Evaluation of the diagnostic potential of antibodies to beta2-glycoprotein 1 domain 1 in Chinese patients with antiphospholipid syndrome

In this study, we evaluated the clinical performance of anti-β2-glycoprotein 1 domain 1 antibodies (aβ2GP1-D1) in the diagnosis of antiphospholipid syndrome (APS). Sera from 229 subjects were tested, including 35 patients with primary APS, 51 patients with APS associated to other diseases, 30 patients with non-APS thrombosis, 32 patients with non-APS pregnancy-related morbidity, 42 patients with systemic lupus erythematosus, and 39 healthy controls (HC). Serum IgG aβ2GP1-D1, IgG/IgM anti-cardiolipin (aCL) and IgG/IgM aβ2GP1 were measured by a chemiluminescence assay. The levels of IgG aβ2GP1-D1 were significantly increased in patients with APS, compared with disease controls and HCs (p < 0.001). Significant correlation was identified between IgG aβ2GP1-D1 and IgG aβ2GP1 (p < 0.0001), indicating IgG aβ2GP1-D1 were the predominant domain-specific antibodies in IgG aβ2GP1 family. Importantly, aβ2GP1-D1, but not aβ2GP1 non-D1, was significantly correlated with thrombotic events. Interestingly, no significant correlation between IgG aβ2GP1-D1 and obstetric complications was observed. Additionally, significantly higher levels of IgG aβ2GP1-D1 were found in patients with triple aPL positivity, compared with patients with double and single aPL positivity. Our findings suggest a potential role of IgG aβ2GP1-D1 in identifying APS patients with high risk of thrombosis, shedding insight on the introduction of IgG aβ2GP1-D1 in China.

Iverson et al. characterized the autoantibodies to each of the 5 β 2GP1 domains and found that most aβ 2GP1 antibodies reacted with epitope(s) in D1, indicating D1 as the main immunogenic epitope targeted by aβ 2GP1 antibodies from patients with APS [5][6][7] . De Laat et al. further demonstrated that the Gly40-Arg43 region in β 2GP1-D1 was the critical epitope, and antibodies against this region were able to interfere with the coagulation process and were strongly correlated with thrombosis 8 . Additionally, Ioannou et al. confirmed D1 as the immunodominant epitope of β 2GP1 in animal models of aPL-induced thrombosis, as treatment of recombinant D1 peptide protected C57BL/6 mice from human aPL-induced pathology 9 . Subsequent work from several groups demonstrated that antibodies to β 2GP1-D1 were associated with increased risk of thrombotic and obstetric manifestations in patients with APS [6][7][8][9][10][11] , suggesting that aβ 2GP1-D1 antibodies could represent a pathogenic subpopulation of aβ 2GPI antibodies.
Although aβ 2GP1-D1 antibodies have attracted particular interest for its prognostic potential for thrombosis and pregnancy complications, the number of studies is still limited, and its clinical value needs to be verified in patients with different ethnic/geographic background. To our knowledge, few, if any, studies have reported the role of aβ 2GP1-D1 antibodies in Chinese patients with APS. It is of paramount importance to evaluate this, as this information will enhance our understanding of the clinical utility of the aβ 2GP1-D1 antibodies.
Currently, the detection of aβ 2GP1-D1 antibodies is mainly based on ELISA assays with different detection strategies 6 . A novel chemiluminescence immunoassay (CIA) based assay for detecting aβ 2GP1-D1 antibodies has recently developed. This novel CIA assay showed good agreements with ELISA [11][12][13] . We and others have showed that CIA assays have good performance in detecting aCL and aβ 2GP1 autoantibodies 13,14 . More importantly, CIA has been considered as a promising tool to improve the reproducibility and reduce inter-laboratory variations. In this study, we utilized the CIA assay to evaluate the role of aβ 2GP1-D1 antibodies in the diagnosis of APS, with a particular interest in their prognostic value for thrombosis and pregnancy complications.

