Agreement between multi-dimensional and renal-specific response criteria in patients with juvenile systemic lupus erythematosus and renal disease

Objectives: To evaluate change over time and level of agreement of renal-specific and multi-dimensional measures in juvenile systemic lupus erythematosus (SLE) with renal disease. Methods: An analysis was made of 205/557 children with baseline 24-hour proteinuria >or=0.5 g. Data were collected at baseline, 6-, 12- and 24-month intervals. Using the Systemic Lupus International Collaborating Clinics (SLICC) renal index (change in proteinuria and urine sediment) as gold standard, responsiveness and discriminative ability analyses were used to identify key renal and multi-dimensional disease activity and damage measures for the evaluation of response to therapy. We also evaluated the kappa agreement between SLICC renal index and PRINTO/ACR juvenile SLE criteria (change in proteinuria, physician and parents evaluations, disease activity, health related quality of life [HRQOL]). Results: Children with renal disease compared to children without renal disease, had a lower female rate and higher disease activity/response rate (p-values <0.01) but similar damage levels. Large responsiveness (standardised response mean >or=0.8) and statistical significant discriminative ability with the SLICC renal index 4 levels of response (improved, partially improved, stable and worsened) were observed for renal specific measures (proteinuria, urine sediment, renal sub-scores, p<0.0001) and for multi-dimensional variables (disease activity level and physician evaluation p<0.001). Agreement between the SLICC renal index and PRINTO/ACR criteria was moderate (0.57; 95% confidence intervals: 0.44-0.71). Conclusions: We propose to incorporate multi-dimensional measures (physician and parents' evaluations, disease activity and HRQOL), in addition to renal specific measures, in future clinical trials in juvenile SLE with renal involvement.