Colorectal Cancer Stem Cells: Mechanisms of Resistance and Emerging Therapeutic Targeting Strategies

Simple Summary

Colorectal cancer is a common and deadly disease, and some patients do not respond well to current treatments or later develop the disease again. This article examines a small but powerful group of tumour cells, colorectal cancer stem cells, which can survive drug treatment and radiotherapy, initiate new tumours, and spread to other organs. We explain how these cells resist treatment, how they interact with their surrounding environment, and which weak points in these cells might be targeted to design better therapies. By bringing together recent laboratory and clinical findings, we aim to guide researchers and clinicians towards new treatment strategies that could make colorectal cancer therapy more effective and reduce the risk of cancer relapse.

Abstract

Colorectal cancer stem cells (CRCSCs) are now recognised as key drivers of tumour growth, therapeutic resistance, and relapse, yet optimal strategies to eradicate them remain unclear. This review explores how targeting distinctive biological, metabolic, and epigenetic features of CRCSCs could yield more effective therapies. It summarises the main traits of CRCSCs, such as self-renewal; quiescence; efficient DNA repair; high ABC transporter activity; activation of the Wnt, Notch, Hedgehog, and PI3K/AKT pathways; a shift toward oxidative phosphorylation; plasticity through the epithelial–mesenchymal transition (EMT); dysregulated microRNA (miRNA) and other non-coding RNA networks; immune evasion; and emerging oncofoetal reprogramming, and links them to therapeutic failure. The review then evaluates emerging therapeutic strategies that exploit these vulnerabilities, including differentiation-inducing agents, signalling pathway inhibitors, modulators of the EMT and cell cycle checkpoints, metabolism-based strategies, miRNA-directed therapy, and advanced drug-delivery technologies, such as nanocarriers. Finally, it discusses how immunotherapies, cell-based treatments, ctDNA-guided strategies, and CRCSC-enriched organoid models, together with single-cell and spatial multi-omics, can support rational biomarker-driven combination regimens. By addressing these issues, the review identifies current knowledge gaps, highlights promising therapeutic targets, and proposes a framework for developing durable treatments that specifically target CRCSCs in colorectal cancer (CRC).

1. Introduction

Colorectal cancer (CRC) is a major worldwide health concern and a leading cause of cancer-related mortality, with over 1.9 million new cases and about 900,000 deaths globally in 2022. High incidence and mortality rates, particularly in metastatic CRC, highlight the vital need for improved therapies [1]. Genetic mutations, such as those in the MUTYH DNA glycosylase gene, and hereditary conditions, including hereditary non-polyposis colorectal cancer (HNPCC), familial adenomatous polyposis (FAP), and Lynch syndrome, markedly increase the risk of CRC [2,3].

Despite advances in radiation, chemotherapy, and surgery, current treatments for CRC remain insufficient for many patients with advanced or metastatic disease. Tumour recurrence and the development of drug resistance persist as major obstacles to a durable clinical response. Given the poor prognosis of metastatic CRC, there is an urgent need for therapies that target the fundamental mechanisms underlying the treatment failure. A deeper understanding of tumour biology is therefore crucial for identifying novel therapeutic targets, as conventional treatments often fail to eradicate the most resilient cancer cell populations [4]. This review consolidates recent evidence on how CRCSCs drive resistance, recurrence, metastasis, and immune evasion through a close interplay with the tumour microenvironment and highlights emerging therapeutic strategies for overcoming these barriers. Unlike earlier reviews that focused on individual markers or single pathways, it presents an integrated, mechanistic and translational framework that connects CRCSC biology, including signalling, metabolic and epigenetic regulation, oncofoetal reprogramming and niche interactions with clinically relevant treatment opportunities.

2. Methods

A structured literature search was conducted in PubMed, Scopus, Web of Science, and Google Scholar to identify relevant studies on colorectal cancer stem cells, treatment resistance, recurrence, metastasis, immune evasion, and emerging therapeutic approaches. The search began in early February 2026 and was completed in March 2026. Only articles published in English were included. Search terms were combined using Boolean operators, and the reference lists of eligible articles were manually screened to identify additional relevant publications.

3. Biology of Colorectal Cancer Stem Cells

3.1. Definition and Concept of Cancer Stem Cells

The cancer stem cell (CSC) theory is widely accepted in oncology as a key framework for explaining tumour heterogeneity. It proposes that tumours comprise diverse cell types and are driven by a small population of undifferentiated CSCs, rather than being uniform masses [5]. In this context, CSCs are defined as a subset of tumour cells with the capacity for self-renewal and multilineage differentiation, capable of sustaining long-term tumour growth and regenerating the heterogeneity of the original lesion [6].

Earlier theories, including the classical “embryonal rest theory”, proposed that cancers originate from dormant embryonic cells persisting in adult tissues. The current CSC model represents a substantial advancement by framing tumours as hierarchically organised structures maintained by a minority of self-renewing stem-like cells. Researchers can now identify and isolate these CSCs in tumours using specific surface protein markers, such as CD133 and CD44, which enrich for highly tumorigenic, stem-like cells [7].

Although CSCs are rare and resilient, they are central to disease progression and treatment outcomes [8]. A detailed understanding of CSC biology is therefore crucial for developing more effective therapeutic strategies. CRCSCs have a unique set of characteristics that distinguish them from non-stem tumour cells and normal intestinal stem cells [9]. Understanding these properties is pivotal for unravelling their malignant potential.

3.2. Phenotypic Markers and Functional Identification

CRCSCs are characterised by the expression of specific surface biomarkers, the capacity for multilineage differentiation that recapitulates tumour heterogeneity, and the capacity for infinite self-renewal. Biomarkers for CRCSCs are critical for achieving therapeutic selectivity and minimising damage to normal stem cell populations, as they can distinguish CRCSCs from other tumour and normal cell types and reflect key functional traits such as self-renewal, invasion, and drug resistance [10].

3.2.1. Identified Surface Markers

CRCSCs are identified by detecting specific combinations of cell surface markers, many of which actively shape cellular behaviour and confer unique stem-like and tumorigenic properties [11].

The transmembrane glycoprotein CD44 is crucial for cell adhesion, migration, and intercellular interactions in CRCSCs. Colon cancer cells expressing CD44 proliferate rapidly, form numerous colonies, evade apoptosis, and exhibit marked recalcitrance to radio-chemotherapy protocols [11].

CD133, also known as AC133 or prominin-1, is widely regarded as a key surface marker for primary CRCSCs. Its expression in colorectal tumours correlates with larger tumour size, poorer differentiation, and reduced sensitivity to radiotherapy and chemotherapy [12].

Leucine-rich repeat-containing G protein-coupled receptor (Lgr5) is expressed on both CRCSCs and normal intestinal stem cells (ISCs) and plays a significant role in CRC biology. High Lgr5 expression is associated with a poorer prognosis, tumour progression, chemoresistance to 5-fluorouracil-based chemotherapy, and increased risk of recurrence. Because Lgr5 is shared by normal and malignant stem cells, selectively targeting cancer cells alone is difficult, so combining multiple markers or focusing on dysregulated signalling pathways is often required for effective therapies [13].

