and settings seem to suggest that the levels of MPs after a thrombotic event remain high despite adequate anticoagulation therapy. Moreover, therapy with low molecular weight heparin seems to reduce MPs levels more effectively than oral anticoagulant [54], [55]. The role of MPs as (ii) prognostic biomarkers for VTE has also been assessed, though mostly in cancer subpopulations [43], [46], [56], [57], [58], [59], [60], [61]. The findings suggest that TF-bearing MPs detected both by flow-cytometry and by the functional MP-TF activity yielded promising results as potential prognostic biomarkers for VTE, especially in pancreatic and brain cancer. Notably, to date the role of MPs as biomarkers in VTE prediction in the presence of triggering acquired conditions such as surgery, trauma, hormonal therapy, pregnancy and puerperium (provoked VTE) has never been addressed.

MPs and hereditary thrombophilia

Established hereditary thrombophilia includes the inherited deficiencies of the natural anticoagulant proteins antithrombin, protein C (PC) and protein S (PS), the gain-of-function mutations in the factor V (FV Leiden) and prothrombin genes (PT G20210A); some dysfibrinogenaemias and increased plasma levels of coagulation factor VIII and factor IX [2], [7], [8]. Although some plasma factors such as levels of TF pathway inhibitor or levels of factor II and increased endogenous thrombin potential are linked to the thrombotic risk in subjects with FV Leiden and PT20210A, no definite risk stratification strategy using laboratory testing has yet been proven to be clinically useful in predicting thrombosis in carriers [62], [63], [64]. The inference that circulating MPs may contribute to the thrombogenic profile of FV Leiden carriers were drawn from the following observations: first, there is an approximately 10- to 20-fold reduction in the rate of activated protein C (APC) catalyzed inactivation of plasma-derived factor Va when bound to synthetic phospholipid vesicles [65]; second, APC can induce endothelial MPs production and APC bound to endothelial MPs is no longer capable of FVa inactivation [66]. For the first time, Flores-Nascimento et al. [67] used flow cytometry to measure levels of annexin V-MPs, endothelial-MPs (CD31), platelet-MPs (CD61), leucocyte-MPs (CD45), TF-bearing MPs (CD142), erythrocyte-MPs (CD235), and monocyte-MPs (CD14) in seven asymptomatic heterozygous carriers of FV Leiden and did not find any difference compared with healthy controls. MPs levels have been measured in a larger cohort of 45 FV Leiden heterozygous individuals by Enjeti and colleagues [68], [69] using specific platelet (CD41a), endothelial (CD62e), and leukocyte (CD45) surface markers. They found that all subsets of MPs were significantly elevated in the FV Leiden group compared with controls, and the most striking disparity was seen in the number of leukocyte-MPs. Moreover, there was no significant difference in MPs levels between FV Leiden subjects with and without a history of thrombosis (Table 1). The authors also performed a functional study of the MP procoagulant activity, but they did not show any significant differences in the prothrombinase activity recorded by the ELISA between FV Leiden carriers and controls. Our group measured flow-cytometric plasma levels of annexin V-MPs, platelet-MPs (CD61), endothelial-MPs (CD62E), and TF-bearing MPs (CD142) and the MP procoagulant activity by the STA^® Procoagulant Phospholipids assay in 142 carriers of factor FV Leiden, 124 of PT G20210A, and in 132 carriers of natural anticoagulant deficiencies (25 antithrombin, 63 protein C, and 64 protein S defect) and compared with age- and gender-matched healthy individuals [70], [71], [72]. We found that each group of thrombophilic carriers showed higher median levels of annexinV-MPs, endothelial-MPs, platelet-MPs, TF-bearing MPs, and MP procoagulant activity than controls. Moreover, homozygous presented higher MPs levels than heterozygous (FV Leiden and PT G20210A group) [70], [71]. Singularly considered, using the 95th percentile of controls as cut-off, carriers of FV Leiden with high levels of annexin V-MPs and TF-bearing MPs (>95th percentile) had an adjusted OR for VTE of 3.08 (95% CI, 1.42–6.69) and 2.28 (95% CI, 1.07–4.85) compared with carriers with MP ≤95th percentile, respectively. Carriers of PT G20210A had an adjusted OR for VTE of 1.95 (95% CI, 1.0–4.33) for high annexin V-MPs and 5.72 (95% CI, 1.96–6.67) for platelet-MPs. Carriers of natural anticoagulant deficiencies with high levels of annexin V-MPs, endothelial-MPs, and platelet-MPs had an adjusted VTE OR of 3.36 (95% CI, 1.59–7.11), 9.26 (95% CI, 3.55–24.1), and 2.72 (95% CI, 1.16–6.38), respectively (Table 1). The OR were adjusted for possible confounders (age, sex, body mass index). In conclusion, the published data show that circulating MPs may contribute to the development of VTE in thrombophilic carriers, both in mild and severe states. In thrombophilia, MPs may act as triggering factors to enhance the global prothrombotic state up to the threshold of the clinical relevant thrombotic event, through different mechanisms: in FV Leiden population, we observed the highest OR for VTE associated with TF-bearing MPs; in PT G20210A, associated with platelet-MPs; and finally in natural anticoagulant deficiencies, in association with endothelial-MPs.

