2.1 Intrusion-Related Reprogramming

The initial step in metastasis is the local invasion of surrounding tissues, requiring tumor cells to undergo changes that enhance motility, invasiveness, and their ability to penetrate the vascular system, either directly or via the lymphatic system, to enter the bloodstream. Cancer stem cells (CSCs) are typically characterized by aggressive proliferation. They possess the capacity for self-renewal, the ability to generate phenotypically diverse tumor cell subpopulations, and tumor-initiating potential, positioning them at the apex of the tumor cell hierarchy [3]. Two key factors enable CSCs to escape from their primary site: epithelial–mesenchymal plasticity (EMP) and alterations within the TME [4] (Figure 1). Biomechanical changes in the extracellular matrix (ECM) also play a significant role in promoting cellular invasiveness [5]. Degradation of the ECM by tumor and stromal cells reduces its structural integrity, creating additional space for tumor expansion. Alterations in ECM composition and increased mechanical stiffness, common in most solid tumors, contribute to resistance against apoptosis and enhance tumor cell invasiveness [6]. EMT is recognized as a pivotal cellular mechanism in malignant tumor progression and metastasis [7]. Under abnormal activation of EMT, tumor cells are rich in a variety of dynamic changes, including decreased adhesion, loss of polarity, and tight junctions, resulting in an enhanced ability to migrate and invade adjacent tissues. EMT is implicated in tumorigenesis, metastasis, and therapeutic relapse across various cancer types, including breast cancer [8] prostate cancer [9], colorectal cancer [10], esophageal cancer [11], gastric cancer [12], and lung cancer [13]. It facilitates the acquisition of tumor heterogeneity and cancer stem cell-like properties, serving as a critical driver of tumor invasion and metastasis [14, 15].

The EMT and mesenchymal-epithelial transformation (MET) programs have traditionally been viewed as binary switches between two distinct static states. However, recent research has revealed that EMT is not an all-or-none process. Instead, it often occurs in an incomplete manner, with individual cancer cells activating EMT to varying extents, resulting in hybrid or partial E/M states. Such tumor cells simultaneously express epithelial and mesenchymal markers at different levels, exhibiting mixed phenotypic, transcriptional, and epigenetic traits characteristic of both epithelial and mesenchymal cells [16]. Emerging evidence indicates that the hybrid E/M state confers enhanced stem cell-like properties, tumor-initiating potential, and increased invasiveness [17]. Moreover, prostate cancer cells in a hybrid E/M state express elevated levels of stem cell markers and form faster-growing, larger tumors in vivo [18]. In triple-negative breast cancer, Bierie et al. identified an E/M hybrid subpopulation expressing integrin β4 (ITGB4), which exhibited greater stem-like traits, tumor-initiating capacity, and poorer prognosis [19]. Collectively, these findings suggest that stem-like properties and tumor-initiating capacity are predominantly associated with the acquisition of a hybrid E/M state rather than a fully epithelial or mesenchymal phenotype.

2.2 Metastatic Circular Transmission-Dissemination

After undergoing invasive transformations, malignant tumor cells can detach from the primary tissue and invade the bloodstream or lymphatic vessels, becoming circulating tumor cells (CTCs). CTCs hold immense clinical value, with current research on liquid biopsy of CTCs primarily focusing on three areas: capturing CTCs to determine their quantity, identifying CTC phenotypes to evaluate tumor status, and analyzing genetic variations in CTCs to reveal tumor heterogeneity [20]. Studies on CTCs provide insights into predicting distant metastases, assessing prognosis, and monitoring therapeutic responses [21-23].

These CTCs typically circulate in microclusters that acquire mesenchymal or stem cell-like properties. They are often encapsulated by platelets, immune cells, or stromal cells of tumor origin, which enhance their stem cell characteristics, proliferative potential, and ability to evade immune detection [24] (Figure 1). CTC clusters, in particular, exhibit greater viability than individual CTCs, clustered structures provide CTCs with a unique hypoxic microenvironment that promotes their survival [25], significantly contributing to their dissemination and subsequent metastasis formation [26]. However, while in circulation, CTCs face numerous environmental stressors, including immune killing, apoptosis, oxidative stress, blood shear forces, and nutrient deprivation. As a result, the majority of CTCs undergo apoptosis, with only a small fraction surviving and successfully initiating metastasis [27]. To navigate these challenges, CTCs must integrate intrinsic cellular modifications with external microenvironmental signals, adapting to the dynamic conditions. This leads to substantial temporal and spatial heterogeneity in the molecular phenotype, transcriptome, translatome, and cytological characteristics of CTCs, even within the same parental tumor population [28]. Interactions between CTCs and circulating microenvironment components further modify the microenvironment, facilitating the migration and long-distance spread of CTCs [29-31]. For instance, Lin et al. demonstrated that in brest cancer (BCa) cells, transforming growth factor-beta (TGF-β) signaling can induce tumor-associated fibroblasts (CAFs) to transform into a (platelet-derived growth factor receptor alpha (PDGFRα⁺) integrin subunit alpha 11 (ITGA11⁺) subpopulation. This subpopulation promotes lymphangiogenesis and ECM remodeling, creating pathways that enhance the invasion of lymphatic vessels by BCa cells and augment their overall invasive potential [32]. Liu et al. discovered that CTCs can bind to platelet-derived regulator of G protein signaling 18 (RGS18), leading to the upregulation of the immune checkpoint molecule major histocompatibility complex, class I, E (HLA-E), thereby enabling CTCs to evade immune surveillance by natural killer (NK) cells [33]. During extravasation, CTCs are restored to an epithelial phenotype by MET and regain their adhesion capacity and stem cell properties, which are important for the metastatic regeneration of CTCs in distant organs [34, 35].

