Allogeneic hematopoietic cell transplantation with cord blood or haploidentical bone marrow extend access to transplantation for patients who lack a human leukocyte antigen (HLA)–matched donor, but results from the BMT CTN 1101 trial favor haploidentical transplant. The findings, published in Blood, revealed no difference in progression-free survival (PFS) in patients who received either donor source, but haploidentical transplant was associated with lower rates of non-relapse mortality and longer overall survival (OS).
“This is the first large-scale, randomized clinical trial comparing two different donor sources for allogeneic hematopoietic cell transplantation,” lead study author Ephraim Fuchs, MD, of John Hopkins Medicine, told ASH Clinical News. “Besides providing guidance in the selection of donor source for patients lacking HLA-matched siblings or HLA-matched unrelated donors, the trial demonstrates the feasibility of conducting phase III randomized trials to answer important questions in the field.”
With the BMT CTN 1101 trial, Dr. Fuchs and colleagues enrolled 368 adult patients with chemotherapy-sensitive Hodgkin or non-Hodgkin lymphoma or acute leukemia whose disease was in remission. Patients were randomized to undergo either two-unit umbilical cord blood transplantation (“double-cord”; n=186) or HLA-haploidentical transplantation (n=182).
Participants in the double-cord arm received conditioning with fludarabine, cyclophosphamide, and total body irradiation (TBI) 200 or 300 cGy. The higher dose of TBI was administered to patients who had not been treated with cytotoxic chemotherapy within a 3-month period of enrollment or an autologous hematopoietic cell transplantation within 24 months of enrollment. The conditioning regimen for those in the haploidentical arm consisted of cyclophosphamide, fludarabine, and 200 cGy TBI.
Prophylaxis for graft-versus-host disease (GVHD) consisted of cyclophosphamide and mycophenolate mofetil in the cord blood group; patients in the haploidentical transplant group also received tacrolimus as well as the other two treatments.
The primary endpoint was 2-year PFS. Secondary endpoints were the incidences of neutrophil and platelet recovery, acute and chronic GVHD, non-relapse mortality, relapse/progression, and OS.
Participants in each group had similar age, sex, disease status, performance status, and self-reported ethnic origin at randomization, the authors report. Median age was 58 years in the cord blood group and 60 years in the haploidentical group. The median follow-up time for surviving patients in each group was 24 months.
The rate of 2-year PFS was higher in the haploidentical group, compared with the cord blood group (41% vs. 35%; p=0.41). While multivariable analysis revealed a trend toward higher risk of death or disease progression in the cord blood transplant group after adjustment for patient and disease characteristics, this finding was not quite statistically significant, the investigators noted (hazard ratio = 1.30; 95% CI 0.99-1.70; p=0.06). There also was no statistically significant interaction between treatment effect and transplant center, they added.
However, the 2-year incidence of non-relapse mortality was higher with cord blood transplantation (18% vs. 11%; p=0.039), which translated to a lower rate of 2-year OS among the cord blood group (46% vs. 57%; p=0.037).
No differences were found between cord blood and haploidentical transplant in terms of the other prespecified secondary outcomes, including neutrophil and platelet recovery and acute and chronic GVHD:
- incidence of grade 2-4 acute GVHD within 180 days of transplant: 35% vs. 28%, respectively (p=0.142)
- median time to platelet recovery: 42 vs. 28 days (p=0.15)
- incidence of relapse/progression at 2 years: 47% vs. 48% (p=0.968)
“The results suggest haploidentical marrow should be considered over cord blood, but with several caveats,” commented Dr. Fuchs. First, he noted, the study findings may apply only to adults with acute leukemia in remission or lymphoma in chemotherapy-sensitive relapse.
“Second, we have not fully examined the potential for center effects, meaning that some centers by experience may achieve better outcomes with cord blood transplants,” he added. Also, haploidentical donors could not be identified for many patients in this study, suggesting other options, such as cord blood, need to be preserved.
When asked which patient-specific clinical factors might warrant consideration of a single treatment over another, Dr. Fuchs said that all patients should be screened for antibodies to donor HLA molecules, as this may dictate donor choice. “Also, some patients may have family members who are unwilling or unable, for example because of obesity, to donate bone marrow,” he added.
The authors report no relevant conflicts of interest.
Fuchs EJ, O’Donnell PV, Eapen M, et al. Double unrelated umbilical cord blood versus HLA-haploidentical bone marrow transplantation (BMT CTN 1101). Blood. 2020 August 31. [Epub ahead of print]
This trial is an important step along the way of identifying the ideal allogeneic transplant donor for patients who do not have a matched family member or matched unrelated donor available. Before this trial, the preferred donor source depended on the transplant center’s active protocols. In patients without a matched family member or matched unrelated donor, both cord blood and haploidentical donors were used; some centers focused more heavily on cord blood and others focused more heavily on haploidentical.
There was no significant difference in the primary endpoint of 2-year PFS whether patients received cord blood or haploidentical donor, but there was better OS with haploidentical donor transplant, primarily due to a lower rate of treatment-related mortality in this group – both of which were secondary endpoints. However, studies are not powered on the basis of secondary endpoints, and rarely are statistical adjustments made for the issue of multiple comparisons.
Other factors are at play in the selection of alternative donors, as well. In general, haploidentical donors are easier to procure, have lower infectious complications, and may have a cost benefit. I believe that most centers will in general favor haploidentical donors over cord blood for these reasons. Cost-effectiveness is an important endpoint to consider; most expect the study will show lower health-care costs with haploidentical donors.
It is important to note the monumental feat of completing this randomized trial. The investigators and the Blood and Marrow Transplant Clinical Trials Network should be lauded for this. However, the low rate of 2-year PFS with both cord blood and haploidentical donors is striking (35% for cord and 41% for haploidentical).
This study allowed for the inclusion of leukemia patients who were beyond first complete remission (CR), which is traditionally thought to increase the risk of post-transplant relapse. Here, 27% of patients with leukemia enrolled in the cord blood donor arm were beyond first CR, which could have decreased the potential benefit of antileukemic efficacy of the cord blood stem cells. Furthermore, transplant is a package consisting of not only donor stem cells, but also conditioning regimen and pretransplant therapy. This study only evaluated stem cell source in combination with reduced-intensity conditioning, not myeloablative conditioning.
The variety of diseases in the study also potentially limits the implications of these findings, though both donor sources appeared to be balanced regarding the various disease entities, which likely minimized the confounding nature of this covariate. In addition, the study was stopped early by the data and safety monitoring board due to slow accrual, making it underpowered to assess a 15% difference in PFS.
Additional questions to be explored include the potential impact of conditioning intensity, disease status at time of transplant, underlying disease, and cost-effectiveness of either approach.
Bart Scott, MD
Fred Hutchinson Cancer Research Center