Cancer is a systemic disease in which the tumor interacts with the entirety of the body to promote tumor pathogenesis and progression. While tremendous efforts have focused on understanding the local immune responses in the primary site because of the noticeable immunosuppressive characteristics in the tumor microenvironment (TME), the interaction between the tumor and the host's systemic response has not received much attention. These interactions influence primary tumor growth and treatment responses and can further facilitate tumor dissemination and condition distant tissues for metastatic colonization. In this review, we summarize the reciprocal interactions between the primary tumor and the systemic environment throughout cancer development, including pre-metastatic niche formation and metastatic progression, with particular emphasis on tumor-draining lymph nodes (TDLNs), bone marrow (BM), and blood circulation. By integrating these perspectives, this review aims to provide a more comprehensive understanding of tumor development within the whole-body context, thereby highlighting the development and optimization of strategies in cancer diagnosis, prognosis, and treatment that account for the systemic nature of the disease.

Breast cancer (BC) metastasis remains a leading cause of mortality despite significant advances in screening and diagnosis. Extracellular vesicles (EVs), secreted by all living cells, carry biologically active cargo such as RNAs and proteins, playing a crucial role in cell-to-cell communication. Distinct EV subpopulations contribute to the tumour microenvironment, pre-metastatic niche formation, immune evasion, extracellular matrix remodelling, and ultimately promote BC organotropism. Beyond their biological roles, growing evidence highlights EVs as potential diagnostic and prognostic biomarkers, as well as viable therapeutic targets. However, the full scope of EV function in BC metastasis remains unclear, and their inherent heterogeneity and molecular complexity present considerable barriers to clinical translation. Addressing these challenges requires both a clear understanding of the distinct functions of each EV subpopulation and the application of advanced single-EV analysis techniques. In this review, we comprehensively examine the functional contributions of EV subpopulations to BC metastasis. Each RNA and protein cargo we discuss is assigned a graded evidence level. We further consider EV lipids, non-vesicular extracellular particles and current EV-directed therapeutic strategies. Finally, we highlight that single-EV analysis is a promising strategy to resolve this complexity and convert EV biology into clinically actionable biomarkers and targets.

Homologous recombination deficiency (HRD) is a major genomic feature of several malignancies and provides a therapeutic rationale for DNA-damaging agents and PARP inhibitors. Beyond its role in genome instability, accumulating evidence suggests that HRD also reshapes the tumor microenvironment (TME). However, the immune consequences of HRD are not uniformly stimulatory. HRD-associated tumors may exhibit DNA damage-driven innate immune activation, increased interferon signaling, and enhanced antigen presentation, while simultaneously developing adaptive immune-evasive programs, including T-cell dysfunction, myeloid and regulatory T-cell recruitment, stromal remodeling, and vascular barriers to immune infiltration. In this review, we summarize current evidence linking HRD to immune activation and immune suppression across tumor types and propose a dual immune framework for understanding HRD-associated TME remodeling. We further discuss why HRD alone is insufficient to predict responses to immune checkpoint blockade and highlight emerging therapeutic strategies that integrate DNA damage response targeting with immune and stromal modulation.

Extracellular vesicles (EVs) are established mediators of long-range intercellular communication in cancer, transferring oncogenic and regulatory cargo that reshapes recipient cells across tissues. Recent integrative multi-omics analyses of small EVs (sEVs) from healthy human donors consistently identify TSAP6/STEAP3 among the most abundant and reproducible sEV components, providing a physiological context for its role in sEV biology. Rather than acting solely as a downstream effector of p53, TSAP6 functions within a broader regulatory network by associating with TPT1/TCTP, which engages DDX3 and promotes the incorporation of RNAs, including microRNAs, into sEVs. The resulting p53-TSAP6-TCTP-DDX3 axis constitutes a wide-reaching regulatory system with non-cell-autonomous impact on cellular communication. Notably, TCTP sustains sEV-mediated signalling even in p53-mutant contexts. We suggest that these findings are consistent with other advances in cancer linking oncogenic transformation with EV biology. Together, these observations support regulated EV-mediated information transfer as a mechanism contributing to cancer progression and systemic cellular reprogramming. Conversely, reduction in EV abundance, altered cargo content, or signalling capacity may contribute to the activation of the tumour reversion programme in TCTP-dependent experimental models.

