The Relapse-Free Survival Benefit Associated with Group B KIR Haplotype Donors for Unrelated Hematopoietic Cell Transplantation is Unique to Acute Myelogenous Leukemia

Abstract

Natural killer (NK) cell alloreactivity correlates with improved survival after T-cell deplete but not T-cell replete hematopoietic cell transplantation (HCT), especially in myeloid diseases. Using samples from the NMDP repository and clinical outcome data from the CIBMTR we evaluated the effect of donor and recipient KIR genotype on outcomes after URD HCT. Using a validated assay we genotyped donors and recipients from 836 HLA-matched and 915 mismatched T-replete URD transplantations performed between 1988 and 2003 for acute myelogenous leukemia (AML: n = 456), acute lymphoblastic leukemia (ALL: n = 334), chronic myeloid leukemia (CML: n = 765) or myelodysplastic syndrome (MDS: n = 196). Multivariate models evaluated the effect of donor and recipient KIR genotypes (A/A: two A KIR haplotypes, or B/x: at least one B haplotype; typically containing more activating KIR) on clinical outcomes.Tabled 1Effect of Donors with B KIR Haplotypes on Relapse-Free Survival and Relapse after URD HCTDonor KIR genotype B/x (reference: A/A)Hazard ratio95% CIp-valueRelapse-Free Survival AML (n=4560.74(0.59, 0.93)0.01 ALL (n=334)0.92(0.69, 1.22)0.56 CML (n=765)1.06(0.86, 1.30)0.61 MDS (n=196)1.05(0.72, 1.54)0.80Relapse AML (n=456)0.71(0.50, 1.02)0.06 ALL (n=334)1.11(0.70, 1.75)0.66 CML (n=765)0.80(0.33, 1.94)0.62 MDS (n=196)0.95(0.42, 2.13)0.90 Open table in a new tab Natural killer (NK) cell alloreactivity correlates with improved survival after T-cell deplete but not T-cell replete hematopoietic cell transplantation (HCT), especially in myeloid diseases. Using samples from the NMDP repository and clinical outcome data from the CIBMTR we evaluated the effect of donor and recipient KIR genotype on outcomes after URD HCT. Using a validated assay we genotyped donors and recipients from 836 HLA-matched and 915 mismatched T-replete URD transplantations performed between 1988 and 2003 for acute myelogenous leukemia (AML: n = 456), acute lymphoblastic leukemia (ALL: n = 334), chronic myeloid leukemia (CML: n = 765) or myelodysplastic syndrome (MDS: n = 196). Multivariate models evaluated the effect of donor and recipient KIR genotypes (A/A: two A KIR haplotypes, or B/x: at least one B haplotype; typically containing more activating KIR) on clinical outcomes.