Unrelated Donor Hematopoietic Stem Cell Transplantation (HSCT) in Children with Acute Leukemia: Risks and Benefits of Umbilical Cord Blood (CB) Versus HLA A, B, C, DRB1 Allele-Matched Bone Marrow (BM).

CB has proven to be an acceptable alternative to BM for transplantation. However, it is unknown how CB mismatched at HLA 0–2 loci compares to HLA matched BM. Therefore, we compared results observed in 503 recipients of CB with those in 116 recipients of allele-matched BM (at HLA A, B, C, DRB1). Of CB recipients, 35 were matched at HLA A, B (antigen-level) and DRB1 (allele-level), 201 were mismatched at 1-locus and 267 were mismatched at 2-loci. All patients (aged <16 years) had acute leukemia and were transplanted between 1995–2003. Median follow up was 45 and 59 months for CB and BM recipients, respectively. Compared to recipients of allele-matched BM, CB recipients were younger, non-Caucasian, more likely to have advanced disease at HSCT, less likley to receive radiation and somewhat more likely to receive cyclosporine. Leukemia-free survival (LFS) rates was superior in recipients of HLA matched CB (p=0.040). Notably LFS and overall survival (OS) at 5-years were comparable between those receiving allele-matched BM and 1 or 2-loci mismatched CB. These results in LFS and OS are in part explained by differences in risks of transplant-related mortality (TRM) and relapse between patient populations. Compared to allele-matched BM transplants, TRM was similar in recipients of matched and 1-locus mismatched/high cell dose (>0.3 x 10⁸/kg) CB and higher in recipients of 1-locus mismatched/low cell dose (≤0.3 x 10⁸/kg) and 2-loci mismatched CB (any cell dose) (p=0.005, p<0.001, respectively). Conversely, relapse rates were lower in recipients of CB mismatched at 1 or 2-loci (p=0.037, p=0.003, respectively). As previously observed, probability of neutrophil recovery (≥500/ul) at day 42 depended on graft type, HLA-match and cell dose (98% with BM; 85% with matched CB; 79% with 1 or 2-loci mismatched CB/high cell dose; 64% with 1 or 2-loci mismatched CB/low cell dose). Compared to allele-matched BM, risks of grade 2–4 and and grade 3–4 acute GVHD were lower after matched CB (relative risk [RR] 0.45, p=0.035, RR 0.51, p=0.035, respectively) and similar after mismatched CB. Risks of chronic GVHD were lower after matched or mismatched CB transplants (RR 0.66, p=0.036). The Table below shows the 5-year probabilities of TRM, relapse, LFS and OS by graft type. These results are the first to support the use of HLA matched or mismatched CB grafts with an adequate cell dose as first line transplant treatment regardless of the availability of a HLA allele-matched BM donor in the setting of actue leukemia in those <16 years.