Natural killer cells in patients with hematologic malignancies, solid tumors and in recipients of hematopoietic stem cell transplantation

Natural killer cells represent the first line of defense against infections and tumors and can be derived from various sources including: bone marrow, peripheral blood, specific types of human stem cells, and certain cell lines. The functions of natural killer cells are influenced by: several cytokines, activating and inhibitory receptors, as well as other immune cells such as dendritic cells and mesenchymal stem cells.

Natural killer cells are attractive candidates for adoptive cellular therapy in patients with hematologic malignancies and solid tumors in addition to recipients of various forms of hematopoietic stem cell transplantation as they enhance antitumor effects without causing graft versus host disease. Several clinical trials have shown safety and efficacy of natural killer cell products obtained from autologous as well as allogeneic sources and used in conjunction with cytotoxic chemotherapy, monoclonal antibodies and novel agents.

The following review, which includes extensive literature review on several aspects of natural killer cells, will give particular attention to: the rising role of natural killer cell therapies in patients with malignant hematological disorders, solid tumors and in recipients of stem cell therapies; preparation and manufacture of natural killer cell products; challenges facing the utilization of this form of cellular therapy including evolution of resistance; and maneuvers that can be employed to enhance the efficacy of natural killer cell therapies as well as suggested solutions to resolve the remaining challenges.

Natural killer (NK) cells are large granular lymphocytes that are: CD3^-, CD56⁺, CD16⁺, CD94⁺ and NKp46⁺ [1-4]. They comprise 5% - 25% of peripheral blood (PB) lymphocytes. Additionally, NK cells are the third population of lymphoid cells and they represent the first line of defense against infections and tumors [3-9]. They develop from common progenitors and differentiate from hematopoietic stem cells (HSCs) in the bone marrow (BM) but diverge into distinct subsets which differ in cytokine production, cytotoxicity, homing and memory traits [10,11]. NK cells can be derived from: BM, PB, cryopreserved umbilical cord blood (UCB), human embryonic stem cells (hESCs), induced pluripotent stem cells (iPSCs), and various cell lines such as NK-92 and KHYG-1 [1,12,13].

NK cells have been traditionally classified as short-lived innate lymphocytes or part of the innate immune system because, unlike T and B cells, NK cells do not express receptors that require gene rearrangements to generate receptor diversity and specificity [6]. Recently, it has been shown that NK cells exhibit many of the features associated with adaptive immunity such as: (1) expansion of pathogen-specific cells, (2) generation of long-lasting memory cells that persist after cognate antigen encounter, (3) ability to mount an enhanced secondary recall response to rechallenge, and (4) having distinct gene regulatory functions by adaptive NK cells [6,14].

Classifications and subsets of NK cells

NK cells can be classified into different subsets according to: function, immunophenotyping or surface markers, and cytokine production as shown in table 1 [7,15-18]. NK cells are usually classified into: naïve CD56^bright CD16^dim CD3^dim cells; and mature CD56^dim CD16^bright CD3^dim cells. However, a third population (lymphoid tissue-resident NK cells: CD69⁺, CXCR6⁺) has been identified recently [15-18]. Also, they are divided into: cytotoxic, tolerant, and regulatory subsets with diverse phenotypes and functions in various body tissues [17].

Invariant NK T (iNKT) cells represent a small population of αβ T lymphocytes that are derived from HSCs [19-21]. Stimulated iNKT cells rapidly release large amounts of cytokines such as interferon (IFN)-γ and interleukin (IL)-4 that activate NK cells, dendritic cells (DCs), CD4 helper and CD8 cytotoxic T cells [19]. iNKT cells protect against graft versus host disease (GVHD) by inhibition of proliferation of alloreactive T cells, and promotion of expansion of T-regulatory cells [20-23]. Also, iNKT cells can protect against cancer despite that cancer patients have reduced numbers and function of iNKT cells [24]. The histone demethylase UTX regulates the development of iNKT cells through multiple epigenetic mechanisms [25].

Cytokine-induced killer (CIK) cells were first described by Schmidt-Wolf in 1991 [26]. CIK cells are NK-like T-cells that can be: derived from PB mononuclear cells (PBMNCs); costimulated and expanded using cytokines for 14-21 days in vitro, and generated from healthy donors or patients with leukemia [27]. CIK cells have a dual functional capacity of both T cells and NK cells ultimately leading to the secretion of perforin and granzymes for the execution of cytotoxicity [9,26]. They have demonstrated cytotoxicity against a variety of malignant or leukemic cells with no or only minor effects on normal hematopoietic progenitor cells [26,27]. Allogeneic CIK cells retain the ability to produce graft versus leukemia (GVL) effect while generating minimal GVHD [27]. CIK cell infusion comprises a safe and a feasible novel immunotherapeutic approach that targets relapse or minimal residual disease (MRD) following hematopoietic stem cell transplantation (HSCT) for hematologic malignancies (HMs) [27,28]. In a recently published study that included 91 patients with various HMs relapsing after allogeneic HSCT, conventional donor lymphocyte infusion (DLI) given to 55 patients was compared to CIK given to 36 patients, the outcome of CIK therapy was superior to that of DLI with higher overall survival (OS), less relapses, and less acute GVHD [26]. However, optimal timing and cell dosage of CIK cells need to be determined [28]. NK cells and CIK cells can be expanded using a variety of clinical-grade approaches before infusion into patients with cancer. Also, CIK cells have the following advantages over other forms of cell therapy: (1) ease of in vitro propagation, and (2) obviating the need of exogenous administration of IL-2 for in vivo priming [9].

NK cells that express inhibitory killer cell immunoglobulin-like receptors (KIRs), for which respective major histocompatibility complex (MHC) class I ligand is absent on leukemic target cells, can exert alloreactivity in vitro and in vivo [2]. The inhibitory KIRs permit NK cells to recognize self-human leukocyte antigen (HLA) class I molecules and provide inhibitory signals to preclude killing of the target cells [29].

Development, maturation and function of NK cells

Development and maturation of NK cells: NK cell maturation is a continuous process that is initiated in the BM and is continued in peripheral tissues [30]. Mature NK cells reside in the BM and secondary lymphoid tissues such as lymph nodes [7]. Thus, BM is not only the place for development, maturation, self-renewal and persistence of NK cells, but also it provides a line of defense against infections and tumors through utilizing the presence of NK cells [30]. The stages of development of NK cells include: NK progenitor cells, immature NK cells, and mature NK cells. Also, NK cell development requires the presence of the family of 6 γ chain containing receptors: IL-2R, IL-4R, IL-7R, Il-9R, IL-15R, and IL-21R [7]. Activation of NK cells is regulated by several receptors including: (1) KIRs, (2) CD94⁺-NKG2 family, (3) leukocyte immunoglobulin-like receptors, (4) natural cytotoxicity receptors, and (5) FcγRIIIa (CD16⁺) [12]. After activation, NK cells have the following 3 main functions to participate in immune defense: (1) ability to mediate contact-dependent killing of target cells through mobilization of highly specialized organelles or lytic granules in NK cells; (2) production of soluble factors such as cytokines, chemokines and other regulators to promote direct antidisease effects and to further induce and regulate immunity; and (3) promotion and regulation of immunity through contact-dependent costimulatory and regulatory mechanisms [31].

The molecular mechanisms that regulate NK cell cytotoxicity can be divided into 3 main processes: target cell recognition, target cell contact and immunological synapse, and NK cell-induced target cell death [7]. NK cells contain cytoplasmic granules that include perforin which is a membrane-disrupting protein and granzyme family of serine proteases and they express death ligands which are members of the tumor necrosis factor (TNF) superfamily including: (1) TNF-related apoptosis-inducing ligand, (2) Fas ligand (FasL) which is expressed on activated NK cells and cytotoxic T lymphocytes, and (3) TNF-like weak inducer of apoptosis [12]. The following factors may potentially influence immune response: expression genes, genetic mutations, alteration of resistance to immunotherapies in leukemic cells, tumor microenvironment consisting of T regulatory cells (T-regs), tumor-associated macrophages, myeloid-derived suppressor cells, and production of cytokines [32]. NK cells selectively kill target cells that downregulate MHC molecules and/or upregulate other activating ligands such as MHC class I chain (MIC)-related antigens: MICA, MICB, and UL-16 binding protein. The use of perforin and granzyme in cytolytic killing is a major mechanism in the elimination of infected cells and tumor cells by NK cells [12].

