Invited commentary by Dr. Parmjeet Randhawa
IgA nephropathy is the most common glomerulonephritis worldwide. Over the years, several pathologic classifications have been devised to determine prognosis in individual patients, but none has gained widespread acceptance.
A good pathology grading system must contribute additional prognostic information beyond that provided by clinical features. The Oxford Classification System for IgA Nephropathy certainly satisfies this criterion. This system was based on actual clinical data, international collaboration, and validation of the reproducibility of defining pathologic lesions. In the original study of 265 patients published by the Oxford group, four lesions, namely endocapillary hypercellularity, mesangial hypercellularity, segmental sclerosis, and tubular atrophy/interstitial fibrosis, were found to predict clinical outcome.
In a recent study published in the American Journal of Kidney Diseases, Zeng et al examine the utility of the Oxford system on a substantially larger cohort of 1026 patients from China. Whereas the prognostic value of mesangial hypercellularity and tubular atrophy/interstitial fibrosis was confirmed, segmental sclerosis and endocapillary hypercellularity were not informative. The difference in results likely reflects different distributions of pathologic lesions in the Chinese and Oxford cohorts. Segmental sclerosis or adhesion had an almost ubiquitous presence in Zeng et al’s study (83% of all biopsies), while endocapillary hypercellularity was uncommon (11% of all biopsies). Such marked differences in frequency can result in specific features losing their statistical significance. Surprisingly, necrosis and crescent formation were not identified as prognostic factors in either the Chinese or Oxford study.
It may be that small numbers of crescents in a biopsy do not have a marked effect on the overall course of a chronic progressive disease. Furthermore, cases with crescents may have received more aggressive immunosuppression, thus obscuring a potential relationship between the two. These confounding variables highlight the need for large prospective studies to better define the contribution of pathology lesions to clinical outcomes.
Parmjeet Randhawa, MD
Professor of Pathology, University of Pittsburgh Medical Center
AJKD Associate Editor