Enteric neural-tumor interactions in gastrointestinal malignancies and therapeutic implications

Abstract

The enteric nervous system (ENS), long recognized for coordinating myenteric-submucosal motility, secretion, barrier function, and intrinsic afferent signaling and chemical coding, has emerged as a participant in gastrointestinal cancer biology. This review synthesizes evidence positioning the ENS as a mechanistic bridge linking tumor microenvironment (TME) dynamics with systemic brain-gut axis regulation in gastrointestinal cancers. Locally, tumors remodel enteric neuroglial networks through neurotrophic and inflammatory cues, promoting axonogenesis, perineural invasion, galanin-related gastric myenteric remodeling, and phenotypic shifts in enteric glia. Conversely, ENS-derived neurotransmitters and glial mediators, including acetylcholine, serotonin, vasoactive intestinal peptide (VIP)/VIP receptor 2 (VIPR2), substance P, nitric oxide, prostaglandin E2, and S100B, influence cancer stemness, proliferation, VIP/VIPR2-mediated ILC3 and macrophage polarization, immune evasion, metabolic adaptation, and stromal remodeling. Systemically, the gut-brain axis relays stress, PER2-related circadian disruption in APC-driven tumorigenesis, and autonomic outflow to modulate tumor progression via ENS-integrated circuits. This bidirectional crosstalk establishes a “neuro-epithelial-immune niche” within the TME. Recognition of this niche has therapeutic implications, as targeting enteric neurotransmitter receptors, neurotrophin signaling, glial mediators, denervation-related and stress-neural axes has shown preclinical efficacy and repurposing potential. Integrating enteric neuroscience into gastrointestinal oncology reframes tumors as ecosystems shaped by malignant cells, stroma, and gut-wall neural circuits, improving outcomes.

Key Words: Enteric nervous system; Gastrointestinal cancer; Tumor microenvironment; Gut-brain axis; Cancer neuroscience

Core Tip: In this review, we propose that the enteric nervous system (ENS) is an active regulator of gastrointestinal malignancies rather than a passive bystander. We summarize how tumors remodel adjacent enteric neuroglial networks and how ENS-derived neurotransmitters and glial signals in turn shape tumor stemness, immune evasion, stromal remodeling, and perineural invasion. We further connect local tumor microenvironment dynamics with systemic brain-gut signaling and highlight druggable neural pathways as potential adjunctive targets for tumor control and symptom management.

INTRODUCTION

Drawing on GLOBOCAN estimates, the combined burden of esophageal cancer, gastric cancer (GC), and colorectal cancers (CRC) reached approximately 3.41 million new cases and approximately 2.01 million deaths in 2022, accounting for approximately 17% of all incident cancers and approximately 21% of global cancer deaths worldwide[1]. While modern chemo-immunotherapy regimens have improved response rates and extended median survival in selected patients with advanced gastrointestinal tract tumors, gains are still incremental and resistance remains common[2-4]. These survival constraints intersect with a second bottleneck. Treatment intensity is frequently limited by gut-specific dysfunction, including disturbances in motility, barrier integrity, inflammation, and nutritional status, which can drive dose reductions, treatment interruptions, and long-term morbidity[5-7].

Against this backdrop, one major determinant of gut-specific functional resilience is the enteric nervous system (ENS), often described as the “brain in the gut” and increasingly recognized as a central player in gastrointestinal cancer research. ENS is a complex system of neurons and glia in the gut wall[8]. It can generate local reflexes and integrate inputs from the epithelium, immune compartment, microvasculature, and luminal environment to coordinate motility, secretion, perfusion, barrier integrity, and mucosal defense[8]. Concurrently, cancer neuroscience has reframed neural elements as active components of the tumor microenvironment (TME), capable of shaping growth, inflammation, immunity, and treatment resistance across malignancies[9]. In gastrointestinal cancers, a recent study highlights nerves as functional constituents of the TME and discusses neural signaling as a potentially druggable dimension of disease biology[10]. Emerging mechanistic studies support an intrinsic enteric angle by showing that enteric neurons and glia can engage tumor cells and myeloid programs, influence tumorigenesis, and modulate metabolic states and drug sensitivity in model systems relevant to colorectal and GC[11]. This dual relevance to symptom-generating gut function and to tumor-promoting microenvironmental programs motivates an ENS-centered framework for gastrointestinal cancers.

In this review, we summarize emerging evidence that positions the ENS as a central integrator in gastrointestinal cancers. We focus on its roles in local TME dynamics and systemic brain-gut axis regulation. We first examine how tumors remodel adjacent enteric neuroglial networks and how ENS-derived neurotransmitters and glial mediators reciprocally shape tumor cell stemness, proliferation, metabolic adaptation, and immune evasion. We then discuss neuroimmune crosstalk that shapes the tumor immune niche through cholinergic, serotonergic, and peptidergic signaling. We also examine broader ENS effects on epithelial barrier function, vascular perfusion, stromal remodeling, and perineural invasion (PNI). Moving beyond the local microenvironment, we consider systemic regulation via stress-activated brain-gut circuits, vagal-enteric integration in GC, and circadian clock influences on intestinal tumor biology. Finally, we highlight emerging therapeutic strategies targeting enteric neurotransmitter receptors, neurotrophin signaling, glial-derived mediators, and stress-neural axes, positioning ENS-directed interventions as a complementary approach to improve both tumor control and symptom management in patients with gastrointestinal malignancies.