Aldehyde dehydrogenase isoform 1 (ALDH1): Elevated ALDH1 levels are associated with colorectal cancer cells exhibiting strong metastatic potential and drug refractoriness. ALDH1 acts as a detoxifying enzyme that protects cancer stem cells from oxidative stress, thereby prolonging their survival and supporting the epithelial–mesenchymal transition (EMT), invasion, and tumour spread [14]. Epithelial cell adhesion molecule (EpCAM) functions as a regulatory transmembrane protein that controls cell motility, proliferation, differentiation, and intercellular communication. By modulating oncogene expression and enhancing signalling pathways that influence cellular behaviour, such as WNT/β-catenin signalling, EpCAM contributes to the initiation of cancer in tissues such as the colon epithelium [15].

CD24 is a cell surface glycoprotein that acts as a cell adhesion molecule, interacting with selectins and integrins to regulate adhesion, migration, and signalling in many cell types. When combined with CD44 expression status, CD24 helps define and isolate CSC-enriched subpopulations, such as the classic CD44+/CD24− phenotype in breast and other carcinomas, which exhibits enhanced self-renewal, tumorigenicity, and sphere-forming capacity compared with unsorted tumour cells [16].

CD166 (ALCAM) is a cell surface adhesion glycoprotein that promotes tumour cell–cell and cell–matrix interactions and is linked to the progression of carcinoma from adenoma and poorer survival in colorectal and other cancers. It also contributes to neural development and axon growth, reflecting a broader role in guiding and supporting nerve growth [17,18].

3.2.2. Functional and Emerging Markers

Additional CSC markers include Nanog, Sox2, Oct-4, CD51, CD26, and CD29, which are described as CSC-associated markers that help maintain stemness and malignant traits in colorectal and other cancers, rather than serving as passive phenotypic markers.

Oct-4, Sox2, and Nanog form a core transcriptional regulator complex that controls pluripotency, self-renewal, and reprogramming. Elevated expression of these genes is linked to advanced stage, therapy resistance, and a poor prognosis in several cancers, including colon cancer [19]. A summary of important biomarkers for CRCSCs, along with their roles in CRCSC biology and key clinical associations, is shown in Table 1.

Because CRCSCs are heterogeneous and plastic, studies of single markers such as ALDH1 have yielded conflicting prognostic results, underscoring the limitations of relying on one biomarker. A more robust strategy uses multiparameter panels that combine functional assays with several surface and intercellular markers to monitor dynamic changes in marker expression. In colon cancer, a combined analysis of ALDH1, EpCAM, and survivin expression has been shown to provide strong, independent prognostic information on recurrence and survival and better capture tumour aggressiveness than any single marker alone [20].

Table 1. Essential biomarkers for CRCSCs and their significance.

Marker	Main Role in CRCSC Biology	Key Clinical Association
CD44	Adhesion; stem-like growth; apoptosis resistance [21].	Radio-/chemoresistance; colony formation [21].
CD133	Primary CRCSC surface marker [21].	Larger tumour size; poor differentiation; therapy resistance [21].
EpCAM	Adhesion molecule; activates Wnt/β-catenin signalling and oncogene signalling [22].	Tumour initiation and progression [22].
LGR5	Normal ISC and CRCSC marker; Wnt target [21].	5-FU resistance, recurrence, and poor prognosis [21].
ALDH1	Detoxification; oxidative stress damage protection [22].	Metastasis; drug resistance; poor differentiation [22].
CD166	Adhesion molecule [23].	Adenoma–carcinoma progression; angiogenesis [23].
Nanog	Core stemness transcription factor [24].	Poor prognosis [24].
Sox2	Stemness regulator [19].	Reduced disease-free survival; recurrence [19]
Oct-4	Stemness and pluripotency regulator [24].	Invasion; metastasis; cancer-related deaths [24].
CD51	Integrin; subunit; motility, and EMT [21].	Tumour development; sphere formation [21].
CD26	Adhesion; migration and invasion [21].	Distant metastasis (CD26+ subpopulations) [21].
CD29	Adhesion; self-renewal and differentiation [21].	Metastasis; tissue repair and immune interactions [21].

Abbreviations: CD44, cluster differentiation 44; CD133, cluster differentiation 133; EpCAM, epithelial cell adhesion molecule; LGR5, leucine-rich repeat-containing G protein-coupled receptor; ALDH1, aldehyde dehydrogenase isoform 1; CD166, cluster differentiation 166; Nanog, homeobox protein Nanog; Sox2, SRY-box transcription factor 2; Oct-4, POU class 5 homeobox 1; CD51, cluster differentiation 51; CD26, cluster differentiation 26; CD29, cluster differentiation 29.

Table 1. Essential biomarkers for CRCSCs and their significance.

Marker	Main Role in CRCSC Biology	Key Clinical Association
CD44	Adhesion; stem-like growth; apoptosis resistance [21].	Radio-/chemoresistance; colony formation [21].
CD133	Primary CRCSC surface marker [21].	Larger tumour size; poor differentiation; therapy resistance [21].
EpCAM	Adhesion molecule; activates Wnt/β-catenin signalling and oncogene signalling [22].	Tumour initiation and progression [22].
LGR5	Normal ISC and CRCSC marker; Wnt target [21].	5-FU resistance, recurrence, and poor prognosis [21].
ALDH1	Detoxification; oxidative stress damage protection [22].	Metastasis; drug resistance; poor differentiation [22].
CD166	Adhesion molecule [23].	Adenoma–carcinoma progression; angiogenesis [23].
Nanog	Core stemness transcription factor [24].	Poor prognosis [24].
Sox2	Stemness regulator [19].	Reduced disease-free survival; recurrence [19]
Oct-4	Stemness and pluripotency regulator [24].	Invasion; metastasis; cancer-related deaths [24].
CD51	Integrin; subunit; motility, and EMT [21].	Tumour development; sphere formation [21].
CD26	Adhesion; migration and invasion [21].	Distant metastasis (CD26+ subpopulations) [21].
CD29	Adhesion; self-renewal and differentiation [21].	Metastasis; tissue repair and immune interactions [21].

Abbreviations: CD44, cluster differentiation 44; CD133, cluster differentiation 133; EpCAM, epithelial cell adhesion molecule; LGR5, leucine-rich repeat-containing G protein-coupled receptor; ALDH1, aldehyde dehydrogenase isoform 1; CD166, cluster differentiation 166; Nanog, homeobox protein Nanog; Sox2, SRY-box transcription factor 2; Oct-4, POU class 5 homeobox 1; CD51, cluster differentiation 51; CD26, cluster differentiation 26; CD29, cluster differentiation 29.

3.3. Stemness Properties: Self-Renewal, Plasticity, and Differentiation

These stemness properties are sustained by developmental signalling pathways such as Wnt/β-catenin, Notch, Hedgehog, and Hippo/YAP, together with PI3K/AKT and microenvironmental factors, including inflammatory cytokines and hypoxia, which collectively govern CRCSC self-renewal, differentiation, and interaction with their niches. Persistent activation or rewiring of these networks maintains the CRCSC pool and seeds minimal residual disease, thereby predisposing to therapeutic resistance and relapse, as discussed in detail in Section 5.4 on signalling-driven resistance mechanisms [25].

