#3823 CAN HLA MOLECULAR MISMATCH GUIDE CLINICAL DECISION IN PAEDIATRIC KIDNEY TRANSPLANT RECIPIENTS: A RETROSPECTIVE STUDY

Abstract

Abstract Background and Aims Optimizing graft survival and diminishing anti-HLA sensitization are essential in paediatric kidney transplant recipients (KTRs) facing multiple retransplantations. Improved pretransplant risk stratification and donor selection using in silico molecular HLA matching could enhance personalized immunosuppression and improve outcomes. HLA EMMA is a new in silico method to predict B cell epitope mismatches, and its clinical applicability needs confirmation in paediatric transplantation. We investigated the association of HLA B cell epitope (HLA EMMA) and predicted HLA T cell epitope (PIRCHE) mismatches with de novo donor specific antibodies (dnDSA), anti-HLA antibodies, rejection, and graft survival. Method We retrospectively analysed 49 consecutive paediatric KTRs aged 1 to 16 transplanted from 2009-2020 at a single Danish centre. The centre has a long tradition of steroid-free kidney transplantations in immunologically uncomplicated recipients, and 80% of the KTRs did not receive steroids. Donors and recipients were all high-resolution HLA typed, and HLA EMMA (v 1.00) and PIRCHE (v 3.3.60) predicted the molecular mismatches. KTRs were screened with mixed and single bead antigen kits after the transplantation. Logistic regression analyses were performed to explore the association of molecular mismatches and dnDSA, sensitization, graft loss, and rejection. Results The median age of the KTRs was 11 years (IQR 8), with a median follow-up time of 5 years (IQR 5). 11.36% developed dnDSA (60% class II, 40% class I and II) within a median of 3.91 years (IQR 2.28; range: 2.22-9.20). 57.45% had detectable anti-HLA antibodies during follow-up. Six KTRs lost their graft. None of the graft losses were related to alloimmunity, and 5-year death censored (one KTR died) graft survival was 88%. The cumulative incidence of rejection was 4% (95% CI: -1.5; 5.5) after one year and 10% (95% CI: 1.6; 18.7) after five years. The mean PIRCHE score was 368.47 (95% CI: 316.79; 420.15), the mean HLA EMMA class I mismatch was 20.33 (95% CI: 17.56; 23.09), and the mean HLA EMMA class II mismatch was 30.02 (95% CI: 24.99; 35.05). We found no association between PIRCHE or HLA EMMA with dnDSA, sensitization, graft loss, or rejection in the logistic regression models. We did see a tendency towards an increased odds ratio in PIRCHE predicting dnDSA (odds ratio: 1.31 (95% CI: 0.96; 1.80)) (Figure 1). Conclusion We did not find evidence to support the inclusion of the prediction of molecular mismatches in our clinical practice. The role of molecular mismatches in a clinical setting has yet to be established.