ORGANIZATION AND FUNCTION OF THE ENS

The ENS is a distributed neural network embedded within the gastrointestinal wall that can generate local reflex activity and coordinate digestive and defensive functions[12,13]. Enteric neurons and enteric glia are organized into ganglia arranged as distinct plexuses linked by interconnecting pathways, enabling precise local control of adjacent smooth muscle, blood vessels, glands, and immune tissues[14]. The myenteric plexus lies between the longitudinal and circular muscle layers, whereas the submucosal plexus lies within the submucosa. Nerve fibers connect ganglia within and between plexuses and project to muscle layers, the muscularis mucosa, arteries, and lymphoid tissues, with dense innervation of the mucosa[15].

Although the ENS extends from the esophagus to the anus, its architecture is region-specific. The two-plexus organization is most fully elaborated in the small and large intestines. In contrast, in the esophagus and stomach the vast majority of enteric neurons reside in the myenteric plexus, while the submucosal plexus is sparse in larger mammals and often absent in common laboratory species[8]. One proposed explanation is that upper gastrointestinal regions receive extensive vagal efferent innervation that strongly regulates secretion[16]. This arrangement supports a functional division of labor. Myenteric circuits primarily regulate motility by controlling contractility in longitudinal and circular muscles[17]. Submucosal circuits regulate secretion, absorption, and blood flow, and in larger mammals an outer submucosal plexus can add muscle control[18-20].

Enterochromaffin cells are among the best-defined epithelial sensors and release serotonin (5-hydroxytryptamine, 5-HT), in response to mechanical or chemical stimuli, thereby initiating reflex programs[21,22]. Reflex control elicited from the lumen involves intrinsic primary afferent neurons (IPANs) that project toward the mucosa and are activated by epithelial sensors, together with interneurons and downstream motor, vasomotor, and secretomotor neurons[23,24]. These outputs are generated by circuits that operate together with enteric glia, macrophages, interstitial cells, and enteroendocrine cells to initiate and shape motility and secretion in space and time[8].

Enteric neurons express a diverse array of neurotransmitters, with individual neurons often co-expressing multiple transmitters-one serving as the primary mediator of neuronal function and others acting as co-transmitters or neuromodulators[25]. This neurochemical coding is highly conserved across intestinal regions and species for functionally equivalent neurons. Cholinergic excitatory motor neurons use ACh as their primary neurotransmitter and co-express tachykinins such as substance P as key co-transmitters to mediate smooth muscle contraction[26]. Inhibitory motor neurons utilize nitric oxide as their primary inhibitory transmitter, with vasoactive intestinal peptide (VIP) and adenosine triphosphate serving as co-transmitters to mediate smooth muscle relaxation[27]. IPANs release ACh and tachykinins to transmit sensory signals from the mucosa to enteric circuits[28]. Enterochromaffin cells, the best-characterized epithelial sensors, release serotonin (5-HT) in response to mechanical or chemical stimuli, thereby activating peristaltic reflexes[29]. VIP is the primary transmitter of non-cholinergic secretomotor neurons that regulate water and electrolyte secretion across the intestinal epithelium[30]. Additional neurotransmitters include gamma-aminobutyric acid, somatostatin, neuropeptide Y (NPY), and galanin, which participate in motility regulation, secretory control, and neuroimmune interactions[31].

LOCAL ENTERIC CONTROL OF THE GASTROINTESTINAL TME

Gastrointestinal tract tumors do not expand in isolation within the gut wall. Instead, they evolve within a TME where malignant cells interact with stromal and immune compartments and also with resident neuroglial elements of the gut wall, particularly enteric glia and enteric neurons that are increasingly recognized as functional constituents of the intestinal TME[32]. ENS forms a dense network of neurons and glia embedded throughout the gastrointestinal wall that can generate local reflex activity and integrate epithelial, immune, microbial, and vascular cues to coordinate motility, secretion, perfusion, barrier function, and mucosal defense[17,33]. Cancer neuroscience increasingly emphasizes that neural and tumor compartments engage in reciprocal crosstalk within the TME, with immune and stromal circuits frequently acting as key intermediaries[34]. In gastrointestinal cancers, neural signaling has been linked to broad malignant programs, including tumor cell proliferation and invasion, remodeling of local immunity and stroma, metabolic adaptation, angiogenesis, and resistance to systemic therapies[35]. Conversely, gastrointestinal tumors can remodel the local neuroglial landscape by promoting axonogenesis and nerve infiltration through neurotrophic and guidance cues, fostering PNI as a route of dissemination, and driving phenotypic shifts in enteric glia within the TME[36]. This section summarizes tumor-driven remodeling of enteric circuits and the downstream ENS signals that reprogram tumor cells, immunity, and other dimensions of the niche.