MPs and antiphospholipid antibody syndrome

Antiphospholipid antibody syndrome (APS) is an autoimmune disorder associated with thrombotic and obstetric complications and persistent aPL [73], [74]. To date, there have been few studies targeting MPs in patients with APS or asymptomatic carriers of aPL and findings have been variable [75], [76], [77], [78], [79], [80], [81], [82], mostly due to the small sample size and the heterogeneity of cases (primary/secondary APS, different aPL subtypes, presence/absence of obstetric complications) (Table 2). Combes et al. [75] pioneered the study of endothelial-MPs (CD51) in 30 patients with primary and secondary APS [only considering positivity for lupus anticoagulants (LAC)] and 30 healthy controls and found an approximately two-fold increased of endothelial-MPs in the LAC patients; mean endothelial-MP levels were also higher in patients with thrombosis (n=13). In another small-cohort study, Joseph et al. [76] first measured platelet-MPs (CD62P) in 20 patients with primary APS and 30 with systemic lupus erythematosus (SLE) (14 of whom had secondary APS). They found that platelet-MPs numbers were not increased neither in primary APS nor in SLE compared to controls. Dignat-George et al. [77] measured endothelial–MPs (CD51) levels in 23 primary APS, 14 secondary APS (SLE-associated), 28 SLE aPL+ (without history of thrombosis), 23 SLE aPL−, 25 thrombosis without aPL and SLE, and 25 healthy controls. Compared to healthy subjects, elevated plasma levels of endothelial-MPs were found in patients with primary and secondary APS and in SLE aPL+ but not in SLE aPL− or in non aPL-related thrombosis. Moreover, no difference between patients with thrombotic complications of aPL (n=37 patients compared to those with asymptomatic aPL were found (n=28 patients). Jy et al. [78] measured endothelial-MPs (CD31+/42−) and platelet-MPs (CD31+/42+) in 60 patients with primary APS and 28 asymptomatic carriers of aPL. They confirmed elevated levels of endothelial-MPs in patients with APS and aPL+ versus controls but did not find any difference between APS and asymptomatic aPL. However, they reported no differences in platelet-MPs in patients with APS and aPL+ compared with controls, although the levels were increased in APS with thrombotic complications compared with asymptomatic aPL. More recently, Vikerfors et al. [79] measured endothelial-MPs (CD144), platelet-MPs (CD42a), monocyte-MPs (CD14), and endothelial-TF bearing MPs (CD144+/CD142+) in 52 patients with APS and 52 healthy controls. They showed increased numbers of endothelial-MPs, endothelial TF-positive MPs, and monocyte-MPs in APS compared with controls. Conversely, platelet-MPs number did not differ between the groups. None of the MPs types differed in numbers between obstetric (20 patients) and thrombotic (42 patients) APS patients. Two latest studies addressed levels of MPs in APS. The study by Chaturvedi et al. [80] found that endothelial-MPs (CD105 and CD144), platelet-MPs (CD41) and TF-bearing MPs were significantly higher in 47 patients with APS (13 with secondary APS) than 144 controls, although the levels of monocyte-MPs (CD14) were not significantly increased. Interestingly enough, levels of endothelial-MPs showed a positive correlation with IgG (R=0.60, p=0.006) and IgM anti-β₂-glycoprotein antibodies (R=0.58, p=0.006). Breen et al. [81] showed that levels of circulating endothelial-MPs (CD51 and CD105) and platelet-MPs (CD41 and CD61) were significantly increased in 37 primary APS patients with thrombotic complications compared to 18 healthy controls. Meanwhile, no differences were observed between patients with obstetric APS or asymptomatic aPL and healthy controls. A small study including only patients with obstetric APS (n=9) showed no difference in endothelial-MPs (CD144+/CD31+) compared to pregnant women with pregnancy loss as control [83]. It is worth mentioning the interesting study by Niccolai et al. [84] that examined the total number of MPs by both Violet Proliferation Dye 450 (VPD450)- and Annexin V-staining, endothelial-MPs (CD31), platelet-MPs (CD41a), and leukocyte-MPs (CD45) in 16 patients with primary APS, 16 asymptomatic aPL, and 16 healthy subjects. They found significantly increased levels of total-MPs (both VPD450+ and annexin V+), endothelial-MPs, platelet-MPs, and leukocyte-MPs in APS compared to asymptomatic aPL and in asymptomatic aPL compared to controls. Interestingly enough, they observed that the majority of the total-MPs measured by VPD450 were negative for annexin V staining.