Currently, various therapeutic strategies targeting the TME or CTCs themselves have been developed to limit CTC survival, thereby reducing metastasis or preventing disease progression into more aggressive phenotypes. These approaches include inhibiting the EMT, reversing the MET, and clearing CTCs from circulation. For example, in breast cancer, Rg3 liposome loading with docetaxel has been used to capture circulating CTCs and disrupt the lung microenvironment, effectively suppressing metastasis [36]. Additionally, eribulin treatment in breast cancer patients has been shown to improve survival by modulating the epithelial–mesenchymal state of CTCs [37]. However, early clinical trials have been terminated due to low recruitment rates or challenges in detecting CTCs, highlighting the critical need for improved CTC isolation and enrichment technologies for future clinical studies.

CTCs serve as a critical bridge linking primary tumors and metastatic sites. They not only represent certain biological characteristics of the primary tumor but also reflect some behaviors of metastatic lesions. Therefore, in-depth investigations into the unique traits of CTCs—such as stemness, epithelial–mesenchymal state, immune evasion, and drug resistance—are essential. Identifying key molecules on or within CTCs and exploring their roles in mediating organ-specific metastasis will be pivotal in uncovering the mechanisms underlying distant organ-specific tumor dissemination [28]. Furthermore, developing targeted therapies against CTC-specific antigens holds great promise for clinically inhibiting metastasis.

2.3 Organ Infiltration and Metastatic Colonization

Disseminated tumor cells (DTCs) have the capacity to spread to nearly all organs, yet distinct cancer types exhibit a preference for specific organs, a phenomenon referred to as metastatic organ tropism. This tendency is shaped by the characteristics of the targeted organs and the organ-specific microenvironment that facilitates DTC seeding and colonization [38]. Upon arrival in distant tissues, the majority of DTCs are swiftly eliminated—either by innate or adaptive immune responses or due to their failure to adapt to the new environment, resulting in cell death [39, 40]. Surviving DTCs may enter a dormant phase, wherein they remain viable but cannot sustain proliferation, as their growth is suppressed by immune mechanisms or constrained by the stromal components of the TME, resulting in minimal net metastatic expansion [2, 41] (Figure 1). In this dormant state, DTCs can persist for extended periods. However, due to their low numbers and minimal activity, detecting micrometastatic clusters clinically remains challenging, making it difficult for healthcare providers to accurately assess patient status at this stage [42, 43].

The TME comprises a variety of immune and stromal cells that co-evolve and interact with tumor cells. Initially, the TME tends to exert anti-tumor effects, inhibiting tumor cell growth during the early phases of colonization. Over time, however, tumor cells adapt and co-opt the TME, evading immune surveillance while promoting colonization through mechanisms such as regeneration, angiogenesis, and the activation of immunosuppressive pathways [44]. For instance, breast cancer cells can alter fibroblast phenotypes by secreting IL-1α and IL-1β, which induces lung fibroblasts to produce C-X-C motif chemokine ligand 9 (CXCL9) and CXCL10 via nuclear factor kappa-B (NF-κB) signaling, thereby promoting the development of lung metastases [45]. Additionally, fibroblasts activated by insulin-like growth factor 2 secreted from inhibitor of DNA binding 1 (ID1)-overexpressing cancer cells can secrete vascular endothelial growth factor (VEGF), which primes the TME and facilitates the involvement of bone marrow cells in supporting tumor growth and distant metastasis [46]. Similarly, (transforming growth factor-beta 1 (TGF-β1) secreted by colorectal cancer (CRC) cells induces anterior gradient 2 release from recruited neutrophils, creating a positive feedback loop that accelerates CRC metastasis [47]. A deeper understanding of how cancer cells hijack host environments is crucial to curbing metastatic colonization. By advancing this knowledge, preventing the establishment of metastatic colonization may remain the most effective strategy against the progression of secondary tumors.

The metastatic cascade is multifaceted, with the same molecules and pathways often playing distinct roles depending on the context—particularly during the primary tumor's proliferation and its dissemination. Phenotypic plasticity, defined as the dynamic, non-genetic adaptation of tumor cells to the metastatic cascade and their ability to respond to shifts in the TME, is now recognized as a key hallmark of metastasis and a critical area of research [19]. Currently, studies have found that circadian rhythms as well as the microbiome play key roles in the immune response against cancer metastasis. The latest study explored the impact of the biological clock on the TME and tumor metastasis and identified an important role for the biological clock-regulated gut microbiota and its metabolites in the evolution of the TME and tumor metastasis [48]. Metastasis represents an ongoing process of cellular and microenvironmental reprogramming, where the selective evolution of tumor cells is able to withstand microenvironmental stresses [49]. As metastasis progresses, unchecked tumor growth leads to systemic organ dysfunction and deterioration, ultimately culminating in the collapse of vital organ function and death.

3.1 Genetic Inheritance of Tumor Metastases

Understanding the relationship between gene mutations and metastasis is essential for advancing knowledge of tumor biology and developing targeted therapies to inhibit metastasis. Chromosomal instability (CIN) is another critical factor, as it leads to abnormal chromosomal alterations that enhance the adaptability of tumor cells to various environments, thus promoting invasion and metastasis [50].