Gastrointestinal stromal tumors (GISTs), the most common mesenchymal neoplasms of the gastrointestinal tract, are molecularly heterogeneous diseases rather than a single entity. Activating mutations in KIT and PDGFRA define the dominant biological dependencies and establish tyrosine kinase inhibition as a therapeutic backbone. Imatinib, followed by later-line tyrosine kinase inhibitors, including sunitinib, regorafenib, ripretinib, and genotype-selected avapritinib, have substantially prolonged survival. However, durable disease control remains limited by clonal evolution, most notably, polyclonal secondary KIT mutations and lesion-to-lesion molecular heterogeneity. Molecular profiling using serial tissue and circulating tumor DNA analyses may improve precision care by identifying spatially and temporally evolving resistance mechanisms and guiding mutation-matched treatment selection. In this review, we summarize the biological and molecular landscape of advanced GIST, as well as the current therapeutic strategies and major mechanisms of resistance. Finally, we highlight promising therapeutic strategies supported by preclinical and clinical evidence that may advance subtype-informed precision medicine.

Castration-resistant prostate cancer (CRPC) represents a lethal stage of prostate cancer that emerges following sustained androgen deprivation therapy (ADT) and remains a major clinical challenge in the era of androgen receptor (AR)-targeted therapies. Although AR reactivation and lineage plasticity are recognized as central mechanisms of resistance, these tumor-intrinsic processes alone cannot fully explain the profound heterogeneity and therapeutic failure observed during CRPC evolution. Here, integrating recent advances in single-cell and spatial profiling, we propose a unified framework in which CRPC is viewed as an adaptive tumor ecosystem shaped by evolutionary selection, cellular plasticity, and microenvironmental co-evolution under therapeutic pressure. ADT not only selects pre-existing resistant populations but also induces cellular reprogramming, expanding tumor state diversity and adaptive capacity. Concurrently, the tumor microenvironment is extensively remodeled to support and stabilize therapy-adapted states. Immune compartments evolve toward myeloid-dominant immunosuppressive configurations characterized by T cell dysfunction and exclusion, whereas cancer-associated fibroblasts (CAFs) promote tumor persistence through stromal remodeling, metabolic coupling, and alternative steroidogenic pathways. Neural remodeling represents an additional regulatory layer that enhances cellular plasticity, stress adaptation, and neuroendocrine differentiation. Together, these alterations establish a multicompartmental and self-reinforcing tumor-microenvironment network that sustains adaptation during AR suppression. This ecosystem framework further extends beyond the primary tumor to distal niches, including bone metastases and the gut microbiome, where tumor-osteoblast-osteoclast interactions and microbial regulation of androgen metabolism and immunity contribute to therapeutic adaptation. By integrating tumor evolution with immune, stromal, neural, and distal ecosystem remodeling, this perspective provides a system-level framework for understanding CRPC progression and highlights therapeutic strategies aimed at reprogramming or destabilizing adaptive tumor ecosystems rather than targeting individual resistance pathways alone.

Circulating cell-free DNA (cfDNA) has attracted increasing attention as a minimally invasive biomarker for assessing radiation-induced tissue injury. Radiation exposure can alter cfDNA concentration, fragmentation profiles, and methylation patterns. These changes may reflect cellular damage, tissue origin, and temporal responses to irradiation. During radiotherapy, however, cfDNA can arise from tumor cells, irradiated normal tissues, endothelial cells, immune cells, and hematopoietic cells. This mixed origin complicates the interpretation of radiation-related signals. Current evidence is derived mainly from preclinical studies and small, heterogeneous clinical cohorts. Sampling schedules, analytical workflows, and clinical thresholds also vary across studies. Recent advances in fragmentomics, tissue-specific methylation analysis, mitochondrial cfDNA, multi-omics integration, and artificial intelligence have expanded the information obtained from cfDNA. In particular, tissue-specific methylation and fragmentation features may help identify the cellular or organ origin of injury. Nevertheless, prospective multicenter validation and methodological standardization are required before routine clinical use. This review summarizes the mechanisms linking radiation injury to cfDNA release, current detection technologies, and evidence for organ-specific applications. It also discusses clinical confounders, translational barriers, and future directions for integrating cfDNA into precision radiation oncology.

