Facts

  • Gastric cancer is a leading cause of death worldwide.

  • Anoikis is programmed cell death activated upon detachment from the extracellular matrix.

  • Anoikis resistance is a hallmark of metastatic cancers.

  • Anoikis-related genes are prognostic factors for gastric cancer progression.

Introduction

Despite recent advancements in the comprehension of the cellular and molecular mechanisms underlying gastric cancer (GC), it still ranks as the fifth leading cause of cancer-related deaths worldwide [1]. Environmental factors such as dietary habits and chronic gastritis due to Helicobacter pylori infection, along with genetic mutations, can affect GC incidence around the world with a higher frequency in Eastern countries [2,3,4]. Currently, the traditional treatments for GC include surgical resection and/or chemotherapy or radiotherapy, although the 5-year survival rate remains extremely low. In addition, the lack of early-stage symptomatology makes this malignancy difficult to diagnose [5]. All differentiated cells grow in proper tissue environments, but when they lose or leave their native environment, apoptosis signal transduction occurs. Anoikis (Greek for “homeless”) is a mechanism of programmed cell death related to apoptosis, sharing similar pathways and components. Particularly, it is activated under specific conditions that lead the cell to detach from the extracellular matrix (ECM). This type of cell death may occur in several diseases, such as tumors, where it is responsible for the detachment and spreading of cancer cells, cardiovascular diseases, in which anoikis triggers the cardiomyocyte detachment in congestive heart failure, and infectious diseases where anoikis disrupts ECM-mediated cell loss of adhesion [6,7,8]. Cancer cells can evade anoikis, allowing them to survive and spread to distant sites, a phenomenon known as anoikis resistance (AR). This mechanism triggers molecular and biochemical alterations that enhance the invasive and metastatic properties of cancer cells [9]. Additionally, AR contributes to drug resistance and cancer recurrence, posing significant challenges for effective treatment [10, 11]. Recent therapeutic strategies aim to target AR by either promoting anoikis in cancer cells or sensitizing them to programmed cell death [12]. While the general biology of anoikis, AR, and their role in cancer have been previously discussed, GC presents a unique microenvironment and dissemination challenges that necessitate a focused investigation. In this review, we provide a comprehensive analysis of the molecular mechanisms, signaling pathways, and the molecular drivers of AR specifically in GC, incorporating insights from genomic and transcriptomic profiling, including GC-specific signatures (e.g., GC-specific anoikis-related gene and long non-coding RNA signatures). We examine how these molecular alterations correlate with clinical outcomes and immune contexture, and we propose a translational roadmap for targeting AR in GC and improve treatment outcomes. The integration of disease-specific biology with emerging therapeutic strategies provided by this work may be a valuable tool to inform both mechanistic research and the development of precision oncology solutions for GC.

Molecular pathways involved in anoikis activation

Activation and regulation of the caspase pathway

Being a type of apoptosis, anoikis relies on the activation of caspases, crucial enzymes in the intrinsic and extrinsic apoptotic pathways. In this context, the Bcl-2 family plays a pivotal role in maintaining cell homeostasis. Particularly, this family includes both pro- and antiapoptotic proteins capable of modulating the cell fate upon different stimuli. Among the proapoptotic members of this family, Bax and Bak modulate outer mitochondrial membrane (OMM) permeability, allowing the release of cytochrome C, leading to apoptosis activation [13]. On the other hand, Bcl-2 protein exerts its anti-apoptotic activity complexing with Bcl-xL, Bad, and the Apoptosis protease-activating factor-1 [14], preventing the release of cytochrome C, thereby repressing the apoptotic pores formation. Pro-apoptotic proteins, including Noxa, Puma, Bad, Bim, Bmf, and Bik, promote the pro-apoptotic signal induced by genotoxic stress, DNA damage, hypoxia, and other forms of intrinsic cellular damage, such as starvation conditions, and defects in calcium flux, hampering Bcl-2 activity [15,16,17,18,19]. Furthermore, variation of OMM permeability, regulated by the interaction of Bax-Bax or Bax-Bak proteins, leads to the subsequent activation of caspases, thereby initiating the pro-apoptotic cascade that involves initiator caspases, including caspase 9 in the intrinsic pathway and caspase 8 in the extrinsic pathway result into caspase 3 activation [20] (Fig. 1). The extrinsic pathway represents a signal transduction mechanism that is modulated by death factors, including FAS, TNF, and death receptors 1–4 (DR receptors) wherein cell death is triggered by external signals [21]. Death receptors are typically expressed on cell membranes and facilitate the formation of multimolecular complexes through various adapter proteins and enzymes [22]. Several stimuli can enhance their transcription, indicating that the regulation of the extrinsic pathway is not solely dependent on the presence of death ligands [23]. Furthermore, apoptosis is also regulated by IAP (Inhibitor of Apoptosis) proteins, which exert their role by binding and inactivating the procaspases. To date, seven members of the IAP family have been identified, including survivin, which can bind and inhibit procaspases 3, 7, and 9, thus affecting both effector and initiator caspases. Modulators of IAP proteins, such as Smac/Diablo, negatively regulate these inhibitors, exhibiting pro-apoptotic properties [24] (Fig. 1). However, the regulation of apoptosis within a cell is influenced by members of the Bcl-2 family and also by the interplay of caspase inhibitors and their antagonists [25]. The concluding stage of apoptosis involves the generation of apoptotic bodies, wherein the cell disintegrates into numerous small vesicles recognizable by immune cells, particularly phagocytes and, more specifically, macrophages. This terminal phase is marked by the presentation of specific signals on the membrane, referred to as “eat me signals”, which include phagocytosis-inducing molecules like phosphatidylserine. Consequently, the apoptotic cell emits both membrane-bound and soluble signals that attract phagocytic cells, thereby promoting the effective clearance of apoptotic bodies [26].

