Mismatched unrelated and haploidentical transplant with post-transplant cyclophosphamide-based graft versus host disease prophylaxis are effective for myelofibrosis

Abstract

Abstract Background Allogeneic hematopoietic cell transplant (HCT) remains the only curative option for myelofibrosis (MF); however, patients with MF historically have had poor survival outcomes due to high rates of graft failure and non-relapse mortality (NRM), especially when using alternative donors. We previously reported worse 2-year outcomes following HCT from 2008-2014 for MF with unrelated donors compared to related donors: non-relapse mortality (NRM) 53% (95% CI 36-78%) vs 21% (95% CI, 9-47%) and overall survival (OS) 38% (95% CI, 20-56%) vs 75% (95% CI, 46-90%) (Keyzner, BBMT 2016). High rates of graft failure with mismatched donors vs matched donors were a significant factor in worse outcomes (60% versus 13%; p=.04). In this study we analyzed the impact of haploidentical donors and the use of post-transplant cyclophosphamide (PTCy)-based graft versus host disease (GVHD) prophylaxis on HCT outcomes in MF in a more recent patient cohort. Method We retrospectively reviewed 65 consecutive patients with primary and secondary MF who underwent HCT at our center from 2015 to 2024. Patients were divided into two groups for comparative analyses: haploidentical and mismatched unrelated donors (100% received PTCy) and matched related/unrelated donors. Outcomes analyzed included 2-year OS, progression-free survival (PFS), GVHD-free relapse-free survival (GRFS), NRM, and engraftment kinetics. For OS, PFS, and GRFS analyses, log-rank p-values were reported based on Kaplan–Meier methodology. For NRM, neutrophil engraftment, and platelet engraftment analyses, the cumulative incidence function (CIF) methodology was applied, and Gray's test p-values were reported. Results The median age at HCT was 63 years (33-73), the median Karnofsky Performance Status (KPS) was 90 (70-100), and 52% were female. High risk characteristics were common: spleen size >22cm (n=18, 28%), high risk by MIPSS-70 scoring (n=55, 85%), high/very high risk by the clinical-molecular myelofibrosis transplant scoring system (MTSS) (n=20, 31%). Patients were either in chronic phase (n=50, 77%), accelerated phase (n=6, 9%), or at blast phase (n=9, 14%). The conditioning regimens were equally divided between Thiotepa, Busulfan, and Fludarabine (TBF) and Fludarabine/Melphalan (Flu/Mel); all but three patients received reduced intensity conditioning. PTCy was used for GVHD prophylaxis for all mismatched recipients, and 34 (52%) patients overall. The mismatched group (n=25) included haploidentical (n=20, 31%) and mismatched unrelated donors (n=5, 8%). The matched group (n=40) included related donors (n=13, 20%) and unrelated donors (n=27, 41%). The 2-year PFS and OS for all patients were 66% (54-77%) and 73% (62-84%), respectively. The 2-year PFS and OS did not significantly differ between groups; mismatched vs matched, PFS: 75% (58-92%) vs 65% (44-75%); OS: 83% (68-98%) vs 67% (52-82%). The 2-year NRM was 24% (16-36%) for the entire cohort, with 8% for mismatched and 30% for matched groups. 2-year GRFS was 40% (28-52%) overall but was markedly better for mismatched vs matched recipients: 67% (48-86%) vs 24% (11-37%); this difference appeared to be driven by greater utilization of PTCy in the mismatched group. 2-year GRFS was more than 2-fold higher for recipients of PTCy prophylaxis: 58% (41-75%) vs 23% (8-37%). There were no meaningful differences in 2-year PFS, OS, GRFS, or engraftment by conditioning regimen (TBF vs Flu/Mel). There were no graft failures in either group, and the proportion of engraftment by standard time points was similar for mismatched vs matched groups: day 28 neutrophil engraftment (80% vs 93%), day 100 platelet engraftment (75% vs 78%). A few patients required stem cell boosts for poor graft function in each group: mismatched (n=3, 5%) vs matched (n=2, 3%). Conclusion The use of mismatched donors, including haploidentical donors, was associated with marked improvements in engraftment, PFS, and OS for HCT in MF in the 2015–2024 era compared to outcomes from 2004–2014. Both TBF and Flu/Mel appear to be equally effective as conditioning regimens for MF regarding engraftment and survival outcomes. The use of PTCy for GVHD prophylaxis appears to be the primary contributor to improved outcomes, and significantly, did not affect rates of engraftment. These data show that the lack of a well-matched donor should not be considered a barrier to curative HCT for MF.