Comparison of 5-Azacytidine Versus Allogeneic Hematopoietic Cell Transplantation In Elderly (≥ 60 years) Patients with De Novo High-Risk MDS - a Retrospective Matched Pair Analysis

Abstract 3500 Background: Myelodysplastic syndromes (MDS) are a disease of the elderly in whom until recently, allogeneic hematopoietic cell transplantation (HCT) has not been considered a reasonable therapeutic option. Best supportive care (BSC), only including transfusion support, has been the standard of care for the majority of these patients (pts). However, innovations in transplant protocols have changed that approach resulting in an increased frequency of allogeneic HCT. Further, several innovative non-transplant therapeutic modalities for MDS pts have been developed. In fact, treatment with DNA-methyltransferase inhibitors (MTI) such as 5-azacytidine (5-aza) has changed the natural course of the disease and prolonged median survival by an average of 9 months compared to BSC. However, allogeneic HCT is currently still the only modality with proven curative potential, although there has been no true randomized trial comparison of HCT to 5-aza therapy only. Methods: Consequently, we performed a retrospective analysis in 126 pts, 60 –77, (median 64) years of age with de novo high-risk MDS according to FAB classification (RAEB n=88, RAEB-t n=20, CMML n=18) undergoing allogeneic HCT at a median of 9 months from initial MDS diagnosis. Cytogenetics were available in 108 pts (good: n=63, intermediate: n=18, poor: n=27), resulting in IPSS classification (in 107 pts) of INT-1 (n=28), INT-2 (n=45) or HIGH (n=34). The ECOG was available in 125 pts, and was either 0 (19%), 1 (66%) or >1 (15%). Prior to HCT, most pts had progressed to a higher stage than present at diagnosis (RAEB n=45, RAEB-t n=10, CMML n=10, AML n=61), the median blast count in the marrow being 12%. Patients were prepared for HCT with one of several reduced (RIC, n=78) or more conventional intensity (n=48) conditioning regimens followed by stem cells from either related (n=50) or unrelated (n=76) donors. The outcome of this high-risk MDS cohort was compared to patients from the French (GFM) 5-aza patient registry matched for age, gender, prior induction chemotherapy, time from diagnosis to therapy (either HCT or 5-aza), disease stage and cytogenetics at diagnosis as well as at start of therapy. The groups could be well matched except for age (median age 5-aza: 67 vs. HCT: 64; p=0.001). Further, the HCT group included more pts with a higher ECOG score (ECOG>0: 81% vs. 59%; p=0.05) prior to start of therapy. Results: With a median follow-up of 20 months for surviving pts from the start of therapy the estimated 3-year overall survival (OS) was 39% (95% CI, 30% to 48%) for HCT and 7% (95% CI, 0 to 16%) for 5-aza, respectively. In multivariate Cox regression analysis, ECOG score (p<0.001), disease stage p=0.021), cytogenetic risk group (p=0.002) and type of treatment (5-aza vs. HCT HR: 1.8; p=0.003) were associated with survival. Conclusion: These retrospective data suggest that allogeneic HCT from related or unrelated donors, using various conditioning approaches, might offer a meaningful survival benefit compared to 5-aza in patients with high-risk MDS, even in the 7th decade of life. There is a need for prospective clinical trials in order to determine the place of this approach within the growing therapeutic opportunities for pts with MDS. Disclosures: Platzbecker: Celgene: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding. Ades:Celgene: Honoraria, Membership on an entity's Board of Directors or advisory committees.