Fig. 1: Intrinsic and extrinsic apoptosis pathways.
figure 1

Cell death by anoikis is induced by the activation of the intrinsic or extrinsic apoptosis pathway. The intrinsic pathway is activated by stimuli that activate pro-apoptotic proteins, also known as BH-3-only proteins. These proteins inhibit the anti-apoptotic proteins Bcl-2/Bcl-xL, altering the permeability of the outer mitochondrial membrane and determining the release of cytochrome c. Thus, the apoptosome activation triggers caspase 9, an initiator caspase, and consequently, caspase 3, effector caspase. The interplay between IAP proteins and their inhibitors, Smac/Diablo, modulates cell death. In the extrinsic pathway, activated death receptors transduce external signals, ending up in the caspase 8 activation, an initiator caspase, and subsequently caspase 3. In the final stage, neo-formed apoptotic bodies are disrupted by immune cells. Integrin-ECM interactions play a key role in maintaining ECM integrity, which is linked to several key survival factors such as FAK and Src. Growth factor receptors, activated both in a ligand-dependent and -independent manner, cooperate with integrin and laminin to promote cell viability. Cell–cell interactions mediated by cadherins or other cell surface molecules activate signaling pathways similar to those triggered by ECM adhesion.

Full size image

Role of integrins in anoikis

Cells sense their location through specific interactions between integrin family receptors and the ECM, where specific signals are recognized to prevent cell anoikis. Humans have 24 different integrins that are expressed on the cell membrane and are implicated in the recognition of ECM ligands [27]. Furthermore, integrins act as anchoring structures for adapter proteins and kinases involved in cell homeostasis maintenance [28].

It is well-known that integrins activate different signaling pathways according to the type of cell and the regulatory mechanisms in which they are involved. Four integrin types—α5β1, αvβ3, α1β1, and α6β1—are particularly rewarded for their contributions to cell survival across different cellular models, protecting against apoptosis and anoikis. The signaling molecules and pathways associated with integrins include pp125 Focal Adhesion Kinase (FAK), Src kinase, integrin-linked kinase, PI3K/Akt, and mitogen-activated protein kinase (MAPK) [6]. Moreover, several integrins bind to pp125-FAK, a non-receptor kinase that is activated upon cell adhesion, which engages with various signaling molecules, including Src, PI3-K, paxillin, and p130CAS, thereby facilitating the activation of multiple signaling pathways that govern cell migration, proliferation, and apoptosis. In addition, several integrins, such as α1β1, α5β1, and αvβ3, mediate the activation of Src, a member of the SRC kinase family, in the focal adhesion complex and determine ERK and PI3K/Akt signaling cascade. Thus, a Src early inactivation is responsible for a loss of cell adhesion, which turns out into anoikis induction [29].

