Serum levels of autoantibodies against C-reactive protein correlate with renal disease activity and response to therapy in lupus nephritis

Introduction Serum levels of C-reactive protein (CRP) seldom reflect disease activity in systemic lupus erythematosus (SLE). We have previously shown that autoantibodies against neoepitopes of CRP often occur in SLE, but that this does not explain the modest CRP response seen in flares. However, we have repeatedly found that anti-CRP levels parallel lupus disease activity, with highest levels in patients with renal involvement; thus, we aimed to study anti-CRP in a material of well-characterized lupus nephritis patients. Methods Thirty-eight patients with lupus nephritis were included. Treatment with corticosteroids combined with cyclophosphamide, mycophenolate mofetil or rituximab was started after baseline kidney biopsy. A second biopsy was taken after >= 6 months. Serum creatinine, cystatin C, complement, anti-dsDNA, anti-CRP and urinalysis were done on both occasions. Biopsies were evaluated regarding World Health Organisation (WHO) class and indices of activity and chronicity. Renal disease activity was estimated using the British Isles Lupus Assessment Group (BILAG) index. Results At baseline, 34/38 patients had renal BILAG-A; 4/38 had BILAG-B. Baseline biopsies showed WHO class III (n = 8), IV (n = 19), III to IV/V (n = 3) or V (n = 8) nephritis. Seventeen out of 38 patients were anti-CRP-positive at baseline, and six at follow-up. Overall, anti-CRP levels had dropped at follow-up (P < 0.0001) and anti-CRP levels correlated with renal BILAG (r = 0.29, P = 0.012). A positive anti-CRP test at baseline was superior to anti-dsDNA and C1q in predicting poor response to therapy as judged by renal BILAG. Baseline anti-CRP levels correlated with renal biopsy activity (r = 0.33, P = 0.045), but not with chronicity index. Anti-CRP levels were positively correlated with anti-dsDNA (fluorescence-enhanced immunoassay: r = 0.63, P = 0.0003; Crithidia luciliae immunofluorescence microscopy test: r = 0.44, P < 0.0001), and inversely with C3 (r = 0.35, P = 0.007) and C4 (r = 0.29, P = 0.02), but not with C1q (r = 0.14, P = 0.24). No associations with urinary components, creatinine, cystatin C or the glomerular filtration rate were found. Conclusions In the present study, we demonstrate a statistically significant correlation between anti-CRP levels and histopathological activity in lupus nephritis, whereas a baseline positive anti-CRP test predicted poor response to therapy. Our data also confirm previous findings of associations between anti-CRP and disease activity. This indicates that anti-CRP could be helpful to assess disease activity and response to therapy in SLE nephritis, and highlights the hypothesis of a pathogenetic role for anti-CRP antibodies in lupus nephritis.