Lastly, a recent study by Willemze et al. addressed the role of TF-bearing MPs measured by a functional assay (MP-TF activity) in plasma samples from 30 patients with APS and 72 asymptomatic aPL [82]. The authors found that MP-TF activity was higher in APS compared to asymptomatic aPL subjects (p=0.001). No differences were observed between APS subjects with thrombosis and/or obstetric complications. Similarly, among subjects with either APS or asymptomatic aPL, MP-TF activity did not differ in the presence or absence of underlying SLE. In conclusion, to date, findings on MPs in APS are conflicting and difficult to interpret, notably because patient numbers have mostly been small, patients included usually had aPL associated with SLE, and APS patients with obstetric complications were rarely studied. It is worth mentioning that four studies evaluated MPs levels in treated APS patients and the use of either anticoagulant (low-molecular-weight heparin or warfarin) or anti-platelet therapy did not affect MPs levels [75], [77], [79], [80]. Summarizing, studies seem to confirm the following: (i) the presence of endothelial-MPs and platelet-MPs in APS and in asymptomatic carriers of aPL; (ii) the direct association of MPs with the presence of aPL (mainly anti-β₂-glycoprotein) but not with underlying autoimmune disorders (mainly SLE) or the thrombotic event itself; (iii) no influence of anticoagulation therapy in MPs levels in APS. We may therefore postulate a similar pathophysiology mechanism as in the hereditary thrombophilia, in which MPs could act as triggering factors to enhance the pre-existing aPL-induced prothrombotic state up to the threshold of the clinical relevant thrombotic event, mainly through the activation of platelets and endothelial cells. Undoubtedly, the functional interplay among endothelial cells, platelets, inflammatory cells, and MPs is involved in the VTE onset in APS.

Discussion and conclusions

Since no definite risk stratification strategy using laboratory testing has yet been proven clinically useful for prediction of thrombosis in carriers of thrombophilia, prognostic markers capable of estimating the individual VTE risk would be of great utility. MPs have been reported to be a promising prognostic markers of thrombotic risk, mainly in cancer population. Indeed, the studies reported showed that MPs might offer one possibile explanation to the variability in VTE phenotypic expression in thrombophilic and aPL carriers. However, the important limitations of our observations are worth mentioning. First, studies conducted so far, by our group also, are mostly retrospective and case-control, and MPs, which rise in response to acute health events (and possibly decline later), were not tested longitudinally in such patients to determine whether they act as a synergistic risk factor for VTE over time. So, at the current time, there is no prospective evidence about the interaction of MP and thrombophilia/aPL on VTE risk. Second, pre-analytical and analytical problems in MP detection play a major role in variability of results and make diffult to compare different studies. In conclusion, in hereditary thrombophilia and in APS, MPs could be considered as another possible, synergistic risk factor for thrombosis, but prospective studies with a longitudinal evaluation of MPs levels over time after the creation of a standardized protocol in pre-analytical and analytical workup are needed to confirm existing data.