Update on immune‐based therapy strategies targeting cancer stem cells

Abstract

Accumulating data reveals that tumors possess a specialized subset of cancer cells named cancer stem cells (CSCs), responsible for metastasis and recurrence of malignancies, with various properties such as self‐renewal, heterogenicity, and capacity for drug resistance. Some signaling pathways or processes like Notch, epithelial to mesenchymal transition (EMT), Hedgehog (Hh), and Wnt, as well as CSCs' surface markers such as CD44, CD123, CD133, and epithelial cell adhesion molecule (EpCAM) have pivotal roles in acquiring CSCs properties. Therefore, targeting CSC‐related signaling pathways and surface markers might effectively eradicate tumors and pave the way for cancer survival. Since current treatments such as chemotherapy and radiation therapy cannot eradicate all of the CSCs and tumor relapse may happen following temporary recovery, improving novel and more efficient therapeutic options to combine with current treatments is required. Immunotherapy strategies are the new therapeutic modalities with promising results in targeting CSCs. Here, we review the targeting of CSCs by immunotherapy strategies such as dendritic cell (DC) vaccines, chimeric antigen receptors (CAR)‐engineered immune cells, natural killer‐cell (NK‐cell) therapy, monoclonal antibodies (mAbs), checkpoint inhibitors, and the use of oncolytic viruses (OVs) in pre‐clinical and clinical studies. This review will mainly focus on blood malignancies but also describe solid cancers.

Targeting of CSCs by novel immune‐based therapeutic strategies.

1. INTRODUCTION

According to the theory of cancer‐initiating cells (CICs), tumors consist of a vast number of cancer cells alongside a population of tumor cells named CSCs with distinctive properties such as self‐renewal, metastatic diffusion and resistance to conventional cancer therapies which comprise 0.01%–10% of the cells present in the tumor microenvironment (TME). ¹ , ² , ³ Besides the expression of unique CD markers such as CD44, CD24, and CD133, the capacity to express biomarkers like drug‐efflux pumps (e.g., ATP‐binding cassette (ABC)), some enzymes such as aldehyde dehydrogenase (ALDH) and transcription factors (e.g., OCT‐4, SOX‐2) CSCs could be distinguished from the entire cells within the tumor. ⁴ This subpopulation was first discovered in acute myeloid leukemia (AML) in 1997. ⁵ CSCs have been further identified in various types of solid tumors. ⁶ Owing to the specific properties that CSCs demonstrate, like differential ability and capacity to xenograft cancer, quiescence, plasticity, heterogenicity, self‐renewal, accounting for drug resistance, tumor recurrence, immune escaping, and metastasis to other organs, there are a range of predicaments in the way of erasing these cells from the TME. ⁷ With concern to the crucial role of CSCs subpopulation in different tumors, in this review, we first elaborated on the CSCs properties and the efficiency of the current cancer therapies including, chemotherapy and radiotherapy. Further, we discussed various immunotherapy approaches for eradicating CSCs.

2. CSCs CHARACTERISTICS

2.1. CSC signaling pathways

One of the proposed methods for preventing the expansion and eradication of CSCs is targeting the implicated pathways. It is commonly observed that multiple signaling pathways such as Hh, Notch, TGFβ, PI3K/Akt, EGFR, JAK/STAT, and Wnt are dysregulated in CSCs, which confer various properties such as chemoresistance, metastasis, and maintaining their stemness. ⁸ The role of the Hh pathway in embryonic development has been highlighted in several solid neoplasms. Depending on the type of tumor, dysregulation of this signaling pathway might cause different clinical outcomes such as chemotherapy resistance in AML cell lines. ⁹ , ¹⁰ Three main inhibitors proposed to abort the Hh pathway include antagonists against the transcription factor Smoothened (SMO) (e.g., Vismodegib, Taladegib, Itraconazole), GLI (e.g., ATO, GANT61, Pirfenidone), and Hh ligands (Robotnikinin, HHIP, 5E1). ¹¹ , ¹² , ¹³ , ¹⁴ , ¹⁵ Notch signaling pathway is another conserved pathway of CSCs with implications in immune evasion, metastasis, and angiogenesis. This signaling pathway has been targeted with inhibitors against transcription mediators (IMR‐1, CB‐103, and SAHM1), mAbs against Notch ligands and receptors (e.g., Tarextumab, Enoticumab, Demcizumab), γ‐secretase inhibitors (GSIs) (e.g., MK‐0752, BMS‐906024), and anti‐nicastrin mAbs. Majority of these inhibitors have been entered into phase I or II clinical trials. ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ In addition, there are some Notch signaling inhibitors with impressive pre‐clinical outcomes. For instance, GSI PF‐03084014 is a Notch signaling inhibitor that has shown promising pre‐clinical results in the treatment of T‐cell acute lymphoblastic leukemia (T‐ALL) when combined with glucocorticoids. ²² The Wnt signaling pathway contributes to the embryogenesis and repair system, and its dysregulation is correlated with the initiation and progression of several tumors such as lung, breast, oral, and colorectal tumors. ²³ Application of Vantictumab targeting the Wnt‐related Frizzled receptor family was associated with a reduction in the abundance of CSCs and suppression of tumor growth in xenograft models. ²⁴ Another inhibitor of the Wnt pathway, CWP232291, with the capacity to facilitate degradation of β‐catenin, demonstrated efficient anti‐cancer activity in mouse models of multiple myeloma (MM). ²⁵

2.2. CSCs surface markers

One of the common approaches for killing CSCs is to target their surface markers. Although the majority of the surface markers expressed on the CSCs are not CSC‐specific, targeting these molecules in combination with other therapies could be effective. CD24, CD44, CD133, CD123, CD47, ALDH, CD34, 5 T4, and EpCAM (ESA) are some examples of the most common surface markers that have been considered for mAbs development. ⁸ , ²⁶ For instance, CD44, a cell adhesion molecule, is one of such important markers that has crucial roles in self‐renewal, tumor proliferation, EMT, niche preparation, apoptosis resistance, and metastasis which can enhance tumor cell stemness through interplaying with extracellular matrix components, growth factors and cytokines. ²⁷ , ²⁸ In a recent study, the impact of anti‐CD44 mAb A3D8 on the growth and apoptosis of sphere‐forming cells (SFCs) from the human ovarian cancer cell line SKOV‐3 was assessed. Data indicated that the cell cycle was arrested in the S phase and the expression level of caspase‐3 was up‐regulated, while CDK2, cyclin A, and Bcl‐2 protein expression levels were down‐regulated in the A3D8‐treated cells. ²⁹ In a pre‐clinical study performed by Jin et al, ³⁰ anti‐CD44 mAb H90 could effectively hinder the homing of leukemic cells including primitive CD34⁺CD38⁻ SL‐ICs (SCID– leukemia‐initiating cells) to bone marrow and spleen. Furthermore, H90 administration substantially abrogated the transmigration process, abolishing the competence of AML LSCs (leukemic stem cells) to access the bone marrow microenvironment. This phenomenon altered the fate of stem cells in NOD/SCID mouse models with human AML. Thus, targeting CD44 using H90 administration decreased leukemic repopulation and demonstrated CD44 as a key player in the regulation of AML LSCs. AML LSCs interact with a stem cell‐supportive niche to maintain their stem cell characteristics, and these findings suggested a therapeutic approach to target quiescent AML LSCs. CD133 or human prominin‐1, which is a pentaspan membrane glycoprotein, is also highly expressed on the surface of CSCs, and its aberrant expression is correlated with poor prognosis and chemo‐ and radio‐therapy resistance. ²⁹ In an experiment conducted by Kato et al, ³¹ CMab‐43, an anti‐CD133 mAb, was administered in nude mice transplanted with Caco‐2 tumor cells (human colon cancer cell line). Following CMab‐43 administration, the size of the tumor was significantly decreased when compared to the control group treated with a mouse IgG on days 12, 14, and 17. EpCAM (ESA) or CD326 is another surface marker with a high expression in CSCs. ³² , ³³ It is primarily expressed in simple epithelia, progenitor cells, stem cells from both healthy and malignant origin, and in multiple carcinomas. ³⁴ EpCAM possesses numerous functions including cell–cell adhesion, proliferation, differentiation, migration, and invasion. However, EpCAM is substantially expressed in carcinoma cells, and due to this fact, the majority of its functions have been primarily identified in malignant cells, and its activity in normal cells is not yet well‐characterized. ³⁵ EpCAM mediates oncogenic functions in tumor cells. For instance, the EPCAM‐positive cell population demonstrated stem cell characteristics including self‐renewal and pluripotency. They also revealed tumorigenic effects following injection into the NOD‐SCID mice with hepatocellular carcinoma (HCC). ³² CD47 is a transmembrane immunoglobulin and a receptor for thrombospondin family members that also acts as the signal regulatory protein alpha (SIRPα) ligand mediating several protein–protein interactions. ³³ Phagocytic cells including macrophages and DCs express SIRPα that commences a signaling cascade leading to the phagocytosis prevention (known as the “don't eat me” signal) after engagement with CD47. ³⁶ Therefore, CD47 overexpression in tumor cells results in the inhibition of phagocytosis by tumor‐associated macrophages (TAM) and is crucial for the survival and proliferation of tumor cells as well as metastasis of hematopoietic malignancies and solid tumors. ³⁶ Studies have indicated that expression of CD47 is much higher in the AML SCs than their normal counterparts, including hematopoietic stem cells (HSCs) and multipotent progenitor cells. ³⁷ Table 1 has listed the frequent surface markers that are recruited for the isolation of CSCs from various cancer types.