Tumor-driven stress and remodeling of the enteric neuroglial network

Within the gut wall, gastrointestinal tumors develop in immediate proximity to the intrinsic enteric neuroglial network. In mouse models of GC, nociceptive nerves undergo prominent nerve growth factor (NGF)-dependent expansion, and gastric NGF overexpression further increases sensory innervation, supporting a tumor-driven program of neural remodeling within the gastric TME[37]. In CRC, enteric glia are likewise reshaped by inflammatory cues in the tumor immune niche. Single-cell and patient-integrated analyses show that tumor-infiltrating monocytes drive a reactive enteric glial phenotypic and functional switch through glial interleukin 1 (IL-1) receptor signaling and that tumor-associated glia can reciprocally promote monocyte differentiation toward SPP1-positive tumor-associated macrophages (TAMs) via interleukin 6 release, positioning enteric glia as actively remodeled, immunomodulatory stromal elements in CRC[32]. Tumor epithelial cells release soluble factors, especially IL-1, which first activate enteric glia into a pro-tumor phenotype. Then, the activated enteric glia secrete prostaglandin E2 (PGE2), which in turn promotes the proliferation and tumorigenesis of colon cancer stem cells through prostaglandin E receptor 4 and EGFR[11]. Consistent with this tumor-associated glial remodeling, population-based pathology demonstrates that glial fibrillary acidic protein (GFAP)-positive enteric glia can be found within carcinoma stroma but are absent in paired normal mucosa or adenoma, and that GFAP immunoreactivity associates with tumor localization and median survival[38]. At the architectural level, human tissue studies further indicate that ENS remodeling is spatially patterned around the invasion margin, where myenteric plexuses become smaller and contain fewer neurons per plexus, yet these changes are not explained by increased caspase 3 or caspase 8 positive neurons, arguing for nonapoptotic mechanisms of plexus atrophy and stress[39]. Analogous decomposition of myenteric plexuses has been reported in cancer-affected human stomach walls, accompanied by an altered caspase profile and reduced galanin signal within myenteric neurons near the tumor margin[40]. Finally, the neuropeptide milieu itself can shift in patients, as galanin levels are increased in serum and in colon wall layers that contain ENS plexuses, as well as in CRC tumor tissue, supporting the idea that tumors reshape enteric neurochemical coding in addition to structure[41]. These tumor-driven structural, cellular, and neurochemical remodeling programs of the enteric neuroglial niche are summarized in Figure 1.

Figure 1 Tumor-driven remodeling of the enteric neuroglial niche in gastrointestinal cancer: This schematic summarizes three representative routes by which gastrointestinal tumors reshape the enteric neuroglial niche. In gastric mucosa, cholinergic and tuft cell signals amplify neurotrophic support and neural expansion. In colorectal cancer, inflammatory crosstalk between infiltrating monocytes and enteric glia establishes a reciprocal myeloid-promoting circuit. Tumor epithelium can also reprogram enteric glia into a prostaglandin-producing state that reinforces stem-like tumor behavior. Together, these interactions depict the enteric neuroglial network as a dynamic stromal compartment that is actively co-opted to support tumor initiation, progression, and local microenvironmental adaptation. NGF: Nerve growth factor; IL-1: Interleukin 1; IL-6: Interleukin 6; PGE2: Prostaglandin E2; DCLK1: Doublecortin-like kinase 1; SPP1: Secreted phosphoprotein 1; TAM: Tumor-associated macrophage.

Although tumor-driven axon outgrowth and increased innervation have been widely documented, evidence for bona fide neurogenesis and the generation of new neurons within tumors remains limited. In prostate cancer models, doublecortin-positive neural progenitors can infiltrate tumors and initiate intratumoral neurogenesis, generating new adrenergic neurons, establishing a proof of principle that cancers can recruit progenitor-like states and build new neuronal elements in situ[42,43]. In gastrointestinal oncology, tumor-associated neurogenesis has been linked to aggressive behavior in CRC, where higher neurogenesis correlated with poorer outcomes and tumor cells promoted neurogenic responses in vitro[44]. In parallel, the adult ENS retains conditional neurogenic capacity. Lineage tracing and injury paradigms show that enteric glia can generate enteric neurons in vivo after disruption of homeostasis, and experimental colitis can trigger enteric neurogenesis through serotonin-dependent 5-HT receptor 4 signaling that drives glia toward neuronal fate[45,46]. Microbial cues further modulate this plasticity, as microbiota reconstitution and short-chain fatty acids can restore enteric neuronal numbers after antibiotic-induced loss and are associated with stimulation of enteric neurogenesis in vivo[47]. Moreover, in vertebrate models with continuous intestinal growth, post-embryonic de novo enteric neurogenesis has been demonstrated from Schwann cell precursors[48]. Together, these observations motivate a testable hypothesis that gastrointestinal tumors, which create sustained niches of inflammation, dysbiosis, and tissue remodeling, may not only attract nerve fibers but may also activate local glial or Schwann-like progenitors to add new enteric neurons. By contrast, while tumor-associated neuroglial remodeling is increasingly supported by human tissue and mechanistic studies, bona fide enteric neurogenesis within gastrointestinal tumors remains an emerging and still largely inferential concept.

ENS to tumor signaling reshapes the local TME

Beyond simply being remodeled by tumors, the ENS actively participates in shaping the gastrointestinal TME through bidirectional signaling with malignant, immune, and stromal compartments. Enteric neurons and glia release diverse neurotransmitters, neuropeptides, and glial-derived mediators. These signals can directly affect tumor cell behavior, including proliferation, stemness, and metabolic adaptation. They can also reprogram immune cell phenotypes and alter vascular and stromal architecture. This section synthesizes current evidence for ENS-derived signals that reshape three key dimensions of the TME: Direct effects on tumor cells, neuroimmune crosstalk that sculpts the immune niche, and broader regulation of epithelial barrier function, luminal chemistry, vascular perfusion, and PNI tracks. Together, these findings position the ENS as an active orchestrator of the pro-TME rather than a passive target of tumor-driven remodeling. Representative ENS-mediated local signaling axes, including their major mediators, target cells, downstream pathways, and biological effects, are summarized in Table 1.

Table 1 Enteric nervous system-mediated local signaling axes shaping the gastrointestinal tumor microenvironment.