Results
Levels of IgG aβ2GP1-D1 Antibodies were Elevated in Patients with APS. The values expressed as chemiluminescent units (CU) of IgG aβ 2GP1-D1 from all subjects is presented in Fig. 1. The levels of IgG aβ 2GP1-D1 were significantly increased in patients with APS, compared with patients with non-APS thrombosis, non-APS PRM, and SLE (p < 0.001), as well as healthy controls (p < 0.001). No significant difference in the levels of IgG aβ 2GP1-D1 antibodies was observed between patients with PAPS and APSAOD. When the manufacturer's recommended cut off of 20 CU was applied, the presence of IgG aβ 2GP1-D1 antibodies in patients with PAPS, APSAOD, non-APS thrombosis, non-APS PRM, and SLE were 48.6%, 45.1%, 0, 0, and 7.1%, respectively (Table 1).

Levels of IgG aβ2GP1-D1 antibodies in APS patients with different aPL profiles. The levels of
IgG aβ 2GP1-D1 antibodies were also evaluated in APS patients with the triple aPL positivity, double aPL positivity, and single aPL positivity, as triple aPL positivity has been considered as a risk factor for aPL-mediated clinical manifestations 6 . Importantly, significantly higher levels of IgG aβ 2GP1-D1 antibodies were found in patients with triple aPL positivity, compared with patients with double and single aPL positivity (Fig. 3). Additionally, patients with double aPL positivity exhibited significantly higher levels of IgG aβ 2GP1-D1 antibodies, compared with patients with single aPL positivity (Fig. 3).

Discussion
Previous studies have shown that abs specific to β 2GP1-D1 are associated with thrombosis and pregnancy morbidity [6][7][8][9][10][11] . Two recent studies provide a proof of concept on the pathogenic role of aβ 2GP1-D1 antibodies 15,16 . Agostinis et al. showed that a single-chain fragment variable (scFv) directed against β 2GP1-D1 induced thrombosis and fetal loss in naïve rats/mice 15 . The other study demonstrated that aβ 2GP1-D1-rich polyclonal IgG fractions from serum of patient with APS induced significantly larger thrombi in vivo compared with aDI-poor counterpart 16 . The pathogenic potential of aβ 2GP1-D1 antibodies may come from their ability to trigger TLR4-NF-κ B pathway, as β 2GP1-D1 shares a high degree of homology with an extracellular epitope of human TLR4 17 . Given the significance of aβ 2GP1-D1 antibodies, it is of paramount importance to characterize the clinical relevance of aβ 2GP1-D1 antibodies in Chinese patients with APS.
In this study, we found that the levels of IgG aβ 2GP1-D1 antibodies were significantly elevated in patients with APS. In addition, IgG aβ 2GP1-D1 antibodies were the predominant domain-specific antibodies in IgG aβ 2GP1    family. More importantly, aβ 2GP1-D1 antibodies, but not aβ 2GP1 non-D1 antibodies, were significantly correlated with thrombotic events. In contrast, no significant correlation between IgG aβ 2GP1-D1 antibodies and obstetric complications was observed. Our findings suggest that aβ 2GP1-D1 antibodies could serve as a promising biomarker to identify patients at risk of thrombosis in China. We used the CIA assay in the entire study, rendering the results more reliable. Previously, we showed that the CIA assay had good performance characteristics and good agreements with a commercial ELISA from the same manufacturer 14 . As a variety of different assays have been used in detecting aβ 2GP1-D1 antibodies (e.g., competitive inhibition ELISA with different D1 antigen, direct ELISA with different D1 antigen), the comparability of results across different studies might result in substantial variations 18 . In the present study, IgG aβ 2GP1-D1 antibodies were detected in 48.6% of patients with PAPS and 45.1% of patients with APSAOD. Mondejar et al. from Spain reported that IgG aβ 2GP1-D1 antibodies were present in 31% of patients with PAPS and 46% of patients with APSAOD using the CIA assay 13 . The prevalence of IgG aβ 2GP1-D1 antibodies in APSAOD patients was similar between the two studies, but the prevalence of IgG aβ 2GP1-D1 antibodies in PAPS patients was higher in our study. Interestingly, a recent meta-analysis on 548 patients with APS from 11 different centers showed that the prevalence of IgG aβ 2GP1-D1 antibodies was 44.0% (241/548) 19 , similar to what we found in this study.
