Background: Among systemic lupus erythematosus (SLE) patients who are anti–double-stranded DNA (anti-dsDNA) ELISA–positive, confirmatory Crithidia luciliae immunofluorescence test (CLIFT) may identify clinically and biologically meaningful sub-phenotypes. We assessed CLIFT-associated clinical–analytical differences in a large cohort and evaluated a CLIFT-associated proteomic signature in a subcohort, including whether proteomic associations persist after accounting for clinical heterogeneity.

Objectives: To compare clinical–analytical features and targeted proteomics between CLIFT-positive and CLIFT-negative anti-dsDNA ELISA–positive SLE patients

Methods: All patients fulfilled the 2019 EULAR/ACR classification criteria for SLE and were anti-dsDNA ELISA–positive. In the clinical cohort (n=287; CLIFT− n=133, CLIFT+ n=154), demographic, clinical, immunological, and treatment variables were collected through structured review of electronic medical records. Group comparisons were performed and multivariable logistic regression was fitted including variables with univariable p<0.15, using backward (Wald) selection. In the proteomic subcohort (n=77; 47 CLIFT+, 30 CLIFT−), peripheral blood samples were collected and processed for targeted proteomic profiling using Olink panels (inflammation and cardiovascular-related proteins). Differential abundance was tested using Welch’s t-test with Benjamini–Hochberg false discovery rate (FDR) correction. To assess independence from clinical differences, protein-wise linear models were fitted (protein ~ CLIFT) and then clinically adjusted for age, serositis, renal and obstetric manifestations, anti-Sm/anti-Ro/anti-RNP, mycophenolate mofetil, SLEDAI, cancer, and accrued damage (SLICC).

Results: In the clinical cohort, CLIFT+ patients were younger than CLIFT− (48.9±15.5 vs 57.7±15.8 years; p<0.01) and had a younger age at diagnosis (32.7±14.8 vs 39.2±16.0; p<0.01). CLIFT+ showed higher disease activity (SLEDAI 1.95±2.29 vs 0.98±1.89; p<0.01) and lower C4 (19.3±9.1 vs 21.8±9.0; p<0.01). Clinically, CLIFT+ had more renal involvement (35.7% vs 17.3%; p<0.01), neuropsychiatric involvement (25.3% vs 13.5%; p=0.03), serositis (27.9% vs 12.0%; p<0.01), and obstetric morbidity (14.9% vs 3.0%; p<0.01). The autoimmune profile in CLIFT+ was enriched for ENA antibodies, including anti-Sm (32.5% vs 9.0%; p<0.01), anti-Ro (44.8% vs 26.3%; p<0.01), and anti-RNP (29.2% vs 9.0%; p<0.01). Treatment exposure also differed, with higher mycophenolate mofetil use (39.6% vs 18.0%; p<0.01), higher “any immunosuppressive ever” (57.8% vs 43.6%; p=0.02), and higher belimumab use (11.7% vs 5.3%; p=0.03). Neoplasms were more frequent in CLIFT− (9.8% vs 3.6%; p=0.01). In multivariable analysis, CLIFT positivity remained independently associated with serositis (OR 3.07; 95%CI 1.50–6.30; p<0.01), obstetric morbidity (OR 8.57; 2.61–28.17; p<0.01), anti-Sm (OR 2.87; 1.25–6.59; p=0.01), anti-Ro (OR 2.14; 1.16–3.92; p=0.01), anti-RNP (OR 2.67; 1.17–6.05; p=0.02), and mycophenolate mofetil exposure (OR 2.67; 1.38–5.14; p<0.01), while current age (OR 0.98 per year; 0.96–0.99; p=0.02) and neoplasms (OR 0.27; 0.07–0.99; p=0.05) were inversely associated with CLIFT positivity. In the proteomic subcohort, the Top12 CLIFT-associated proteins (volcano ranking) were TFRC, IL10, AZU1, CPA1, MMP9, TNFRSF10C, CNTN1, CCL24, PGLYRP1, CPB1, COL1A1, and IL6R . Clinically adjusted protein-wise models indicated that AZU1 remained independently associated with CLIFT status (p=0.022), with borderline adjusted associations for IL10 (p=0.051) and TFRC (p=0.057), suggesting that part of the proteomic divergence is not fully explained by clinical differences captured in the cohort. Enrichment analysis of the Top12 proteins yielded clinically interpretable themes consistent with immune trafficking/chemotaxis, cytokine signaling and regulation, innate host-defense responses, and tissue remodeling/motility.

Conclusions: In anti-dsDNA ELISA–positive SLE, CLIFT positivity identifies a distinct clinical phenotype and a coherent proteomic signature, with AZU1 persisting after clinical adjustment.

REFERENCES: NIL.

Acknowledgments: NIL.

Disclosure of Interests: Ignacio Gómez-García GSK, AstraZeneca, Boehringer, Santiago Dans-Caballero: None declared, Carlos Perez-Sanchez: None declared, Alejandro Escudero-Contreras: None declared, Rafaela Ortega-Castro: None declared, María Ángeles Aguirre-Zamorano: None declared, Chary Lopez-Pedrera: None declared.