Background Anti-myeloperoxidase (MPO) antibodies have been shown to predict the development of proliferative lupus nephritis (LN) suggesting anti-neutrophil cytoplasmic antibodies (ANCA) may have a pathogenic and prognostic role in LN. This study compared the type of LN, renal function, and systemic lupus erythematosus (SLE)-related and antiphospholipid autoantibodies between LN patients who were ANCA (anti-proteinase 3 (PR3) and anti-MPO antibodies) positive and negative.

Methods Patients fulfilling the ACR or SLICC Classification Criteria for SLE were enrolled in a local cohort. We retrospectively identified patients with Class 2, 3, 4, or 5 LN on renal biopsy who also had an ANCA, plasma creatinine, and urine protein creatinine ratio (UPCR) at time of biopsy. ANCA by IIF was performed on ethanol and formalin-fixed polymorphonuclear leukocytes and a HEp-2 cell biochip (EuroPattern, Euroimmun GmbH, Luebeck, Germany) while antibodies to MPO and PR3 were determined by multiplex immunoassay (Bio-Rad, Hercules, CA: BioPlex 2200, cutoff ≥2 KEU/L). Using sera collected at enrollment, SLE-related autoantibodies (dsDNA, Sm, U1RNP, Sm, Ro52/TRIM21, Ro60/SSA, SS-B/La, Scl-70, Jo-1, RiboP, PCNA, PM/Scl) were performed by laser bead immunoassay (Euroimmune), lupus anticoagulant by tissue thromboplastin inhibition test and dilute Russell viper venom time, and anti-cardiolipin IgG and anti-2 glycoprotein-1 IgG by ELISA. Comparisons were performed with Fishers exact or Mann-Whitney U.

Results 23 SLE patients with LN were included; 82.6% were female. Most patients (20/23, 87.0%) were ANCA positive by IIF while only 5/23 (21.7%) had antibodies to MPO (3/23, 13.0%) or both MPO and PR3 (2/23, 8.7%). Anti-MPO/PR3 positive patients had p-ANCA (2/5, 40%) or an atypical pattern (3/5, 60%) on ANCA IIF. When comparing anti-MPO/PR3 positive (5) to negative (18) patients, there was no difference in LN class, creatinine, UPCR, or the presence of SLE-related autoantibodies. Anti-cardiolipin IgG antibodies were more common in anti-MPO/PR3 positive patients (60.0% vs 5.6%, p=0.021), while a nuclear pattern on ANCA IIF was more common in anti-MPO/PR3 negative patients (55.6% vs. 0%, p=0.046). Of note, three of the five LN patients with cerebrovascular accidents or venous thrombosis were anti-MPO/PR3 positive.

Conclusions Positive ANCA by IIF was common in LN, however, only a fifth of LN patients had anti-MPO/PR3 antibodies which were associated with anti-cardiolipin IgG antibodies. Anti-MPO/PR3 negative patients were more likely to have a positive ANCA IIF with nuclear staining. We are currently comparing ANCA positivity between SLE patients with and without LN.

Funding Source(s): The Arthritis Society Chair in Rheumatic Diseases at the Cumming School of Medicine, University of Calgary