Background: NK cell cytotoxicity is modulated by interactions between killer Ig-like receptors (KIR) and their ligands. While previous studies have examined the impact of inhibitory KIR (iKIR) after allogeneic transplantation in patients with myeloid leukemia, the effects in patients with lymphoid malignancies and the impact of activating KIR (aKIR) is poorly studied. The current study examines the differences in outcomes of allogeneic hematopoietic stem cell transplantation (HSCT) between patients with lymphoid and myeloid malignancies with respect to the impact of both aKIR and iKIR.

Methods: Thirty four patients with lymphoid malignancies (LM) and 46 patients with myeloid malignancies (MM) transplanted with CD34 selected hematopoietic stem cells from related haploidentical donors following non myeloablative conditioning were studied. KIR genes in the donor and recipient were typed by Luminex-rSSOP. Each patient-donor pair was evaluated for

  1. donor iKIR and lack of corresponding class-I ligand;

  2. donor and recipient HLA ligand incompatibility; and

  3. presence or absence of donor and recipient aKIR genes and the effects of these differences on transplant outcomes were studied.

Transplant outcomes including overall survival (OS), acute graft versus host disease (aGVHD) and relapse related mortality (RRM) were evaluated by comparison of Kaplan-Meier analysis using log-rank test.

Results: LM patients with lymphoma and acute lymphoid leukemia had a median age of 44.1 (range 18–69) yrs and 64.7% had relapsed/refractory disease. Donor-recipient HLA match was 3 or 4 of 6 in 85.1% and 5 to 8 of 10 in 97.1%. MM patients with acute myeloid and chronic myelomonocytic leukemia and myelodysplastic syndrome had a median age of 49.7 (range 21–72) yrs, and 45.7% had relapsed/refractory disease. Donor-recipient HLA match was 3 or 4 of 6 in 84.8% and 5 to 8 of 10 in 80%. In the patients with myeloid malignancies, the OS was significantly higher in patients lacking HLA-C2 ligand to their donor KIR as compared to those expressing HLA C2 ligand (p=0.047). RRM was lower if patients lacked A3/11 ligand against donor KIR (p=0.024). Patients with higher numbers of absent ligands (2 or 3) had significantly lower relapses (p=0.033). However, absent ligand(s) to donor KIR did not influence the incidence of aGvHD. There was no impact of ligand-ligand incompatibility or donor-recipient activating KIR genotype on OS, RRM, or aGVHD. In contrast, in patients with lymphoid malignancies, OS and RRM were similar whether the patients expressed or lacked the ligand to donor KIR genes. Incompatibilities in the ligand-ligand model analyses did not modify OS and RRM. However, OS was better in the presence of activating receptor aKIR 2DS2 (p=0.012) in the donor and the presence of aKIR 2DS1 (p=0.017), 2DS5 (p=0.049) and 3DS1 (p=0.017) in the recipient. In addition, RRM was lower with donor aKIR D-2DS2 (p=0.044).

Conclusions: These results suggest that certain missing HLA ligands in patient to donor KIR contribute to graft-versus-leukemia effect without an increase in GVHD in high-risk MM patients undergoing non-myeloablative haploidentical related donor transplants. In contrast, patients with LM undergoing similar transplants are not impacted by missing HLA ligand in patient to donor KIR or ligand to ligand incompatibility. Importantly, presence of certain activating KIR in the patients and/ or donors impact the OS and RRM in LM patients undergoing these transplants.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author