In normal tissue, self-renewal is tightly regulated, but in cancer, this control is lost, allowing CRCSCs and their progeny to expand while retaining stem cell properties. Unlike rapidly dividing cancer cells, CRCSCs divide slowly or remain quiescent, rendering them less vulnerable to standard chemotherapy that targets proliferating cells and contributing to treatment resistance. Effective therapies may need to target vulnerabilities independent of the cell cycle or induce CRCSCs to re-enter the cell cycle to increase their sensitivity [9]. Evidence suggests that CRCSCs can arise from either normal stem cells or mature cells that reacquire stem-like properties, enabling tumours to adapt and recur even after existing CRCSCs are eliminated. Therefore, treatment strategies must both eradicate existing CRCSCs and prevent the formation of new CRCSC populations to reduce the risk of recurrence [26].

3.4. Relationship Between Intestinal Stem Cells and Colorectal Cancer Stem Cells

Distinguishing CRCSCs from ISCs is difficult because both share core stem cell features: infinite replicative potential, telomerase activity, and multipotency. However, CRCSCs are characterised by persistent, abnormal activation of signalling pathways (including Notch, Hedgehog, transforming growth factor-β (TGF-β), MAPK, and PI3K) that sustain their stem-like state and promote uncontrolled self-renewal. This overlap in key characteristics complicates the development of targeted therapies that eliminate cancer cells but spare normal ISCs. Therefore, effective treatments must specifically block these dysregulated pathways in CRCSCs while maintaining the signalling necessary for the survival and function of normal stem cells, thereby minimising damage to vital regenerative cell populations. A detailed comparison between ISCs and CRCSCs is summarised in

Table 2. Comparison of normal intestinal stem cells and colorectal cancer stem cells.

Feature	Normal Intestinal Stem Cells (ISCs)	Colorectal Cancer Stem Cells (CRCSCs)
Self-Renewal	Controlled symmetric/asymmetric division; maintains crypt homeostasis [28].	Uncontrolled self-renewal; excess symmetric division driving tumour growth [29].
Proliferation Rate	Mostly quiescent/slow cycling; inducible proliferation for tissue renewal [30].	Often quiescent but can switch to rapid, deregulated proliferation, supporting expansion and relapse [29].
Differentiation Potential	Generate all normal intestinal lineages; support organogenesis and homeostasis [30].	Recreates heterogenous tumour cell types; sustains tumour hierarchy [29].
Key Signalling Pathway	Wnt, Notch, Hedgehog, and related pathways are transiently tightly regulated [30].	Persistent/aberrant Wnt/β-catenin, Notch, Hedgehog, PI3K/AKT, and Hippo/YAP activation [29].
Telomerase Activity	Controlled activity maintaining telomeres and finite self-renewal [31].	Often upregulated; enables near limitless replicative potential [31].
DNA Repair Mechanisms	Efficient repair; preserves genomic stability [31].	Altered/enhanced repair; promotes survival after genotoxic stress [31].
Drug Efflux Mechanisms	Low-normal ATP-binding cassette (ABC) transporter expression [32].	High ABC transporter expression; multidrug resistance [33].
Carcinogenicity	Maintains tissue homeostasis; non-tumorigenic under physiological conditions [30].	Highly tumorigenic; initiates and sustains CRC, recurrence, and metastasis [29].

[27].

4. The Role of Colorectal Cancer Stem Cells in Tumorigenesis and Disease Development

CRCSCs are proposed to drive tumour initiation, progression, metastasis, and therapy resistance in CRC, consistent with the CSC hypothesis, which posits that a small, specialised CSC pool sustains tumour growth and relapse. These cells can arise either from the dedifferentiation of mature somatic cells or from the accumulation of genetic and epigenetic alterations within normal stem or progenitor cells [29].

4.1. Development and Growth

Because of their high tumorigenicity and ability to self-perpetuate, CRCSCs can initiate colorectal tumours and accelerate their progression. When transplanted into immunodeficient mice, they regenerate heterogeneous tissues that recapitulate the cellular complexity of the original lesion. This intrinsic heterogeneity underlies variable treatment responses and the frequent emergence of resistant clones, so effective therapies must target the CRCSC compartment to limit tumour regrowth and recalcitrance [34].

4.2. Metastasis Contribution

A metastasis-competent subset of CD26⁺ CRCSCs is responsible for establishing distant metastases: CD26⁺, but not CD26⁻, cells formed secondary tumours after injection into the mouse caecal wall, and detection of CD26⁺ cells in primary lesions is predictive of future metastatic disease. Moreover, some CRCSCs express CD44v6, and high CD44v6 expression correlates with an epithelial–mesenchymal transition phenotype, increased migration and invasion, and poorer disease-free and overall survival in CRC [35].

Studies that selectively deplete LGR5⁺ CRCSCs have demonstrated that, while this approach can restrict primary tumour growth, it does not fully prevent local relapse after treatment cessation because LGR5⁻ cells can repopulate the LGR5⁺ stem cell pool. In contrast, eliminating LGR5+ cells in metastatic lesions effectively blocks regrowth. Most circulating and disseminated CRC cells are LGR5⁻, whereas LGR5⁺ CSCs subsequently emerge in metastases to support long-term metastatic development. This bidirectional plasticity between LGR5⁻ and LGR5⁺ cells indicates that early and established metastatic stages may depend on distinct CSC populations. Therefore, effective therapy may require targeting both populations and their transitions, which complicates therapeutic design. Metastasis, particularly to the liver, is further promoted by complex interactions between CSCs and their microenvironment, including tumour-associated pericytes and circulating tumour cells. Additionally, this metastatic cascade is largely facilitated by cellular and molecular processes that modulate the EMT [34].

4.3. Contribution to Relapse and Recurrence of Tumours

Tumour recurrence and relapse after the initial response are primarily attributed to CRCSCs, which critically influence disease outcomes. Owing to their ability to self-renew and differentiate into multiple lineages, these cells frequently persist as minimal residual disease (MRD) following standard therapies. Although initial treatments may be effective, resistance—driven mainly by the resilience of CSCs—often leads to therapeutic failure. Chemotherapy can enrich the CSC subpopulation, as conventional treatments preferentially eradicate rapidly dividing bulk tumour cells while sparing quiescent, drug-tolerant CSCs, thereby allowing resistant clones to expand, repopulate the tumour, and give rise to more aggressive recurrences [36].

This highlights a major limitation of current treatment strategies: by favouring the survival of the most therapy-resistant CSC subclones, these treatments may inadvertently promote the development of more aggressive, therapy-resistant tumours, thereby increasing the risk of recurrence and poor long-term outcomes. For this reason, there is an urgent need for therapeutic strategies that specifically target and eradicate CRCSCs to achieve more durable clinical responses [37].

5. Therapeutic Resistance Mechanisms Mediated by Colorectal Cancer Stem Cells

CRCSCs are widely recognised as the primary contributors to CRC treatment resistance and disease recurrence, exploiting multiple molecular and cellular mechanisms to evade and withstand conventional cancer therapies [38].

5.1. Quiescence and Modified Checkpoints in the Cell Cycle

The quiescent or slow-cycling phenotype of CRCSCs is a core mechanism of therapeutic resistance. These cells often remain in G0-like cell cycle phases, making them less susceptible to chemotherapeutic agents that preferentially target rapidly dividing cells. These dormant CSCs can subsequently re-enter the cell cycle following treatment, thereby accelerating tumour regeneration and increasing the risk of relapse [37].

