Infusions of Allogeneic Natural Killer Cells as Cancer Therapy

Dynamic equilibrium. After cell-to-cell contact, NK cell integrates signals from its surface receptors in seconds, resulting in either target-cell attack or no response and continual immunosurveillance of other cells. A, healthy cells express normal amount of MHC class I ligands with no activating “stress” ligands. B, downregulation or absence of MHC ligand for cognate inhibitory receptors is insufficient to trigger NK cells. This happens in the physiologic setting with red blood cells and autologous KIR receptor–ligand mismatch cells, and in pathologic conditions with adult lymphoblastic leukemia. C, sufficient activating ligands must be expressed on target cells to induce NK cell activity. D, if self-MHC ligands are expressed in normal amount, the reactivity of the NK cell is ultimately dependent on the balance of activating and inhibitory signals. Successful NK cell therapy relies on clinical strategies that optimize activation and avoid inhibition by cancer cells.

Three basic steps of allogeneic NK cell therapy. A, the first step includes donor selection based on health history and examination, infectious biomarkers, and KIR typing, because KIRs are highly polymorphic. All three levels of KIR typing should be done, including genotyping, phenotyping, and allelotyping (depicted from left to right are representative outputs from flow cytometry, RT-PCR, and allelotyping analyses). B, the second step is NK cell purification and processing to enrich NK cells and to improve NK cell functions. The process needs vigorous steps for quality assurance (QA) and quality control (QC). C, after cell infusion, antitumor activity could be enhanced in vivo using various augmentation agents (Aug).

KIR haplotype map, B scoring, mismatch model, and typing algorithm. A, simplified genomic maps of A and B haplotypes. KIR genes that are present in some but not all cen-B motifs are color-coded in green and the two loci that are used for simplified B-scoring are in red. B, calculation of total B scores. C, all three mismatch models are biologically sound but are fundamentally different in principles and in clinical usages, including the relationship in consideration, the blood tests required, the definition of mismatches, and the applicability to antibody therapy and autologous settings. D, donor typing and selection algorithms.