Haploidentical Peripheral Blood Stem Cell Combined with Bone Marrow or Unrelated Cord Blood As Grafts for Acute Leukemia/Myelodysplastic Syndrome: An Open-Label, Multicenter, Randomized, Phase 3 Trial

Background Retrospective studies suggested that haploidentical transplant combined with unrelated cord blood (haplo-cord transplant) might improve survival for hematological malignancies. We aimed to prospectively compare outcomes of haplo-cord transplant with haplo-transplant in this population. Methods We did an open-label, randomized, phase 3 trial at seven transplant centers in China. Eligible patients were aged 18-65 years, had a diagnosis of hematologic malignancies, received haploidentical peripheral blood stem cell combined with bone marrow (haplo-PBSC+BM) or unrelated cord blood (haplo-PBSC+UCB) as grafts. Patients were randomly assigned (1:1) to receive haplo- PBSC+BM or haplo-PBSC+UCB transplant. The primary endpoint was 1-year disease-free survival (DFS). Secondary endpoints included 1-year overall survival (OS), engraftment, GVHD, relapse, TRM, and adverse events (AEs). All efficacy and safety endpoints were assessed in the intention-to-treat population. This trial is registered with ClinicalTrials.gov, NCT05290545 and complete. Single cell RNA sequence method was used to analyze the difference in immune reconstruction between the PBSC+UCB and PBSC+BM groups. Results Between February 15, 2022, and January 2, 2023, 357 patients were screened for eligibility, 314 of whom were randomly assigned to PBSC+UCB group (n=157) or PBSC+BM group (n=157). There were 180 (57.3%) male and 134 (42.7%) female, with median age of 38 (IQR, 27-52). Two hundred and sixty-nine (85.7%) patients were diagnosed with acute leukemia (AL) and 45 (14.3%) with myelodysplastic syndrome (MDS). With a median follow-up of 17.2 months (IQR 10.0-20.8) after random assignment, the 1-year DFS was 82.2% (95% CI 75.2-87·3) in the PBSC+UCB group and 65.6% (57.6-72.5) in the PBSC+BM group (HR 0·47 [95% CI 0·30-0·74]; p=0·001). The 1-year cumulative incidence of relapse was 10.2% (6.1-15.5) and 19.7% (95% CI 13.9-26.3) (HR 0·50 [95% CI 0·27-0.91]; p=0·024) in the two groups, respectively. The 1-year TRM was 7.6% (4.2-12.5) and 14.7% (95% CI 9.6-20.7) (HR 0·50 [95% CI 0·25-0.99]; p=0·049) in two groups, respectively. The 1-year OS was 86.6% (95% CI 80.2-91.1) and 75.2% (67.6-81.2) (HR 0·50 [95% CI 0·29-0.85]; p=0·01; Fig. 2D) in two groups, respectively. Within 100 days posttransplant, the most common grade 3-5 AEs in the two groups were infections (58 [36.9%] and 77 [49.0%], p=0.03), acute GVHD (49 [31.2%] and 61 [38.9%]), and gastrointestinal disorders (38 [24.2%] and 38 [24.2%]). Seven (4.5%) patients in the PBSC+UCB group and 17 (10.8%) patients in the PBSC+BM group died of transplant-related mortalities within 100 days posttransplant (p=0.034). As for immune reconstruction, CD62L+B cells were reconstructed rapidly in PBSC+UCB transplant, and the function of CD8+T cells changed obviously. Cell communication found that the effect of CD62L+B cells on CD8+T cells was enhanced, suggesting that this transplant model might enhance GVT by driving CD62L+B cells to act on CD8+T cells. Furthermore, our data suggested that the reconstituted CD62L+B cells in haplo-PBSC+UCB transplant were mainly derived from UCB grafts. C o nclusions Haplo-PBSC+UCB transplant achieves superior DFS compared with haplo-PBSC+BM transplant in AL and MDS with more satisfactory safety profile. This transplant mode can be a suitable option for this population.