Adverse Immunologic, Cytogenetic, and Molecular Features of AML in Elderly Patients Could Be Overcome by Allogeneic Hematopoietic Cell Transplantation Following Reduced Intensity Conditioning

Abstract 3065

Allogeneic hematopoietic cell transplantation (HCT) following reduced-intensity-conditioning (RIC) is a curative therapeutic option in elderly patients with AML. Yet, hematological relapse (HR) and non-relapse mortality (NRM) remain major issues. The impact of AML characteristics, post-induction consolidation chemotherapy (PCC) in patients with complete remission (CR), type of donor [unrelated (UD) vs. related (RD)], graft cell count, and Donor-cell-chimerism (DCC) on long-term outcome and management of relapse after HCT following 200 cGy TBI + fludarabine 30 mg/m² for 3 days followed by mycophenolate mofetil and cyclosporine were analysed in 245 consecutive patients with AML [132 male/113 female; median age 62 years] transplanted at the University of Leipzig. De novo and secondary AML were diagnosed in 151 (62%) and 94 (38%) patients respectively. A positive leukemic CD34-phenotype > 15% was present in 60%.Cytogenetics were high and intermediate (IR)-risk in 64 (26.7%) and 166 (69%) patients respectively. FLT3 -mutations (FLT3 mut) were present in 32 (28%) of the 115 patients with known FLT3 status. CR at HCT was present in 85% (CR1, n=155; CR2, n=53). The number of PCC applied was 0 in 88 (42%), 1 in 93 (45%), and 2 in 25 (12%). Donors were UD in 197 (80%) and RD in 48 (20%) patients. DCC in flow-sorted CD34+-marrow cells at days 28, 56, 84, and at 3 months interval thereafter was monitored by PCR of polymorphic micro satellite regions. After a median follow-up of 3.6 years, survival (OS), leukemia-free-survival (DFS), NRM, and Relapse (RI) at 5-years were 39%, 34%, 32% and 51% respectively. Engraftment was 95.5%. Incidence of acute GvHD > grade 3, limited and chronic GvHD was 22.5%, 20%, and 44.6% respectively. In multivariate analysis, type of AML, cytogenetics, CD34+-phenotype, and graft cell counts (CD3+, CD34+- and natural killer-cells) had no impact on outcome. Irrespective of the number of PCC applied, outcome was similar for CR1 and CR2. For the entire cohort and also for patients with IR-cytogenetics in CR, FLT3 mut did not adversely affect OS or RI. The lower RI after UD-HCT (39%) compared to RD-HCT (63%) (p=0.04) was opposed by a higher NRM after UD-HCT (36%) vs. 13% for RD-HCT (p=0.05) so that long-term OS and LFS were similar for both donor types. Chronic GvHD was associated with a superior OS, LFS and lower RI compared to patients without GvHD or with acute GvHD only (p<0.0005). Irrespective of leukemic CD34+-phenotype, CD34+-DCC day28 <90% was highly predictive of inferior OS (12%) and higher RI (95%) vs. OS of 50% and RI of 39% if CD34+-DCC day28 was > 90% (p<0.0001). In multivariate analysis, UD (p=0.007), female donor/male patient (p=0.02), and a higher CD34+- (p=0.01) but not CD3+-cells in the graft correlated with CD34+-DCC day28 was > 90%. HR and CD34+-DCC <90% without HR were managed by immunomodulation + chemotherapy. Overall, CR was achieved in 29% of HR and 36.4% of CD34+-DCC <90%. Relapse beyond day 100, a declining CD34+-DCC rather than a HR, and induction of GvHD correlated with a superior response and OS after relapse. Donor cells rather than the usual AML prognostic features predict long-term survival and relapse after RIC-HCT. In particular, CD34+-DCC is very effective in predicting outcome and identifying patients at risk of relapse thereby allowing early immunomodulation to enhance the graft-versus-leukemia effect. Yet, research is needed to further optimize the graft-versus-leukemia effect without the injurious effects of GvHD.