Prevalence, persistence and clinical correlations of classic and novel antiphospholipid antibodies in systemic lupus erythematosus

Prevalence, persistence and clinical correlations of classic and novel antiphospholipid... Abstract Objectives aPL are frequently present in SLE. In a well characterized SLE cohort we aimed at investigating the prevalence of aPL and assessing their analytical performance and clinical association by testing criteria specificities including LA, aCL IgG and IgM, anti-β2-glycoprotein 1 (antiβ2GP1) IgG and IgM, as well as the non-criteria aPS-PT IgG and IgM and anti-β2GP1 domain 1 (aD1) IgG. Methods We included 178 patients satisfying the ACR SLE classification criteria, from whom 283 samples and thrombotic events were collected longitudinally. Each sample was tested for criteria and non-criteria aPL using validated techniques in a single centre. Results All assays provided highly reproducible results. Of the samples, 42.5% were positive for at least one criteria assay, 20.5% showed double positivity and 12.6% triple positivity. All criteria and non-criteria specificities persisted over time. Most antibody titres were only moderately correlated; however, strong correlation was observed on one hand between aD1 IgG, antiβ2GP1 IgG and aCL IgG, and on the other between aPS-PT IgG and LA. aD1 IgG titres were extremely elevated in triple-positive samples. aPS-PT IgG by itself, and jointly with LA, was associated with thrombosis, an association mostly driven by venous thrombotic events. Conclusions In this SLE cohort, the non-criteria aPL aD1 IgG and aPS-PT IgG performed differently. aD1 IgG was highly enriched in triple-positive samples, and aPS-PT IgG, jointly with LA, was associated with thrombotic events. systemic lupus erythematosus, antiphospholipid, lupus anticoagulant, beta 2-glycoprotein 1, phosphatidylserine/prothrombin complex, domain 1 Rheumatology key messages High titers of anti-β2-glycoprotein 1 domain 1 IgG are characteristic of triple-positive SLE sera. Anti-phosphatidylserine/prothrombin complex IgG are associated with lupus anticoagulant in SLE. Anti-phosphatidylserine/prothrombin complex IgG by itself and jointly with lupus anticoagulant are associated with thrombotic events in SLE. Introduction SLE is a chronic inflammatory condition due to altered immunologic tolerance, biologically characterized by the presence of auto-antibodies directed against ubiquitous autoantigens [1, 2]. aPL are often present and part of the ACR classification criteria for SLE [3, 4]. aPL represent a heterogeneous family of antibodies directed against phospholipids and phospholipid-binding proteins [5, 6]. The presence of aPL in SLE has been associated with an increased risk of thrombosis, and poor pregnancy outcomes as in primary APS [7–9]. In addition, aPL are linked to more severe SLE features, including valve disease, pulmonary hypertension, livedo reticularis, thrombocytopaenia, haemolytic anaemia, acute/chronic renal vascular lesions and moderate/severe cognitive impairment, worse quality of life and higher risk of organ damage [10, 11]. The current biological criteria for APS classification rely on the identification of LA using coagulation assays and/or by detecting anti-β2-glycoprotein 1 (antiβ2GP1) and aCL IgG or IgM with solid phase assays [12]. Several additional antigenic specificities for aPL antibodies have recently attracted great interest [13]. The β2GP1 protein is composed of five domains, and domain 1 (D1), which is a cryptic epitope when β2GP1does not bind to anionic phospholipids, is thought to be preferentially recognized by antibodies in individuals with APS [14–16]. Interestingly, conflicting results have been reported on the association of anti-D1 (aD1) IgG and thrombotic events, with some reports indicating an increased risk of thrombosis in individuals with aD1 IgG [14] and others not confirming such an association, particularly in SLE [17]. Furthermore, IgG or IgM directed against the aPS-PT have been variably associated with thrombotic events. However, in a systematic review of published evidence, Sciascia et al. [18] report that the presence of aPS-PT IgG or IgM increases the risk (odds ratio = 5.11; 95% CI: 4.2, 6.3) of arterial and/or venous thrombosis. Thus, while aD1 and aPS-PT IgG and IgM are not currently among the criteria antibodies for APS, we felt it to be of interest to investigate whether the presence of these non-criteria antibodies, in association or not with criteria antibodies, could provide additional information on the clinical picture in SLE patients. In this study, taking advantage of the Swiss SLE cohort [19], we aimed to investigate: the prevalence at entry into the cohort of criteria and non-criteria aPL antibodies, tested in a central laboratory; their persistence over time; whether their presence and titres were correlated; and which combination of autoantibodies was most strongly associated with thrombotic events. Methods Plasma samples were obtained from 178 individuals satisfying at least four ACR SLE criteria and consecutively enrolled in the prospective Swiss SLE Cohort Study (SSCS) from April 2007 to January 2015 [19]. A second, follow-up sample was obtained from 105 individuals of the original cohort for a total of 283 distinct samples. Median interval (interquartile range) between the first and second samples was of 13.1 (3.6–61.4) months. Ethical clearance was obtained in all participating SSCS centres. Informed, written consent was obtained from all participants according to the Declaration of Helsinki. All samples (N = 283) were tested in a single laboratory in Geneva (Hemostasis Unit, University Hospital, Switzerland). LA was detected using ACL TOP (Instrumentation Laboratory, Paris, France), which detected the ability to prolong two phospholipid-dependent tests: the silica clotting time (SCT) and the dilute Russell’s viper venom time (dRVVT). The presence and titre of aCL and antiβ2GP1 IgG and IgM antibodies was assessed using the HemosIL AcuStar Coagulation Analyzer (Instrumentation Laboratory, Bedford, MA, USA) [20, 21]. The cut-off values for these tests were based on the 99th percentile and the corresponding 95% CI for each antibody of 626 healthy blood bank donors [21]. aPS-PT IgG and IgM antibodies were determined by ELISA using QUANTA Lite kits using the DSX automated apparatus (DSX/best 2000; INOVA, San Diego, CA, USA). aD1 antibodies were determined with the HemosIL AcuStar Coagulation Analyzer using QUANTA Flash. Reagent Kits were kindly provided kindly supplied by INOVA (Bettlach, Switzerland). All tests were carried in duplicate [22]. aCL IgG and IgM cut-offs were 14 and 19 IU/ml, respectively; antiβ2GP1 IgG and IgM cut-offs were 17 and 12 IU/ml, respectively [21]; for both aPS-PT IgG and IgM the cut-off was 30 IU/ml, according to manufacturer recommendations. The aD1 cut-off was 14.4 IU/l, based on the 99th percentile obtained by testing the plasma of 195 healthy women as previously reported [23]. When biological duplicate samples were assessed, infraclass correlation coefficients were >90% for all tests (supplementary Table S1, available at Rheumatology online). Thrombotic events Venous and arterial thrombotic events were recorded at inclusion into SSCS and at subsequent visits. Consistent with the SLE Damage Index (SDI) [24] as used in the PROFILE study [25], arterial vascular events were documented if myocardial infarction, angina pectoris and/or a vascular procedure for myocardial infarction (coronary artery bypass graft), cerebral vascular accident and claudication lasting ⩾6 months and/or evidence of gangrene or significant tissue loss (loss of a digit or a limb) had occurred. Venous thrombotic events were deep venous thrombosis and or pulmonary embolism. Only the first incident event was taken into consideration. By definition, these events were recorded in SDI only if 6 months had elapsed since the diagnosis of SLE had been made. Statistical analysis We computed the proportions of patients who were positive for any of the nine aPL tests. We also defined categories of positivity as single, double or triple, according to the number of positive results for antibodies against cardiolipin (IgG or IgM), β2GP1 (IgG or IgM) and the LA (dRVVT or SCT). We examined the levels of non-criteria antibodies across categories of positivity. We obtained Pearson correlation coefficients between the nine test titres or results, which were logarithm-transformed to reduce the skewness of the distributions. To explore the pattern of associations between the tests, we conducted an exploratory factor analysis on the logarithm-transformed variables, followed by varimax rotation [26]. We examined the sampling adequacy of the exploratory factor analysis using the Kaiser-Meyer-Olkin statistic (desirable value >0.8), retained factors whose eigenvalue was >1, and obtained the proportion of total variance explained by the retained factors. To explore the clinical relevance of the retained factors, we examined their associations with the presence of vascular events (venous, arterial or both). To examine the reliability of the laboratory procedures, we obtained intraclass correlation coefficients on the titres obtained for doubly processed samples; this was done for only a subset of the samples [27]. To examine the stability of aPL over time, we computed kappa statistics (for results defined as positive or negative) and intraclass correlation coefficients (for continuous titres) for pairs of assessments obtained over time in the same patient, usually distant by several months. All tests were two sided and assessed at the 5% significance level. Statistical analyses were performed using SPSS version 22 (IBM Corp, Armonk, NY). Results Characteristic of the study population In this study, we included 178 individuals of whom 156 (87.6%) were women. The SLE clinical characteristics of the individuals included are summarized in Table 1. The mean (s.d.) age at SLE diagnosis was 34.7 years (15.2) and 44.1 years (14.6) at inclusion in the cohort. Some 42.5% of the samples were positive for at least one classical aPL, 20.5% showed double and 12.6% triple positivity (Table 2). Of interest, six samples positive for aPS-PT IgG and 12 samples positive for aPS-PT IgM were negative for criteria aPL. Indeed, the frequency of single-positive aPS-PT IgM and aPS-PT IgG was higher than the frequency of single-positive samples for the other assessed specificities, being >7.4% for aPS-PT and <3.7% for the others. Concerning specifically LA and aPS-PT IgG, 19 samples were positive for both, 10 positive for aPS-PT IgG alone and 21 for LA alone. All but one aD1 IgG (15%) positive sera were also positive for antiβ2GP1 IgG. No statistically significant differences in the ACR criteria, disease activity as captured by (Safety of Estrogens in Lupus National Assessment, SELENA-SLEDAI) and Physician Global Assessment as well as damage as captured by SLICC-SDI were observed between positive and negative sera for criteria aPL (data not shown). In addition, no statistically significant differences were observed in the frequency of positive samples, for all nine aPL tests performed, whether or not the patients were under CS treatment or whether disease activity captured by SELENA-SLEDAI was 0–6 or >6 (data not shown). Table 1 Baseline characteristics of the 178 patients with SLE   N (%)  Female sex  156 (87.6)  Caucasiana  139 (78.1)  Age at SLE diagnosis, mean (s.d.), years  34.7 (15.2)      Range  11.9–73.4  Age at first assessment, mean (s.d.), years  44.1 (14.6)      Range  16.6–85.3  ACR criteria at inclusion        Malar rash  76 (42.7)      Discoid rash  34 (19.2)      Photosensitivity  83 (46.6)      Nasopharyngeal ulcers  42 (23.6)      Arthritis  146 (82.0)      Pleuritis  43 (24.2)      Pericarditis  41 (23.0)      Renal disorder  81 (45.5)      Seizures  14 (7.9)      Psychosis  10 (5.6)      Haematologic disorder  112 (62.9)      Anti-Sm antibody positive  40 (22.7)      Anti-dsDNA antibodies positive  121 (68.4)      aPL positiveb  85 (48.3)      ANA positive  174 (98.3)  Disease activity at first assessment        PGA, mean (s.d.)  0.61 (0.78)      SELENA-SLEDAI, mean (s.d.)  6.56 (7.78)      SDI, median (IQR)  0 (0–2)  Treatment at first assessment        Systemic CS  96 (53.9)      Antimalarials  102 (57.6)      Immunosuppressant agents  65 (36.5)      Anticoagulants/low-dose aspirin  59/172 (34.3)    N (%)  Female sex  156 (87.6)  Caucasiana  139 (78.1)  Age at SLE diagnosis, mean (s.d.), years  34.7 (15.2)      Range  11.9–73.4  Age at first assessment, mean (s.d.), years  44.1 (14.6)      Range  16.6–85.3  ACR criteria at inclusion        Malar rash  76 (42.7)      Discoid rash  34 (19.2)      Photosensitivity  83 (46.6)      Nasopharyngeal ulcers  42 (23.6)      Arthritis  146 (82.0)      Pleuritis  43 (24.2)      Pericarditis  41 (23.0)      Renal disorder  81 (45.5)      Seizures  14 (7.9)      Psychosis  10 (5.6)      Haematologic