Reproducibility studies on arteriolar hyaline thickening scoring in calcineurin inhibitor-treated renal allograft recipients.

Arteriolar hyaline thickening (AH) is the most characteristic lesion of chronic calcineurin inhibitor nephrotoxicity. This study was performed to compare the inter-observer reproducibility of AH scoring using Banff criteria and a newly proposed criterion. Forty-five nonprotocol post-transplant biopsies from 38 patients immunosuppressed with tacrolimus or cyclosporine A (CsA) were included. The severity of AH was blindly scored by three observers. According to the new criteria, AH is graded based on circular vs. noncircular involvement and the number of arterioles involved. The kappa statistics were used to assess the inter-observer reproducibility. Twenty-seven (60%) biopsies showed AH. The AH grades by both criteria were correlated with serum creatinine at biopsy and inversely correlated with estimated glomerular filtration rate (GFR) (p < 0.05). The recent AH criteria improved the mean pairwise agreement (79.4% vs. 68%) and the overall kappa value (0.67 vs. 0.52) (p = 0.02) compared to Banff criteria. The mean inter-slide variation using Banff and the new criterion were 23% and 27.6%, respectively (p > 0.05). The new AH criterion results in better inter-observer reproducibility, and is clinically validated against serum creatinine and estimated GFR. There is substantial intra-biopsy variation, therefore, evaluation of more than one section is crucial to determine severity of arteriolar damage more accurately.