Single Cord Blood Transplantation Strongly Inhibits Leukemia Relapse and Is Associated with Significantly Better Leukemia-Free and Overall Survival Than HLA-Matched Related Donor in Allo-HSCT for Patients with Acute Myeloid Leukemia in Non-Remission

Background: The outcomes of allogeneic hematopoietic cell transplantation (allo-HSCT) in patients with non-remission acute myeloid leukemia (AML) remain unsatisfactory. Although allo-HSCT from HLA-matched related donor (MRD) have been the first choice of graft when available, superiority of MRD over other grafts is controversial for patients with non-remission AML. Cord blood (CB) is an alternative graft and several studies suggested that CB may have great graft-versus-leukemia effects. So far, few studies have directly compared the outcomes of CB and MRD for patients with AML in non-remission. Methods: We report on patients diagnosed with AML who received first allo-HSCT using CB or MRD in non-remission status at our center between 2008 and 2023. Patients who were younger than 66 years old and were in ECOG PS 0-2 were included, and those with active infections during the conditioning were excluded. The primary endpoint was leukemia-free survival (LFS), and secondary endpoints were overall survival (OS) and cumulative incidence of relapse, non-relapse mortality (NRM), and neutrophil engraftment. The probability of LFS and OS were estimated using the Kaplan-Meier method, and between-group differences were analyzed using the log-rank test. Multivariate analyses were performed using the Cox proportional hazards model, and hazard ratios (HRs) together with their corresponding 95% confidence intervals (CIs) were calculated. Relapse, NRM, and neutrophil engraftment were evaluated by accommodating the competing risks, and the Fine-Gray model was used for univariate and multivariate analyses. Results: A total of 335 patients were enrolled and donors were CB in 309 (92.2%) patients and MRD in 26 (7.8%) patients. Median age was 55 years (range 19 to 65 years) and 41.8% of patients were female. Disease status before HSCT was primary induction failure in 29.9%, relapse in 34.3% and blast control in 35.8%. Cytogenetic risk was favorable in 5.7%, intermediate in 56.1%, adverse in 35.8%, and unevaluable in 8 (2.4%) patients. Conditioning regimens were fludarabine-busulfan +/- melphalan or TBI-based myeloablative conditioning in 89.3% and reduced intensity conditioning in 10.7%. Patients in the CB group were older than the MRD group (median age: 56 vs. 48.5 years). The CB group also included a greater proportion of males (59.9% vs. 38.5%). There were no significant differences in cytogenetic risk, disease status before HSCT and PS between the groups. Median follow-up of survivors after HSCT were 1938 days in CB group and 2370 days in MRD group. At 5 years, the probability of LFS was statistically significantly better in the CB group than the MRD group (42.0% [95% CI, 36.1-47.7%] vs 16.8% [95% CI, 5.3-33.7%], p<0.01). The 5-year OS was 45.5% (95% CI, 39.5-51.4%) in the CB group and 32.6% (95% CI, 15.6-50.8%) in the MRD group (p=0.40). Extremely lower incidence of relapse was observed in CB recipients (24.6% vs 51.7% at 5 years, p<0.01). The 5-year cumulative incidence of NRM was 33.5% in CB and 31.5% in MRD (p=0.80). Although there was a trend of earlier neutrophil engraftment in the MRD group compared to CB, no difference in the neutrophil engraftment rate between the groups (93.5% vs 88.5%, p=0.19). In multivariate analyses, in comparison with CB recipients, MRD recipients had lower LFS (HR=2.51, 95% CI: 1.57-4.03, p<0.01), lower OS (HR=1.74, 95% CI: 1.03-2.92, p=0.037) and higher incidence of relapse (HR=3.52, 95% CI: 1.79-6.93, p<0.01). Other factors affecting LFS and OS in multivariate analyses were age >50 years at HSCT (LFS: HR=1.89, OS: HR=2.22), favorable cytogenetic risk (LFS: HR=0.19, OS: HR=0.22, both compared to intermediate risk), adverse cytogenetic risk (LFS: HR=1.59, OS: HR=1.51, both compared to intermediate risk), and PS 2 (LFS: HR=2.21, OS: HR=2.04, both compared to PS 0-1). Conclusions: In allo-HSCT for patients with AML in non-remission status, LFS was better in the CB group compared to the MRD group. In multivariate analysis, CB was also a statistically significant factor affecting better OS compared to MRD. This result was considered to be due to greater reduction in post-transplant relapse in the CB group and the fact that there was no significant difference in NRM between the two groups, possibly owing to practical advances in CB transplantation. Our data might support selecting CB donor in allo-HSCT for non-remission AML even when MRD is available.