“Out damned spot”: ex vivo purging revisited

The use of autologous transplantation in malignant disease is beset with 2 problems: the perceived lack of “graft-versus-tumor” effect and the potential contamination of the graft with residual tumor cells. Many approaches have been employed to purge autologous grafts including chemotherapy, freeze/thawing, direct removal or lysis of tumor cells with specific monoclonal antibodies, and indirect purging where CD34⁺ cells have been selected from grafts. It is fair to say that no universally successful purging strategy has been developed. Koh and colleagues (page 4067) report their latest findings using allogeneic natural killer (NK) cells to purge autologous grafts ex vivo. NK-cell activation is controlled by inhibitory and activating signaling molecules expressed on the cell surface.¹  Murine and human NK cells are inhibited from lysis by ligation of killer immunoglobulin-like receptors (KIRs) with specific class I major histocompatibility complex (MHC) molecules on the target cells. In certain MHC-mismatched combinations, the MHC molecule for the specific KIRs on the NK cells is absent from the target cell. This lack of inhibitory signal has been considered sufficient to induce NK-cell–mediated lysis of the target cell. Koh et al have confirmed that KIR-mismatched allogeneic NK cells can indeed lyse tumor cells seeded into autologous grafts and that additional blocking of the KIRs enhanced this. What is notable is that the lysis appears restricted to the tumor cells since the normal hematopoietic stem cells (HSCs) within the graft were unaffected and gave rise to complete donor engraftment even when additional KIR blockade was provided. Thus, simple lack of KIR ligation is insufficient to provoke NK-cell killing and some degree of activating signals are required; furthermore, these activating signals are present on many different tumor cell types.

Is this clinically relevant? Apart from the application to ex vivo purging there is the possibility of in vivo use of NK cells for the immunotherapy of leukemia. KIR incompatibility in recipients of haploidentical transplants has already been shown to be associated with survival advantage in acute myeloid leukemia (AML)²  in the absence of graft-versus-host disease (GVHD). This may be interpreted as tumor-restricted lysis though there was no evidence of mixed chimerism and thus the lysis may have been broadly antihematopoietic. Donor KIR-mismatched NK cells can mediate graft-versus-leukemia (GVL) in vivo after allogeneic bone marrow transplantation (BMT) without GVHD.³ 

The evidence that normal HSCs and tumor cells have differential susceptibility to NK-cell lysis is also significant in the autologous setting, where KIR mismatch is absent. The importance of autologous NK cells in maintenance of remission after chemotherapy alone in AML has recently been demonstrated⁴  and must be due to NK-cell activation by the tumor cell, possibly via Hsp70.⁵ 

The use of NK cells in tumor immunotherapy is resurgent and the study by Koh et al is informative for ex vivo, in vivo, allogeneic, and autologous studies.