disorder  112 (62.9)      Anti-Sm antibody positive  40 (22.7)      Anti-dsDNA antibodies positive  121 (68.4)      aPL positiveb  85 (48.3)      ANA positive  174 (98.3)  Disease activity at first assessment        PGA, mean (s.d.)  0.61 (0.78)      SELENA-SLEDAI, mean (s.d.)  6.56 (7.78)      SDI, median (IQR)  0 (0–2)  Treatment at first assessment        Systemic CS  96 (53.9)      Antimalarials  102 (57.6)      Immunosuppressant agents  65 (36.5)      Anticoagulants/low-dose aspirin  59/172 (34.3)  a Additional ethnic groups were African 8.4%, Asian 9.0%, Others 4%, unknown 0.6%. b Positivity defined by historical chart data. PGA: Physician Global Assessment; SDI: SLICC/ACR SLE Damage Index; IQR: interquartile range. Table 1 Baseline characteristics of the 178 patients with SLE   N (%)  Female sex  156 (87.6)  Caucasiana  139 (78.1)  Age at SLE diagnosis, mean (s.d.), years  34.7 (15.2)      Range  11.9–73.4  Age at first assessment, mean (s.d.), years  44.1 (14.6)      Range  16.6–85.3  ACR criteria at inclusion        Malar rash  76 (42.7)      Discoid rash  34 (19.2)      Photosensitivity  83 (46.6)      Nasopharyngeal ulcers  42 (23.6)      Arthritis  146 (82.0)      Pleuritis  43 (24.2)      Pericarditis  41 (23.0)      Renal disorder  81 (45.5)      Seizures  14 (7.9)      Psychosis  10 (5.6)      Haematologic disorder  112 (62.9)      Anti-Sm antibody positive  40 (22.7)      Anti-dsDNA antibodies positive  121 (68.4)      aPL positiveb  85 (48.3)      ANA positive  174 (98.3)  Disease activity at first assessment        PGA, mean (s.d.)  0.61 (0.78)      SELENA-SLEDAI, mean (s.d.)  6.56 (7.78)      SDI, median (IQR)  0 (0–2)  Treatment at first assessment        Systemic CS  96 (53.9)      Antimalarials  102 (57.6)      Immunosuppressant agents  65 (36.5)      Anticoagulants/low-dose aspirin  59/172 (34.3)    N (%)  Female sex  156 (87.6)  Caucasiana  139 (78.1)  Age at SLE diagnosis, mean (s.d.), years  34.7 (15.2)      Range  11.9–73.4  Age at first assessment, mean (s.d.), years  44.1 (14.6)      Range  16.6–85.3  ACR criteria at inclusion        Malar rash  76 (42.7)      Discoid rash  34 (19.2)      Photosensitivity  83 (46.6)      Nasopharyngeal ulcers  42 (23.6)      Arthritis  146 (82.0)      Pleuritis  43 (24.2)      Pericarditis  41 (23.0)      Renal disorder  81 (45.5)      Seizures  14 (7.9)      Psychosis  10 (5.6)      Haematologic disorder  112 (62.9)      Anti-Sm antibody positive  40 (22.7)      Anti-dsDNA antibodies positive  121 (68.4)      aPL positiveb  85 (48.3)      ANA positive  174 (98.3)  Disease activity at first assessment        PGA, mean (s.d.)  0.61 (0.78)      SELENA-SLEDAI, mean (s.d.)  6.56 (7.78)      SDI, median (IQR)  0 (0–2)  Treatment at first assessment        Systemic CS  96 (53.9)      Antimalarials  102 (57.6)      Immunosuppressant agents  65 (36.5)      Anticoagulants/low-dose aspirin  59/172 (34.3)  a Additional ethnic groups were African 8.4%, Asian 9.0%, Others 4%, unknown 0.6%. b Positivity defined by historical chart data. PGA: Physician Global Assessment; SDI: SLICC/ACR SLE Damage Index; IQR: interquartile range. Table 2 Prevalence of aPL in 178 SLE patients at entry   N (%)  Criteria aPL        LA (SCT)  35/175a (20.0)      LA (dRVVT)  40/172 (23.3)      aCL IgM  13/176 (7.4)      aCL IgG  37/176 (21.0)      Antiβ2GP1 IgM  12/176 (6.8)      Antiβ2GP1 IgG  40/176 (22.7)      Single positive (including double and triple positive)  75/169 (44.4)      Double positive (including triple positive)  35/171 (20.5)      Triple positive  22/175 (12.6)      Negative for all  94/169 (55.6)  Non-criteria aPL        aPS-PT IgM  47/178 (26.4)      aPS-PT IgG  29/178 (16.3)      aD1 IgG  26/164 (15.9)    N (%)  Criteria aPL        LA (SCT)  35/175a (20.0)      LA (dRVVT)  40/172 (23.3)      aCL IgM  13/176 (7.4)      aCL IgG  37/176 (21.0)      Antiβ2GP1 IgM  12/176 (6.8)      Antiβ2GP1 IgG  40/176 (22.7)      Single positive (including double and triple positive)  75/169 (44.4)      Double positive (including triple positive)  35/171 (20.5)      Triple positive  22/175 (12.6)      Negative for all  94/169 (55.6)  Non-criteria aPL        aPS-PT IgM  47/178 (26.4)      aPS-PT IgG  29/178 (16.3)      aD1 IgG  26/164 (15.9)  a Number of available determinations. Antiβ2GP1: anti-β2-glycoprotein 1; aD1: anti-domain 1 of β2GP1; dRVVT: dilute Russell’s viper venom test; SCT: silica clotting time. Table 2 Prevalence of aPL in 178 SLE patients at entry   N (%)  Criteria aPL        LA (SCT)  35/175a (20.0)      LA (dRVVT)  40/172 (23.3)      aCL IgM  13/176 (7.4)      aCL IgG  37/176 (21.0)      Antiβ2GP1 IgM  12/176 (6.8)      Antiβ2GP1 IgG  40/176 (22.7)      Single positive (including double and triple positive)  75/169 (44.4)      Double positive (including triple positive)  35/171 (20.5)      Triple positive  22/175 (12.6)      Negative for all  94/169 (55.6)  Non-criteria aPL        aPS-PT IgM  47/178 (26.4)      aPS-PT IgG  29/178 (16.3)      aD1 IgG  26/164 (15.9)    N (%)  Criteria aPL        LA (SCT)  35/175a (20.0)      LA (dRVVT)  40/172 (23.3)      aCL IgM  13/176 (7.4)      aCL IgG  37/176 (21.0)      Antiβ2GP1 IgM  12/176 (6.8)      Antiβ2GP1 IgG  40/176 (22.7)      Single positive (including double and triple positive)  75/169 (44.4)      Double positive (including triple positive)  35/171 (20.5)      Triple positive  22/175 (12.6)      Negative for all  94/169 (55.6)  Non-criteria aPL        aPS-PT IgM  47/178 (26.4)      aPS-PT IgG  29/178 (16.3)      aD1 IgG  26/164 (15.9)  a Number of available determinations. Antiβ2GP1: anti-β2-glycoprotein 1; aD1: anti-domain 1 of β2GP1; dRVVT: dilute Russell’s viper venom test; SCT: silica clotting time. Correlation between criteria and non-criteria aPL antibodies titres We assessed whether there was a relationship in the titres of aPL tests obtained with the nine assays performed. While they all were positively correlated (P < 0.001), the level of correlation was very high only for a few of them (Table 3). Thus, the strongest correlations with a Pearson correlation coefficient (r) value of about 0.8 were observed for aD1 IgG with antiβ2GP1 IgG and aCL IgG (denoted with the superscript b in Table 3). Similarly, aCL IgM and antiβ2GP1 IgM was also strongly correlated. Lower levels of correlation (Pearson’s r between 0.48 and 0.59) were found between dRVVT and IgG titres of aCL, antiβ2GP1, aD1, and aPS-PT IgG and IgM aPS-PT IgM (denoted with the superscript c in Table 3). The correlations were weakest between aPS-PT IgG/IgM, aCL IgG/IgM and antiβ2GP1 IgG, as well as aD1 IgG. Table 3 Pairwise Pearson correlation coefficients among aPL titres in 178 patients with SLE Test  SCT  aCL IgMa  aCL IgGa  Antiβ2GP1 IgMa  Antiβ2GP1 IgGa  aPS-PT IgMa  aPS-PT IgGa  aD1 IgGa  dRVVT  0.77b  0.37  0.49c  0.43  0.59c  0.43  0.50c  0.56c  SCT    0.25  0.39  0.27  0.47c  0.29  0.50c  0.46c  aCL IgMa      0.38  0.86b  0.41  0.51c  0.19  0.37  aCL IgGa        0.43  0.81b  0.30  0.39  0.79b  Antiβ2GP1 IgMa          0.49  0.52c  0.26  0.41  Antiβ2GP1 IgGa            0.35  0.49c  0.87b  aPS-PT IgMa              0.36  0.32  aPS-PT IgGa                0.46c  Test  SCT  aCL IgMa  aCL IgGa  Antiβ2GP1 IgMa  Antiβ2GP1 IgGa  aPS-PT IgMa  aPS-PT IgGa  aD1 IgGa  dRVVT  0.77b  0.37  0.49c  0.43  0.59c  0.43  0.50c  0.56c  SCT    0.25  0.39  0.27  0.47c  0.29  0.50c  0.46c  aCL IgMa      0.38  0.86b  0.41  0.51c  0.19  0.37  aCL IgGa        0.43  0.81b  0.30  0.39  0.79b  Antiβ2GP1 IgMa          0.49  0.52c  0.26  0.41  Antiβ2GP1 IgGa            0.35  0.49c  0.87b  aPS-PT IgMa              0.36  0.32  aPS-PT IgGa                0.46c  a Logarithm transformation. b Pearson correlation coefficient above 0.75. c Pearson correlation coefficient around 0.5. Antiβ2GP1: anti-β2-glycoprotein 1; aD1: anti-domain 1 of β2GP1; dRVVT: dilute Russell’s viper venom test; SCT: silica clotting time. All P < 0.001. Numbers of pairs vary between 251 and 283. Table 3 Pairwise Pearson correlation coefficients among aPL titres in 178 patients with SLE Test  SCT  aCL IgMa  aCL IgGa  Antiβ2GP1 IgMa  Antiβ2GP1 IgGa  aPS-PT IgMa  aPS-PT IgGa  aD1 IgGa  dRVVT  0.77b  0.37  0.49c  0.43  0.59c  0.43  0.50c  0.56c  SCT    0.25  0.39  0.27  0.47c  0.29  0.50c  0.46c  aCL IgMa      0.38  0.86b  0.41  0.51c  0.19  0.37  aCL IgGa        0.43  0.81b  0.30  0.39  0.79b  Antiβ2GP1 IgMa          0.49  0.52c  0.26  0.41  Antiβ2GP1 IgGa            0.35  0.49c  0.87b  aPS-PT IgMa              0.36  0.32  aPS-PT IgGa                0.46c  Test  SCT  aCL IgMa  aCL IgGa  Antiβ2GP1 IgMa  Antiβ2GP1 IgGa  aPS-PT IgMa  aPS-PT IgGa  aD1 IgGa  dRVVT  0.77b  0.37  0.49c  0.43  0.59c  0.43  0.50c  0.56c  SCT    0.25  0.39  0.27  0.47c  0.29  0.50c  0.46c  aCL IgMa      0.38  0.86b  0.41  0.51c  0.19  0.37  aCL IgGa        0.43  0.81b  0.30  0.39  0.79b  Antiβ2GP1 IgMa          0.49  0.52c  0.26  0.41  Antiβ2GP1 IgGa            0.35  0.49c  0.87b  aPS-PT IgMa              0.36  0.32  aPS-PT IgGa                0.46c  a Logarithm transformation. b Pearson correlation coefficient above 0.75. c Pearson correlation coefficient around 0.5. Antiβ2GP1: anti-β2-glycoprotein 1; aD1: anti-domain 1 of β2GP1; dRVVT: dilute Russell’s viper venom test; SCT: silica clotting time. All P < 0.001. Numbers of pairs vary between 251 and 283. Given the substantial heterogeneity in aPL titres in SLE patients, we conducted an exploratory factor analysis to identify patterns of antibody responses. This analysis suggested three factors with an eigenvalue >1, which together explained 79.5% of total variance (Table 4). The measure of sampling adequacy was 0.82, above the threshold of 0.8. Factor 1 grouped aCL IgG, antiβ2GP1 IgG and aD1 IgG. Factor 2 grouped dRVVT, SCT and aPS-PT IgG. Factor 3 grouped aCL IgM, antiβ2GP1 IgM and aPS-PT IgM. Thus, the nine tests performed provide results that can be grouped into three homogeneous ensembles. Interestingly, these groups respect isotypes differences and only aPS-PT IgG significantly aggregates with LA. Table 4 Factor analysis of nine aPL tests identifies three groups of antibodies Test  Factor     1  2  3  dRVVTa  0.30  0.77  0.29  SCTa  0.18  0.83  0.12  aCL IgMa  0.22  0.06  0.91  aCL IgGa  0.88  0.21  0.20  Antiβ2GP1 IgMa  0.29  0.10  0.88  Antiβ2GP1 IgGa  0.87  0.32  0.23  aPS-PT IgMa  0.04  0.38  0.69  aPS-PT IgGa  0.29  0.72  0.07  aD1 IgGa  0.87  0.29  0.18  Test  Factor     1  2  3  dRVVTa  0.30  0.77  0.29  SCTa  0.18  0.83  0.12  aCL IgMa  0.22  0.06  0.91  aCL IgGa  0.88  0.21  0.20  Antiβ2GP1 IgMa  0.29  0.10  0.88  Antiβ2GP1 IgGa  0.87  0.32  0.23  aPS-PT IgMa  0.04  0.38  0.69  aPS-PT IgGa  0.29  0.72  0.07  aD1 IgGa  0.87  0.29  0.18  Factor loadings after varimax rotation are shown. Primary factor loadings are in bold. Data from 162 patients with complete data. a Logarithm transformation. Antiβ2GP1: anti-β2-glycoprotein 1; aD1: anti-domain 1 of β2GP1; dRVVT: dilute Russell’s viper venom test; SCT: silica clotting time. Table 4 Factor analysis of nine aPL tests identifies three groups of antibodies Test  Factor     1  2  3  dRVVTa  0.30  0.77  0.29  SCTa  0.18  0.83  0.12  aCL IgMa  0.22  0.06  0.91  aCL IgGa  0.88  0.21  0.20  Antiβ2GP1 IgMa  0.29  0.10  0.88  Antiβ2GP1 IgGa  0.87  0.32  0.23  aPS-PT IgMa  0.04  0.38  0.69  aPS-PT IgGa  0.29  0.72  0.07  aD1 IgGa  0.87  0.29  0.18  Test  Factor     1  2  3  dRVVTa  0.30  0.77  0.29  SCTa  0.18  0.83  0.12  aCL IgMa  0.22  0.06  0.91  aCL IgGa  0.88  0.21  0.20  Antiβ2GP1 IgMa  0.29  0.10  0.88  Antiβ2GP1 IgGa  0.87  0.32  0.23  aPS-PT IgMa  0.04  0.38  0.69  aPS-PT IgGa  0.29  0.72  0.07  aD1 IgGa  0.87  0.29  0.18  Factor loadings after varimax rotation are shown. Primary factor loadings are in bold. Data from 162 patients with complete data. a Logarithm transformation. Antiβ2GP1: anti-β2-glycoprotein 1; aD1: anti-domain 1 of β2GP1; dRVVT: dilute Russell’s viper venom test; SCT: silica clotting time. To further address the correlation and concomitant positivity between criteria and non-criteria antibodies we assessed the titre of aD1 IgG and aPS-PT IgG and IgM in samples with no, single, double or triple positivity according to criteria assays. As expected, the titre of the three non-criteria antibodies increased with the increase of the number of positive criteria assays. However, we observed an exquisite enrichment for high aD1 IgG titres in triple-positive samples, increasing by more than two orders of magnitude from no positive to triple-positive criteria sera (supplementary Table S2, available at Rheumatology online). For comparison, in the same groups aPS-PT titres increased by 5-fold only for both IgG and IgM (supplementary Table S2, available at Rheumatology online). Thus, in SLE sera there is a preferential enrichment for high titre aD1 IgG in triple-positive samples. Persistence of aPL The next question we asked was whether the presence or absence of individual aCL antibody specificities was sustained over time by assessing sera taken on average 1 year apart in 105 individuals. The clinical characteristics of these individuals did not significantly differ from the initial population (supplementary Table S3, available at Rheumatology online). Overall, positivity or negativity was sustained over time (supplementary Table S4, available at Rheumatology online). Kappa statistics revealed that SCT was the least persistent. In addition, aCL IgG as well as aPS-PT IgG and IgM had a kappa statistic ranging from 0.69 to 0.74, thus indicating some variation over time. However, the intra-class correlation coefficients on continuous measurements were very high for all tests (except for SCT), ranging from 0.83 to 0.99 (supplementary Table S4, available at Rheumatology online), thus indicating that results located near the cut-off values could explain most of the variations observed for dichotomous results. Association of aPL tests with thrombotic events We next addressed the question of whether aPL was associated with thrombotic events in our SLE cohort. The overall frequency of thrombotic events was low. Twenty-two venous, 20 arterial and 37 either venous or arterial events were recorded. Among the nine distinct assays performed to detect aPL antibodies, only aPS-PT IgG titres were associated significantly with venous and composite venous or arterial events (supplementary Table S5, available at Rheumatology online). To substantiate these findings, we used the presence or absence of aPS-PT antibodies as a dichotomous variable and assessed their association with thrombotic events. aPS-PT IgG but not aPS-PT IgM was strongly associated (P = 0.012) with venous thrombotic events and to a lesser extent to the composite arterial and venous vascular events (P = 0.077) (Table 5). In addition, when using the three factors obtained by exploratory factor analysis only the factor that grouped LA and anti-PS-PT IgG was positively associated with thrombotic events, and the association was statistically significant for the composite of venous and arterial events (Table 6). The factor grouping aCL IgG, antiβ2GP1 IgG and aD1 IgG had much weaker association, while the factor grouping aCL IgM, antiβ2GP1 IgM and aPS-PT IgM had a negative association with vascular events (factor scores were lower in patients with events); neither one of the latter was statistically significant. Of interest, when triple (data not shown) or quadruple (supplementary Table S6, available at Rheumatology online) positive samples were taken into consideration, the association with vascular thrombotic events was statistically significant only for IgG responses. Thus, the non-criteria aPS-PT IgG—by itself and in conjunction with LA—appears to be associated with thrombotic events in Swiss SLE patients. Table 5 Association of aPS-PT complex IgG with thrombotic events in SLE   Venous vascular events (n = 20)   Arterial vascular events (n = 17)   Venous or arterial vascular events (n = 34)   Test  Yes  No  P-value  Yes  No  P-value  Yes  No  P-value  aPS-PT IgG (pos = 29)  8  21  0.012  2  27  0.537  10  19  0.077  aPS-PT IgM (pos = 47)  3  43  0.445  6  41  0.440  10  37  1.000  aPS-PT IgG or aPS-PT IgM (pos = 59)  9  50  0.470  8  51  0.615  17  42  0.078    Venous vascular events (n = 20)   Arterial vascular events (n = 17)   Venous or arterial vascular events (n = 34)   Test  Yes  No  P-value  Yes  No  P-value  Yes  No  P-value  aPS-PT IgG (pos = 29)  8  21  0.012  2  27  0.537  10  19  0.077  aPS-PT IgM (pos = 47)  3  43  0.445  6  41  0.440  10  37  1.000  aPS-PT IgG or aPS-PT IgM (pos = 59)  9  50  0.470  8  51  0.615  17  42  0.078  Number of aPS-PT IgG and/or IgM positive samples with thrombotic events in SLE. P-value determined by the Fisher’s exact test. pos: positive. Table 5 Association of aPS-PT complex IgG with thrombotic events in SLE   Venous vascular events (n = 20)   Arterial vascular events (n = 17)   Venous or arterial vascular events (n = 34)   Test  Yes  No  P-value  Yes  No  P-value  Yes  No  P-value  aPS-PT IgG (pos = 29)  8  21  0.012  2  27  0.537  10  19  0.077  aPS-PT IgM (pos = 47)  3  43  0.445  6  41  0.440  10  37  1.000  aPS-PT IgG or aPS-PT IgM (pos = 59)  9  50  0.470  8  51  0.615  17  42  0.078    Venous vascular events (n = 20)   Arterial vascular events (n = 17)   Venous or arterial vascular events (n = 34)   Test  Yes  No  P-value  Yes  No  P-value  Yes  No  P-value  aPS-PT IgG (pos = 29)  8  21  0.012  2  27  0.537  10  19  0.077  aPS-PT IgM (pos = 47)  3  43  0.445  6  41  0.440  10  37  1.000  aPS-PT IgG or aPS-PT IgM (pos = 59)  9  50  0.470  8  51  0.615  17  42  0.078  Number of aPS-PT IgG and/or IgM positive samples with thrombotic events in SLE. P-value determined by the Fisher’s exact test. pos: positive. Table 6 The combination of LA with aPS-PT IgG is associated with thrombotic events in SLE   Venous vascular events   Arterial vascular events   Venous or arterial vascular events   Test  Yes (N = 20)  No (N = 142)  P-value  Yes (N = 17)  No (N = 145)  P-value  Yes (N = 34)  No (N = 128)  P-value  Factor 1: aCL IgG, antiβ2GP1 IgG and aD1 IgG  0.23  −0.03  0.27  −0.06  0.01  0.77  0.13  −0.04  0.38  Factor 2: LA and aPS-PT IgG  0.36  −0.05  0.084  0.34  −0.04  0.14  0.35  −0.09  0.02  Factor 3: aCL IgM, antiβ2GP1 IgM and aPS-PT IgM  −0.22  0.03  0.30  −0.18  0.02  0.42  −0.17  0.05  0.25    Venous vascular events   Arterial vascular events   Venous or arterial vascular events   Test  Yes (N = 20)  No (N = 142)  P-value  Yes (N = 17)  No (N = 145)  P-value  Yes (N = 34)  No (N = 128)  P-value  Factor 1: aCL IgG, antiβ2GP1 IgG and aD1 IgG  0.23  −0.03  0.27  −0.06  0.01  0.77  0.13  −0.04  0.38  Factor 2: LA and aPS-PT IgG  0.36  −0.05  0.084  0.34  −0.04  0.14  0.35  −0.09  0.02  Factor 3: aCL IgM, antiβ2GP1 IgM and aPS-PT IgM  −0.22  0.03  0.30  −0.18  0.02  0.42  −0.17  0.05  0.25  Exploratory factor analysis performed according to Bartko [27]. Three factors with an eigenvalue >1, explaining 79.5% of total variance, are reported. Mean values of three factors combining aPL in the presence and absence of thrombotic events, in 162 patients (16 had missing data). All factors had a mean of 0 and s.d. of 1 in the whole sample. The measure of sampling adequacy was 0.82, above the threshold of 0.8. Antiβ2GP1: anti-β2-glycoprotein 1; aD1: anti-domain 1 of β2GP1. Table 6 The combination of LA with aPS-PT IgG is associated with thrombotic events in SLE   Venous vascular events   Arterial vascular events   Venous or arterial vascular events   Test  Yes (N = 20)  No (N = 142)  P-value  Yes (N = 17)  No (N = 145)  P-value  Yes (N = 34)  No (N = 128)  P-value  Factor 1: aCL IgG, antiβ2GP1 IgG and aD1 IgG  0.23  −0.03  0.27  −0.06  0.01  0.77  0.13  −0.04  0.38  Factor 2: LA and aPS-PT IgG  0.36  −0.05  0.084  0.34  −0.04  0.14  0.35  −0.09  0.02  Factor 3: aCL IgM, antiβ2GP1 IgM and aPS-PT IgM  −0.22  0.03  0.30  −0.18  0.02  0.42  −0.17  0.05  0.25    Venous vascular events   Arterial vascular events   Venous or arterial vascular events   Test  Yes (N = 20)  No (N = 142)  P-value  Yes (N = 17)  No (N = 145)  P-value  Yes (N = 34)  No (N = 128)  P-value  Factor 1: aCL IgG, antiβ2GP1 IgG and aD1 IgG  0.23  −0.03  0.27  −0.06  0.01  0.77  0.13  −0.04  0.38  Factor 2: LA and aPS-PT IgG  0.36  −0.05  0.084  0.34  −0.04  0.14  0.35  −0.09  0.02  Factor 3: aCL IgM, antiβ2GP1 IgM and aPS-PT IgM  −0.22  0.03  0.30  −0.18  0.02  0.42  −0.17  0.05  0.25  Exploratory factor analysis performed according to Bartko [27]. Three factors with an eigenvalue >1, explaining 79.5% of total variance, are reported. Mean values of three factors combining aPL in the presence and absence of thrombotic events, in 162 patients (16 had missing data). All factors had a mean of 0 and s.d. of 1 in the whole sample. The measure of sampling adequacy was 0.82, above the threshold of 0.8. Antiβ2GP1: anti-β2-glycoprotein 1; aD1: anti-domain 1 of β2GP1. Discussion Taking advantage of the Swiss SLE cohort [19] as well as of the availability of novel commercial kits to assess the presence of aPS-PT IgG and IgM as well as aD1 IgG, our study examined the analytical performance of nine tests aiming at detecting aPL antibodies and their clinical associations. Our results show that: the tests employed here provide reliable results; aPL tend to persist with time; the titres of criteria and non-criteria aPL all correlate, but vary considerably in the strength of correlations; exploratory factor analysis identifies three groups of specificities segregating aPL isotypes and grouping LA only with aPS-PT IgG; high titre aD1 IgG are highly enriched in triple-positive samples; and aPS-PT IgG—by itself and in combination with LA—stands out in patients with an history of thrombotic events. The reproducibility of results is, of course, essential to provide reliable conclusions. While all samples were tested with technical duplicates, we tested also biological duplicates in a random subgroup of about 70 patients (supplementary Table S1, available at Rheumatology online). The results obtained were robust and satisfied the requirements of consensus experts [12]. This allowed the study of the persistence of aPL in time. As expected, and in agreement with published data [28], most negative samples remained negative 1 year apart, while somehow, less frequently positive samples remained positive. Low titre antibody positivity explained most of this variability, which was independent of treatment (data not shown). Most importantly, techniques of dosage and cut-off values used here were validated in previous studies by us and similar to those used by others [20, 21, 23, 29–32]. The investigation of concordance and discordance among the aPL assays provided interesting data that complement previous studies. Overall, the correlation coefficients between the tests expressed as continuous variables were statistically highly significant, which is in line with previous reports by us and others [14, 17, 20, 23, 32, 33]. However, the correlation coefficients varied substantially among pairs of tests. The exploratory factor analysis identified the underlying three factors that associated the tests in three distinct groups: one grouping aCL IgG, antiβ2GP1 IgG and aD1 IgG, a second grouping LA and aPS-PT IgG and the third grouping aCL IgM, antiβ2GP1 IgM and aPS-PT IgM. The association of specificities we found in the first group has already been documented in the literature [20, 32, 33], and indeed D1 has been previously shown to be the immune-dominant epitope of β2GP1 IgG. Previous studies, however, focused on the performance of the tests in an APS patient population with a high prevalence of thrombotic events [34], which is not the case in our SLE cohort. Our factor three grouped IgM responses, with very high correlation coefficients, much higher than those correlating IgM with IgG specific for the same antigen. This indicates that some SLE patients are developing IgM responses against distinct phospholipid (PL) specificities that are sustained in time and do not tend to switch to IgG. Of interest, IgM specificities had similar correlation coefficients with dRVVT to those of their IgG counterpart. In this perspective, our results are in agreement with previous data indicating association of aCL IgM, antiβ2GP1 IgM and aPS-PT IgM with LA [17] and support the contention that IgM may also contribute to LA. Finally, our factor two grouped LA (both dRVTT and/or SCT) and aPS-PT IgG. Thus, aPS-PT IgG in our study appears to have distinct characteristics compared with the other IgG aPL assays. This, may be reflected in biological properties which may then have clinical impact as discussed here below. Several studies have reported that patients with triple-positive aPL carry a particularly high risk of thrombotic events [8, 32, 35–37]. Furthermore, aD1 IgG was shown to be elevated in triple-positive sera [38, 39]. Our study confirms that aD1 IgG are particularly high, in triple-positive sera, possibly more than aPS-PT IgG (and aPS-PT IgM) in SLE patients. In contrast, we did not confirm the previous observation that among antiβ2GP1-positive patients aD1 positivity implies an increased risk of thrombotic events [14]. This is consistent with other reports in SLE [17]. Two non-mutually exclusive hypotheses may explain this discrepancy. First, the epitopes within aD1 recognized by SLE IgG may differ from those recognized by primary APS. This may affect their pro-thrombotic activity. Second, thrombotic events in SLE may be due, at least in part, to factors independent form aPL, which contribution may lessen the importance of aD1 IgG. In our cohort, the presence of aPS-PT IgG was particularly associated with venous thrombotic events, and interestingly the combination of aPS-PT IgG with LA—identified in the exploratory factor analysis—was associated in a statistically significant manner when venous and arterial events were grouped. This confirms data by others in primary APS [18] and in SLE [17]. However, many other studies identified also aCL IgG/IgM and antiβ2GP1 IgG/IgM to be associated