Source component	Key mediator or molecular cue	Key receptor or binding partner	Target cells	Downstream pathway or mode of action	Biological effect	Ref.
Enteric serotonergic neurons	5-HT	HTR1B/1D/1F	CRC stem cells	Wnt/β-catenin signaling	Promotes CSC self-renewal and enhances tumorigenesis	[49]
Enteric neurons	NPY	Not specified	Intestinal epithelial cells	PI3-K/pAkt/β-catenin and miR-375 suppression	Promote inflammation-induced tumorigenesis	[50]
Enteric glia-associated S100B	S100B	RAGE	colon carcinoma cells	RAGE/p38 MAPK/NF-κB	Promotes a pro-inflammatory, pro-angiogenic, and anti-apoptotic microenvironment	[53]
Cholinergic signal axis	ACh	α7 nicotinic ACh receptor	TAMs	JAK2/STAT3	Reduce CRC cell migration and invasion	[59]
VIP signaling axis	VIP	VPAC/VPAC2	RAW264.7 macrophages	Macrophage repolarization	Increasing the M1/M2 ratio enhances the phagocytosis of tumor cells by macrophages	[60]
Platelets	Serotonin/5-HT	/	Tumor cells	Serotonylation	Inhibits tumor growth, increases functional CD8+ T cell infiltration, and reduces PD-L1	[62]
Cholinergic signaling axis	ACh	M3R	CT-26 murine colon cancer cells	EGFR/AKT/ERK	ACh-M3R signaling supports tumor growth and immunosuppressive/cholinergic/angiogenic markers	[63]
Tumor-associated cholinergic machinery	ACh	M3R	LIM-2405, HT-29 cell	EGFR/ERK	Supports PD-L1/PD-L2 expression, proliferation, and migration; blockade reverses these changes and induces apoptosis	[64]
VIP axis	VIP	/	Hepatic sinusoidal endothelial cells (CRC with liver metastasis)	cAMP/PKA	Inhibits hepatic sinusoidal endothelial cells invasion, migration, and capillary-like structure formation, thereby suppressing intratumoral angiogenesis	[72]
Enteric neurons	Neuronal L1CAM/N-cadherin	/	Colorectal tumor epithelial cells	/	Facilitates adhesion to and migration along enteric neurons, supporting local dissemination/PNI	[75]
Enteric neurons	N-cadherin-dependent adhesion		GC cells	/	Diffuse-type GC cells show stronger adhesion to enteric neurons; N-cadherin blockade reduces adhesion, supporting a potential dissemination/PNI route	[76]

5-HT: 5-hydroxytryptamine; HTR1B/1D/1F: 5-hydroxytryptamine receptor 1B/1D/1F; CRC: Colorectal cancer; CSC: Cancer stem cell; NPY: Neuropeptide Y; TAMs: Tumor-associated macrophages; VIP: Vasoactive intestinal peptide; VPAC: Vasoactive intestinal peptide receptor; VPAC2: Vasoactive intestinal peptide receptor 2; PD-L1: Programmed death-ligand 1; PD-L2: Programmed death-ligand 2; PNI: Perineural invasion; GC: Gastric cancer.

Enteric neuro-glia signaling to tumor cells: Enteric neurons and glia can reshape tumor cell programs, including stemness and stress-adaptive metabolism, in part through neurotransmitter and glial mediator signaling. A clear stemness axis has emerged from CRC models. A recent study identified enteric serotonergic neurons as a required niche input[49]. Neuron-derived 5-HT supports CRC stem cell self-renewal and tumorigenesis via 5-HT receptors enriched on stem-like cells and activation of Wnt and beta-catenin programs[49]. In inflammation-driven intestinal tumorigenesis, NPY promotes epithelial proliferation via PI3K Akt and reduces miR-375-linked apoptosis, favoring neoplastic progression[50]. Complementing these transmitter pathways, functional depletion of GFAP-positive enteric glia reduces early tumor burden and dysplasia in CRC models, consistent with glial support for permissive epithelial states[51].

ENS cues also tune tumor metabolic and survival programs. Enteric neuron-gastric tumor organoid co-culture revealed lipid metabolic dependencies and shifted drug sensitivity to lipid-pathway inhibitors, highlighting neuron-dependent metabolic vulnerabilities[52]. In human colon carcinoma biopsies, the enteric glia-associated protein S100B activates RAGE-p38-NF-kappaB signaling and opposes wt p53-dependent apoptosis[53]. Pharmacologic inhibition of S100B dampens this survival pathway ex vivo[53]. Denervation studies provide causal support at the system level. Intrinsic colonic denervation reduces preneoplastic markers in a chemical carcinogenesis model, and myenteric denervation decreases gastric adenocarcinoma incidence and burden in MNNG-treated rats[54,55].

Enteric neuroimmune signaling shapes the tumor immune niche: Under physiological conditions, the ENS can regulate macrophage homeostasis through neurotransmitters such as 5-HT, VIP, and calcitonin gene-related peptide (CGRP), as well as nutrient factors such as CSF1, supporting tissue repair and neural development[56]. Optogenetic studies have revealed that different firing patterns of enterocholinergic neurons differentially regulate immune gene expression and epithelial barrier function[57]. VIP can also confer intestinal mucosal immunity by regulating the activity of ILC3[58]. These intricate neuro-immune circuits, finely tuned for homeostasis and host defense, provide a physiological template that can be co-opted in malignancy.

