Impact of Pre-Transplant Minimal Residual Disease in Patients with AML Undergoing Myeloablative Cord Blood Transplantation

Background: Pre-transplant minimal residual disease (MRD) is associated with inferior outcomes in patients with acute myeloid leukemia (AML) undergoing myeloablative hematopoietic cell transplantation (HCT) from related and unrelated donors¹. However data on the impact of pre-transplant MRD after cord blood transplantation (CBT) are limited. Using prospectively collected data, we reviewed 80 consecutive patients with AML who received a myeloablative CBT between 2006 and 2013. Methods: The myeloablative preparative regimen consisted of TBI (1320 cGy), Cytoxan and Fludarabine (FLU) (n=50; 62%) or Treosulfan (Treo), FLU and 200 cGY TBI (n=30; 38%). Graft-versus-host-disease prophylaxis consisted of cyclosporine and mycophenolate mofetil. Ten-color multiparameter flow cytometry studies on bone marrow aspirates were performed before HCT to determine the presence of MRD. Any level of residual disease was considered MRD-positive (MRD^pos). Results: Pre-transplant MRD was identified in 20 (25%) patients. The majority of patients (n=67; 84%) received a double-CB graft; the rest received a single-CB graft. Pre-transplant demographic and clinical characteristics were similar between the 2 groups (Table 1). Disease free survival (DFS) at 5 years was 69% (95% CI: 55-79) and 52% (95% CI: 29-72) for MRD^neg and MRD^pos patients, respectively (p=0.08) (Figure1A). Figure 1B and 1C show DFS by MRD for patients receiving different conditioning regimen. After adjusting for age, disease risk (low/intermediate vs. high/very high), and CR status (CR1 vs. ≥CR2), the relative hazard of overall mortality (OM) in the MRD^pos group compared to the MRD^neg group was suggestively different with a HR=2.36 (95%CI: 1.02-5.43; p=0.04). However, the impact of MRD on OM was only significant in patients receiving a Treo based regimen [(HR=9.5; 95% CI 1.84-49.5) p=0.007)] while no difference were seen in patients receiving high-dose TBI [(HR=1.28; 95% CI 0.44-3.74) p=0.64)]. Cumulative incidence of relapse was significantly higher in MRD^pos patients, 31% (95%CI: 13-52) vs. 10% (95%CI: 4-20) in MRD^neg (p=0.01) for the group as a whole. An higher proportion of relapses were observed in MRD^pos  patients receiving Treo [34% vs. 8% (p=0.03)] while no differences were observed in patients receiving high-dose TBI [29% vs. 12% (p=0.10)]. Incidence of non-relapse mortality was 15% (95% CI: 4-34) and 19% (95%CI: 10-31) for MRD^pos and MRD^neg patients, respectively (p=0.81). Conclusion: Herein we demonstrate that the presence of pre-transplant MRD after myeloablative CBT is associated overall with adverse outcomes.  However, the presence of pre-transplant MRD impacted outcomes less if the patient received a high-dose TBI-based preparative regimen. Our study suggests furthering extending the use of CBT with high-dose TBI in AML MRD^pos patients.