NK cell function and dysfunction: The development and functions of NK cells are controlled by various cytokines which operate at different stages by regulating distinct signaling pathways. However, IL-15 and IL-21 are instrumental in driving NK cell differentiation and maturation [9,17]. NK cells are heterogeneous with respect to functional activity and cell-surface antigen presentation [15]. Functions of NK cells are greatly influenced by the cellular microenvironment mainly due to cytokines, chemokines and cell-to-cell interaction [17,33]. The main functions of NK cells are: elimination of infected cells during the adoptive phase of immune responses, recognition and destruction of cancer cells in patients with HMs and solid tumors that have evaded cytotoxic T-lymphocytes, reducing the incidence of GVHD following HSCT, and regulation of the outcome of pregnancy [16,17,33]. NK cell function is finely tuned by receptors that transduce inhibitory or activating signals and that recognize both foreign and self-antigens expressed by NK cell-susceptible targets [9,6]. The activating and inhibitory receptors of NK cells recognize molecular structures on cell surfaces [29]. NK cell dysfunction, which predisposes to infections by herpes viruses, is associated with: genetic or hereditary disorders; chronic disorders such as autoimmune diseases and metastatic cancer; exposure to occupational chemicals; and certain viral infections such as human immunodeficiency virus (HIV) [3,8]. The causes of dysfunctional NK cells can be broadly divided into two main categories: (1) quantitative deficiencies with low numbers or absent NK cells and (2) qualitative deficiencies with abnormal function of NK cells. The etiological causes and the associations of quantitative as well as qualitative NK cell deficiencies are shown in tables 2 and 3 [31,34-52].

Control of NK cell functions: NK cells: kill susceptible targets, functionally interact with different immune cells, sense pathogens using certain Toll-like receptors (TLRs), adapt their responses to the local environment, and mount some degree of immunological memory [53]. Several elements in cell metabolism such as: glucose-driven glycolysis, and oxidative metabolism play critical roles in NK cell development, education, memory generation, and antitumor as well as antiviral effector functions [54].

The Foxo family of genes is critical to many aspects of cellular physiology. Foxo1 acts as a negative checkpoint on NK cell maturation as it represses NK cell specification and proliferation. In a mouse model, hematopoietic-specific deletion of Foxo1 has been found to promote NK cell specification and proliferation [55]. Hypoxic environment may profoundly influence the nature of NK cell infiltrate and its effects on immune-mediated responses within tumor tissues by: influencing the NK cell transcriptome, affecting the immunoregulatory functions of NK cells, and changing the chemotactic responses of different NK cell subsets [56].

The development and functions of NK cells are controlled by various cytokines such as: fms-like tyrosine kinase 3 ligand, kit ligand, IL-3, IL-10, IL-18, transforming growth factor-β (TGF-β), and common-γ chain family cytokine which operate at different stages by regulating distinct signaling pathways [17]. IL-2, IL-12, IL-15, and IL-18 regulate to phenotype, proliferation, and function of human cytokine-induced memory-like (CIML) NK cells, while inhibition of NK cell activation and function is achieved by TGF-β, and IL-10 [57,58]. Exposure of CIML-NK cells to ruxolitinib produces very high levels of cytokines such as: TNF-α and IFN-γ [58]. Members of the emerging NK cell checkpoint family including: IL-1R8, IL-15 signaling, cytokine-induced SHz (CIS) containing protein, and TGF-β RIL act as potential drug targets to boost global NK cell function [59].

The micro-RNA, has-miRNA-146a-5p, may be involved in the regulation of KIR expression. Targeting has-miRNA-146a-5p downregulates KIR and may improve NK-mediated antitumor activity [60]. Vitamins: A, B, C, D, and E as well as natural compounds such as: genistein, curcumin, ginseng extract, garlic extract, resveratrol, ashwagandha extract, ingenol mebutate, kumquat pericarp extract, prostratin, lectins, and polysaccharides affect NK cell function by: increasing cytotoxicity, enhancing proliferation, and stimulating cytokine production [61]. Strategies that can be employed to enhance NK cell function are included in table 4 [1,2,32,62-73].

Education, licencing and memory responses of NK cells

Education and licensing of NK cells refer to the process of acquiring killing or cytokine production after encountering and recognizing self-HLA molecules [9]. The aim of NK cell education is to distinguish self from non-self [7]. Alterations in cellular metabolism or cellular metabolic pathways play a role in NK cell education and immune cell functions. Thus, educated and uneducated NK cell subsets can be distinguished by their metabolic profile regarding glucose metabolism [74]. Approximately 10% - 20% of NK cells remain unlicensed and functionally hyporesponsive due to lack of receptors for self-MHC. However, unlicensed NK cells become alloreactive upon encountering cytokines in a recipient environment after adoptive transfer into recipients of HSCT [9].

Immunological memory is the ability of the immune system to respond rapidly and provide protection against a previously encountered pathogen. After initial infection, long-lived memory cells are generated and they display heightened responses upon secondary challenge with the same pathogen [75]. The process of memory formation in T cells is generally divided into 3 main phases: expansion, contraction, and memory phases [75,76]. Also, there are 3 major types of NK cell memory or memory-like responses: (1) antigen-specific memory responses, (2) cytomegalovirus (CMV)-adaptive memory responses, and (3) cytokine-induced antigen-independent memory-like responses. However, NK cells might adapt their inhibitory responses for memory [77].

Interactions between NK cells and other immune cells

NK cells interact with other immune cells including: DCs, mesenchymal stem cells (MSCs), macrophages, T cells, and endothelial cells [78]. Activated NK cells regulate the following actions of DCs: cytokine-producing capacity; Th-cell polarizing ability; chemokine expression; migration, editing and maturation; and stimulatory functions. On the other hand, activated DCs are required for the execution of innate and effector functions of NK cells including NK cell differentiation [79-84]. NK cells and DCs mutually influence each other and the bidirectional crosstalk between the 2 components of the innate immune system, that may be well coordinated, plays a pivotal role in the regulation of immune defense against viruses such as CMV, parasites such as Leishmania amazoneasis, and tumors [80,82,85-90]. Activated NK cells are capable of killing MSCs while MSCs can: alter the phenotype of NK cells and modulate NK cell function; increase release of IFN-γ from NK cells in order to enhance defense against infections at the sites of injury; and inhibit IL-2 induced NK cell proliferation [91-97]. The crosstalk or interaction between BM-MSCs and NK cells is complicated and can impact the immunobiology of both cell types [98,99].

GVT effects and GVHD in HSCT

GVT effects and GVHD in allogeneic HSCT: Allogeneic HSCT is an established and a potentially curative therapeutic modality not only for high-risk (HR) HMs such as relapsed/refractory (R/R) acute leukemia but also for several benign conditions [100-102]. GVHD, disease relapse, and infectious complications represent the main causes of morbidity and mortality associated with the use of HSCT [100,103]. Allogeneic HSCT benefits HMs by providing graft versus tumor (GVT) or GVL effects which are mediated by donor-derived T cells as well as NK cells [100,101,104,105]. Thus, the success of allogeneic HSCT depends upon: (1) the infusion of benign stem cells and lymphocytes that are capable of inducing GVT/GVL effect, and (2) the ability of the engrafted immune system to remove residual leukemia or tumor cells through the GVT/GVL effect [104,105].

T lymphocytes are responsible for the development of both acute and chronic GVHD post-HSCT. However, removal or depletion of these cells from allografts results in higher rates of graft failure and relapse of HMs [102]. T cells and NK cells are important in providing GVL effect and both cell types are amenable to ex vivo manipulation and clinical manufacture in order to make them more versatile immunotherapeutics [106]. The immunopathophysiology of chronic GVHD is more complicated than that of acute GVHD as it involves complex interactions between several alloreactive and dysregulated cells including: T lymphocytes, B cells, macrophages, neutrophils, DCs, and NK cells [107,108]. Although donor T lymphocytes mainly contribute to the pathophysiology of chronic GVHD, donor B cells contribute to the development of this complication but to a lesser extent [103]. However, specific immune effector cells within the BM allograft have been recognized to correlate with: GVHD, relapse of HM, and OS [103,104].