TABLE 1.

CSCs biomarkers in different cancer types.

Tumor type	Biomarkers	References
Colorectal cancer	CD133, CD24, CD29, CD44, CD166, EpCAM, Lgr5, ALDH, ESA	38, 39, 40
Gastric cancer	CD133, CD44, CD24, CD166, EpCAM	41, 42, 43, 44
Head and neck cancer	SSEA‐1, CD44, CD133, CD166	45, 46, 47, 48
Melanoma	ABCB5, CD20	49, 50
Pancreatic cancer	CD133, CD44, CD24, ABCG2, ALDH, EpCAM, ESA	41, 51, 52
Lung cancer	CD133, CD44, ABCG2, ALDH, CD87, SP, CD90	41, 50, 53
Liver cancer	CD133, CD44, CD49f, CD90, ALDH, ABCG2, CD24, ESA, CD13, OVC, EpCAM	38, 49
Leukemia (AML)	CD34, CD38, CD123,	52, 54
Prostate cancer	CD133, CD44, α2β1, ABCG2, ALDH, Integrins, CD166, Trop2, CD117	41, 49, 54
Breast cancer	CD133, CD44, CD24, EpCAM, ALDH‐1	49, 52
Ovarian cancer	CD24, CD44, CD117, CD133, ABCG2, EpCAM	55

2.3. Microenvironment of CSCs

The cross‐talk between the CSCs and the TME is one of the game‐changing factors that could dictate the behavior of these cells and modify the activities of not only the CSCs subpopulation but other cells including immune cells to optimize the environment in favor of tumor cells and promote tumor growth. ⁵⁶ , ⁵⁷ Targeting suppressive cells in the TME including regulatory T‐cells (T‐reg) and myeloid‐derived suppressor cells (MDSCs) can enhance the functionality of anti‐tumor effector immune cells, as these suppressive cells release inhibitory factors that lead to the suppression and exhaustion of functional immune cells. ⁵⁸ Moreover, the acidic pH of TME is a hallmark of malignant tumor cells, which could also have a deterrence effect on activated immune cells. Hypoxia is a feature of tumor cells in TME which induces the expression of hypoxia‐inducible factor‐1 (HIF‐1) regulating the transcription of various angiogenic factors, ⁵⁹ and CSC proliferation and self‐renewal. ⁶⁰ Hypoxic condition, hereby, induces tumor neovascularization. Angiogenesis is among the most vital properties of TME as the relentless division of tumor cells requires increased blood flow. Hence, crucial nutrients and oxygen are provided to sustain rapid tumor cell proliferation. ⁵⁹ Notably, the activities and the network of CSCs and other tumor cells could induce hypoxia and angiogenesis in TME. ⁶¹ Several experiments have focused on the correlation between CSCs activities and angiogenesis. For instance, it is found that higher level of VEGF is produced by CSCs both in normal and hypoxic conditions compared to the non‐CSC populations. The increased levels of VEGF eventually result in new vascular formation. ⁶² In a recent study on colorectal cancer (CRC), an anti‐angiogenic substance, Ginsenoside Rg3, was investigated and its suppressor impact on the stemness and growth of CSCs confirmed both in vitro and in vivo. Besides, the mRNA expression level of several genes participating in the angiogenesis including EGF, FGF‐2, PGF, and PIGF, were deceased following treatment with the RG3 that led to the disrupted vascularization of CRC xenograft. ⁶³ Last but not least, the complication of the metabolism of CSCs that could be altered by the impact of TME has become a topic of interest recently. In a more recent study, Wang and colleagues revealed that downregulation of molecular pathways participating in the Arf1‐mediated lipid metabolism was associated with defects in the mitochondria metabolism, enhanced ER stress, and the release of damage‐associated molecular patterns (DAMPs). These changes recruited DCs involved in the activation of IFN‐γ‐secreting cytotoxic T lymphocytes that eventually resulted in the eradication of CSCs and tumor suppression. More intriguingly, the killing of CSCs led to the induction of a tumor‐specific immune response through imposing CSC‐specific proteins into antigen presentation pathways, which in turn resulted in durable treatment efficacy. ⁶⁴

2.4. CSCs‐induced drug resistance

Drug resistance which is attributed to the presence of CSCs is one of the major issues encountered in cancer therapies. This resistance can stem from low proliferation rate, expression of some transporters like drug‐efflux proteins, autophagy, DNA repair mechanism, and upregulation of ALDH. ⁶⁵ , ⁶⁶ ALDHs facilitate stem cell maintenance and their proper development and differentiation. They protect the drug‐tolerant cells against elevated levels of reactive oxygen species (ROS) ⁶⁷ and also mediate retinoic acid (RA) biosynthesis. ⁶⁸ Data report high levels of ALDH activity in various solid tumor types. ⁶⁹ Notably, an elevated rate of ALDH activity generally is considered as a negative prognostic indicator. ⁷⁰ It is indicated that a pharmacologic decrease of ALDH activity results in toxic levels accumulation of ROS, which consequently eventuates in DNA damage and apoptosis within the drug‐resistant cancer stem cells. ⁶⁷ Several experiments have shown that chemotherapy that targets cancer cells with a high proliferation rate, by causing DNA damage or inhibiting the mitotic process, fails to remove CSCs which have a slow rate of proliferation. ⁷¹ , ⁷² , ⁷³ Moreover, the expression of ATP‐binding cassette (ABC) transporters (e.g., P‐gp, MDR1, ABCA1, ABCB1, ABCC11 (MRP8), and ABCG2 (BCRP1)) which take part in transporting conventional chemotherapeutic agents from cytosol to the environment through the energy gained from hydrolysis of ATP, is responsible for the resistance of CSCs to a range of chemotherapeutics and various molecular‐targeted agents leading to lower drug levels in the resistant cells below the amount needed for induction of cell death. ⁷⁴ , ⁷⁵ , ⁷⁶ , ⁷⁷ Finally, high expression of apoptotic inhibitors, low levels of Fas and Fas ligand (Fas‐L), and expression of Fas‐associated death domain‐like IL‐1β‐converting enzyme (FLICE)‐inhibitory protein (c‐FLIP) on CSCs could contribute to their resistance to apoptosis ⁷⁸ , ⁷⁹ , ⁸⁰ (Figure 1). C‐FLIP is considered a critical regulator of the death receptor (DR) networks and is a catalytically inactive caspase‐8/‐10 homolog. c‐FLIP protein significantly induces anti‐apoptotic activities through inhibiting cytokine‐ and chemotherapy‐mediated apoptosis, which results in resistance to these agents. ⁸¹ Therefore, targeting drug resistance to increase the sensitivity of CSCs to chemotherapeutic substances or radiotherapy administration could be a promising approach. Cell death regulators such as BCL‐2 have been also targeted in hematopoietic malignancies. However, a proportion of patients treated with venetoclax, a BCL‐2 inhibitor, exhibited minimal residual disease (MRD) which hindered the complete remission rates with the therapeutic strategies based on venetoclax resulting in relapse. ⁸² Investigations demonstrated that the retention of leukemic stem cells in the protective niches of bone marrow induced MRD. ⁸³ C‐X‐C Motif Chemokine Ligand 12 (CXCL12)‐ C‐X‐C Chemokine Receptor Type 4 (CXCR4) signal transduction drives an increase in the cell populations that express high levels of embryonic stem cell core transcription factors (ESC‐TFs: Sox2, Oct4, Nanog) in AML. ⁸⁴ Data showed that CD44 was involved in CXCL12‐induced venetoclax resistance of human AML cell lines and AML patient samples. ⁸⁴ Yu et al ⁸⁴ introduced a novel AML xenograft model in zebrafish and indicated that loss of function of CD44, which physically associates with CXCR4 at the cell membrane upon CXCL12 induction, sensitizes AML cells to the venetoclax through abrogating CXCL12‐mediated survival signaling. This suggests that inhibition of CD44 can be potentially considered to overcome venetoclax‐based therapy resistance in AML.