In gastrointestinal tumors, ENS-linked neurotransmitter and neuropeptide pathways can remodel the immune niche by shifting macrophage functional states. In CRC, α7 nicotinic ACh receptor signaling in TAMs restrains metastasis through a JAK2-STAT3 axis, supporting the idea that cholinergic tone can imprint anti-metastatic myeloid programs[59]. In CT26 murine colon cancer models, blocking VIP signaling increases macrophage phagocytosis and shifts macrophages toward an M1-like state, leading to tumor regression[60]. Similarly, the VIPR2-selective antagonist KS-133 repolarizes macrophages and provides additional benefit when combined with anti-programmed death-1 (PD-1) therapy[61]. Peripheral serotonin provides a second, highly translatable neurochemical lever. In syngeneic colorectal and pancreatic tumors, TPH1-driven serotonin promotes immune escape by increasing programmed death-ligand 1 (PD-L1) via serotonylation and small G protein activation, while genetic depletion of peripheral serotonin or pharmacologic depletion with fluoxetine or the TPH1 inhibitor telotristat boosts functional CD8 T cell accumulation and augments PD-1 blockade to achieve durable control[62]. Muscarinic receptor 3 blockade in an orthotopic CRC model reduces tumor growth and immunosuppressive markers, and human data link cholinergic signaling to PD-L1 and programmed death-ligand 2 expression that can be suppressed by cholinergic antagonists through EGFR-ERK pathway inhibition[63,64].

Due to structural differences in the anatomical distribution of the ENS, there is limited evidence that the ENS regulates immunity in gastric and esophageal cancers. Some studies in GC focus on the vagus nerve, which is physiologically closely coupled with the ENS. For example, vagus nerves infiltrating GC release ACh, upregulate ABHD16A in GC cells, and promote the production and secretion of LysoPS, thereby affecting ILC-related immune responses and enhancing PD-L1-mediated immune tolerance[65]. Upper tract neuroimmune evidence from nonmalignant disease suggests plausible templates for gastric tumors. In the stomach, muscularis macrophages modulate gastric myenteric networks by changing the proportion of nitrergic neurons, indicating that immune tone can rewire ENS output and downstream epithelial repair[66]. The role of the ENS in regulating immunity in upper gastrointestinal tumors requires further investigation. Overall, ENS-linked neuroimmune regulation is strongly supported in preclinical CRC models, but translation to patient stratification or clinically validated immunomodulatory targets remains preliminary.

ENS control of additional TME dimensions: Beyond direct modulation of immune cells, the ENS profoundly influences the gastrointestinal TME by orchestrating changes in epithelial physiology, luminal chemistry, and stromal architecture. This section explores these non-immune, non-tumor cell dimensions.

Enteric glial cells (EGCs) regulate epithelial barrier integrity, a process with dual implications for tumorigenesis. Glial release of S-nitrosoglutathione enhances mucosal barrier function and limits inflammatory injury, while EGCs promote mucosal healing through focal adhesion kinase activation via proEGF release[67]. Similarly, glial cell line-derived neurotrophic factor supports epithelial wound healing and barrier maturation[68]. This protective role, however, is context-dependent. Importantly, aberrant glial programs can sustain epithelial vulnerability; for instance, glial adenosine 2B receptor signaling contributes to persistent barrier dysfunction after acute inflammatory injury[69]. Consistently, a functional subepithelial EGC network is required for efficient epithelial barrier restitution after ischemic injury, highlighting that barrier recovery is an active, glia-dependent process rather than a passive consequence of reduced inflammation[70].

Beyond the physical barrier, ENS output shapes the luminal chemical environment. Experimental models indicate that ENS activity regulates luminal pH, which in turn structures microbial community composition and mucosal inflammation. This provides an indirect route to tune epithelial stress programs and microbe-derived metabolites that impinge on the tumor niche[71]. Extending its influence from the luminal surface deep into the tumor stroma, the ENS also intersects with vascular and connective tissue compartments. VIP suppresses angiogenesis within tumor masses and reduces metastatic growth in a murine colon carcinoma model, indicating that neuropeptides can modulate perfusion and oxygen/nutrient delivery[72]. At the stromal level, CRC exhibits a bidirectional neuro-mesenchymal loop in which norepinephrine induces ADRB2-dependent NGF secretion from cancer-associated fibroblasts (CAF), which then increases sympathetic innervation and promotes tumor progression[73]. Complementing this, PRRX2-driven MMP2-positive myofibroblast-like CAF programs are linked to PNI through TGF-beta signaling and extracellular matrix remodeling, further coupling neural niches to matrix-based invasion tracks[74]. Collectively, these findings illustrate that the neural regulation of the TME actively shapes the physical and chemical landscape that dictates tumor progression.

PNI in gastrointestinal cancers is increasingly viewed as an active tumor-nerve interaction rather than passive spread. In CRC, PNI is defined histologically by tumor cells invading the nerve sheath or encircling a substantial portion of the nerve and is associated with local invasion, metastasis, pain, and worse clinical outcomes[36]. Mechanistically, primary colorectal tumor epithelial cells preferentially adhere to enteric neural structures and migrate along the enteric neuronal network, with L1CAM and N-cadherin contributing to neuron-guided motility, supporting the concept that ENS fibers can act as permissive tracks for dissemination[75]. In GC, co-culture with primary gastric ENS similarly shows stronger adhesion when neurons are present, particularly for diffuse-type cell lines, and this adhesion is reduced by N-cadherin blockade[76].

SYSTEMIC AND CROSS-ORGAN REGULATION OF GASTROINTESTINAL TUMORS VIA THE GUT-BRAIN AXIS

Systemic regulation of gastrointestinal tumors is increasingly framed through the gut-brain axis, a bidirectional network in which central nervous system (CNS) states shape gut physiology through autonomic pathways and humoral signals, with the ENS as the primary local integrator and effector within the organ wall. This chapter synthesizes how vagal, sympathetic, stress, and circadian signals converge on the ENS, the dominant intramural integrator and effector, to shape gastric and colorectal tumor trajectories across organs.

