Allogeneic Hematopoietic Stem Cell Transplantation for Myelofibrosis: Role of Disease Status and Chimerism on Long-Term Outcomes

Introduction. Allogeneic hematopoietic stem cell transplantation (allo-HCT) is the only curative option for Myelofibrosis (MF), a myeloproliferative neoplasm characterized by severely reduced survival. Despite a recognized superiority of transplant compared to medical options in higher risk MF, short-term complications substantially limit its use. Also, relapse can occur in a significant proportion of patients after allo-HCT. In this regard, controversial results have been reported on the predictive role of mixed chimerism. While CD33+ mixed chimerism was reported to correlate with worse OS due to higher risk of both relapse and non-relapse mortality (NRM), incomplete CD3+ chimerism was associated to an improved relapse-free survival. However, CD33 and CD3 antigens may not represent optimal markers for MF, since their expression, in contrast to CD34 antigen, has never been extensively described in MF. Herein we present a single Center experience on MF allo-HCT patients periodically monitored with CD34+, CD34- and mononuclear cells (MNC) donor chimerism with the aim to evaluate predictive factors for transplant outcome.

Patients and methods. Overall, 31 MF patients were transplanted at the Unit of Bone Marrow Transplant of ASST Spedali Civili of Brescia from May 2014 to June 2022. Patients’ characteristics at time of transplant are detailed in table 1. Regarding transplant features, 15 patients received a myeloablative conditioning and 20 double-alkylators regimen (TBF in 19 and BuCy in 1 case). Donor was sibling, unrelated or haploidentical in 22.6%, 51.6% and 25.8% of cases. HCT-CI was high in 45.2%, FIL score unfit in 9.7% and KPS was 80 or less in 6.5% of cases, respectively. After transplant, the large majority of surviving patients underwent 1, 3, 6, 9, 12, 18 and 24-month donor chimerism, by using single tandem repeats (STR) PCR on patients’ bone marrow samples.

Kaplan Meier was employed for computing OS. The Fine and Gray method was used for NRM and incidence of relapse (CIR). Three- and 6-month landmark analysis was utilized in order to evaluate the effect of mixed chimerism on outcomes. Univariate and multivariate analyses for OS were carried out using the Cox regression model. Logistic regression was employed for identifying factors associated with full donor chimerism.

Results. After a median follow-up of 2.33 years (95%CI 1.02-3.20), 15 patients had died (6 relapse, 3 acute GVHD, 3 infections, 3 other causes) for a 5y CIR, NRM and OS of 21.7% and 30.5% and 48.6%. In Univariate analysis, high risk DIPSS (HR 3.91, 95%CI 1.40-10.95, p=0.009), FIL unfitness (HR 7.00, 95%CI 1.71-28.73, p=0.007), unfavorable cytogenetics (HR 4.88, 95%CI 4.88-1.36-17.49, p=0.015), accelerated phase disease (HR 3.84, 95%CI 1.32-11.17, p=0.013) and spleen palpable >15 cm below LCM (HR 4.50, 95%CI 1.49-13.53, p=0.008) were associated to reduced survival. In Multivariable analysis, accelerated phase disease (HR 8.03, 95%CI 1.68-38.52, p=0.009), splenomegaly (HR 15.64, 95%CI 3.12-78.47, p<0.001) and FIL unfitness (HR 33.43, 95%CI 3.97-281.40, p=0.001) maintained prognostic information. Importantly, high risk mutations did not prove any ability to predict survival after transplant (HR 1.60, 95%CI 0.51-4.98, p=0.420).

Figure 1 illustrates the evolution of chimerism over time. Interestingly, 3 and 6-month MNC, CD34+ and CD34- mixed donor chimerism (<97.5%) was associated with significantly reduced survival (3-month landmark OS: 92.9% vs 20.0% [p=0.001], 100% vs 33.3% [p=0.003] and 100% vs 0% [p<0.001] at 1y; and 6-month landmark OS: 100% vs 0% [p<0.001], NA [only one patient with <97.5% donor chimerism], and 100% vs 0% [p=0.002] at 1 year, respectively), caused by increased CIR (data not shown).

Double alkylator (HR 14.70, 95%CI 1.16-185.00, p=0.038) was the only parameter predicting an higher probability of reaching full donor chimerism after allo-HCT.

Conclusions. Our study confirms the efficacy and safety of early transplant in MF. Patients submitted for transplant with massive splenomegaly or advanced stage disease appear to have a poor outcome. In addition, multidimensional geriatric assessment may represent an effective method to select MF allo-HCT candidates. The early achievement of full donor chimerism, as evaluated on MNC and CD34 lineages can predict a reduced risk for relapse and long-term survival. In this regard, use of double alkylating conditioning regimen could be relevant.