Prevention of GVHD can be provided by: calcineurin inhibitors, various T-cell depletion strategies, and immunomodulatory agents [106]. Strategies to balance immune responses favoring the development of GVT/GVL effects without inducing harmful GVHD are required to: protect against relapse of HMs, treat persistent disease, and improve disease-free survival (DFS) [104]. A thorough flow cytometric analysis of donor cells for phenotypic and allogeneic activity may be valuable in evaluating the pretransplant risk of development of severe acute GVHD [109]. DLI is an effective therapy for patients with acute leukemia relapsing post-allogeneic HSCT, but the development of severe acute GVHD after DLI infusions remains an obstacle to the success of this therapeutic procedure [100]. The use of oncolytic viruses such as myxoma virus is associated with control of GVHD while preserving or even augmenting GVT/GVL effects [101]. In a murine system, adenoviral vectors have been found to stimulate NK cells and ultimately cause enhanced antitumor activity in the absence of transgene expression [110]. Although GVT/GVL effects occur spontaneously following allogeneic HSCT, such effects can be induced by the following maneuvers: (1) optimal donor selection and optimal conditioning therapy; (2) administration of GVHD prophylaxis; (3) DLI infusion; (4) use of allogeneic mismatched NK cells; (5) activated DCs of leukemic origin; (6) generation of multi-leukemia antigen specific T cells or chimeric antigen receptor (CAR)-modified T cells; (7) use of pharmacological agents such as: tyrosine kinase inhibitors (TKIs), fms-like tyrosine kinase 3 (FLT3) inhibitors, and immune check point inhibitors (ICPIs); and (8) allograft engineering by using: negative CD3/CD19 depletion, as well as CD34 positive selection [102,105,106,111,112].

Autologous GVT effects and autologous GVHD: In the setting of autologous HSCT, not only GVT effect but also biopsy proven GVHD can be encountered [113-118]. Hence, GHVD is an increasingly recognized complication of autologous HSCT [113,118]. The risk factors for the development of autologous GVHD include: (1) primary disease as multiple myeloma (MM), non-Hodgkin lymphoma (NHL), and breast cancer; (2) second autologous HSCT; (3) heavily pretreated patients; (4) female gender; (5) high CD34⁺ stem cell dose infused; (6) achievement of high levels of absolute lymphocyte count (ALC) after autologous HSCT; and (7) use of certain novel agents such as bortezomib, lenalidomide, pomalidomide and alemtuzumab [113-115,117-119]. Autologous GHVD may occur spontaneously, but can be induced by a number of immunosuppressive drugs [113]. It resembles GVHD occurring after allogeneic HSCT not only clinically but also histologically and can be treated with the same immunosuppressive therapies [113,116,117,119].

The role of NK cells in GVT/GVL and GVHD: Immune reconstitution is a critical process following HSCT [120]. NK cells, regulatory T cells (T regs), MSCs, regulatory B cells (B regs), and myeloid cells play significant roles in posttransplant immune regulation, but NK cells are the first donor-derived lymphocyte population to recover following allogeneic HSCT [120,121]. However, spontaneous and full recovery of the numbers of NK cells as well as their cytotoxic function generally occurs within the first month after allogeneic HSCT [121,122]. Interestingly, patients having acute and chronic GVHD have delayed constitution of NK cells [123]. Acute GVHD has been shown to impair the constitution of total and CD56dim NK cells at 3 months after allogeneic HSCT [124].

NK cells mediate a strong GVT/GVL effect and play an important role in GVHD as adoptively transferred NK cells can potentially reduce acute GVHD by killing host antigen-presenting cells [121]. Various subsets of NK cells play specific functions in the development and manipulation of GVHD. Examples include: (1) NKG2A⁺ NK cells play a crucial role during early stages of GVHD following HSCT; (2) NKG2C-expressing NK cells have a direct role in the early control of CMV reactivation after allogeneic HSCT; (3) c-Kit- CD 27^- CD11b⁺ NK cell population is capable of significantly dimishing GVHD in the setting of fully mismatched allogeneic HSCT by supporting GVL effects whilst providing protection against the development of GVHD; (4) CD56^bright NK regulatory cells have been shown to have a much stronger impact on filgrastim-stimulated PB apheresis products than on filgrastim-stimulated BM products as lower proportions of CD56^bright NK regulatory cells result in higher rate of chronic GVHD seen in filgrastim-stimulated PB apheresis; (5) iNK cell dose in allogeneic grafts is associated with significant improvement in GVHD and GVHD-free progression-free survival; (6) failure to reconstitute iNK cells following allogeneic HSCT is associated with higher risk of primary disease relapse; (7) low doses of adoptively transferred donor CD4⁺ iNK cells protect from GVHD while preserving GVT effects as these cells inhibit proliferation of alloreactive T cells and promote robust expansion of donor T regs; and (8) higher doses of CD4^– iNKT cells in PB stem cell allografts are associated with protection from acute GVHD [21,22,121,124-127].

The administration of ex vivo expanded NK cells in the early post-transplant period has been shown to be safe, feasible, and encouraging in lowering the rate of relapse of HMs following HSCT [128]. The number of reinfused autologous graft NK cells in the apheresis product affects the ALC recovery following autologous HSCT [129]. Also, the counts of early lymphocyte subsets following allogeneic HSCT have an association with not only acute GVHD but also with post-HSCT outcome [130]. In recipients of HLA-identical allogeneic HSCT, a well-defined donor-activating NK cell receptor genotype protects against relapse of leukemia after HSCT [131]. Mismatch of NK cell receptors and ligands during allogeneic HSCT may be utilized to prevent relapse of leukemia post-HSCT [132]. The following factors have been found to be important in enhancing NK cell alloreactivity: (1) high stem cell dose, (2) extensive T cell depletion, (3) no GVHD prophylaxis, (4) myeloid malignancy as primary disease, (5) donor source: related versus unrelated and haploidentical, (6) ethnicity, (7) disease status at transplantation, and (8) differences in the definition of alloreactivity: phenotypic, genotypic, and mismatching algorithm [133].

Antitumor effects of NK cells

IFN-γ, a product of activated T lymphocytes and NK cells, is a major immunoregulatory cytokine as it produces: TNF-α, IL-1 and IL-2 [134]. In animal models, bortezomib has been shown to enhance the antitumor effects of adoptively infused NK cells, ultimately resulting in delayed tumor growth and prolongation of survival [135]. The antitumor effects of NK cells are mediated by the synergistic actions of IFN-γ and IL-2 [134]. However, the antitumor effects of bortezomib can be potentiated by the administration of IL-2 [135]. Allogeneic NK cells have been found to mediate more potent antitumor effects than H2-matched syngeneic NK cells without causing adverse hematologic effects thus allogeneic NK cells can be used to purge tumor cells contaminating the BM [136].

Antibody-dependent cell-mediated cytotoxicity (ADCC) is a cell-mediated immune defense mechanism in which effector immune cells actively lyse antibody-coated target cells [137]. The ADCC of tumor cells is utilized in the treatment of various malignancies that overexpress unique antigens such as: neuroblastoma, breast carcinoma, and B-cell NHL [137,138]. Among immune cells, only NK cells are known to be major effectors of antibody-mediated ADCC [137]. Thus, ADCC mediated by NK cells is presumed to be a key effector function [139]. However, NK cell education does not influence ADCC levels, but does contribute to antibody-dependent NK cell activation [140]. NK cells harbor the activating differentiation molecule FcγRIIIa, which is also known as CD16⁺a or CD32c, on their cell surfaces. Hence, NK cells are considered the most important effectors of ADCC in humans [141]. Monoclonal antibodies such as rituximab and the humanized anti-CD123 monoclonal antibody utilize different mechanisms including ADCC to destroy cancer cells [138,142-144]. ADCC may be improved by drugs that are able to enhance activity of NK cells as monoclonal antibodies have a significant role in the induction of ADCC against tumor cells. However, it is difficult to distinguish between the target effect that many monoclonal antibodies exert against specific cell membrane receptors and the ADCC effect that can be induced by monoclonal antibodies [141]. Despite the success of the approved second generation monoclonal antibodies for the treatment several malignant diseases, efforts are made to further augment ADCC in vivo by antibody engineering [139].

