Comparison of melphalan- And busulfan-based myeloablative conditioning in children undergoing allogeneic transplantation for acute myeloid leukemia or myelodysplasia.

Background The optimal conditioning regimen for alloHCT in children with myeloid malignancies remains undefined. Procedure We performed a retrospective review of children undergoing alloHCT for AML and MDS over a 10-year period (2008-2018) at our institution, comparing the outcomes of recipients of either a myeloablative busulfan- or reduced toxicity mel/thio-based conditioning regimen. Results A total of 49 patients underwent alloHCT for AML/MDS (mel/thio, N = 21; busulfan, N = 28). Mel/thio recipients were selected due to pretransplant comorbidities. Recipients of mel/thio were more likely to have t-AML, and less likely to have MRD <0.1% at the time of alloHCT (57.1% vs 82.1%). Graft failure was more common in busulfan recipients; engraftment kinetics were similar between groups. Sinusoidal obstructive syndrome was diagnosed in 21% of busulfan and no mel/thio recipients (P = .03). One patient in each group died from TRM. Relapse incidence was comparable (mel/thio-29% vs busulfan-32%); however, relapse occurred significantly later in recipients of mel/thio conditioning (median d + 396 vs d + 137; P = .01). As a result, there was a trend toward improved OS at 1 and 3 years in mel/thio recipients (95% vs 74%, P = .06; and 75% vs 50%, P = .11; respectively). Conclusion In our single institution, when compared to myeloablative busulfan-based conditioning, use of a mel/thio-based reduced toxicity regimen resulted in comparable outcomes, despite higher risk patient and disease characteristics. Mel/thio recipients had both more comorbidities and higher risk disease profile, which did not translate into higher rates of either TRM or relapse.