Cytotoxic cells are the major components of the immune system that counterattack tumor cells. The overexpression of co-inhibitory ligands for specific receptors on cancer cell surfaces to disrupt T cell response is one of the primary mechanisms developed to hijack the immune system. The cell surface molecules CD28, cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), inducible co-stimulator (ICOS), PD1, and PD-1L are basic ligands that induce co-stimulatory or inhibitory signals in T cells to maintain immune system homeostasis ( Figure 2A ) (105). In the early immune response stages, CD28 facilitates and maintains the CD4+ and CD8+ T cell response. CTLA-4 arrests T cell activation by triggering an inhibitory signal within the T cell, affecting critical peripheral T cell tolerance and function (106). CD28 and CTLA-4 share ligands and are necessary to avoid inappropriate or prolonged CD4+ and CD8+ T cell activation. Both molecules have been found to be constitutively expressed in acute myeloid leukemia (AML) blasts at diagnosis and have an increased expression in leukocytes from the peripheral blood of these patients compared with that of healthy controls, likely favoring AML cell escape from T cell activation and its effector functions (107). An evaluation of the CTLA-4 expression revealed that this co-inhibitory molecule was elevated in T cells in patients with high-risk ALL (108); in addition, CTLA-4 solubility was significantly elevated in 70% of B-ALL pediatric patients with active disease (109). Furthermore, CTLA-4 overexpression has been correlated with the percentage of leukemic B cells and poor prognosis in pediatric patients (108, 110), and a high serum CTLA-4 level has been detected in patients with B-ALL who died from the disease (111). Thus, the disputed CTLA-4 expression from ALL cells could be a potential mechanism of immune surveillance escape (109, 112).

PD-1 and PD-L1 overexpression has been reported to evade the host immune system in numerous cancer types (105, 110, 113). PD-L1 and PD-L2 are ligands of PD-1, which is an important inhibitory immune checkpoint that suppresses T cell activity after antigen activation. In fact, CTL function inhibition by PD-1 expression has been observed in patients with AML (113). CTLA-4 and PD-1 expression in hematological malignant cells has been suggested as an immune evasion strategy to promote leukemia blast survival and prevent efficient recognition and destruction by anti-tumor T cells (107, 110). Studies using a mouse model of disseminated AML and in transplanted patients before relapse have shown that sustained inhibitory signaling mediated by CTLA-4 and PD-1 on T cells correlates with a T cell exhaustion stage, reduced T cell effector function, and lower cytotoxicity (114, 115). Data from ALL evidenced a decrease in PD-1 expression on CD4+ and CD8+ T cells after the inhibition of myeloid–epithelial–reproductive tyrosine kinase, a gene associated with the induction of an antiapoptotic gene expression signature in B-ALL cells, leading to increased T cell activation (116). Abnormal expression of checkpoint molecule PD-1 has been reported in BM biopsies from adult patients and in T cells of pediatric cases with ALL (108, 117). In addition to the observations in ALL, upregulation of both malignant and infiltrating immune cells from B cell lymphomas and T cells of peripheral blood mononuclear cells from patients with chronic myeloid leukemia (CML) exhibits the relevance of PD-L1 in hematological malignancies (118–120). PD-L1 overexpression is correlated with poor prognosis in ALL (113) and has been found to be one of the most expressed inhibitory markers in pediatric ALL blast, whose expression is increased in relapsed patients with ALL (108, 115). Other checkpoint molecules are T cell immunoglobulin and mucin domain-containing protein 3 (TIM-3) and lymphocyte-activation gene 3 (LAG-3). TIM-3 is involved in apoptosis, and Tregs expressing TIMP3 have a higher suppressor function than Tregs negative to TIMP3 (121). LAG-3 expression has been detected in highly immunosuppressive T cells and correlates with an increased expression of IL-10 production by Tregs (122, 123). TIM-3 and LAG-3 have been found to be subexpressed in the BM of patients with ALL, in contrast to healthy subjects (117).

ICOS (CD275) is involved in maintaining immune reactions and has a relevant role in Tregs function and differentiation, which protects tumor cells from immune cells in the TME (124, 125). Significant Tregs accumulation in the BM-TME and upregulation of ICOS ligand (ICOS-L) have been observed in AML, suggesting that ICOS-L contributes to the conversion and expansion of Tregs and preserves the immunosuppressive environment. Additionally, ICOS-L and ICOS expression was found to be a predictor of OS and disease-free survival in patients with AML (126). Although there is no information regarding ICOS or ICOS-L in ALL, the relevance of ICOS in ALL immune escape is supported by studies showing that ICOS is part of the intracellular region of the signaling domain complexes that activate and induce cytotoxicity against target cells during chimeric antigen receptor (CAR) T cell immunotherapy (127).

Inflammation is an immune response to body damage and is mediated by cytokines, which are relevant players involved in oncogenic processes, such as cell proliferation, apoptosis inhibition of mutated cells, and promotion and progression of cancer development (128). Cytokines are classified as pro-inflammatory (IL-1β, IL-6, IL-15, IL-17, IL-23, IFN-α, and TNF-α) and anti-inflammatory (IL-4, IL-10, IL-13, and TGF-β). Interestingly, an inflammatory marker analysis of neonatal blood reported that children who developed pre-B ALL had a cytokine signature (lower concentrations of the cytokine IL-8, soluble IL-6 receptor α, and TGF-β1 and higher concentrations of IL-6, IL-17, and IL-18) (129). Increased CCL2 and IL-8 concentrations of T cell-polarizing cytokines (IFN-γ and IL-12) and cytokines associated with infectious processes, such as TNF-α and IL-6, have been detected in patients with ALL at diagnosis, suggesting a pro-inflammatory state (130–133). These findings could be associated with immune cell activation by endogenous molecules that are released after tissue injury or cell death to generate an immune response against cancer (133, 134). The pro-inflammatory environment in the BM of patients with leukemia is facilitated by hematopoietic and stromal cells. However, studies indicate that cancer cells hamper immune activation by creating an anti-inflammatory TME by overproducing anti-inflammatory cytokines and by blocking the release of pro-inflammatory cytokines, thus successfully evading immune surveillance (128, 135). In CML, AML, and ALL, cancer cells express TGF-β and IL-10 to reduce immunogenicity (136, 137). Studies in mouse models with B-ALL have shown that TNF-α is secreted by B-ALL cells, and this leads to increased invasiveness and significant prolongation of surviving leukemia cells (138), which is an important mediator of leukemia-induced NK cell dysfunction. Thus, it is fundamental for NK cell immune evasion in childhood B cell ALL (139). Although IL-4 has shown antitumor effects and ALL cell suppression (140), it has been suggested that IL-4 expression in leukemia cells could reduce immunological recognition by decreasing HLA-class II molecule expression (132).

IFN-γ and interleukin 6 (IL-6) are among the most important cytokines associated with immune response in cancer (141). IFN-γ gene expression is reduced in patients with ALL, suggesting that the immune system is disrupted and leukemia cells may take advantage of defective IFN-γ production to promote escape from immune surveillance (142, 143). IL-6 contributes to lymphocyte and monocyte differentiation and induces antibody secretion by B cells. The low antibody production and decreased cellular immunity derived from abnormal IL-6 expression detected in ALL cases ( Figure 2B ), in addition to the association between single-nucleotide polymorphisms in the IL-6 gene and susceptibility to ALL (the genotype of which correlated with IL-6 serum levels), are evidence of the relevance of this cytokine in this malignant disease (141, 144, 145).

Other cytokines and chemokines, such as IL-1, IL-7, IL-8, CCL2, CXC-10, and CXCL-12, could contribute to immunotolerance (112, 146). In fact, IL-1, IL-7, and CXCL12 expression favors ALL cell surveillance in the BM-TME (146, 147). Through the induction of CCL2 by periostin, this molecule stimulates the proliferation and dissemination of ALL (146, 148).

The abnormal proliferation of immune cell populations is another important mechanism for preventing immune attacks in cancer. Two distinct T cell subsets are involved in the immune system against cancer. The first is CTLs that kill cancer cells, and the second is cells required for the activation and proliferation of APCs.

Tregs (CD4+ CD25+ Foxp3+) are involved in tumor development and progression by inhibiting anti-tumor immunity in the TME (93, 149). Under physiological conditions, Tregs play an essential role in self-tolerance and immune homeostasis processes by suppressing normal and pathological immune responses and by eliminating a broad range of pathogenic microorganisms and malignant cells (85, 150–152). The correlation between tumor-infiltrating Treg levels and prognosis has been described in several malignancies, including ALL, suggesting that Tregs may be involved in the immune evasion process (149, 153–156). Indeed a high number of Tregs in the BM and peripheral blood of ALL cases has been associated with poor prognosis (153, 154, 157, 158). Studies in BM-TME have shown that immunosuppressive cytokines, such as IL-10 and TGF-β, are secreted by Tregs ( Figure 2C ) (159).

One of the biological features of patients with ALL is the presence of severe cytopenia and poor reconstitution of the innate and adaptive immune system. Although normal lymphoid and myeloid cells are present in ALL BM, the early compartment of progenitor hematopoietic cells is reduced in number and activity, including NK cells, MDSCs, and macrophages (93, 130, 160).

NK cells represent 5–20% of the lymphocytes in peripheral blood and are relevant in early antitumor immune response by lysing the tumor cells due to cytokine release. Based on CD56 and CD16 expression, two principal subpopulations of NK cells were identified: cytotoxic NK cells or cNK (CD56dim CD16+) and regulatory NK cells or NKregs (CD56highCd16-). cNK are abundant in peripheral blood (95% of the NK) and inflammation sites and show a higher cytotoxic capacity than NKregs, which predominate in lymphoid nodes (33). Natural cytotoxic receptor expression is a relevant mechanism to stimulate responses against tumor cells and has been observed to be downregulated in ALL BM (33). In recent years, several studies have provided evidence of the fundamental role of NK cells in the onset, development, and establishment of ALL (59, 139, 161, 162). ALL NK cells at diagnosis had an inhibitory phenotype associated with impaired function due to abnormal NK ligand expression (139). Torrelli et al. (59) observed a higher expression of the ligands for NK cell-activating receptors, Nec2, ULBP1, and UBLP3, on the surface of the blasts from children in contrast to adults with ALL, which could be associated with the worse clinical evolution of ALL in adults than in children (59). Differences in NK cell activity among molecular ALL subtypes have been described, with increased NK cell-activating ligand expression (NKG2D and DNAM1) in patients with ALL carrying the fusion gene BCR-ABL (Philadelphia chromosome: Ph+), in contrast to Ph negatives (59). Additionally, Ph+ cells are more susceptible to NK cell killing activity than ALL cells carrying no known molecular markers and were enhanced in Ph+ adult cases; B-ALL with MLL-AF4 gene and T-ALL cases displayed a high density of the NKG2D ligand and UL16-binding protein (ULBP-1) (59). NK cells from child and adult patients have shown aberrant functions, such as low degranulation of granzyme B (117, 139). In a cohort of child patients with B cell ALL sampled at diagnosis, end induction, and maintenance, evidence of altered NK phenotype and function compared to age-matched controls was revealed. It should be emphasized that the NK abnormalities were partially corrected during the maintenance phase of the ALL treatment and were inducible in healthy NK cells after co-culture with ALL blasts in vitro by TGF-β1 release (139). In fact, leukemic cells secrete IL-10 and TGF-β in order to evade the effect of CTLs and NK cells (137). Furthermore, a direct contribution of the TME to the exhaustion of NK cell functions by the CRTAM/Necl-2 interaction was reported in ALL. Indeed the decreased NK cell content and their depleted cytotoxic capacity in peripheral blood are two of the predominant immune surveillance problems in acute leukemia (139). Current investigations focusing on in vitro activation and NK cell expansion protocols to treat ALL are underway.

MDSCs are a heterogeneous population of regulatory immature cells derived from monocytes or granulocytes that are involved in immunosuppression in patients with cancer (163). MDSCs consist of two main subpopulations (monocytic MDSCs—MO-MDSCs and polymorphonuclear MDSCs—PMN-MDSCs) that suppress the activation, proliferation, and cytotoxicity of effector T and NK cells and induce the differentiation and expansion of Tregs (163–165). The role of MDSCs in ALL remains incomplete. Recently, it was reported that patients with B cell ALL at diagnosis have a higher number of MDSCs than healthy subjects, which was even higher during induction chemotherapy (166). Furthermore, Zahran et al. (163) detected increased MDSCs in pediatric patients with B cell ALL compared to healthy controls. Moreover, these authors observed a relationship between PMN-MDSCs and the levels of peripheral and BM blast cells and CD34 + cells, suggesting that PMN-MDSC cells provide a suitable immune-suppressive state for B cell ALL tumor progression. A reduction in PMN-MDSC population was related to complete post-induction remission (163). The high number of these cells in pediatric patients with B cell ALL suggests that the increased levels and activity of MDSCs and Tregs could explain the immunosuppression state observed in this malignancy (163). MDSCs secrete TGF-β and IL-10 that have direct immunosuppressive effects and induce Treg expansion, which suppressed tumor-specific T cell responses (167) ( Figure 2C ).

Macrophages are other essential immune cell populations of the host and are composed of two subtypes: M1, which has antitumor effects, and M2 (anti-inflammatory cells with protumoral properties), which supports TME through the induction of angiogenesis, metastasis, and immune suppression (168–170). Macrophages have anti-tumoral activities at the initial stages of solid and hematological tumor development; however, TME impairs macrophage function, transforming them into immunosuppressive cell types with pro-tumoral activities (171). The frequency of M1 macrophages has been reported to be notably reduced in adult patients with B-ALL compared to controls, while M2 macrophages are increased (117). M2 is divided into several subtypes, where TAMs, which are relevant in solid tumor cell invasion, are included. Knowledge about the role of macrophages in hematopoietic malignancies has been obtained mainly from the study of lymphomas, where an association between the number of TAMs in lymph node biopsy and the prognosis of patients with classical Hodgkin lymphoma (cHL) was found (172). In ALL, the production of M2 macrophages with immunosuppressive/tolerogenic properties can be induced by tumor-mediated mechanisms (tumor-derived cytokines and growth factors, etc.) (171). Furthermore, it has been found that spleen leukemia-associated macrophages (LAMs) stimulate the proliferation of T cell ALL and have high migration, and their functional and phenotypic characteristics are modified by an organ-specific microenvironment (169, 173). Most LAMs have an M2-like phenotype. It has been reported that this type of immune cells also secrete immunosuppressive cytokines such as IL-10 and TGF-β (159).

During malignant hematopoietic disorders, such as acute leukemias, intrinsic and extrinsic signals (including those participating in immune surveillance) influence the cell differentiation pathway and cooperate in abnormal fate decisions, highlighting the relevance of a continuous homeostatic control to produce elements of tumor suppression (174, 175). Switching of CML to ALL in the blast crisis, AML cases relapsing as ALL, ALL converting AML after chemotherapy, and mixed phenotypes (simultaneous expression of both myeloid and lymphoid antigens) suggest that linage-associated molecule expression contributes to immune response disruption and facilitates cancer progression (174, 176–179). Lineage switching in leukemia is more frequent in children than in adults, and most cases are ALL converting to AML (180). Studies have proposed that this process is a consequence of stem cell plasticity (capacity to cell fate conversion in defined cells adopting biological properties to the same or different lineages) since the evidence shows that cancer cells are derived from the same founder clone in leukemia lineage switching (180). Leukemias with lineage switching appear to be more common in specific genetic subtypes, such as those with KMT2A (MLL) gene rearrangements (180). The absence of EBF1 expression in ALL allows early lymphoid progenitors to differentiate into the myeloid lineage, and deletion of PAX5 in mature B cells can induce conversion to different fates, including macrophages and T cells (176, 181). Low PAX5 expression has been reported in patients with ALL and very early relapse-expressing AML genes, such as MPO and FLT3 (174). A single-cell RNA-seq study revealed that plasticity coexists with oncogenic and immune evasion programs in early T progenitor ALL (174), suggesting that specific features acquired during lineage conversion could contribute to immune evasion response in ALL. It has been proposed that the plasticity of leukemic blasts in early progenitor T cell ALL can modulate the treatment based on inhibitors of the Notch pathway due to the coexistence of transcriptional programs that are characteristic of lymphoid and myeloid lineages. Additionally, immunoevasion signatures were found to be activated in the TME—for example, the interaction between hepatitis A cellular virus receptor 2 and galectin 9 is associated with CD8+ T cell dysfunction (174). Studies aimed at understanding leukemic blast plasticity could contribute to the identification of potential therapeutic targets based on the reversion of T cell depletion and consequently improve OS rates in patients with ALL.

Antibodies are the basis for many new anti-cancer treatment strategies due to their immunomodulatory properties and capacity to promote the induction of anti-tumor immune responses. These antibodies target self-tumor antigens or the TME to inhibit tumor growth by increasing host immune responses to antigens expressed by the tumor itself or by reducing pro-tumorigenic factors generated in the tumor stroma (206). CTLA-4-specific mAbs have been used in human cancers, such as melanoma (206). Using the anti-CTLA-4 mAb (ipilimumab) in combination with IgG4 mAb (nivolumab), which disrupts the interaction between PD-1 and PD-L1/PD-L2, in patients with relapsed/refractory cHL, non-Hodgkin lymphoma, or multiple myeloma showed no significant improvement in efficacy over single-agent nivolumab (207). The anti-PD-1 anti-leukemic treatment is based on the maintenance and expansion of tumor-specific memory T cells and NK cell activation. This approach has been explored in diverse tumors, including relapsed/refractory lymphoid malignancies; however, its clinical application in ALL remains unknown. In ALL, CD-38 and CD-52 have been identified as target antigens of mAbs for the treatment of relapsed T cell ALL. Currently, there are ongoing clinical trials testing the efficacy of anti-CD38 mAbs (isatuximab and daratumumab) and anti-CD52 mAbs (alemtuzumab) with favorable results for better disease prognosis (127).

BiTEs are bispecific recombinant glycoproteins with two single-chain variable fragments (scFvs) connected by a flexible linker, whose targets are membrane molecules (costimulators, coinhibitors, adhesion, etc.) from both T cells and malignant cells. BiTEs favor immune responses by creating an immune synapse among tumor antigens and T cells (204). The distribution of BiTEs depends on factors such as the diffusion of the vascular endothelium, laminar flux rate, and interaction with the target. Blinatumomab (anti-CD19/anti-CD3; AMG103) is a BiTE that binds to CD3+ lymphocyte T and CD19+ B lymphocytes. Although blinatumomab induces selective lysis of tumoral cells, its half-life is short, and constant administration is necessary for effect maintenance. The Children’s Oncology Group has incorporated blinatumomab in clinical trials in patients with B cell ALL with a standard risk (1–9.99 years and leukocyte count <50,000/ml) classification. Good results and acceptable toxicity were observed; in addition, half of the population had a significant twofold improvement in median OS compared to patients with standard chemotherapy regimens (208).

Studies in patients with relapse/refractory pre-B cell ALL indicate that Treg proportion could determine the prognosis of blinatumomab treatment. It was observed that blinatumomab responders had a lower percentage of Tregs (4.82%) in peripheral blood compared to non-responders (10.25%). Additionally, the restoration of the activated T cell population was detected after the in vitro depletion of Tregs in leukemic blasts, thus highlighting the regulatory role of Tregs in the development of the immune response in ALL (209).

The main disadvantage of BiTEs is the induction of cytokine release syndrome (CRS) through proinflammatory cytokines and aplasia of lymphocytes B (205). Although collateral effects in clinical assays with blinatumomab have been reported (fever, nausea, headache, neurological, and hepatic adverse effects), a lower percentage of minimal residual disease in patients treated with blinatumomab in contrast to patients with high-risk ALL treated with conventional regimens was reported (204, 208, 210).

CAR is a synthetic construct formed by an extracellular scFvc (that recognizes the tumor antigen) fused to a transmembrane domain and to intracellular activating/co-stimulator motives (CD3ζ, CD28, 4-1BB) (204, 211–213). CAR is transduced to T cells (CAR T cell), and after the recognition of the tumoral antigen, it promotes a cytotoxic effect on the target cell (205). It has been reported that CAR T cells act independently of HLA recognition. Therefore, this approach could be used in different cases to overcome the lower HLA density of ALL malignant cells. Furthermore, it is feasible to use CD4+ T and CD8+ T cells to generate CAR T cells, which increases the effector and cytotoxic potential of T cells (213).