Stress-activated brain-gut circuits drive CRC progression

Although psychological symptoms are often viewed as primarily driven by the CNS, the ENS frequently contributes to psychological or chronic stress-induced intestinal inflammation, barrier dysfunction, and motility disturbances[77,78]. Some CRC patients with a stoma report psychosocial burden, including body image problems and higher anxiety and depression[79]. In CRC, a systematic review and meta-analysis showed an increased risk of depression after diagnosis, supporting the clinical relevance of this comorbidity[80]. More broadly, recent work has proposed that microbiota-gut-brain axis disruption, chronic inflammation, barrier dysfunction, and neuroendocrine stress signaling may link depressive states to gastrointestinal tumor progression[81]. A recent orthotopic CRC study identified an anterior cingulate cortex circuit whose modulation alleviated depression-like behaviors and reduced tumor progression[82]. Chronic psychological stress exacerbates intestinal inflammation through ENS-mediated mechanisms. Sustained glucocorticoid (GC) elevation generates an inflammatory enteric glial subset that secretes CSF1, thereby promoting monocyte recruitment and TNF-mediated inflammation[83]. GCs also induce TGFβ2-dependent transcriptional immaturity in enteric neurons, leading to ACh deficiency and dysmotility[83]. Such ENS-driven inflammation and barrier dysfunction may create a pro-tumor niche that accelerates CRC progression. In mice, GABAergic neurons in the lateral septal nucleus connect to enteric cholinergic neurons via a brain-gut multisynaptic pathway. The nerve fibers of the latter enter the TME and are “hijacked” by CRC cells to maintain tumor growth. Chronic restraint stress can enhance the activity of this circuit and aggravate tumor progression[84]. The endocrine, enteric glial, and polysynaptic neural pathways through which chronic stress promotes intestinal inflammation, dysmotility, and CRC progression are illustrated in Figure 2.

Figure 2 Chronic stress drives intestinal inflammation, dysmotility, and tumor growth through neuroendocrine and neuroenteric pathways. This schematic integrates findings from studies of stress-associated intestinal inflammation and colorectal cancer progression. In the inflammatory pathway, chronic stress activates the hypothalamic-pituitary-adrenal axis, in which corticotropin-releasing hormone from the hypothalamus stimulates adrenocorticotropic hormone release from the pituitary, leading to glucocorticoid (GC) secretion by the adrenal glands. Sustained GC signaling promotes inflammatory enteric glial responses, inducing colony-stimulating factor 1-dependent monocyte recruitment and tumor necrosis factor-mediated intestinal inflammation. In parallel, GC signaling acts on enteric neurons through a transforming growth factor beta 2-associated program, driving enteric neuronal immaturity, reducing mature cholinergic and nitrergic neurons, lowering ACh availability, and contributing to dysmotility. In the neural-tumor pathway, chronic stress activates gamma-aminobutyric acid (GABA)-ergic neurons in the lateral septum, which signal through a septo-enteric polysynaptic circuit involving the sacral parasympathetic nucleus and tumor-innervating enteric cholinergic neurons. These enteric cholinergic neurons release GABA into the tumor microenvironment, activating epsilon-subunit-containing gamma-aminobutyric acid type A receptors encoded by gamma-aminobutyric acid type A receptor subunit epsilon in colorectal cancer (CRC) cells and promoting tetraspanin 1-associated tumor growth. Together, these pathways illustrate how chronic stress can engage endocrine, immune, enteric neuronal, and brain-gut neural circuits that converge on intestinal inflammation, motility dysfunction, and CRC progression. CRH: Corticotropin-releasing hormone; ACTH: Adrenocorticotropic hormone; CSF1: Colony-stimulating factor 1; TNF: Tumor necrosis factor; LS: Lateral septum; SPN: Sacral parasympathetic nucleus; TSPAN1: Tetraspanin 1; GABA: Gamma-aminobutyric acid; GABRE: Gamma-aminobutyric acid type A receptor subunit epsilon; GCs: Glucocorticoids.

Integrated vagal and enteric control in GC

In GC, most infiltrating nerve fibers originate from the vagus nerve[85]. This does not mean that there are only vagal fibers in the tumor and no ENS components. The vagus nerve mainly alters local function in the stomach wall by activating neurons in the myenteric plexus of the stomach. Therefore, many “vagus nerve effects” are essentially amplified and transmitted by the ENS at the tissue level[86]. In mouse gastric antrum imaging, the frequency of vagal nerve stimulation was linearly related to the activation of myenteric plexus neurons[86]. There is a clear intramural neural amplification mechanism in the process of GC. As noted above, NGF overexpression in gastric epithelial cells enhances intramural innervation and promotes carcinogenesis[87]. This kind of “neural amplification” may include the remodeling of vagal terminals and intramural intestinal neural networks at the same time, and the distinction between “vagal axons” and “enteric nerve axons” using commonly used immunomarkers is not always absolutely clear.

Accumulating evidence supports that vagal input promotes gastric carcinogenesis and progression. In an orthotopic xenograft model, unilateral vagotomy reduces tumor burden and implicates cholinergic signaling and the M3 muscarinic receptor in vagus-driven oncogenic effects[88]. Population-based data also suggest a lower long-term GC risk after prior vagotomy[89]. Notably, the axis linking cholinergic signaling, M3 receptor activation, and Wnt signaling can be reproduced in vitro using a functional replacement model that relies only on enteric neurons, in which neurons promote gastric organoid growth in a Wnt-dependent manner and M3 inhibition or deletion suppresses this effect[90]. The vagus provides upstream parasympathetic drive, whereas enteric neural elements can locally relay and amplify this cholinergic program to shape a pro-tumor stem cell niche and microenvironment. However, compared with vagotomy, which is a mature and spatially definable intervention, the field still lacks in vivo strategies that target gastric enteric circuits with comparable specificity and controllability, which may partly explain why direct ENS-focused studies and translational targeting efforts in GC remain relatively limited.

Circadian timing and enteric clocks in gastrointestinal tumor biology

Gastrointestinal circadian timing is coordinated by both central and local clocks. The CNS master clock in the suprachiasmatic nucleus synchronizes gut peripheral clocks through systemic timing cues[91]. Local clocks within the ENS and intestinal epithelium then regulate rhythmic motility, absorption, secretion, and mucosal integrity[91]. Gastrointestinal tissues harbor peripheral circadian clocks that rhythmically gate epithelial proliferation, DNA repair, metabolism, and mucosal immunity, so clock disruption can shift the balance toward tumorigenesis[92].

Light entrains the central clock in the SCN, which then synchronizes gastrointestinal peripheral clocks via autonomic, hormonal, temperature, and behavioral outputs, providing a systemic time cue for the gut[93]. ENS contains its own local clock machinery and helps translate timing signals into rhythmic gut functions, coordinating motor and secretory programs across the intestinal wall[94]. In parallel, feeding time is a potent local zeitgeber that can phase shift gastrointestinal clock gene expression largely independently of the central clock, and gastric entrainment is not mediated by the vagus nerve, highlighting gut intrinsic control routes in which ENS circuits are well positioned to integrate luminal cues[95].

In ApcMin/⁺ mice, inactivation of the core clock gene PER2 increases intestinal and colonic tumor burden, consistent with a tumor-suppressive function of an intact gut clock[96]. More broadly, both genetic clock disruption and environmental misalignment accelerate APC-driven CRC by promoting APC loss of heterozygosity and hyperactivating Wnt signaling[97]. Intestinal clock disruption also drives dysbiosis and impaired barrier function, changes that may facilitate CRC pathogenesis[98]. In metastasis models, circadian dysfunction can act systemically through microbiota-derived signals that expand myeloid-derived suppressor cells and enhance immune evasion[99]. In summary, ENS-related gastrointestinal circadian rhythm dysregulation may be associated with the development and progression of gastrointestinal tumors, but this association still requires further experimental validation.

THERAPEUTIC IMPLICATIONS

Pharmacologic strategies that modulate enteric neural signaling can be broadly organized into three therapeutic categories: Cholinergic circuit blockade, inhibition of neural remodeling and stress-related inputs, and repurposing of drugs targeting neurotransmitter and neuropeptide receptors.

Targeting cholinergic enteric circuits in gastrointestinal cancers

In GC models, pharmacologic disruption of intramural cholinergic signaling has been explored as a therapeutic strategy, including approaches that block ACh release and downstream effector programs and can be combined with systemic therapy[100]. For example, a strategy targeting nerve-cancer metabolic crosstalk using intratumoral botulinum toxin A together with systemic pathway inhibition improved overall survival in mice, supporting the feasibility of drug-based interference with gastric neural drive in established disease[100]. Mechanistically, ACh can promote GC cell proliferation through M3 muscarinic receptor activation of EGFR ERK and AKT signaling, and M3R antagonists such as 4-DAMP and darifenacin inhibit gastric xenograft growth in vivo[101]. Extending this logic to the distal gut, darifenacin blocks M3R-dependent cholinergic signaling in CRC models and reduces tumor growth in vivo, placing muscarinic blockade among the most straightforward examples of drug-based interference with enteric-type cholinergic programs in gastrointestinal tumors[102].

Targeting neural remodeling and stress-related inputs

A complementary therapeutic angle targets both neural remodeling programs that maintain tumor-associated innervation and stress-related neural inputs that reinforce tumor progression. In GC, high tumor NGF expression is associated with increased sympathetic nerve density and poorer survival, and GC cells with high NGF levels promote neurite elongation in co-culture, suggesting that suppressing sympathetic nerve elongation or activation may represent a therapeutic strategy[103]. In CRC, sympathetic and stromal pathways can converge on NGF-dependent innervation, and recent work highlights a beta 2 adrenergic NGF loop that can be interrupted with TRK pathway inhibition in preclinical settings, providing a route to drug targeting of nerve stromal coupling without requiring physical denervation[104]. Systemic adrenergic blockade can also be positioned within an ENS-centered framework as a way to dampen central stress outputs that impinge on gut wall circuits, since propranolol enhances irinotecan efficacy and reshapes anti-tumor immunity in mouse CRC[105].

Repurposing drugs targeting neurotransmitter and neuropeptide receptors

Beyond ACh and neurotrophins, repurposed drugs targeting enteric neurotransmitter and neuropeptide receptors may offer additional therapeutic opportunities. In CRC, pharmacologic inhibition of the 5-HT2B receptor suppresses tumor growth through ERK-linked signaling, supporting serotonin receptor antagonism as a feasible strategy to modulate tumor-relevant serotonergic signaling within the bowel wall[106]. In GC cells, clinically used 5-HT3 antagonists, including ondansetron, palonosetron, and ramosetron, show anti-proliferative effects and induce cell-cycle and autophagy programs, suggesting a potential drug-repositioning strategy targeting this receptor family[107]. Along the substance P-NK1 axis, the NK1 antagonist NKP608 inhibits CRC cell proliferation through Wnt/beta-catenin signaling. The approved NK1 antagonist aprepitant also shows preclinical antitumor activity, with efficacy influenced by dosing time, and additional synergy has been reported when it is combined with 5-fluorouracil[108,109].

Importantly, most ENS-directed therapeutic strategies discussed here remain at the preclinical stage, with limited prospective clinical evidence and no established biomarker-guided framework for patient selection. Nevertheless, the converging evidence for neural regulation of tumor growth, immune remodeling, and treatment response supports ENS-targeted intervention as a promising translational direction, particularly in combination with systemic therapy, immunotherapy, and biomarker-guided patient stratification.

LIMITATIONS AND FUTURE RESEARCH

The ENS is increasingly recognized as an active regulatory layer in gastrointestinal malignancies rather than a passive structure displaced by tumor growth. The evidence reviewed here supports a bidirectional model in which gastrointestinal tumors reshape adjacent enteric neuroglial networks through inflammatory, neurotrophic, and stromal cues, while ENS-derived neurotransmitters, neuropeptides, and glial mediators feed back to regulate cancer stemness, proliferation, immune escape, metabolic adaptation, stromal remodeling, and PNI. This framework extends the conventional view of the gastrointestinal TME by adding a neural dimension that links tumor biology with organ-level functions such as motility, barrier integrity, secretion, perfusion, pain, nutritional tolerance, and treatment-related gastrointestinal morbidity. A deeper implication of this model is that the ENS should not be viewed as a uniformly tumor-promoting system. Enteric neural and glial signals are context-dependent. The same pathways that maintain epithelial barrier repair, immune homeostasis, and tissue resilience under physiological conditions may be co-opted by tumors to support growth, invasion, and immune suppression. Conversely, some neural signals may restrain metastasis or promote anti-tumor immunity in specific settings. Therefore, the central challenge is not simply to inhibit enteric neural activity, but to define which neuroglial cell states, transmitters, receptors, and circuit outputs are pathogenic in a given tumor type, anatomical region, disease stage, and treatment context.

Several limitations define the current field. First, the evidence base is uneven across gastrointestinal cancer types. Mechanistic data are strongest in CRC, where enteric glia, serotonergic neurons, cholinergic signaling, and neuroimmune interactions have been examined in preclinical and patient-linked systems. In GC, many studies implicate vagal or cholinergic input, but the relative contribution of intrinsic enteric circuits vs extrinsic autonomic fibers remains difficult to resolve. In esophageal cancer, ENS-focused evidence remains particularly limited. Second, many studies rely on denervation, co-culture, organoid, or xenograft approaches that establish biological plausibility, but do not always distinguish causality from adaptive remodeling. Third, common neural markers often lack the resolution needed to separate sensory, autonomic, and intrinsic enteric fibers in human tumor tissues. These limitations argue for more precise anatomical, molecular, and functional interrogation of tumor-associated neural niches.

Future studies should move from descriptive neuroanatomy toward causal and spatially resolved cancer neuroscience. Single-cell and spatial transcriptomic profiling should be integrated with multiplex imaging, neural tracing, electrophysiology, organoid-neuron co-culture, and genetically or pharmacologically controlled perturbation models. Such approaches could clarify whether tumor-associated neuroglial remodeling reflects axonogenesis, phenotypic switching, selective neuronal loss, glial reprogramming, or true local neurogenesis. They could also identify the cellular sources and targets of key mediators such as ACh, serotonin, VIP, CGRP, NGF, PGE2, and S100B within defined neuro-epithelial-immune niches. Translational development will require a biomarker-guided framework. Candidate neural biomarkers may include nerve density and topology, receptor expression, glial activation states, neurotrophin signatures, neuroimmune ligand-receptor pairs, PNI features, and circulating or tissue-based neurotransmitter-related metabolites. These markers should be evaluated not only as prognostic indicators, but also as tools for patient selection, pharmacodynamic monitoring, and prediction of response to immunotherapy, chemotherapy, or ENS-directed adjunctive treatment. Importantly, clinical translation must account for the essential physiological roles of the ENS. Therapeutic strategies that target enteric neurotransmitter receptors, neurotrophin signaling, glial mediators, or stress-neural axes should be designed to preserve gut motility, barrier function, secretion, and mucosal defense while disrupting tumor-supportive neural programs.

ENS-directed therapy is therefore best viewed as a complementary strategy rather than a stand-alone replacement for established oncologic treatment. Repurposed agents targeting muscarinic, serotonergic, adrenergic, neurotrophin, or neuropeptide pathways may be particularly attractive because their safety profiles and gastrointestinal effects are partly known. However, prospective validation, rational dose selection, tissue-specific delivery, and combination timing with chemotherapy or immune checkpoint blockade remain unresolved. Addressing these questions could transform ENS-related mechanisms from an emerging biological concept into a clinically actionable dimension of gastrointestinal oncology.

CONCLUSION

In summary, integrating enteric neuroscience into gastrointestinal cancer research reframes tumors as organ-embedded ecosystems shaped by malignant cells, immune and stromal compartments, microbial and systemic cues, and the neural circuits of the gut wall. This perspective offers a more comprehensive explanation for the coexistence of tumor progression, treatment resistance, gastrointestinal dysfunction, and impaired quality of life. It also creates a path toward therapies that aim not only to suppress tumor growth, but also to preserve the neural and epithelial functions that determine patient resilience during cancer treatment.