Document resource
Objectives The discovery of novel biomarkers associated with systemic lupus erythematosus (SLE) phenotypes and outcomes may enable more effective use of targeted therapeutics. Although serum proteomics shows promise for biomarker discovery, few studies have examined relationships between proteomic profiles and detailed longitudinal SLE phenotypes and outcome measures. We aimed to generate unbiased serum mass spectrometry (MS) proteomic profiles in SLE patients, and to explore associations with socio-demographics, SLE phenotypes and outcomes.Methods Untargeted label-based proteomics was performed on serum from 60 SLE patients enrolled in the Australian Lupus Registry and Biobank, using liquid chromatography-tandem MS (LC-MSMS). The cohort was balanced for self-reported ethnicity (Asian and European) and disease activity (SLEDAI-2K). Proteins with missing values in >50% of patients were excluded; imputation was performed in those with missing values in <=50%. Associations between proteins and clinical features were assessed using differential abundance (DA) analysis with regression models, and supervised learning with least absolute shrinkage and selection operator (LASSO). Age, sex, ethnicity and renal impairment (eGFR <60 ml/min) were included as covariates in multivariable linear regression and LASSO models.Results We quantified 1,092 proteins in 59 SLE patients (1 dialysis patient excluded), with a median (IQR) age of 42 (35-52) years. 92% of patients were female, 49% were of Asian ethnicity, 51% were of European ethnicity and 49% had active disease (SLEDAI-2K >4). Multiple overlapping protein associations between univariable and multivariable DA and LASSO were found. Eleven proteins (AMY1B, CFHR2, CFHR5, F12.1, FN1, FUCA1, MASP1.2, PVR, QPCT, QSOX1, ZMYND15) were associated with Asian ethnicity. APOC3, CSPG4, CTSS, FABP4 and LDHB were associated with organ damage. C3.4 and TNXB were associated with haematological disease across the study period. Eleven proteins (ACAN, AMBP, APOA4, ART3, CDH11, HYOU1, IGFBP6, LCAT, P4HB, PRKCSH, RBP4) showed associations with SLEDAI-2K defined active lupus nephritis at the time of serum sample collection using univariable DA and LASSO, but not multivariable DA.Conclusions Untargeted quantitative proteomics identified distinct protein patterns associated with Asian ethnicity, organ damage, haematological disease and active lupus nephritis in SLE, highlighting the potential for this approach to generate clinically useful biomarkers that could refine disease stratification and guide precision management.