Analysis of the Impact of KIR B Haplotype Donors on Outcomes after Haploidentical (HI) and Matched Related (MR) Hematopoietic Stem Cell Transplantation (HSCT)

Abstract

Introduction:Natural killer (NK) cell alloreactivity has been associated with graft versus leukemia effects in HSCT. These effects are in part mediated through killer cell immunoglobulin-like receptors (KIR) on the surface of NK cells. The absence of inhibition of NK cells via inhibitory KIR in KIR ligand mismatched T-cell depleted HI HSCT, and the use of unrelated donors possessing activating KIR haplotypes (KIR-B) have been associated with improved overall survival (OS) due to relapse protection. Inhibitory KIR ligand mismatch in the setting of T-cell replete HI HSCT has not consistently been shown to be beneficial, and data regarding the impact of using KIR-B donors in T-replete HI and MR HSCT is limited. In this preliminary retrospective review, we compared outcomes between patients with KIR-B donors versus KIR A haplotype (KIR-A) donors after HI and MR HSCT to help inform donor selection.Methods:All patients underwent HSCT on our 2-step approach in which after conditioning, a fixed T cell dose (2 x 108/kg) is administered (step 1), followed 2-3 days later by cyclophosphamide (CY) for bidirectional tolerization. Two days after CY, patients receive a CD 34-selected donor product (step 2). To be included in the analysis, patients had to be alive for at least 1 month post HSCT and be without evidence of graft rejection. Recipients were divided into 2 groups based on whether they were transplanted from a KIR-B (possessing one or more non-framework activating KIR genes) versus a KIR-A donor (genotypically negative for non-framework activating KIR genes). The relationship between the use of a KIR-B donor and the end points of OS, relapse, grades II-IV acute graft versus host disease (GVHD), chronic GVHD and non-relapse mortality (NRM) was examined. Probability of OS was estimated by Kaplan-Meier (KM) method with log-rank analysis and multivariate analyses using Cox proportional hazards. Donor and recipient age and gender, HCT-CI score, Revised Disease Risk Index, presence of any inhibitory KIR ligand mismatch between donor and recipient and CMV reactivation after HSCT were considered as confounding variables. Cumulative incidence functions with competing risk analyses were used to estimate the probabilities of relapse, NRM, acute and chronic GVHD.Results:The study group consisted of 128 consecutive patients with myeloid malignancies (AML=91 MDS=27, CML=1, MPD=9) undergoing HI [n=84] or MR [n=44] HSCT. The median recipient age and follow-up of the study group was 58 (range 19-78) years and 40.6 (range: 1.07-117.27) months respectively. Only 13 (10.2%) patients had grade III or higher acute GVHD and only 7 (5.5%) patients developed severe, chronic GVHD (score of 3) respectively. Seventy-four (58%) and 54 (42%) of recipients had KIR-B versus KIR-A donors respectively. Fifty-five (74.3%) KIR-B donors possessed HLA C1 or C1/2 alleles in the presence of KIR2L2/3/2DS2. On multivariate Cox regression analysis, OS (HR, 0.579; 95% CI, 0.337-0.995; P= 0.048) was significantly higher in patients who received HSCT from KIR-B donors. On competing risk analysis, relapse (HR, 0.756; P=0.403) and NRM (HR, 0.466; P=0.128) were lower in patients who had KIR haplotype B donors though not statistically significant. In contrast, the risk of grades II to IV acute GVHD (with death as a competing event) was higher in patients with KIR-B donors (HR, 1.972; 95% CI, 1.009-3.852; P=0.047). Similarly transplants from KIR-B donors were associated with a higher risk of chronic GVHD (HR, 2.89; P=0.053).Discussion:Our results show that within the context of the 2-step approach, the use of a KIR-B donor is associated with improved OS and an increased risk of acute and chronic GVHD in patients with myeloid neoplasms receiving HI and MR HSCT. While not reaching significance, the results for relapse and NRM suggest that the OS benefit in patients with KIR-B donors was derived from a combination of lower relapse rates and NRM that was not adversely impacted by GVHD which was mostly lower grade in this population. Our findings support the concept of KIR-B donors as more alloreactive than KIR-A donors, specifically those possessing a C1-C1/2/2DL2/3/2DS2 repertoire, the primary type possessed by KIR-B donors in this analysis. The data argues for the continued use of KIR genotyping at our institution and analyses of HSCT outcomes based on donor KIR repertoire and other donor characteristics to optimize donor selection in HI and MR HSCT. [Display omitted] [Display omitted] [Display omitted] DisclosuresNo relevant conflicts of interest to declare.