Allogeneic Blood or Marrow Transplantation (BMT) with Post-Transplantation Cyclophosphamide (PTCy) Based Graft Versus Host Disease (GVHD) Prophylaxis for Myelodysplastic Syndrome/ Myeloproliferative Neoplasm-Overlap Syndromes (MDS/MPN)

INTRODUCTION: MDS/MPN are a group of clonal hematopoietic disorders with overlapping features of MDS and MPN, and include chronic myelomonocytic leukemia (CMML), atypical chronic myeloid leukemia (aCML), MDS/MPN-unclassifiable along with juvenile myelomonocytic leukemia and MDS/MPN with ring sideroblast and thrombocytosis. Therapeutic options for MDS/MPN remain limited and BMT remains the only modality with curative potential. However, the overall survival of these patients following BMT remains poor marked by a high incidence of non-relapse mortality (NRM) and relapse. Moreover, most of the published cohorts consisted of patients receiving matched related donor grafts, with limited data on the outcomes of patients receiving grafts from alternative donors. In the current study, we aim to describe outcomes of BMT for MDS/MPN using PTCy as the GVHD prophylaxis platform and primarily haploidentical related (haplo) donors.

PATIENTS AND METHODS: We performed a retrospective analysis of patients with MDS/MPN who underwent BMT with PTCy at Johns Hopkins between 1/2011-1/2019. Kaplan-Meier analysis was used to evaluate overall survival (OS), relapse-free survival (RFS), and GVHD/relapse-free survival (GRFS). Cumulative incidence of relapse and non-relapse mortality (NRM) were estimated non-parametrically in presence of competing events where NRM was a competing event of relapse, and vice versa. Univariate analysis was used to evaluate the impact of donor type, HCT-CI, and karyotype on these clinical outcomes via Cox proportional hazard models for OS/RFS/GRFS, and Fine-Gray’s sub-distribution (s) hazard models for relapse/NRM/GVHD.

RESULTS: Thirty-four consecutive patients (11 women and 23 men) with a median age of 63 years (range 51-74) at the time of BMT were identified. Of these, 21 (61.8%) had haplo donors and 31 (91%) received non-myeloablative conditioning with fludarabine, cyclophosphamide and total body irradiation. Median follow up was 3.9 years (range: 24 days – 8.4 years) based on reverse Kaplan-Meier approach. Other patient, disease and BMT related details are shown in Table 1.

Time to neutrophil and platelet recovery was 22 days (range 19 – 28 days) and 31 days (range 22 – 67 days), respectively. Two (6%) patients had primary graft failure, both had received a marrow graft. At 3-years, probability of OS was 46% (95% CI: 31% - 68%), 3 RFS was 30% (17% - 51%), GRFS was 26% (14%-48%), cumulative incidence of relapse was 59% (95% CI: 41% - 77%) and incidence of NRM was 12% (95% CI: 1% - 23%). The incidence of grade 2-4 acute GVHD at 1 year following HCT was 24% but no patients had grade 3-4 acute GVHD. The cumulative incidence of any chronic GVHD at 3 years was 20% with 3 (9%) patients requiring systemic therapy. Univariate analysis showed that HCT-CI score at the time of HSCT was associated with higher incidence of relapse (sHR 5.22, 95% CI 2.26-12.08) and inferior RFS (HR 3.09, 95% CI 1.16-8.2). Other patient-, disease- and HCT-related factors were not statistically significantly associated with these clinical outcomes.

The outcomes of patients with haplo donors (n=21) were also estimated at 3 years and estimates of OS were 56% (95% CI 38% - 82%), RFS 42% (95% CI 26% - 70%), GRFS 37% (21% - 66%), cumulative incidence of relapse 43% (95% CI 21% - 66%), NRM 14% (95% CI 1% - 30%), chronic GVHD 21% (95% CI 2% - 40%). No patients developed grade 3-4 acute GVHD. While the sample size is limited, comparison of K-M estimates for RFS and relapse with haplo donor (n=21) and matched related/unrelated donor (n=12), is shown in Figure 1.

CONCLUSIONS: We report, for the first time, clinical outcomes of BMT for MDS/MPN with the PTCy based GVHD platform, and outcomes appear comparable with those previously described in other retrospective studies using conventional approach. Although limited by sample size for a statistical analysis, the outcomes with haplo donor BMT appear promising. The relatively high incidence of relapse in these patients appears to be the major factor contributing adversely to outcomes, while NRM and GVHD incidence with our non-myeloablative conditioning PTCy platform are low. We are currently exploring strategies to decrease relapse rates to further improve BMT outcomes in these patients.