The disappointing contribution of anti-human leukocyte antigen donor-specific antibodies characteristics for predicting allograft loss.

Abstract

Pathogenicity of donor-specific antibodies (DSAs) can be assessed using the single-antigen flow beads (SAFB) assays through mean fluorescence intensity (MFI) with or without serum ethylenediaminetetraacetic acid (EDTA) treatment, measurement of C1q or C3d binding and/or their intragraft detection [graft-bound donor-specific antibody (gDSA)]. We aimed to investigate which of these markers best associates with antibody-mediated rejection (ABMR) and kidney allograft loss at the time of a for-cause biopsy. This retrospective, single-centre study included 77 kidney transplant recipients who underwent a for-cause biopsy between December 2004 and July 2013. All displayed serum DSAs were identified on the same day as the biopsy. Sera were tested in parallel with the classical SAFB assay with or without serum EDTA treatment, C1q- and C3d-binding assays. gDSAs were eluted from biopsy fragments and identified with SAFB. The median time between transplantation and biopsy was 25 months (range 0.5-251). The median follow-up was 36 months (range 0-140). ABMR was histologically proven in 40% of recipients. The sensitivity and specificity of C1q, C3d and gDSA assays for predicting ABMR were 68% and 61%, 52% and 70% and 64.5% and 56.5%, respectively. At the time of biopsy, only the DSA MFI after EDTA treatment and C3d positivity were associated with graft loss. In multivariate analyses, glomerular filtration rate, transplant glomerulopathy and C4d positivity were the only factors associated with graft loss. Our findings weaken the rationale for systematically implementing C1q, C3d or gDSA assays in this situation, because they do not independently predict ABMR and graft loss.