with thrombosis [40–42]. These differences may be due to the clinical characteristic of the SLE patients enrolled in our cohort and in particular to the relatively low incidence of thrombotic events (21%) when compared with other study populations where the prevalence of such events ranged from 25 to 50% [17, 42, 43], thus reducing the power of the statistical analysis we conducted. This may further explain why in our cohort aD1 IgG was not associated with thrombotic events in contrasts with the findings of others in APS [37]. A further point of interest is that, due to their strong association with LA and with vascular events, the assessment of of aPS-PT IgG in substitution of LA may particularly useful in anticoagulated patients in which the determination of LA is technically difficult. It has to be stressed that our study population was followed in tertiary centres across Switzerland, patients attended by internal medicine, nephrology, clinical immunology and rheumatology specialists. This may represent a blend of patients that may differ from those attended to in dedicated, single centres. A weakness of our study is that ours is not an inception cohort, and most thrombotic events were recorded by review of the clinical charts. Others have found that thrombotic events accumulate with time, particularly in triple-positive individuals [35]. It is possible that the prospective follow-up of our cohort will reveal other associations. As an additional limitation, we did not include in our study aPL IgA or other non-criteria PL specificities beside PS-PT and D1, thus limiting the potential for detecting important clinical associations. However, the analytical performance of our study met the highest standards and results were highly reproducible. In conclusion, our study stresses the importance of aPS-PT IgG in identifying patients at risk for thrombosis in SLE and suggests the usefulness of adding this specificity when searching for aPL antibodies. Prospective studies are needed to support this contention. Acknowledgements We thank D. Wahl (Nancy, France), J.-C. Gris (Nîmes, France) and K.M. Devreese (Gent, Belgium) for their critical reading of the manuscript. We acknowledge INOVA (Switzerland) for kindly supplying reagents for aPL antibody determination (QUANTA Lite and QUANTA Flash) free of charge. Authors’ contribtions: T.M., C.R., T.P., P.d.M. and C.C. analysed the data and drafted the manuscript; M.T. and U.H.-D. provided critical reading; C.R., M.T., U.H.-D. and C.C. acquired the clinical data and collected samples; T.M. and P.d.M. acquired the laboratory data; and T.P. performed the statistical analysis. Funding: This work was supported in part by a Gebert Ruf unrestricted grant to SSCS, and the ISTH2007 presidential fund (to C.C.). Disclosure statement: The authors have declared no conflicts of interest. Supplementary data Supplementary data are available at Rheumatology online. References 1 Liu Z, Davidson A. Taming lupus-a new understanding of pathogenesis is leading to clinical advances. Nat Med  2012; 18: 871– 82. Google Scholar CrossRef Search ADS PubMed  2 D’Cruz DP, Khamashta MA, Hughes GR. Systemic lupus erythematosus. Lancet  2007; 369: 587– 96. Google Scholar CrossRef Search ADS PubMed  3 Tan EM, Cohen AS, Fries JF et al.   The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum  1982; 25: 1271– 7. Google Scholar CrossRef Search ADS PubMed  4 Hochberg MC. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum  1997; 40: 1725. Google Scholar CrossRef Search ADS PubMed  5 Meroni PL, Borghi MO, Raschi E, Tedesco F. Pathogenesis of antiphospholipid syndrome: understanding the antibodies. Nat Rev Rheumatol  2011; 7: 330– 9. Google Scholar CrossRef Search ADS PubMed  6 Giannakopoulos B, Krilis SA. The pathogenesis of the antiphospholipid syndrome. N Engl J Med  2013; 368: 1033– 44. Google Scholar CrossRef Search ADS PubMed  7 Habe K, Wada H, Matsumoto T et al.   Presence of antiphospholipid antibodies as a risk factor for thrombotic events in patients with connective tissue diseases and idiopathic thrombocytopenic purpura. Intern Med  2016; 55: 589– 95. Google Scholar CrossRef Search ADS PubMed  8 Yelnik CM, Laskin CA, Porter TF et al.   Lupus anticoagulant is the main predictor of adverse pregnancy outcomes in aPL-positive patients: validation of PROMISSE study results. Lupus Sci Med  2016; 3 e000131. Google Scholar CrossRef Search ADS PubMed  9 Bertolaccini ML, Sanna G. Recent advances in understanding antiphospholipid syndrome. F1000Res  2016; 5: 2908. Google Scholar CrossRef Search ADS PubMed  10 Unlu O, Zuily S, Erkan D. The clinical significance of antiphospholipid antibodies in systemic lupus erythematosus. Eur J Rheumatol  2016; 3: 75– 84. Google Scholar CrossRef Search ADS PubMed  11 Pons-Estel GJ, Andreoli L, Scanzi F, Cervera R, Tincani A. The antiphospholipid syndrome in patients with systemic lupus erythematosus. J Autoimmun  2017; 76: 10– 20. Google Scholar CrossRef Search ADS PubMed  12 Miyakis S, Lockshin MD, Atsumi T et al.   International consensus statement on an update of the classification criteria for definite antiphospholipid syndrome (APS). J Thromb Haemost  2006; 4: 295– 306. Google Scholar CrossRef Search ADS PubMed  13 Meroni PL, Chighizola CB, Rovelli F, Gerosa M. Antiphospholipid syndrome in 2014: more clinical manifestations, novel pathogenic players and emerging biomarkers. Arthritis Res Ther  2014; 16: 209. Google Scholar CrossRef Search ADS PubMed  14 de Laat B, Pengo V, Pabinger I et al.   The association between circulating antibodies against domain I of beta2-glycoprotein I and thrombosis: an international multicenter study. J Thromb Haemost  2009; 7: 1767– 73. Google Scholar CrossRef Search ADS PubMed  15 Broen JC, Dieude P, Vonk MC et al.   Polymorphisms in the interleukin 4, interleukin 13, and corresponding receptor genes are not associated with systemic sclerosis and do not influence gene expression. J Rheumatol  2012; 39: 112– 8. Google Scholar CrossRef Search ADS PubMed  16 Mahler M, Norman GL, Meroni PL, Khamashta M. Autoantibodies to domain 1 of beta 2 glycoprotein 1: a promising candidate biomarker for risk management in antiphospholipid syndrome. Autoimmun Rev  2012; 12: 313– 7. Google Scholar CrossRef Search ADS PubMed  17 Akhter E, Shums Z, Norman GL et al.   Utility of antiphosphatidylserine/prothrombin and IgA antiphospholipid assays in systemic lupus erythematosus. J Rheumatol  2013; 40: 282– 6. Google Scholar CrossRef Search ADS PubMed  18 Sciascia S, Sanna G, Murru V et al.   Anti-prothrombin (aPT) and anti-phosphatidylserine/prothrombin (aPS-PT) antibodies and the risk of thrombosis in the antiphospholipid syndrome. A systematic review. J Thromb Haemost  2014; 111: 354– 64. Google Scholar CrossRef Search ADS   19 Ribi C, Trendelenburg M, Gayet-Ageron A et al.   The Swiss Systemic lupus erythematosus Cohort Study (SSCS) - cross-sectional analysis of clinical characteristics and treatments across different medical disciplines in Switzerland. Swiss Med Wkly  2014; 144: w13990. Google Scholar PubMed  20 Van Hoecke F, Persijn L, Decavele AS, Devreese K. Performance of two new, automated chemiluminescence assay panels for anticardiolipin and anti-beta2-glycoprotein I antibodies in the laboratory diagnosis of the antiphospholipid syndrome. Int J Lab Hematol  2012; 34: 630– 40. Google Scholar CrossRef Search ADS PubMed  21 Fontana P, Poncet A, Lindhoff-Last E, de Moerloose P, Devreese KM. Refinement of the cutoff values of the HemosIL AcuStar assay for the detection of anticardiolipin and anti-beta2 glycoprotein-1 antibodies. J Thromb Haemost  2014; 12: 2034– 7. Google Scholar CrossRef Search ADS PubMed  22 Devreese KM. Antiphospholipid antibody testing and standardization. Int J Lab Hematol  2014; 36: 352– 63. Google Scholar CrossRef Search ADS PubMed  23 Marchetti T, de Moerloose P, Gris JC. Antiphospholipid antibodies and the risk of severe and non-severe pre-eclampsia: the NOHA case-control study. J Thromb Haemost  2016; 14: 675– 84. Google Scholar CrossRef Search ADS PubMed  24 Gladman D, Ginzler E, Goldsmith C et al.   The development and initial validation of the Systemic Lupus International Collaborating Clinics/American College of Rheumatology damage index for systemic lupus erythematosus. Arthritis Rheum  1996; 39: 363– 9. Google Scholar CrossRef Search ADS PubMed  25 Bertoli AM, Vilá LM, GS A et al.   Factors associated with arterial vascular events in PROFILE: a Multiethnic Lupus Cohort. Lupus  2009; 18: 958– 65. Google Scholar CrossRef Search ADS PubMed  26 Joliffe IT, Morgan BJ. Principal component analysis and exploratory factor analysis. Stat Methods Med Res  1992; 1: 69– 95. Google Scholar CrossRef Search ADS PubMed  27 Bartko JJ. The intraclass correlation coefficient as a measure of reliability. Psychol Rep  1966; 19: 3– 11. Google Scholar CrossRef Search ADS PubMed  28 Erkan D, Derksen WJ, Kaplan V et al.   Real world experience with antiphospholipid antibody tests: how stable are results over time? Ann Rheum Dis  2005; 64: 1321– 5. Google Scholar CrossRef Search ADS PubMed  29 de Moerloose P, Reber G, Musial J, Arnout J. Analytical and clinical performance of a new, automated assay panel for the diagnosis of antiphospholipid syndrome. J Thromb Haemost  2010; 8: 1540– 6. Google Scholar CrossRef Search ADS PubMed  30 Sciascia S, Bertolaccini ML. Thrombotic risk assessment in APS: the Global APS Score (GAPSS). Lupus  2014; 23: 1286– 7. Google Scholar CrossRef Search ADS PubMed  31 Hata K, Andoh A, Shimada M et al.   IL-17 stimulates inflammatory responses via NF-kappaB and MAP kinase pathways in human colonic myofibroblasts. Am J Physiol Gastrointest Liver Physiol  2002; 282: G1035– 44. Google Scholar CrossRef Search ADS PubMed  32 De Craemer AS, Musial J, Devreese KM. Role of anti-domain 1-beta2 glycoprotein I antibodies in the diagnosis and risk stratification of antiphospholipid syndrome. J Thromb Haemost  2016; 14: 1779– 87. Google Scholar CrossRef Search ADS PubMed  33 Hoxha A, Ruffatti A, Mattia E et al.   Relationship between antiphosphatidylserine/prothrombin and conventional antiphospholipid antibodies in primary antiphospholipid syndrome. Clin Chem Lab Med  2015; 53: 1265– 70. Google Scholar CrossRef Search ADS PubMed  34 Mahler M, Albesa R, Zohoury N et al.   Autoantibodies to domain 1 of beta 2 glycoprotein I determined using a novel chemiluminescence immunoassay demonstrate association with thrombosis in patients with antiphospholipid syndrome. Lupus  2016; 25: 911– 6. Google Scholar CrossRef Search ADS PubMed  35 Pengo V, Ruffatti A, Legnani C et al.   Incidence of a first thromboembolic event in asymptomatic carriers of high-risk antiphospholipid antibody profile: a multicenter prospective study. Blood  2011; 118: 4714– 8. Google Scholar CrossRef Search ADS PubMed  36 Otomo K, Atsumi T, Amengual O et al.   Efficacy of the antiphospholipid score for the diagnosis of antiphospholipid syndrome and its predictive value for thrombotic events. Arthritis Rheum  2012; 64: 504– 12. Google Scholar CrossRef Search ADS PubMed  37 Pengo V, Ruffatti A, Tonello M et al.   Antiphospholipid syndrome: antibodies to Domain 1 of β2-glycoprotein 1 correctly classify patients at risk. J Thromb Haemost  2015; 13: 782– 7. Google Scholar CrossRef Search ADS PubMed  38 Banzato A, Pozzi N, Frasson R et al.   Antibodies to domain I of beta(2)glycoprotein I are in close relation to patients risk categories in antiphospholipid syndrome (APS). Thromb Res  2011; 128: 583– 6. Google Scholar CrossRef Search ADS PubMed  39 Iwaniec T, Kaczor MP, Celińska-Löwenhoff M, Polański S, Musiał J. Clinical significance of anti-domain 1 beta2-glycoprotein I antibodies in antiphospholipid syndrome. Thromb Res  2017; 153: 90– 4. Google Scholar CrossRef Search ADS PubMed  40 Sciascia S, Murru V, Sanna G et al.   Clinical accuracy for diagnosis of antiphospholipid syndrome in systemic lupus erythematosus: evaluation of 23 possible combinations of antiphospholipid antibody specificities. J Thromb Haemost  2012; 10: 2512– 8. Google Scholar CrossRef Search ADS PubMed  41 Sciascia S, Sanna G, Murru V, Khamashta MA, Bertolaccini ML. Validation of a commercially available kit to detect anti-phosphatidylserine/prothrombin antibodies in a cohort of systemic lupus erythematosus patients. Thromb Res  2014; 133: 451– 4. Google Scholar CrossRef Search ADS PubMed  42 Taraborelli M, Lazzaroni MG, Martinazzi N et al.   The role of clinically significant antiphospholipid antibodies in systemic lupus erythematosus. Reumatismo  2016; 68: 137– 43. Google Scholar CrossRef Search ADS PubMed  43 Sciascia S, Sanna G, Murru V et al.   GAPSS: the Global Anti-Phospholipid Syndrome Score. Rheumatology (Oxford)  2013; 52: 1397– 403. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Rheumatology Oxford University Press

Prevalence, persistence and clinical correlations of classic and novel antiphospholipid antibodies in systemic lupus erythematosus

Loading next page...
