The Comparison of Allogeneic Stem Cell Transplantation From Family-Mismatched/Haploidentical Donors with Unrelated Donors in Adults with High-Risk Acute Myeloid Leukemia,

Abstract 4164

Since August 2008, we have been continuously doing allogeneic stem cell transplantation (SCT) from family-mismatched/haploidentical donors (FMT) in high-risk acute myeloid leukemia (AML) lacking of HLA-identical siblings, well-matched unrelated donors (WM-UD), and partially-matched unrelated donors (PM-UD), and a pilot study about the encouraging outcomes of FMT using unmanipulated donor cells and less aggressive conditioning regimen without T-cell depletion were reported at last ASH (2010;116:Abstract 3533). We herein investigated the clinical outcomes of FMT in high-risk AML compared to SCT from WM-UD and PM-UD in order to weigh up the pros and cons of FMT.

Between August 2008 to December 2010, 69 adult patients with high-risk AML underwent allogeneic SCT from WM-UD (n=33), PM-UD (n=13), and FMT (n=23). Conditioning regimen of FMT consisted of total body irradiation (800 cGy), fludarabine (150 mg/m²/day), busulfex (6.4 mg/kg/day), and ATG (thymoglobulin, Genzyme, 5.0 mg/kg). Unmanipulated granulocyte colony stimulating factor-mobilized peripheral blood stem cells (PBSCs) were used for FMT. Patients (n=46) transplanted from WM-UD and PM-UD received myeloablative (n=23, 70%) or reduced-intensity conditioning (n=14, 30%), and bone marrow (n=11, 23%) or PBSCs (n=36, 77%). Graft-versus-host disease (GVHD) prophylaxis consisted of tacrolimus and short course of methotrexate, which was identical for the three donor groups.

The type of AML was de novo (n=59, 86%) and secondary (n=10, 14%). The majority of patients (n=66, 95%) had intermediate (n=52, 75%) or unfavorable (n=14, 20%) cytogenetic and molecular features based on criteria of 2011 NCCN guidelines. Three patients with favorable cytogenetics were second complete remission (n=2) or had persistent disease (n=1) at transplantation. All but one patients (99%) and 49 donors (71%) were cytomegalovirus (CMV)-seropositive. Offspring (n=12), mothers (n=7), and siblings (n=4) were donors for FMT. There was no significant difference in patients and transplant characteristics, such as age and pre-transplant disease status, among the three donor groups.

All patients were successfully engrafted. The median time to neutrophil (>0.5^10⁹/L) and platelet (>20^10⁹/L) recovery were 12 days (range, 10–21) and 13 days (range, 8–60) for WM-UR, 12 days (range, 9–22) and 13 days (range, 9–25) for PM-UR, and 11 days (range, 10–17) and 12 days (range, 8–40) for FMT, respectively. There was no significant difference in the neutrophil (P=0.370) and platelet (P=0.487) recovery times between groups. With 24 months (range, 6–34) of median follow-up for surviving patients, overall survival (OS), disease-free survival (DFS), relapse, and non-relapse mortality were 82.7%, 82.9%, 9.4%, and 7.6% for WM-UD, 61.5%, 61.5%, 30.8%, and 7.7% for PM-UD, and 68.9%, 70.0%, 17.9%, and 12.1% for FMT. Multivariate analyses revealed that PM-UD had significantly inferior OS (HR 3.68, 95% CI 1.04–13.1) and DFS (HR 3.1, 95% CI 0.90–10.73), and higher relapse rate (HR 6.09, 95% CI 1.28–29.01) than WM-UD, whereas no significant difference in FMT [OS (HR 2.08, 95% CI 0.59–7.30), DFS (HR 1.95, 95% CI 0.60–6.40), and relapse (HR 1.19, 95% CI 0.55–2.57)]. The occurrence of acute GVHD (aGVHD) with grade II-IV (WM-UD vs. PM-UD vs. FMT; 48.5% vs. 38.5% vs. 52.2%; P = 0.65) and chronic GVHD (cGVHD; WM-UD vs. PM-UD vs. FMT; 51.9% vs. 38.5% vs. 47.8%; P = 0.25) were not significantly different in three donor groups. CMV reactivation was more common in FMT (91.3%, P=0.047) than WM-UD (66.7%) and PM-UD (61.5%), but the occurrence of CMV disease was not different (WM-UD vs. PM-UD vs. FMT; 6.1% vs. 7.7% vs. 17.4%; P=0.390). Causes of death were the recurrence of leukemia (n=3) and refractory cGVHD (n=1) for WM-UD, the recurrence of leukemia (n=4) and aGVHD combined with CMV enteritis (n=1) for PM-UD, and the recurrence of leukemia (n=4) and aGVHD combined with CMV enteritis (n=1), and CMV enteritis combined with veno-occlusive disease (n=1) for FMT.

We demonstrate that survival outcomes of FMT using a Korean-adapted protocol are comparable to SCT from WM-UD in contrast to PM-UD showing inferior outcomes than WM-UD. Our data suggest that FMT may be a better option than PM-UD for high-risk AML in the absence of WM-UD as well as HLA-identical sibling donors if complemented by more efficient CMV prophylactic measures, which needs to be validated with more large scale prospective trials in the future.