3.2 Epigenetic Regulation of Tumor Metastasis

For primary tumors to metastasize, a cascade of adaptive changes driven by both genetic and epigenetic modifications is required. Epigenetics encompasses regulatory mechanisms that influence gene expression without altering the underlying DNA sequence [63]. This dynamic regulation plays a pivotal role in metastasis [64], as aberrant activation of epigenetic elements is a hallmark of tumor cells, contributing to both drug resistance and metastatic potential [65]. Therefore, epigenetic regulation is fundamental to tumor progression. Mechanisms such as chromatin remodeling, DNA modification, RNA modification, and non-coding RNA play significant roles in tumor metastasis and merit further exploration [66].

4.1 Phosphorylation

Phosphorylation is one of the most evolutionarily conserved PTMs, notable for its dynamic and reversible characteristics, leading to diverse functional outcomes. This process involves the addition of phosphate groups to amino acids such as serine, threonine, or tyrosine, regulated by ATP, protein kinases, and protein phosphatases [110, 111]. The reversible nature of phosphorylation is mediated by phosphatases, including specific serine and threonine phosphatases [108]. Phosphorylation is integral to numerous cellular functions, including protein–protein interactions, stability, signal transduction, transcriptional regulation, intracellular localization, and apoptosis [112]. Aberrant phosphorylation has been recognized as a key factor in disease development and progression, particularly in cancer. For instance, mitogen-activated protein kinase 1 (MAPK) signaling-mediated phosphorylation and the subsequent subcellular translocation of pyruvate dehydrogenase E1 component subunit alpha (PDHE1α) have been linked to tumor immune evasion. Furthermore, low PDHE1α phosphorylation levels are associated with poor prognosis in patients with lung cancer [113]. Research indicates that mutations in kinases or phosphatases can alter phosphorylation patterns, disrupting the intricate balance of cancer-related proteins and acting as both contributors to and consequences of tumorigenesis [114]. These insights emphasize the importance of unraveling the complex interplay between phosphorylation and disease pathogenesis, particularly in cancer.

4.2 Acetylation

Acetylation is a fundamental mechanism for gene expression regulation, primarily governed by lysine acetyltransferases (KATs) and lysine deacetylases [115, 116]. KATs, also known as histone acetyltransferases (HATs), consist of a catalytic subunit and accessory proteins that transfer acetyl groups to lysine residues [117, 118]. Histone deacetylases (HDACs), classified into four groups—Class I, Class IIA, Class IIB, Class III, and Class IV—remove acetyl groups, thereby modulating acetylation [119]. Acetylation plays a vital role in chromatin remodeling and transcriptional activation, influencing essential biological processes such as tumor cell proliferation, apoptosis, metastasis, and stem cell behavior. Research underscores the significance of histone acetylation regulators in managing HCC and improving patient outcomes, offering promising potential for clinical application [120]. NAT10 has also been implicated in promoting colon cancer progression by inhibiting ferroptosis through N4-acetylation and stabilization of ferroptosis suppressor protein 1 (FSP1) mRNA [121]. Recent findings highlight histone H2B N-terminal multisite lysine acetylation (H2BNTac) as an active enhancer, regulated by CBP/p300 and HDACs 1 and 2, marking active enhancers and serving as a predictor for CBP/p300 target genes [122]. Acetylation modifications also influence immune function and tumor immunity [123], suggesting important implications for developing novel cancer therapies.

4.3 Ubiquitination

Ubiquitination is a key post-translational modification involved in regulating numerous cellular processes, including proteasomal degradation, DNA damage repair, and cell cycle progression [108]. The attachment of ubiquitin to target proteins determines their fate by modulating interactions with proteins containing various ubiquitin-binding domains. The role of ubiquitination in cancer is multifaceted. For instance, SMAD-specific E3 ubiquitin protein ligase 2 (SMURF2) has been shown to act as a tumor suppressor by regulating genomic stability through ubiquitin-dependent modification of ring finger protein 20 (Rnf20). Additionally, SMURF2-mediated ubiquitination and degradation of NAD-dependent deacetylase sirtuin 1 (SIRT1) inhibit CRC cell proliferation and tumor growth [124]. Linear ubiquitination also plays a significant role in modulating NF-κB-dependent inflammatory signaling and immune responses by mediating epidermal growth factor receptor (EGFR)-induced NF-κB activation, which contributes to tumorigenesis [125]. Recent studies have also highlighted the importance of non-lysine ubiquitination in various biological processes, such as endoplasmic reticulum-associated degradation, immune signaling, and neuronal functions. This expands the scope of ubiquitination beyond protein substrates to non-protein targets, revealing a novel ubiquitination cascade that regulates the integrated stress response (ISR). Activation of ISR has emerged as a potential therapeutic strategy [126]. In addition to ubiquitination, studies on the deubiquitinase family can offer new insights into the understanding of cancer pathology [127]. These insights are expected to propel significant advances in ubiquitin research, offering new opportunities for therapeutic intervention [128].

4.4 Ubiquitin-Like (UbL)

The PTM of UbL proteins constitutes a complex signaling language that governs nearly all cellular processes [129]. The UbL pathway has become a promising therapeutic target due to its central role in cancer, and research into UbL modifications has been abundant in several areas.

