Continuous indices to assess the phenotypic spectrum of kidney transplant rejection

The international Banff classification for kidney transplant pathology discretizes the rejection continuum into distinct diagnostic categories, introducing artificial dichotomization and threshold effects. To better reflect the underlying disease spectrum, we developed, in this cohort study, two novel indices for quantifying antibody-mediated (AMR) and T-cell mediated rejection (TCMR) from histological lesion scores, and calculated indices for overall activity and chronicity. These indices were evaluated in one derivation cohort and two independent validation cohorts, totaling 19,500 biopsies from 8,873 kidney transplant patients across 10 centers worldwide. The AMR, TCMR, and activity indices demonstrated hierarchical ordering between No rejection, intermediate and complete rejection histology. The chronicity index showed limited association with the major diagnostic categories. In the derivation cohort, the AMR and TCMR indices discriminated AMR from absence of AMR, and TCMR from absence of TCMR, with an AUC of 0.98 (95% confidence interval 0.97 to 0.98) and 0.99 (0.99 to 1.00) respectively). This excellent discrimination was confirmed in the validation cohorts. Those indices strictly confined intermediate phenotypes to a range of low index values and related to graft failure rates even within the diagnostic categories, thus reflecting the underlying rejection continuum. The four continuous indices offer an implementable and interpretable global evaluation of kidney transplant biopsy histology while eliminating the need for intermediate diagnostic categories and enable more probabilistic reasoning in the diagnostic approach to the spectrum of kidney transplant rejection.