Abstract 3487


Increasing evidence supports the notion that aplastic anemia (AA) defined as peripheral blood pancytopenia and bone marrow hypoplasia is an autoimmune disease mediated by T-cells attacking hematopoietic progenitor cells. Around 10% of acquired AA is associated with acute hepatitis, yet a causative virus has not been identified in most patients. Like other autoimmune disorders, adult AA is associated with human leukocyte antigen (HLA)-DR15. However, HLA-DR15 is not significantly associated in children with AA (Fuhrer et al. Pediatric Blood Cancer 2007 and Yoshida et al. Br J Haematol 2008). Human leukocyte antigen associations might differ between children and adults with AA. We therefore investigated the HLA types among patients with idiopathic aplastic anemia, hepatitis-associated aplastic anemia and fulminant hepatic failure who underwent liver transplantation.

Patients and methods:

We investigated data from 321 children (male, n = 202; female, n = 119; median age, 10 [range 0 – 18] years) with severe and very severe AA. All patients were enrolled in the AA-92 or AA-97 study from 1992 to 2010 and received immunosuppressive therapy (IST) using antithymocyte globulin and cyclosporine. Two hundred and sixty nine children had idiopathic AA and 52 had AA associated with hepatitis. The HLA alleles of 25 children with fulminant hepatic failure who underwent liver transplantation between 2006 and 2011 were characterized. Differences in all HLA alleles between the patients and healthy controls in the Japanese population were statistically analyzed at the phenotype level by χ2 and t tests using PRISM software.


The prevalence of HLA-B61 was significantly higher in idiopathic AA (21.9%, P < 0.0001), hepatitis associated AA (19.2%, P = 0.0089) and fulminant hepatic failure (20%, P = 0.0008) than in healthy Japanese controls (10.7%). No other HLA class I and II alleles including HLA-DR15 significantly differed between children with AA and healthy controls. Furthermore, HLA distribution did not significantly differ between patients who responded completely or partially to immunosuppressive therapy (IST) and non-responders.


We did not uncover a significant association between AA and HLA-DR15 in children. Associations between HLA and disease might differ between children and adults with AA. HLA-B61 was positively associated with idiopathic aplastic anemia, hepatitis-associated aplastic anemia and fulminant hepatic failure in pediatric patients. These findings suggest that a common autoimmune mechanism is responsible for both hepatic- and bone marrow-related pathologies in children.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.