5.1 Glucose

Glucose is a critical energy source for the survival and proliferation of tumor cells. Due to mitochondrial dysfunction and hypoxic conditions, cancer cells primarily rely on aerobic glycolysis to extract glucose from their surroundings [143]. Additionally, tumor cells can absorb glycolytic byproducts such as lactate and pyruvate, which further promote invasion and migration. Lactate activates the NF-κB signaling pathway, thereby enhancing metastasis, and also regulates DNA repair mechanisms, contributing to both tumor invasion and chemotherapy resistance [144, 145]. Elevated plasma pyruvate levels stimulate the hypoxic response via hypoxia-inducible factor 1α (HIF1α), which sustains circulating tumor cells and influences angiogenesis, as well as EMT processes [144]. Pyruvate also inhibits HDAC3, facilitating tumor cell migration and promoting metastasis [146, 147]. Tumor cells display markedly increased expression of pyruvate kinase 2, which activates EMT and signal transducer and activator of transcription 3 (STAT3) signaling pathways [148] and enhances the activity of fructose-2,6-bisphosphatase (PFKFB2), sustaining glycolysis and driving cancer progression [149, 150]. Further investigation into the metabolic regulation of the microenvironment reveals that glucose uptake and metabolism profoundly influence tumor-associated macrophages (TAMs), which in turn promotes metastasis and heightens resistance to chemotherapy [142, 151]. Elevated fructose levels also contribute to cancer progression through increased glycosylation [152]. Moreover, oxidative phosphorylation plays a pivotal role in cancer metabolism [153]. Understanding these metabolic adaptations and their effects on the TME sheds light on the mechanisms underlying cancer progression and treatment resistance.

5.2 Lipid

Excessive lipid uptake, coupled with abnormal expression or hyperactivation of key enzymes involved in de novo lipid synthesis and catabolism, supports tumor growth, invasion, and metastasis [154]. For instance, the transmembrane protein cluster of differentiation 36 (CD36), responsible for fatty acid transport, is upregulated during metastasis, leading to substantial fatty acid absorption. This accumulation drives the EMT and accelerates cancer progression [155]. The lipid synthesis pathway, particularly mediated by fatty acid synthase, has been identified as a vital driver of metastasis [156]. Fatty acid-binding proteins (FABPs) are pivotal in regulating lipid metabolism and energy balance, with several FABPs overexpressed in tumors and implicated in tumorigenesis. Notably, FABP4 facilitates metabolic crosstalk between ovarian cancer cells and adipocytes during omental metastasis [157]. FABP5 promotes lymph node metastasis in cholangiocarcinoma by activating lipolysis and fatty acid synthesis, increasing intracellular fatty acids and triggering NF-κB signaling [158]. Moreover, fatty acid oxidation generates mitochondrial reactive oxygen species, which promote EMT in tumor cells [159]. Recent studies have also highlighted the significant role of cholesterol metabolism in tumor development [160], and tissue-specific lipid metabolism has been linked to the organ selectivity of metastasis [161]. These findings on lipid metabolism adaptations offer novel insights into the mechanisms driving cancer progression.

5.3 Amino Acid

Reprogramming of amino acid metabolism similarly orchestrates energy production, redox homeostasis, and other metabolic pathways critical to cancer metastasis. Arginine metabolism, for example, is intertwined with fatty acid metabolism and regulates ovarian tumor progression [162]. Additionally, arginine acts as a feedback regulator of tumor metabolism, functioning similarly to a second messenger, thereby furthering tumor progression [163]. Variations in branched-chain amino acids (BCAAs) significantly affect tumor cell behavior. The enzyme branched-chain amino acid transferase 1, which catalyzes BCAA metabolism, is highly expressed in many malignant tumors and is linked to increased tumor proliferation, invasion, and metastasis [164]. Elevated amino acid metabolism is especially evident in tumor cells, particularly for glutamine. Glutamine metabolism produces α-ketoglutarate, tricarboxylic acid cycle intermediates, and ATP, which metastatic tumors rely on to modify cell adhesion and activate invasion signaling [144]. Additionally, tryptophan metabolism is frequently upregulated in tumors, enhancing the invasion-metastasis cascade [165]. This increase in tryptophan metabolism fosters tumor stemness, mediated by BicC family RNA binding protein 1, and contributes to drug resistance [166]. Glycine metabolism, regulated by glycine decarboxylase and serine hydroxymethyltransferase 2, plays a significant role in tumor progression [167, 168]. Furthermore, serine metabolism promotes perineural invasion and metastasis by modulating S-adenosylmethionine levels [169, 170]. These insights into the reprogramming of amino acid metabolism deepen our understanding of the metabolic adaptations that underlie cancer metastasis.

5.4 Nucleotide

The invasive behavior of cancer cells, regardless of type or genetic background, consistently depends on metabolic processes, particularly the aberrant activation of nucleotide synthesis and utilization [171]. Nucleotide metabolism plays a multifaceted role in cancer metastasis. Dysregulation of key signaling pathways, such as MAPK/ERK (extracellular regulated protein kinases) and mammalian target of rapamycin, in tumor cells enhances nucleotide synthase activity, which in turn supports rapid cell proliferation and increases invasive potential [172]. Elevated nucleotide synthesis also aids in repairing DNA damage caused by the oxidative and stressful conditions within cancer cells, thereby improving the survival of migrating cells and contributing to treatment resistance [173]. Moreover, nucleotide metabolism influences the TME by regulating immune evasion. For example, increased adenosine synthesis by tumor cells inhibits T-cell cytotoxicity [174], while the secretion of uracil nucleotides facilitates the recruitment of TAMs and further promotes immune escape [175]. Targeting these metabolic pathways offers a promising therapeutic approach to inhibit tumor metastasis.

