Background Childhood-onset systemic lupus erythematous (cSLE) patients are at higher risk for renal disease than those with adult-onset disease. Mycophenolate mofetil (MMF) and intravenous cyclophosphamide (IV CTX), commonly used for induction therapy of proliferative lupus nephritis (LN), are considered equally efficacious in adults. Comparative data in the paediatric population are lacking. To reduce treatment variability and facilitate comparative effectiveness studies, the Childhood Arthritis and Rheumatology Research Alliance (CARRA) published a consensus treatment plan (CTP) for induction therapy in childhood proliferative LN. The CTP recommended treatment with MMF or IV CTX and one of three steroid regimens: primarily oral, primarily IV, or mixed oral/IV. We report physician decision-making and 6-month response rates in a multi-centre pilot feasibility study.

Materials and methods This observational study enrolled 41 cSLE patients from 10 CARRA sites. Subjects had new-onset biopsy proven class III or IV proliferative LN and were starting MMF or IV CTX. Complete renal response (CRR), defined as normal renal function, inactive urine sediment, and spot urine protein/creatinine ratio of <0.2, was measured at 6 months. Subjects were followed for up to 24 months. Baseline demographics, disease-related features, physician decision-making and achievement of CRR were compared according to induction treatment group and among steroid regimens.

Results The majority of participants were female (83%) with a mean age of 14 years. There were no significant differences in demographics between MMF or IV CTX groups or among steroid regimens. Those with class IV nephritis (35.3% vs 73%, p = 0.015) and hematuria (36% vs 74% p < 0.001) were more likely to be treated with IV CTX. Physicians more often reported compliance concerns as a reason for selecting treatment for the CTX group compared to MMF (22% vs 0%, p = 0.04). Overall, physicians reported “this is what I or my group always does” as the most common reason for choice of induction agent and steroid regimen. Induction agent use did not differ significantly according to study site. Steroid regimen differed significantly by study site and induction agent. CRR at 6 months was achieved for 56% with MMF and 64% with IV CTX (p = 0.6); the study was not powered to evaluate treatment efficacy.

Conclusions Class IV nephritis, hematuria and patient adherence influenced selection of induction agent. Steroid regimens differed by study site and induction regimen. To evaluate comparative effectiveness, future larger studies will be needed.

Acknowledgements CARRA Registry, Lupus Foundation of America, Arthritis Foundation, Duke Clinical Research Institute. NIH T32GM00756.