Furthermore, integrins have the ability to either promote or inhibit actin-based structures, suggesting that the cytoskeleton also delineates the various pathways that connect integrins to the cytoskeletal scaffold. In particular, the cytoskeleton plays a key role in extending the integrin junction, enabling adhesive contacts to develop sufficient strength to survive the contractile forces associated with cellular movement and functionality. Alterations in these mechanical forces can modify the signaling pathways linked to adhesion, thus affecting cell survival [30]. Epithelial cell survival is also associated with the linkage between growth factors and adhesion mechanisms, and the lack of these conditions triggers the apoptotic pathway. A recent study reported that mammary epithelial cells depend on the adhesion of the ECM to laminin and, at the same time, to the insulin-like growth factor 1 (IGF-1) signal to guarantee correct tissue development [31, 32]. Moreover, increased levels of reactive oxygen species (ROS) lead to the activation of the tyrosine kinase Src by integrins that, as previously described, are involved in cell adhesion [33]. Further evidence highlighted the integrins and GFR crosstalk, ensuring target-dependent oligodendrocyte survival under growth factor-limiting conditions [34]. Finally, delineating the roles of integrins and growth factors in apoptosis regulation proved to be a complex attempt.

Cell–cell interactions

As previously reported, anoikis mechanisms are not only related to cell–ECM interactions but also to cell-cell adhesion. Cell–cell interactions are facilitated by a group of membrane proteins known as cadherins, which establish either homotypic or heterotypic bonds between cells in a calcium-dependent manner [35, 36]. The disruption of these interactions can lead the anoikis activation. A recent study demonstrated that the adhesion mediated by E-cadherin activates the PI3K signaling pathway, which is essential for promoting cell survival. Moreover, cadherins may also play a role in cell survival through their indirect interactions with integrins. It has been noted that integrins, such as α2β1 and α3β, are often found at sites of cell–cell contact, where they can engage with EGF receptors, thereby activating downstream signaling pathways that inhibit apoptotic processes [6]. Additionally, recent investigations revealed that other molecules, including P- and L-selectin as well as NCAM, can initiate intracellular signaling cascades like those activated by ECM adhesion, involving pathways such as FAK, Src, PI3K/Akt, and MAPK [37].

Targeting anoikis resistance in cancer

In physiological conditions, anoikis is an essential process for maintaining tissue homeostasis and preventing aberrant cell growth. Indeed, this programmed cell death acts as a safeguard to eliminate displaced cells, ensuring they do not colonize inappropriate locations and compromise tissue integrity. By eliminating detached or misplaced cells, anoikis assumes the role of a barrier to uncontrolled cell proliferation, tumor formation, and metastasis [38]. On the contrary, in cancer, cells acquire resistance to anoikis, enabling them to survive and proliferate in non-adherent conditions. AR represents a pivotal step in the metastatic cascade. During metastasis, cancer cells detach from the primary tumor, survive in suspension within the bloodstream or lymphatic system, and colonize distant tissues [39]. Therefore, tumor cells acquiring AR have enhanced invasive and migratory capacities. As a consequence, AR facilitates the dissemination process and promotes the establishment of secondary tumors [40, 41].

It is well known that the ability to evade anoikis is a hallmark of metastatic cancer and has been identified in numerous malignancies, including breast, lung, ovarian, pancreatic, and colorectal cancers (CRCs). It is a multifactorial process involving intrinsic genetic changes and extrinsic pathways [42, 43]. Integrins play a central role in mediating the interaction between cells and the ECM [44]. A key facilitator of this process is epithelial-to-mesenchymal transition (EMT), a phenotypic transformation that reduces cell adhesion and enhances mobility [33, 45]. EMT-associated transcription factors such as Snail, Twist, and ZEB1 suppress E-cadherin expression and activate survival pathways like PI3K/Akt, aiding in the evasion of anoikis [46]. Cancer cells often alter the expression of specific integrins, such as β1 and αvβ3, to promote survival under detachment conditions [47]. In addition to integrin signaling, AR is driven by the metabolic plasticity of cancer cells. Detached cells often undergo metabolic reprogramming, switching from oxidative phosphorylation to aerobic glycolysis (Warburg effect) to meet their energy demands in low-attachment states [48]. For example, in breast and pancreatic cancers, the detachment from ECM was linked to an increased reliance on autophagy and glycolysis, which provide energy and mitigate stress induced by ECM deprivation [49]. Autophagy, regulated by pathways such as PERK-AMPK, is activated in detached cells to promote survival by recycling intracellular components and reducing ROS production [50].

