Impact of Azacitidine (AZA) Prior to Allogeneic Stem Cell Transplantation (allo-SCT) for Myelodysplastic Syndromes (MDS): A Large-Scale Study on Behalf of the SFGM-TC and GFM Groups

Abstract 160 Background: Until the early 2000s, treatment prior to transplant in candidates to allo-ST varied according to cytogenetic score and percentage of BM blasts, from best supportive care (BSC) to AML-type chemotherapy (intensive CT). Recently AZA has become a reference treatment of IPSS higher risk MDS not candidates for allo-SCT, but its role prior to transplant in order to reduce the tumor burden remains uncertain. Methods: We report a retrospective study on 470 consecutive MDS patients who underwent allo-SCT between Jan 1999 and Dec 2009 in 26 French and Belgian centers. Inclusion criteria were: age > 18 and allo-CST from either sibling (n=285) or HLA-A, -B, -C, -DRB1 and -DQB1 allele matched unrelated donor (10/10) (n=172) for MDS or AML/RAEB-t (with 20–30% BM blasts). Data quality was ensured using computerized discrepancy errors and vigorous on-site data verification of every single file. HLA matching was double-checked by the French Bone Marrow Donor Registry. Results: The first 405 files analyzed until now are presented, including 250 males and 155 females. At diagnosis, WHO was: 102 RA/RARS/RCMD, 136 RAEB-1, 146 RAEB-2 and 18 RAEB-t/AML; Cytogenetic IPSS were favorable (n=205), intermediate (n=104), unfavorable (n=86) and missing (n=10); 216 patients had IPSS Int-2 or High. One hundred and therty-three patients had progressed to a more advanced disease before allo-SCT. Prior to transplant, 77 patients had received AZA, either alone (AZA-alone-group; n=45) or AZA preceded or followed by intensive CT (AZA-chemo-group; n=32), generally due to failure of the first treatment given. The 328 remaining patients received BSC (n=162) or intensive CT (n=166) (no-AZA-group). In AZA groups, the drug was started 38 to 941 days after diagnosis (median 150) and discontinued 6 to 438 days before transplant (median 60) with a median number of 4 cycles (range 1–26). Overall, 178 patients (45%) were considered responders at transplant (in CR or PR), including 32/45 (71%) treated with AZA-alone, 19/32 (60%) with AZA-chemo, 6/162 (4%) with BSC and 121/166 (73%) with intensive CT, while 222 were transplanted with progressive disease (untreated, stable without hematological improvement, relapsed or refractory disease). Median age at allo SCT was 54 (range18–70). Patients received myeloablative (n=145) or nonmyeloablative (NMA) (n=260) conditioning with bone marrow (n=118) or PBSC (n=287) as source of stem cells. Compared with other treatment groups, patients belonging to AZA-alone-group were older (p =.025) and had more often Int-2 and high IPSS (p=.013) and poor cytogenetic IPSS (p<.001) and received more often NMA conditioning (p =.005) from an unrelated donor (p=.007). As of April 1st, 2011, median FU was 4.6 years (range, 0.2–12.2). The estimated 3-year OS was 60%, 28%, 52% and 49% in the AZA-alone, AZA-chemo, BSC and intensive CT groups, respectively (p=.033); The estimated 3-year TRM was 13%, 29%, 34% and 20% in the AZA-alone, AZA-chemo, BSC and intensive CT groups, respectively (p=.055) and 3-year relapse was 31%, 41%, 29% and 38% in the AZA-alone, AZA-chemo, BSC and intensive CT groups respectively (p=.169). Multivariate analyses confirmed the influence of prior treatment with an unfavorable outcome in patients of the AZA-chemo group (who received CT before or after AZA) on OS and EFS (p=.003 and p=.0013, respectively), in spite of the fact that, 60% of the patients in that group had achieved at least PR at transplant. Conclusion: With the goal of down-staging underlying disease before allo-SCT, AZA treatment appears to be a valid therapeutic approach, but mainly in patients receiving AZA alone since allo-SCT in patients who required both AZA and chemotherapy had less satisfactory outcomes, possibly reflecting additional toxicity and/or more resistant disease. Disclosures: Michallet: Celgen: Honoraria. Mohty: celgene: Honoraria. Fenaux: Celgene: Honoraria, Research Funding. Yakoub-Agha: celgene: Honoraria, Research Funding.