FIGURE 1.

Targeting CSCs. Several strategies have been developed to eradicate CSCs within the tumor, nonetheless, due to specific characteristics of this subpopulation, including overexpression of multidrug resistance (MDR) proteins, multiple signaling pathways as well as interaction with different players of the tumor microenvironment (TME), CSCs exhibit resistance to conventional chemotherapeutics. Novel immune‐based therapy approaches, however, are introduced to efficiently target CSCs. These strategies include: DC vaccines, chimeric antigen receptor (CAR)‐engineered immune cells, natural killer‐cell (NK‐cell) therapy, monoclonal antibodies (mAbs), immune checkpoint inhibitors (ICIs), and oncolytic viruses (OVs).

3. CONVENTIONAL STRATEGIES TARGETING CSCs

3.1. Chemotherapy

Chemotherapy is one of the cancer treatment strategies. The observed efficacy of the chemotherapy drugs represented that, if a feature is vitally important to tumor biology, its inhibition leads to the prevention of tumor growth. However, the observed clinical responses to these targeted therapies are generally transient and often followed by recurrence. One interpretation, supported by growing empirical evidence, is that each of the main characteristic features is regulated by a set of redundant partial signaling pathways. Thus, tumor cells regulate other pathways following the targeting of a specific pathway. For example, recent pre‐clinical data suggest that hypoxia induction by anti‐angiogenic agents could activate several pathways leading to an aggressive phenotype of tumor through triggering alternative pro‐angiogenic factors and drug resistance. ⁸⁵ , ⁸⁶ , ⁸⁷ Chemotherapy is functional in many cancers at early and advanced stages to eliminate cancer cells and improve survival. However, the effectiveness of conventional chemotherapy drugs on CSCs is limited implying the CSC resistance against various chemotherapy treatments. For instance, imatinib is a BCR‐ABL inhibitor and has shown remarkable efficacy in the recovery and survival of chronic myeloid leukemia (CML) patients. ⁸⁸ Nonetheless, in the bone marrow of many patients, BCR‐ABL‐expressing cells fail to be fully eradicated following the imatinib regimen. ⁸⁹ , ⁹⁰ Besides, in some cases, after achieving complete remission, patients are relapsed following imatinib withdrawal. ⁹⁰ Notably, according to the available data, the persistency of CML stem cells resistant to imatinib explains these clinical observations. ⁹⁰ , ⁹¹ In the CML‐like mouse model introduced by Oravecz‐Wilson and colleagues, two various CSCs subpopulations including leukemogenic and non‐leukemogenic cells were distinguished through their differences in the expression of specific surface markers. Intriguingly, the leukemogenic cells were scarce and significantly demonstrated higher resistance to imatinib therapy compared to the non‐leukemogenic population. ⁹² Accordingly, the properties of CSCs in various cancers play a key role in the progression of cancer and drug resistance. It is of note that the origin of CSCs is not fully understood and might be various in each type of tumor. Furthermore, it is not well known if the population of CSCs is increased upon tumor progression. ⁹³

3.2. Radiotherapy

In addition to chemotherapy, radiotherapy is one of the current cancer treatment options. Radiation therapy destroys cancer cells directly by ionizing radiation that causes DNA damage and indirectly by producing ROS. However, CSCs are considered radioresistant which is associated with an increased ability to repair DNA damage and reduced ROS. ⁹⁴ , ⁹⁵ Congruently, CSCs in glioma exhibit radiation resistance through induction of DNA damage response, ⁹⁶ while in breast cancer, CSCs disclose lower levels of ROS which was described by their anti‐oxidant expression profile leading to radiotherapy resistance. ⁹⁷ Radiation is delivered in multiple fractions and doses with improved radiosensitivity by different mechanisms. By fractionations, tumor cells have the opportunity to move into more sensitive phases during cell cycling. Moreover, remaining hypoxic cells will become reoxygenated, which is critical for ionization ⁹⁷ ; although, CSCs are believed to reside in perivascular and hypoxic niches, providing the required condition for them to escape from the radiation effect. ⁵⁶ The disadvantage of multi‐fractionated radiotherapy is that the mobilization of CSCs into cell cycling causes the tumor to repopulate, albeit most CSCs are in the G₀ phase of the cell cycle and proliferate slowly. In conventionally fractionated radiotherapy, the repopulation of tumors is considered the most common reason for therapy failure; the condition in which following the treatment with sublethal doses the re‐progression rate of a tumor is higher than the growth rate of the untreated tumor. ⁹⁵ , ⁹⁷ Interestingly, local pre‐treatment with radiotherapy could increase the NK‐cell transfer effects. ⁹⁸ Moreover, synergistic effects of CSCs‐ targeting T‐cells with radiotherapy on CSCs elimination could be favorable. ⁹⁹

4. NOVEL IMMUNE‐BASED THERAPEUTIC MODALITIES AGAINST CSCs

With the advances in the understanding of cancer cells and their molecular mechanisms and cell immune cycle that regulate immune response, immunotherapy has been developed for the treatment of hematological malignancies as well as solid tumors. These therapy strategies consist of passive (also known as adaptive) immunotherapy or active immunotherapy. In passive/adaptive approaches, mAbs or tumor‐reactive cells including genetically modified T‐cells, and NK‐cells are used, while in active immunotherapy cells like genetically modified B cells, DCs, and macrophages are applied to elicit an immune response against tumors. ¹⁰⁰ Checkpoint inhibitors and mAbs have provided powerful tools in treating cancer, especially in the advanced and metastatic stages. Cancer immunotherapy is an artificial procedure to stimulate immune system cells to erase cancer cells. Despite the remarkable advances and outcomes of immunotherapy, immunotherapy application against CSCs remains a challenge due to the potential of this subpopulation to escape the immune system by different mechanisms like impaired and downregulation of antigen presentation, modulation of the activity of immune cells, release and expression of immune suppressive factors and receptors, recruiting suppressive cells, and modulation of TME. ¹⁰¹ , ¹⁰² Subsequently, concerning the urgent necessity to improve the current immunotherapy approaches, the new procedures are focused on CSCs' specific properties in order to beat them. In the following sections, we review the immunotherapy methods introduced for the eradication of CSCs (Figure 1).

