Allogeneic Hematopoietic Cell Transplantation (HCT) for Myelodysplastic Syndrome (MDS) after Pretransplant 5-Azacitidine.

Reduced intensity conditioning has decreased mortality in HCT and has increased the age of eligible patients (pts), but has increased risk of post-transplant relapse. One strategy to decrease the risk of relapse in MDS pts with IPSS > Int-1 who benefit from immediate HCT is pre-transplant treatment with 5-azacitidine. We analyzed outcomes of 22 pts who received a median of 3 (1–7) cycles of 5-azacitidine (AZA) and 13 pts who received supportive care alone (NO-AZA) prior to allogeneic HCT. Pts who received other chemotherapy for treatment of MDS were excluded. Pts received targeted busulfan (3500–6000 AUC) and fludarabine conditioning followed by peripheral blood HCT from an HLA sibling or unrelated donors between July 2004 and April 2007. Pts received GVHD prophylaxis consisting of tacrolimus plus either methotrexate or mycophenolate mofetil. AZA pts were significantly older [median age 56 (range 40–69) yr] than the NO-AZA pts [median age 49 (range 34–63) yr] (p=0.03). At referral to HCT and before starting 5-azacytidine, 20 AZA pts had MDS at Int-1 (8), Int-2 (9), or High (3) IPSS risk category; and 1 pt each had CMML-1 and CMML-2. The 12 NO-AZA pts had MDS Low (1), Int-1 (3), and Int-2 (5); 2 pts did not have karyotype information available, and one had CMML-1. Prior to HCT, AZA pts had MDS Low (4), Int-1 (6), Int-2 (7) or High (3), and 1 pt each had CMML-1 and CMML-2; NO-AZA pts had MDS Low (1), Int-1 (4), Int-2 (5) and High (1), and one had CMML-2 (p=0.8). Cytogenetics were not available in 1 pt. Donors for AZA pts were HLA-identical siblings (6), HLA-A, B, C, DRB1 matched unrelated (13), or single locus mismatched unrelated (3). Donors for NO-AZA pts, were HLA-identical siblings (8), matched unrelated (4), or a single locus mismatched unrelated (1) (p=0.06). Post-transplant follow-up for surviving pts is a median of 13.6 (range 3.8–31.2) vs 25.6 (range 3.8–34.7) months, respectively for AZA vs. NO-AZA pts (p=0.12). All evaluable AZA pts achieved complete remission posttransplant. In the NO-AZA pts, 1 pt had persistent MDS at the time of his death on day 62; the remainder achieved complete remission. At 100 days, the cumulative incidence of grade 2–4 GVHD was 92% vs 68% and grade 3–4 GVHD 36% vs. 8% in AZA vs NO-AZA, respectively, but the apparent differences can be at least in part explained by the higher prevalence of unrelated donors used for AZA pts and their higher age. Using a Cox proportional hazards regression model including age and donor, the use of 5-azacitidine was not a significant predictor for GVHD. At 2 years after HCT, the cumulative incidence of relapse was 45% vs 37%, in AZA vs NO-AZA (p=ns), and overall survival was 43% vs. 59%, in AZA vs NO-AZA (p=ns). Causes of non-relapse death in AZA include relapse of pre-existing NHL (1), GVHD (2), infection (1 viral, 1 fungal), interstitial pneumonitis (1), sudden cardiac death (1), and unknown (1) and in NO-AZA pts include GVHD (1), and multiorgan failure (1). From these observations it appears that pretransplant treatment with 5-azacitidine may not be associated with a lower risk of relapse posttransplant. However, 5-azacitidine may be of value in stabilizing disease thereby allowing time for pts who may otherwise experience progressive disease to reach transplant. Further appropriately designed studies will be needed to address the safety of this strategy.