Comparison of Allogeneic Hematopoietic Stem Cell Transplant from HLA-Matched Related Donors and from Unrelated Donors for Intermediate- to High-Risk Adult Acute Myeloid Leukemia in Remission

Introduction: Allogeneic hematopoietic stem cell transplant (allo-HSCT) is currently the most powerful method to treat intermediate- to high-risk acute myeloid leukemia (AML). Human leukocyte antigen (HLA) matched related donors (MRD) are considered the best donors in allo-HSCT, but they are only available for a small proportion of patients. With the progress of transplant skills over decades, outcome of unrelated donor (URD) HSCT has improved significantly, and the difference between MRD and URD HSCT has been steadily declining. But the role of URD HSCT has still not been well defined in AML, especially for different risk groups. We therefore compared the results of URD HSCT with MRD HSCT in adult AML patients with intermediate- to high-risk in our center.

Methods: During January 2005 to December 2012, 108 adult patients with intermediate- or high-risk AML receiving allo-HSCT in complete remission (CR) in our center were admitted in this study. All patients received a standard-dose cytarabine with anthracycline or homoharringtonine as induction therapy. Consolidation courses include high-dose cytarabine or standard-dose cytarabine with anthracycline, homoharringtonine, or mitoxantrone. After consolidation, all patients received allo-HSCT in CR. High resolution DNA typing for HLA-A, -B, -C, -DRB1, and -DQB1 was performed during CR. Patients with HRD would receive HRD HSCT; while for those without HRD, URD with at least 8/10 locus match were selected as donor source. Among the 108 patients, 51 received grafts from HRD and 62 from URD. The conditioning regimen mainly involved BuCy without total body irradiation. All patients received graft-versus-host disease (GVHD) prophylaxis consisting of cyclosporine A, methotrexate and low-dose mycophenolate mofetil. Rabbit antithymocyte globulin (Thymoglobulin, IMTIX-SangStat, Lyon, France) was also administered to some patients receiving URD HSCT (4.5–6 mg/kg total dose).

Results: Most of the baseline characteristics were well balanced between two groups. However, The interval from diagnosis to transplantation differed significantly between two groups, with a median interval of 7 months for HRD group versus 10 months for URD group (P=0.008). Bone marrow was used as graft source in 100% and 80.7% of patients in HRD and URD groups, respectively. Another 19.3% of patients in URD group received peripheral blood stem cell as graft source. The median follow-up was 37.5 months in the whole cohort.

The cumulative incidence of acute grade 2 to 4 GVHD was significantly higher in URD group compared with HRD group (40.3% vs. 13.8%, P=0.002). Chronic GVHD and extensive chronic GVHD were also significantly more frequent in URD group (chronic GVHD: 49.2% vs. 29.6%, P=0.037; extensive chronic GVHD: 22.0% vs. 3.7%, P=0.005). The treatment-related mortality (TRM) was significantly higher in the URD group (17.9% vs. 4.3%, P=0.015). While the overall survival (OS), relapse-free survival (RFS), and relapse rate were comparable between two groups (OS at 3 years: HRD 79.9% vs. URD 70.3%, P=0.146; RFS at 3 years: HRD 73.1% vs. URD 66.7%, P=0.239; relapse rate at 3 years: HRD 23.5% vs. URD 18.0%, P=0.643). Clinical characteristics were subjected into univariate and multivariate Cox analysis for OS to find prognostic factors. Cytogenetics, status at transplant (CR1 vs. CR2 to CR3), and MRD levels at transplant were identified as factors that were significantly associated with OS, whereas donor type had no significant impact.

When the analysis was done in patients divided by risk stratification, we found that for intermediate-risk patients, the HRD group had a 3-year improved survival rate than URD group (RFS: 85.2% vs. 70.5%, P=0.043; OS: 85.2% vs. 72.7%, P=0.062) with a significantly reduced TRM rate (3.8% vs. 18%, P=0.025). The relapse rate were comparable between two groups (HRD 11.3% vs. URD 12.7%, P=0.649). While for high-risk patients, no differences were observed between two groups in terms of OS, RFS, TRM and relapse rate.

Conclusions: For intermediate-risk AML patients, compared with URD HSCT, the results from HRD HSCT were better with improved survival rate and reduced TRM; while for high-risk AML patients, donor types had no effect on transplant outcome. Cytogenetics, status at transplant, and MRD levels at transplant were found to be independent prognostic factors for OS.