CAR T cells have recently been approved by the FDA to treat patients with leukemia and lymphoma (203–205), and several clinical trials of CD19 CAR T cell therapy are being carried out in ALL relapse patients, which have shown favorable results with remission after 6 months in up to 90% of the patients, with 78% OS and 67% event-free survival (214). However, in a group of refractory/relapsed patients with B cell ALL, CD22 CAR T cell therapy treatment achieved a complete remission of 70.5% in comparison with those previously treated with CD19 CAR T cells without success. CD22 CAR T cell-treated patients only exhibited a moderate grade of CRS and neurotoxicity, and better results were observed in patients undergoing alloHSCT (215). An important advantage of CAR T cells is their capacity to interact with malignant cells displaying different antigens, such as CD19, CD20, and CD22 (212).

Although CAR T cell implementation in B cell ALL has obtained favorable results, its implementation in T cell ALL presents limitations because CAR T cells and malignant T cells share similar expression profiles of target antigens, which gives rise to a non-specific cytotoxic activity incapable of discriminating CAR T cells from malignant T cells, leading to T cell aplasia and eventual immunodeficiency. Notwithstanding, the identification of antigens, such as CD4, CD5, and CD7, on T cell ALL shows promise for the use of CAR T cell technology, the clinical trials of which are ongoing (127).

Limitations of CAR T cell therapy in ALL involve the following (1): poor expansion and limited persistence in vivo caused by defects in the design and manufacturing of CAR T cell therapy (2), internalization of the CD19 glycoprotein and resurgence in tumor cells (3), toxicity (CAR T cells present numerous cellular interactions that could promote the cytokine-mediated systemic inflammatory response), and (4) aplasia of B cells and humoral deficiency that might promote infections (204, 205, 216).

ACT is based on TIL expansion and infusion in patients following lymphodepletion. ACT aims to generate a robust immune-mediated antitumor response via infusion of ex vivo-manipulated T cells. Studies have suggested that clinical outcomes correlate with tumor mutational and neoantigen load (217–219). Although ALL has been described as a malignancy with low mutational load, a recent analysis reported that it is possible to obtain immunodominant neoantigens that could be used to develop neoepitope-CD8+ T cells and treat patients with ALL (83, 220). To explore the effectiveness of this strategy, 36 putative neoantigens from the ETV6–RUNX1 fusion were tested, and 31 neoantigens were immunogenic. The co-culture of HLA-specific APCs with neoepitopes and isolated CD8+ tumor-infiltrating lymphocytes results in TNF-α and IFN-γ production. Therefore, this strategy provides a possibility to consider the adoptive transfer of neoepitope-CD8+ T cells as immunotherapy in leukemia and could be used in the consolidation phase or subsequent treatment (83).

The suppression of cell proliferation in leukemic lineages has significant challenges. On one hand, mAb treatment ablates the main elements of the adaptive immune response, including T and B cells, which could favor infection burst or immune-mediated disease development (221, 222). On the other hand, to increase tumor-specific T cell responses, it is necessary to promote leukemic cell immunogenicity. To date, the primary goal of several research groups has been to promote apoptosis in malignant cells; however, this type of cell death is immune-tolerogenic. Recent studies have shown that a new class of targeted drugs (second mitochondrial activator of caspases, SMAC) antagonizes diverse anti-apoptotic proteins (inhibitor of apoptosis proteins) and, in combination with dexamethasone, promotes increased immunogenic cell death (necroptosis) in ALL (223, 224). Necroptosis is a regulated inflammatory mode of cell death that is caspase-independent and presents highly regulated necrotic features (225). Necroptosis produces the release of DAMPs and proinflammatory cytokines, allowing a better cytotoxic function by tumor-specific T cells (225–227). The activation of necroptosis has been explored as an anti-leukemic therapy, and several SMAC mimetic compounds are currently in phase I or II clinical trials to treat hematological malignancies, including ALL (225, 228). The leading problem with active necroptosis in ALL therapy is its potential to induce immunogenicity. Observations from solid tumors suggest that necroptosis is not always pro-inflammatory or immunogenic; however, there are no reports of necroptosis in ALL (223, 229).