Failure to remove de novo donor-specific HLA antibodies is influenced by antibody properties and identifies kidney recipients with late antibody-mediated rejection destined to graft loss - a retrospective study.

Current research is focusing on identifying bioclinical parameters for risk stratification of renal allograft loss, largely due to antibody-mediated rejection (AMR). We retrospectively investigated graft outcome predictors in 24 unsensitized pediatric kidney recipients developing HLA de novo donor-specific antibodies (dnDSAs), and treated for late AMR with plasmapheresis + low-dose IVIG + Rituximab or high-dose IVIG + Rituximab. Renal function and DSA properties were assessed before and longitudinally post treatment. The estimated GFR (eGFR) decline after treatment was dependent on a negative % eGFR variation in the year preceding treatment (P = 0.021) but not on eGFR at treatment (P = 0.74). At a median follow-up of 36 months from AMR diagnosis, 10 patients lost their graft. Altered eGFR (P < 0.001) and presence of C3d-binding DSAs (P = 0.005) at treatment, and failure to remove DSAs (P = 0.01) were negatively associated with graft survival in the univariable analysis. Given the relevance of DSA removal for therapeutic success, we analyzed antibody properties dictating resistance to anti-humoral treatment. In the multivariable analysis, C3d-binding ability (P < 0.05), but not C1q-binding, and high mean fluorescence intensity (P < 0.05) were independent factors characterizing DSAs scarcely susceptible to removal. The poor prognosis of late AMR is related to deterioration of graft function prior to treatment and failure to remove C3d binding and/or high-MFI DSAs.