Studies of clinical and haemostatic features associated with antiphospholipid antibodies

Abstract

Background: Antiphospholipid antibodies (aPL) are associated with a heterogeneous range of clinical conditions with predisposition for thrombosis, pregnancy morbidity and other autoimmune disorders, most frequently systemic lupus erythematosus (SLE). Both the clinical manifestations themselves and their treatment may be harmful to the affected individuals and aPL testing is still controversial. There is thus an urgent need for improved aPL diagnostics and individual risk assessments. Exploring the largely unknown disease aetiology may provide clues to new biomarkers. From previous studies of other pro-thrombotic conditions, we hypothesized that fibrin clot properties, fibrinolytic function, microparticle (MP) profile and smoking could be important in aPL-related disease.

Aims: To study clinical, serological and haemostatic features of the antiphospholipid syndrome (APS) and aPL-related disorders, including SLE, in order to improve diagnostic methods and enhance understanding of disease pathogenesis.

Materials and methods: Two patient groups were cross-sectionally studied. In paper I, IV, and V, the Karolinska SLE cohort was studied; in paper I together with patients from the Swedish SLE network (n=712 patients) and in paper IV in comparison with matched population controls (n=290 patients). In paper II and III, APS patients (n=49 and 52 patients, respectively) from Karolinska University Hospital were compared both to healthy controls (paper II and III) and to non- APS thrombotic controls (paper II). Laboratory assays included: different ELISAs for aPL detection, liquid permeation technique and scanning electron microscopy (SEM) for examination of fibrin network permeability and structure, turbidimetric clotting and lysis assays, and finally MP measurement by flow cytometry.

Results: In paper I, we report only moderate agreement (Kappa-values 0.16-0.71) but similar, modest association with previous thrombotic events when comparing a new automated aPL method with standard assays in SLE patients. Antibody isotype and titer influenced the association with clinical events. Odds ratios for lupus anticoagulant (LA) for the associations with vascular events were generally higher than for the specific aPL immunoassays. In paper II, we report that fibrin clots formed in vitro in samples from APS patients have a decreased permeability compared to the clots formed in samples from healthy controls and non-APS thrombosis controls (p<0.0001 for both). SEM images visually confirmed denser fibrin structure in APS. No clear difference in fibrinolysis function between the APS patients and controls were observed (p>0.05 for two of the three investigated parameters) but APS patients with previous arterial thrombosis had prolonged clot lysis times compared to the controls (p<0.05). In paper III, we report that the number of lactadherin positive MPs, endothelial MPs, Tissue factor (TF)-positive endothelial MPs and monocyte MPs in APS samples were increased compared to healthy controls (p<0.001 for all analyses). There was no significant difference in the number of platelet MPs (p=0.13). APS patients had a high proportion of MPs negative to lactadherin. In paper IV, we report that the number of lactadherin positive MPs, endothelial MPs, platelet MPs and leukocyte MPs in SLE samples were increased compared to matched population controls (p<0.0001 for all analyses). Moreover, MPs exposing inflammation and/or activation markers such as CD40L, TF, vascular cell adhesion molecule 1(VCAM-1), high-mobility group protein B1 (HMGB1) and C4d also were also clearly increased in the SLE patients (p<0.0001 for all analyses). However, MP number could neither be used to distinctively discriminate between SLE patients with or without APS nor between SLE patients with or without previous vascular events in multivariable models (p-values for all analyses >0.05 except for VCAM-1- positive endothelial MPs which had p-values of 0.044 and 0.047 for positive associations in the vascular event models). In paper V we report a positive association between cigarette smoking (ever smoking and, most strikingly, former smoking) and the presence of aPL among patients with SLE (OR ≈3 for aPL of the IgG isotype and for LA and former smoking vs. never smoking in a multivariable model). A positive interaction between ever smoking and aPL was noted for the association with previous vascular events. The additive interaction analysis demonstrated a significant interaction between ever smoking and LA (attributable proportion due to interaction, AP=0.80, 95% CI 0.5 - 1.0) and ever smoking and triple aPL positivity (AP=0.85, 95% CI 0.6 - 1.0) for the association with presence of previous vascular events.

Conclusions: Automated aPL assays are promising for facilitating aPL testing. However, it is crucial to also evaluate LA positivity, antibody titers and isotypes together with smoking status when aPL results are used for risk assessment in individual patients. Smoking, abnormal fibrin clot properties, fibrinolytic function, and MP profile may contribute to the pathogenesis of aPL-associated disease. MP formation is likely to play a role in SLE as well, which is a disorder closely related to APS. However, the exact relationship between aPL, vascular events and MPs in SLE patients needs to be further studied. Further mechanistic studies are needed to address how the different variables studied in this thesis are related to the pathogenesis of aPL-associated disease. The findings and conclusions in this thesis also need to be prospectively studied to clarify their prognostic value. With such additional studies, our observations could guide the search for new biomarkers and risk scores in aPL-related disease.