4.1. Dendritic cell (DC) vaccines

Using the process of antigen presentation by antigen‐presenting cells (APCs) to T‐cells is one of the most studied strategies among adoptive immunotherapy approaches. DCs are among the APCs that elicit both innate and adaptive responses of the host immune system. Via MHC‐II antigen presentation, DCs are capable of activating CD4⁺ T‐cells. CD8⁺ T‐cells are further activated by DCs through a process known as cross‐presentation, in which exogenous antigens are presented on MHC‐I. ¹⁰³ Transferring DCs carrying tumor‐associated antigens (TAA) to patients results in the activation of T‐cells, forming immunity against the used antigens and elimination of cancer cells. Applying this approach or cancer vaccination, CSCs could be used as antigens to elicit immune responses. ¹⁰⁴ , ¹⁰⁵ Current advances in CSCs‐targeting vaccination classify DC vaccines into 4 various groups including CSC‐lysates (or inactivated CSC‐based vaccines, CSC‐lysate‐loaded DC vaccines, cytotoxic T‐cell (CTL) vaccines generated with CSC‐lysate‐loaded DC vaccines, plus prophylactic and therapeutic experimental models for combinational strategies. ¹⁰⁶ Experiments on animal models of cancer have illustrated that CSCs lysates could lead to greater responses from T‐cells as it provides the possibility of targeting multiple antigens. ¹⁰⁷ In an experiment using immunocompetent murine tumor models, CSCs‐based vaccination caused the prevention of metastasis of melanoma cells to the lung and inhibition of expansion of the squamous carcinoma. ¹⁰⁸ Additionally, vaccination of mice with CSC lysate‐pulsed DCs significantly increased the lifespan of mice and caused tumor suppression. CSC lysate‐loaded DCs could specifically target CSCs and indicate anti‐tumor potential compared to C57BL/6 mice immunized with the murine melanoma cell line B16F10‐pulsed DCs. ¹⁰⁹ In mouse models of breast cancer, transferring DCs loaded with CSCs lysates could stimulate responses from CD8 and CD45⁺ T‐cells and cause their expansion. ¹¹⁰ In animal models of D5 melanoma and SCC7 squamous cell cancer, vaccination with ALDH^high SCC7 CSC‐DC following surgical excision was effective in reducing tumor recurrence and increasing host survival. In the D5 melanoma, the murine model establishment of ALDH^high SCC7 CSC‐DC vaccination was followed by inhibition of tumor growth alongside a reduction in metastasis to the lungs and prolonged survival. ¹¹¹ Of note, the combination of the CSC‐DC vaccine and dual blockade of programmed death‐ligand 1(PD1) and cytotoxic T‐lymphocyte‐associated protein (CTLA‐4), in murine models of melanoma, Zheng et al ¹¹² Proved that triple combination therapy not only could promote the expansion of T‐cell responses against CSCs and inhibit the release of TGF‐β, but it was also more beneficial in the elimination of ALDH CSCs. Of note, the combination of conventional chemotherapy and immunotherapy could be applied to increase the therapy response and also overcome tumor resistance. The administration of dual therapy of CSCs‐DC vaccine and cisplatin against Solid Ehrlich carcinoma in mice was accompanied by increased apoptosis, reduction in tumor growth, and expansion of MDR and Bcl‐2 and was represented as a promising combinational approach. ¹¹³ In a 9L CSC brain tumor model established by Xu et al, ¹⁰⁴ CSCs were recruited as sources of antigens to prime DCs for human GBM vaccination. Data indicated that CSC‐loaded DCs substantially triggered cytotoxic T lymphocytes against CSCs, with tumor‐bearing animals exhibiting a prolonged survival rate. In a more recent experiment on breast cancer, total RNA was gained from the whole breast cancer cell population as well as CSCs, separately. Intriguingly, DCs pulsed with the total RNA of CSCs significantly acted as a better source for activation of effector T‐cells, consequently, resulting in effective apoptosis of breast cancer cells. Nevertheless, there was a report on resistance from CSCs to apoptosis by effector T‐cells due to high expression of PD1 by the CSCs population which in turn eventuates in the higher apoptosis rate of these effector T‐cells. ¹¹⁴ Although the preclinical studies on DC vaccines for cancer treatment are promising, the better understanding of DC subtypes, approaches to prevail over the immunosuppressive TME, and novel biomarkers recognition are highly required to achieve more efficient DC vaccines. ¹¹⁵ Several findings from clinical trials studying CSC‐based DC vaccines also indicate significant tumor‐specific immune responses, some of which correlated with the survival benefits for treated cases. ¹⁰⁶ , ¹¹⁶ Notably, all clinical studies are ongoing in the first two phases, which are short‐term studies with a time span of 10–13 months. Moreover, these trials are mostly aimed at side effects (safety), vaccine dose identification, and the impact of the vaccine on provoking cellular and humoral immune responses. ¹⁰⁶ However, up until now, no clinical trial has been posted to clinical registries. Thus, more high‐quality clinical studies are needed prior to coming to universal conclusions regarding the clinical efficacy of the CSC‐based DC vaccines. Evidently, due to the suboptimal efficacy of CSC‐based DC vaccination, it is consequently suggested to be applied in combination with other therapy strategies including immunotherapies and conventional approaches. ¹⁰⁶ For instance, findings from a randomized phase II clinical trial indicated significant extended survival rates plus elevated CCL22 and IFN‐γ plasma levels in GBM cases following surgical tumor excision and treatment with conventional chemotherapy or radiation therapy in combination with CSC‐DC vaccine. ¹¹⁷

4.2. CAR‐based immune cell therapy

4.2.1. CAR‐T‐cell therapy

Genetically modified immune cells demonstrate remarkable potential for targeting CSC among all the immunotherapy strategies. T‐cells are crucial in cell‐mediated immunity against tumor cells, and engineered T‐cells, including chimeric antigen receptors (CAR) T and T cell receptor (TCR)‐T‐cells, have indicated promising clinical results that exhibit their therapeutic potential in eliminating tumor progression. Currently, this therapy strategy has exhibited more efficiency specifically in the treatment of hematopoietic malignancies. CAR‐T‐cells are the conventional immune cells that were designed and introduced to direct the CD19 molecule on the human leukemia cell. ¹⁰⁰ Cells engaged in this strategy to trigger anti‐tumor responses could be obtained from patients' blood or healthy donors, and following specific ex vivo modification are infused back into the patients while they have achieved the capacity to express specific receptors for the recognition of TAA and subsequently eliminate them. ¹¹⁸ Beating the need for MHC‐restricted antigen recognition could enable CAR cells to be engineered to target a vast range of antigens. One of the proposed targets for CAR cells could be CSCs as they play a key role in tumor development, and thus, CSCs‐targeting CAR‐T therapy would effectively accelerate the cancer prognosis. ¹¹⁹ Data from preclinical studies have displayed encouraging results regarding the CD133⁺ CSCs targeting in solid tumors. Accordingly, CAR‐T‐cell therapy is either in monotherapy strategies in glioblastoma ¹²⁰ or by applying combinational chemotherapy in ovarian ¹²¹ and gastric cancer stem cells. ¹²² In a clinical trial, the efficacy of CD133‐directed CAR‐T‐cells in patients with ALL, AML, breast, brain, liver, colorectal cancer, and pancreatic and ovarian cancers was studied. ¹²³ , ¹²⁴ In 23 patients (14 diagnosed HCC, 7 with pancreatic carcinomas, and 2 with colorectal carcinomas), CD133‐targeting CAR‐T‐cells were administered which later on illustrated the elimination of CD133⁺ cells, alongside managed toxicity and effective disease stability. Interestingly, in all patients, the duration of response ranged from 9 to 63 weeks. Particularly, sorafenib‐resistant hepatocellular carcinoma patients indicated a median progression‐free survival of 7 months. ¹²⁴ Recently, the study by Sangsuwannukul et al ¹²⁵ demonstrated promising results indicating the efficacy of the fourth‐generation anti‐CD133‐CAR‐T‐cells in eliminating the CD133‐expressing cholangiocarcinoma stem cells dose‐dependently. EpCAM, also known as CD326, ESA, or EGP40 is an adhesion molecule and has critical participation in cell–cell and cell‐to‐cellular matrix interplay. Clinical and preclinical evaluations engaging CAR‐T‐cells directing EpCAM have been conducted ¹²⁶ , ¹²⁷ and revealed remarkable efficiency in targeting and elimination of EpCAM‐expressing cells in an ovarian cancer cell line (SKOV3) ¹²⁸ and xenografts. ¹²⁹ Besides, in a colorectal cancer xenograft model, tumor progression was significantly blocked by EpCAM‐CAR‐T‐cell administration, followed by the increase of cytotoxic cytokines, including interferon‐γ (IFN‐γ) and tumor necrosis factor‐alpha (TNF‐α) that were confirmed via in vitro evaluation. ¹³⁰ Moreover, Chondroitin Sulfate Proteoglycan 4 (CSPG4), Disialoganglioside (GD2), CD44v6, Interleukin‐13 Receptor α2 (IL13Rα2) as well as CD133 plus CD33 (both of which are leukemic stem cell markers (LSC)) are multiple CSCs markers that are targeted by CAR‐T‐cells in different clinical trials and preclinical studies ¹³¹ , ¹³² , ¹³³ (NCT04097301). Beard et al ¹³¹ reported for the first time that glioblastoma CSCs express CSPG4 on their surface. They generated glioblastoma CSCs from resected human tumors and indicated that anti‐CSPG4 CAR‐ T‐cells have significant capability to recognize and eliminate these cells. The generation of anti‐GD2 CAR‐T‐cells against breast cancer stem‐like cells and administration of them in the xenograft model of TNBC was accompanied by the prevention of local tumor growth and lung metastasis. ¹³³ Notably, multi‐target CAR‐T‐cells also have been introduced. For example, in primary GBM samples, CAR‐T‐cells engineered to express multiple antigens, including Her2, IL13Rα2, and Ephrin‐A2 (EphA2) confirmed to overcome antigenic heterogeneity and to enhance the therapeutic efficacy in xenograft models. ¹³⁴ Although it should be noted that, as much as the generation of CAR immune cells engineered to target specific antigens might seem promising, in practice, severe toxicities have been reported upon injection. Accordingly, targeting multiple antigens probably enhances the challenge of on‐target/off‐tumor toxicity, as the majority of the antigens are expressed on both malignant and healthy cells. ¹³⁵ , ¹³⁶ As reported, almost 27% of CSC surface markers are also expressed by normal cells. ¹³⁵ Advancement of methodologies recognizing tumor‐specific antigens would increase the efficacy of CAR‐based immunotherapies. Furthermore, the development of other engineered T‐cells, including TCR‐engineered T‐cells, TCR‐like CARs, and TCR‐CARs ¹³⁷ designed to target CSCs would be of high interest for future investigations of CSCs immunotherapy strategies. Since CSCs are mutated cells and mutated cells commonly present most of these mutations, including neoepitopes, on their surface via MHC class I molecules, the limitation of MHC recognition by CAR‐designed cells might be overcome through the development of TCR‐based CARs, thus, directing CSCs‐specific neoepitopes can substantially reduce off‐tumor toxicities. Other potential strategies that would be applied to increase safety through bypassing off‐tumor toxicity include the modification of CAR affinity, which remarkably contributes to binding and cytotoxicity induction. Accordingly, only high‐density TAAs are recognized by CARs, designed with lower affinity to antigens, while low‐density TAAs in healthy cells are ignored. ²² Another approach could be the administration of antigen‐specific inhibitory CAR‐T cells. In this strategy, CAR‐T cells express inhibitory receptors against normal antigens along with TAA‐specific receptors. Thus, in the case of interacting with a healthy cell expressing TAA, inhibitory signals hinder cytotoxicity. ¹²⁷ Furthermore, employing CAR‐T cells expressing an anti‐CAR would be another potential strategy. In line with this approach, in a more recent experiment, Ruella and colleagues specifically depleted anti‐CD19‐CAR‐T cells using a cellular antidote. ³⁸ However, we still lack a deeper characterization of these not‐yet well‐studied strategies for the reduction of toxicity following CSCs‐targeting CAR‐T cell therapy.

4.2.2. CAR‐natural killer (NK) cell therapy

NK‐cells are a vital member of innate immunity and exhibit advantages for cancer immunotherapy compared to CAR‐T‐cells. For example, CAR‐NKs kill tumor cells with a lower risk of graft‐versus‐host disease (GvHD) induction. Moreover, CAR‐NK‐cells have a shorter lifespan than T‐cells which leads to a decrease in off‐target toxicities. ¹³⁸ CD123 and CD33 are common identification targets for leukemia. Recently, CD33‐CAR‐NK‐92‐cells have entered clinical trials for relapsed/refractory AML (NCT02944162) and case reports investigated the safety and indicated the encouraging tolerability of these AML‐specific CAR‐NK‐cells. ¹³⁹ Moreover, in ovarian cancer, targeting CD133 by third‐generation CAR‐NK92 cells significantly prevented tumor progression. ¹⁴⁰ Intriguingly, cisplatin combinational therapy resulted in a higher cell‐killing impact compared to a single treatment strategy, either CAR therapy or chemotherapy. ¹⁴⁰ In another study, researchers developed CAR‐NK‐cells targeting both CD24 and mesothelin which simultaneously directs ovarian CSCs and non‐stem cell tumor cells efficiently. ¹⁴¹ CAR‐NK92 cells targeting EpCAM on colorectal CSCs have revealed remarkable efficacy in suppressing CRC cell growth. ¹³¹ Notably, a combination of regorafenib with CAR‐NK92 immunotherapy elevated the anti‐cancer effects of therapy in CRC mouse models compared to monotherapy. ¹⁴² Besides, the administration of CAR‐NKs showed effectiveness in the induction of apoptosis in CSCs of MM, as in an experiment on MM patients, the combination therapy of Daratumumab (anti‐CD38) and CAR‐NK‐cells targeting CS1, which is highly expressed in MM CSCs compared to any other cell type, demonstrated potential anti‐tumor effect and inhibited MM relapse through the elimination of MM stem cells. ¹⁴³ Figure 1 schematically illustrates novel CAR‐based immune cell therapy against CSCs.

4.2.3. CAR‐macrophage (M) cell therapy

The poor infiltration rate of immune effector cells in the TME is considered an underlying challenge for immunotherapy, specifically for solid tumors. Notably, monocyte‐derived macrophages are the chief players of innate immunity and are the main participants in the TME due to their capacity for penetration into tumor lesions. The enhanced knowledge of TME has introduced novel approaches for applying manipulated macrophages to alleviate immunosuppressive TME, for example, ectopic expression of IL‐21, which stimulates activation of NK/T‐cells at TME. ¹⁴⁴ Data from multiple preclinical studies have confirmed the efficacy of macrophage‐based immunotherapies on tumor suppression, however, still many attempts should be made to optimize the efficacy and safety of CAR‐M in clinical treatment. ¹⁴⁵ Klichinsky et al ¹⁴⁶ utilized primary human macrophages and engineered them to represent sustained pro‐inflammatory phenotypes (M1). Overexpression of pro‐inflammatory chemokines and cytokines by these CAR macrophages equipped them with the accelerated antigen presentation processes and resistance to immunosuppressive cytokines. These CAR‐MS demonstrated significant capacity for alleviation of tumor burden and prolonging overall survival in the xenograft models. Macrophage‐editing‐based immunotherapy strategy, particularly against solid tumors, is a promising direction for future research. Nonetheless, data for CSC‐targeting CAR‐M immunotherapy is not currently available. Thus, this research idea would be a potential avenue in forthcoming research.