5.5 Other Metabolites

Beyond the three major metabolites—glucose, lipids, and amino acids—tumor metastasis is intricately regulated by other metabolites. Acetic acid, for instance, upregulates c-MYC expression, thereby reshaping metabolic pathways and promoting immune evasion [176]. Succinic acid has been shown to drive metastasis both directly and indirectly [177, 178]. Similarly, d-2-hydroxyglutarate (D-2-HG) induces epithelial–mesenchymal transition and is linked to distant metastasis in colorectal cancer [179]. Interestingly, the precursor α-ketoglutarate (α-KG) can act synergistically with SLC25A (mitochondrial carrier subfamily of solute carrier protein 25) inhibitors to reprogram metabolism and enhance tumor sensitivity to radiotherapy [180]. These metabolites in the TME often exhibit dual functions, reflecting the complexity and diversity of their roles. Understanding their precise mechanisms remains a significant challenge in unraveling the intricacies of tumor metastasis.

6.1 Overview of the TME

The TME is an intricate and dynamic system that plays a pivotal role in promoting tumor development, progression, metastasis, and resistance to treatment [181]. It comprises diverse components, including tumor cells, non-tumor cells, ECM, and exosomes. Key non-tumor cells include endothelial cells, pericytes, CAFs, immune-inflammatory cells [182], and lipids [183] (Figure 5). The TME is defined by several hallmark features: acidity, hypoxia, angiogenesis, chronic inflammation, and immunosuppression. These conditions create a complex interplay of processes that drive tumor progression [184, 185].

6.2 Implication of Cellular Components of TME on Metastasis

The TME consists of a variety of cell types, including stromal and immune cells, which interact to create a dynamic environment that can either promote or inhibit tumor development and metastasis.

6.3 TME Heterogeneity

However, the composition and behavior of the TME exhibit significant variability across different cancer types, ultimately influencing the metastatic patterns and treatment responses unique to each tumor. In breast cancer, tumors frequently establish a pre-metastatic niche through interactions with fibroblasts and immune cells. TAMs predominantly adopt an M2-like phenotype, which fosters immunosuppression and angiogenesis, thereby facilitating metastasis to the lungs and bones [224, 225]. Conversely, CRC is characterized by heightened activity of CAFs, which promote ECM remodeling that supports liver metastasis. Notably, CRC cells secrete TGF-β, which drives the transformation of neutrophils into pro-tumor phenotypes, enhancing metastatic dissemination [5, 47]. Prostate cancer cells have a distinct propensity for bone metastasis, a phenomenon attributed to their capacity to exploit the osteoblastic niche. These malignant cells engage with osteoblasts and osteoclasts, thereby altering bone remodeling processes to form metastatic lesions [226, 227]. Recent advancements introduced by Zuo et al. have introduced a deep learning approach termed stKeep, designed to parse TME heterogeneity from spatial omics data. This innovative methodology integrates multimodal and molecular network data, facilitating the construction of cell, gene, and communication modules, which allows for an in-depth analysis of tumor ecosystem heterogeneity [228]. The advancement of emerging technologies offers new perspectives for understanding the mechanisms by which the heterogeneity of the TME influences metastasis, thereby providing a scientific basis for future individualized treatments.

The heterogeneity of the TME not only dictates the mechanisms underlying metastasis among different cancer types but also profoundly impacts the microenvironmental differences observed in metastatic target organs within the same cancer type. The establishment of metastatic lesions in distant sites is supported by pre-metastatic niche, which display diverse regulatory functions that promote metastasis across various tissues [229]. Lymphatic metastasis generally involves lymphangiogenesis and the remodeling of high endothelial venules. In contrast, pulmonary metastasis may promote the generation of Tregs, thereby creating an immunosuppressive microenvironment that further supports metastatic dissemination [230]. In contrast, pulmonary metastasis may engage alveolar epithelial cells to promote the generation of Tregs, forming an immunosuppressive microenvironment [231]. The liver's abundant vascular supply enhances angiogenic pathways, increasing the likelihood of metastasis [232]. In comparison, brain metastasis is closely associated with the activity of microglial cells and extracellular vesicles (Evs), while bone metastasis relies on the physiological activities of osteoclasts and osteoblasts [233-235]. Researchers often use target organ metastasis models based on the specificity of different cancer types to explore the mechanisms of tumor metastasis, for example, lung and breast cancers tend to metastasize to the brain [236, 237], whereas prostate cancer primarily metastasizes to the bone [238]. Commonly, there are in vitro models, such as cell lines and organoid systems, which can simulate the interaction between tumor cells and the host microenvironment. Studies have shown that breast cancer cells exhibit specific growth and metastatic abilities in brain-like microenvironments [239]. Meanwhile, in vivo models (e.g., xenograft models and genetically engineered mouse models) allow observation of the behavior of tumor cells in vivo, providing an opportunity to study the gene regulatory mechanisms of prostate cancer in skeletal metastasis [240]. However, each of these models has its own advantages and disadvantages: in vitro models may not be able to fully mimic the complex microenvironment in vivo, while in vivo models, although biologically relevant, are costly and cumbersome. Therefore, integrating the results from different models will contribute to a comprehensive understanding of tumor metastasis mechanisms and provide new perspectives and directions for subsequent studies.

7.1 Applications of Biological Markers

Tumor markers, whether produced by neoplastic cells or triggered by the host in response to neoplastic stimuli, are critical biomolecules that act as objective indicators in the assessment of normal physiological processes, pathogenic mechanisms, and therapeutic responses. These markers, with their distinct molecular, histological, radiological, and physiological characteristics, enable precise and quantitative evaluations [241]. Tumor markers have become essential tools in clinical practice, particularly for monitoring treatment efficacy and detecting metastatic recurrence, making them a key component in dynamic observation and guiding clinical decision-making.