 
/lp/ou_press/prevalence-persistence-and-clinical-correlations-of-classic-and-novel-KpZh9qcFZJ
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For permissions, please email: journals.permissions@oup.com
ISSN
1462-0324
eISSN
1462-0332
D.O.I.
10.1093/rheumatology/key095
Publisher site
See Article on Publisher Site

Abstract

Abstract Objectives aPL are frequently present in SLE. In a well characterized SLE cohort we aimed at investigating the prevalence of aPL and assessing their analytical performance and clinical association by testing criteria specificities including LA, aCL IgG and IgM, anti-β2-glycoprotein 1 (antiβ2GP1) IgG and IgM, as well as the non-criteria aPS-PT IgG and IgM and anti-β2GP1 domain 1 (aD1) IgG. Methods We included 178 patients satisfying the ACR SLE classification criteria, from whom 283 samples and thrombotic events were collected longitudinally. Each sample was tested for criteria and non-criteria aPL using validated techniques in a single centre. Results All assays provided highly reproducible results. Of the samples, 42.5% were positive for at least one criteria assay, 20.5% showed double positivity and 12.6% triple positivity. All criteria and non-criteria specificities persisted over time. Most antibody titres were only moderately correlated; however, strong correlation was observed on one hand between aD1 IgG, antiβ2GP1 IgG and aCL IgG, and on the other between aPS-PT IgG and LA. aD1 IgG titres were extremely elevated in triple-positive samples. aPS-PT IgG by itself, and jointly with LA, was associated with thrombosis, an association mostly driven by venous thrombotic events. Conclusions In this SLE cohort, the non-criteria aPL aD1 IgG and aPS-PT IgG performed differently. aD1 IgG was highly enriched in triple-positive samples, and aPS-PT IgG, jointly with LA, was associated with thrombotic events. systemic lupus erythematosus, antiphospholipid, lupus anticoagulant, beta 2-glycoprotein 1, phosphatidylserine/prothrombin complex, domain 1 Rheumatology key messages High titers of anti-β2-glycoprotein 1 domain 1 IgG are characteristic of triple-positive SLE sera. Anti-phosphatidylserine/prothrombin complex IgG are associated with lupus anticoagulant in SLE. Anti-phosphatidylserine/prothrombin complex IgG by itself and jointly with lupus anticoagulant are associated with thrombotic events in SLE. Introduction SLE is a chronic inflammatory condition due to altered immunologic tolerance, biologically characterized by the presence of auto-antibodies directed against ubiquitous autoantigens [1, 2]. aPL are often present and part of the ACR classification criteria for SLE [3, 4]. aPL represent a heterogeneous family of antibodies directed against phospholipids and phospholipid-binding proteins [5, 6]. The presence of aPL in SLE has been associated with an increased risk of thrombosis, and poor pregnancy outcomes as in primary APS [7–9]. In addition, aPL are linked to more severe SLE features, including valve disease, pulmonary hypertension, livedo reticularis, thrombocytopaenia, haemolytic anaemia, acute/chronic renal vascular lesions and moderate/severe cognitive impairment, worse quality of life and higher risk of organ damage [10, 11]. The current biological criteria for APS classification rely on the identification of LA using coagulation assays and/or by detecting anti-β2-glycoprotein 1 (antiβ2GP1) and aCL IgG or IgM with solid phase assays [12]. Several additional antigenic specificities for aPL antibodies have recently attracted great interest [13]. The β2GP1 protein is composed of five domains, and domain 1 (D1), which is a cryptic epitope when β2GP1does not bind to anionic phospholipids, is thought to be preferentially recognized by antibodies in individuals with APS [14–16]. Interestingly, conflicting results have been reported on the association of anti-D1 (aD1) IgG and thrombotic events, with some reports indicating an increased risk of thrombosis in individuals with aD1 IgG [14] and others not confirming such an association, particularly in SLE [17]. Furthermore, IgG or IgM directed against the aPS-PT have been variably associated with thrombotic events. However, in a systematic review of published evidence, Sciascia et al. [18] report that the presence of aPS-PT IgG or IgM increases the risk (odds ratio = 5.11; 95% CI: 4.2, 6.3) of arterial and/or venous thrombosis. Thus, while aD1 and aPS-PT IgG and IgM are not currently among the criteria antibodies for APS, we felt it to be of interest to investigate whether the presence of these non-criteria antibodies, in association or not with criteria antibodies, could provide additional information on the clinical picture in SLE patients. In this study, taking advantage of the Swiss SLE cohort [19], we aimed to investigate: the prevalence at entry into the cohort of criteria and non-criteria aPL antibodies, tested in a central laboratory; their persistence over time; whether their presence and titres were correlated; and which combination of autoantibodies was most strongly associated with thrombotic events. Methods Plasma samples were obtained from 178 individuals satisfying at least four ACR SLE criteria and consecutively enrolled in the prospective Swiss SLE Cohort Study (SSCS) from April 2007 to January 2015 [19]. A second, follow-up sample was obtained from 105 individuals of the original cohort for a total of 283 distinct samples. Median interval (interquartile range) between the first and second samples was of 13.1 (3.6–61.4) months. Ethical clearance was obtained in all participating SSCS centres. Informed, written consent was obtained from all participants according to the Declaration of Helsinki. All samples (N = 283) were tested in a single laboratory in Geneva (Hemostasis Unit, University Hospital, Switzerland). LA was detected using ACL TOP (Instrumentation Laboratory, Paris, France), which detected the ability to prolong two phospholipid-dependent tests: the silica clotting time (SCT) and the dilute Russell’s viper venom time (dRVVT). The presence and titre of aCL and antiβ2GP1 IgG and IgM antibodies was assessed using the HemosIL AcuStar Coagulation Analyzer (Instrumentation Laboratory, Bedford, MA, USA) [20, 21]. The cut-off values for these tests were based on the 99th percentile and the corresponding 95% CI for each antibody of 626 healthy blood bank donors [21]. aPS-PT IgG and IgM antibodies were determined by ELISA using QUANTA Lite kits using the DSX automated apparatus (DSX/best 2000; INOVA, San Diego, CA, USA). aD1 antibodies were determined with the HemosIL AcuStar Coagulation Analyzer using QUANTA Flash. Reagent Kits were kindly provided kindly supplied by INOVA (Bettlach, Switzerland). All tests were carried in duplicate [22]. aCL IgG and IgM cut-offs were 14 and 19 IU/ml, respectively; antiβ2GP1 IgG and IgM cut-offs were 17 and 12 IU/ml, respectively [21]; for both aPS-PT IgG and IgM the cut-off was 30 IU/ml, according to manufacturer recommendations. The aD1 cut-off was 14.4 IU/l, based on the 99th percentile obtained by testing the plasma of 195 healthy women as previously reported [23]. When biological duplicate samples were assessed, infraclass correlation coefficients were >90% for all tests (supplementary Table S1, available at Rheumatology online). Thrombotic events Venous and arterial thrombotic events were recorded at inclusion into SSCS and at subsequent visits. Consistent with the SLE Damage Index (SDI) [24] as used in the PROFILE study [25], arterial vascular events were documented if myocardial infarction, angina pectoris and/or a vascular procedure for myocardial infarction (coronary artery bypass graft), cerebral vascular accident and claudication lasting ⩾6 months and/or evidence of gangrene or significant tissue loss (loss of a digit or a limb) had occurred. Venous thrombotic events were deep venous thrombosis and or pulmonary embolism. Only the first incident event was taken into consideration. By definition, these events were recorded in SDI only if 6 months had elapsed since the diagnosis of SLE had been made. Statistical analysis We computed the proportions of patients who were positive for any of the nine aPL tests. We also defined categories of positivity as single, double or triple, according to the number of positive results for antibodies against cardiolipin (IgG or IgM), β2GP1 (IgG or IgM) and the LA (dRVVT or SCT). We examined the levels of non-criteria antibodies across categories of positivity. We obtained Pearson correlation coefficients between the nine test titres or results, which were logarithm-transformed to reduce the skewness of the distributions. To explore the pattern of associations between the tests, we conducted an exploratory factor analysis on the logarithm-transformed variables, followed by varimax rotation [26]. We examined the sampling adequacy of the exploratory factor analysis using the Kaiser-Meyer-Olkin statistic (desirable value >0.8), retained factors whose eigenvalue was >1, and obtained the proportion of total variance explained by the retained factors. To explore the clinical relevance of the retained factors, we examined their associations with the presence of vascular events (venous, arterial or both). To examine the reliability of the laboratory procedures, we obtained intraclass correlation coefficients on the titres obtained for doubly processed samples; this was done for only a subset of the samples [27]. To examine the stability of aPL over time, we computed kappa statistics (for results defined as positive or negative) and intraclass correlation coefficients (for continuous titres) for pairs of assessments obtained over time in the same patient, usually distant by several months. All tests were two sided and assessed at the 5% significance level. Statistical analyses were performed using SPSS version 22 (IBM Corp, Armonk, NY). Results Characteristic of the study population In this study, we included 178 individuals of whom 156 (87.6%) were women. The SLE clinical characteristics of the individuals included are summarized in Table 1. The mean (s.d.) age at SLE diagnosis was 34.7 years (15.2) and 44.1 years (14.6) at inclusion in the cohort. Some 42.5% of the samples were positive for at least one classical aPL, 20.5% showed double and 12.6% triple positivity (Table 2). Of interest, six samples positive for aPS-PT IgG and 12 samples positive for aPS-PT IgM were negative for criteria aPL. Indeed, the frequency of single-positive aPS-PT IgM and aPS-PT IgG was higher than the frequency of single-positive samples for the other assessed specificities, being >7.4% for aPS-PT and <3.7% for the others. Concerning specifically LA and aPS-PT IgG, 19 samples were positive for both, 10 positive for aPS-PT IgG alone and 21 for LA alone. All but one aD1 IgG (15%) positive sera were also positive for antiβ2GP1 IgG. No statistically significant differences in the ACR criteria, disease activity as captured by (Safety of Estrogens in Lupus National Assessment, SELENA-SLEDAI) and Physician Global Assessment as well as damage as captured by SLICC-SDI were observed between positive and negative sera for criteria aPL (data not shown). In addition, no statistically significant differences were observed in the frequency of positive samples, for all nine aPL tests performed, whether or not the patients were under CS treatment or whether disease activity captured by SELENA-SLEDAI was 0–6 or >6 (data not shown). Table 1 Baseline characteristics of the 178 patients with SLE   N (%)  Female sex  156 (87.6)  Caucasiana  139 (78.1)  Age at SLE diagnosis, mean (s.d.), years  34.7 (15.2)      Range  11.9–73.4  Age at first assessment, mean (s.d.), years  44.1 (14.6)      Range  16.6–85.3  ACR criteria at inclusion        Malar rash  76 (42.7)      Discoid rash  34 (19.2)      Photosensitivity  83 (46.6)      Nasopharyngeal ulcers  42 (23.6)      Arthritis  146 (82.0)      Pleuritis  43 (24.2)      Pericarditis  41 (23.0)      Renal disorder  81 (45.5)      Seizures  14 (7.9)      Psychosis  10 (5.6)      Haematologic disorder  112 (62.9)      Anti-Sm antibody positive  40 (22.7)      Anti-dsDNA antibodies positive  121 (68.4)      aPL positiveb  85 (48.3)      ANA positive  174 (98.3)  Disease activity at first assessment        PGA, mean (s.d.)  0.61 (0.78)      SELENA-SLEDAI, mean (s.d.)  6.56 (7.78)      SDI, median (IQR)  0 (0–2)  Treatment at first assessment        Systemic CS  96 (53.9)      Antimalarials  102 (57.6)      Immunosuppressant agents  65 (36.5)      Anticoagulants/low-dose aspirin  59/172 (34.3)    N (%)  Female sex  156 (87.6)  Caucasiana  139 (78.1)  Age at SLE diagnosis, mean (s.d.), years  34.7 (15.2)      Range  11.9–73.4  Age at first assessment, mean (s.d.), years  44.1 (14.6)      Range  16.6–85.3  ACR criteria at inclusion        Malar rash  76 (42.7)      Discoid rash  34 (19.2)      Photosensitivity  83 (46.6)      Nasopharyngeal ulcers  42 (23.6)      Arthritis  146 (82.0)      Pleuritis  43 (24.2)      Pericarditis  41 (23.0)      