Donor-Derived Natural Killer Cell Infusion after Human Leukocyte Antigen-Haploidentical Hematopoietic Cell Transplantation in Patients with Refractory Acute Leukemia

- Early additional donor natural killer cell infusion after HLA-haploidentical hematopoietic cell transplantation is not associated with an enhanced antileukemia effect
- Early donor natural killer cell infusion after HLA-haploidentical hematopoietic cell transplantation is associated with significant toxicities related to cytokine release
- Higher NKp30 expression on donor natural killer cells is an independent predictor for higher complete remission and less leukemia progression after HLA-haploidentical hematopoietic cell transplantation and donor natural killer cell infusion

The optimum method of donor natural killer cell infusion (DNKI) after allogeneic hematopoietic cell transplantation (HCT) remains unclear. Fifty-one patients (age range, 19 years to 67 years) with refractory acute leukemia underwent HLA-haploidentical HCT and underwent DNKI on days 6, 9, 13, and 20 of HCT. Median DNKI doses were .5, .5, 1.0, and 2.0 × 108/kg cells, respectively. During DNKI, 33 of the 45 evaluated patients (73%) developed fever (>38.3°C) along with weight gain (median, 13%; range, 2% to 31%) and/or hyperbilirubinemia (median, 6.2 mg/dL; range, 1.0 mg/dL to 35.1 mg/dL); the toxicity was reversible in 90% of patients. After transplantation, we observed cumulative incidences of neutrophil engraftment (≥500/µL), grade 2 to 4 acute graft-versus-host disease (GVHD), chronic GVHD, and nonrelapse mortality of 84%, 28%, 30%, and 16%, respectively. The leukemia complete remission rate was 57% at 1 month after HCT and 3-year cumulative incidence of leukemia progression was 75%. When analyzed together with our historical cohort of 40 patients with refractory acute leukemia who underwent haploidentical HCT and DNKI on days 14 and 21 only, higher expression of NKp30 (>90%) on donor NK cells was an independent predictor of higher complete remission (hazard ratio, 5.59) and less leukemia progression (hazard ratio, .57). Additional DNKI on days 6 and 9 was not associated with less leukemia progression (75% versus 55%).