4.3. NK‐cell therapy

NK‐cells were identified to be naturally cytotoxic to the tumor and damaged cells in 1970. They protect the body against infectious pathogens and tumor cells owing to their intrinsic diverse characteristics and active interplay with adaptive immune cells, including B and T‐cells. NK‐cell populations are categorized into two different subsets of CD56^bright and CD16^Dim and include 5% to 15% of peripheral blood mononuclear cells (PBMCs). ¹⁴⁷ , ¹⁴⁸ Cytotoxicity effects of allogenic NK‐cells have been evaluated in hematological cancers, while both allogeneic and autologous NK‐cells are efficient against solid tumors. Recently, multiple studies have revealed NK‐cell therapy's capacity to eliminate targeted CSCs. ¹⁴⁷ Experiments on human breast, colon, melanoma, and glioblastoma reveal that IL‐2 and/or IL‐15 activated NK‐cells could identify solid tumor CSCs by involving the receptor‐dependent mechanism, eventually leading to CSC‐elimination in these tumors. ¹⁴⁹ , ¹⁵⁰ , ¹⁵¹ The Increased sensitivity of CSCs to NK‐cells has been shown in models of colorectal cancer. In this study, allogeneic NK‐cells were recruited to distinguish and eradicate colorectal CSCs. Of note, non‐CSC differentiated tumor cells were less sensitive to NK‐cells, which was correlated with lower expression of NKp30 and NKp44 ligands (in the natural cytotoxicity receptor (NCR) group of activating NK receptors) compared to CSCs. Besides, lower levels of MHC class I are expressed on the CSCs surface, which was indirectly linked to the more efficient targeting of them by NK‐cells. ¹⁵² In our recent experiment on glioblastoma multiform (GBM), the anti‐tumor effect of HSP70/Il‐2‐treated NK‐cells was analyzed and confirmed through both in vitro study and in vivo rat GBM models. Accordingly, NK‐cells could effectively cross the blood–brain barrier following systemic injection and subsequently, target the GMB tumor cells. ¹⁵³ Evidence indicates that the combination of drug therapies such as mAbs specific to CSCs markers with NK‐cell therapies could improve the results of cancer treatments and the elimination of CSCs. ⁶ For instance, Grossenbacher et al ¹⁵⁴ revealed that co‐incubation of human pancreatic cells with cetuximab could affect against CSCs more efficiently due to its antibody‐dependent‐cell‐mediated cytotoxicity (ADCC) ability. A bispecific fully humanized anti‐CD133 ScFV (single‐chain variable fragment) which bound to CD16 on NK‐cells and CD133 on colorectal cancer CSCs simultaneously, accelerated NK‐cell therapy efficiency. ¹⁵⁵

4.4. Monoclonal antibodies (mAbs)

In the last two decades, various mAbs have been approved by FDA for the treatment of cancers such as Rituximab (anti‐CD20), Daratumumab (anti‐CD38) cetuximab (anti‐EGFR), trastuzumab (anti‐HER2) and bevacizumab (anti‐VEGF‐A/anti‐angiogenic) for the treatment of lymphoma, multiple myeloma, epithelial cancer, HER2‐positive breast cancer, respectively. ³² , ¹⁵⁶ , ¹⁵⁷ MAbs engage the host's immune system for the eradication of targeted cells through triggering humoral and cellular mechanisms, including ADCC, complement‐dependent cytotoxicity (CDC), induction of apoptosis, prevention of receptor‐mediated signal transduction, and activating immune effector cells. ³² Recent studies in the field of mAbs targeting CSCs have introduced novel approaches. Morita and colleagues introduced CD271 as a CSC biomarker of hypopharyngeal cancer and developed an anti‐CD271 mAb, targeting CD271‐positive cells in xenograft models, eventuated in the decrease of CD271‐expressing CSCs through ADCC mechanism. ¹⁵⁸ Delta‐like ligand 4‐Notch (DLL4‐Notch) signaling plays a key role in protecting chemotherapy‐resistant CSCs. The feasibility of combining standard chemotherapy and anti‐CLLF mAb, demcizumab, and improving the anti‐tumor efficacy was investigated in a phase IB clinical trial. ¹⁹ ROR1 is an oncoembryonic orphan receptor for Wnt5a which is aberrantly expressed in CSCs. Specifically, neoplastic B cells in 95% of CLL patients exhibit ROR1 overexpression. The gene expression signatures of stemness in CLL were prevented following the administration of anti‐ROR1 mAb, cirmtuzumab in vivo. Besides, clinical recruitment of cirmtuzumab effectively suppressed the ROR1 pathway in CLL cases. ¹⁵⁹ MM CSCs aberrantly overexpress ABCG2, known to be the underlying player involved in drug‐efflux and consequent chemotherapy resistance in MM. In vitro and in vivo experiments of Shi and co‐workers indicated that anti‐ABCG2 mAb, conjugated with Epirubicin significantly induced apoptotic signaling pathways in MM CSCs through downregulation of PCNA, Bcl‐2, and CD31 and increased the expression of caspase‐3 and Bax. Various CSCs overexpress the CXCR4, which is correlated with tumorigenicity, angiogenesis, invasion, and drug resistance. Oriuchi et al ¹⁶⁰ developed a radioimmunotherapy strategy using a mAb targeting CXCR4 on AML CSCs in tumor xenografted mice. The strategy demonstrated significant feasibility in targeting and eradicating CSCs of AML. More functionalized mAbs against different targets on CSCs have been categorized in Table 2.

TABLE 2.

Functionalized mAbs against CSCs.

Name of mAb	Target	In vitro/in vivo	Effects on CSCs	References
Cetuximab	EGFR	In vitro	Killing human pancreatic CSCs by targeting EGFR	154
GV5	CD44	In vivo	Elimination of CSCs	161
Trastuzumab	HER2	In vitro	Killing of CSCs in breast cancer cells (MCF‐7 and ZR75)	162
H4C4	CD44	In vivo	Inhibition of self‐renewal capacity of pancreatic CSCs	163
CSL362	CD123	In vitro and in vivo	Lysis of leukemic CSCs	164
Adecatumumab (MT201)	EPCAM	In vitro	Killing ovarian cancer CSCs	165
H90	CD44	In vivo	Elimination of AML LSC	166
P245	CD44	In vivo	Killing of Breast cancer CSCs	167
A1MCMMAF	5T4	In vitro and in vivo	Induction of tumor regression Decrease in CSCs population	168
Solitomab (MT110)	EpCAM	In vitro and in vivo	Killing of pancreatic and colon CSCs	169, 170
7G3	CD123	In vivo	Killing of AML CSCs	171
Fusion of anti‐CD123 ScFV and anti‐CD3 ScFV	CD123 CD3	In vitro	Killing of AML CSCs	172
BH6H12	CD47	In vitro and in vivo	Affecting CSCs of brain tumors	32, 173
OMP‐52 M51	Notch 1	In vivo	Killing of breast cancer CSCs	174
Demcizumab	DLL4	Phase IB	Metastatic non‐squamous NSCLC	19
12C7 and 9B8 from mAb library	–	In vitro	NSCLC	175
Figitumumab	IGF	In vitro and in vivo	Reduction in colon cancer cell population and tumor growth	176
AVE1642	IGF	In vitro and in vivo	Killing of colon CSCs	177