Epigenetics and NK cells

Cancer arises as a result of an accumulation of genetic mutations and epigenetic abnormalities that are regulated by histone deacetylases (HDACs) which control chromatin condensation and gene expression [145]. HDACs are enzymes that regulate diverse cellular processes such as: (1) expression of genes that are implicated in tumor initiation, progression, and antitumor responses; and (2) cell proliferation, survival, and immune pathways through deacetylation of their protein targets and modulation of gene expression and protein activity [146,147]. However, HDAC activities are frequently dysregulated in patients with cancer [146].

Treatment with epigenetic agents is a novel therapeutic approach that modulates gene expression by targeting the: DNA methylation machinery, histone covalent modification, and micro-RNAs. However, major limitations of epigenetic therapy are the lack of specificity and the consequent global induction of epigenetic changes [148,149]. Treatment with epigenetic drugs can reduce chemotherapy resistance in patients with HMs and solid tumors. Hence, epigenetic agents can be added to cytotoxic chemotherapy or targeted therapies in order to derive chemosensitization benefit [148,150,151]. HDAC inhibitors have been developed as novel anticancer agents in the treatment of myelodysplastic syndromes (MDSs), acute myeloid leukemia (AML), lymphomas, as well as solid tumors and they can be used alone as monotherapies or in combination with other anticancer therapies [145,148,152]. However, HDAC inhibitors can exert deleterious effects on NK cell function, which may weaken immune surveillance and facilitate relapse of the primary malignant disease in patients treated with HDAC inhibitors [145]. Valproic acid, a HDAC inhibitor, has been found to impede the lytic activity of NK cells against leukemic cells in a dose-dependent manner [152].

NKG2D is a major receptor of NK cells and plays a critical role in tumor immunosurveillance. Expression of NKG2D by NK cells is inhibited by valproic acid and enhanced by entinostat which is a narrow-spectrum HDAC inhibitor [153]. NK cells express the activating receptor NKG2D which provides one mechanism by which NK cells recognize their targets [154]. Cancer cells which survive the direct induction of cell death by HDAC inhibitors become targets for NKG2D-expressing cells such as NK cells, γδ cells and CD8 T cells [155]. Treatment of NK cells with HDAC inhibitors can recover histone acetylation, and restore NK cell activity as well as interferon-γ production [156]. The H3K27me3 histone demethylase UTX regulates development of iNKT cells through multiple epigenetic mechanisms [25]. Combination immunotherapy using HDAC inhibitors and NK cells activated by IL-15 could improve the immune recognition of both therapy-sensitive and therapy-resistant cells of Ewing’s sarcoma and sensitize tumor cells for NK cell cytotoxicity [157].

Preparation of NK cells and Improvement of their potency

Preparation of NK cells for clinical use is a complicated process that involves multiple steps and depends on the following factors: (1) the desired primary source of NK cells such as PB, BM, UCB, hESCs, and iPSCs; (2) the specific cell line to be used; and (3) whether to use autologous or allogeneic NK cells. However, details of the steps involved in preparation, manufacture, delivery and tracing of NK cells are shown in table 5 [12,158-162]. The types of available NK cell-based immunotherapies are shown in table 6 [12].

Strategies that can be used to improve NK cell immunotherapies include: (1) optimal donor selection; (2) combination with cytokine stimulation or ICPIs; (3) use of drugs that enhance NK cell antitumor activity or sensitize malignant cells to NK cells; (4) use of bispecific or trispecific killer engagers; (5) use of adoptively infused allogeneic NK cells in haploidentical transplantation; (6) advancing the field of ex vivo manipulation and genetic engineering; (7) and priming NK cells and using extracellular vesicles (ECVs) derived from NK cells. The details are included in table 4 [1,2,32,70-73].

Current and future therapeutic uses of NK cells

The potential indications for NK cell therapies: The main potential therapeutic uses for NK cell therapies include: (1) autologous and allogeneic HSCT by providing GVT effect and enhancing engraftment; (2) various HMs including: MM, acute and chronic leukemia, as well as lymphomas; (3) various solid tumors by providing antitumor effects; (4) solid organ transplantation; (5) autoimmune and inflammatory disorders; (6) pregnancy and reproduction; (7) various types of infections; and (8) bronchial asthma [9,12,71,78,163-168].

Examples of non-specific adoptive immunotherapies include: CD3/CD28-activated DLIs; lymphokine-activated killer cells; anti-CD3^-activated killer cells; tumor-infiltrating lymphocytes; in vitro-generated or selected tumor cytotoxic T lymphocytes; γ/δ-T cells; and NK cells [26]. However, NK cells are attractive candidates for adoptive cellular therapy in patients with HMs and solid tumors, as well as in recipients of allogeneic HSCT as they enhance GVT/GVL effects without causing GVHD [1,26,97,169-174].

NK cells in MM: MM is characterized by gradual immune dysregulation which impairs function of: T and B cells, NK cells, as well as antigen presenting or DCs thus allowing malignant plasma cells to escape immunosurveillance [175]. MM cells exhibit specific immunoevasive strategies in order to circumvent and attenuate NK cell function [176]. Transformed plasma cells in MM are susceptible to NK cell-mediated killing by engagement of tumor ligands for activating receptors or missing self-recognition [174-176]. Despite the advancements in novel therapies and autologous HSCT, MM remains an incurable and difficult-to-treat HM due to drug resistance predisposed to by the immunosuppressive microenvironment and clonal evolution which favor disease progression [175]. However, allogeneic HSCT which is associated with significant morbidity and mortality is the only potentially curative therapeutic modality due to its potent graft versus myeloma effect [172].

In MM, NK cell function has been shown to be diminished by specific factors that are active in the tumor microenvironment (TME) [177]. Also, relatively high levels of HLA molecules are expressed in MM [178]. The recognition of plasma cells in patients with MM by NK cells is regulated by: HLA class I/HLA-E, NKG2D receptor and possibly NKG2A receptors, and natural cytotoxicity [178-180]. HLA class I may be involved in the resistance of myeloma cells to NK cell lysis thus contributing to the immune escape and consequently drug resistance in R/R-MM [181]. Tumor progression in patients with MM is associated with decreased expression of activating receptors [182]. However, infusion of large numbers of expanded NK cells which has been shown to be feasible and safe may be critical to boost their activity in vivo [183]. In patients with MM, NK cells have been used in several trials in the setting of autologous as well as allogeneic HSCT as NK cells elicit cytotoxic effects against MM cells and as KIR-ligand mismatch m tables 2 and ay improve the outcome of allogeneic HSCT [172,176,184-186]. NK cell killing of tumor cells in MM can be augmented by: ICPIs, therapeutic antibodies such as daratumomab, immunomodulatory agents such as lenalidomide, indoleamine 2,3 dioxygenase inhibitors, and adoptive transfer of unmanipulated or CAR-engineered NK cells [171,175].

Daratumumab is a CD38-specific monoclonal antibody that induces death of MM cells via various mechanisms including: ADCC, and complement-dependent cytotoxicity [187]. On one hand, it has been shown that daratumumab augments NK cell cytotoxicity against target cells having high expression of CD38 surface markers [177]. On the other hand, it has been shown that daratumumab treatment may induce NK cell depletion thus making daratumumab-treated myeloma patients susceptible to infectious complications such as bacterial infections and reactivation of viruses belonging to the herpes group [188]. Additionally, daratumumab-medicated ADCC can be significantly improved by lenalidomide mainly due to the potent capacity of the latter to activate NK cells [187].