Renal disorder  81 (45.5)      Seizures  14 (7.9)      Psychosis  10 (5.6)      Haematologic disorder  112 (62.9)      Anti-Sm antibody positive  40 (22.7)      Anti-dsDNA antibodies positive  121 (68.4)      aPL positiveb  85 (48.3)      ANA positive  174 (98.3)  Disease activity at first assessment        PGA, mean (s.d.)  0.61 (0.78)      SELENA-SLEDAI, mean (s.d.)  6.56 (7.78)      SDI, median (IQR)  0 (0–2)  Treatment at first assessment        Systemic CS  96 (53.9)      Antimalarials  102 (57.6)      Immunosuppressant agents  65 (36.5)      Anticoagulants/low-dose aspirin  59/172 (34.3)  a Additional ethnic groups were African 8.4%, Asian 9.0%, Others 4%, unknown 0.6%. b Positivity defined by historical chart data. PGA: Physician Global Assessment; SDI: SLICC/ACR SLE Damage Index; IQR: interquartile range. Table 1 Baseline characteristics of the 178 patients with SLE   N (%)  Female sex  156 (87.6)  Caucasiana  139 (78.1)  Age at SLE diagnosis, mean (s.d.), years  34.7 (15.2)      Range  11.9–73.4  Age at first assessment, mean (s.d.), years  44.1 (14.6)      Range  16.6–85.3  ACR criteria at inclusion        Malar rash  76 (42.7)      Discoid rash  34 (19.2)      Photosensitivity  83 (46.6)      Nasopharyngeal ulcers  42 (23.6)      Arthritis  146 (82.0)      Pleuritis  43 (24.2)      Pericarditis  41 (23.0)      Renal disorder  81 (45.5)      Seizures  14 (7.9)      Psychosis  10 (5.6)      Haematologic disorder  112 (62.9)      Anti-Sm antibody positive  40 (22.7)      Anti-dsDNA antibodies positive  121 (68.4)      aPL positiveb  85 (48.3)      ANA positive  174 (98.3)  Disease activity at first assessment        PGA, mean (s.d.)  0.61 (0.78)      SELENA-SLEDAI, mean (s.d.)  6.56 (7.78)      SDI, median (IQR)  0 (0–2)  Treatment at first assessment        Systemic CS  96 (53.9)      Antimalarials  102 (57.6)      Immunosuppressant agents  65 (36.5)      Anticoagulants/low-dose aspirin  59/172 (34.3)    N (%)  Female sex  156 (87.6)  Caucasiana  139 (78.1)  Age at SLE diagnosis, mean (s.d.), years  34.7 (15.2)      Range  11.9–73.4  Age at first assessment, mean (s.d.), years  44.1 (14.6)      Range  16.6–85.3  ACR criteria at inclusion        Malar rash  76 (42.7)      Discoid rash  34 (19.2)      Photosensitivity  83 (46.6)      Nasopharyngeal ulcers  42 (23.6)      Arthritis  146 (82.0)      Pleuritis  43 (24.2)      Pericarditis  41 (23.0)      Renal disorder  81 (45.5)      Seizures  14 (7.9)      Psychosis  10 (5.6)      Haematologic disorder  112 (62.9)      Anti-Sm antibody positive  40 (22.7)      Anti-dsDNA antibodies positive  121 (68.4)      aPL positiveb  85 (48.3)      ANA positive  174 (98.3)  Disease activity at first assessment        PGA, mean (s.d.)  0.61 (0.78)      SELENA-SLEDAI, mean (s.d.)  6.56 (7.78)      SDI, median (IQR)  0 (0–2)  Treatment at first assessment        Systemic CS  96 (53.9)      Antimalarials  102 (57.6)      Immunosuppressant agents  65 (36.5)      Anticoagulants/low-dose aspirin  59/172 (34.3)  a Additional ethnic groups were African 8.4%, Asian 9.0%, Others 4%, unknown 0.6%. b Positivity defined by historical chart data. PGA: Physician Global Assessment; SDI: SLICC/ACR SLE Damage Index; IQR: interquartile range. Table 2 Prevalence of aPL in 178 SLE patients at entry   N (%)  Criteria aPL        LA (SCT)  35/175a (20.0)      LA (dRVVT)  40/172 (23.3)      aCL IgM  13/176 (7.4)      aCL IgG  37/176 (21.0)      Antiβ2GP1 IgM  12/176 (6.8)      Antiβ2GP1 IgG  40/176 (22.7)      Single positive (including double and triple positive)  75/169 (44.4)      Double positive (including triple positive)  35/171 (20.5)      Triple positive  22/175 (12.6)      Negative for all  94/169 (55.6)  Non-criteria aPL        aPS-PT IgM  47/178 (26.4)      aPS-PT IgG  29/178 (16.3)      aD1 IgG  26/164 (15.9)    N (%)  Criteria aPL        LA (SCT)  35/175a (20.0)      LA (dRVVT)  40/172 (23.3)      aCL IgM  13/176 (7.4)      aCL IgG  37/176 (21.0)      Antiβ2GP1 IgM  12/176 (6.8)      Antiβ2GP1 IgG  40/176 (22.7)      Single positive (including double and triple positive)  75/169 (44.4)      Double positive (including triple positive)  35/171 (20.5)      Triple positive  22/175 (12.6)      Negative for all  94/169 (55.6)  Non-criteria aPL        aPS-PT IgM  47/178 (26.4)      aPS-PT IgG  29/178 (16.3)      aD1 IgG  26/164 (15.9)  a Number of available determinations. Antiβ2GP1: anti-β2-glycoprotein 1; aD1: anti-domain 1 of β2GP1; dRVVT: dilute Russell’s viper venom test; SCT: silica clotting time. Table 2 Prevalence of aPL in 178 SLE patients at entry   N (%)  Criteria aPL        LA (SCT)  35/175a (20.0)      LA (dRVVT)  40/172 (23.3)      aCL IgM  13/176 (7.4)      aCL IgG  37/176 (21.0)      Antiβ2GP1 IgM  12/176 (6.8)      Antiβ2GP1 IgG  40/176 (22.7)      Single positive (including double and triple positive)  75/169 (44.4)      Double positive (including triple positive)  35/171 (20.5)      Triple positive  22/175 (12.6)      Negative for all  94/169 (55.6)  Non-criteria aPL        aPS-PT IgM  47/178 (26.4)      aPS-PT IgG  29/178 (16.3)      aD1 IgG  26/164 (15.9)    N (%)  Criteria aPL        LA (SCT)  35/175a (20.0)      LA (dRVVT)  40/172 (23.3)      aCL IgM  13/176 (7.4)      aCL IgG  37/176 (21.0)      Antiβ2GP1 IgM  12/176 (6.8)      Antiβ2GP1 IgG  40/176 (22.7)      Single positive (including double and triple positive)  75/169 (44.4)      Double positive (including triple positive)  35/171 (20.5)      Triple positive  22/175 (12.6)      Negative for all  94/169 (55.6)  Non-criteria aPL        aPS-PT IgM  47/178 (26.4)      aPS-PT IgG  29/178 (16.3)      aD1 IgG  26/164 (15.9)  a Number of available determinations. Antiβ2GP1: anti-β2-glycoprotein 1; aD1: anti-domain 1 of β2GP1; dRVVT: dilute Russell’s viper venom test; SCT: silica clotting time. Correlation between criteria and non-criteria aPL antibodies titres We assessed whether there was a relationship in the titres of aPL tests obtained with the nine assays performed. While they all were positively correlated (P < 0.001), the level of correlation was very high only for a few of them (Table 3). Thus, the strongest correlations with a Pearson correlation coefficient (r) value of about 0.8 were observed for aD1 IgG with antiβ2GP1 IgG and aCL IgG (denoted with the superscript b in Table 3). Similarly, aCL IgM and antiβ2GP1 IgM was also strongly correlated. Lower levels of correlation (Pearson’s r between 0.48 and 0.59) were found between dRVVT and IgG titres of aCL, antiβ2GP1, aD1, and aPS-PT IgG and IgM aPS-PT IgM (denoted with the superscript c in Table 3). The correlations were weakest between aPS-PT IgG/IgM, aCL IgG/IgM and antiβ2GP1 IgG, as well as aD1 IgG. Table 3 Pairwise Pearson correlation coefficients among aPL titres in 178 patients with SLE Test  SCT  aCL IgMa  aCL IgGa  Antiβ2GP1 IgMa  Antiβ2GP1 IgGa  aPS-PT IgMa  aPS-PT IgGa  aD1 IgGa  dRVVT  0.77b  0.37  0.49c  0.43  0.59c  0.43  0.50c  0.56c  SCT    0.25  0.39  0.27  0.47c  0.29  0.50c  0.46c  aCL IgMa      0.38  0.86b  0.41  0.51c  0.19  0.37  aCL IgGa        0.43  0.81b  0.30  0.39  0.79b  Antiβ2GP1 IgMa          0.49  0.52c  0.26  0.41  Antiβ2GP1 IgGa            0.35  0.49c  0.87b  aPS-PT IgMa              0.36  0.32  aPS-PT IgGa                0.46c  Test  SCT  aCL IgMa  aCL IgGa  Antiβ2GP1 IgMa  Antiβ2GP1 IgGa  aPS-PT IgMa  aPS-PT IgGa  aD1 IgGa  dRVVT  0.77b  0.37  0.49c  0.43  0.59c  0.43  0.50c  0.56c  SCT    0.25  0.39  0.27  0.47c  0.29  0.50c  0.46c  aCL IgMa      0.38  0.86b  0.41  0.51c  0.19  0.37  aCL IgGa        0.43  0.81b  0.30  0.39  0.79b  Antiβ2GP1 IgMa          0.49  0.52c  0.26  0.41  Antiβ2GP1 IgGa            0.35  0.49c  0.87b  aPS-PT IgMa              0.36  0.32  aPS-PT IgGa                0.46c  a Logarithm transformation. b Pearson correlation coefficient above 0.75. c Pearson correlation coefficient around 0.5. Antiβ2GP1: anti-β2-glycoprotein 1; aD1: anti-domain 1 of β2GP1; dRVVT: dilute Russell’s viper venom test; SCT: silica clotting time. All P < 0.001. Numbers of pairs vary between 251 and 283. Table 3 Pairwise Pearson correlation coefficients among aPL titres in 178 patients with SLE Test  SCT  aCL IgMa  aCL IgGa  Antiβ2GP1 IgMa  Antiβ2GP1 IgGa  aPS-PT IgMa  aPS-PT IgGa  aD1 IgGa  dRVVT  0.77b  0.37  0.49c  0.43  0.59c  0.43  0.50c  0.56c  SCT    0.25  0.39  0.27  0.47c  0.29  0.50c  0.46c  aCL IgMa      0.38  0.86b  0.41  0.51c  0.19  0.37  aCL IgGa        0.43  0.81b  0.30  0.39  0.79b  Antiβ2GP1 IgMa          0.49  0.52c  0.26  0.41  Antiβ2GP1 IgGa            0.35  0.49c  0.87b  aPS-PT IgMa              0.36  0.32  aPS-PT IgGa                0.46c  Test  SCT  aCL IgMa  aCL IgGa  Antiβ2GP1 IgMa  Antiβ2GP1 IgGa  aPS-PT IgMa  aPS-PT IgGa  aD1 IgGa  dRVVT  0.77b  0.37  0.49c  0.43  0.59c  0.43  0.50c  0.56c  SCT    0.25  0.39  0.27  0.47c  0.29  0.50c  0.46c  aCL IgMa      0.38  0.86b  0.41  0.51c  0.19  0.37  aCL IgGa        0.43  0.81b  0.30  0.39  0.79b  Antiβ2GP1 IgMa          0.49  0.52c  0.26  0.41  Antiβ2GP1 IgGa            0.35  0.49c  0.87b  aPS-PT IgMa              0.36  0.32  aPS-PT IgGa                0.46c  a Logarithm transformation. b Pearson correlation coefficient above 0.75. c Pearson correlation coefficient around 0.5. Antiβ2GP1: anti-β2-glycoprotein 1; aD1: anti-domain 1 of β2GP1; dRVVT: dilute Russell’s viper venom test; SCT: silica clotting time. All P < 0.001. Numbers of pairs vary between 251 and 283. Given the substantial heterogeneity in aPL titres in SLE patients, we conducted an exploratory factor analysis to identify patterns of antibody responses. This analysis suggested three factors with an eigenvalue >1, which together explained 79.5% of total variance (Table 4). The measure of sampling adequacy was 0.82, above the threshold of 0.8. Factor 1 grouped aCL IgG, antiβ2GP1 IgG and aD1 IgG. Factor 2 grouped dRVVT, SCT and aPS-PT IgG. Factor 3 grouped aCL IgM, antiβ2GP1 IgM and aPS-PT IgM. Thus, the nine tests performed provide results that can be grouped into three homogeneous ensembles. Interestingly, these groups respect isotypes differences and only aPS-PT IgG significantly aggregates with LA. Table 4 Factor analysis of nine aPL tests identifies three groups of antibodies Test  Factor     1  2  3  dRVVTa  0.30  0.77  0.29  SCTa  0.18  0.83  0.12  aCL IgMa  0.22  0.06  0.91  aCL IgGa  0.88  0.21  0.20  Antiβ2GP1 IgMa  0.29  0.10  0.88  Antiβ2GP1 IgGa  0.87  0.32  0.23  aPS-PT IgMa  0.04  0.38  0.69  aPS-PT IgGa  0.29  0.72  0.07  aD1 IgGa  0.87  0.29  0.18  Test  Factor     1  2  3  dRVVTa  0.30  0.77  0.29  SCTa  0.18  0.83  0.12  aCL IgMa  0.22  0.06  0.91  aCL IgGa  0.88  0.21  0.20  Antiβ2GP1 IgMa  0.29  0.10  0.88  Antiβ2GP1 IgGa  0.87  0.32  0.23  aPS-PT IgMa  0.04  0.38  0.69  aPS-PT IgGa  0.29  0.72  0.07  aD1 IgGa  0.87  0.29  0.18  Factor loadings after varimax rotation are shown. Primary factor loadings are in bold. Data from 162 patients with complete data. a Logarithm transformation. Antiβ2GP1: anti-β2-glycoprotein 1; aD1: anti-domain 1 of β2GP1; dRVVT: dilute Russell’s viper venom test; SCT: silica clotting time. Table 4 Factor analysis of nine aPL tests identifies three groups of antibodies Test  Factor     1  2  3  dRVVTa  0.30  0.77  0.29  SCTa  0.18  0.83  0.12  aCL IgMa  0.22  0.06  0.91  aCL IgGa  0.88  0.21  0.20  Antiβ2GP1 IgMa  0.29  0.10  0.88  Antiβ2GP1 IgGa  0.87  0.32  0.23  aPS-PT IgMa  0.04  0.38  0.69  aPS-PT IgGa  0.29  0.72  0.07  aD1 IgGa  0.87  0.29  0.18  Test  Factor     1  2  3  dRVVTa  0.30  0.77  0.29  SCTa  0.18  0.83  0.12  aCL IgMa  0.22  0.06  0.91  aCL IgGa  0.88  0.21  0.20  Antiβ2GP1 IgMa  0.29  0.10  0.88  Antiβ2GP1 IgGa  0.87  0.32  0.23  aPS-PT IgMa  0.04  0.38  0.69  aPS-PT IgGa  0.29  0.72  0.07  aD1 IgGa  0.87  0.29  0.18  Factor loadings after varimax rotation are shown. Primary factor loadings are in bold. Data from 162 patients with complete data. a Logarithm transformation. Antiβ2GP1: anti-β2-glycoprotein 1; aD1: anti-domain 1 of β2GP1; dRVVT: dilute Russell’s viper venom test; SCT: silica clotting time. To further address the correlation and concomitant positivity between criteria and non-criteria antibodies we assessed the titre of aD1 IgG and aPS-PT IgG and IgM in samples with no, single, double or triple positivity according to criteria assays. As expected, the titre of the three non-criteria antibodies increased with the increase of the number of positive criteria assays. However, we observed an exquisite enrichment for high aD1 IgG titres in triple-positive samples, increasing by more than two orders of magnitude from no positive to triple-positive criteria sera (supplementary Table S2, available at Rheumatology online). For comparison, in the same groups aPS-PT titres increased by 5-fold only for both IgG and IgM (supplementary Table S2, available at Rheumatology online). Thus, in SLE sera there is a preferential enrichment for high titre aD1 IgG in triple-positive samples. Persistence of aPL The next question we asked was whether the presence or absence of individual aCL antibody specificities was sustained over time