Value of the Oxford classification of IgA nephropathy in children with Henoch–Schönlein purpura nephritis

Ke Xu,Lili Zhang,Jie Ding,Suxia Wang,Baige Su,Huijie Xiao,Fang Wang,Xuhui Zhong,Yanming Li
DOI: https://doi.org/10.1007/s40620-017-0457-z
IF: 4.4061
2017-01-01
Journal of Nephrology
Abstract:Background The widely used International Study of Kidney Disease in Children (ISKDC) classification for Henoch-Schönlein purpura nephritis (HSPN) does not completely correlate with the clinical presentation and long-term prognosis of this disease. Primary IgA nephropathy (IgAN) and HSPN share common features; thus, the Oxford classification of IgAN might be useful in predicting the long-term outcomes of HSPN. However, its value has not been confirmed in children with HSPN. Methods We selected children with HSPN diagnosed between 2003 and 2015, and reclassified their renal biopsies according to the Oxford classification scoring system. The primary outcome was impaired renal function, and remission of proteinuria and clinical remission were secondary outcomes. Results We included 104 patients (58 males, 46 females) with a median age of 10 (4–17) years. Mesangial hypercellularity (M1) was strongly associated with proteinuria, and tubular atrophy/interstitial fibrosis (T1&2) and C2 (with crescents in > 25% of glomeruli) were associated with reduced estimated glomerular filtration rate (eGFR) at the time of biopsy. Patients with M1, endocapillary proliferation (E1), segmental glomerulosclerosis (S1), and crescents (C1&2) were more likely to have been treated with high-dose methylprednisolone. At univariate time-dependent analyses, S1 was strongly associated with the primary outcome (p = 0.025), whereas T1&2 was significantly negatively associated with proteinuria remission (p = 0.035) and clinical remission (p = 0.038). Conclusions Our findings suggest that the Oxford classification is valid for children with HSPN. S and T lesions, which are ignored in the ISKDC classification, can be used to assess renal outcomes of HSPN, and such assessments are not affected by currently available treatments. The value of M, E and C lesions in predicting response to therapy and renal outcome warrants further study.
What problem does this paper attempt to address?