Umbilical cord blood reduced relapse but increased non-relapse mortality compared to matched unrelated donor transplantation in pediatric acute myeloid leukemia with active disease: A CIBMTR 2008-2017 analysis of donor source and residual disease.

Abstract

Umbilical cord blood (UCB) and matched unrelated donors (MUD) are common alternative donor options in children with high-risk acute myeloid leukemia (AML). Emerging evidence suggests an augmented graft-versus-leukemia (GVL) effect of UCB, but uncertainties persist due to the heterogeneity of the hematopoietic cell transplantation (HCT) characteristics in the previous studies. We reviewed 1148 patients aged ≤18 years with AML, who underwent the first HCT between 2008-2017, using a publicly available dataset from the Center for International Blood and Marrow Transplantation Research (CIBMTR) registry data. Multivariable analyses evaluated predictors of DFS and other clinical outcomes, factoring in graft source, conditioning regimen, patient age, cytogenetic risk, and HCT year (significance at p <0.01). Residual disease status was assessed both as a covariate and as a stratifying factor. Additionally, the differential effects of conditioning regimens were analyzed specifically within the UCB cohort. UCB was used most frequently (33.8%) followed by MUD (29.1%), both of which had comparable DFS and overall survival. In patients with minimal residual disease or not in remission prior to HCT, human-leukocyte antigen (HLA) ≤5/8 matched UCB was associated with lower relapse rates than MUD (hazard risk [HR]: 0.25 and 0.29, p = 0.005 and 0.006, respectively) but with increased nonrelapse mortality (HR: 32.8 and 7.5, p = 0.001 and 0.012, respectively). Conditioning regimens varies by graft type; total body irradiation (TBI)-based regimens, primarily combined with cyclophosphamide and fludarabine, were more common in the UCB cohort (45% in UCB vs. 19% in the other grafts, p <0.001). Within the 388 patients received UCB, multivariable analysis demonstrated comparable DFS and OS across variable busulfan- and TBI-based regimens, with no trend of superiority for either approach. In conclusion, highly HLA-mismatched UCB reduced relapse in pediatric AML with higher disease burden but increased non-relapse mortality, resulting in similar DFS to MUD. Improved supportive care and toxicity mitigation may improve the outcomes of UCB transplant. Overall, UCB should be considered a viable alternative graft source with equally favorable outcomes to MUD. Further research is warranted to refine conditioning regimen, including TBI- and busulfan-based strategies, mitigate toxicity, and improve supportive care to optimize UCB HCT outcomes.