In a one-year follow-up study on patients with MM, it has been shown that lenalidomide therapy neither activated NK cells nor it improved their capacity to degranulate or secrete cytokines and that discontinuation of the drug did not reduce the effector function of NK cells [189]. Immunomodulatory agents and proteasome inhibitor have been shown to upregulate the expression of the activating receptors NKG2D and DNAM-1 on NK cells [182]. Studies have shown that: (1) bortezomib, the first generation proteasome inhibitor, can significantly enhance the sensitivity of MM cells to allogeneic as well as autologous NK cell-mediated lysis, and (2) carfilzomib, the second generation proteasome inhibitor, can also enhance NK cell degranulation and significantly enhance the sensitivity of myeloma cells to NK cell-mediated lysis through downregulation of the expression of newly formed HLA class I on MM cells [186,190].

The excellent safety and feasibility profiles of NK cells make them interesting candidates in combination therapy with novel agents in order to enhance their clinical efficacy in the treatment of MM patients [178]. Blockade or inhibition of KIR2D in patients with smoldering myeloma by IPH2101 monoclonal antibody has been shown to enhance NK cell killing of myeloma cell lines [191]. In a meta-analysis that included 12 clinical trials and 592 patients, adjuvant immunotherapy with DCs and CIK cells has been shown to be safe and efficacious in enhancing the efficacy of chemotherapy administered to patients with MM [192]. Studies have shown that early recovery of donor derived lymphocytes and NK cells after autologous HSCT is associated with improved long-term outcome in patients with MM [193,194]. In a phase I clinical trial that included 10 patients with MM, the administration of allogeneic NK cells derived from UCB has been shown to be safe and feasible [184]. CAR-transduced NK cells and bispecific antibodies utilizing NK cells hold great promise and potential against MM [182].

NK cells in AML: In patients with AML, NK cells are frequently defective thus leading to tumor escape where the continuous cross-talk between AML and NK cells predisposes to immune escape of leukemia and eventually disease relapse [195,196]. Various mechanisms are potentially involved in the inhibition of NK cell function in AML patients and these include: (1) defects in the normal lymphopoiesis, (2) reduction in the expression of activating receptors through cell-to-cell contacts, and (3) production of immunosuppressive soluble agents by AML blast cells [195]. Therefore, it is of vital importance to restore NK cell activity in AML patients by: (1) stimulating immunosurveillance mediated by NK cells, (2) combining conventional chemotherapy with immune mediators that include NK cells, and (3) genetic modification of CIK cells with chimeric receptors specific for the CD33 myeloid antigen [195,197]. A novel sβ3-integrin protein has been identified in patients with AML and this protein has been shown to enhance the cytotoxic activity of NK cells against AML blasts after stimulation with IL-2 and IL-15 [198]. Cytokine-induced memory-like NK cells have been shown to exhibit enhanced antitumor effects against AML blasts [199].

So far, allogeneic HSCT is the only curative therapeutic intervention in patients with AML. However, the most common cause of death in AML patients subjected to allogeneic HSCT is disease relapse [32]. NK cell immunotherapies using: adoptive NK cells, cytokines-based immunotherapies, ICPIs, and bispecific as well as trispecific engagers have the potential to significantly enhance the ability of conventional therapies to eliminate AML after HSCT [32,199]. Initial reports of haploidentical HSCT in AML patients showed favorable effects of alloreactive NK cells on disease relapse and survival by promoting engraftment, enhancing GVL effect and reducing the incidence of GVHD. Subsequently, studies have shown either no difference in the incidence of GVHD or adverse outcomes related to GVHD, infections and disease relapse. Therefore, selecting the most appropriate alloreactive NK cell model and selective expansion of a particular NK cell subset may become vital in restoring NK cell function in the post-HSCT period [133].

Donor KIR-group B profiles and the homozygous of centromeric motif B are the most preferable KIR gene content motifs for HSCT [200]. The benefits of unidirectional graft versus host (GVH)-KIR ligand incompatibility in T-cell replete haploidentical HSCT may be masked by the relatively favorable transplantation outcomes of bidirectionally KIR-ligand matched recipients. So, AML patients with unidirectional host versus graft KIR ligand incompatibility have experienced significantly higher relapse rates and decreased DFS compared to patients with bidirectionally KIR ligand matched group of patients [201].

In AML patients receiving allogeneic HSCT, alloreactive T cells and NK cells mediate GVL effect thus improving disease outcome by enhancing eradication of leukemia and preventing AML relapse [202]. The use of the combination of cytosine arabinoside cytotoxic chemotherapy and ex vivo activation of NK cells has the potential to be a feasible approach to treat AML relapsing after HSCT [203]. A study that included 112 patients with HR-AML, subjected to haploidentical HSCT: 51 patients received allografts from donors with alloreactive NK cells and 61 patients received allografts from donors with no alloreactive NK cells. The study showed that transplantation of NK-alloreactive donors was associated with significantly: lower relapse rate in patients transplanted in complete remission (CR), better event free survival (EFS) in patients transplanted in relapse or in CR, and lower risk of mortality [204].

Studies have shown that IL-2-activated haploidentical NK cell treatment can induce CR in 30% - 50% of patients with R/R-AML and thus can be used as effective bridging therapy to the potentially curative allogeneic HSCT [205-207]. In the first-in-human phase I and phase II trials using either (IV) intravenous (26 patients) or (SC) subcutaneous (16 patients) recombinant human (rh) IL-15 administered with haplo-NK cell therapy after lymphodepletion with cyclophosphamide and fludarabine (CY-Flu) to treat R/R-AML, the following results were obtained: (1) CR was achieved in 35% of patients with CR rates of 32% in recipients of IV IL-15 and 40% in recipients of SC IL-15, (2) cytokine release syndrome (CRS) was reported in 56% of recipients of SC Il-15, (3) SC dosing of rh IL-15 after lymphodepletion prolonged drug exposure thus leading to CRC and neurotoxicity, and (4) strategies to augment in vivo expansion of NK cells included: lymphodepletion with CY-Flu before NK cell infusion, depletion of T regs with an IL-2 Diphtheria toxin, and in vivo use of recombinant cytokines [207].

NK cells in Acute Lymphoblastic Leukemia (ALL): In patients with malignant diseases, NK cell immunosurveillance may be impaired resulting in tumor escape and disease progression [208]. Precursor B-lineage ALL is associated with immune deficiencies that can be further exacerbated by cytotoxic chemotherapy [209]. The classical prognostic factors in patients with ALL include: (1) age of patient; (2) white blood cell count at presentation; (3) cytogenetic profile; (4) early treatment response; and (5) MRD evaluated by flow cytometry on day 15 of induction chemotherapy [210]. However, presence of NK cells in the BM can serve as an additional prognostic factor in patients with ALL [210]. In these patients, post-induction MRD is a very important prognostic marker. Hence, therapies targeting MRD have been shown to improve the outcome of ALL patients [211]. The following NK-related factors have been found to strongly correlate with post-induction MRD: Fas ligand, granzyme B, NKp46, and KIR2DL5A in NK cells; as well as PI-9 in blast cells [211]. Compared to AML, only a minority of ALL blasts are susceptible to NK cell-mediated killing [212]. Hence, ALL has been considered resistant to NK cell-mediated lysis [208,213]. Not all NK cells are equally cytotoxic against leukemic cells because of differences in receptor gene content and surface expression [211]. Additionally, adult and pediatric ALL blasts show a difference in the expression of the known ligands for NK cell activating receptors. Also, specific phenotypic patterns of expression are associated with molecularly defined subgroups of ALL patients such as Philadelphia chromosome positive ALL patients [212].

CD56⁺ NK cells obtained from PB or BM of patients with ALL at various stages of disease, including diagnosis, remission, and relapse, can be expanded ex vivo. The expanded NK cells have been shown to kill autologous ALL blasts obtained from the same patient spontaneously or through antibody-dependent cytotoxicity [209]. Strategies that have been employed to sensitize ALL blasts to NK cell killing include: (1) stimulation of NK cells by TLR-9 activated plasmacytoid DCs which have the capacity to reinforce the GVL effect of HSCT; (2) drugs such as bortezomib, valproate, and troglitzaone; and (3) genetic modification of NK cells to overcome ALL tumor cell antigens [208,213,214].

