Abstract
Introduction
Allogeneic hematopoietic stem Cell transplantation (HSCT) is the only potentially curative approach in patients with myelofibrosis. However, data on long-term outcome after HSCT are relatively limited.
Methods
This is a retrospective analysis of HSCT outcomes of all patients with primary or secondary myelofibrosis given an HSCT in a single institution, between the years 2002 and 2025. Patients with transformation to acute myeloid leukemia (AML) before HSCT were not included. The analysis focused on relapse patterns after HSCT and on long-term survival.
Results
The study included 102 patients, median age 60 years (range, 27-74), 56 males, 46 females. The donor was an HLA- matched sibling (n=37), matched unrelated (n=50) or mismatched unrelated/ haploidentical (n-15). The conditioning regimen was fludarabine and reduced dose intravenous busulfan (FluBu, total 6.4-8 mg/kg, n=25), the combination of fludarabine thiotepa, busulfan (TBF, n=61) or other regimens (n=16). Graft-versus-host disease (GVHD) prophylaxis included cyclosporine and methotrexate or mycophenolate. All patients also recieved Anti thymocyte globulin (ATG, Grafalon) during conditioning.
With a median follow-up of 5.6 years (range, 0.1-22 years), 56 are alive and 46 have died, 30 of non-relapse causes and 16 of relapse. The estimated 10-year overall survival (OS) is 48% (95%CI, 35-60). The estimated 10-year non-relapse mortality (NRM) and relapse mortality rates were 34% (95%CI, 24-48) and 18% (95%CI, 11-29), respectively. 10-year OS tended to improve in more recent years (at or after 2015), 61% (95%CI, 49-74) compared to earlier year of HSCT, 36% (95%CI, 20-53), respectively (P=0.16). 10-year OS also tended to improve with TBF conditioning, 63% (95%CI, 50-76) compared to FluBu conditioning, 28% (95%CI, 10-45), respectively (P=0.15).
A total of 32 patients were alive and disease free 5-years after HSCT. Six patients subsequently relapsed 5.3 to 18.9 years after HSCT and 1 of them died. Four patients had NRM more than 5 years after HSCT, 2 of second malignancies, one of chronic GVHD and one of a cardiovascular event. For patients who were alive and disease-free 5 years after HSCT, the estimated subsequent 5-year OS and disease-free survival were 77% (95%CI, 56-98) and 56% (95%CI, 33-78), respectively.
In all, 26 patients relapsed after HSCT. The estimate rate was 22% (95%CI, 15-32) and 31% (95%CI, 21-45) 5 years and 10 years after HSCT, respectively. The median time to relapse was 1.2 years (range, 0.2-18.9). Ten patients are alive a median of 1.3 years (range, 0.3-18.1) after relapse. The 5-year OS after relapse was 30% (95%CI, 9-51). There were 3 patterns of relapse with different outcomes. Four patients relapsed as Polycythemia Vera (PV). Their median time to relapse was 2.4 years (range, 0.3-4.1). One achieved CR with donor lymphocyte infusion (DLI) and is in CR for more than 18 years. The 3 others progressed again to myelofibrosis 38, 64 and 156 months later and are all alive at a median of 9.6 years (range, 5.3-18.1) years after first relapse. Thirteen patients relapsed as myelofibrosis in chronic phase. Their median time to relapse was 2.9 years (range, 0.3-18.9). Five are alive and 8 died. The median survival after relapse was 2.4 years (range, 0.2 to >7.5). Nine patients relapsed as MDS/ AML. The median time to relapse was 0.7 years (range, 0.2-1.4). Only one is alive with a median survival of 0.8 years (range, 0.1-1.5). Most patients were given supportive care and JAK inhibitors. Azacytidine and venetoclax was the most frequent treatment for patients relapsing as AML. A second HSCT was given to 10 patients, 5 achieved CR but 3 of them subsequently died due to NRM (n=2) or relapse.
Conclusions:
Allogeneic HSCT can allow long-term survival and possible cure for patients with myelofibrosis. However, long-term survivors are at continuous risk for late relapse and late NRM. For those who are not cured, HSCT may reset the pattern of progression of the myeloproliferative process. Patients may relapse as PV or myelofibrosis in chronic phase that can allow long-term OS after relapse. However, the outcome of patients relapsing in MDS/AML (blastic phase) is dismal.
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal