Innate lymphoid cells (ILCs) are lymphocytes that do not belong to the T or B cell lineages and are relatively rare in lymphoid tissues and peripheral blood. They are defined by the absence of adaptive antigen receptors [1]. One of their principal functions is to initiate an early immune response upon pathogen invasion, preceding the activation of antigen-specific lymphocytes [2]. ILCs are predominantly located at mucosal surfaces of non-lymphoid organs [3]. Remarkably, even in the absence of recombination-activating genes (Rag-1 and Rag-2) and without the expression of conventional T and B cell antigen receptors, ILCs are still capable of maturing. Although they do not possess immunological memory, they are classified as components of the innate immune system [4, 5]. The mechanisms underlying their activation, however, remain largely undefined [6, 7].

ILCs are categorized into four major subgroups. Natural killer (NK) cells and type 1 ILCs represent two subsets of ILCs that express the T-box transcription factor TBX21 (T-bet). Although NK cells and type 1 ILC1s both belong to the family of ILCs, they exhibit distinct functional, phenotypic, and developmental characteristics.

These cells exert their functions primarily through the production of interferon-gamma (IFN-γ). NK cells are involved in type 1 immune responses and are classified as cytotoxic ILCs. In contrast, ILC1s are widely distributed across various tissues, including the liver, adipose tissue, intestine, and salivary glands. They are activated by soluble cytokines such as interleukin (IL)-15, IL-12, and IL-18. ILC1s contribute to host defense against viral and intracellular bacterial infections by producing effector cytokines and initiating a rapid, first-line immune response [8].

The second group of ILCs (ILC2s) is defined by the expression of the GATA-binding protein 3 (GATA3) and is highly prevalent in mucosal tissues such as the gastrointestinal (GI) tract, lungs, tonsils, and skin. These cells mediate type 2 immune responses by producing IL-5, IL-13, IL-4, and members of the epidermal growth factor family. They also secrete cytokines such as amphiregulin, which play a role in combating helminth infections and regulating tissue repair [9].

ILC3s are characterized by the presence of RAR-related orphan receptor gamma T (RORγt). ILC3s are further subdivided into NKp46⁺ and NKp46^– subsets based on surface marker expression. These cells secrete a range of cytokines and growth factors, including IFN-γ, tumor necrosis factor-alpha (TNF-α), IL-22, IL-17, granulocyte-macrophage colony-stimulating factor (GM-CSF), and heparin-binding epidermal growth factor-like growth factor (HB-EGF) [10, 11]. ILC3s are abundant in the skin, lungs, intestinal mucosa, and mesenteric lymph nodes, where they play a key role in initiating rapid immune responses against extracellular microorganisms and maintaining tissue homeostasis [12].

The final group of ILCs, known as lymphoid tissue inducer (LTi) cells, comprises lymphoid tissue-derived cells that are also dependent on RORγt and originate from fetal liver progenitor cells. These cells contribute to the development of secondary lymphoid organs by promoting lymphoid tissue proliferation, a process mediated by lymphotoxin, a member of the TNF superfamily. LTis are present in various organs and tissues during early embryonic development. By regulating adaptive immune responses and supporting the formation of secondary lymphoid structures, they play a critical role in the establishment of both primary and secondary lymphoid tissues [13, 14].

The differentiation of ILCs from a common progenitor cell is illustrated in Figure 1.

Although the classification of ILCs provides a valuable theoretical framework for understanding their diversity, immune responses can introduce additional complexity into their functional programming. Evidence suggests that certain ILC subsets exhibit functional plasticity—a feature well-documented in T cells [15, 16]—which may be critical for adapting and modulating immune responses to diverse pathogenic stimuli.

In vitro, human RORγt⁺ ILC3s stimulated with IL-2 or IL-15 can differentiate into ILC1-like cells, characterized by upregulation of the transcription factor T-bet and the interleukin-12 receptor β2 (IL-12Rβ2) [17] (Figure 2). These cells subsequently produce IFN-γ in response to IL-12 stimulation.

Furthermore, culturing ILC3s with IL-23 promotes their conversion into ILC1s. Interestingly, IL-23 is also the primary stimulus that induces IL-22 secretion by ILC3s. This seemingly paradoxical effect is enabled by the constitutively high expression of the transcription factor signal transducer and activator of transcription 4 (STAT4) in ILC3s [18, 19]. As a result, sustained exposure to IL-23 activates STAT4 and drives the polarization of ILC3s toward a type 1 phenotype. Some studies also suggest that IL-23 may facilitate the reverse transition—from ILC1s back to ILC3s [17]—although the molecular mechanisms underlying this bidirectional plasticity remain to be elucidated.

Acute myeloid leukemia (AML) is a hematological disease characterized by the growth and proliferation of immature cells. AML utilizes unique immune evasion strategies like those of solid cancers. This section provides an update on recent advances in understanding how AML affects each group of ILCs [20].

Although the anti-tumor effects of NK cells are well established, the functional roles of other ILC subsets in cancer remain incompletely understood [44]. In particular, the role of ILC1s in AML is still unclear, and ongoing research is focused on elucidating their contribution. One of the major challenges in studying ILC1s is the lack of unique, definitive surface markers for their identification.

Early investigations of peripheral blood and bone marrow from AML patients revealed an enrichment of functionally impaired, lineage-negative ILC1s compared to healthy donors (HDs) [45]. More recently, a study identified a phenotypically distinct ILC1-like subset (Lin^–CD56⁺CD94⁺CD16^–CD127⁺) with reduced cytotoxic potential at diagnosis, which appeared to be restored in patients who achieved remission [46]. In this study, only surface markers were used to differentiate conventional NK cells from ILC1 or ILC1-like cells, and the phenotypic profile of the latter overlapped with that of circulating CD56 bright NK cells [46, 47].

A key distinction between human and murine NK cells and ILC1s lies in their transcription factor expression: NK cells typically co-express EOMES and T-bet, whereas ILC1s express only T-bet. In mice, the inhibitory receptor CD200R1 has been proposed as a selective surface marker for liver-resident ILC1s and is notably absent in NK cells [44, 48]. Interestingly, CD200 expression on human AML blasts has been shown to suppress IFN-γ production and reduce cytotoxic activity by engaging CD200R1 on human NK cells [49]. Although the expression pattern of CD200R1 in human ILCs is still under investigation, this interaction may reflect an increased presence of ILC1s in CD200 Hi AML cases [20].

Finally, in murine models, NKp46 expressed on ILC1s mediates direct interactions with tumor cells, enhancing cytotoxicity and promoting the production of TNF and IFN-γ. Deletion of NKp46 results in reduced survival and impaired ILC1-mediated tumor control in a mouse model of AML. This phenotype can be reversed by adoptive transfer of NKp46⁺ ILC1s into NKp46-deficient mice. In humans, NKp46⁺ ILC1s produce higher levels of cytokines and exhibit greater cytotoxicity compared to their NKp46^– counterparts, suggesting that NKp46 plays a similarly critical role in human ILC1 function [50].

Recent studies have revealed a tumor-promoting role for ILC2s. While ILC2s are primarily known for their involvement in allergic responses and anti-helminthic immunity, emerging evidence suggests they may also contribute to leukemogenesis. A study focusing on AML demonstrated that prostaglandin D2 (PGD2), secreted by mesenchymal stem cells (MSCs), activates ILC2s via the chemoattractant receptor-homologous molecule expressed on Th2 cells (CRTH2) [51]. This activation induces the secretion of IL-5, which in turn promotes the expansion of regulatory T cells (Tregs) and the proliferation of hematopoietic stem and progenitor cells (HSPCs). In an AML mouse model, this Treg expansion was associated with reduced survival and accelerated leukemia progression.

In addition, tumor-derived PGD2 and the NKp30 ligand B7-H6 have been shown to activate ILC2s, leading to the secretion of IL-13. This cytokine stimulates the activity of myeloid-derived suppressor cells (MDSCs), whose immunosuppressive functions are well-documented in promoting tumor progression [52]. These findings suggest that therapeutic strategies targeting the PGD2-ILC2-Treg or PGD2-ILC2-MDSC axes may hold promise in the treatment of AML.

However, contrasting evidence has emerged from another study, which found no significant differences in ILC2 frequency or IL-5 and IL-13 levels in the peripheral blood of untreated AML patients compared to healthy controls [53]. These discrepancies may be attributed to differences in the tissue microenvironment, as PGD2-producing mesenchymal cells are primarily located in the bone marrow rather than in peripheral blood. This highlights the importance of investigating the primary tissue microenvironment in systemic diseases such as AML, where distinct tissues may exhibit divergent immunological profiles.

Taken together, these investigations highlight the intricate and situation-specific function of ILC2 cells in the development of AML. Clinical data from peripheral blood studies do not support the tumor-promoting axis suggested by preclinical models, which include PGD2-mediated ILC2 activation and the downstream proliferation of immunosuppressive Tregs and MDSCs. This disparity highlights the crucial role that the bone marrow microenvironment (BMME) plays in determining ILC2 function and most likely reflects the compartmentalized nature of immune regulation. To resolve these conflicting findings and improve therapeutic targeting of the PGD2-ILC2 axis in AML, future research should give priority to tissue-specific analyses.

Despite growing evidence that ILC3s play a significant role in post-chemotherapy prognosis and in the pathogenesis of graft-versus-host disease (GVHD), studies specifically investigating ILC3s in AML remain limited. One study reported a marked reduction in natural cytotoxic receptor-positive (NCR⁺) ILC3s—but not NCR^– ILC3s—in the peripheral blood of treatment-naïve AML patients [54]. In this study, ILC3s were defined as Lin^–CD127⁺CRTH2^–CD117⁺NKp46⁺/^– cells, a phenotype that overlaps with that of immature NK cells and ILC precursors circulating in the bloodstream. Interestingly, no significant differences in IL-17A or IL-22 levels were observed between the ILC compartments of AML patients and HDs [55].

Emerging evidence suggests that NKp44 may be a more reliable marker than CD117 for identifying human ILC3 populations, as NKp44⁺ cells are typically absent from the circulation of healthy individuals. In AML patients who responded to standard chemotherapy, the frequency of Lin^–CD127⁺CRTH2^–CD117⁺NKp46⁺ cells was comparable to that of healthy controls. In contrast, non-responders exhibited a reduced percentage of these cells, suggesting a potential prognostic value for this ILC3 subset.

Beyond their role in AML pathophysiology, ILC3s have also been detected during post-induction chemotherapy and hematopoietic stem cell transplantation (HSCT). A study evaluating the reconstitution of ILC subsets following induction chemotherapy and allogeneic HSCT (allo-HSCT) found that donor-derived ILC1s, ILC2s, and NKp44^– ILC3s reconstituted more rapidly and at higher levels than other ILC subsets. NKp44⁺ ILC3s were also observed in the peripheral blood of AML patients undergoing these treatments. These ILC populations expressed activation markers such as CD69 and homing receptors for the gut and skin, including α4β7 integrin, CCR6, CCR10, and cutaneous lymphocyte-associated antigen (CLA). Their presence following chemotherapy or allo-HSCT was associated with a reduced incidence of GVHD [56].

In murine models, NCR⁺ ILC3s have been shown to promote intestinal tissue regeneration and prevent bacterial translocation—mechanisms that contribute to the mitigation of GVHD [57]. These effects are mediated through IL-22-dependent pathways. Indeed, IL-22 has been shown to confer protection against GVHD in mouse models of allo-HSCT [58]. Further research is needed to clarify the role of ILC3s in AML biology and to assess their potential as biomarkers for treatment response.

An overview of innate lymphoid cell abnormalities in AML is provided in Table 1.

Recent advances in the understanding of multiple myeloma (MM) have highlighted the pivotal role of the immune system in disease progression. The BMME consists of both cellular components—including bone marrow stromal cells (BMSCs), endothelial cells, osteoclasts, osteoblasts, fibroblasts, T lymphocytes, and dendritic cells—and non-cellular components, such as the extracellular matrix (ECM) and soluble factors including chemokines, cytokines, and growth factors.

The primary function of the bone marrow stroma is to regulate and support the proliferation and differentiation of hematopoietic cells. During MM progression, interactions between microenvironmental cells—particularly endothelial cells and MSCs—and tumor clones are mediated by surface adhesion molecules, receptors, and soluble mediators secreted by these cells. These interactions promote MM cell survival, proliferation, and differentiation [74, 75].

Initial studies of tumor-infiltrating immune cells in MM suggested a dynamic relationship between immune system components and the tumor [76]. While these immune cells were initially believed to exert anti-tumor effects, more recent evidence indicates that they may also contribute to tumor progression. Interactions between ILCs and microenvironmental cells are critical for tumor development and dissemination, as both immune and tumor cells are influenced by cytokines, adhesion molecules, and metalloproteinases [77].