Adaptive NK cells have the following distinguishing features: (1) they are highly differentiated NK cells, (2) they expand naturally in vivo in response to human CMV infection, (3) they carry unique repertoires of inhibitory KIRs, and (4) they display strong cytotoxicity against tumor cells. Hence, adaptive NK cells hold promise for treatment of refractory ALL either as a bridge to HSCT or as a form of treatment for patients who lack stem cell donors as these cells are capable of eradicating residual blast cells [215]. Several models of adoptive transfer of mature allogeneic NK cells have been used in transplant and non-transplant settings in patients with ALL and AML. Safety and feasibility of such models have been determined but their effectiveness has not been found to be uniform [2].

NK cells in chronic leukemias: Currently, TKIs are the standard therapeutic modality in patients with chronic myeloid leukemia (CML). However, the use of these novel agents is associated with numerous adverse effects [216]. The majority of patients with CML have quantitative as well as qualitative defects in the NK cell compartment of their immune system. Therefore, developing strategies to exploit NK cells for immunotherapy in patients with CML is of vital importance [216,217]. Successful discontinuation of imatinib in patients with CML is associated with high proportion of mature cytokine-producing NK cells [218].

Chronic lymphocytic leukemia (CLL) patients have severe immune defects that predispose them to a variety of infectious complications [219]. Also, chronic exposure of NK cells to a significant tumor burden has its own consequences on the phenotype and function of these cells, thus rendering NK cells unable to counteract not only infections but also the chronic leukemia itself [220]. Strategies to augment NK cell function, which are being evaluated in clinical trials, will have positive effects on both CLL and the associated infectious complications [219,220]. Blocking NKG2A on NK cells of CLL by monalizumab can enhance NK cell activity by restoring the direct cytotoxicity function [221].

NK cells in lymphomas: In patients with B-cell NHL: the number of NK cells in PB may affect the outcome of patients receiving anti-CD 20-based immunotherapy, functional NK cells infiltrate tissue biopsies and their presence is tissues correlates with survival of these patients [222,223]. The combined genetic and microfluidic assays can evaluate the sensitivity of cells of B-cell NHL to NK cell-based cytotoxicity [224]. T-regs, iNK cells, and B-regs are involved in the pathogenesis of NHL [225]. Additionally, iNK cells and Th17 (T cells that produce IL-17) inhibit tumor growth of B-cell NHL while T regs support tumor growth in this category of lymphoma [226].

The combination of rituximab and lirilumab, which is a NK cell agonist that causes KIR blockade, has potent antilymphoma activity [227]. Also, a recent study showed that prior to the administration of combined chemotherapy and rituximab, an anti-CD 20 monoclonal antibody, patients with NHL had lower levels of B-regs, and to a lesseer extent T-regs but not iNK cells in the PB compared to controls, while following complete remission of NHL, the levels of circulating T-regs, iNK cells, and B-regs were elevated [225].

Expansion of PD-1⁺ NK cells and PD-L1⁺ monocytes/macrophages is more prominent in classical Hodgkin lymphoma (cHL) than in diffuse large B-cell lymphoma (DLBCL). Also, programmed cell death protein 1 (PD-1) blockade reverses the immune evasion mediated by the interaction between PD-1⁺ NK cells and programmed death-ligand 1 (PD-L1)⁺ monocytes/macrophages [228,229]. NK cells are inhibited directly by malignant B-cells and indirectly by PD-L1/PD-L2 expressing tumor-associated macrophages. Thus, cells of cHL are more sensitive to PD-1 blockade than DLBCL cells [228,229]. In patients with HL, studies have shown that: (1) the immunosuppressive nature of the TME specifically inhibits proliferation and activity of NK cells, (2) malignant Reed-Sternberg cells and other components of the TME express ligands to inhibit NK cells, and (3) although NK cell deficiency begins at the tumor site, ultimately it progresses systematically in patients with advanced disease as the secretion of cytokines and chemokines mediates the systemic immunosuppression. Thus, strategies to reactivate NK cell function or those aimed at blocking the evasive mechanisms displayed by the TME may ultimately identify new immunotherapeutic targets [230,231]. In patients with HL, CD 123 (IL-3Rα) is frequently expressed by malignant cells and the combination of NK cells and the fully humanized anti-CD 123 monoclonal antibody (CSL362) represents a promising future therapeutic strategy [144].

NK cells in HSCT: Allogeneic HSCT has revolutionized the treatment of HMs, but the use of this potentially curative therapy is limited by: GVHD, infections and relapse of the primary disease [27,106,170,232]. Relapse of HM remains the leading cause of treatment failure of allogeneic HSCT [26,32]. Calcineurin inhibitors, T-cell depletion and immunomodulators prevent GVHD but have negative impact on GVL effects [106].

Separating GVL effects from GVHD is of special interest in non-specific cell-based immunotherapy which may eradicate molecular disease and prevent relapse following allogeneic HSCT particularly when leukemia burden is low [26,32]. The recognition of missing-self on target cells is crucial for promoting NK cell-mediated GVL effects [2]. NK cells have a central role in tumor-cell surveillance but leukemic cells have great capacity to escape NK cell recognition and killing thus limiting the use of NK cells in cellular immunotherapy [72].

NK cells are the first subset of donor-derived lymphocytes to reconstitute after HSCT thus they may protect against relapse in the early months following HSCT by providing GVL effect without causing GVHD [1,29,169,232]. In recipients of HSCT, NK cells provide protection against bacterial infections at mucosal barriers and viral infections particularly those caused by CMV both of which are associated with significant morbidity and mortality in this group of patients [233]. CMV infection stimulates and expands a distinctive population of NK cells that expresses the NKG2C receptors and exhibits enhanced effector functions [234]. However, the initial wave of NK cells that reconstitutes in the early post-transplantation period is rather dysfunctional [234,235]. Nevertheless, rapid immune recovery after allogeneic HSCT predicts clinical outcome [236]. Although immune recovery post-HSCT has been evaluated in the PB, it can also be evaluated by BM examination. Also, evaluation of absolute lymphocyte subsets, particularly the status of NK cell recovery in the BM on day 21 after HSCT predicts clinical outcome [236]. Presence of a KIR B haplotype in donors and lack of recipient HLA-C epitope in recipients provide protection against relapse of AML following allogeneic HSCT. Additionally, strategies combining NK cell infusions with CD16⁺-binding antibodies or immune engagers could make NK cell antigen specific [234].

Although the initial studies on the use of autologous NK cells were disappointing, the use of allogeneic NK cells has resulted in favorable outcomes in both transplant and non-transplant settings and this led to the advancement of NK immunotherapy over the last decade [1,169]. Donor NK cells play significant roles in: promotion of hematopoietic engraftment following HSCT, preventing relapse of HM post-allogeneic HSCT by mediating GVL effects, and regulation of GVHD by suppression of alloreactive T-cell responses [169,232]. Enhancement of GVL without increasing the incidence of GVHD can be achieved by adopting the following maneuvers: optimal donor selection, optimal conditioning therapy, administration of GVHD prophylaxis, and administration of T-cells and donor-derived NK cells which are amenable to ex vivo manipulation and clinical manufacture [106]. Studies have shown that: augmentation of T-cell alloreactivity may be influenced by NK cells in recipients of T-cell deleted allografts, and while immunosuppression with sirolimus and expansion of T-regulatory cells may decrease the incidence of acute GVHD by suppressing the development of T-cell mediated alloreactivity [170,237,238].

NK cell infusions derived from PB and UCB contain contaminating T-cells whose stimulation by cytokines produced by NK cells may trigger GVHD in vivo thus limiting the safety and efficacy of NK cell infusions in allogeneic HSCT. However, NK cells obtained from iPSCs, hESCs, and NK cell lines are free of contamination with T and B cells thus offering alternative sources of NK cells that can be used in adoptive immunotherapy [164]. Owing to the short-lived and limited in vivo activity of the effectors involved, non-specific immunotherapy is dependent on repeat administrations [26].

