Preformed Donor-Specific HLA Antibodies in Living and Deceased Donor Transplantation: A Multicenter Study

Abstract

Background and objectives: The prognostic value of preformed donor-specific HLA antibodies (DSA), which are only detectable by sensitive methods, remains controversial for kidney transplantation.

Design, setting, participants, & measurements: The outcome of 4233 consecutive kidney transplants performed between 2012 and 2015 in 18 German transplant centers was evaluated. Most centers used a stepwise pretransplant antibody screening with bead array tests and differentiation of positive samples by single antigen assays. Using these screening results, DSA against HLA-A, -B, -C, -DRB1 and -DQB1 were determined. Data on clinical outcome and possible covariates were collected retrospectively.

Results: Pretransplant DSA were associated with lower overall graft survival, with a hazard ratio of 2.53 for living donation (95% confidence interval [95% CI], 1.49 to 4.29; P<0.001) and 1.59 for deceased donation (95% CI, 1.21 to 2.11; P=0.001). ABO-incompatible transplantation was associated with worse graft survival (hazard ratio, 2.09; 95% CI, 1.33 to 3.27; P=0.001) independent from DSA. There was no difference between DSA against class 1, class 2, or both. Stratification into DSA <3000 medium fluorescence intensity (MFI) and DSA ≥3000 MFI resulted in overlapping survival curves. Therefore, separate analyses were performed for 3-month and long-term graft survival. Although DSA <3000 MFI tended to be associated with both lower 3-month and long-term transplant survival in deceased donation, DSA ≥3000 MFI were only associated with worse long-term transplant survival in deceased donation. In living donation, only strong DSA were associated with reduced graft survival in the first 3 months, but both weak and strong DSA were associated with reduced long-term graft survival. A higher incidence of antibody-mediated rejection within 6 months was only associated with DSA ≥3000 MFI.

Conclusions: Preformed DSA were associated with an increased risk for graft loss in kidney transplantation, which was greater in living than in deceased donation. Even weak DSA <3000 MFI were associated with worse graft survival. This association was stronger in living than deceased donation.

Keywords: ABO-incompatible transplantation; Antibodies; Fluorescence; HLA-A Antigens; Incidence; Prognosis; Retrospective Studies; Tissue Donors; donor-specific HLA antibodies; graft survival; kidney transplantation; preformed HLA antibodies.

Copyright © 2019 by the American Society of Nephrology.

Figures

Graphical abstract

Figure 1.

The graft survival of patients with DSA was worse compared with patients without DSA, but not different according to DSA class. Overall graft survival is shown for (A) the total cohort according to the presence of pretransplant DSA, (B) living donors according to the presence of pretransplant DSA, (C) deceased donors according to the presence of pretransplant DSA, (D) the total cohort according to HLA class of pretransplant DSA, (E) living donors according to HLA class of pretransplant DSA, and (F) deceased donors according to HLA class of pretransplant DSA. Numbers of patients at risk are given below the Kaplan–Meier-curves. Log-rank test after exclusion of patients without DSA: (D) P=0.74; (E) P=0.89; (F) P=0.84.

Figure 2.

Both DSA <3000 MFI and DSA ≥3000 MFI were associated with impaired overall graft survival, while only DSA ≥3000 MFI were associated with early antibody-mediated rejection. Strength of pretransplant DSA are depicted in relation to (A) graft survival for the total cohort, (B) incidence of antibody mediated rejections during the first 6 months after transplantation (*P<0.001 compared with patients without DSA), (C) graft survival for living donors, (D) graft survival for deceased donors, (E) 3-month graft survival for living donors, (F) 3-month graft survival for deceased donors, (G) long-term graft survival starting at the fourth month for living donors, and (H) long-term graft survival starting at the fourth month for deceased donors. Graft survival has not been censored for patient death.

Figure 2.

Both DSA <3000 MFI and DSA ≥3000 MFI were associated with impaired overall graft survival, while only DSA ≥3000 MFI were associated with early antibody-mediated rejection. Strength of pretransplant DSA are depicted in relation to (A) graft survival for the total cohort, (B) incidence of antibody mediated rejections during the first 6 months after transplantation (*P<0.001 compared with patients without DSA), (C) graft survival for living donors, (D) graft survival for deceased donors, (E) 3-month graft survival for living donors, (F) 3-month graft survival for deceased donors, (G) long-term graft survival starting at the fourth month for living donors, and (H) long-term graft survival starting at the fourth month for deceased donors. Graft survival has not been censored for patient death.

Figure 3.

ABO-incompatibility was associated with decreased graft and patient survival after living donation. ABO-incompatible living kidney transplantations are compared with ABO-compatible living donations for (A) overall graft survival, (B) patient survival, and (C) death-censored graft survival. GS, graft survival; tx, transplantation.