Patients with High-Risk Myelodysplastic Neoplasms Do Not Benefit from Remission Induction with Azacitidine Prior to Hematopoietic Cell Transplantation: A Retrospective, Single-Center Analysis with Extended Follow-up

Background:

The value of remission induction in high-risk myelodysplastic neoplasms (MDS) with increased blasts prior to allogeneic hematopoietic cell transplantation (HCT) remains controversial. The new ELN recommendations include the re-designation of patients with 10-20% blasts as AML/MDS, raising the question of whether these patients should receive induction chemotherapy before HCT. An alternative is treatment with hypomethylating agents (HMA) such as azacitidine, for which allo-sensitizing effects have been proposed. In contrast, bridging to HCT with HMA may increase drop-out rate for allogeneic HCT, while at the same time not further benefiting patients who subsequently receive HCT.

Methods:

We report on 135 patients diagnosed with MDS, chronic myelomonocytic leukemia or myelodysplastic/myeloproliferative neoplasm who received HCT at our center between 2010 and 2022. Median age was 60 years (range 26 to 77 years) and 44.4% of patients were female. IPSS-R score was high in 25.8% of patients and very high in 44.8% of patients. 33.3% of patients received remission induction prior to HCT, 28.8% of all patients with HMA. Conditioning regimens were myeloablative in 8.1%, FLAMSA-based reduced intensity conditioning (RIC) in 26.7%, and RIC in 60.7% of cases. Donors were matched (8/8) related in 23.7%, matched (8/8) unrelated in 53.3%, and mismatched (7/8 or haploidentical) in 23%.

Relationships between outcomes and patient, treatment and disease parameters were analyzed using multivariate Cox proportional hazards regression and, in the case of relapse, non-relapse mortality (NRM) and graft-versus-host disease (GVHD), competing risk analysis. P-values were adjusted for multiple comparisons using false discovery rate adjustment.

Results:

Median follow-up after HCT was 46 months. Kaplan-Meier estimates for OS and RFS at 5 years were 56.6% and 43%, respectively. The cumulative incidences of relapse, NRM, acute GVHD grade II to IV, and chronic GVHD at 5 years were 39%, 17.8%, 36.4%, and 40%, respectively. OS was significantly correlated with the IPSS-R cytogenetic risk category but not with the percentage of blasts in bone marrow before treatment initiation.

In multivariate regression analysis, OS was significantly worse in patients with poor or very poor IPSS-R cytogenetic category (HR 1.56; p<0.001) but not in patients with a pre-treatment bone marrow blast count exceeding 10% (HR 1.25; p=0.19). Similarly, cytogenetic risk, but not bone marrow blast count, was significantly associated with worse RFS (HR 1.5 vs. 1.2; p<0.001 vs. 0.199) and higher cumulative hazard of relapse (HR 1.74 vs. 1.36; <0.0001 vs. p=0.31). Bridging or induction treatment with azacitidine before HCT by trend showed a higher cumulative hazard of relapse (HR 2.3; p=0.057) and worse RFS (HR 1.7; p=0.088). In exploratory subgroup analyses of patients with a bone marrow blast count above 10% (n=69), OS was not significantly different between patients who received remission induction prior to HCT and those who did not (HR 1.3; p=0.5), while the cumulative hazard of relapse by trend even increased in the first group (HR 1.85; p=0.13). Furthermore, there was no significant difference in OS between patients with a bone marrow blast count above 10% who received intensified sequential conditioning with FLAMSA-RIC versus fludarabine/treosulfan (n=44; HR 1.45; p=0.52).

Donor mismatch was the only parameter significantly associated with higher NRM (HR 2.4 and p=0.044), and GVHD prophylaxis with anti-thymocyte globulin and/or tacrolimus the only parameters associated with a significant reduction of chronic GVHD severity (f²=0.05 and 0.06; p=0.03 and 0.02), while no single parameter was associated with a significant difference in acute GVHD severity.

Conclusions:

This study reports promising long-term outcomes for MDS patients undergoing HCT in the real-world setting. In our analysis, cytogenetic risk, but not bone marrow blast count, was the main determinant of transplantation outcomes. Furthermore, remission induction prior to HCT with azacitidine, which is commonly used in patients with increased bone marrow blasts, was not associated with a survival benefit but by trend even with higher relapse risk and worse RFS in multivariate analysis. This suggests that the perceived role of azacitidine as a bridge-to-transplant should be questioned. Rather, eligible patients should receive HCT as soon as possible.