The early immune response to tumor formation involves the activation of cytotoxic mechanisms, recruitment of immune cells, and secretion of cytokines that induce tumor cell apoptosis [78]. However, as neoplastic cells proliferate and dominate the microenvironment, ILCs begin to produce factors that support tumor growth [3, 79]. The ECM plays a dual role in adaptive immunity: it facilitates T-cell migration into tissues while also exerting inhibitory effects on T-cell proliferation [80–82]. Interactions between ILCs and stromal cells are also evident, with ILCs potentially contributing to tumor tissue formation and expansion [83].

As previously discussed, a defining feature of ILCs is their plasticity—the ability to differentiate into various subtypes in response to environmental cues [81]. In MM, this plasticity can modulate ILC function, resulting in either anti-tumor or tumor-promoting activities [84] (Figure 3).

Recent research has shed light on how ILCs may play a role in preventing MM. These insights could lead to novel therapeutic strategies aimed at leveraging the immune system to combat MM.

ILC1s contribute to cancer prevention primarily through the production of IFN-γ [89–92]. IFN-γ induces apoptosis in malignant cells by upregulating the expression of Fas and Fas ligand (FasL) on their surface. Additionally, IFN-γ enhances tumor antigen presentation by increasing the expression of MHC molecules, thereby improving immune recognition and targeting of cancer cells [93]. It also promotes a shift toward Th1-type immune responses while suppressing Th2 cell activity [94].

In the context of MM, IFN-γ has been shown to modulate oncogenic transcription factors within malignant cells. Notably, several studies have demonstrated that IFN-γ can inhibit MM progression with efficacy comparable to that of the corticosteroid dexamethasone [94–101].

Therefore, although early research focused on IFN-γ’s anti-proliferative and immunostimulatory actions in MM, more recent research has shown that these effects can be enhanced or inhibited by complex regulatory networks, depending on the tumor microenvironment (TME) and epigenetic variables.

Although more direct comparison clinical data would support this assertion, IFN-γ’s ability to decrease IL-6 signaling and modify transcription factors supports its therapeutic equivalency to dexamethasone, as suggested in any works.

In patients with NHL, the cytotoxic activity of NK cells is significantly reduced compared to HDs. Multiple studies have identified three NK cell subtypes in the peripheral blood of NHL patients: CD56 bright NK cells, CD16⁺ NK cells, and unconventional CD56 dim (uCD56 dim) NK cells [116–117]. Among these, CD16⁺ NK cells—known for their potent cytotoxic capabilities—are notably less prevalent in NHL patients than in HDs, suggesting a downregulation of this critical subset. Furthermore, CD16⁺ NK cells in NHL patients exhibit increased expression of CD73 and CD39, ectoenzymes involved in ATP hydrolysis and commonly upregulated in inflammatory conditions. Conversely, expression of activation markers such as CD69 and KLRG1 is reduced [116]. In CD56 dim NK cells, only CD69 expression is significantly lower in patients compared to HDs. These findings suggest that the TME in NHL may impair NK cell-mediated antitumor responses.

Within neoplastic lymph nodes of NHL patients, NK cells—particularly CD56 bright and CD16⁺ subsets—show elevated expression of immunosuppressive markers CD73 and CD39 compared to their counterparts in peripheral blood, further supporting the notion that the TME contributes to NK cell dysfunction [116].

In the context of B-cell lymphomas, both type I and type II natural killer T (NKT) cells have been investigated for their roles in lymphoma pathogenesis. Type I NKT cells appear to exert protective antitumor effects, whereas type II NKT cells may suppress immune responses. Although research on NKT cell involvement in hematologic tumor immune evasion is still in its early stages, findings from a murine T-cell lymphoma model indicate that type I NKT cells can inhibit tumor growth. CD1d, a molecule expressed on various human hematopoietic cells, plays a role in antitumor immunity, although its precise function remains to be fully defined. Some hematologic malignancies secrete glycolipids that interfere with CD1d-mediated antigen presentation to NKT cells, thereby facilitating immune escape. Notably, recent studies have shown that type I NKT cells can eliminate EL-4 T-cell lymphoblastic lymphoma cells both in vitro and in vivo in a CD1d-dependent manner [117].