AR mechanisms vary across cancer types due to differences in tumor biology and microenvironmental factors [10]. In Table 1, a summary of recent studies on AR divided by cancer type is provided. The models used, as well as the mechanisms examined, are reported.

Table 1 Recent studies on the role of anoikis in cancer.
Full size table

In breast cancer, the loss of E-cadherin induces AR by disrupting cell-cell adhesion and activating survival pathways [51]. Similar results were obtained for human ovarian cancer cells via the MEK/ERK pathway, which has been reported to help cells escape anoikis [52]. Integrins, such as β1 and α6, mediate anchorage-independent growth via FAK signaling [53]. EMT transcription factors like Snail and Twist are upregulated, while proteins such as HER2 and αB-crystallin promote survival in non-adherent states [33]. Triple-negative breast cancers often exhibit AR due to enhanced metabolic flexibility [54, 55].

Similarly, in lung cancer, proteins like caveolin-1 (CAV-1) and IL-13 receptor α2 contribute to AR by modulating autophagy and ROS responses. In addition, mitochondrial proteins like Bit1 and pathways such as PERK-mediated autophagy and IL13Rα2/PI3K/Akt signaling enhance AR, further driving metastatic potential [56,57,58]. In a study on ovarian cancer, cell aggregates in ascitic fluid from 15 patients use cell-cell adhesion and integrin signaling to evade anoikis. Moreover, αSMA and fibronectin are key components for the formation of 3D metastasizing units and plasticity maintenance necessary for implant and seeding into the peritoneal lining. These 3D structures are protected from anoikis and chemotherapeutics [59]. In CRC preclinical and clinical studies, tumor suppressors such as HCRP-1 [60] and transcription factors like KLF5 have been implicated in promoting survival under detachment conditions, suggesting their potential role as novel targets in CRC progression and metastatic cascade inhibition [10, 61, 62]. In pancreatic ductal adenocarcinoma (PDAC), AR is closely linked to the PI3K/Akt pathway and its downstream effectors. Overexpression of CDCP1 and the loss of tumor suppressors such as p16INK4a enhance survival and metastatic potential. The metabolic plasticity of PDAC cells, particularly the Warburg phenotype, also supports survival in non-adherent conditions [63].

AR is described in other cancer types, including hepatocellular carcinoma (HCC), esophageal squamous cell carcinoma (ESCC), and glioblastoma [64,65,66]. For instance, proteins like CAV-1, integrins α(5) β(1), and TGFβ/TGFβRI/Smad2 pathway in glioblastoma are critical for enabling survival and metastasis [67]. Malignant melanoma cells exhibit unique adaptations, including bleb formation, which functions as a survival hub during detachment [68].

Understanding the diversity of these mechanisms across cancer types is essential for developing targeted therapies. In fact, given its pivotal role in metastasis, targeting AR has emerged as an attractive field of study, which may lead to the discovery of new druggable targets [12, 69]. In several studies, most of them conducted employing various databases, Anoikis-triggered EMT genes and Anoikis-related genes (ARGs) were obtained and analyzed, and different signatures were developed and correlated with patients’ prognosis, therapeutic response, and chemoresistance acquisition [70, 71]. Indeed, Li et al., Tang et al., and Chen et al. identified five, ten, and seven ARG-prognostic signatures, respectively, in breast cancer by integrating Machine Learning, transcriptomic, clinical data, and experimental validation [72,73,74]. For this reason, therapeutic strategies targeting AR have shown promise in preclinical and early clinical studies. Agents that disrupt integrin signaling, such as cilengitide (targeting αvβ3 and αvβ5 integrins), can restore sensitivity to anoikis [75]. Other classes of anoikis inducer and sensitizer agents include FAK inhibitors. As an example, defactinib prevents the integrin-mediated activation of different downstream signal transduction pathways, including those involving RAS/MEK/ERK and PI3K/Akt. In this way, this inhibits tumor cell migration, proliferation, survival, and tumor angiogenesis [76]. In addition, drugs targeting EMT regulators, including Snail and Twist, are also considered a new therapeutic direction to overcome AR.