Biomarkers specific to tumor metastasis not only facilitate early disease detection but also inform the development of targeted therapies, thereby providing significant clinical value. Circulating biomarkers such as cell-free mitochondrial DNA (mtDNA), cell-free viral DNA, RNA, and EVs are commonly employed as tumor indicators. For example, a pan-gastrointestinal cancer early detection model, based on circulating free DNA (cfDNA) methylation analysis from 1781 tumors, demonstrated superiority over conventional detection methods [242]. Chemokines and their receptors are essential in the metastatic process, aiding in the identification of metastatic sites and the establishment of pre-metastatic niches [243]. Additionally, matrix metalloproteinases, which are involved in remodeling the TME, serve as important biomarkers for metastasis [5]. Despite the growing application of biomarkers in clinical diagnosis and treatment, challenges such as specificity and variability remain.

Immunotherapy has made significant advances in treating metastatic tumors, yet certain limitations persist, resulting in suboptimal outcomes for some patients. Biomarkers like PD-L1 expression and tumor mutational burden are currently used to predict responses to immunotherapy, but their predictive power is still under investigation [244, 245]. There is a critical need for more precise biomarkers to accurately predict immune responses, identify resistance, and foresee potential side effects. Emerging biomarkers in this field include tumor-specific antigens, immune checkpoint proteins, and immune cell subset characteristics, all of which show potential for improving predictive accuracy and therapeutic optimization. A recent study identified CD8-fit T cells as predictive biomarkers for immunotherapy response through single-cell functional assessment [246]. The development of novel biomarkers for predicting immunotherapy outcomes, particularly using next-generation sequencing technologies, is essential for advancing metastatic tumor treatments in the future.

7.2 Noninvasive Biomarker of Metastasis

Tissue-based molecules face significant limitations due to tumor heterogeneity, making the reliance on single-tissue biopsies potentially inadequate for guiding treatment decisions [247]. The American Society of Clinical Oncology assessed the prognostic utility of various molecular biomarkers from prostate cancer tissue but refrained from recommending their routine use. The lack of prospective validation and insufficient evidence of improved long-term outcomes were key reasons behind this decision [248]. In June 2016, the United States Food and Drug Administration (US FDA) approved the first liquid biopsy to detect EGFR mutations in patients with non-small cell lung cancer, marking the beginning of a new era in tumor liquid biopsies [249].

In response to the growing need for noninvasive diagnostic tools, several biomarkers have been explored, including protein-based, DNA-based, mRNA-based, and circulating miRNAs [250, 251]. A recent study developed and validated a noninvasive multidimensional biomarker assay combining pretreatment circulating tumor DNA (ctDNA) and peripheral T cell signatures to predict immunotherapy response in advanced non-small cell lung cancer [252]. Moreover, findings from the TRACERx study using ctDNA to track early metastatic dissemination in lung cancer have supported advancements in (neo)adjuvant trials and provided insights into metastasis detection through low-ctDNA-level liquid biopsies [253]. The noninvasive 5hmC-Seal technique, which evaluates 5hmC in cfDNA, has also shown promise as a tumor marker for glioma screening and disease monitoring [254]. Additionally, plasma snoRNA SNORD33 has been identified as a predictor of sensitivity to platinum-based chemotherapy in triple-negative BC, aiding in the selection of effective treatments [255]. Innovations like fluorine isotope 19 MRI have enabled the identification of TAMs in glioma genesis and allowed for prognosis monitoring and drug resistance tracking, previously achievable only through invasive methods [256]. Furthermore, a fecal microbial-based classification model combined with CA19-9 has shown 94% accuracy in pancreatic cancer identification, highlighting the potential of noninvasive screening methods [257]. Despite these advantages, large-scale clinical trials are still needed to substantiate the accuracy and cost-effectiveness of these biomarkers.

Exosomes, due to their transfer properties, hold significant promise for developing therapeutic vectors in oncology and drug applications. Their ability to cross the blood–brain barrier makes them particularly valuable for diagnosing central nervous system (CNS) diseases and monitoring CNS status [258]. Consequently, exosomes are emerging as promising diagnostic biomarkers in immunotherapy contexts [259]. The exosome protein hyaluronan-binding protein has been identified as a predictor of brain metastasis progression and patient survival, offering a novel target for preventing and treating brain metastases [260]. Although exosomes demonstrate remarkable transcellular permeability and biocompatibility, which are advantageous for drug delivery, the cost of exosome detection remains high, and evidence supporting co-targeted assays is still limited.

Research into metabolic products in tumor metastasis is unveiling new biological mechanisms and offering critical insights for developing novel diagnostics and therapies. Studies show that alterations in specific metabolic pathways, along with the activation of related signaling pathways, are closely linked to tumors’ invasive and metastatic capabilities, as well as their survival in distant organs. For example, changes in levels of lactate [261], fatty acids [262], and pyruvate [263] can reflect the metabolic state of metastasizing tumors and serve as biomarkers for tumor dissemination.

8.1 Small Molecule Drugs

Small molecule inhibitors, recognized for their selectivity, small size, and ability to target intracellular molecules, are considered an ideal first-line treatment for relapse or metastasis due to their minimal side effects [266]. However, the cost of developing a new drug that gains US FDA approval can exceed US$2 billion, making drug development a costly and intricate process requiring extensive efficacy screening [267]. Fortunately, discovering new functions and indications for existing drugs offers a cost-effective alternative to de novo drug synthesis [268]. Table 1 compares small-molecule inhibitors and biologic drugs across key factors like mechanism of action, delivery methods, and clinical use, offering insight into their roles in cancer treatment and metastasis.

