Graft failure is a majorcause of severe aplasia after bone marrow transplant (BMT) in patients with acquired bone marrow failure (aBMF). However, we have encountered several patients with aBMF who showed persistent aplasia with full donor chimerism after BMT (donor-type aplasia). The mechanism and risk factors of donor-type aplasia are not clarified.

To clarify risk factors for donor-type aplasia, we reviewed clinical data for 50 patients with aBMF who received allogeneic BMT at our hospital from August 1997 to July 2016. Out of the 50 patients, there were 32 males and 18 females. The median age at BMT was 11 years (range, 1-19 years). We retrospectively reviewed BM smears and histopathological samples at the time of BMT according to the 2008 WHO classification. Out of the 50 patients, 26 were classified as aplastic anemia (AA) and 24 as refractory cytopenia of childhood (RCC). Twenty patients received BMT from related donors including 5 human leukocyte antigen (HLA)-mismatched donors and 30 patients from unrelated donors including 15 HLA-mismatched donors. Conditioning regimens including cyclophosphamide (CY) and antithymocyte globulin (ATG) with or without low dose total body irradiation (TBI) were used for 31, fludarabine (FLU) and half dose of CY with low dose TBI were used for 15, and FLU and melphalan (MEL) with low dose TBI were used for 4 patients. Donor-type aplasia was defined as hypocellular BM with persistent cytopenia (hemoglobin <80 g/L, neutrophils <1x109/L, or platelets <50x109/L) for >6 months after achieving engraftment with full donor chimerism. To reveal risk factors of donor type aplasia, we analyzed variables as follows; age at BMT, gender, BM morphology at BMT, year performed BMT, period from diagnosis to BMT, times of red blood cell (RBC) transfusion, HLA disparity, transfused BM cell dose, conditioning regimen, graft versus host disease (GVHD), and viral infection.

Out of the 50 patients, 10 patients developed late graft failure. Eight out of them developed donor-type aplasia. Out of the 8 patients, there were 5 males and 3 females. The median age at BMT was 10 years (range, 5-13 years). Bone marrow findings of all patients showed RCC at the time of BMT. Two of them received BMT from HLA-mismatched donors and 3 from unrelated donors. FLU/CY regimen was used for 6 patients, while CY/ATG regimen was used for 2 patients. All patients received low dose TBI. Three patients received BMT with <3x108/kg cell dose. Acute GVHD were observed in 2 patients and one of them developed chronic GVHD. Five patients showed cytomegalovirus antigenemia in their blood and received ganciclovir. Two patients developed Epstein-Barr virus reactivation and received rituximab. The onset of donor-type aplasia ranged from day 46 to day 710 after BMT. Five patients needed RBC or platelet transfusion, or granulocyte-colony stimulating factor (G-CSF). Donor-type aplasia continued for more than 3 years in 5 patients. Donor-type aplasia was improved spontaneously in 6 patients, ranging from 460 days to 3539 days after BMT. Only one patient underwent haploidentical stem cell transplant (SCT) from her father because of the development of severe pneumonia. She achieved engraftment after second SCT. All patients who developed donor type aplasia are alive. Cumulative incidence (CI) of donor-type aplasia was significantly higher in patients who had RCC (32%) than in patients who had AA (4%) (p = 0.013), and CI of donor-type aplasia was also significantly higher in patients who received FLU/CY regimen (44%) than in patients who received non-FLU/CY regimen (6%) (p = 0.003). Multivariate analysis confirmed that RCC (hazard ratio 8.8, 95% CI 1.1-68, p = 0.037) and FLU/CY regimen (hazard ratio 7.7, 95% CI 1.5-38, p = 0.013) were significant risk factors for donor-type aplasia.

In conclusion, FLU/CY regimen is not suitable for children with RCC as a conditioning regimen of BMT. It is important to evaluate BM morphology to decide conditioning regimens for children with aBMF. The establishment of the optimal conditioning regimens for children with aBMF on the basis of the BM morphology is essential.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution