Development Of The Risk Prediction Model To Predict The Risk Of Venous Thromboembolism In Patients With Systemic Lupus Erythematosus And Glycoanalysis Of Immunoglobulin G

Part ?:Development of the risk prediction model to predict the risk of venous thromboembolism in patients with systemic lupus erythematosusObjective:To explore the clinical characteristics and risk factors of patients with systemic lupus erythematosus(SLE)combined with venous thromboembolism(VTE)and establish a VTE risk prediction model for SLE patients.Methods:We enrolled the national multi-center SLE patients who participated in the Chinese SLE treatment and research group(CSTAR)registration cohort,and collected basic baseline information and clinical laboratory indicators at the time of registration for all patients.All patients were followed up every 3-6 months and VTE events were recorded.Electronic questionnaires were used to review and supplement VTE events occurred during follow-up.Through single-factor COX univariate analysis,the risk factors of VTE in SLE were found,combined with the clinical background.The candidate variables were then selected by Lasso regression,and the prediction model is established by multi-factor COX regression.The C-Index was used to evaluate the discrimination of the model.And the calibration curve was used to evaluate the calibration of the model.Finally the optimal risk model was determined and the Nomogram was established according to the optimal risk prediction model.The optimal risk prediction model was then validated in the external data.Results:A total of 4752 patients were collected,and 198 patients had venous thrombotic events.Choose 7/10 as the training data and 3/10 as the validation data.The 9 variables included in the optimal risk prediction model include:Male,age of onset,BMI?25kg/m2,no use of hydroxychloroquine,hyperlipidemia,renal involvement,nervous system involvement,anti-? 2GPI antibody positive,LA positive.The C-Index of the optimal prediction model was 0.776 in the training data and 0.772 in the validation data.According to the optimal risk prediction model,the Nomogram was established and could predict the 1 year,3 years,and 5 years risk of VTE after the onset of SLE.Conclusion:A variety of factors are related to the occurrence of VTE in SLE.The first COX regression prediction model can accurately predict the risk of VTE in SLE.Part ?:Glycoanalysis of immunoglobulin G in systemic lupus erythematosus and antiphospholipid syndromeObjective:To explore the expression of IgG glycoanalysis in patients with SLE and antiphospholipid syndrome(APS)using a lectin chip;to explore the association between IgG glycosylation and phenotypes of SLE and APS.Methods:The lectin chip was used to detect the serum IgG glycosylation level of SLE patients(N=75)(divided into SLE secondary APS group(n=25),SLE antiphospholipid antibody positive group(n=25),SLE antiphospholipid antibody negative group(n=25))and APS patient(divided into venous thrombosis APS group(n=12),arterial thrombosis APS group(n=13),obstetric APS group(n=15)and mixed APS group(n=28)).Sjogren syndrome patients(N=26)were chosen as the disease control,and healthy human(N=26)as the healthy control.The levels of 56 kinds of lectins were compared between groups.Results:Decreased mannosylation levels were observed in patients with primary APS.The levels of galactosylation and core fucosylation were significantly reduced in SLE-APS patients.Acetamidogalactose levels in arterial thrombosis APS patients were reduced.Acetamidogalactose levels were increased and fucose levels were reduced in obstetric APS patients.Conclusion:The IgG glycome in SLE and APS patients is significantly altered,suggesting that abnormal IgG glycosylation may be potential biomarkers in SLE and APS.