Abstract 3396

Poster Board III-284


Allogeneic hematopoietic stem cell transplantation (HSCT) is a successful curative therapy for a variety of hematological malignancies. The success of allogeneic HSCT partly depends on the graft-versus-leukemia (GVL) effect caused by donor lymphocytes. Apart from alloreactive T cells, alloreactive natural killer (NK) cells also play an important role in the induction of GVL effect after HSCT. However, the effect of NK cell alloreactivity on the outcome of unrelated hematopoietic stem cell transplantation is controversial. NK cell alloreactivity after allogeneic HSCT is regulated by killer cell immunoglobulin-like receptors (KIRs). So, our study is to investigate the impact of KIRs on the outcome after unrelated hematopoietic stem cell transplantation.


116 cases of unrelated transplants between Jan. 2001 to May. 2008 were involved in the study. The inhibitory KIRs identified with HLA ligands are as follows: KIR2DL2/KIR2DL3 bind HLA-C1; KIR2DL1 recognizes HLA-C2; and KIR3DL1 recognizes HLA-Bw4. KIR-ligand mismatch was defined as the absence of one or more recipient HLA epitopes for the corresponding donor-inhibitory KIRs. Samples containing either one or two group B haplotypes were assigned the genotype designation B/x. Samples that lacked all KIR B loci were assigned the genotype A/A.


We found that KIR-ligand mismatch was significantly associated with a decreased leukemic relapse risk (p=0.019; HR=0.329, 95% CI 0.131-0.830), mainly in myeloid disease (p=0.003; HR=0.193, 95% CI 0.066-0.563). In myeloid disease, the cumulative risk of relapse was lower in the patients with KIR-ligand mismatch than KIR-ligand match (10.7% vs. 41.5%, p=0.002). This effect of the KIR-ligand mismatch on relapse was comparably significant between AML/MDS subgroup (12.6% vs. 40%, p=0.03) and CML subgroup (4% vs. 41.7%, p=0.008). For standard-risk myeloid leukemia, KIR-ligand mismatch resulted in a lower relapse rate than KIR-ligand match, although no statistical significance was reached. (11.3% vs. 26%, p=0.356). For high-risk myeloid leukemia, the risk of relapse in patients with KIR-ligand mismatch was significantly lower than that in patients with KIR-ligand match (9.1% vs. 71.4%, p=0.002). Furthermore, the impact of KIR-ligand mismatch was found mainly in the HLA-match group (13.9% vs. 62.5%, p= 0.001) and KIR haplotype, Bx group (11.5% vs. 55%, p=0.01). In myeloid disease, KIR-ligand mismatch also improved 5-year overall survival (OS) (p=0.034; HR=0.430, 95% CI 0.194-0.924) and disease-free survival (DFS) (p=0.024; HR=0.445, 95% CI 0.221-0.897). No effect was seen in patients with lymphoid disease. AA in the donor significantly decreased the cumulative incidence of relapse compared with Bx (12.9% vs. 28.6%, p=0.024). We further found that more activating KIR genes in the donor resulted in an increased relapse risk (p=0.005; HR=1.463, 95% CI 1.123-1.906). Meanwhile, the presence of donor activating KIR2DS3 gene was associated with increased relapse risk (p=0.003; HR=5.046, 95% CI 1.746-14.575) and decreased OS (p=0.004; HR=3.181, 95% CI 1.432–7.064) and DFS (p=0.003; HR=2.919, 95% CI 1.430-5.959) in myeloid disease.


Our study indicated that, in unrelated HSCT for myeloid leukemia, selection of donors with KIR-ligand mismatch offered a lower relapse risk and a long-term survival advantage. The presence of KIR2DS3 in the donor was an important risk factor for myeloid leukemia. If donors carry low number of activating KIR genes, the risk of relapse is reduced.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.