Impact of KIR and HLA Genotypes on Outcome in Nonmyeloablative Hematopoietic Cell Transplantation (HCT) Using HLA Matched Related Donors.

Killer immunoglobulin-like receptor (KIR)-HLA ligand mismatches in the graft-versus host direction (i.e. the recipient lacks an inhibitory HLA ligand for one or more donor KIRs) appears to be associated with improved outcome in patients (pts) receiving HLA-matched or mismatched T cell-depleted HCT. We investigated the impact of genotypic KIR-ligand mismatches in pts undergoing T cell-replete HCT from an HLA-identical or 5/6 antigen-matched related donor following nonmyeloablative conditioning. Donor KIR genotype (KIR2DL1,KIR2DL2/3,KIR3DL1) and HLA-B and C genotypes were analyzed in 92 pts who received an unmanipulated, G-CSF mobilized peripheral blood HCT from a 6/6 (n=88) or 5/6 (n=4) HLA-matched related donor following conditioning with cyclophosphamide (120mg/m²) and fludarabine (125mg/m²). The conditioning regimen included anti-thymocyte globulin (160mg/kg) for pts at increased risk for graft rejection (heavily transfused, 5/6 HLA-matched donors, or non-sibling related donors, n=19). Indications for transplantation included metastatic renal cell cancer (RCC, n= 37), treatment refractory solid tumors (n=26), hematologic malignancies (n=9) and non-malignant hematologic disorders (n=20). Cyclosporine, alone (n=11) or in combination with either mycophenolate mofetil (n=33) or methotrexate (n=48) was used as graft versus host disease (GVHD) prophylaxis. Sixty-eight donors (73%) demonstrated genotypic evidence of at least one KIR for which the pt lacked a corresponding inhibitory HLA C or B ligand (‘missing KIR ligand’). There was a trend towards a higher incidence of grades 3–4 acute GVHD (37% vs. 17%, p=0.08) and a significantly higher incidence of steroid refractory GVHD (16% vs 0%, p=0.02) in pts lacking an HLA ligand for one or more donor KIR. The ‘missing KIR ligand’ effect on response rate was further evaluated in a subset of pts with metastatic RCC undergoing HCT. An HLA ligand for one or more donor KIR was absent in 28/38 (76%) RCC pts. A higher objective response rate (PR or CR) was seen in pts with one or more missing KIR-ligands (12/28, 43%) compared to those whose HLA genotype would predict inhibition of all donor KIRs (2/9, 22%, p=0.24). This effect was most pronounced in pts homozygous for HLA Bw6 who lacked HLA Bw4 ligands capable of inhibiting donor KIR3DL1. Compared to pts not homozygous for HLA-Bw6, Bw6 homozygous RCC pts had a higher response rate (24% vs 58%; p=0.047) and significantly prolonged survival (median 419 vs 1496 days; p=0.005) following HCT. A univariate analysis considering patient age, acute and chronic GVHD, and KIR-HLA ligand mismatches (C mismatch, B mismatch, or > 1 mismatch) showed that a missing HLA-Bw4 ligand and the absence of a ligand for > 1 donor KIR were each associated with improved survival in RCC patients; in a multivariate analysis, absence of an HLA Bw4 ligand for donor KIR3DL1 was the only factor that remained significantly associated with improved survival (p=0.015, OR 0.276, CI 0.097–0.781). Conclusion: KIR-HLA ligand mismatches appear to impact multiple transplant outcomes including the incidence and severity of acute GVHD, tumor response and survival following T-cell replete nonmyeloablative transplantation from HLA-matched related donors. Evaluation of a larger cohort of patients to confirm these findings is currently underway.