Quiescence is actively regulated rather than a purely passive phenomenon. In CRCSCs, the tyrosine kinase c-Yes is linked to dormancy, while the chromatin regulator high mobility group A1 (HMGA1) is downregulated in quiescent cells. Its silencing has been shown to promote a more quiescent, less proliferative stem cell phenotype. Chemotherapy enriches for quiescent CSCs partly through upregulation of the transcription factor zinc finger E-box binding homeobox 2 (ZEB2), which is elevated in quiescent, chemoresistant CRCSCs and helps maintain anti-apoptotic signalling. These findings indicate that CRCSCs’ quiescence is a dynamically controlled survival mechanism and suggest that targeting its key regulators could force dormant CSCs back into the cell cycle, sensitising them to conventional chemotherapy and enabling effective combination strategies [39].

In addition to quiescence, CRCSCs modify cell cycle checkpoints to withstand drug-induced DNA damage. Aberrant activation of checkpoint kinases, such as Chk1 and Chk2, together with dysregulation of p53, allows replication-stressed, hyper-diploid CRCSCs to survive genotoxic insults; conversely, pharmacologic CHK1 inhibition reduces this ability [37].

5.2. Enhanced DNA Damage Repair Capacity

Due to their highly effective DNA repair mechanisms, CRCSCs can withstand the cytotoxic effects of chemotherapy and radiation, both of which rely on inducing DNA damage. Key DNA repair genes, such as excision repair cross-complementation group 1 (ERCC1) and O (6)-methylguanine-DNA methyltransferase, are frequently upregulated in these cells, and ERCC1 overexpression has been linked to radio-resistance and poor responses to chemotherapy in CRC. Functionally, CRCSCs exhibit fewer γH2AX foci and improved survival after irradiation, consistent with enhanced repair capacity [37].

Moreover, CRCSCs limit the accumulation of reactive oxygen species (ROS) and possess enhanced antioxidant defences that efficiently scavenge therapy-induced ROS, thereby providing additional protection from genotoxic stress. Through the combination of robust DNA repair mechanisms and increased ROS detoxification, CRCSCs establish a dual, highly redundant defence system against DNA-damaging treatments. Addressing this resistance will require therapeutic strategies that simultaneously target both protective mechanisms, such as combining DNA-damaging agents with inhibitors of DNA repair pathways or with drugs that disrupt antioxidant systems specific to cancer stem cells [40].

5.3. Enhanced Drug Efflux via ATP-Binding Cassette Transporters

In CRCSCs, chemotherapy resistance is strongly linked to abnormal expression of ABC transporters. Members of this protein family, including ABCB1/MDR1/P-gp, ABCG2/BCRP1, ABCB5, and ABCC1, use ATP to actively pump toxic and chemotherapeutic substances out of cells, thereby conferring multidrug resistance (MDR) and rendering many traditional treatments ineffective [32].

CRCSCs frequently display a Hoechst dye-excluding side population (SP) phenotype, which reflects high efflux pump activity. Due to elevated ABC transporter expression, therapeutic agents may be expelled before reaching cytotoxic concentrations, so simply escalating drug doses is unlikely to overcome resistance. Promising strategies include directly inhibiting ABC transporters, designing drugs that these pumps cannot readily efflux, or using advanced drug-delivery systems, such as nanoparticle-based formulations, to bypass efflux mechanisms and improve intracellular drug accumulation in CRCSCs [41].

5.4. Signalling Pathway Activation

Dysregulated activation of several key signalling pathways that drive excessive clonogenicity and survival is a hallmark of CRCSCs’ resilience and lack of responsiveness to treatment. In CRC, aberrant Wnt/β-catenin signalling is common; stabilised nuclear β-catenin accumulation promotes proliferation and survival, while fibroblast-derived exosomes containing Wnt ligands can induce the dedifferentiation of bulk tumour cells into a CSC-like phenotype and enhance chemoresistance [42].

Notch signalling also plays a vital role in CRCSC refractoriness, as colonospheres and chemoresistant cells exhibit increased Notch1 activity, which supports their growth and reduces sensitivity to oxaliplatin and 5-fluorouracil (5-FU); pharmacologic or genetic Notch inhibition can restore chemosensitivity [43].

The Hedgehog pathway similarly contributes to refractoriness: the transcription factor GLI1 is required for colorectal carcinogenesis and is upregulated in 5-FU-resistant cells, and high GLI1 expression in patients correlates with relapse and shorter survival, while GLI1/GLI2 knockdown resensitises cells to 5-FU [44].

The Hippo/YAP signalling pathway constitutes a significant mechanism of resistance. Elevated YAP expression and that of its target genes in 5-FU-treated CRC patients are associated with increased recurrence rates as well as reduced survival. Experimental studies show that simultaneous inhibition of EGFR and YAP can overcome 5-FU insensitivity in CRC models [45].

In addition, the phosphatidylinositol 3-kinase/protein kinase B (PI3K/AKT) signalling pathway is activated by metastasis-associated colon cancer 1 (MACC1), which is upregulated in 5-FU-resistant colon cells and promotes chemoresistance, CSC-like properties, and MDR expression, whereas MACC1 knockdown restores drug sensitivity and reduces stemness [46].

Figure 1 illustrates the main signalling pathways that converge in CRCSCs and contribute to treatment resistance. Key developmental pathways, including Wnt/β-catenin, Notch, Hedgehog, TGF-β, and PI3K/AKT/mTOR, intersect at the MAPK/ERK and NF-kB nodes. These pathways support the evasion of apoptosis, DNA damage repair, drug efflux via ABC transporters, and the maintenance of quiescence or slow proliferation [47]. Solid black arrows represent primary regulatory signals, while curved arrows indicate extensive crosstalk among pathways. This network generates redundant survival signals, suggesting that durable therapeutic responses will require combination therapies targeting multiple interconnected pathways [48]. In contrast to their more differentiated progeny, CRCSCs display distinct metabolic programmes, such as altered lipid metabolism, alternative substrate utilisation, and a shift toward oxidative phosphorylation (OXPHOS), which supports survival under chemotherapy-induced stress [49]. For instance, 5-fluorouracil (5-FU)-resistant colonospheres become dependent on OXPHOS through the activation of sirtuin-1 (SIRT1) and its coactivator PGC1, as well as increased aldehyde dehydrogenase (ALDH) activity. This activity detoxifies reactive aldehydes and reduces oxidative damage, thereby sustaining CRCSC viability [22,50,51]. These metabolic adaptations highlight the plasticity of CRCSCs and indicate that the metabolic profiling of patient tumours could inform personalised combinations of pathway-targeted therapies, metabolic inhibitors, and standard chemotherapy to overcome resistance [52].

5.5. MicroRNA Dysregulation

MicroRNAs (miRNAs) represent an additional layer of post-transcriptional regulation that significantly influences CRCSC biology. Dysregulated miRNA expression in CRCSCs alters sensitivity to anti-cancer therapies and contributes to chemoresistance by modulating networks governing cell growth, proliferation, and survival. Tumour-suppressive miRNAs, such as miR-93, miR-45, miR-215, miR-328, and miR-497, are frequently downregulated and typically function to restrict CSC proliferation or enhance treatment responsiveness. In contrast, oncomiRs, including miR-21, miR-40, and miR-203, are commonly upregulated in colorectal CSC-enriched populations and have been associated with CSC expansion, the EMT, and chemotherapy resistance. The extensive dysregulation of miRNAs in CRCSCs indicates a complex regulatory network controlling stemness and therapeutic recalcitrance. The therapeutic modulation of miRNAs, either through anti-miRs to inhibit overexpressed oncomiRs or miRNA mimics to restore tumour-suppressive miRNAs, is currently being investigated as a strategy to reduce stemness and resensitise CRCSCs to conventional therapies [53].

5.6. Cellular Plasticity and Immune Evasion

The capacity of CRCSCs to undergo the EMT demonstrates considerable cellular plasticity and is strongly associated with invasion, metastasis, and the development of stem-like, therapy-resistant phenotypes. EMT processes facilitate the conversion of non-cancer stem cell (non-CSC) epithelial cells into CSC-like cells, thereby expanding the CRCSC population and contributing to tumour recurrence following therapy [54].

Beyond their inherent plasticity, CRCSCs demonstrate a pronounced capacity to evade the host immune system. This immune evasion is primarily attributed to elevated programmed death-ligand 1 (PD-L1) expression, reduced expression of molecules involved in antigen presentation to T cells, and the ability to establish an immunosuppressive tumour microenvironment. The combination of cellular plasticity and immune evasion renders CRCSCs as highly resilient cancer cell populations and presents significant challenges for both conventional and emerging immunotherapies [55].

6. Tumour Microenvironment and Maintenance of Colorectal Cancer Stemness

The TME provides conditions that support CRCSCs and modulate their stemness, adaptability, and resistance to therapy. Key components of the TME include tumour-associated macrophages, cancer-associated fibroblasts, cytokine and inflammatory signalling networks, hypoxic regions, the gut microbiome, and stromal elements, each of which contributes to CRC progression and metastasis [56]. Recent reviews demonstrate that CRCSCs are influenced by both the TME and their interactions with the tumour immune microenvironment. These observations highlight the need to develop niche-directed therapeutic strategies [57].

6.1. Tumour-Associated Macrophages and CRCSC Stemness

Tumour-associated macrophages (TAMs) represent key immune components within the colorectal TME and play a significant role in supporting cancer stem cell populations. CRCSCs recruit and polarise monocytes toward M2-like TAMs, which reinforce stemness, survival, and therapy resistance through factors such as interleukin-6 (IL-6), TGF-β, and vascular endothelial growth factor (VEGF). In CRC, TAM-derived signals have been shown to promote stem-like traits, the EMT, and immune suppression, thereby enhancing tumour initiation, progression, and metastatic potential. Consequently, targeting TAM recruitment or reprogramming their polarisation is an emerging strategy for disrupting the CRCSC-TAM axis and improving therapeutic outcomes [58].

6.2. Cancer-Associated Fibroblasts and Extracellular Matrix Remodelling

Cancer-associated fibroblasts (CAFs) are the predominant stromal cells in CRC that modulate the CRCSC niche through extracellular matrix (ECM) remodelling and paracrine signalling. By depositing and reorganising ECM components, CAFs establish a rigid, pro-invasive microenvironment that facilitates CRCSC maintenance, the EMT, and tissue invasion. Additionally, CAFs secrete growth factors, cytokines, chemokines, and exosomes that promote stemness, angiogenesis, and resistance to chemotherapy. Elevated stromal and CAF contents in CRC are associated with increased disease aggressiveness and poor prognosis, underscoring the significance of CAFs as therapeutic targets within the CRCSC niche [59].

6.3. Cytokines, Inflammatory Signalling, and Niche Support

Chronic inflammation and disrupted cytokine networks are central to CRC development and support the maintenance of CRCSCs within the TME. Tumour cells, immune cells, and CAFs release proinflammatory cytokines and chemokines, like IL-6, IL-8, tumour necrosis factor (TNF), and CXCL family members. These factors create an environment that suppresses immune responses, promotes stem-like traits, enhances cell survival, and impairs responses to therapy. These signalling cascades activate pathways including STAT3, NF-kB, and Wnt/β-catenin in CRCSCs, increasing their self-renewal, adaptability, and metastatic potential. Targeting cytokine–receptor interactions or downstream pathways may disrupt the inflammatory environment that supports CRCSCs and improve tumour responses to systemic therapies [60].

6.4. Hypoxic Niches and Metabolic Adaptation

Hypoxia is a common feature of solid tumours, and hypoxic regions within CRC create specialised niches that maintain CRCSC stemness and reduces response to treatment. The stabilisation of hypoxia-inducible factors (HIF-1α, HIF-2α) increases the expression of genes involved in glycolysis, angiogenesis, and survival, and promotes the EMT and stem cell transcription factors. Hypoxia-driven metabolic reprogramming, including increased glycolysis, altered lipid and amino acid metabolism, and redox adaptation, enables CRCSCs to survive nutrient and oxygen stress and to resist chemo- and radiotherapy. Targeting hypoxia-responsive pathways or metabolic vulnerabilities in these niches is a promising strategy for disrupting CRCSC maintenance in vivo [61].

7. Current and Emerging Therapeutic Strategies Targeting Colorectal Cancer Stem Cells

Because CRCSCs have unique biological features and several resistance mechanisms, new treatments beyond standard chemo-radiotherapy are urgently needed to achieve lasting results in CRC [62].

7.1. Traditional Treatments and Their Disadvantages

Standard management of CRC, particularly in stage III and high-risk stage II, includes surgical resection with fluoropyrimidine-based chemotherapy and, in some cases, radiotherapy. Treatment regimens may involve monotherapies such as capecitabine or irinotecan, or combination protocols including CAPOX (capecitabine + oxaliplatin), FOLFOX4 (5-FU + calcium folinate + oxaliplatin), FOLFOXIRI (5-FU + calcium folinate + oxaliplatin + irinotecan), and LVFU2 (5-FU + calcium folinate). Preoperative radiotherapy, delivered as either a short or long course, is commonly used for rectal cancer. Monoclonal antibodies such as bevacizumab, cetuximab, panitumumab, abituzumab, dalotuzumab, and ramucirumab target growth factors or their receptors and are often combined with chemotherapy to improve response rates and survival in advanced disease [63].

Despite these advances, 20–30% of patients with stage II-III CRC experience recurrence within 5 years of curative surgery, and metastasis and relapse remain the main cause of death. A key limitation of conventional therapies is their predominant activity against rapidly proliferating, differentiated tumours, while sparing slow-cycling or quiescent CRCSCs. These residual CRCSCs can survive treatment, repopulate the tumour, and drive more aggressive recurrences owing to their self-renewal capacity, multilineage differentiation potential, and robust defence mechanisms against both chemotherapy and immune attack [64].

Chemoresistance in CRCSCs arises from multiple intrinsic properties, including telomerase activity, high expression of ABC drug efflux transporters, reduced sensitivity to pro-apoptotic signals, efficient DNA damage repair mechanisms, quiescence, the EMT, and immune evasion. Collectively, these characteristics reduce the effectiveness of conventional therapies and highlight the need for therapeutic strategies specifically designed to target and eliminate CRCSCs [65].

