Risk Factors for Graft Failure with Busulfan/Fludarabine-Based Conditioning in Children Undergoing Allogeneic Hematopoietic Cell Transplantation for Nonmalignant Disorders

Background: Allogeneic hematopoietic cell transplantation (aHCT) is curative for a number of pediatric nonmalignant disorders. Toxicities related to conditioning, failure of engraftment, and GVHD remain obstacles to successful transplantation for these patients. Since 2000, our institution has prospectively studied busulfan/fludarabine (Bu/Flu)-based conditioning regimens with a goal to reduce complications associated with myeloablative conditioning while maintaining adequate, stable donor-derived hematopoiesis. While overall survival has remained excellent, graft failure rates remain high. Therefore, we performed a retrospective analysis to identify risk factors for graft failure in patients receiving Bu/Flu-based conditioning. In addition, we report results for second aHCTs following graft failure.

Methods: Since 2000, our institution has performed 74 aHCTs with Bu/Flu-based conditioning for nonmalignant disorders. The most common diagnoses include Wiskott-Aldrich syndrome (n = 10), thalassemia (n = 9), Hurler syndrome (n = 7), hemophagocytic lymphohistiocytosis (HLH) (n = 6), sickle cell disease (n = 5), and aplastic anemia (n = 5). Three aHCTs were excluded due to relapse of the underlying disease (e.g. HLH) despite > 90% donor chimerism. Conditioning included a backbone of busulfan and fludarabine as previously described (Law J, et al, Biol Blood Marrow Transplant, 2012). Starting in 2012, four doses of clofarabine 10 mg/m2 (0.33 mg/kg for patients < 12 kg) were added for recipients of mismatched unrelated donors (MMUD). In addition, conditioning included either rabbit antithymocyte globulin (rATG) or alemtuzumab.

Results: Twelve graft failures (17%) occurred between day +18 and +445 (median +66). Of the variables analyzed, only donor chimerism for whole blood and the CD3 subset were significantly correlated with graft failure (CD14/15 and CD19 subsets were not). Cord blood aHCTs were excluded from conditioning analysis since these transplantations exclusively used rATG rather than alemtuzumab.

Among the 12 patients with aHCTs that resulted in graft failure, 7 were successfully rescued with a second transplantation, 1 declined a second transplantation, 1 is awaiting a second transplantation, and 3 died due to sequelae of graft failure despite a second transplantation.

Conclusions: We have previously shown that reduced toxicity conditioning with busulfan, fludarabine, and either rATG or alemtuzumab is well tolerated with low rates of transplant-related morbidity and mortality. In our cohort, low early donor chimerism was significantly correlated with increased graft failure. Close monitoring and early interventions may benefit patients at high risk for graft failure. Notably, there was a trend towards reduction in graft failure with alemtuzumab (9%) compared to rATG (26%), which did not reach statistical significance possibly due to insufficient power. While graft failure remains an issue with Bu/Flu-based conditioning, rescue with second transplantation is feasible with a reasonable chance for success.