The data show that NHL patients have a significant deficit in NK and NKT cell-mediated immune surveillance. A tumor-driven reprogramming of innate immunity is suggested by the decreased frequency and changed phenotype of cytotoxic CD16⁺ NK cells, which are characterized by decreased activation markers (CD69, KLRG1) and increased expression of inhibitory ectonucleotidases (CD73, CD39). In the TME, where NK cells display an even more marked immunosuppressive character, this dysfunction is further aggravated. A complicated regulatory axis is further highlighted by the distinct roles that type I and type II NKT cells play in regulating antitumor responses, with type I NKT cells showing promise as antineoplastic immune effectors. All of these results suggest that the TME plays a crucial role in regulating the activity of innate lymphocytes.

A recent study has demonstrated direct interactions between ILCs and malignant cells in HL [137]. The research showed that a dual cytokine fusion protein, IL-12–IL-2, activates T cells and NK cells more effectively than single-cytokine formulations. CD3⁺ T cells and CD16⁺ NK cells isolated from peripheral blood exhibited enhanced proliferation in response to the IL-12–IL-2 fusion protein compared to IL-12 alone, underscoring the synergistic effect of these cytokines in amplifying immune responses.

Further analysis revealed that the fusion protein could stimulate T and NK cells upon binding to the membrane of CD30⁺ target cells. Notably, the IL-12–IL-2 fusion protein did not induce IFN-γ release from T cells in the absence of target cells, indicating that its immunostimulatory effects are context-dependent. When targeted to CD30⁺ cells via a combined antibody, the fusion protein retained its biological activity and effectively stimulated T cells, reinforcing its therapeutic potential [137].

The effects of NK cells in both NHL and HL are summarized in Table 3.

While chimeric antigen receptor (CAR) T-cell therapy has shown considerable success in treating B-cell malignancies expressing CD19, its application in AML has been limited. This is due to fundamental biological differences between these diseases, a narrow therapeutic window, the risk of severe adverse effects, and the challenge of identifying universal surface antigens suitable for targeted therapy [138–145].

Allogeneic NK cell therapy represents a distinct and promising form of adoptive cell therapy, particularly when combined with allo-HSCT in patients with relapsed or refractory AML. Notably, one study demonstrated the successful use of IL-2-stimulated allogeneic NK cells [144]. In another approach, NK cells were induced and expanded ex vivo from CD34⁺ HSPCs derived from HLA-matched umbilical cord blood (UCB) [145]. In this study, six out of ten AML patients relapsed at a median of 364 days post-infusion, while four patients remained alive. Importantly, the infused NK cells continued to mature in vivo, acquiring KIRs and CD16 expression.

These findings underscore both the safety and therapeutic potential of NK cell-based therapies. Adverse events of grade 2 or higher were rare [146]. Unlike CAR T-cell therapy, clinical trials involving CAR-engineered NK cells have not reported dose-limiting toxicities, even at doses as high as five billion cells per patient [147]. However, a limitation of NK-92 cells—a continuously growing NK cell line used in some therapies—is their requirement for irradiation prior to infusion to prevent uncontrolled proliferation, necessitating repeated dosing for sustained efficacy. Future strategies may involve genetically engineering NK-92 cells with a “kill switch” to eliminate the need for irradiation.

In a phase I clinical trial, NK cells expanded ex vivo using K562 feeder cells expressing membrane-bound IL-21 (mbIL-21) were well tolerated, with only minor injection-related reactions and limited GVHD symptoms reported [148]. At a median follow-up of 14.7 months, all 13 patients remained in remission, with only one experiencing relapse. The use of mbIL-21 significantly enhanced NK cell proliferation and in vivo persistence. These encouraging results have led to a phase II trial evaluating CSTD002, a haploidentical donor-derived NK cell product generated ex vivo using PM21 nanoparticles in combination with mThbIL-21 and 4-1BB ligand (4-1BBL) [149].

Additionally, research into induced pluripotent stem cell (iPSC)-derived anti-CD19 CAR-NK cells has shown promising results in preclinical models of CD19-expressing lymphoid malignancies [150].

In conclusion, given their intrinsic and particular anti-tumor activity, accessibility as an “off the shelf” cellular therapy, lower costs, and enhanced safety, CAR-NK cells may have advantages over CAR-T cells [151].

Future efforts will focus on customizing CAR-NK cells to specifically target AML cells (Table 4).