The interactions of variable killer-cell immunoglobulin-like receptors (KIR) with polymorphic HLA class I ligands form an extraordinary immunogenetic system that influences NK cell biology, human susceptibility to disease, and the outcome of hematopoietic cell transplantation (HCT). The independent segregation of KIR and HLA genes, on chromosomes 19 and 6, respectively, increases the functional diversity of the system. Previously we found that patients undergoing unrelated donor (URD) HCT for acute myelogenous leukemia (AML) had superior leukemia-free survival (LFS) and less relapse with KIR B/x genotype donors than with KIR A/A donors. At higher resolution we identified the significant protective effect of donors with “Better” (≥2 B-motifs) and “Best” (≥2 B-motifs with Cen-B/B) KIR gene content compared to “Neutral” (0 or 1 B-motif) donors, supporting a donor selection strategy that is being tested in an ongoing prospective multicenter trial. The heterogeneity of transplant cohorts (preparative regimens, graft source, T cell depletion, HLA match status) has complicated the evaluation of the mechanisms underlying the beneficial effects mediated by NK cells. We analyzed 2404 URD transplants for AML and found a strong interaction between the protection associated with KIR B donors and the conditioning intensity (myeloablative [MA] vs. reduced intensity [RI]) for relapse (p=0.0002) and LFS (p=0.043). “Better” and “Best” donors were associated with significantly enhanced LFS and relapse protection in MA transplants, but appears to have an opposite effect in RI transplants. Therefore, to further explore the mechanisms of NK cell mediated protection, we focused our evaluation on a cohort of 1007 MA, T cell replete URD transplants for AML. We evaluated the interaction of donor KIR with recipient and donor HLA C1, C2 and Bw4. Superior LFS (RR 0.78 [0.66-0.92], p=0.0024) and relapse protection (RR 0.50 [0.38-0.67], p<0.0001) were observed with “Better+Best” donors (n=299) compared to “Neutral” donors (n=656). The benefit associated with “Better+Best” donors was enhanced in recipients expressing 1 or 2 HLA-C allotypes carrying the C1 epitope (C1/x (n=245) vs. C2/C2(n=51)) (RR 0.65 [0.41-1.01); p=0.05). No significant beneficial interactions were found between donor KIR and donor HLA-C group KIR ligands or with recipient or donor Bw4. Importantly, we did not identify any specific KIR B-haplotype defining gene as associated for the LFS advantage and relapse protection observed in C1/x recipients compared to C2/C2 recipients. Individual analyses for KIR B/x donors with or without 2DS1, 2DS2, 2DS3, 2DS5, 2DL2, 2DL5, 3DS1 showed equally significantly improved outcome (RR ranged from 0.71-0.80, p=0.002-0.03 for LFS and RR 0.56-0.63; p=0.0003-0.0008 for relapse). There was no clinical advantage of any KIR B genes in C2/C2 recipients. Importantly, the components of the relapse protection associated with interactions between “Better+Best” donors and C1/x recipients may depend upon the HLA-match status of the transplant. (P=0.018 for the interaction). In the <10/10 HLA-matched transplants (n=513), the “Better+Best” donors conferred potent relapse protection in C1/x (n=122) vs. C2/C2 (n=28) recipients (RR 0.35 [0.16-.76]; P <0.01) (Figure 1; right). This relapse protection, of equivalent strength in HLA-C or other HLA mismatches, resulted in improved LFS (RR 0.55 [0.34-.90]; P=0.02). Conversely, in the 10/10 HLA-matched transplants (n=494), treatment related mortality (TRM) was reduced when “Better+Best” donors could engage recipient HLA-C1 (n=123, RR 0.40 [0.20-.0.81]; P=0.01) although LFS was not statistically different, due in part to a lack of enhanced relapse protection (RR 3.00 [0.69-13.01]; P=0.14)(Figure 1; left). Interactions between KIR B-haplotype donors and homozygous C1/C1 in the recipient were associated with less acute GVHD when all transplants were analyzed. In summary, analysis in this homogeneous cohort demonstrates that the outcome of URD transplantation for AML is improved by interactions between recipient HLA-C1 and “Better+Best” KIR donors, which reduce relapse in <10/10 HLA-matched and reduce TRM in 10/10 HLA-matched transplants. This protection is observed in MA but not RI conditioning, suggesting that NK cell reconstitution and function differs in these settings.

Can't read 3B2 tag because stream don't exist.Tag: Figure_12

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution