Scholar Presentation: Prediction of Antibody-mediated Rejection by Quantifying Eplets Potentially Targeted by Preformed Donor-specific Antibodies

Aim: Preformed donor-specific antibodies (DSA) may cause antibody-mediated rejection (ABMR) despite potent
desensitization protocols. Mean fluorescence intensity (MFI) in single antigen bead assays (SAB) serves as surrogate for
DSA “strength” and often associates with antibody-mediated rejection (ABMR). HLA molecular mismatch studies have
mostly focused on the formation of de novo DSA. However, the potential for molecular mismatch calculations to aid in
post-transplant risk stratification for preformed DSA remains unexplored.

Methods: This retrospective study evaluated deceased-donor kidney transplant (KTX) recipients who underwent peritransplant desensitization for preformed DSA between 2012 and 2020. DSAs were defined by an MFI threshold of 2000.
Only donor and recipient pairs with available four-digit HLA typing were included. The HLA Eplet Registry mismatch
calculator was used to calculate an overall eplet mismatch score and an eplet mismatch score including only the
mismatched eplets potentially targeted by the preformed DSA. The primary outcome was biopsy-proven ABMR
according to the Banff 2022 classification.

Results: Of 168 patients, 63 had available data for inclusion: 32 were female (51%) and 43 (68%) had received an earlier
KTX. Median HLA (A/B/DR) mismatch was three (IQR 2-4). Patients had a median number of two DSA (IQR 1-3; 45
(71%) class I, 33 (52%) class II) and median follow-up was 5.2 years. Twenty-three individuals showed ABMR (37%),
which was diagnosed at a median of 1.3 years (IQR 0.1-3.8) after transplant. At time of transplant, overall eplet mismatch
was not different between the groups (64 [IQR 41-76] vs 56 [IQR 32-66]; p=0.186). However, when the eplet analysis
was restricted to the specific HLA targets of the preformed DSA(s), patients who subsequently showed ABMR had a
higher eplet score compared to those who did not (21 [IQR 14-27] vs 11 [6-19]; p=0.002; Figure 1). Notably, in a
multivariable logistic regression model DSA eplet mismatch score predicted ABMR after correction for MFI max (AUC
0.74; OR 1.08, 95% [CI 1.01, 1.15]; Figure 2).

Conclusion: Our data suggests that DSA-restricted eplet mismatch scores predict ABMR. This implies that the number
of epitope targets for pre-existing DSA contributes to the degree of DSA-mediated damage and translates into histologic
ABMR. Larger cohorts are needed to validate this finding.