Int J Immunogenet. 2020;00:1–7. wileyonlinelibrary.com/journal/iji | 1 © 2020 John Wiley & Sons Ltd 1 | INTRODUCTION Human Leukocyte antigens (HLA) donor-specific antibodies (DSA) developed de novo after transplant, in particular against HLA class II, remain a major cause of chronic allograft dysfunction (Terasaki & Cai, 2008; Wiebe et al., 2012). HLA mismatching is a well-known risk factor for de novo DSA (dnDSA) (Opelz & Döhler, 2013), alongside with early T cell-mediated rejection, young age and inadequate im- munosuppression due to minimization or nonadherence. HLA mismatching traditionally is determined considering the donors HLA molecules not shared by the recipient. However, after the development of HLAMatchmaker by Rene Duquesnoy (Duquesnoy, 2002), several authors have been demonstrating the importance of matching at the epitope level (Dankers et al., 2004; Received: 4 August 2020 | Revised: 26 September 2020 | Accepted: 13 October 2020 DOI: 10.1111/iji.12519 ORIGINAL ARTICLE HLA class II eplet mismatch load improves prediction of dnDSA development after living donor kidney transplantation Sandra Tafulo 1,2 | Jorge Malheiro 2,3 | Sofia Santos 2,3 | Leonídio Dias 3 | Manuela Almeida 2,3 | La Salete Martins 2,3 | Sofia Pedroso 2,3 | Cecília Mendes 1 | Luísa Lobato 2,3 | António Castro-Henriques 3 1 Blood and Transplantation Center of Porto, Portuguese Institute for Blood and Transplantation, Porto, Portugal 2 Unit for Multidisciplinary Research in Biomedicine (UMIB), ICBAS, Porto, Portugal 3 Department of Nephrology, Hospital de Santo António, Centro Hospitalar Universitário do Porto, Porto, Portugal Correspondence Sandra Tafulo, Oporto Blood and Transplantation Center, Portuguese Institute for Blood and Transplantation, Rua de Bolama, 133, 4200-139 Porto, Portugal. Email: sandra.tafulo@ipst.min-saude.pt Funding information Disclosure of any funding received for this work. Abstract HLA donor-specific antibodies developed de novo after transplant remain a major cause of chronic allograft dysfunction. Our study main purpose was to determine whether HLA MM, assessed traditionally and by HLA total and AbVer eplet mismatch load (EptMM and EpvMM) assessed with HLAMatchMaker, had impact on dnDSA development after living donor kidney transplantation (LDKT). We retrospectively analysed a cohort of 96 LDKT between 2008 and 2017 performed in Hospital Santo António. Seven patients developed dnDSA-II and EpvMM and EptMM were greater in dnDSA-II group compared to the no dnDSA-II (18.0 ± 8.7 versus 9.9 ± 7.9, p = .041 and 41.3 ± 18.9 versus 23.1 ± 16.7, p = .018), which is not observed for AgMM (2.29 versus 1.56; p = .09). In a multivariate analysis, we found that preformed DSA (HR = 7.983; p = .023), living unrelated donors (HR = 8.052; p = .024) and retrans- plantation (HR = 14.393; p = .009) were predictors for dnDSA-II (AUC = 0.801; 0.622–0.981). HLA-II EpvMM (HR = 1.105; p = .028; AUC = 0.856) showed to be a superior predictor of dnDSA-II, when compared to AgMM (HR = 1.740; p = .113; AUC = 0.783), when adjusted for these clinical variables. Graft survival was signifi- cantly lower within dnDSA-II patient group (36% versus 88%, p < .001). HLA mo- lecular mismatch analysis is extremely important to minimize risk for HLA-II dnDSA development improving outcome and increasing chance of retransplant lowering allosensitization. KEYWORDS antibodies, histocompatibility, HLA, immune response, Immunology, matching, medicine, transplantation