Early gastric cancer is still commonly conceptualized as the endpoint of a linear sequence from chronic gastritis to intestinal metaplasia, dysplasia, and invasion. Yet recent single-cell, spatial, genomic, and functional studies indicate that this model incompletely captures the biology of early gastric carcinogenesis. Malignant potential is established progressively within a precancerous gastric field already shaped by somatic evolution, chronic inflammatory injury, and epithelial lineage distortion. Within this field, progression is concentrated in a restricted set of precursor states, particularly incomplete, hybrid, and stem-like metaplastic populations that display plasticity, persistence, and increasing compatibility with a supportive microenvironment. Fibroblast niche remodeling, immune protection loss, endothelial rewiring, genomic instability, epigenetic drift, and selective retention of advantageous molecular alterations further promote malignant commitment. In parallel, diffuse gastric cancer appears to follow a distinct route that may arise independently of conventional intestinal metaplasia through E-cadherin-deficient epithelial transformation and downstream chromatin reprogramming. Here, we synthesize recent evidence to propose an updated framework for early gastric carcinogenesis based on field evolution, lineage instability, ecosystem support, and pathway divergence. Rather than replacing the classical Correa cascade, this framework seeks to refine it by shifting the unit of risk assessment from histologic stage alone to biologically defined precursor states shaped by lineage instability, clonal persistence, niche permissiveness, and pathway-specific molecular constraints. This perspective shifts the emphasis of prevention from detecting smaller cancers to identifying and intercepting biologically committed precursor states before invasion occurs.

Langerhans cell histiocytosis (LCH) is a nonhereditary myeloid neoplasm driven by somatic alterations in the mitogen-activated protein kinase (MAPK) signaling pathway, most commonly involving BRAF and MAP2K1 (MEK1). Oncogenic BRAF alterations activate MAPK signaling through diverse structural mechanisms. The prominent BRAF(V600E) mutation traps BRAF in a constitutively active status by forming the K507-E600 salt bridge independent of upstream RAS signaling. In contrast, non-V600E BRAF mutants exhibit variable kinase activity, among which kinase-dead variants rely on BRAF/CRAF heterodimerization and transactivate CRAF. In parallel, oncogenic MEK1 mutations disrupt autoinhibition or enhance MEK1 homodimerization, sustaining MAPK activation. Furthermore, RAF/MEK heterodimerization modulates pathway output and persistence. These mechanistic insights provide a strong biological rationale for therapeutic targeting of the MAPK pathway in LCH, thereby driving the clinical use of BRAF inhibitors (BRAFi) including vemurafenib and dabrafenib, as well as MEK inhibitors (MEKi) such as trametinib and cobimetinib. However, the efficacy is frequently limited by acquired resistance associated with MAPK pathway reactivation through RAF and MEK dimerization-dependent mechanisms. Collectively, these altered molecular interactions contribute to the pathway dysregulation and signaling persistence, thereby providing a mechanistic basis for adaptive resistance and for developing next-generation RAFi and MEKi. This review therefore summarizes the current understanding of the molecular mechanisms underlying MAPK pathway activation and therapeutic resistance in LCH, with particular emphasis on conformational switch-based regulations of RAF and MEK proteins, and discusses emerging strategies involving next-generation RAFi and RAFi plus MEKi combinations aimed to achieve more durable MAPK pathway suppression and improving long-term outcomes of LCH treatment.

The seed and soil hypothesis has long explained organotropic metastasis through biochemical compatibility between tumor cells and distant tissues. However, accumulating evidence points to the role of mechanobiology in this process. This review proposes a framework in which cellular mechanical memory, the persistent adaptation to physical cues, functions as a mechanical compass that influences metastatic destination. We analyze how cancer cells, conditioned by the physical properties of the primary tumor (such as stiffness and viscosity), are epigenetically programmed to colonize distant organs with compatible mechanical signatures. We examine stiffness matching in bone metastasis, where cells conditioned by fibrotic tumors preferentially home to rigid skeletal niches, mediated by pathways including YAP/TAZ and RUNX2. We then address the paradox of soft tissue metastasis (brain, liver), proposing that a dual mechanical memory enables adaptation to compliant, viscoelastic environments. The role of dynamic forces, including fluid shear stress and cyclic strain, is examined in the context of lung colonization. By synthesizing these findings, we establish a comparative framework in which multidimensional mechanical memory, spanning static stiffness, viscoelasticity, and dynamic forces, contributes to organotropism as a complementary determinant alongside biochemical signaling. This framework generates testable predictions: that pharmacological disruption of specific mechanosensors or epigenetic erasure of stiffness memory should alter organ-specific metastatic patterns in preclinical models. Such a mechanobiological and biomechanical perspective may open therapeutic opportunities aimed at disrupting these physical memories to reduce metastatic dissemination.

Immune checkpoint inhibitor (ICI)-based combinations represent the standard first-line treatment for metastatic clear cell renal cell carcinoma (mRCC), although robust biomarkers for treatment selection remain undefined. We conducted a systematic review to identify biomarkers assessed in randomized clinical trials (RCTs) on first-line ICI-based regimens. Following PRISMA guidelines, we searched PubMed, Web of Science and Scopus (January 2018-October 2025) for phase III RCTs investigating first-line ICI-based therapies in mRCC with molecular or circulating biomarker analyses. Given heterogeneity across studies, a qualitative synthesis was performed. Sixteen reports were included. Biomarkers were assessed using immunohistochemistry, transcriptomics, genomic sequencing and blood analyses. PD-L1 expression did not reliably discriminate benefit across ICI-based combinations, although it revealed a negative prognostic influence with sunitinib and a potential predictive role for nivolumab-ipilimumab. Tumours with angiogenic signatures were consistently associated with improved outcomes, suggesting prognostic relevance, while derived limited additional benefit from ICIs. Immune-related signatures were associated with ICI response, whereas proliferation and MYC-related signatures identified disease with poor prognosis across treatments. Individual mutations (e.g. PBRM1, VHL, BAP1, PTEN) showed heterogeneous and mainly prognostic associations, whereas composite gene panels (e.g. rDM) may offer better predictive value. Circulating biomarkers, including inflammatory cytokines and KIM-1 dynamics, demonstrated promising prognostic and early predictive signals. No validated biomarker currently supports treatment selection among first-line ICI-based combinations in mRCC. Future research should prioritize adaptive, biomarker-guided trial designs integrating longitudinal multi-omic profiling and circulating biomarkers to generate clinical evidence and support personalized treatments.

Epithelial-mesenchymal transition (EMT), historically described as a binary morphological change from an epithelium to a mesenchymal state, no longer adequately accounts for carcinoma cell phenotypes observed in patients. Single-cell sequencing, spatial-omics, and lineage tracing converge on a different picture as carcinoma cells rarely undergo a unidirectional transition ending as sarcoma-like cells. Instead, carcinoma cells distribute across a continuum of intermediate states in which partial EMT often as hybrid epithelial/mesenchymal phenotypes couple adhesion and survival to motility, stress tolerance, and adaptive reprogramming. While EMT signatures remain a recurrent hallmark of metastasis, treatment failure, and disease relapse in epithelial cancers, functional cell plasticity is fundamentally shaped by distinct histotypes, tumor stages, malignancy grades, and local microenvironmental ecosystems. Carcinoma cells may not need to acquire a fixed position in the EMT spectrum to invade and metastasize, and "more mesenchymal" does not necessarily translate to "more dangerous." The clinical translation of EMT-directed interventions faces steep methodological hurdles, primarily driven by unverified causal dynamics and unresolved therapeutic safety profiles. Overcoming these hurdles will require dynamic, multi-pronged strategies tailored to specific carcinoma cell states and spatial niches. Effective regimens must simultaneously deploy oncogene inhibition, state-plasticity restriction, and microenvironmental, metabolic, or extracellular matrix-directed remodeling.

