Matched Unrelated Donor Hematopoietic Cell Transplant with Increased Donor Inhibitory KIR Improves Survival in a Pediatric Acute Leukemia: A CIBMTR Cohort Analysis

Natural killer (NK) cells potentially mediate innate immunity early after hematopoietic cell transplantation (HCT). NK cells posses an array of activating and inhibitory receptors on their surface, including killer immunoglobulin-like receptor (KIR) and which interact with KIR-ligands (KIRL) on target cells. The impact of NK cell mediated alloreactivity in pediatric HCT is not well characterized. In this study, we explored a mathematical approach to quantify KIR-KIRL combinations to assess clinical outcomes in a cohort of pediatric patients who underwent HLA-matched unrelated donor (MUD) HCT for acute leukemia including AML and ALL.

This retrospective study was performed on deidentified donor and recipient demographic and clinical outcomes data provided by the Center for International Blood and Marrow Transplant Research (CIBMTR). KIR-KIRL interaction values were assigned as previously (Krieger et. al 2021, Bone Marrow Transplantation). Briefly donor inhibitory (iKIR) and activating (aKIR) KIR- recipient KIRL interactions, accounting for missing KIR-L (mKIR), were summed up, using a system of matrices.

The study cohort was comprised of 686 pediatric donor-recipient pairs (DRP) who underwent MUD HCT for AML (N=355) and ALL (N=333) and had neutrophil engraftment (Table 1). This cohort of pediatric patients was aged 0-18 years (average 11). All patients received T cell replete grafts; 26% received in vivo T cell depletion, with either anti-thymocyte globulin (ATG) or alemtuzumab. The majority, 77% received a bone marrow graft and 90% received myeloablative conditioning.

The effect of all 3 KIR-KIRL interaction scores (absolute values of iKIR, mKIR and aKIR) on clinical outcomes were evaluated. Higher donor iKIR score was associated with improvement in overall survival (OS) SHR 0.73 (multivariate P=0.003) and relapse-free survival (RFS), SHR 0.8 (P=0.019); grade 3-4 acute GVHD was lower in patients transplanted with higher iKIR or mKIR scores (iKIR; SHR 0.58 (P=0.013); mKIR SHR 0.6 (P=0.041)). Higher aKIR scores, were not significantly associated with any clinical outcomes examined. As in previous analysis, the iKIR and mKIR scores were added to give an IM-KIR score, which was dichotomized for analysis into those DRP with an IM KIR score of 5 (IM = 5), and those with an IM-KIR score of 2, 3 or 4 (IM < 5). Of note, KIR haplotype A patients always have an IM KIR score of 4, and are thus categorized among IM<5 donors. In this cohort of pediatric patients transplanted for both AML and ALL, those transplanted with IM=5 donors demonstrated improved OS, SHR 0.75 (P = 0.027), RFS, SHR 0.79 (P=0.05) and grade 3-4 aGVHD incidence, SHR 0.37 (P= 0.003) in multivariate analysis. We next examined outcomes in patients with ALL and AML to understand disease specific effects. Patients transplanted for AML with IM=5 donors had an improvement in OS, SHR 0.62 (P=0.01), as well as RFS, SHR 0.70 (p=0.033) (Fig. 1). There was a trend towards relapse risk reduction in this cohort SHR 0.67 (univariate P=0.048). Patients transplanted for ALL with IM=5 donors demonstrated a marked decrease in grade 3-4 acute GVHD incidence SHR, 0.31 (P=0.017).

This large pediatric study confirms that unrelated DRPs with donor haplotype B genotypes and a higher inhibitory KIR content (regardless of KIR ligand status), confer significant improvement in overall survival and severe aGVHD after MUD HCT. Those patients transplanted for AML experienced an improvement in OS and RFS despite the lower event rate observed in pediatric-age patients, the majority of whom receive myeloablative preparatory regimens without the use of in vivo T cell depletion. This challenges the notion that KIR are irrelevant to pediatric donor selection. Future clinical trials evaluating donor selection for URD HCT should include this measure to evaluate its value prospectively.

No relevant conflicts of interest to declare.