4.5. Immune checkpoint inhibitors (ICIs)

Available evidence has revealed that CSCs express more immunosuppressive molecules, including programmed death ligand‐1 (PD‐L1) and T‐lymphocyte antigen‐4 (CTLA‐4), known as immune checkpoints (ICs), compared to their differentiated counterparts. ¹⁷⁸ , ¹⁷⁹ , ¹⁸⁰ ICIs are a promising type of immunotherapy strategy that acts by re‐invigorating the productivity of the host's immune system and suppressing overexpressed receptors or ICs in the microenvironment of cancer cells. ¹⁸¹ , ¹⁸² CTLA‐4, programmed death receptor‐1 (PD‐1) and PDL‐1 are some of the most well‐known ICs that suppress immune system response which have been already recruited in several clinical studies. ¹⁸¹ Since CSCs evade the immune system by activating ICs, inhibiting ICs might increase immunity functions and attack CSCs. ¹⁸³ CTLA‐4 is a T‐cell receptor that regulates T‐cell activation. Ipilimumab is a known human mAb that blocks CTLA‐4 activation. Clinical studies showed the recruitment of Ipilimumab after chemotherapy could improve the therapeutic efficacy in lung cancer patients. ¹⁸¹ PD‐1 is an inhibitory T‐cell or B‐cell surface receptor that belongs to the CTLA‐4 family, inflicting the immune system functions once it has bound to its ligand PDL‐1 in various cancers. ¹⁸¹ Some ICIs such as nivolumab, cemiplimab, and pembrolizumab specifically inhibit PD‐1 and PDL‐1 interactions and enhance T‐cell toxicity against cancer cells. ¹⁸⁴ PDL‐1 is the other checkpoint molecule that causes T‐cell anergy and attenuates immune system functions in different cancers. ¹⁸¹ It could promote stemness properties, including self‐renewal, tumorigenesis, and drug resistance in CSCs which is linked to the interplay between PDL‐1 and HMGA1 and subsequent activation of PI3K/AKT and MAPK pathways. ¹⁸⁵ Accordingly, in breast cancer, expression of PDL‐1 eventuates in the upregulation of the embryonic stem‐cell markers OCT‐4A, NANOG, and BMI1 which is dependent on the PI3K/AKT signaling pathway. ¹⁸⁶ Furthermore, PDL‐1 expression in gastric CSCs has been related to a higher proliferation rate and chemoresistance. ¹⁸⁷ Zheng et al ¹¹² examined the CSC‐directing impact of the CSC‐DC vaccine which was combined with a dual suppression of PD‐L1 and CTLA‐4 in an in vivo model for the melanoma tumor. They reported that dual blockade of PD‐L1 and CTLA‐4 significantly augmented the anti‐tumor effect of the CSC‐DC vaccine. Collectively, novel strategies targeting ICs in CSCs or signaling pathways contributed to ICs expression in CSCs would be of high interest for future investigations for improvement of therapeutic efficacy in cancer. Recent clinical trials of immunotherapeutics for CSCs targeting have been listed in Table 3.

TABLE 3.

Clinical trials of some novel therapeutic modalities targeting CSCs.

Phase	ID number	Approach	Target	Cell‐based therapy	Condition
III	NCT03434379	Checkpoint inhibitors	PDL‐1	–	HCC
–	NCT04977791	Checkpoint inhibitors	PD‐1 and PDL‐1		Non‐small lung cell carcinoma
I/II	NCT02176746	DC vaccine	Colorectal cancer stem cells	–	Colorectal cancer
I/II	NCT00846456	DC vaccine	Tumor stem cells	–	Glioblastoma CSCs
I	NCT01358903	mAbs	CD44	–	Malignant solid tumors
II	NCT03651271	mAbs	CD8	–	Advanced metastatic cancers
I/II	NCT02944162	NK‐cell therapy	CD33	–	Relapsed CD33 AML
I/II	NCT02944162	CAR‐NK‐cell therapy	CD33	NK‐92‐cells	Relapsed/refractory AML
I/II	NCT02541370	CAR‐T‐cell therapy	CD133	Autologous or donor‐derived T‐cells	Liver cancer pancreatic cancer brain tumor
I	NCT03423992	CAR‐T‐cell therapy	CD133, EGFRvIII, IL13RvIII2, Her‐2,EphA2, GD2	Autologous CAR‐T‐cells	Recurrent malignant glioma
I	NCT03563326	CAR‐T‐cell therapy	EpCAM	WCH‐GC‐CAR‐T	Neoplasm, stomach metastases
I	NCT02915445	CAR‐T‐cell therapy	EpCAM	CAR‐T‐cells	Malignant neoplasm of nasopharynx TNM staging distant metastasis (M), Breast cancer recurrent
I	NCT03766126	CAR‐T‐cell therapy	CD123	Autologous CAR‐T‐cells	Relapsed/refractory AML
I	NCT03672851	CAR‐T‐cell therapy	CD123	Autologous CAR‐T‐cells	Relapsed/refractory AML
I	NCT02159495	CAR‐T‐cell therapy	CD123	Autologous/allogeneic CAR‐T‐cells	AML (various) or blastic plasmacytoid dendritic cell neoplasms
I	NCT03114670	CAR‐T‐cell therapy	CD123	Donor‐derived CAR‐T‐cells	Recurred AML after allogeneic hematopoetic stem cell transplantation
I	NCT03126864	CAR‐T‐cell therapy	CD33	Autologous CAR‐T‐cells	Relapsed/refractory AML
I	NCT02799680	CAR‐T‐cell therapy	CD33	Allogeneic CAR‐T‐cells	Relapsed/refractory AML
I/II	NCT04097301	CAR‐T‐cell therapy	CD44v6	Autologous CAR‐T‐cells	AML, MM
I/II	NCT03356782	CAR‐T‐cell therapy	CD133	Autologous CAR‐T‐cells	Sarcoma, osteoid sarcoma, ewing sarcoma
I/II	NCT03013712	CAR‐T‐cell therapy	EpCAM	Autologous CAR‐T‐cells	Colon cancer; esophageal carcinoma; pancreatic, prostate cancer; gastric cancer, hepatic carcinoma
I/II	NCT03556982	CAR‐T‐cell therapy	CD123	Autologous/allogeneic CAR‐T‐cells	Relapsed/refractory AML
I/II	NCT03222674	Multi‐CAR‐T‐cell therapy	CD33, CD38, CD123, CD56, MucI, CLL‐1	Autologous CAR‐T‐cells	Relapsed/refractory AML
I/II	NCT04010877	Multi‐CAR‐T‐cell therapy	CLL‐1, CD33, and/or CD123	Autologous/allogeneic CAR‐T‐cells	AML
I/II	NCT04109482	CAR‐T‐cell therapy	CD123	Autologous CAR‐T‐cells	Relapsed or refractory blastic plasmacytoid dendritic cell neoplasm, acute myeloid leukemia, and high‐risk myelodysplastic syndrome
I/II	NCT01864902	CAR‐T‐cell therapy	CD33	Autologous or donor‐derived T‐cells	Relapsed/refractory AML
II	NCT02725125	CAR‐T‐cell therapy	EpCAM	Autologous CAR‐T‐cells	Relapsed or refractory stomach cancer
II/III	NCT03631576	CAR‐T‐cell therapy	CD123/CLL‐1	CAR‐T‐cells	Relapsed/refractory AML
II	NCT02729493	CAR‐T‐cell therapy	EpCAM	Autologous CAR ‐T‐cells	Relapsed or refractory liver cancer
‐	NCT03473457	Single or double CAR‐T‐cell therapy	CD33, CD38, CD56, CD123, CD117, CD133,CD34, or Mucl	CAR‐T‐cells	Relapsed/refractory AML

