Risk Factors for Donor-Type Aplasia After Bone Marrow Transplant in Children with Acquired Bone Marrow Failure Syndrome

Abstract 3474

Graft rejection is a major cause of aplasia after bone marrow transplant (BMT) in patients with bone marrow failure syndrome (BMFS). However, we have experienced several patients with acquired BMFS who developed persistent aplasia with full donor chimerism after BMT (donor-type aplasia).

To clarify risk factors for donor-type aplasia, we reviewed the clinical data of 58 children (male, n = 36; female, n = 22) with acquired BMFS who received allogeneic BMT at Nagoya University Hospital and Japanese Red Cross Nagoya First Hospital between January 1997 and December 2010. The median age at diagnosis was 10 (range, 1 – 20) years. The median interval between diagnosis and BMT was 30 (range, 0.70 – 105) months. Twenty-five of the 58 patients received > 20 units of red blood cell (RBC) transfusions. We retrospectively reviewed BM smears and samples of trephine biopsies according to the 2008 WHO classification. Of the 58 patients, 26 and 32 were diagnosed with aplastic anemia (AA) and refractory cytopenia of childhood (RCC), respectively. Twenty, five and 33 patients received BMT from HLA-matched and mismatched related donors and unrelated donors, respectively. Conditioning regimens consisting of cyclophosphamide (CY), antithymocyte globulin (ATG) with or without TBI and of fludarabine, CY, ATG and TBI (fludarabine regimen) were administered to 38 and 18 patients, respectively. Marrow grafts (median dose, 4.2 ×10⁸ nucleated cells/kg) were not T cell–depleted. Prophylaxis against graft versus host disease (GVHD) consisted of cyclosporine/methotrexate (MTX) for 26 patients and tacrolimus/MTX for 32 others. Donor-type aplasia was defined as hypocellular BM with persistent cytopenia (hemoglobin < 8 g/dl; neutrophils < 1 ×10⁹/L or platelets < 50 ×10⁹/L) for > 6 months after achieving engraftment with full donor chimerism. We evaluated the following variables to determine risk factors for donor type aplasia; age at BMT, sex, BM morphology at BMT, year of BMT, period between diagnosis and BMT, units of RBC transfusion, history of immunosuppressive therapy (IST) before BMT, HLA disparity, BM cell dose, conditioning regimen, GVHD and viral infection.

Eleven (19%) of the 58 children developed donor-type aplasia. All received BMT after 2003 and had the BM morphology of RCC. Three and five of them received BMT from HLA-mismatched and unrelated donors, respectively. Fludarabine regimens were administered to eight patients. Seven patients received ganciclovir to treat CMV reactivation and two received rituximab to treat Epstein-Barr virus reactivation before aplasia developed. The onset of donor-type aplasia ranged from 46 – 710 days after BMT. Donor-type aplasia continued for over three years in five patients. Eight patients required transfusions with either RBC or platelets or granulocyte stimulating factor (G-CSF). Donor-type aplasia improved in five patients within 203 – 1789 days. All patients with donor-type aplasia remained alive for 20 – 103 months after developing donor-type aplasia. Univariate analysis revealed BM morphology, year of BMT and the fludarabine regimen as risk factors for donor-type aplasia and multivariate analysis confirmed these findings. The cumulative incidence (CI) of donor-type aplasia was significantly higher in children who received the fludarabine regimen (45%) or had RCC (35%) than in those who received a non-fludarabine regimen (5%) or had AA (0%) (p = 0.001, respectively). The CI of donor-type aplasia was significantly higher in children with RCC who received the fludarabine regimen (62%) than in those who did not (15%). In contrast, none of the children with AA developed donor-type aplasia irrespective of the fludarabine regimen.

The fludarabine regimen was identified as a risk factor for donor-type aplasia in children with RCC. Children with RCC should not be conditioned with the regimens used for children with AA.