Drug repositioning has opened new possibilities for anticancer treatments, as demonstrated by several cases. For instance, researchers identified the transcription factor KLF11 as a negative regulator of sarcoma CSCs through a CRISPR-CAS9 library screen, while thiazolidinedione, an oral hypoglycemic agent, has shown promise in combination therapies for multiorgan sarcomas [273]. Similarly, the tyrosine kinase receptor inhibitor mubritinib, targeting ERBB2, appears to be a viable treatment option for acute myeloid leukemia [274]. Additionally, recent research identified neurokinin-1 receptor (NK-1R) as a target for combating drug resistance and recurrence in colorectal cancer, with alprazolam acting as a NK-1R antagonist [275]. The antimalarial drug Atovaquone has been shown to inhibit the growth of primary and drug-resistant BC by blocking human epidermal growth factor receptor 2 (HER2)/β-catenin signaling [276]. Despite strong evidence supporting its therapeutic potential, regulatory approval and intellectual property issues have delayed its market entry [277].

Combining epigenetic drugs with conventional oncology therapies offers a promising approach to cancer treatment. Several classes of epigenetic modulators are currently being investigated for their potential in targeting metastasis pathways. Among them, histone deacetylase inhibitors (HDACis), BET protein inhibitors, and EZH2 inhibitors are at the forefront of clinical trials (Table 2). These agents work by modulating chromatin structure and gene expression to interfere with key processes involved in metastasis, such as EMT, cell migration, and invasion. Table 2 summarizes the ongoing clinical trials of epigenetic drugs targeting tumor metastasis pathways [281, 282]. Additionally, isocitrate dehydrogenase (IDH) enzyme inhibitors have been approved by the US FDA for the treatment of acute myeloid leukemia (AML) patients with IDH1 and IDH2 mutations, but these drugs have not been evaluated in the context of metastatic solid tumors [283]. Similarly, although BRD4 inhibitors such as OTX015 [284] and TEN-010 [285] have shown promise in preclinical models of solid tumors, their role in inhibiting metastasis has not yet been extensively tested in clinical trials targeting metastatic disease [286].

The future of small molecule inhibitors looks promising, particularly with the use of predictive non-clinical models to identify potential drug candidates for human trials. Additionally, dynamic indicators from liquid biopsies could play a vital role in monitoring tumor recurrence and metastasis, complementing research into drug-refractory targets. In summary, these advancements may significantly enhance the effectiveness of small molecule inhibitors in cancer treatment [287].

8.2 Macromolecular Drugs

Recent advancements in cancer therapy have underscored the potential of chimeric antigen receptor T (CAR-T) cell therapy and other innovative strategies in addressing various malignancies. A screening study utilizing a protein antibody library has identified a promising CAR-T target for multiple myeloma [288]. Additionally, MSA-2, a non-nucleotide agonist of human STING, was discovered through high-throughput screening and has shown encouraging results in tumor regression and the induction of durable anti-tumor immune responses [289]. Comprehensive knowledge of a tumor's systemic metastatic status is vital for effective drug development. To address this, researchers developed the DeepMACT pipeline, an automated system that quantifies tumor metastasis and evaluates therapeutic antibody targeting, systematically detailing tumor size, shape, spatial distribution, and targeting efficiency [290]. Advancing these treatments requires a multidisciplinary approach that integrates protein engineering, tumor serology, and a deep understanding of immune-tumor cell interactions [291, 292]. Future improvements must focus on enhancing target precision, increasing cytotoxic payloads, refining antibody-drug linker technology, and optimizing combination therapies, including immunotherapy, once underlying mechanisms are fully understood [291].

Nanotechnology is playing a pivotal role in advancing drug delivery systems. For instance, the synthesis of nanostructured MUV-2, a hierarchical mesoporous iron-based metal-organic framework, enables high payload storage of the macromolecular drug paclitaxel, improving its selectivity for HeLa cancer cells while sparing noncancerous HEK cells. NanoMUV-2 also enhances cell permeation, cytosolic delivery, and cancer-targeting effects, thereby increasing paclitaxel's efficacy against cancer [293]. Furthermore, bispecific antibodies targeting CD3 and CAR-T cell therapies are capable of redirecting T cells toward B cell maturation antigen (BCMA)-expressing multiple myeloma (MM) cells, inducing cytotoxicity in target cells [294].

Despite the success of antibody-based therapies targeting programmed cell death protein 1 (PD-1)/PD-L1, cytotoxic T lymphocyte-associated antigen-4, HER2, CD20, and CD38 some treatments have been linked to serious adverse effects [295]. Recent studies have shown that tumor-infiltrating immune cells and CD8 + T cell function exhibit circadian rhythmic fluctuations, with tumor endothelial cells regulating immune cell infiltration through intercellular cell adhesion molecule-1 expression. The research also found that altering the timing of cell therapy and anti-PD-1 antibody treatment, especially in the evening, can significantly enhance therapeutic efficacy. By optimizing treatment timing, the effectiveness of immunotherapies can be improved, providing new insights for personalized treatment and clinical trial design [296]. Consequently, ongoing monitoring of both tumor and systemic immune responses following immune checkpoint inhibitor treatment is essential to gain deeper insights into the mechanisms of efficacy and resistance. Such knowledge can guide the development of novel immunotherapies or combination treatments with broader, more durable efficacy and improved safety profiles [295].

