The gold standard for the classification of lupus nephritis (LN) is based on renal histology but variation in interpretation exists. The aim of this multicentric international study was to assess the inter-observer variability of the 2003 International Society Nephrology/Royal Pathology Society (ISN/RPS) LN histological classification criteria in children.

Expert histopathologists from a reference centre (USA) and three tertiary paediatric centres (UK) independently scored digitalized histology slides from percutaneous kidney biopsies of 55 children with LN. Histological ISN/RPS Class and additional features [LN-activity (scored 0–24), LN-chronicity (0–12), and tubulo-interstitial activity (TIA: 0–21)] were scored. Statistical analysis (Kappa score and intra-class correlation, ICC) quantified the level of agreement, interpreted as: 0.01–0.20 slight; 0.21–0.40 fair; 0.41–0.60 moderate; 0.61–0.80 good; 0.81–1.00 excellent agreement.

In the cohort (73% females), the mean±standard error (SE) age at the time of biopsy was 15.5±0.39 years. Based on the reference centre, 42% (23/55) had ISN/RPS class IV with a LN-activity score of 4.23±0.50, LN-chronicity 1.81±0.18, and TIA 4.45±0.35. There were between 4–54 (mean 16.7) glomeruli per biopsy. Pathologists had only fair agreement for LN ISN/RPS assignment (kappa 0.26±0.12). LN-chronicity and TIA scoring were also only fair (ICC 0.36±0.09 and 0.22±0.09 respectively). Conversely, there was good agreement for scoring LN-activity features (ICC 0.69±0.06). When the biopsy findings were categorized into proliferative and non-proliferative disease, poor agreement remained (kappa 0.24±0.11).

Despite unified criteria for the interpretation of histological features of LN, marked reporting variation remains and this may influence treatment decisions. Adding activity features to the biopsy report improves inter-observer agreement. Supplementary measures of predicting the extent of LN, such as the use of non-invasive urine biomarkers, are required.