Revisiting the changes in the Banff classification for antibody-mediated rejection after kidney transplantation

The Banff classification for antibody-mediated rejection (ABMR) has undergone important changes, mainly by inclusion of C4d-negative ABMR in Banff'13 and elimination of suspicious ABMR (sABMR) with the use of C4d as surrogate for HLA-DSA in Banff'17. We aimed to evaluate the numerical and prognostic repercussions of these changes in a single-center cohort study of 949 single kidney transplantations, comprising 3662 biopsies that were classified according to the different versions of the Banff classification. Overall, the number of ABMR and sABMR cases increased from Banff'01 to Banff'13. In Banff'17, 248 of 292 sABMR biopsies were reclassified to No ABMR, and 44 of 292 to ABMR. However, reclassified sABMR biopsies had worse and better outcome than No ABMR and ABMR, which was mainly driven by the presence of microvascular inflammation and absence of HLA-DSA, respectively. Consequently, the discriminative performance for allograft failure was lowest in Banff'17, and highest in Banff'13. Our data suggest that the clinical and histological heterogeneity of ABMR is inadequately represented in a binary classification system. This study provides a framework to evaluate the updates of the Banff classification and assess the impact of proposed changes on the number of cases and risk stratification. Two alternative classifications introducing an intermediate category are explored.