Another valuable field of investigation is the use of natural compounds as anoikis modulators [77]. As an example, curcumin, a polyphenol from turmeric, inhibits EMT and integrin-mediated survival pathways. In several studies it has been shown that resveratrol, a stilbenoid found in high concentrations in red grapes, among its pleiotropic functions, suppresses the PI3K/Akt pathway and inhibits cell growth, invasion, and proliferation by targeting NF-κB, Sirt1, Sirt3, LDH, PI-3K, mTOR, PKM2, R5P, G6PD, TKT, talin, and PGAM [78]. Flavonoids such as quercetin modulate ROS to sensitize cells to anoikis, and epigallocatechin gallate, a green tea polyphenol, reduces integrin expression to promote detachment-induced apoptosis [79]. Xu et al. screened 19 phytochemicals and identified apigenin as a potent anoikis sensitizer with anti-metastatic properties in both mouse and in vitro models of breast cancer [80]. Despite this preliminary evidence and fascinating data on the pro-anokis activity of natural compounds, the studies in this field meet several limitations, including a lack of biokinetic studies, poor bioavailability, and specificity. This emphasizes the need for innovative drug delivery systems or combination regimens to enhance efficacy. Nanomedicine, based on novel composite nanocarrier agents, may be considered another precious tool for treating AR. Among most recent studies Wang et al. developed a novel composite nano-enzyme based on mesoporous silica/nano-cerium oxide and obtained promising data for treating AR in HCC cells [81].

These novel approaches, both stand-alone or combined with standard treatment strategies, offer several advantages, including improved delivery, specificity, and a reduced toxicity profile; however, the assurance of their success passes through a deeper understanding of the tumor-specific determinants of AR. Moreover, the simultaneous modulation of multiple anoikis-related pathways, necessary to overcome the robust survival programs adopted by detached tumor cells, necessitates integrative analyses that combine high-throughput genomics, transcriptomics, and patient-derived data. With a deeper investigation of AR molecular machinery, in the future, we may significantly improve outcomes for patients with metastatic cancers, particularly those like GC where peritoneal metastasis leads to a very poor prognosis for patients.

Prognostic value of anoikis in gastric cancer

GC represents malignant neoplasia of the gastrointestinal system characterized by heterogeneous features among the patients [82]. Despite the development of surgical procedures and techniques for diagnosis aimed at improving the patient’s quality of life, there is still a lack of GC diagnosis and treatment in clinics. Therefore, the identification of novel biomarkers, as well as the enhancement of the accuracy of this prediction for GC, is a crucial point. As previously reported, anoikis modulates the biological behaviors of several cancers, affecting cell viability and metastasis [6]. Understanding the role of anoikis is pivotal in cancer research, particularly in the dissemination and metastasis process. Cancer cells must overcome a critical challenge: surviving independently of their attachment to a substrate to successfully enter the circulatory system and disseminate. This requires the development of resistance to anoikis, which allows the cells to evade death upon detachment. Recently, the analysis of ARGs has been used as a novel prognostic strategy. Indeed, Li et al. identified SNCG as a specific ARG in GC by analyzing the Cancer Genome Atlas (TCGA) database. Particularly, they constructed a risk signature for GC patients to evaluate the role of ARGs in prognostic prediction to provide an original predictive tool for GC patients [83]. However, this study had some limitations, as both the gene expression analysis and the clinical GC cohort were derived from TCGA and the Gene Expression Omnibus repositories. Furthermore, these data should be validated by an in vitro/in vivo approach aimed at clarifying the SNGC role in GC progression.

Another study described the identification of further signatures of ARGs to predict the survival outcomes of GC patients, which includes PDK4, NOX4, SERPINE1, MMP11, DNMT1, EZH2, and SNCG genes. Specifically, GC patients were divided into high- and low-risk groups based on the ARGs identified, and the follow-up analyses disclosed that the high-risk score GC patients correlated with a poorer prognosis, which was confirmed by testing several cohorts [84]. The univariable and multivariable Cox analysis results displayed peculiar prediction values of the anoikis-related genes prognostic score within the high-score patients’ group, characterized by a worse overall survival. Further, by analyzing the immune landscape between high- and low-risk groups, the amount of immune cell infiltration was significantly increased in the high-risk group, a phenomenon probably related to the anoikis, as it could regulate cancer progression by affecting immune cell infiltration [84]. Although the authors performed a remarkable analysis of the public databases, the lack of in vitro/in vivo validation gives this study some limitations. Similarly, Cao et al. described another ARG signature constituted of nine genes, including CD36, MSLN, SERPINE1, EZH2, MMP11, OLFM3, TRAF2, TFDP1, and PDK4, whose overexpression is associated with a poor prognosis [85]. In this study, the ARG signature constructed is derived by analyzing public databases. Preliminary data confirm that the ARG signature is upregulated in GC cells; thus, it is necessary to corroborate these results using GC clinical specimens. In addition, another preliminary study identified a further ARG signature including 10 differentially expressed genes involved in both intracellular pathways, such as cell cycle, drug metabolism, and extrinsic pathways, including cGMP-PKG signaling, ECM, and receptor–ligand interaction [86]. Meng et al. described a further ARG signature based on lncRNA (AR-lncRNA) that is associated with GC prognosis and immune infiltration prediction [87]. This model included LINC02241, AL356417.2, AC012073.1, AL391152.1, PVT1, LINC01711, and CYMP-AS1 lncRNAs that are correlated with other tumor prognoses, including breast cancer, melanoma, HCC, glioblastoma, and GC [88,89,90,91,92,93]. However, as these studies have been performed only using an in silico approach, the validation by experimental research is crucial for deepening the AR-lncRNA role in GC progression. Another AR-lncRNAs-based GC prognostic model, constructed using retrospective data, includes AC091057.1, ADAMTS9.AS1, AC090825.1, AC084880.3, EMX2OS, HHIP.AS1, AC016583.2, EDIL3.DT, DIRC1, LINC01614, and AC103702.2, screened from AnoRGs-related lncRNAs [94]. According to other reports, most of the lncRNAs identified directly or indirectly affect tumor progression by regulating signaling pathways associated with cell proliferation and migration [95].

However, despite the convincing evidence provided in these reports, the lack of clinical validation of the above-mentioned prognostic biomarkers represents a crucial limitation.

Mechanisms of anoikis resistance in gastric cancer

The proliferation of tumor cells in GC is facilitated by their AR, which results in metastatic dissemination processes like contact loss, extravasation, invasion, and colonization of districts far from the primary sites typically associated with solid tumors (Fig. 2). In the early stages, cancer cells change their phenotype and cellular plasticity to survive. Growth factors, cytokines, and stress signals often cause the cell to activate survival pathways, leading to this change [96]. In GC, Integrin beta-like 1 (ITGBL1) via the Akt/Fibulin-2 axis (FBLN2) can enhance the metastatic capacity of the gastric tumor cell by determining AR. As evidenced both in vivo and in vitro, ITGB1 depletion leads to a decrease in AR, ultimately leading to metastatic spread [97]. Wang et al. have discovered how CAV1 regulates the process of avoiding death and growth-independent anchoring in EGFR-ITGB1 signaling. CAV1 is responsible for regulating AR by activating Src and pathways like PI3K/Akt and MEK/ERK [98]. A study on AGS cells revealed that a triterpenoid saponin called Platycodin D induces anoikis by activating stress-activated protein kinases like c-JUN and p38 MAPK [99]. Activating the Src and Akt signaling occurs also when claudin-1 (CLDN1) is upregulated. CLDN1 is a member of the tight junction family and is required for cell-cell adhesion. In GC, CLDN1 is overexpressed and acts as an anti-apoptotic protein by binding to β-catenin and promoting the survival of cells [100]. Furthermore, the tumor mass requires a higher supply of oxygen and nutrients and thus activates the process of new vessel formation, known as angiogenesis. The activation of angiogenesis in GC is linked to an increase in platelet-derived growth factor B (PDGFB) after a self-feedback loop from transcription factor C/EBPβ. Once released, PDGFB activates the MAPK signaling pathway in endothelial cells, resulting in a boost in metastatic progression and angiogenesis in GC cells [101].

Fig. 2: Overview of genes-inducing anoikis resistance in gastric cancer metastasis.
figure 2

Deregulation of Anoikis-Related Genes modulates transcriptional and metabolic pathways involved in cell detachment from ECM and promotes tumor dissemination.

Full size image

Once cells detach and survive, they leave the primary site by activating chemotaxis, invasion, and migration processes. In GC, it was observed that the regulation of the plasminogen activating receptor, involved in the fibrinolytic process, is dependent on the transcription factor TCF7L2. Overexpression of TCF7L2 has been shown to promote AR by inhibiting the 3/7 caspase pathway in vivo [102]. A further transcription factor, transcription factor 2 (ATF2), promotes the transcription of growth differentiation factor 15 (GDF15). GDF15, also named the macrophage-1 inhibitor cytokine (MIC-1), is a secreted protein belonging to the superfamily of TGF-β. This increase in GDF15 transcription is associated with the activation of AR by the activation of protective autophagy [103]. Moreover, Wu et al. identified a novel pathway associated with autophagy and AR in GC, which involves ATG4B-circSPECC1-ELAVL1. Particularly, being ATG4B, an RNA-binding protein overexpressed in GC detached cells, it undergoes ubiquitination and degradation upon interaction with circSPECC1. However, as the circSPECC1 is sequestered by the binding to ELAVL1, the inhibition of circSPECC1-ELAVL1 linkage by the FDA-approved compound Lopinavir reestablishes circSPECC1 abundance necessary for GC metastasis suppression [104]. Additionally, it has been demonstrated that nuclear factor MYH9, via its DNA-binding domain, also increases CTNNB1 transcription by inducing AR in both in vivo and in vitro studies. This crosstalk occurs due to the activation by CTNNB1 of the canonical pathway of Wnt/β-catenin, which increases gastric metastasis [105].

Inflammation is also a cellular mechanism that is associated with AR, in addition to protective autophagy. DBC1 is an overexpressed nuclear protein in GC and is responsible for reduced survival rates of patients with GC. The activation of the inflammatory pathway of NF-κB is controlled by this protein, which leads to cell AR [106]. Glucose-regulated protein 94 (GRP94) is also closely associated with cellular resistance to apoptosis, inflammation, proliferation, and invasion [107]. In GC, the transcription factor SP1 was found to induce overexpression of GRP94, which contributes to the onset of peritoneal metastases through the activation of YAP/TEAD1 and therefore of the Hippo pathway. Inhibition of YAP reduced the effects of GRP94 on AR and metastasis [108]. In addition, another mechanism implicated in the regulation of invasion and metastasis of AGS cells includes the α-fetoprotein, a pleiotropic tumor marker, which exerts its function by modulating the anoikis sensitivity, conferring resistance to cell death [109]. Further evidence identified the Endothelial cell-specific molecule 1 (ESM1), a secreted glycoprotein, as a novel biomarker negatively correlated with a poorer outcome in GC patients. Functionally, ESM1 expression promotes GC progression by regulating the EGFR/HER3-Akt/ANGPT2 pathway, which increases cell proliferation, migration/invasion and AR in GC cells [110]. Moreover, using a multi-omics approach in diffuse-type GC mice lacking E-cadherin, p53, and Smad4, Park et al. identified novel effectors such as osteopontin, which is involved in regulating EMT, and Tβ4, a potential E-cadherin repressor, which is involved in promoting cancer cell AR [111]. A summary of all the above-mentioned studies, including biomarkers and AR mechanisms and models, is reported in Table 2.

Table 2 Molecular factors related to anoikis resistance in GC.
Full size table

While the studies summarized in this section highlight a wide range of molecular players implicated in AR and GC, it is important to underline that most of these findings are based on in vitro models or animal studies. To date, there is a striking lack of validation in patient-derived systems or clinical studies. This limits the translational applicability of these biomarkers and mechanisms. A deeper mechanistic investigation, particularly within clinically relevant models, is urgently needed to confirm the role of these pathways in metastatic GC.

Anoikis resistance and metabolism in gastric cancer

GC cells change their metabolism compared to their healthy counterparts. This “metabolic remodeling” (or metabolic reprogramming) is essential to supporting their rapid proliferation and survival in unfavorable conditions like hypoxia and a lack of nutrients in the tumor microenvironment [112, 113]. Activation of cell death programs like anoikis can result from changes in metabolic homeostasis, such as nutrient deficiency or ROS accumulation [10]. Conversely, tumor gastric cells possess a greater capacity to survive in metabolic stress conditions and develop mechanisms of AR. Upregulation of the NADPH oxidase 4 (NOX4) gene has been found to play a crucial role in the development of AR in GC cells. NOX4 is responsible for transporting electrons from NADPH to oxygen, which leads to the generation of H2O2 and ROS. An increase in protein levels causes an up-regulation of EGFR and a higher production of ROS, yielding higher AR in clustered cells than their adherent counterpart. In vivo, the inhibition of NOX4 led to a reduction in AR and metastatic progression of GC [114].

SMAD4 is another GC-related gene whose depletion is associated with an increase in ROS production. Particularly, SMAD4 downregulation correlates with the malic enzyme 2 (ME2) suppression. This latter is involved in lowering the oxidative stress by raising levels of NADPH. In the absence of ME2, the malic enzyme 1 (ME1) isoenzyme is upregulated by the ROS through the ME1-ETV4 axis, which results in resistance to death by anoikis and a rise in the invasion process [115].

In contrast to NOX4, Glutathione Peroxidase 4 (GPX4) is the enzyme responsible for preventing oxidative damage caused by iron accumulation. The gastric cell, via the intrinsic anoikis pathway, induces an increase in mitochondrial ROS production dependent on iron accumulation. As a result, the cell initiates the death process known as anoikis-induced ferroptosis. Sun et al. showed that the increase in GPX4 levels in cluster-detached cancer cells inhibited the ferroptosis process by reducing ROS levels, which promoted the migration of GC cells [116]. Adipose tissue and metabolism associated with fatty acids play a role in the metastasis and growth of gastric tumors related to AR. Obesity and/or a high-fat diet have been shown to increase lipid droplets in the peritoneum and the oxidation of fatty acids, as evidenced by multiple studies [117,118,119]. Diacylglycerol acyltransferase 2 (DGT2) deregulation was associated with C/EBPα-dependent increases in AR, leading to metastatic dissemination at the peritoneum level in the GC [117]. Fatty acid synthase (FASN) is another enzyme that contributes to AR, and it is involved in the synthesis of neonatal fat. In detail, the increase of FASN levels leads to AR through an increase in phosphorylation of ERK1/2 and activation of the anti-apoptotic protein Bcl-xL [120] (Table 3). In response to AR, several cancer types, including GC, experience an increase in proliferation, migration, and invasion processes. Even though comprehensive comparative studies are still lacking in GC, the evidence gathered so far suggests that metabolic reprogramming, along with integrin-mediated signaling and EMT, may constitute major mechanisms leading to AR in this malignancy. These processes converge to support survival under anchorage-independent conditions, maintain redox homeostasis, enhance migratory and invasive properties, and activate pro-survival pathways such as PI3K/Akt and MAPK/ERK [121]. Further investigations using multi- and pan-omics technologies, as well as advanced study models, are necessary to clarify this evidence, particularly in metastatic GC.

Table 3 Metabolic enzymes involved in the anoikis resistance in GC.
Full size table

Conclusions and future perspectives

Anoikis has become a prominent topic in research, with a growing number of studies demonstrating the complex mechanisms cancer cells use to avoid this innate cell death, stressing the necessity for novel targeted approaches in developing effective therapies for anoikis-resistant tumors. Despite being a relatively recent area of study, intensive research on anoikis has yielded significant progress over a decade. In addition, from these investigations, AR appears to be a major driver of metastatic dissemination and therapeutic failure in GC. Through this review, we have examined the complex network of molecular mechanisms that allow GC cells to evade anoikis, including integrin-mediated signaling, EMT, autophagy, inflammation, and metabolic reprogramming. These adaptations enable detached tumor cells to survive under anchorage-independent conditions, promoting peritoneal dissemination and resistance to standard therapies. Nevertheless, our analysis has revealed that most of the current evidence on AR in GC derives from in vitro experiments or animal models. Few of the studies examined have employed patient-derived models or clinical specimens, revealing a significant translational gap in the field. Moreover, while multi-omics analyses have begun to identify prognostic ARG and lncRNA signatures, these have yet to be validated in large, well-annotated clinical cohorts.

To move the field forward, it is essential to integrate functional validation of druggable targets within patient-relevant models, such as patient-derived organoids, microfluidic systems, and xenografts. Moreover, investigating the crosstalk between AR and other regulated cell death mechanisms, such as ferroptosis, necroptosis, or pyroptosis, may uncover synergistic vulnerabilities. In the view of therapeutic applications, targeting AR could improve the efficacy of current regimens and reduce recurrence in metastatic GC. However, the development of such approaches will require a deeper mechanistic understanding of tumor- and context-specific AR pathways. Consolidating the current knowledge on AR in GC and validating the mechanisms, along with the identification of research gaps, remain critical points for translating these findings into effective, patient-tailored therapies.