It is worth mentioning that 13 patients with APS (7 patients with PAPS and 6 patients with APSAOD) were LAC positive but aβ 2GP1 negative. This discrepancy may be due to the existence of other antibodies 20 . Indeed,   It has been suggested that the majority of the IgG aβ 2GP1 antibodies bind to epitopes located in β 2GP1-D 21,22 . In our study, we found IgG aβ 2GP1-D1 antibodies were present in 81.4% of APS patients with positive aβ 2GP1 antibodies, supporting D1 as the major epitope in β 2GP1. Interestingly, in a multicenter study on patients with APS from Europe and the United States, IgG aβ 2GP1-D1 antibodies were detected in 55% of patients with positive aβ 2GP1 antibodies 10 , which is lower than that in our study. It is likely that the discrepancies are caused by different assays on IgG aβ 2GP1 and IgG aβ 2GP1-D1 antibodies detection, as we utilized the CIA assay, while they used the ELISA assay 10 . Interestingly, another study from Italy using CIA for IgG aβ 2GP1 and IgG aβ 2GP1-D1 determination showed that IgG aβ 2GP1-D1 antibodies accounted for 69% of IgG aβ 2GP1 antibodies 11 . Interestingly, one patient was negative for IgG aβ 2GP1, but was positive for IgG aβ 2GP1-D1. A possible explanation for this discrepancy could be due to the different epitopes recognized by IgG aβ 2GP1 and IgG aβ 2GP1-D1, as D1 epitope becomes available when β 2GP1 transitions from a circular form to a fish-hook conformation 4 .
Increasing evidence suggest that multiple positivity of aPLs are important parameters for risk assessment 29,30 . Interestingly, we observed that, 4 patients with PAPS had multiple thrombosis, and three of them exhibited triple-positive aPL profile and one showed positive LAC. Additionally, 3 patients with PAPS had both thrombosis and obstetric complications, and 2 of them exhibited triple-positive aPL profile and one showed positive LAC. In patients with APSAOD, 12 patients showed multiple thrombosis, and 5 of them showed triple-positive aPL profile, 5 of them displayed double-positive aPL profile, and the rest 2 patients exhibited positive LAC. In addition, 11 patients with APSAOD had both thrombosis and obstetric complications, and 5 of them exhibited triple-positive aPL profile, and the rest 6 patients showed double-positive aPL profile (data not shown). Notably, in patients with PAPS, 2 patients out of 35 (5.7%, one patient with LAC+ /aCL+ and the other patient with aCL+ /aβ 2GP1+ ) exhibited double-positive aPL profile, while in patients with APSAOD, 15 patients out of 51 (29.4%, 11 patients with LAC+ /aβ 2GP1+ and 4 patients with aCL+ /aβ 2GP1+ ) showed double-positive aPL profile (data not shown). More importantly, in this study, we found significantly higher levels of IgG aβ 2GP1-D1 antibodies in patients with triple-positive aPL profile, further supporting the importance of IgG aβ 2GP1-D1 antibodies in evaluation of the APS clinical risks.
In contrast to the association between aβ 2GP1-D1 antibodies with thrombosis, we did not observe any significant correlation between IgG aβ 2GP1-D1 antibodies and obstetric complications, which differs from previous studies [6][7][8]10,11 . Moreover, no significant associations were observed between IgG aCL, IgG aβ 2GP1, or LAC and obstetric complications. Different ethnic/geographic backgrounds might contribute to this discrepancy. Further studies with more APS patients with obstetric complications are needed.
It should be noted, however, that several limitations exist in this study. First, the diagnosis of patients with APS in this study requires the presence of at least one of the aPLs (LA, aCL, and aβ 2GP1 autoantibodies) 2 , which might exclude the seronegative APS patients 31 . Second, we used sera from homogenous Chinese Han population. A multicenter study with different ethnic backgrounds is needed for generalizing our data to wider populations. Third, thrombosis is unusual in young non-APS subjects. Thus, patients with non-APS thrombosis were younger than patients with APS, as we wanted to reflect the real epidemiology in patients with non-APS thrombosis. Last, as mentioned before, more APS patients with obstetric complications are needed to assess the association of IgG aβ 2GP1-D1 antibodies and obstetric complications.
