Donor–recipient HLA mismatch is widely assessed with the number of HLA antigenic mismatches between donor and recipient. However, the epitope load, i.e. the number of epitope mismatches, seems to be more informative. We ought to evaluate the capacity of epitope load to predict development of Donor Specific Anti-HLA antibodies (DSA) and occurrence of antibody-mediated rejection (AMR) in a paediatric renal transplant population.

In this retrospective monocentric study, we included the 76 consecutive children who had undergone renal transplantation at Necker Children's Hospital between 2010 and 2015, without any pre-transplant DSA. The number of HLA antigenic mismatch, the epitope load (determined with the HLA-Matchmaker software), the occurrence of de novo DSA (MFI>500) and AMR were analysed. Non-adherence and immunising events were also assessed as potential confounding factors.

We observed a relationship between epitope load and development of de novo DSA. Epitope load was significantly greater in patients who developed DSA: 45 mean±16.9 vs. 34±18.7, P=0.038). A threshold >20 class I epitope mismatches was associated to a relative risk of developing class I DSA of 3.79 [(1.23–11.73)]. Similarly, a global epitope load >45 conferred a relative risk of 2.18 [95% CI (1.05–3.69)] of developing class I or II DSA. The model was not predictive for class II only DSA occurrence, maybe due to difficulties in determining the de novo character of some class II DSA. We did not show any direct relationship between epitope load and antibody-mediated rejection, whereas there was a strong link between de novo DSA and humoral rejection as expected (P<0.001).

Our study shows in a paediatric population, a relationship between epitope load and de novo DSA known to impair renal allograft survival. We have also found significant epitope load thresholds allowing identification of patients at the highest risk of DSA development, for example after immunosuppression minimization. Envisioned as a graft allocation tool, HLA epitope load, by avoiding highly immunogenic epitope-mismatches, could be a potent strategy to minimize post-transplantation immunisation, deleterious in paediatric patients.