Impact of Graft Source [Cord Blood Supported By CD34+ Selected Cells (Haplo-Cord) Vs. Matched Donor] on Transplant Outcomes in Myelofibrosis

Introduction: Allogeneic hematopoietic cell transplantation (allo-HCT) is curative for patients with myelofibrosis (MF). Here, we evaluated the impact of donor type [cord blood transplant supported by CD34+ cells (haplo-cord) vs adult matched donor] on outcomes in MF.

Methods: All adult patients with primary or secondary (post-ET or post-PV) MF, including those progressing to acute leukemia (AML) who underwent their first allo-HCT at NewYork-Presbyterian Hospital/Weill Cornell Medicine between 01/2015 and 06/2022 were included.

Results: 40 patients were transplanted. Median age was 63 years (range 21-77). 23% had a Karnofsky performance scale (KPS) of <80. Two had undergone splenectomy prior to HCT. The median spleen size was 18cm (10-28). Median number of pre-HCT molecular mutations was 2 (range 0-7). 16 (40%) had MF-3 (Table 1).

24 (60%) underwent a matched sibling donor (MSD) (n=9) or unrelated donor (MUD) (n=15) HCT, and 16 (40%) underwent a cord-based (CB) HCT. 39 received Fludarabine/Melphalan +/- low dose total body irradiation (TBI), one underwent myeloablative conditioning with TBI/etoposide based on extensive extramedullary disease. The median time to neutrophil engraftment in the MSD/MUD group was 14 days (IQR 14-18) vs 30 days in the CB group (IQR 15-36) and (p=0.1). The median time to platelet engraftment was 22 days (IQR 17-38) in MSD/MUD vs 49 days (IQR 26-60) in the CB group (P=0.07). The CI of absolute neutrophil count (ANC) and platelet engraftment by day 60 for MSD/MUD vs CB was 92% (95% CI 65-98%) vs 80% (95% CI 45-94%) (p=0.09) and 75% (95% CI 51%- 88%) vs 59% (95% CI 27-81%) (p=0.36), respectively. One patient in the MSD/MUD died without neutrophil engraftment compared to 3 in CB group. 5 patients in each group died without platelet engraftment.

Median follow-up for all alive patients was 53 months. Overall survival (OS) for the cohort was 62% at 1 year (95% CI 49-79) and 34% at 3 years (95% CI 21-55). OS at 3 years was similar between MSD/MUD group and CB (32% vs 37%, P=0.9) (figure 1). Graft versus host disease relapse free survival (GRFS) at 3 years was not significantly different in MSD/MUD group, compared to CB group (15% vs 33%, P=0.64). Cumulative incidence of relapse at 3 years in MSD/MUD group was 28% (95% CI 10-49), and 12% (95% CI 1.8-34) in CB group (p=0.356). Non-relapse mortality (NRM) at 1 year was 33% (95% CI 15-52) in MSD/MUD group and 38% (95% CI 15-61) in CB group. NRM at 3 years was 54% (95% CI 29-73) in MSD/MUD group and 48% (95% CI 19-72) on CB group (P=0.95).

Discussion: In our small cohort, we showed that OS after haplo-cord transplant was comparable to adult matched donors' grafts (MSD or MUD). While the haplo-cord platforms have been shown to decrease the time to engraftment (11 days for neutrophils and 22 days for platelets) compared with double cord transplants or haploidentical transplants (van Besien K et al . Haematologica 2016, van Besien K et al. Blood Adv 2019), in patients with MF, we saw a trend towards delayed engraftment compared to patients transplanted with adult matched donors' grafts (MSD or MUD). This suggests that the CD34 selected cells may not be able to provide an adequate myeloid bridge and overcome the low cellular content in the CB for adult patients with extensive marrow fibrosis. Larger studies are needed to compare best alternative graft (CB-based versus haploidentical or mismatched unrelated donors) for patients with MF that lack a matched donor.