NK cells are negatively regulated by MHC class I-specific inhibitory receptors [239]. NK cells lacking inhibitory receptors for self MHC class I ligands are hyporesponsive. However, responsive NK cells in the donors of allogeneic HSCT may become aberrantly activated and functionally competent in the recipients of HLA-identical allografts [240]. Because NK cell education after HSCT is driven by donor ligands, NK cell specific to a ligand present in the donor but absent in the recipient could remain responsive even long after transplantation and may exert a long-term GVL effect [241]. Donor selection for the KIR B haplotype of the centromeric motifs can improve survival in recipients of HLA-identical sibling allografts [242]. KIRs recognize groups of HLA class I alleles and missing expression of the KIR ligand on mismatched allogeneic cells can trigger NK cell alloreactivity [239]. Most NK cells express the inhibitory KIRs [243]. Allowing a greater T-cell content in the allograft could reduce the infection-related morbidity and mortality that are associated with extensive T cell depletion in mismatched transplants thus facilitating the use of mismatched allogeneic HSCT in elderly individuals and in heavily pretreated patients [239]. In a study the included 2062 patients with AML, CML, and MDSs subjected to unrelated allogeneic HSCT, missing KIR ligands were found to be associated with lower rates of relapse and higher incidences of GVHD [243].

Although, the use of post-transplant cyclophosphamide (PTC) as GVHD prophylaxis has revolutionized haploidentical HSCT, PTC eliminates most mature donor NK cells infused in the graft including alloreactive NK cells [244,245]. However, recovery of NK cells after haploidentical HSCT is greatly influenced by other subsets of immune cells and by drugs used in the post-transplant period [244]. Although NK cell infusion given to 10 patients with AML in CR1 led to a 2 year EFS of 100%, a prospective phase II study that was performed in 2 centers and that included 16 patients with HR leukemia and multiply relapsed tumors subjected to haploidentical HSCT showed that preemptive NK cell infusions had no apparent effect on the rates of graft failure or disease relapse [246,247]. NK cells generated after haploidentical HSCT are blocked at an immature state that has specific phenotype and impaired functioning and this has negative impact on immune responsiveness and clinical outcome following HSCT [248]. Haploidentical HSCT offers the benefits of rapid and nearly universal donor availability and has been accepted worldwide as an alternative therapeutic modality for patients with HR-HMs who do not have HLA-identical sibling donors [249]. A phase I clinical trial using membrane bound IL-21 ex vivo-expanded donor derived NK cells given to patients with HR-HMs receiving haploidentical HSCT showed that NK cell infusions were associated with: improved NK cell function, lower relapse rate, and low incidence of viral infections in the post-transplant period [250]. In patients with acute leukemia subjected to haploidentical allogeneic grafts: early CMV reactivation and expression of CD56^bright CD16^dim DNAM1⁺ NK cells were associated with GVL effect reflected by lower relapse rates [251]. Also, in a retrospective study that included 246 patients with HMs subjected to allogeneic HSCT, NK cell reconstitution was associated with lower rate of disease relapse after HSCT particularly in patients with CMV reactivation [252].

Unidirectional KIR ligand incompatibility in the host versus graft effect has a detrimental effect on the outcome of T-cell-replete haploidentical HSCT in adults with AML [201]. Studies have shown that: (1) NK cells mediate GVL effect which is crucial to cure patients with HR-acute leukemia subjected to haploidentical HSCT; (2) graft manipulation based on depletion of αβ T cells and B cells allows infusion of fully mature and alloreactive NK cells; and (3) reconstitution of iNK cells in PB following haploidentical HSCT in children with HMs is associated with disease control [253,254].

In a study that included 45 recipients of HLA matched HSCT grafts from related and unrelated donors subjected to reduced intensity conditioning (RIC) including antithymocyte globulin, the following results were obtained: NK cells recovered quickly after HSCT regardless the donor type; rapid quantitative reconstitution of the NK cell compartment despite the administration of potent immunosuppression in the conditioning therapy and in the post-HSCT period; the rapidly reemerging NK cells remained immature for > 6 months; and the rapid reconstitution of cytokine production correlated with lower relapse rates and prolongation of OS [235]. Manipulation of type of lymphocytes may be instrumental in reducing the relatively high relapse rate following allogeneic HSCT with RIC [234]. In a study that included 909 patients with AML and MDSs subjected to RIC-allogeneic HSCT from unrelated donors, it was shown that KIR-HLA combinations recapitulated some but not all KIR-HLA effects observed in myeloablative allografts [255]. Also, in a study that included 282 patients with HMs subjected to allogeneic HSCT with non-myeloablative conditioning therapy: engraftment of donor NK cells correlated with lower risk of relapse and no GVHD, while engraftment of donor T cells correlated with higher risk of GVHD [256].

NK cells in solid tumors: Cancer cells frequently produce platelet derived growth factor receptor (PDGFR)-β which, through autocrine and paracrine PDGFR-β signaling, promotes tumor growth, cell proliferation, metastasis, stromal recruitment, angiogenesis, and epithelial-mesenchymal transition [257]. NK cells play a major role in the immune response to certain malignancies by several mechanisms that include: (1) directly by secretion of potent immune mediators such as targeted secretion of cytokines or cytotoxic granules to cause cytolysis of transformed cells, (2) indirectly by orchestrating anti-tumor immune responses to prevent metastatic spread by engagement of the activating receptor NKp46 on NK cells, (3) the human immunoreceptor NKp44 expressed on NK cells and the innate lymphoid cells recognize PDGF-DD produced by tumor cells and this plays a major part in the control of tumor growth by NK cells, and (4) NK cells recruit conventional type I DCs into the TME to promote immune control of tumors [6,7,72,257-259]. Thus, NK cells play key roles in innate and adaptive responses through unique NK cell activation mechanisms during early host defense against viruses and tumors by performing 2 major roles: contact dependent cytotoxicity and cytokine production for immune modulation [6,258,260].

Target cell apoptosis is primarily mediated by perforin and genzyme B. Also, the regulation of the immune responses is mediated by secretion of cytokines such as: IFN-γ, TNF-α, IL-1, IL-3, and granulocyte-macrophage colony-stimulating factor [3,260]. NK cells are attractive candidates for adoptive cellular therapy in: (1) cancer: HMs such as acute leukemia, and solid tumors, with either CAR-engineered NK cells or combining NK cells with CD16⁺ binding antibodies or immune engagers; and (2) allogeneic HSCT including haploidentical allografts to protect against disease relapse by enhancing GVL effect without causing GVHD [1,2,10,72,106,234,242,261]. NK cells have crucial role in protection against cancer relapse [13,232]. The antitumor activity of NK cells is regulated by a sophisticated network of activating and inhibitory receptors [13]. Viral and non-viral methods have been adopted to genetically engineer NK cells in order to improve their antitumor activity [13].

The metastatic spread of malignant cells to distant sites, which is regulated by TME and systemic processes that include immunosurveillance, is a principal cause of cancer-related death [262]. NK cells are crucial for immunosurveillance and are essential for controlling metastatic dissemination of cancer cells [59,262]. However, NK cells are highly dysfunctional and reduced in number in patients with solid tumors [71]. Also, antitumor response of NK cells faces plenty of limitations [73]. Adoptive transfer of large numbers of cytolytic NK cells to induce antitumor responses is widely explored in cancer immunotherapy [71]. In preclinical studies: (1) ECVs derived from NK cells have shown promising antitumor effects, and (2) genetic engineering of NK cells to express CARs to redirect their antitumor specificity has shown significant promise [73,263]. Monoclonal antibodies and bispecific killer engagers may enhance specificity by inhibition of CD16⁺ shedding and enhance NK cytotoxicity [264]. PDGF-DD-induced NK cytokines such as TNF-α and IFN-γ can trigger tumor cell-cycle arrest in mouse models [257].

Several studies have shown the following effects of NK cells in different cancers: NK cells have been shown to be effective in patients with advanced lung cancer; Herceptin-treated NK cells have been found to have therapeutic potential in the treatment of patients with human epidermal growth factor receptor 2 (Her2⁺) and Herceptin-intolerant breast cancer; expanded and cryopreserved NK cells are promising candidates for cellular immunotherapy in patients with pancreatic carcinoma; NK cells inhibit metastesis of ovarian carcinoma cells in murine models; CIK cells may become a novel therapeutic modality for rhabdomyosarcoma after allogeneic HSCT; CAR-engineered CIK cells may become valuable in the treatment of HR soft tissue sarcoma in children; and expanded and CIK cells are safe and well tolerated and they enhance cytotoxicity against gastric carcinoma [162,265-271]. A meta-analysis, that included 29 clinical trials involving 2610 patients, has been shown that the combination of CIK/DC-CIK immunotherapy and cytotoxic chemotherapy: enhances the immune function of patients, alleviates the adverse effects of chemotherapy, and improves the overall response, disease control, and quality of life [272].

Glioblastoma (GB) is the most aggressive primary brain malignancy in adults and it carries poor prognosis as it is still incurable [273,274]. Studies have shown that CAR-engineered and PD-1 inhibited NK cells exhibit antitumor effects and can induce apoptosis of GB cells [273-276]. GB stem-like cells are more susceptible to lysis by NK cells than differentiated GB cells [277]. In animal models, pretreatment of GB with bortezomib has been shown to promote NK cell cytotoxicity and to inhibit tumor growth ultimately leading to prolonged survival [278]. Varicella zoster virus (VZV) is the only virus with negative correlation with GB [169,279]. NK cells represent a significant barrier to oncolytic herpes simplex virus therapy of GB [280]. Oncolytic virotherapy provokes the antitumor activity of NK cells by triggering antiviral immune responses [281]. NK cell immunotherapy has shown promising effects in the treatment of childhood solid tumors such as: rhabdomyosarcoma, Ewing’s sarcoma, and neuroblastoma due to prominent GVT effects both in transplant setting and in combination with antibodies [282].

NK cells and infections: NK cells are involved in the host immune response against infections caused by viral, bacterial, and fungal pathogens which are significant causes of morbidity and mortality in immunocompromised hosts. Hence, there is interest in strengthening the immune response in immunocompromised individuals by: (1) use of cytokines or growth factors, or (2) adoptive cellular therapies including: donor granulocytes, pathogen-specific T cells, and adoptively transferred or transfused NK cells [163]. In polymicrobial sepsis, there is immunoparalysis that includes impairment of both the number as well as the function of NK cells [283].

NK cells can act as rheostats, regulating T cell-mediated support for the antiviral CD8 T cells that control viral pathogenesis and persistence [284]. NK cells play a vital role in host defense against HIV, herpes viruses, hepatitis B virus (HBV), and hepatitis C virus (HCV) [3]. The hepatic NK cell populations that are involved in controlling HCV infection may also be involved in the control of HCV-associated liver damage [285]. NKG2C-expressing NK cells are involved in the early control of CMV reactivation following allogeneic HSCT [124]. Two NK cell subsets might play a critical role in the immune response against Dengue virus infection [286]. NK cells are important in herpes virus infections as patients with deficiencies of NK cells experience systemic and life-threatening infections that are caused by herpes viruses [287]. VZV productively infects human NK cells and actively manipulates their phenotype thus NK cells play a potential role in the pathogenesis of VZV infections [288].

In immunocompromised hosts including recipients of HSCT and cancer patients, recovery of the immune system has a major impact on the outcome of infectious complications that represent a significant cause of morbidity and mortality in this group of patients [163,287]. Strategies that can be employed to strengthen the host immune response to counteract infections include the use of: growth factors; specific cytokines and adoptive cellular therapies such as granulocyte transfusions, pathogen-specific T-cells and adoptive transfer of NK cells [163]. NK cells have a major role to play in controlling various infectious complications including: (1) viral infections such as: HIV, and HBV; (2) parasitic infections such as: toxoplasmosis, malaria, leishmaniasis, and trypanosomiasis; and (3) bacterial sepsis due to: Streptococcus pneumoniae, and Escherichia coli [3,78,283].

NK cells play a major role in the immune response to certain viral infections by: (1) direct cytolysis or killing of virus-infected cells in order to rapidly control viral infection, and (2) secretion of potent immune mediators such as IFN-γ and other cytokines [3,8,289,290]. NK cells share features with long-lived adaptive immune cells and this can impact disease pathogenesis through inhibition of adaptive immune responses by virus-specific T and B cells as NK cells are potent regulators of antiviral T and B cell responses [290]. NK cells can produce persistent memory in response to certain viral infections particularly those caused by CMV [14]. NK cells have multiple mechanisms to kill virus-infected cells through the engagement of extracellular death receptors, and through exocytosis of cytotoxic granules. However, mediation of cytolysis occurs through: engagement of death receptors expressed on target cells, and expression by NK cells of multiple extracellular ligands including fas ligand and TNF-related apoptosis-induced ligand ultimately resulting in apoptosis of the target cells [8].

VZV actively manipulates the NK cell phenotype through productive infection. NK cells have a potential role in VZV pathogenesis and they are implicated in controlling infections caused by VZV [288]. Decreased NK cell function is associated with: (1) several genetic or hereditary disorders, (2) several chronic disorders such as: chronic fatigue syndrome, depression, autoimmune diseases, metastatic cancer, and exposure to occupational chemicals, and (3) certain viral infections such as HIV [3]. Although NK cell deficiencies are rare, they predispose to infections by herpes viruses [8]. VZV-infects NK cells using multiple entry mechanisms and causes: (1) cell to cell interaction with VZV-infected epithelial cells during early encounter or entry, and (2) subsequent modulation of NK cell function and phenotype resulting in stimulation of chemokine receptors and CD57 expression and inhibition of the expression of CD56, CD 16 and FcVRIII [8,288].

Challenges facing the clinical utilization of NK cells

Acquisition of large numbers of mature and functional NK cells that can be derived and differentiated from UCB-CD3⁺HSCs is easily accessible, but optimal clinical protocols for NK cell therapies in leukemia and other cancers are still lacking [291]. Although the use of viral vectors has achieved the highest level of efficiency of gene transfer of NK cells, the utilization of nonviral vectors and other gene transfer approaches such as: electroporation, lipofection, nanoparticles, and trogocytosis are emerging [13,261]. Despite the extensive number of preclinical studies only a handful of NK cell-based therapies have progressed to the clinic. As of mid-2018, there were only 8 registered clinical trials utilizing genetically engineered NK cells [13]. For successful translation of genetically modified NK cells, issues related to: viral vector safety, efficacy, and compliance with regulatory guidelines become vitally important [13,73]. Despite the numerous challenges associated with the preclinical and clinical development of NK cell-based therapies for cancer, NK cells have several unique immunological properties that enable them to be potentially effective means for cancer immunotherapy [292].

Maturation of NK cells starts in the BM then continues in peripheral tissues. Upon differentiation, mature NK cells migrate outside the BM and peculiar subsets of NK cells home back to or localize in the BM compartment to perform specific functions [30]. Human BM-resident NK cells have a unique transcriptional profile and resemble resident memory CD8⁺ T cells [18]. Persistence of NK cells in vivo can be improved by: IL-2 and IL-15 [12]. Homing of NK cells to target tissues can be improved by: (1) transfection with CCR7mRNA electroporation, (2) transfer of CCR7 protein from feeder cells using trogocytosis, and (3) genetic modification that targets homing receptors [12]. The major contributing factors for resistance to NK cell therapies are shown in table 7 [164,293].

The role of NK cell therapy in patients with HMs, solid tumors and in recipients of various forms of HSCT is evolving rapidly. The utilization of NK cells in conjunction with cytotoxic chemotherapy, and novel therapies including monoclonal antibodies has increased the response rates of patients with HMs and solid tumors to the standard therapeutic modalities. In general, cancer patients have qualitative and quantitative NK cell deficiencies, but there are several maneuvers to enhance NK cell function in these immunocompromised individuals. There are several limitations facing the clinical utilization of NK cell-base immunotherapies that need to be resolved. These challenges include: evolution of resistance to therapy; having control quality and safety measures; design of specific protocols for preparation and manufacture; banking and cryopreservation of harvested NK cell products; administration and therapeutic use of each type and source of NK cells; and finally tracing of the infused NK cells.