by assessing sera taken on average 1 year apart in 105 individuals. The clinical characteristics of these individuals did not significantly differ from the initial population (supplementary Table S3, available at Rheumatology online). Overall, positivity or negativity was sustained over time (supplementary Table S4, available at Rheumatology online). Kappa statistics revealed that SCT was the least persistent. In addition, aCL IgG as well as aPS-PT IgG and IgM had a kappa statistic ranging from 0.69 to 0.74, thus indicating some variation over time. However, the intra-class correlation coefficients on continuous measurements were very high for all tests (except for SCT), ranging from 0.83 to 0.99 (supplementary Table S4, available at Rheumatology online), thus indicating that results located near the cut-off values could explain most of the variations observed for dichotomous results. Association of aPL tests with thrombotic events We next addressed the question of whether aPL was associated with thrombotic events in our SLE cohort. The overall frequency of thrombotic events was low. Twenty-two venous, 20 arterial and 37 either venous or arterial events were recorded. Among the nine distinct assays performed to detect aPL antibodies, only aPS-PT IgG titres were associated significantly with venous and composite venous or arterial events (supplementary Table S5, available at Rheumatology online). To substantiate these findings, we used the presence or absence of aPS-PT antibodies as a dichotomous variable and assessed their association with thrombotic events. aPS-PT IgG but not aPS-PT IgM was strongly associated (P = 0.012) with venous thrombotic events and to a lesser extent to the composite arterial and venous vascular events (P = 0.077) (Table 5). In addition, when using the three factors obtained by exploratory factor analysis only the factor that grouped LA and anti-PS-PT IgG was positively associated with thrombotic events, and the association was statistically significant for the composite of venous and arterial events (Table 6). The factor grouping aCL IgG, antiβ2GP1 IgG and aD1 IgG had much weaker association, while the factor grouping aCL IgM, antiβ2GP1 IgM and aPS-PT IgM had a negative association with vascular events (factor scores were lower in patients with events); neither one of the latter was statistically significant. Of interest, when triple (data not shown) or quadruple (supplementary Table S6, available at Rheumatology online) positive samples were taken into consideration, the association with vascular thrombotic events was statistically significant only for IgG responses. Thus, the non-criteria aPS-PT IgG—by itself and in conjunction with LA—appears to be associated with thrombotic events in Swiss SLE patients. Table 5 Association of aPS-PT complex IgG with thrombotic events in SLE   Venous vascular events (n = 20)   Arterial vascular events (n = 17)   Venous or arterial vascular events (n = 34)   Test  Yes  No  P-value  Yes  No  P-value  Yes  No  P-value  aPS-PT IgG (pos = 29)  8  21  0.012  2  27  0.537  10  19  0.077  aPS-PT IgM (pos = 47)  3  43  0.445  6  41  0.440  10  37  1.000  aPS-PT IgG or aPS-PT IgM (pos = 59)  9  50  0.470  8  51  0.615  17  42  0.078    Venous vascular events (n = 20)   Arterial vascular events (n = 17)   Venous or arterial vascular events (n = 34)   Test  Yes  No  P-value  Yes  No  P-value  Yes  No  P-value  aPS-PT IgG (pos = 29)  8  21  0.012  2  27  0.537  10  19  0.077  aPS-PT IgM (pos = 47)  3  43  0.445  6  41  0.440  10  37  1.000  aPS-PT IgG or aPS-PT IgM (pos = 59)  9  50  0.470  8  51  0.615  17  42  0.078  Number of aPS-PT IgG and/or IgM positive samples with thrombotic events in SLE. P-value determined by the Fisher’s exact test. pos: positive. Table 5 Association of aPS-PT complex IgG with thrombotic events in SLE   Venous vascular events (n = 20)   Arterial vascular events (n = 17)   Venous or arterial vascular events (n = 34)   Test  Yes  No  P-value  Yes  No  P-value  Yes  No  P-value  aPS-PT IgG (pos = 29)  8  21  0.012  2  27  0.537  10  19  0.077  aPS-PT IgM (pos = 47)  3  43  0.445  6  41  0.440  10  37  1.000  aPS-PT IgG or aPS-PT IgM (pos = 59)  9  50  0.470  8  51  0.615  17  42  0.078    Venous vascular events (n = 20)   Arterial vascular events (n = 17)   Venous or arterial vascular events (n = 34)   Test  Yes  No  P-value  Yes  No  P-value  Yes  No  P-value  aPS-PT IgG (pos = 29)  8  21  0.012  2  27  0.537  10  19  0.077  aPS-PT IgM (pos = 47)  3  43  0.445  6  41  0.440  10  37  1.000  aPS-PT IgG or aPS-PT IgM (pos = 59)  9  50  0.470  8  51  0.615  17  42  0.078  Number of aPS-PT IgG and/or IgM positive samples with thrombotic events in SLE. P-value determined by the Fisher’s exact test. pos: positive. Table 6 The combination of LA with aPS-PT IgG is associated with thrombotic events in SLE   Venous vascular events   Arterial vascular events   Venous or arterial vascular events   Test  Yes (N = 20)  No (N = 142)  P-value  Yes (N = 17)  No (N = 145)  P-value  Yes (N = 34)  No (N = 128)  P-value  Factor 1: aCL IgG, antiβ2GP1 IgG and aD1 IgG  0.23  −0.03  0.27  −0.06  0.01  0.77  0.13  −0.04  0.38  Factor 2: LA and aPS-PT IgG  0.36  −0.05  0.084  0.34  −0.04  0.14  0.35  −0.09  0.02  Factor 3: aCL IgM, antiβ2GP1 IgM and aPS-PT IgM  −0.22  0.03  0.30  −0.18  0.02  0.42  −0.17  0.05  0.25    Venous vascular events   Arterial vascular events   Venous or arterial vascular events   Test  Yes (N = 20)  No (N = 142)  P-value  Yes (N = 17)  No (N = 145)  P-value  Yes (N = 34)  No (N = 128)  P-value  Factor 1: aCL IgG, antiβ2GP1 IgG and aD1 IgG  0.23  −0.03  0.27  −0.06  0.01  0.77  0.13  −0.04  0.38  Factor 2: LA and aPS-PT IgG  0.36  −0.05  0.084  0.34  −0.04  0.14  0.35  −0.09  0.02  Factor 3: aCL IgM, antiβ2GP1 IgM and aPS-PT IgM  −0.22  0.03  0.30  −0.18  0.02  0.42  −0.17  0.05  0.25  Exploratory factor analysis performed according to Bartko [27]. Three factors with an eigenvalue >1, explaining 79.5% of total variance, are reported. Mean values of three factors combining aPL in the presence and absence of thrombotic events, in 162 patients (16 had missing data). All factors had a mean of 0 and s.d. of 1 in the whole sample. The measure of sampling adequacy was 0.82, above the threshold of 0.8. Antiβ2GP1: anti-β2-glycoprotein 1; aD1: anti-domain 1 of β2GP1. Table 6 The combination of LA with aPS-PT IgG is associated with thrombotic events in SLE   Venous vascular events   Arterial vascular events   Venous or arterial vascular events   Test  Yes (N = 20)  No (N = 142)  P-value  Yes (N = 17)  No (N = 145)  P-value  Yes (N = 34)  No (N = 128)  P-value  Factor 1: aCL IgG, antiβ2GP1 IgG and aD1 IgG  0.23  −0.03  0.27  −0.06  0.01  0.77  0.13  −0.04  0.38  Factor 2: LA and aPS-PT IgG  0.36  −0.05  0.084  0.34  −0.04  0.14  0.35  −0.09  0.02  Factor 3: aCL IgM, antiβ2GP1 IgM and aPS-PT IgM  −0.22  0.03  0.30  −0.18  0.02  0.42  −0.17  0.05  0.25    Venous vascular events   Arterial vascular events   Venous or arterial vascular events   Test  Yes (N = 20)  No (N = 142)  P-value  Yes (N = 17)  No (N = 145)  P-value  Yes (N = 34)  No (N = 128)  P-value  Factor 1: aCL IgG, antiβ2GP1 IgG and aD1 IgG  0.23  −0.03  0.27  −0.06  0.01  0.77  0.13  −0.04  0.38  Factor 2: LA and aPS-PT IgG  0.36  −0.05  0.084  0.34  −0.04  0.14  0.35  −0.09  0.02  Factor 3: aCL IgM, antiβ2GP1 IgM and aPS-PT IgM  −0.22  0.03  0.30  −0.18  0.02  0.42  −0.17  0.05  0.25  Exploratory factor analysis performed according to Bartko [27]. Three factors with an eigenvalue >1, explaining 79.5% of total variance, are reported. Mean values of three factors combining aPL in the presence and absence of thrombotic events, in 162 patients (16 had missing data). All factors had a mean of 0 and s.d. of 1 in the whole sample. The measure of sampling adequacy was 0.82, above the threshold of 0.8. Antiβ2GP1: anti-β2-glycoprotein 1; aD1: anti-domain 1 of β2GP1. Discussion Taking advantage of the Swiss SLE cohort [19] as well as of the availability of novel commercial kits to assess the presence of aPS-PT IgG and IgM as well as aD1 IgG, our study examined the analytical performance of nine tests aiming at detecting aPL antibodies and their clinical associations. Our results show that: the tests employed here provide reliable results; aPL tend to persist with time; the titres of criteria and non-criteria aPL all correlate, but vary considerably in the strength of correlations; exploratory factor analysis identifies three groups of specificities segregating aPL isotypes and grouping LA only with aPS-PT IgG; high titre aD1 IgG are highly enriched in triple-positive samples; and aPS-PT IgG—by itself and in combination with LA—stands out in patients with an history of thrombotic events. The reproducibility of results is, of course, essential to provide reliable conclusions. While all samples were tested with technical duplicates, we tested also biological duplicates in a random subgroup of about 70 patients (supplementary Table S1, available at Rheumatology online). The results obtained were robust and satisfied the requirements of consensus experts [12]. This allowed the study of the persistence of aPL in time. As expected, and in agreement with published data [28], most negative samples remained negative 1 year apart, while somehow, less frequently positive samples remained positive. Low titre antibody positivity explained most of this variability, which was independent of treatment (data not shown). Most importantly, techniques of dosage and cut-off values used here were validated in previous studies by us and similar to those used by others [20, 21, 23, 29–32]. The investigation of concordance and discordance among the aPL assays provided interesting data that complement previous studies. Overall, the correlation coefficients between the tests expressed as continuous variables were statistically highly significant, which is in line with previous reports by us and others [14, 17, 20, 23, 32, 33]. However, the correlation coefficients varied substantially among pairs of tests. The exploratory factor analysis identified the underlying three factors that associated the tests in three distinct groups: one grouping aCL IgG, antiβ2GP1 IgG and aD1 IgG, a second grouping LA and aPS-PT IgG and the third grouping aCL IgM, antiβ2GP1 IgM and aPS-PT IgM. The association of specificities we found in the first group has already been documented in the literature [20, 32, 33], and indeed D1 has been previously shown to be the immune-dominant epitope of β2GP1 IgG. Previous studies, however, focused on the performance of the tests in an APS patient population with a high prevalence of thrombotic events [34], which is not the case in our SLE cohort. Our factor three grouped IgM responses, with very high correlation coefficients, much higher than those correlating IgM with IgG specific for the same antigen. This indicates that some SLE patients are developing IgM responses against distinct phospholipid (PL) specificities that are sustained in time and do not tend to switch to IgG. Of interest, IgM specificities had similar correlation coefficients with dRVVT to those of their IgG counterpart. In this perspective, our results are in agreement with previous data indicating association of aCL IgM, antiβ2GP1 IgM and aPS-PT IgM with LA [17] and support the contention that IgM may also contribute to LA. Finally, our factor two grouped LA (both dRVTT and/or SCT) and aPS-PT IgG. Thus, aPS-PT IgG in our study appears to have distinct characteristics compared with the other IgG aPL assays. This, may be reflected in biological properties which may then have clinical impact as discussed here below. Several studies have reported that patients with triple-positive aPL carry a particularly high risk of thrombotic events [8, 32, 35–37]. Furthermore, aD1 IgG was shown to be elevated in triple-positive sera [38, 39]. Our study confirms that aD1 IgG are particularly high, in triple-positive sera, possibly more than aPS-PT IgG (and aPS-PT IgM) in SLE patients. In contrast, we did not confirm the previous observation that among antiβ2GP1-positive patients aD1 positivity implies an increased risk of thrombotic events [14]. This is consistent with other reports in SLE [17]. Two non-mutually exclusive hypotheses may explain this discrepancy. First, the epitopes within aD1 recognized by SLE IgG may differ from those recognized by primary APS. This may affect their pro-thrombotic activity. Second, thrombotic events in SLE may be due, at least in part, to factors independent form aPL, which contribution may lessen the importance of aD1 IgG. In our cohort, the presence of aPS-PT IgG was particularly associated with venous thrombotic events, and interestingly the combination of aPS-PT IgG with LA—identified in the exploratory factor analysis—was associated in a statistically significant manner when venous and arterial events were grouped. This confirms data by others in primary APS [18] and in SLE [17]. However, many other studies