Since the approval of hypomethylating agents for the treatment of myelodysplastic syndromes (MDS) increasing numbers of patients have received these agents prior to haematopoietic stem cell transplant (HSCT). A number of small studies have suggested that this approach results in similar outcomes to conventional chemotherapy. In light of this we utilised the EBMT dataset to retrospectively analyse the outcomes of hypomethylating agent (HMA) therapy compared with conventional chemotherapy (CC) or cytarabine (ara-C) alone pre-HSCT

Because hypomethylating agents were approved in early 2000, we selected MDS patients who received allogeneic stem cell transplantation between 2004 and 2011 reported to EBMT. In order to include a homogeneous group of patients with blasts at time of diagnosis we included only patients classified as RAEB or RAEB-T at time of diagnosis with sufficient data on anthracycline containing chemotherapy, hypomethylating agents or cytarabine. 234 MDS patients treated with hypomethylating agents (N=77), conventional anthracycline-based chemotherapy (N=132) or cytarabine (N-25) receiving myeloablative (N=94) or reduced intensity conditioned (RIC) HSCT (140), from sibling (N=101) or matched unrelated donors (N=133) were identified. 136 (58%) male and 98 (32%) females with a median age of 56 years (range 19-69) were included. The median follow-up of the cohort was 22.1 months (95%CI 0.7-94.3). Patients with >30% blasts were excluded. Cytogenetics were abnormal in 57%, 51% and 26% of patients receiving HMA, CC and ara-C respectively(p=0.04). Of the 77 patients treated with HMA, 73 had RAEB and 4 had RAEB-T, of the 132 patients receiving CC, 117 had RAEB and 11 had RAEB-T (missing data in 4) and of the 25 patients receiving ara-C 15 had RAEB and 1 RAEB-T (missing data in 9). Overall 127 patients (54%) achieved CR prior to HSCT with more CR in the CC group (68%) than either the HMA (32%) or ara-C (48%) groups (p<0.01). A number of patients achieved partial but not complete remission (non-CR) including 22% of patients receiving CC, 48% of patients receiving HMA and 24% of patients receiving ara-C. Primary refractory disease was noted in 10%, 19% and 28% of patients in CC, HMA and ara-C groups.

On univariate analysis overall survival (OS), relapse free survival (RFS), and non-relapse mortality (NRM) did not differ significantly between those receiving HMA, CC or ara-C at three years; OS (42% vs 41% vs 54%), RFS (29% vs 36% vs 49%) and NRM (26% vs 26% vs 31%) respectively. The median OS for those treated with HMA was 31 months (10-not reached) vs 18months (13-45) for CC (p=NS). Results for sibling vs unrelated donor HSCT were similar for OS, RFS and relapse with worse outcomes for URD with respect to NRM (HR 1.74 (0.98-3.11) p=0.06). No significant difference in outcomes for standard vs RIC HSCT or normal vs abnormal cytogenetics was observed. When compared to patients in remission those in non-CR had significantly worse outcomes with regard to relapse (HR 1.75 (95%CI 1.05-2.91) p=0.03) and approached significance for RFS (HR 1.47 (95%CI(0.99-2.19) p=0.06). This was not evident for OS (HR 1.29(95%CI 0.83-2.01p=NS), or NRM (HR1.12 (0.58-2.16) p=NS). Patients with primary refractory disease had significantly worse outcomes across all parameters; OS(HR 2.81(95%CI 1.74-4.56) p=<0.01), RFS (HR 2.63 95%CI 1.67-4.16) p=<0.01), relapse (HR 2.32(95%CI 1.22-4.40) p=0.01) and NRM (HR 3(95%CI1.56-5.79) p<0.01). On multivariate analysis when adjusted for CR status at transplantation those receiving CC had inferior outcomes for OS compared to those receiving HMA (HR 1.54 (0.97-2.45), p = 0.07). For patients with primary refractory disease MVA demonstrated significantly worse outcomes for OS, RFS, CIR and NRM (HR 3.4, 3, 2.62 and 3.5 respectively p<0.01 for all).

Whilst the inherent limitation of retrospective data analysis are acknowledged these findings support the growing body of evidence that HMA therapy is comparable to both CC and ara-C alone when used as pre-transplant induction treatment. Disease status is clearly important for best outcomes and the results of prospective studies which clarify the ability of hypomethylating agents vs other methods to achieve this are awaited.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution