The following post was originally published in the November 2015 edition of The Vector, the online newsletter for the American Society of Gene and Cell Therapy.
Chimeric antigen receptors (CARs) have provided a powerful molecular mechanism to target T cells to specific surface antigens on tumor cells. The most promising results have been obtained by groups at the University of Pennsylvania, Memorial Sloan Kettering, and in a few additional small studies around the world, by achieving complete remissions in the 90% range in relapsed acute lymphoblastic leukemia (ALL) using T cells armed with CARs targeting the CD19 antigen. While the efficacy and durability of this cell therapy has been nothing short of astounding, we are now tasked to look beyond CD19 and apply this powerful approach to using cytotoxic cells to kill other types of tumors, including solid tumors which present many additional challenges for a CAR engineered cell.
In this brief overview, I would like to discuss several ways in which NK cells can offer us a path forward as a delivery platform for engineered cell (and ex vivo gene) therapy. The purpose of this discussion is not to provide a detailed overview of NK cell biology. The apparent complexity of NK cell activation as well as the relative dearth of biomolecular insights as to the quantitative dynamics of NK cell activation also remains a challenge. NK cells are often compared to their “cousins,” the T cells, which have historically tended to hog the limelight, as well as take up a large share of government grant and investor resources for academic and industry research, respectively. I would simply like to focus on four different aspects of cytotoxic cell therapy, and provide some contrasts where I feel NK cells can offer promise.
1. Durability
T cells have evolved to maintain their exquisite antigen specificity, as well as maximize the ability of the immune system to retain the benefits of the costly selection mechanism for the best clones by retaining a very small but critical subset as memory T cells to enable a potent recall response, even decades after the original expansion event. NK cells, for the most part, have eschewed such a purpose, as their primary role seems to be surveillance, early detection and essentially to hold the line until the reinforcements can arrive.
As such, NK CAR therapies are unlikely to achieve the durability we have recently seen with CAR-T therapies for ALL. This apparent deficiency may actually prove to be an advantage for cell therapy. The flipside of durability can be toxicity (e.g., durable B cell depletion observed in some CD19 CAR-T studies), and periodic administration of NK CAR therapies according to a regular dosing regimen may provide a more accessible pharmaceutical model for the development of safer treatment options for a broader patient population.
2. Control
NK cell activation is negatively controlled by a diverse set of receptor families, such as KIR, NKG2A, and PD-1. In addition, tumor cells may downregulate ligands for these receptors, or upregulate ligands for activating NK receptors such as DNAM-1 and NKG2D, rendering certain tumors uniquely susceptible to NK cell lysis. Does this natural mode of NK cell-mediated tumor recognition potentially provide an additional mechanism to modify and control the specificity of NK cell therapies? Do allogeneic NK CAR therapies offer a unique advantage? Since NK cells do not promote classic graft versus host disease (GVHD), a certain amount of “mismatch” between donor NK cells and host MHC class I and KIR receptor repertoires may provide enhanced efficacy for NK cell therapies. Matching of donor NK haplotypes with patient genotype may offer a level of personalized cell therapy not accessible by CAR-T cell therapies. Guidelines established for ongoing clinical studies of adoptive transfer of related donor haploidentical allogeneic NK cells may be helpful in guiding future matching of donor and recipient haplotypes for engineered NK cell therapies.
