Outcomes for allogeneic transplant in myelofibrosis remain suboptimal in the modern era despite increased use of post-transplant cyclophosphamide and pre-transplant JAK inhibition.

Abstract Allogeneic transplant remains the only curative therapy for patients with primary or secondary myelofibrosis (MF), but outcomes remain suboptimal. Disease-specific characteristics such as splenomegaly, an unfavorable marrow environment, transfusion burden and advanced age can result in delayed engraftment and higher non-relapse mortality (NRM). We hypothesized that more recent transplant trends, including increased use of JAK inhibition pre-transplant and post-transplant cyclophosphamide (PTCy) for GVHD prophylaxis, could result in better transplant outcomes in the modern era. To this end, we analyzed 42 consecutive patients with primary or secondary MF transplanted at a single institution from 2012 – 2024. Donor type was MRD, MUD, HAPLO and cord blood in 9 (21%), 15 (36%), 17 (40%) and 1 (2%). Conditioning intensity was non-myeloablative, reduced intensity and myeloablative in 9 (21%), 31 (74%) and 2 (5%). Median age was 64 years (39, 75), male 52% and HCT-CI ≥3 in 86%. MF was primary in 22 (52%), Jak2-mutated in 26 (62%) and unfavorable cytogenetics in 14 (33%). Pre-transplant Jak inhibition was used in 30 (71%). Median follow-up was 107 months (7, 161). Graft rejection was seen in 3 patients (7%); all of which received HAPLO transplant (18%). The incidence of grade 2-4 and 3-4 acute GVHD were 38% and 24%; all-grade and moderate-severe chronic GVHD were 17% and 15%. Median days-to-ANC and -PLT recovery was 20 and 43 days respectively. One-year NRM and overall survival (OS) were 31% and 67%. Five-year OS, disease-free survival (DFS), NRM and relapse were 51%, 46%, 36% and 17%. Unexpectedly, more recent transplantation (2017-2024 vs. 2011-2016) did not improve NRM (5-yr NRM was 38.9% vs. 33.3% respectively). In multivariable analysis, favorable cytogenetics was the only significant predictor of improved OS (HR 0.31, p=0.005) and DFS (HR 0.36, p=0.012). Predictors of improved relapse risk include favorable cytogenetics (HR 0.25, p=0.035) and receipt of MRD/MUD vs HAPLO (HR 0.26, p=0.043). Neither age, HCT-CI, date of transplant, conditioning intensity, use of pre-transplant Jak inhibition or use of PTCy significantly affected any transplant outcome. In summary, high NRM remains a significant barrier to transplant success in MF despite more frequent use of pre-transplant Jak inhibition and improvements in transplant techniques that have resulted in better outcomes in other hematologic malignancies. Additional strategies, such as earlier transplantation, improved patient selection, pre-transplant management of splenomegaly, choice of conditioning regimens, and optimizing pre- and post-transplant use of Jak inhibition, are likely required to improve post-transplant survival.