8.3 Cellular Therapy

The field of cellular therapy, particularly CAR-T cell therapy, has transformed cancer treatment, with approvals for refractory pre-B-cell acute lymphoblastic leukemia and diffuse large B-cell lymphoma (DLBCL) [297]. Clinical trials have demonstrated the effectiveness of CAR-T cell therapy in adult patients with refractory or relapsed DLBCL [298, 299]. Notably, the US FDA has approved CD19 CAR-T cell therapy Yescarta (axicabtagene ciloleucel) for treating adult patients with large B-cell lymphoma (LBCL) who are refractory to or have relapsed within 12 months of first-line chemoimmunotherapy. Despite these achievements, CAR-T therapy faces significant obstacles, including life-threatening toxicity, limited antitumor efficacy, antigen evasion, and restricted tumor interactions [300].

Macrophages are also emerging as promising targets for cancer immunotherapy. Sipuleucel-T (Provenge, Dendreon), the first FDA-approved therapeutic cancer vaccine for metastatic castration-resistant prostate cancer, uses a fusion protein that combines a tumor-targeting antigen with granulocyte-macrophage colony-stimulating factor to stimulate antigen-specific T cells and boost antitumor activity [301]. A notable development in this field is the recent initiation of clinical trials for HER2-targeted chimeric antigen receptor macrophage (CAR-M) therapy by Carisma Therapeutics [302]. NK cells, essential for tumor destruction, are also being explored in CAR-based therapies. CAR-NK therapy offers several advantages over CAR-T, including increased safety and the ability to activate multiple cytotoxic mechanisms [303, 304]. In a clinical trial using HLA-mismatched CAR-NK cells derived from cord blood, 11 patients with relapsed or refractory CD19-positive cancers responded positively to the treatment without significant toxicities [305]. CAR-NK therapy leverages the principles of CAR-T therapy while incorporating the innate biological advantages of NK cells, attracting significant attention in scientific and industrial circles. Neutrophils, as frontline defenders and key immune regulators, play dual roles in cancer, either promoting or inhibiting tumor growth through N1/N2 polarization, a topic of substantial interest. Recent studies have shown that CAR-engineered neutrophils significantly enhance antitumor activity and can act as vehicles for targeted drug delivery [306]. Additionally, targeting NETs has emerged as a promising strategy to combat cancer metastasis [307], although effective systemic delivery of these therapies remains a considerable challenge.

8.4 Other Drugs

Traditional cytotoxic antitumor drugs exert their effects by disrupting DNA synthesis or repair in cancer cells, thereby inhibiting cell division and proliferation. Although chemotherapy has broad antitumor efficacy, it is often associated with significant side effects and the development of drug resistance, which can ultimately lead to tumor metastasis. In contrast, endocrine therapy has become the standard of care for estrogen receptor (ER)-positive BC and has achieved substantial clinical success. When combined with cyclin-dependent kinases 4/6 (CDK4/6) inhibitors, it demonstrates direct antitumor activity; however, resistance and distant metastasis still occur in some patients [308]. Increasing evidence underscores the importance of immune modulation in overcoming treatment resistance. Recent advances in immuno-nanomaterials—either with intrinsic immunogenicity or designed to efficiently deliver immune modulators—have shown promise in enhancing immune responses and effectively targeting metastatic tumors [309]. Manganese-based and virus-like nanomaterials, in particular, have been found to significantly improve antitumor immune responses against metastatic disease [310, 311]. Despite these advancements, emerging chemical immune modulators require further optimization to improve targeting specificity and minimize side effects.

8.5 Emerging Therapies and Combination Approaches

The evolution of precision medicine has spurred the development of novel therapies and combination strategies designed to overcome the limitations of conventional treatments in metastatic cancers. Recent advances have focused on targeting new molecular pathways, including those involved in TME and metabolic reprogramming, which have shown promising therapeutic potential. Immunotherapies, particularly immune checkpoint inhibitors, when combined with targeted therapies or chemotherapy, have demonstrated significant improvements in efficacy and the potential to reduce resistance. For instance, the combination of pembrolizumab with chemoradiotherapy has notably enhanced progression-free survival in patients with newly diagnosed, high-risk, locally advanced cervical cancer [312]. Similarly, the NALIRIFOX regimen, which combines liposomal irinotecan, oxaliplatin, fluorouracil, and leucovorin, was recently approved by the FDA as a first-line treatment for metastatic pancreatic cancer (NCT04083235) [313]. Furthermore, metabolic reprogramming of both cancer cells and the immune cells within the TME is emerging as a promising therapeutic approach [314, 315], and combination therapies utilizing oncolytic viruses have emerged as a promising strategy for advancing innovative and effective tumor treatments in the future [316]. While these strategies are in early stages of development, they offer substantial potential for enhancing therapeutic outcomes. However, no combination therapy involving metabolic modulators and immunotherapy has yet received FDA approval. To optimize clinical trial designs and accelerate the drug discovery process, the integration of robust biomarker-driven patient stratification is critical.

Artificial intelligence (AI) is increasingly recognized as a powerful tool in identifying novel therapeutic targets, improving patient selection, and refining clinical trial methodologies, which could significantly enhance the success rates of emerging therapies [312, 317, 318]. AI has significantly increased the efficiency of drug discovery, especially in handling large-scale experimental data. It helps researchers identify potential bioactive compounds from high-throughput screening. By utilizing public databases and machine learning for virtual screening, AI can identify molecules from various compound libraries that may bind to target proteins, thereby accelerating the translation of computational predictions into viable clinical candidates [319].