Donor-Specific HLA Antibodies Are Associated with Graft Failure and Delayed Hematologic Recovery after Unrelated Donor Hematopoietic Cell Transplantation

Abstract

Graft failure (GF) is one of the major concerns after allogeneic hematopoietic cell transplantation (allo-HCT) and remains a significant cause of morbidity and mortality. Although earlier reports have associated the presence of donor-specific HLA antibodies (DSAs) with increased risk of GF after unrelated donor allo-HCT, recent studies have failed to confirm this association. We sought to validate the presence of DSAs as a risk factor for GF and hematologic recovery in the unrelated donor allo-HCT setting. We retrospectively evaluated 303 consecutive patients who underwent their first unrelated donor allo-HCT at our institution from January 2008 to December 2017. DSA evaluation was performed using 2 Single Antigen Beads (SAB) assays, DSA titration with 1:2, 1:8, and 1:32 dilutions, C1q-binding assay, and absorption/elution protocol to assess possible false-positive DSA reactivity. The primary endpoints were neutrophil and platelet recovery and GF, whereas the secondary endpoint was overall survival. Multivariable analyses were performed using Fine-Gray competing risks regression or Cox proportional hazards regression models. The median patient age was 14 years (range, 0-61 years), 56.1% were male, and 52.5% were transplanted for nonmalignant diseases. Eleven patients (3.63%) were DSA-positive. Of them, 10 had preexisting DSAs, and one showed post-transplant de novo DSA. Nine patients had 1 DSA, 1 had 2 DSAs, and 1 had 3 DSAs, with a median MFI of 4334 (range, 588-20,456) and 3581 (range, 227-12,266) in LABScreen and LIFECODES SAB assays, respectively. Overall, 21 patients experienced GF. Of them, 12 had primary graft rejection, 8 had secondary graft rejection, and 1 had primary poor graft function. The cumulative incidences of GF at 28, 100, and 365 days were 4.0% (95% CI, 2.2%-6.6%), 6.6% (95% CI, 4.2%-9.8%), and 6.9% (95% CI, 4.4%-10.2%), respectively. In the multivariable analyses, DSA-positive patients had significantly delayed neutrophil (subdistribution hazard ratio [SHR] = 0.48; 95% CI, 0.29-0.81; P = .006) and platelet recovery (SHR = 0.51; 95% CI, 0.35-0.74; P = .0003) than patients without DSAs. In addition, only DSAs were significant predictors of primary GF at 28 days (SHR = 2.78; 95% CI, 1.65-4.68; P = .0001). The Fine-Gray regression also demonstrated that the presence of DSAs was strongly associated with a higher incidence of overall GF (SHR = 7.60; 95%CI, 2.61-22.14; P = .0002). DSA-positive patients with GF had significantly higher median MFI values than DSA-positive patients who achieved engraftment in LIFECODES SAB assay using neat serum (10,334 vs. 1250; P = .006) and in LABScreen SAB at 1:32 dilution (1627 vs. 61; P = .006). All 3 patients with C1q-positive DSAs failed to engraft. DSAs were not predictive of inferior survival (hazard ratio = 0.50; 95% CI, 0.20-1.26, P = .14). Our results validate the presence of DSAs as a significant risk factor for GF and poor hematologic recovery after unrelated donor allo-HCT. Thus, careful pre-transplant DSA evaluation may optimize unrelated donor selection and improve allo-HCT outcomes.