Abstract

KIR ligand incompatibility in the graft-versus-host (GvH) direction has been associated with a significant reduction in relapse, graft rejection and graft-versus-host disease (GvHD) in patients with high-risk acute myeloid leukemia (AML)undergoing full haplotype-mismatched, T-cell depleted transplants. The effect in UDT has been less consistent. This study investigates the effect of KIR ligand mismatching on the outcome of UDT in a large combined data set from the National Marrow Donor Program and the European Group for Blood and Marrow Transplantation, comparing the outcome of 1,816 KIR ligand matched and mismatched transplants for AML (n=501), chronic myelogenous leukemia (n=1024), and myelodysplasia (n=291). All cases had high-resolution HLA typing, and were matched for HLA-A, and -DRB1 alleles. Based on HLA typing for -B and -C alleles, cases were divided into one of 4 groups for comparison of outcome: KIR ligand incompatible in GvH direction (n=156), KIR ligand incompatible in host-versus-graft (HvG) direction (n=185), HLA mismatched for -B and/or -C, KIR ligand compatible (n=301), and fully HLA matched (n=1174). All received myeloablative preparative regimens, and ex-vivo T-cell depletion of the graft was performed in 18%, 22%, 16% and 15% of patients in the 4 groups, respectively. Overall, a beneficial effect of KIR ligand incompatibility in the GvH direction could not be demonstrated. KIR ligand incompatibility was associated with increased risk of grade III/IV acute GvHD and worse overall survival (OS). However, the effect varied according to whether or not ex-vivo T-cell depletion of the graft was performed (see Table). Our results suggest a detrimental effect of KIR ligand incompatibility in unmanipulated UDT, whereas this negative effect is lost with ex-vivo T-cell depletion. With ex-vivo T-cell depletion, KIR ligand incompatibility in the GvH direction may be associated with a reduced risk of severe acute GvHD and improved OS compared to HLA mismatched, KIR ligand matched transplants, with the outcome approaching that of fully HLA matched transplants. Therefore, full MHC class I matching remains the best option in UDT. KIR ligand mismatching in the GvH direction may be considered if only HLA-B and or -C incompatible donors are available and ex-vivo T-cell depletion is performed. This requires validation in prospective studies.

 GvH KIR ligand mismatch HvG KIR ligand mismatch No KIR ligand mismatch, HLA mismatch HLA matched P-value 
T-cell depleted UDT 
Treatment mortality 51% ± 17% 64% ± 14% 61% ± 13% 47% ± 7% 0.12 
Grade III/IV acute GvHD 14% ± 13% 25% ± 13% 37% ± 13% 17% ± 6% 0.04 
Chronic GvHD at 3 years 50% ± 21% 39% ± 17% 32% ± 15% 53% ± 8% 0.09 
Relapse 18% ± 14% 12% ± 9% 22% ± 11% 20% ± 6% 0.49 
OS 31% ± 17% 27% ± 14% 18% ± 11% 38% ± 7% 0.06 
T cell replete UDT 
Treatment mortality 63% ± 8% 63% ± 8% 54% ± 6% 44% ± 3% <0.0001 
Grade III/IV acute GvHD 47% ± 9% 45% ± 8% 31% ± 6% 28% ± 3% <0.0001 
Chronic GvHD at 3 years 45% ± 10% 67% ± 9% 61% ± 7% 63% ± 3% 0.005 
Relapse 22% ± 7% 12% ± 5% 15% ± 4% 17% ± 2% 0.09 
OS 19% ± 7% 28% ± 8% 33% ± 6% 43% ± 3% <0.0001 
 GvH KIR ligand mismatch HvG KIR ligand mismatch No KIR ligand mismatch, HLA mismatch HLA matched P-value 
T-cell depleted UDT 
Treatment mortality 51% ± 17% 64% ± 14% 61% ± 13% 47% ± 7% 0.12 
Grade III/IV acute GvHD 14% ± 13% 25% ± 13% 37% ± 13% 17% ± 6% 0.04 
Chronic GvHD at 3 years 50% ± 21% 39% ± 17% 32% ± 15% 53% ± 8% 0.09 
Relapse 18% ± 14% 12% ± 9% 22% ± 11% 20% ± 6% 0.49 
OS 31% ± 17% 27% ± 14% 18% ± 11% 38% ± 7% 0.06 
T cell replete UDT 
Treatment mortality 63% ± 8% 63% ± 8% 54% ± 6% 44% ± 3% <0.0001 
Grade III/IV acute GvHD 47% ± 9% 45% ± 8% 31% ± 6% 28% ± 3% <0.0001 
Chronic GvHD at 3 years 45% ± 10% 67% ± 9% 61% ± 7% 63% ± 3% 0.005 
Relapse 22% ± 7% 12% ± 5% 15% ± 4% 17% ± 2% 0.09 
OS 19% ± 7% 28% ± 8% 33% ± 6% 43% ± 3% <0.0001 

Author notes

Corresponding author