#1134 Molecular rejection phenotype patterns differentiate between probable AMR, DSA-negative microvascular inflammation, and AMR

Abstract Background and Aims Biopsy-based transcripts associated with antibody-mediated rejection (AMR) hold promise as substitutes for C4d positivity, according to the latest Banff Meeting Report of 2022. However, patterns and strength of rejection phenotype scores remain inadequately studied within the new subgroups along the AMR continuum. Method We analyzed 365 kidney allograft biopsies by histology and the Molecular Microscope Diagnostic System (MMDx) at the University Hospital of Zurich from July 2021 to November 2023. Histologic findings were classified according to Banff 2022 into (1) probable AMR (n = 32), (2) donor-specific antibodies (DSA)-negative C4d-negative microvascular inflammation (MVI, n = 32), and (3) AMR (n = 60). A control group of 118 cases without MVI was used for comparison. Results 10 out of 32 cases (31%) with probable AMR, 20 of 32 cases (63%) with DSA-negative C4d-negative MVI, and 31 of 60 cases (52%) with AMR showed molecular AMR. Among cases with molecular AMR the sum of the TCMR-related (R2) and late-stage AMR-related (R6) phenotype scores was significantly higher in probable AMR and DSA-negative MVI cases compared to AMR cases (p = 0.029). In contrast, the sum of the mixed rejection (R3), early-stage AMR (R4), fully-developed AMR (R5) phenotype scores was higher in AMR compared to probable AMR (p = 0.063). Among cases without molecular AMR the all AMR phenotype score was significantly higher in AMR and DSA-negative MVI compared to probable AMR and the control group without MVI (p = 0.012, p < 0.001). Conclusion Our results identify different molecular rejection phenotype patterns between cases with probable AMR, DSA-negative MVI, and AMR. Sub-threshold findings appear relevant along the AMR continuum.