Programme Hospitalier Recherche Clinique, Institut Pasteur, Inserm, French Public Health Agency.

Although large retrospective studies have identified the presence of donor-specific antibodies (DSAs) to be a risk factor for rejection and impaired survival after liver transplantation, the long-term predicted pathogenic potential of individual DSAs after liver transplantation remains unclear. We investigated the incidence, prevalence and consequences of DSAs in maintenance liver transplant (LT) recipients. Two hundred sixtyseven LT recipients, who had undergone transplantation at least 6 months previously and had been screened for DSAs at least twice using single-antigen bead technology, were included and tested annually for the presence of DSAs. At a median of 51 months (min-max: 6-220) after an LT, 13% of patients had DSAs. At a median of 36.5 months (min-max: 2-45) after the first screening, 9% of patients have developed de novo DSAs. The sole predictive factor for the emergence of de novo DSAs was retransplantation (OR 3.75; 95% CI 1.28-11.05, p ¼ 0.025). Five out of 21 patients with de novo DSAs (23.8%) developed an antibody-mediated rejection. Fibrosis score was higher among patients with DSAs. In conclusion, monitoring for the development of DSAs in maintenance LT patients is useful in case of graft dysfunction and to identify patients with a high risk of developing liver fibrosis.

Summary The incidence and consequences of de novo donor‐specific anti‐HLA antibodies (DSAs) after liver transplantation (LT) are not well known. We investigated the incidence, risk factors, and complications associated with de novo DSAs in this setting. A total of 152 de novo liver‐transplant patients, without preformed anti‐HLA DSAs, were tested for anti‐HLA antibodies, with single‐antigen bead technology, before, at transplantation, at 1, 3, 6 and 12 months after transplantation, and thereafter annually and at each time they presented with increased liver‐enzyme levels until the last follow‐up, that is, 34 (1.5–77) months. Twenty‐one patients (14%) developed de novo DSAs. Of these, five patients had C1q‐binding DSAs (24%). Younger age, low exposure to calcineurin inhibitors, and noncompliance were predictive factors for de novo DSA formation. Nine of the 21 patients (43%) with de novo DSAs experienced an acute antibody‐mediated rejection (AMR). Positive C4d staining was more frequently observed in liver biopsies of patients with AMR (9/9 vs. 1/12, P < 0.0001). Eight patients received a B‐cell targeting therapy, and one patient received polyclonal antibodies. Only one patient required retransplantation. Patient‐ and graft‐survival rates did not differ between patients with and without DSAs. In conclusion, liver‐transplant patients with liver abnormalities should be screened for DSAs and AMR.