identified also aCL IgG/IgM and antiβ2GP1 IgG/IgM to be associated with thrombosis [40–42]. These differences may be due to the clinical characteristic of the SLE patients enrolled in our cohort and in particular to the relatively low incidence of thrombotic events (21%) when compared with other study populations where the prevalence of such events ranged from 25 to 50% [17, 42, 43], thus reducing the power of the statistical analysis we conducted. This may further explain why in our cohort aD1 IgG was not associated with thrombotic events in contrasts with the findings of others in APS [37]. A further point of interest is that, due to their strong association with LA and with vascular events, the assessment of of aPS-PT IgG in substitution of LA may particularly useful in anticoagulated patients in which the determination of LA is technically difficult. It has to be stressed that our study population was followed in tertiary centres across Switzerland, patients attended by internal medicine, nephrology, clinical immunology and rheumatology specialists. This may represent a blend of patients that may differ from those attended to in dedicated, single centres. A weakness of our study is that ours is not an inception cohort, and most thrombotic events were recorded by review of the clinical charts. Others have found that thrombotic events accumulate with time, particularly in triple-positive individuals [35]. It is possible that the prospective follow-up of our cohort will reveal other associations. As an additional limitation, we did not include in our study aPL IgA or other non-criteria PL specificities beside PS-PT and D1, thus limiting the potential for detecting important clinical associations. However, the analytical performance of our study met the highest standards and results were highly reproducible. In conclusion, our study stresses the importance of aPS-PT IgG in identifying patients at risk for thrombosis in SLE and suggests the usefulness of adding this specificity when searching for aPL antibodies. Prospective studies are needed to support this contention. Acknowledgements We thank D. Wahl (Nancy, France), J.-C. Gris (Nîmes, France) and K.M. Devreese (Gent, Belgium) for their critical reading of the manuscript. We acknowledge INOVA (Switzerland) for kindly supplying reagents for aPL antibody determination (QUANTA Lite and QUANTA Flash) free of charge. Authors’ contribtions: T.M., C.R., T.P., P.d.M. and C.C. analysed the data and drafted the manuscript; M.T. and U.H.-D. provided critical reading; C.R., M.T., U.H.-D. and C.C. acquired the clinical data and collected samples; T.M. and P.d.M. acquired the laboratory data; and T.P. performed the statistical analysis. Funding: This work was supported in part by a Gebert Ruf unrestricted grant to SSCS, and the ISTH2007 presidential fund (to C.C.). Disclosure statement: The authors have declared no conflicts of interest. Supplementary data Supplementary data are available at Rheumatology online. References 1 Liu Z, Davidson A. Taming lupus-a new understanding of pathogenesis is leading to clinical advances. Nat Med  2012; 18: 871– 82. Google Scholar CrossRef Search ADS PubMed  2 D’Cruz DP, Khamashta MA, Hughes GR. Systemic lupus erythematosus. Lancet  2007; 369: 587– 96. Google Scholar CrossRef Search ADS PubMed  3 Tan EM, Cohen AS, Fries JF et al.   The 1982 revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum  1982; 25: 1271– 7. Google Scholar CrossRef Search ADS PubMed  4 Hochberg MC. Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus. Arthritis Rheum  1997; 40: 1725. Google Scholar CrossRef Search ADS PubMed  5 Meroni PL, Borghi MO, Raschi E, Tedesco F. Pathogenesis of antiphospholipid syndrome: understanding the antibodies. Nat Rev Rheumatol  2011; 7: 330– 9. Google Scholar CrossRef Search ADS PubMed  6 Giannakopoulos B, Krilis SA. The pathogenesis of the antiphospholipid syndrome. N Engl J Med  2013; 368: 1033– 44. Google Scholar CrossRef Search ADS PubMed  7 Habe K, Wada H, Matsumoto T et al.   Presence of antiphospholipid antibodies as a risk factor for thrombotic events in patients with connective tissue diseases and idiopathic thrombocytopenic purpura. Intern Med  2016; 55: 589– 95. Google Scholar CrossRef Search ADS PubMed  8 Yelnik CM, Laskin CA, Porter TF et al.   Lupus anticoagulant is the main predictor of adverse pregnancy outcomes in aPL-positive patients: validation of PROMISSE study results. Lupus Sci Med  2016; 3 e000131. Google Scholar CrossRef Search ADS PubMed  9 Bertolaccini ML, Sanna G. Recent advances in understanding antiphospholipid syndrome. F1000Res  2016; 5: 2908. Google Scholar CrossRef Search ADS PubMed  10 Unlu O, Zuily S, Erkan D. The clinical significance of antiphospholipid antibodies in systemic lupus erythematosus. Eur J Rheumatol  2016; 3: 75– 84. Google Scholar CrossRef Search ADS PubMed  11 Pons-Estel GJ, Andreoli L, Scanzi F, Cervera R, Tincani A. The antiphospholipid syndrome in patients with systemic lupus erythematosus. J Autoimmun  2017; 76: 10– 20. Google Scholar CrossRef Search ADS PubMed  12 Miyakis S, Lockshin MD, Atsumi T et al.   International consensus statement on an update of the classification criteria for definite antiphospholipid syndrome (APS). J Thromb Haemost  2006; 4: 295– 306. Google Scholar CrossRef Search ADS PubMed  13 Meroni PL, Chighizola CB, Rovelli F, Gerosa M. Antiphospholipid syndrome in 2014: more clinical manifestations, novel pathogenic players and emerging biomarkers. Arthritis Res Ther  2014; 16: 209. Google Scholar CrossRef Search ADS PubMed  14 de Laat B, Pengo V, Pabinger I et al.   The association between circulating antibodies against domain I of beta2-glycoprotein I and thrombosis: an international multicenter study. J Thromb Haemost  2009; 7: 1767– 73. Google Scholar CrossRef Search ADS PubMed  15 Broen JC, Dieude P, Vonk MC et al.   Polymorphisms in the interleukin 4, interleukin 13, and corresponding receptor genes are not associated with systemic sclerosis and do not influence gene expression. J Rheumatol  2012; 39: 112– 8. Google Scholar CrossRef Search ADS PubMed  16 Mahler M, Norman GL, Meroni PL, Khamashta M. Autoantibodies to domain 1 of beta 2 glycoprotein 1: a promising candidate biomarker for risk management in antiphospholipid syndrome. Autoimmun Rev  2012; 12: 313– 7. Google Scholar CrossRef Search ADS PubMed  17 Akhter E, Shums Z, Norman GL et al.   Utility of antiphosphatidylserine/prothrombin and IgA antiphospholipid assays in systemic lupus erythematosus. J Rheumatol  2013; 40: 282– 6. Google Scholar CrossRef Search ADS PubMed  18 Sciascia S, Sanna G, Murru V et al.   Anti-prothrombin (aPT) and anti-phosphatidylserine/prothrombin (aPS-PT) antibodies and the risk of thrombosis in the antiphospholipid syndrome. A systematic review. J Thromb Haemost  2014; 111: 354– 64. Google Scholar CrossRef Search ADS   19 Ribi C, Trendelenburg M, Gayet-Ageron A et al.   The Swiss Systemic lupus erythematosus Cohort Study (SSCS) - cross-sectional analysis of clinical characteristics and treatments across different medical disciplines in Switzerland. Swiss Med Wkly  2014; 144: w13990. Google Scholar PubMed  20 Van Hoecke F, Persijn L, Decavele AS, Devreese K. Performance of two new, automated chemiluminescence assay panels for anticardiolipin and anti-beta2-glycoprotein I antibodies in the laboratory diagnosis of the antiphospholipid syndrome. Int J Lab Hematol  2012; 34: 630– 40. Google Scholar CrossRef Search ADS PubMed  21 Fontana P, Poncet A, Lindhoff-Last E, de Moerloose P, Devreese KM. Refinement of the cutoff values of the HemosIL AcuStar assay for the detection of anticardiolipin and anti-beta2 glycoprotein-1 antibodies. J Thromb Haemost  2014; 12: 2034– 7. Google Scholar CrossRef Search ADS PubMed  22 Devreese KM. Antiphospholipid antibody testing and standardization. Int J Lab Hematol  2014; 36: 352– 63. Google Scholar CrossRef Search ADS PubMed  23 Marchetti T, de Moerloose P, Gris JC. Antiphospholipid antibodies and the risk of severe and non-severe pre-eclampsia: the NOHA case-control study. J Thromb Haemost  2016; 14: 675– 84. Google Scholar CrossRef Search ADS PubMed  24 Gladman D, Ginzler E, Goldsmith C et al.   The development and initial validation of the Systemic Lupus International Collaborating Clinics/American College of Rheumatology damage index for systemic lupus erythematosus. Arthritis Rheum  1996; 39: 363– 9. Google Scholar CrossRef Search ADS PubMed  25 Bertoli AM, Vilá LM, GS A et al.   Factors associated with arterial vascular events in PROFILE: a Multiethnic Lupus Cohort. Lupus  2009; 18: 958– 65. Google Scholar CrossRef Search ADS PubMed  26 Joliffe IT, Morgan BJ. Principal component analysis and exploratory factor analysis. Stat Methods Med Res  1992; 1: 69– 95. Google Scholar CrossRef Search ADS PubMed  27 Bartko JJ. The intraclass correlation coefficient as a measure of reliability. Psychol Rep  1966; 19: 3– 11. Google Scholar CrossRef Search ADS PubMed  28 Erkan D, Derksen WJ, Kaplan V et al.   Real world experience with antiphospholipid antibody tests: how stable are results over time? Ann Rheum Dis  2005; 64: 1321– 5. Google Scholar CrossRef Search ADS PubMed  29 de Moerloose P, Reber G, Musial J, Arnout J. Analytical and clinical performance of a new, automated assay panel for the diagnosis of antiphospholipid syndrome. J Thromb Haemost  2010; 8: 1540– 6. Google Scholar CrossRef Search ADS PubMed  30 Sciascia S, Bertolaccini ML. Thrombotic risk assessment in APS: the Global APS Score (GAPSS). Lupus  2014; 23: 1286– 7. Google Scholar CrossRef Search ADS PubMed  31 Hata K, Andoh A, Shimada M et al.   IL-17 stimulates inflammatory responses via NF-kappaB and MAP kinase pathways in human colonic myofibroblasts. Am J Physiol Gastrointest Liver Physiol  2002; 282: G1035– 44. Google Scholar CrossRef Search ADS PubMed  32 De Craemer AS, Musial J, Devreese KM. Role of anti-domain 1-beta2 glycoprotein I antibodies in the diagnosis and risk stratification of antiphospholipid syndrome. J Thromb Haemost  2016; 14: 1779– 87. Google Scholar CrossRef Search ADS PubMed  33 Hoxha A, Ruffatti A, Mattia E et al.   Relationship between antiphosphatidylserine/prothrombin and conventional antiphospholipid antibodies in primary antiphospholipid syndrome. Clin Chem Lab Med  2015; 53: 1265– 70. Google Scholar CrossRef Search ADS PubMed  34 Mahler M, Albesa R, Zohoury N et al.   Autoantibodies to domain 1 of beta 2 glycoprotein I determined using a novel chemiluminescence immunoassay demonstrate association with thrombosis in patients with antiphospholipid syndrome. Lupus  2016; 25: 911– 6. Google Scholar CrossRef Search ADS PubMed  35 Pengo V, Ruffatti A, Legnani C et al.   Incidence of a first thromboembolic event in asymptomatic carriers of high-risk antiphospholipid antibody profile: a multicenter prospective study. Blood  2011; 118: 4714– 8. Google Scholar CrossRef Search ADS PubMed  36 Otomo K, Atsumi T, Amengual O et al.   Efficacy of the antiphospholipid score for the diagnosis of antiphospholipid syndrome and its predictive value for thrombotic events. Arthritis Rheum  2012; 64: 504– 12. Google Scholar CrossRef Search ADS PubMed  37 Pengo V, Ruffatti A, Tonello M et al.   Antiphospholipid syndrome: antibodies to Domain 1 of β2-glycoprotein 1 correctly classify patients at risk. J Thromb Haemost  2015; 13: 782– 7. Google Scholar CrossRef Search ADS PubMed  38 Banzato A, Pozzi N, Frasson R et al.   Antibodies to domain I of beta(2)glycoprotein I are in close relation to patients risk categories in antiphospholipid syndrome (APS). Thromb Res  2011; 128: 583– 6. Google Scholar CrossRef Search ADS PubMed  39 Iwaniec T, Kaczor MP, Celińska-Löwenhoff M, Polański S, Musiał J. Clinical significance of anti-domain 1 beta2-glycoprotein I antibodies in antiphospholipid syndrome. Thromb Res  2017; 153: 90– 4. Google Scholar CrossRef Search ADS PubMed  40 Sciascia S, Murru V, Sanna G et al.   Clinical accuracy for diagnosis of antiphospholipid syndrome in systemic lupus erythematosus: evaluation of 23 possible combinations of antiphospholipid antibody specificities. J Thromb Haemost  2012; 10: 2512– 8. Google Scholar CrossRef Search ADS PubMed  41 Sciascia S, Sanna G, Murru V, Khamashta MA, Bertolaccini ML. Validation of a commercially available kit to detect anti-phosphatidylserine/prothrombin antibodies in a cohort of systemic lupus erythematosus patients. Thromb Res  2014; 133: 451– 4. Google Scholar CrossRef Search ADS PubMed  42 Taraborelli M, Lazzaroni MG, Martinazzi N et al.   The role of clinically significant antiphospholipid antibodies in systemic lupus erythematosus. Reumatismo  2016; 68: 137– 43. Google Scholar CrossRef Search ADS PubMed  43 Sciascia S, Sanna G, Murru V et al.   GAPSS: the Global Anti-Phospholipid Syndrome Score. Rheumatology (Oxford)  2013; 52: 1397– 403. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

RheumatologyOxford University Press

Published: Apr 17, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off