Allogeneic Hematopoietic Cell Transplantation (HCT) Outcomes in Myelodysplastic/Myeloproliferative Neoplasms (MDS/MPN)

MDS/MPN are hematopoietic malignancies with both dysplastic and proliferative features. Therapeutic options are limited, with low probabilities and durations of remission. Allogeneic HCT is the only potentially curative treatment; however few data on outcomes and risk factors for survival in MDS/MPN have been reported. We thus evaluated our HCT experience with MDS/MPN.

From 2000 to 2013, we retrospectively identified 142 patients who underwent HCT for MDS (n=84), MDS/MPN (n=20) or secondary acute myeloid leukemia (sAML) (n=38) from a prior MDS or MDS/MPN, as defined by WHO criteria. Median age at transplant was 55 yrs (range 41-67) for MDS/MPN, 52 yrs (20-70) for MDS and 56 yrs (29-73) for sAML. Among MDS/MPN patients, 14 had CMML and 6 had MDS/MPN unclassifiable, 2 of which evolved from an MPN. Prior to HCT, 44% of MDS patients received hypomethylating agents (HMA), 12% intensive chemotherapy (IC), and the remaining other or no treatment. A minority received a combination of therapy. Similarly, 40% of MDS/MPN patients received HMA, 15% IC, 20% no treatment and 30% additional treatment such as hydrea or ruloxitinib. Most (79%) sAML patients received IC.  36% MDS, 15% MDS/MPN and 34% sAML patients had intermediate or high (35%, 45%, 42%, respectively) comorbidity scores. At time of HCT, 13% MDS, 20% MDS/MPN and 24% sAML had >10% blasts. The majority (74% MDS, 65% MDS/MPN, 66% sAML) received standard HCT with busulfan (Bu)/cyclosphosphamide, and the remaining had reduced-intensity HCT with fludarabine (Flu)/total body irradiation or Bu/Flu. All MDS/MPN patients had an HLA matched related or unrelated donor, with a minority of MDS (7%) and sAML (16%) patients receiving cord blood.

There were no differences in time to hematopoietic recovery or GVHD in all three disease groups. There was a higher incidence of relapse at 5 yrs in sAML (47%) and MDS/MPN (43%) compared to MDS (32%) however this was not significant. 100-day non-relapse mortality (NRM) was similar, 15% MDS, 20% MDS/MPN and 13% sAML, and most commonly due to infection. Overall survival (OS) was worse in sAML compared to MDS/MPN and MDS but not statistically significant (Table). Multivariable analysis identified intermediate/high risk comorbidity scores (HR 1.67, 95% CI 1.04-2.69, p=0.04) and blast count >10% (HR 1.69, 95% CI 1.02-2.81, p=0.04) as significant risk factors for mortality. A blast count of 5-10% (HR 1.01, 95% CI 0.6-1.7, p=0.97) was not found to be a significant factor, along with diagnosis, IPSS score, cytogenetics, prior therapy, or conditioning regimen. 

This analysis reinforces that HCT remains a valid treatment option for patients with MDS/MPN. Relapse is still a major cause of treatment failure and patients with high blast count or comorbidity scores are unlikely to benefit from HCT.