Cancer-associated fibroblasts sit at the center of gastrointestinal metastasis
Metastasis remains the dominant cause of death across gastrointestinal cancers, and the stromal compartment has moved from background noise to a central biologic driver of that process. In this review, cancer-associated fibroblasts, or CAFs, emerge as the most influential nonmalignant cells shaping the metastatic trajectory of colorectal, gastric, pancreatic, biliary, and esophageal tumors. Their influence spans the full chain of dissemination: local invasion, vascular entry, immune evasion, distant seeding, and eventual resistance to treatment.
The picture is not one of a single stromal villain. CAFs are heterogeneous, plastic, and deeply context dependent. Some subsets promote tumor spread, while others may restrain it under certain conditions. In pancreatic ductal adenocarcinoma, for example, specific myCAF populations can suppress progression, while inflammatory CAFs tend to reinforce invasion and immune suppression. Antigen-presenting CAFs add another layer of complexity because their immunologic behavior shifts with tumor context and location.
Matrix remodeling creates the physical route for invasion
One of the most immediate ways CAFs accelerate metastasis is by reorganizing the extracellular matrix. Dense stromal fibrosis is not just a structural feature of gastrointestinal tumors; it actively changes how cancer cells move, survive, and interact with surrounding tissues. CAFs deposit matrix proteins, alter collagen alignment, and reshape tissue stiffness, creating a microenvironment that favors migration and invasion.
This process matters because tumor cells do not metastasize in isolation. They navigate a mechanically remodeled scaffold that can direct movement along aligned fibers and facilitate penetration through tissue boundaries. A stiff, fibrotic matrix also changes mechanotransduction signals inside cancer cells, helping them adopt more invasive phenotypes. In pancreatic and other highly desmoplastic GI tumors, this stromal architecture is often so pronounced that it becomes part of the disease biology rather than a passive bystander effect.
Paracrine signaling pushes tumor cells toward dissemination
CAFs also act as signaling hubs. Through paracrine secretion, they provide growth factors, cytokines, and chemokines that sustain malignant behavior and help tumors adapt to hostile conditions. These signals can promote epithelial-mesenchymal transition, survival under stress, and enhanced motility, all of which support metastatic spread.
Inflammatory CAFs are especially important here. Their secretory profile supports a pro-tumor immune milieu and can amplify oncogenic pathways in neighboring cancer cells. The biologic consequence is a tumor compartment that is not merely protected by stroma, but actively instructed by it. In a metastatic setting, such signaling can help cells detach, survive transit, and colonize distant organs.
Immune exclusion and immune suppression blunt anti-tumor defenses
The review places immune exclusion among the most consequential CAF-mediated mechanisms in GI cancer metastasis. CAFs contribute to a tumor microenvironment in which effector immune cells struggle to penetrate the malignant core or function effectively once they arrive. This can occur through physical barriers created by matrix deposition as well as through cytokine-driven suppression of immune activity.
The result is a sanctuary for tumor growth and spread. Immune cells that would normally recognize and eliminate disseminating cancer cells are kept at the periphery, reprogrammed, or rendered ineffective. Antigen-presenting CAFs may influence this process in different ways depending on tumor type, which reinforces the need to avoid treating all CAFs as a single biologic category. A one-size-fits-all stromal strategy is unlikely to succeed when the immune consequences differ so widely across CAF subsets.
Angiogenesis, lymphangiogenesis, and pre-metastatic niches extend the reach of disease
CAFs do not simply support established tumor nests. They help prepare the routes cancer cells use to leave the primary site and establish colonies elsewhere. The review details roles in angiogenesis and lymphangiogenesis, both of which expand access to circulation and lymphatic pathways. These vascular programs increase the probability that tumor cells will enter transport channels and seed distant organs.
Even more striking is the CAF contribution to pre-metastatic niches. Before tumor cells arrive, stromal and immune conditioning can alter distant tissue environments to make them more receptive to metastatic colonization. CAF-derived factors participate in that preparation, effectively priming remote sites so that incoming cancer cells encounter a more permissive soil. In GI cancers, where metastatic spread often determines survival, the ability to build these niches is biologically and clinically consequential.
Therapy resistance grows out of the stromal environment
CAFs also shape how tumors respond to treatment. Their influence extends to therapeutic resistance, a major barrier in advanced gastrointestinal malignancies. Dense stromal signaling can limit drug penetration, blunt apoptosis, and sustain survival pathways after systemic therapy begins. In parallel, immune suppression created by CAF-rich microenvironments can weaken the activity of immunotherapies.
This explains why eliminating cancer cells without addressing the surrounding stroma often leaves residual disease behind. The fibroblast compartment can serve as a reservoir of resistance, helping tumors rebound after chemotherapy or targeted therapy. In practice, the stromal niche may allow small surviving populations to persist long enough to seed recurrence and metastasis.
Broad CAF depletion has failed because the biology is not uniform
The review argues against indiscriminate CAF elimination, and the evidence summarized makes that caution hard to ignore. Broad depletion strategies have produced limited clinical success, in part because CAFs are not uniformly malignant. Some subsets protect against progression in specific contexts, and removing them can strip away structural or immunologic constraints that were, at least partially, containing the tumor.
This is where subtype specificity becomes essential. MyCAF, iCAF, and apCAF populations should not be treated as interchangeable. Their distribution varies by organ site, spatial niche, and metastatic stage. A therapy aimed at one stromal function may suppress spread in one setting while worsening disease in another. The biology demands precision, not blanket stromal ablation.
A blunt anti-fibroblast strategy can misfire when the tumor depends on both tumor-promoting and tumor-restraining stromal programs.
Subtype-aware targeting is the most credible path forward
The most clinically relevant message in the review is the need for context-aware CAF-targeted strategies. Future therapies will likely need to identify which CAF subsets are active, where they sit within the tumor, and what stage of metastasis they are influencing. Blocking a pro-metastatic inflammatory circuit may be useful in one cancer type, while preserving or even harnessing restraining fibroblast populations may matter in another.
That approach fits the current understanding of GI tumor ecology. Metastasis is not driven by a single cell population but by coordinated interactions among cancer cells, fibroblasts, immune cells, vascular structures, and matrix components. CAFs sit at the intersection of those systems. Their heterogeneity makes them harder to target, but also makes them a rational focus for next-generation therapy.
For now, the central limitation is clear: therapies built around broad CAF removal do not account for fibroblast diversity, tissue context, or metastatic stage. The field will need interventions that map CAF subsets with greater precision before stromal targeting can move from concept to reliable clinical strategy.