Single Cord Blood Transplantation Versus HLA-Haploidentical Related Donor Transplantation Using Post-Transplant Cyclophosphamide in Patients with Hematological Malignancies

Introduction: Unrelated cord blood (UCB) and haploidentical related donor transplantation using post-transplant cyclophosphamide (PTCy-haplo) have become alternative options to treat patients with hematological malignancies without a human leukocyte antigen (HLA)-matched donor. Although a previous phase III study showed better outcomes after bone marrow transplantation (BMT) using PTCy-haplo than those after double unit UCB, a donor selection algorithm including peripheral blood stem cell transplant (PBSCT) using PTCy-haplo or high-risk patients remains to be clarified. With the aim to assess the relative efficacies of UCB and PTCy-haplo transplant, we performed a multi-center retrospective analysis to compare the clinical outcomes in patients with hematologic malignancies.

Methods: We included 517 patients aged 16 years to 70 years with hematological malignancies who received a first stem cell transplantation using a single UCB unit or PTCy-haplo between 2013 and 2019 in Kyoto Stem Cell Transplantation Group (KSCTG), which is a multi-center group of 18 transplantation centers in Japan. The primary endpoint was to compare overall survival (OS) between the UCB and PTCy-haplo groups. Secondary endpoints were relapse-free survival (RFS), GVHD- and relapse-free survival (GRFS), relapse, non-relapse mortality (NRM), neutrophil engraftment, platelet engraftment, grade II-IV acute graft-versus-host disease (GVHD), grade III-IV acute GVHD, chronic GVHD, and extensive chronic GVHD.

Results: Among 517 patients, 460 received single UCB transplant and 57 received PTCy-haplo related donor transplant (53 PBSCT and 4 BMT). The median age of the UCB and PTCy-haplo groups was 55 and 52 years, respectively (P= 0.770). Myeloablative conditioning (MAC) was used more often in the UCB group than in the PTCy-haplo group (P< 0.001). In the PTCy-haplo cohort, the total dose of Cy was 80mg/m ² (43.9%) or 100mg/m ² (56.1%), and the median duration of tacrolimus and mycophenolate mofetil use was 173 and 34 days, respectively. The median follow-up periods of survivors were 2.2 and 2.4 years, respectively. OS in the UCB group was comparable to that in the PTCy-haplo group (2-year OS; 53.7%, 53.6%, P= 0.71 and adjusted hazard ratio (aHR) 1.00, P= 0.99), with similar risks of relapse (aHR 0.91, P= 0.99) and NRM (aHR 0.93, P= 0.69). This result was consistent regardless of disease risk. Neutrophil and platelet engraftment were significantly lower (92.2% vs 94.7%, P= 0.042 and 79.8% vs 82.5%, P= 0.003) and the incidence of acute GVHD in the UCB group tended to be higher (gradeⅡ-Ⅳ; aHR 1.64, p= 0.055, grade III-IV; aHR 3.60, P= 0.073) in the UCB group than those in the PTCy-haplo group. However, the incidences of chronic or extensive chronic GVHD after UCB transplant were comparable (chronic; aHR 0.89, p=0.68, extensive chronic; aHR 0.69, P= 0.35) to those after PTCy-haplo transplant. In the subgroup analysis of disease-specific comparison, acute leukemia patients have a significantly lower risk of relapse with UCB transplant (UBC vs PTCy-haplo, HR 0.55, P= 0.030) and lymphoma patients tended to have good results after PTCy-haplo transplant in terms of higher OS (UCB vs PTCy-haplo, aHR 2.13, P= 0.162) and lower relapse (aHR 3.43, P= 0.100). In terms of Cy toxicity in the PTCy-haplo group, transplant with a reduced dose of Cy (80 mg/m ²) tended to show higher OS (2-year OS, 64.8% vs 45.0%) and significantly lower NRM (2-year NRM, 8.0% vs 31.2%) without an increase in gradeⅡ-IV acute GVHD compared to those after standard-dose Cy (100 mg/m ²).

Conclusion: Our findings suggest that UCB transplant gives outcomes comparable to PTCy-haplo transplant for patients without an HLA-matched sibling or unrelated donor.