4.6. Oncolytic virotherapy

Oncolytic viruses (OVs) could reproduce and destroy cells, making them a promising immunotherapy modality in cancer treatment strategies. While unmodified OVs may harm both normal and tumor cells, genetically engineered OVs could distinguish between cancerous and healthy cells, leading to the selective infection and eventual destruction of tumor cells. ¹⁸⁸ Besides, faulty interferon pathways in tumor cells make them more susceptible to infection, especially with some viruses, including vesicular stomatitis virus and myxoma virus. ¹⁸⁹ Evidence suggests multiple viruses possessing oncolytic capacity, that include Poxviridae, Herpesviridae (HSV), Reoviridae (REO), Adenoviridae (AD), Paramyxoviridae, Picornaviridae, and Togaviridae. ¹⁸⁹ , ¹⁹⁰ Talimogene Laherparepvec (TVEC), or Imlygic is one of the FDA‐approved OVs for melanoma treatment, which is a modified HSV. ¹⁹⁰ The other one is Oncorine (H101), a modified AD, which has been applied in the treatment of head and neck cancer. ¹⁹¹ OVs could be delivered to cancer cells via intratumoral injection or systemically. They function through two direct and indirect pathways. OVs can directly infect and eliminate tumor cells by identifying special biomarkers that are overexpressed on cancer cells like laminin and CD64 or causing immune attack via cytolytic cells indirectly, or both ways together ¹⁸⁹ , ¹⁹² (Figure 1). Current data have shown OVs' ability in targeting CSCs in different types of cancers including brain tumors. ¹⁹³ Accordingly, Gp73‐regulated oncolytic AD exhibited toxic effects on hepatic CSCs and induced apoptosis in vitro and in vivo. ¹⁹⁴ In the other study, Jiang et al ¹⁹⁵ examined the anti‐tumor capability of Delta‐24‐RGD, targeted to the abnormal p16INK4/Rb pathway in CSCs of glioblastoma. They reported induced autophagic cell death of CSCs, which was due to the accumulation of Atg5 and LC3‐II protein in cells. In breast cancer patients, CSCs were susceptible to REo, and engineered REO OV was successful in eliminating and lysing both CSCs and non‐CSCs in vitro and in vivo. ¹⁹⁶ Zika virus (ZIKV) has shown a therapeutic oncolytic effect on glioblastoma stem cells (GSCs). It preferentially eliminated patient‐derived GSCs compared with other GBM tumor cells in culture, tumor organoids, and slice cultures. ¹⁹⁷ Recent evidence has indicated that ZIKV targets GSCs and stem‐like cells in medulloblastoma and ependymoma through directing the SOX2‐ integrin αvβ5 pathway. ¹⁹⁸ However, further studies on oncolytic virotherapy require to focus on strategies to improve viral delivery into the tumor, specific targeting of CSCs, as well as enhancing the bioactivity of viruses to survive in the patient's circulation, to reach tumor cells even in distant organs. ¹⁹⁹ Table 4 summarizes various immunotherapy strategies targeting CSCs for a better comparison.

TABLE 4.

Different immunotherapy approaches for targeting CSCs.

Immunotherapy approach	Advantages	Disadvantages	References
DC vaccines	Cause immunologic memory Targeting a broad range of antigens	Requiring present single antigen Alteration in efficacy after the transition from in vitro to in vivo	200, 201, 202
Engineered T‐cell therapy	Binding with high affinity Production of antigen‐specific, patient‐derived T‐cells Killing cancer cells repeatedly	Short‐term persistence of modified T‐cells in vivo Individual engineering of patient T‐cells is required High risk of autoimmunity	203
Nk‐Cell therapy	Easy isolation and expansion ex vivo Recognition of several ligands Antigen non‐specific	Reduced accumulation and activation in the microenvironment of solid tumors	202, 204
mAbs and checkpoint inhibitors	Provide long‐term anti‐tumor response The high survival rate compared to chemotherapy agents	Higher doses are related to a higher risk of treatment‐related death Cause hepatotoxicity or skin‐related troubles Critical neurologic interaction in children Requiring present single antigen	202, 205
Oncolytic viruses	Viral‐mediated cytotoxicity Increased tumor selectivity Enhanced anti‐tumor activity	Efficacy of ADs in vivo is not tested Causing antiviral immunity	206

5. CHALLENGES AND PERSPECTIVES

Researchers have encountered many obstacles to develop new remedial fields to target CSCs. Perspectives should engage issues, including developing novel delivery methods, preventing toxicity to normal cells, and increased the specificity of targeting CSCs. ³⁸ Despite the efficacy of immunotherapy approaches on cancer cells and CSCs, the inefficiency of innate and acquired immune systems may happen in some CSCs. ⁹⁸ Moreover, the use of specific immunotherapy approaches like receptor/ligand‐based targeting CSCs eliminates some CSCs populations, however, heterogenicity of the present population which has no relevant ligand/receptor could lead to the escape from the antigen‐dependent immunotherapy. ²⁰⁷ To overcome this problem, some studies suggest that antibodies that are not able to detect surface agents, yet have an elimination effect on CSCs are needed. ³³ Moreover, The main challenge with the CAR‐T‐cell approach is the on/off‐target tumor toxicity which may have toxicity against normal cells. ²⁰⁷ In addition to them, NK‐cell therapy also is faced with some obstacles including dysfunctionality of autologous NK‐cells and released longevity of NK‐cells in vivo. ⁹⁸ In this regard, adaptive delivery of NK‐cells into tumor sites could be more beneficial. Extracellular vesicles (EVs) derived from NK‐cells have the capacity to tolerate the acidic pH of the TME and their nano size may cause encouraging outcomes in visceral tumor treatments. ¹⁴⁸ Furthermore, anti‐cancer drugs or imaging probes‐loaded nanoparticles could pave the way for the treatment and diagnosis of CSCs in a targeted manner. ⁴⁹ Yao et al ²⁰⁸ illustrated that salinomycin‐loaded chitosan‐coated carbon nanotubes could target gastric cancer CSCs and inhibit their self‐renewal potency, migration, and invasion. Since novel treatment paradigms as mentioned in this article could eradicate CSCs, as well as conventional methods' effects on the bulk tumor, combination therapy with immunotherapeutic approaches and conventional treatments may improve the cancer treatment results. ³² , ³³ Data have shown that a combination of chemotherapy agents with OVs can be a better solution as OVs might overcome the chemoresistance of CSCs, furthermore, chemotherapy drugs may accelerate the cytotoxic activity of OVs. ²⁰⁹ In light of CSCs targeted treatments, more studies on CSCs characteristics and their related signaling pathways are of high importance. Notably, high throughput sequencing strategy and evaluation of expression patterns of CSCs may help to develop novel drugs targeting CSCs. ²¹⁰

6. CONCLUSION

CSCs are a subpopulation of cancer cells with self‐renewal feature, responsible for tumor recurrence. Conventional treatments such as chemotherapy and radiotherapy function against the bulk of tumors, but relapse of the tumor may happen in some cases as the CSCs still remain. In this regard, developing novel and effective therapeutic options is an urgent need. The use of immunotherapy methods, including DC vaccines, CAR‐T‐cells, NK‐cells, mAbs, checkpoint inhibitors, and OVs could improve the current cancer treatments' effects and specifically eradicate the CSCs. Like all treatment methods, they have some challenges, mentioned during this review. Of note, more preclinical and clinical studies are required for a better understanding and advancement of these new treatment paradigms.

AUTHOR CONTRIBUTIONS

Amirhossein Izadpanah: Conceptualization (equal); data curation (equal); investigation (equal); methodology (equal); project administration (equal); validation (equal); visualization (equal); writing – original draft (equal); writing – review and editing (equal). Niloufar Mohammadkhani: Conceptualization (equal); data curation (equal); investigation (equal); project administration (equal); resources (equal); visualization (equal); writing – original draft (equal); writing – review and editing (equal). Mina Masoudnia: Investigation (equal); methodology (equal); visualization (equal); writing – original draft (equal); writing – review and editing (equal). Mahsa Ghasemzad: Methodology (equal); project administration (equal); resources (equal); writing – original draft (equal); writing – review and editing (equal). Arefeh Saeedian: Project administration (equal); validation (equal); writing – original draft (equal); writing – review and editing (equal). Hamid Mehdizadeh: Conceptualization (equal); methodology (equal); project administration (equal); writing – original draft (equal); writing – review and editing (equal). Mansour Poorebrahim: Conceptualization (equal); data curation (equal); supervision (equal); validation (equal); writing – original draft (equal). Marzieh Ebrahimi: Conceptualization (supporting); data curation (supporting); supervision (lead); validation (lead); visualization (supporting).

CONFLICT OF INTEREST STATEMENT

All authors declare that they have no conflict of interests.

AVAILABILITY OF DATA AND MATERIALS

The authors declare that [the/all other] data supporting the findings of this study are available within the article [and its supplementary information files].

Izadpanah A, Mohammadkhani N, Masoudnia M, et al. Update on immune‐based therapy strategies targeting cancer stem cells. Cancer Med. 2023;12:18960‐18980. doi: 10.1002/cam4.6520

DATA AVAILABILITY STATEMENT

Data sharing is not applicable for this study.