7.2. Techniques for Colorectal Cancer Stem Cell Differentiation Induction

Inducing CRCSC differentiation is a promising treatment approach. By forcing CSCs to develop into more differentiated tumour cells, which are more vulnerable to traditional chemotherapy because they cannot proliferate indefinitely, this strategy aims to reduce the number of CSCs.

Retinoic acid (RA) has been effective in treating acute promyelocytic leukaemia by upregulating RA receptors and acting as a steroid hormone receptor agonist, thereby promoting the terminal differentiation of abnormal blasts [66]. Similarly, inducing the differentiation of CRCSCs into more differentiated tumour cells with reduced proliferative capacity and increased chemosensitivity may help deplete the self-renewing CSC pool. In CRC, RA differentiation therapy reduces the ALDH⁺ CSC population, inhibits proliferation, and promotes differentiation along a benign neuroendocrine lineage, which decreases sphere formation and stemness. In APC-mutant CRC, persistent Wnt activation increases expression of the retinoid-degrading enzyme CYP26A1, resulting in RA deficiency, blocked differentiation, and the expansion of ALDH⁺ stem cells. Inhibiting this Wnt-RA interaction or treatment with all-trans RA restores RA signalling, reduces ALDH1A1 expression, and promotes the differentiation of quiescent CRCSCs into neuroendocrine cells [67].

Bone morphogenetic protein 4 (BMP4), a TGF-β superfamily member, is a key regulator of CRCSC differentiation; by lowering nuclear β-catenin levels, BMP4 drives CSC maturation, differentiation, and apoptosis and markedly increases the sensitivity of CRC-derived CSC tumours to 5-FU and oxaliplatin in mouse models. The BMP signalling co-receptor Dragon (RGMb) is upregulated in CRC and enhances BMP4 signalling; genetic knockdown of RGMb reduces proliferation and tumour growth and decreases CD133+ CSCs, suggesting that Dragon is a promising target for modulating BMP4-driven differentiation [68].

Triiodothyronine (T3) also affects this pathway. Treating colospheres with T3 lowers the number of CD133+ CD29+ CD44+ stem-like cells, reduces their ability to form spheres and survive, and leads to increased apoptosis and G0/G1 arrest. This suggests that thyroid-hormone signalling can decrease CSC potential and may work with BMP4/Wnt modulation to make cells more sensitive to chemotherapy. These results support the use of differentiation-inducing agents, such as RA, BMP4 agonists, RGMb inhibitors, and T3, in combination therapies to target CRCSCs and improve the outcomes of standard chemotherapy [69].

7.3. Cell Cycle Checkpoint Impairment in Colorectal Cancer Stem Cells

Targeting cell cycle checkpoint proteins offers a promising way to overcome CRCSC-mediated resistance. Disrupting these proteins can lead to uncontrolled cell division, replication stress, and eventually cell death in tumour cells. In human CRC models enriched for CSC traits (CD44+/CD133+ tumourspheres), blocking both STAT3 and telomerase with the flavonoid morin and the telomerase inhibitor MST-312 reduces CSC traits. This approach reduces the number of CD44+/CD133+ cells, decreases tumoursphere formation, and alters DNA damage signalling, including the phosphorylation of p53 and checkpoint kinases. In addition, APC-mutant CRC cells, which usually resist 5-FU, become much more sensitive when the replication checkpoint kinases Ck1 or Ck2 are blocked with drugs or silenced by siRNA. This suggests that blocking these checkpoints can make this specific group of cells more responsive to chemotherapy. Overall, these strategies aim to weaken the checkpoint defences that help CRCSCs survive DNA damage and cell cycle arrest, thereby increasing their likelihood of responding to standard chemotherapy [70].

7.4. Epithelial to Mesenchymal Transition Inhibition

The EMT is a key driver of invasion, metastasis, and the acquisition of stem-like phenotypes in CRC and is largely orchestrated by the TGF-dependent induction of transcription factors such as Snail1/2, Twist, and Zeb1/2, which downregulate E-cadherin and disrupt epithelial polarity. Non-CSC tumour cells can gain CSC-like properties during the EMT, so the pharmacologic inhibition of TGFβ signalling is being explored to block both metastatic spread and de novo CSC formation [71].

The TGFβ1-blocking peptides P17 and P144 significantly reduce liver metastasis from CRC in mouse models by attenuating the TGF-β-driven EMT, proliferation, and angiogenesis; enriching CD44+/SOX2+ stem-like cells; and impairing tumour–fibroblast interactions. Kindlin-1, an integrin-associated focal adhesion protein, promotes CRC progression by recruiting SARA and Smad3 to TGF receptor 1, thereby activating TGF/Smad3 signalling; its overexpression correlates with advanced stage and poor prognosis, identifying kindlin-1 as another potential target for restraining the EMT and metastasis [72].

Overall, EMT inhibition via TGF-β pathway antagonists such as P17/P144, small-molecule ALK5 inhibitors, or the blockade of mediators like kindlin-1 aims to limit tumour invasion and metastatic colonisation and prevent differentiated cancer cells from re-entering a CSC-like state, complementing direct CRCSC-targeted therapies [73].

7.5. Targeting Important Colorectal Cancer Stem Cell-Related Signalling Pathways

Targeting aberrant signalling is essential for eliminating CRCSCs, as pathways such as Wnt/β-catenin and Notch maintain stemness, survival, and resistance to therapy in CRC [74].

The Wnt/β-catenin pathway is constitutively activated in approximately two-thirds of CRCs, which sustains intestinal crypt stem cells and promotes CSC-like properties [75].

Pyrvinium pamoate. This anthelmintic acts as a potent Wnt inhibitor. It reduces the viability of colon cancer cells, downregulates Wnt target genes such as c-MYC, impairs cellular migration, and decreases tumour growth in vivo, particularly in cells with APC or β-catenin mutations [76].

Tankyrase (Tnk) inhibitors (JW74 and XAV939) directly target CRCSCs by stabilising Axin2 and suppressing Wnt-driven c-KIT signalling. In CD44+ cancer stem cell-enriched populations derived from human CRC cell lines, tankyrase inhibition reduced cancer stem cell proliferation and self-renewal in vitro. Additionally, c-KIT knockdown produced effects mimicking of those observed with c-KIT inhibition, confirming c-KIT as a critical downstream effector. In xenograft models, the combination of a low-dose tankyrase inhibitor with irinotecan significantly inhibited tumour growth. These findings indicate that tankyrase inhibition may increase CRCSC sensitivity to standard chemotherapy [77].

PRI-724 (CBP/catenin antagonist). PRI-724 and its predecessor ICG-001 disrupt the β-catenin–CBP complex, shifting β-catenin towards p300-mediated differentiation, eliminating tumour-initiating cells, and sensitising tumours to cytotoxic agents; early-phase trials have shown acceptable safety and pharmacodynamic target engagement [78].

LF3 (β-catenin/TCF4 blocker). LF3 specifically interferes with β-catenin–TCF4 binding; suppresses Wnt target gene expression; reduces the motility and self-renewal of colon, head, and neck CSCs; and slows xenograft growth while promoting differentiation [79].

Dkk1 and soluble Wnt antagonists. Dickkopf-1 (Dkk1) binds to LRP5/6 co-receptors to inhibit Wnt signalling; Dkk1 overexpression in colon cancer cells reverses the EMT, lowers stem cell markers (CD133, LGR5), and diminishes tumour-initiating capacity; high Dkk1 levels are associated with less aggressive disease [80].

Natural Wnt-modulating compounds. Flavonoids such as quercetin, luteolin, apigenin, silibinin, and wogonin inhibit Wnt/β-catenin signalling in CRC models through mechanisms including GSK-3 inhibition, decreased nuclear accumulation of β-catenin, downregulation of Wnt target genes, reduced proliferation, the induction of apoptosis, and the inhibition of xenograft growth [81,82].

Notch Signalling Pathway. CRCSCs often depend on aberrant Notch activity for survival and self-renewal [83].

Monoclonal antibodies that block DLL4 (delta-like 4 ligand) suppress tumour growth and significantly reduce the tumour-initiating cell frequency in colorectal xenografts, with even greater depletion of CSCs when combined with irinotecan [84].

The γ-secretase inhibitor diabenzazepine (DBZ) blocks the Notch cascade and induces goblet cell differentiation in intestinal crypts and APC-mutant adenomas, leading to the loss of proliferative progenitors and the arrest of adenoma growth [85].

Jagged1 is a Notch ligand transcriptionally activated by the β-catenin/TCF4 complex in CRC cells; targeting Jagged1 or its downstream Notch activation can interrupt the Wnt–Notch feed-forward loop that supports tumour growth and stemness [86].

The γ-secretase inhibitor DAPT, as well as related GSIs, can deplete CD33+CD44+ CSCs, reverse EMT markers, and markedly inhibit xenograft growth in gastrointestinal cancer models, illustrating the potential of γ-secretase-mediated Notch blockade to eliminate therapy-resistant CSCs [87].

7.6. Metabolism-Based Treatment Strategies

CRCSCs exhibit distinctive metabolic characteristics that can be therapeutically exploited [49].

Targeting Oxidative Phosphorylation (OXPHOS): Metformin, a mitochondrial complex I inhibitor commonly prescribed for diabetes, suppresses OXPHOS, downregulates Wnt/β-catenin signalling, and reduces both the self-renewal and xenograft-forming capacity of CRCSCs. These effects support the use of metformin as an adjuvant to enhance chemotherapy efficacy and potentially permit dose reduction. However, chronic exposure may result in adaptive resistance [88].

Targeting Mitochondrial Translation and Biogenesis: Antibiotics such as doxycycline and tigecycline selectively inhibit mitochondrial protein synthesis and biogenesis, preferentially killing OXPHOS-dependent CSCs and blocking sphere formation across multiple tumour types, including CRC models [89,89].

Targeting Mitochondrial Dynamics: Enhanced mitochondrial fission is linked to stemness and invasiveness; the DRP1 inhibitor mdivi-1 reduces oxidative phosphorylation, impairs tumoursphere formation, and limits migration in several CSC-enriched cancer cell lines, suggesting similar potential in CRCSCs [90].

Targeting Mitophagy: In CSCs, stress-induced mitophagy supports survival and drug resistance. Inhibitors such as liensinine and radionuclide-loaded liposomal formulations that block late-stage mitophagy or autophagosome–lysosome fusion increase ROS, amplify mitochondrial damage, and resensitise cancer cells to chemotherapy. Although these studies were performed mainly in breast cancer and mixed solid tumour models rather than exclusively in CRC, they support the broader concept that inhibiting stress-induced mitophagy can resensitise CSCs to conventional drugs, a strategy that may be translatable to oxaliplatin-treated CRC [91].

Redox Homeostasis: CRCSCs maintain tight redox control via glutathione and related antioxidant systems, a feature linked to therapy resistance in CSCs across multiple tumours. Pharmacologic glutathione depletion with the γ-glutamylcysteine synthetase inhibitor buthionine sulfoximine (BSO) increases oxidative stress and enhances the efficacy of cytotoxics; in a rat model of CRC hepatic metastasis, BSO-induced GSH depletion approximately doubled the antitumour effect of melphalan compared with chemotherapy alone. These findings support glutathione targeting as a strategy to reduce clonogenic survival and resensitise CRC cells, with potential applicability to CRCSC populations [47].

Disulfiram. The alcohol-aversion drug disulfiram, especially in combination with copper, perturbs redox balance and autophagy signalling (e.g., via ULK1), inhibiting CRC cell proliferation and tumour growth, and decreasing stem-like populations in preclinical models [92].

Lipid Metabolism: CRCSCs rely on altered lipid desaturation, lipogenesis, uptake, and fatty acid oxidation. Inhibitors of SCD1, FASN, CD36, or FAO (such as etomoxir) suppress CSC viability and sphere formation, whereas omega-3 PUFAs like eicosapentaenoic acid (EPA) reduce CD133 expression, promote differentiation, and increase 5-FU sensitivity in colon cancer cells [93,94,95,96].

Collectively, these data highlight metabolic reprogramming as a promising avenue for combination therapies designed to selectively eradicate CRCSCs while augmenting standard chemotherapy [97].

7.7. Immunotherapies and Cell-Based Therapeutics

Immune checkpoint inhibitors harness the host immune system to attack tumour cells. They are now the standard of care for metastatic CRC with microsatellite instability-high (MSI-H) or mismatch-repair-deficient (dMMR) biology, with agents such as nivolumab, ipilimumab, and pembrolizumab showing durable responses in this subset. Recently, the phase III ATOMIC trial demonstrated that adding the PD-L1 inhibitor atezolizumab to adjuvant FOLFOX in stage III dMMR colon cancer reduced the risk of recurrence or death by about 50% and improved 3-year disease-free survival from 76.6% to 86.4%, establishing chemo-immunotherapy as a new standard for these patients [98]. Cellular immunotherapies are being explored to more selectively eliminate CRCSCs.

CART cell therapy modifies a patient’s T cells to express a chimeric antigen receptor (CAR) that recognises CRC-associated antigens such as CEA, NKG2D ligands, DCLK1, HER2, and guanylyl cyclase C (GUCY2C). Early-phase trials of CEA and GUCY2C-targeted CAR-T cells in heavily treated metastatic CRC have reported manageable safety signs and clinical activity, including tumour shrinkage and disease stabilisation, supporting the further development of CAR-T strategies against CRCSC niches [99].

Mesenchymal stem cells (MSCs) are being studied as tumour-homing vehicles to deliver cytokines, interferons, prodrug-converting enzymes, or oncolytic viruses directly into the CRC microenvironment, aiming to increase local cytotoxicity while sparing normal tissues. MSCs can preferentially migrate toward CD133⁺/CD44⁺ CSC-rich regions, but they also secrete immunosuppressive mediators (e.g., TGF-β, IDO, NO) and upregulate PD-L1, which can dampen effector T cell function and, if not carefully controlled, potentially promote tumour growth or relapse. These dual roles highlight the need to engineer MSC-based therapies that retain tumour-targeting and payload-delivery while minimising protumorigenic and immunosuppressive effects [100].