Microcalcifications are among the most reliable mammographic signs of breast cancer, particularly of non-palpable and in situ disease, yet for decades they were regarded as inert by-products of cellular degeneration-a view that delayed inquiry into their biology. A growing body of in vitro, ex vivo, and tissue-based evidence now reframes breast microcalcification as a process that can be actively regulated and cell-driven, recapitulating physiological and pathological mineralization elsewhere in the body. In this review, we synthesize the mechanistic basis of this paradigm shift and argue that mineral formation is not a passive marker but a participant in tumor biology. Two mineral species-calcium oxalate and calcium hydroxyapatite-carry different biological meaning and are separable on routine histopathology by birefringence and histochemistry. Hydroxyapatite can promote proliferation, matrix remodeling, and invasion, yet it is also the mineral of most benign calcifications, so its significance lies in composition and cellular origin, not presence alone. We discuss the osteomimetic program through which mammary tumor cells acquire bone-like mineralizing capacity, including the roles of alkaline phosphatase activity, ion transport, and PI3K-Akt signaling, and we distinguish osteomimicry-a phenotype of the tumor cell-from the bone metastatic niche that such cells may exploit. We then consider how mineral composition and microstructure encode the surrounding microenvironment and connect to clinical observations: association with HER2-positive disease, behavior under neoadjuvant chemotherapy, prognostic associations, and an emerging link with bone metastasis. We close with an integrated framework and a research agenda positioning mineral biology at the interface of the tumor microenvironment and metastatic progression.

Anaplastic lymphoma kinase (ALK) is a receptor tyrosine kinase with structural criteria shared among members of the insulin receptor subfamily. ALK plays crucial physiological roles such as in neural tissue development as well as in thinness and body weight regulation. In addition, ALK is a key driver of systemic inflammation that may lead to lethal sepsis. The oncogenic potential of ALK was first described in 1994 when 2 independent research groups discovered that ALK is aberrantly expressed in a rare aggressive subtype of T-cell lymphoma descriptively known as anaplastic large-cell lymphoma. Later, it was found that the aberrant expression of ALK extends to include other neoplasms such as non-small cell lung cancer, neuroblastoma, inflammatory myofibroblastic tumors, and others. In these neoplasms, ALK plays a central oncogenic role by upregulating the activity of a comprehensive network of upstream and downstream survival systems, which causes tumor cell survival and proliferation. The development of selective ALK inhibitors has led to paradigm shift in how ALK-expressing tumors, particularly non-small cell lung cancer, are treated. More recently, ALK has been implicated in disturbing the integrity of the immune system and sustaining immune evasion. In this review, we highlight 3 decades of ALK research, briefly discuss its physiological and pathological roles with a focus on cancer and immunity, and provide an update on the clinical utilization and mechanisms of resistance to ALK inhibitors.

Pancreatic ductal adenocarcinoma (PDAC) is shaped by a mechanically abnormal tumor microenvironment (TME) in which dysregulated mechanotransduction promotes malignant progression and therapeutic resistance. Two coupled but distinct pressure states dominate this landscape: solid stress and interstitial fluid pressure (IFP). Solid stress arises from constrained tumor growth, stromal contractility, and extracellular matrix remodeling. By contrast, IFP reflects hydrostatic pressure within the interstitial fluid compartment and is elevated by vascular leakage, impaired drainage, and low tissue hydraulic conductivity. Together, these abnormalities compress vessels, disrupt transport, and activate mechanotransduction programs that reinforce malignant adaptation. Here we integrate solid stress and IFP within a unified pressure-state framework for PDAC. We examine how these forces shape tumor progression, drug transport, and therapeutic response. We then evaluate spheroid, organoid, hydrogel, bioprinted, and microfluidic models according to what they truly control, directly measure, or merely infer. This distinction separates pressure-relevant systems from pressure-reconstructing models. We also discuss stromal normalization and the emerging role of artificial intelligence and machine learning (AI/ML) in model engineering and patient stratification. Current computational approaches can optimize mechanically defined models and infer pressure-related tumor states from multimodal data. However, they still rely largely on surrogates rather than direct measurements of solid stress or IFP. Our framework defines the biomechanical validation required to develop clinically predictive models of PDAC mechanobiology.