Identification of Biomarkers That Predict Response to Treatment of Lupus Nephritis With Mycophenolate Mofetil or Pulse Cyclophosphamide

Objective. There is a need to identify clinical characteristics and/or biomarkers that can predict treatment outcome in lupus nephritis. To this end, we utilized data from the Aspreva Lupus Management Study to identify possible baseline and early predictors of renal response to mycophenolate mofetil (MMF) or intravenous (IV) cyclophosphamide (CYC). Methods. Patients with class III-V lupus nephritis were randomized to MMF or IV CYC. We assessed predictors of renal response, including baseline demographic, clinical, laboratory, and histologic characteristics, as well as early clinical and laboratory data, obtained within the first 2 months of therapy. Odds ratios (ORs) and 95% confidence intervals for renal response were calculated for each putative predictor. Results. Normalization of C3, C4, or both by week 8 was strongly predictive of renal response at week 24 (ORs 2.5, 2.6, and 2.9, respectively; P < 0.05). Reduction in proteinuria by >= 25% by week 8 was predictive of renal response at week 24 (OR 3.2, P < 0.05). Reduction in anti-double-stranded DNA (anti-dsDNA) by week 8 was not predictive of renal response. Only 3 baseline characteristics (C4 level, time since diagnosis of lupus nephritis, and estimated glomerular filtration rate [GFR]) were predictive of renal response; the remaining characteristics (age, age at lupus nephritis onset, time since diagnosis of systemic lupus erythematosus, sex, histopathologic class, anti-dsDNA antibody level, C3 level, level of proteinuria, and use of angiotensin-converting enzyme inhibitors, statins, or hydroxychloroquine) were not. Conclusion. This study demonstrates that baseline C4 level, time since diagnosis of lupus nephritis, baseline estimated GFR, early normalization of complement, and reduction in proteinuria independently predict renal response to therapy at 6 months.