In summary, our data suggest a potential role of IgG aβ 2GP1-D1 antibodies in identifying APS patients with high risk of thrombosis, and thus could serve as a promising biomarker in clinical and therapeutic decision-making process. Our findings might shed insight on the introduction of IgG aβ 2GP1-D1 antibodies in the laboratory diagnosis of APS in Chinese hospitals.

Methods
Subjects and Specimen Collections. Sera from 229 subjects were collected and analyzed in this study (Table 1). All the subjects were Chinese Han population. These subjects included 35 patients with primary APS (PAPS), 51 patients with APS associated to other diseases (APSAOD) (43 patients with SLE, 1 patient with both SLE and Sjögren's syndrome (SS), 4 patients with connective tissue diseases, 1 patient with primary SS, 1 patient with Waldenstrom macroglobulinemia and 1 patient with tuberculous pleurisy), 30 patients with non-APS thrombosis, 32 patients with non-APS pregnancy-related morbidity (PRM), 42 patients with systemic lupus erythematosus (SLE), and 39 healthy controls (HC). HC were defined as no signs of infection or inflammation or other significant illnesses. APS was diagnosed according to the Sydney revised Sapporo guidelines 2 .
Specifically, a combination of one positive clinical criterion and one positive laboratory criterion (LAC, aCL or aβ 2G1 antibodies determined by ELISA) on two different occasions separated by 12 weeks were used for the diagnosis 2 . For patients with PAPS, treatments of patients at the time of serum collection include Aspirin ( 29.4%). The median time intervals between clinical events and the time of serum collection were 3 years (0.2-8 years) for patients with thrombosis and 1.5 years (0.3-6 years) for patients with obstetric complications. Clinical and laboratory features were collected from all the subjects. The presence of arterial and venous thrombosis in patients with PAPS, APSAOD, non-APS thrombosis, non-APS PRM, and SLE were 25.7% and 40.0%, 37.3% and 51.0%, 16.7% and 86.7%, 0 and 3.0%, and 2.3% and 0, respectively. The incidence of obstetric complications in patients with PAPS, APSAOD, non-APS thrombosis, non-APS PRM, and SLE were 52.6%, 51.4%, 0, 100%, and 0, respectively. LAC was determined by updated guidelines, as previously described 21 . The presence of LAC in patients with PAPS, APSAOD, non-APS thrombosis, non-APS PRM, and SLE were 71.4%, 78.4%, 6.7%, 3.1%, and 11.9%, respectively. Study protocols were reviewed and approved by the Ethical Committee of Peking Union Medical College Hospital (PUMCH) and informed consents were obtained from all participants. The study was conducted in accordance with the approved guidelines. All sera were stored at − 20 °C until analysis.
Serum aPL Antibodies Determination. Serum IgG and IgM aCL and IgG and IgM aβ 2GP1 antibodies were determined by CIA (QUANTA Flash ® assays, INOVA Diagnostic, Inc, San Diego, CA) according to the manufacturer's instructions, as previously described 14,32 . Serum IgG aβ 2GP1-D1 antibodies were measured by CIA from QUANTA Flash ® β 2GPI Domain 1 (INOVA Diagnostic, Inc, San Diego, CA). The principle and procedures of the QUANTA Flash ® β 2GPI Domain 1 was previously described by Pengo et al. 14  Statistical Analysis. Prism 5.02 (GraphPad Software, San Diego, California, USA) was utilized for all statistical tests. Data of IgG aβ 2GP1 (CU) and IgG aβ 2GPI-D1 antibodies were transformed into log10 to create the Gaussian distribution. One-way ANOVA was used to calculate the difference between groups. Spearman's correlation test was performed to analyze the correlation between IgG aβ 2GP1 and IgG aβ 2GPI-D1 antibodies. p values of less than 0.05 were considered statistical significant.