7.8. Nanotechnology for Targeting Colorectal Cancer Stem Cells

Nanotechnology offers a promising approach for targeting CRCSCs more precisely while limiting systemic toxicity. Thiolated chitosan (TMC) nanoparticles, for instance, can be engineered to improve the oral delivery of drugs, such as 5-fluorouracil and irinotecan, by enhancing stability, mucoadhesion, and receptor-mediated uptake via CD44, which is overexpressed on tumour and stem-like cells. These nanocarriers efficiently encapsulate chemotherapeutics or genetic payloads, protect them from degradation, and enable controlled, tumour-localised release, thereby increasing intratumoural drug concentrations, overcoming resistance, and reducing off-target effects, thus having strong potential in combination regimens to overcome resistance [101,102].

7.9. miRNA-Directed Therapy

MicroRNA (miRNA)-directed therapy aims to modulate dysregulated miRNAs in CRCSCs using synthetic anti-miRs or replacement mimics, often delivered by advanced carriers such as nanoparticles, to diminish stemness and overcome drug resistance [103].

7.10. ctDNA-Guided Strategies

ctDNA-guided strategies use circulating tumour DNA as a real-time marker of minimal residual disease to personalise CRC treatment, escalating or de-escalating adjuvant chemotherapy or immunotherapy, and refining surveillance based on molecular relapse risk rather than clinicopathologic features alone [104].

8. Challenges and Future Directions for Targeting Colorectal Cancer Stem Cells

Despite major advances, effectively eliminating CRCSCs remains difficult. Their quiescent state, enhanced DNA repair, and overexpression of ABC transporters confer multidrug resistance, while concurrent activation of Wnt, Notch, hedgehog, and P13K/AKT pathways provides redundant survival signals that blunt single-target therapies. Additional barriers include metabolic plasticity, EMT-driven metastasis and CSC regeneration, and the creation of an immunosuppressive, PD-L1-rich microenvironment. Tumour heterogeneity and extensive crosstalk with stromal and immune cells, together with molecular similarities between CRCSCs and normal intestinal stem cells, further complicate the development of selective, non-toxic therapies [29].

Future strategies must therefore employ rational combinations that target both bulk tumour cells and CRCSC survival mechanisms. Promising avenues include (i) modulating quiescence, for example, via regulators such as ZEB2, to force CRCSCs into a more therapy-sensitive cycling state; (ii) metabolic reprogramming, exploiting OXPHOS and lipid metabolism dependencies; (iii) miRNA-based approaches to disrupt the stemness network; (iv) EMT inhibition using TGF-β/kindlin-1-directed agents to curb plasticity and metastasis; (v) advanced nanocarriers that concentrate drugs in CRCSC niches and bypass efflux pumps; and (vi) enhanced immunotherapy, including ctDNA-guided chemo-immunotherapy and CAR-T cells targeting CSC-associated antigens. Organoid and stem cell-based models will be crucial for deconvoluting CRCSC heterogeneity and prioritising combinations for clinical translation [65].

Future research is needed to clarify how epigenetic and epitranscriptomic mechanisms, particularly miRNAs and other non-coding RNAs, regulate CRCSC self-renewal, plasticity, and therapy resistance [105]. Recent studies show that specific miRNAs influence key signalling pathways in CRCSCs, including Wnt, Notch, and PI3K/AKT, while long non-coding RNAs (lncRNAs) and circular RNAs (circRNAs) coordinate transcriptional and post-transcriptional networks that maintain stem-like traits and metastatic potential [105,106]. CRCSCs also display pronounced metabolic plasticity, shifting between glycolysis, oxidative phosphorylation, and alternative substrates to adapt to microenvironmental stress and cytotoxic therapies, thereby supporting their survival, progression, and dissemination [107]. Advanced high-resolution methods, such as single-cell multi-omics and spatial transcriptomics combined with metabolomics, will be essential to map CRCSC hierarchies, niches, and metabolic states in patients’ tumours and to identify actionable vulnerabilities with spatial and temporal precision. Translating CRCSC-targeted therapies into clinical practice will require overcoming major challenges, including the development of reliable biomarkers and companion diagnostics, effective patient stratification, safe and efficient delivery of RNA-based or metabolism-targeting treatments, and integration of CRCSC-specific endpoints into clinical trials [107].

An additional emerging challenge in targeting CRCSCs is the phenomenon of oncofoetal reprogramming. This constitutes a significant obstacle to the durable eradication of CRCSCs. During this process, tumour cells and their microenvironment reacquire foetal-like transcriptional, epigenetic, and metabolic characteristics, forming an oncofoetal system [108]. Initially described in pan-cancer studies, this concept links developmental signalling and tissue-patterning pathways to tumour initiation, progression, and immune evasion [108,109]. Recent investigations in WNT-dependent CRC have identified an oncofoetal cell state that cooperates with canonical LGR5⁺ stem cells to drive tumour growth, cellular plasticity, and resistance to standard chemotherapy [110]. Oncofoetal reprogramming elucidates how cancer cells and their niche acquire foetal-like properties to withstand therapeutic interventions. These findings indicate that effective CRCSC-targeted therapies must address both canonical stemness pathways and oncofoetal programmes. Furthermore, oncofoetal markers may facilitate patient stratification and prediction of therapy resistance [109,110,111].

9. Conclusions

Recognition of the CSC model has fundamentally redefined our understanding of CRC biology. CRCSCs constitute a small but critical subpopulation that drives tumour initiation, progression, metastasis, and, importantly, resistance to therapy and disease recurrence. The defining traits include indefinite self-renewal, multilineage differentiation, quiescence, enhanced DNA damage repair, ABC transporter-mediated drug efflux, aberrant activation of multiple survival pathways, metabolic plasticity, miRNA dysregulation, and immune escape, collectively underpinning therapeutic failure. Substantial overlap between CRCSCs and normal intestinal stem cells, especially in surface marker expression and signalling circuits, continues to hinder the development of highly selective, non-toxic therapies. Nevertheless, aberrantly activated pathways and discrete metabolic vulnerabilities in CRCSCs provide attractive targets for intervention.

Conventional treatments can debulk tumours but often spare or even enrich CRCSCs, contributing to minimal residual disease and relapse. Further progress will depend on multidimensional strategies that directly address CRCSC biology, including differentiation-inducing approaches, cell cycle-checkpoint disruption, EMT inhibition, targeting key signalling and metabolic networks, and the deployment of advanced immunotherapies, cell-based treatments, and nanotechnology-enabled delivery systems. Parallel advances in ctDNA-based minimal residual disease monitoring, organoid and stem cell models, and molecularly stratified clinical trials, exemplified by ATOMIC and emerging ctDNA-guided immunotherapy studies, are beginning to translate these concepts toward the clinic.

Sustained efforts to delineate CRCSC heterogeneity, niche interactions, and resistance mechanisms, coupled with rational, biomarker-driven combination trials, will be essential to convert CRCSC-targeted strategies into durable survival benefits for patients with CRC. Overall, this review offered an integrated, mechanistic, and translational perspective that links CRCSC signalling, metabolic and epigenetic regulation, oncofoetal programmes, and tumour microenvironment (TME) interactions to clinically relevant, biomarker-guided treatment strategies.