Introduction: The Banff 2013 criteria for diagnosing antibody-mediated rejection (ABMR) includes three lesions: glomerulitis (g), peritubularitis (PTC), and chronic glomerulopathy (cg). The poor reproducibility of these scores and the resultant diagnosis of ABMR may limit their use in clinical trials. The goal of this study was to determine the variability of diagnosing ABMR.

Methods: We evaluated the reproducibility of individual scores and the resulting diagnosis of ABMR by 4 renal pathologists at 3 institutions. Seventy-two biopsies with a spectrum of lesions were rescored (from 46 patients obtained at 3 months to 2 years after transplantation). We hypothesized that a “majority rules” approach in which 3 pathologists' scores were used to reach an agreement would reduce variability.

Results: Using the original C4d scoring available for 67 biopsies, agreement among all 4 pathologists on diagnosis (chronic ABMR (cABMR), active ABMR (aABMR), and no ABMR) was only 55.2% (37/67). The distribution of diagnoses for different pathologists ranged from: 16.4% vs 22.4% for no ABMR; 45.1% vs 29.8% for aABMR; and 35.8% vs 47.7% for cAMBR.

When any 3 pathologist's scores were combined (resulting in 4 different teams of pathologists), the same diagnosis was reached in 90% of cases. A typical result of agreement on diagnosis between 2 pathologists and 2 3-person groups is shown in the Table.

Conclusion: The Banff 2013 ABMR schema has high variability and this variability might have a major impact on its use in a clinical trial. Thus, a majority rules approach is needed for scoring in clinical trials.

CITATION INFORMATION: Smith B., Cornell L., Smith M., Cortese C., Geiger X., Schinstock C., Kremers W., Stegall M. “Majority Rules”: A Method to Reduce Variability in Diagnosing Antibody Mediated Rejection in Renal Allograft Biopsies Using Banff 2013 Am J Transplant. 2017;17 (suppl 3).

« Back to 2018 American Transplant Congress