From Combination Early Detection to Multicancer Testing: Shifting Cancer Care toward Proactive Prevention and Interception

Abstract

Identifying the presence of tumors at a very early stage or deciphering the processes underlying their development can enable the interception of promalignant mechanisms underpinning cancer emergence, facilitating more effective prevention. Advances in molecular profiling allow the detection of genetic, epigenetic, immune, and microenvironmental alterations associated with cancer risk. Liquid biopsy permits noninvasive analysis of circulating tumor cells, nucleic acids, immune cells, extracellular vesicles, proteins, cytokines, and metabolites, whereas metagenome analysis facilitates gut microbiota profiling. Multicancer early detection assays broaden this approach, capturing signals from multiple malignancies using a single blood sample. These technologies go beyond genomics, addressing immune dysregulation and metabolic shifts, and may help identify gut microbiota imbalances. Clonal hematopoiesis of indeterminate potential is gaining increasing recognition as a biomarker. Cardiovascular risk scores based on multiple parameters are an inspiring example. The analysis of a combination of cancer drivers and enablers should provide a more sensitive and personalized measure of cancer prodromic profiles. Artificial intelligence will further support this transition by integrating molecular, immune, and metabolic data to develop individualized risk profiles. This shift from single-cancer detection to integrated, mechanism-based screening fosters a more proactive prevention model. This combination of early detection empowers cancer interception by using strategies, including lifestyle modification, nutritional optimization, drug repurposing, pharmacologic interventions, and targeted chemoprevention. Moving beyond single parameters analysis and organ-specific screening, this multidimensional approach advances early detection and interception as practical clinical goals, facilitating the fundamental aim of positioning prevention at the forefront of oncology.

Introduction

Strategies to reduce cancer incidence and mortality include early detection paired with timely access to high-value preventive or interceptive treatments. Traditionally, early detection has focused on identifying (pre)cancerous lesions or early-stage cancers arising from defined organs, framed within population-based or individual risk–informed screening interventions (1). However, a deeper understanding of carcinogenesis has shifted attention toward identifying early molecular and cellular alterations and intercepting key tumorigenic mechanisms before clinical lesions emerge (2). By targeting these cancer drivers and enablers, more effective prevention, interception, and intervention strategies can be developed (3). In cardiology, this concept is largely and historically established (4, 5). Multiple drivers of cardiovascular risk refer to the combination of factors, both biological and behavioral, that interact to significantly raise the chances of developing cardiovascular diseases. Some of these factors include hypertension; dyslipidemia, which presents as elevated low-density lipoprotein and high triglycerides, hyperglycemia, diabetes or insulin resistance, smoking, poor dietary habits, obesity, and inflammation. These drivers often act together, amplifying the overall risk.

Cancer interception is an emerging approach aiming to prevent malignancy by identifying and targeting genetic, epigenetic, immune, and microenvironmental mechanisms that drive cancer initiation and progression in asymptomatic populations through lifestyle, pharmacologic, and biologic interventions (2, 6–10). Interception (Fig. 1) bridges early detection and secondary prevention by targeting interconnected genetic and nongenetic cancer drivers and enablers, such as inflammation, immune imbalance, and microbiota alterations, to prevent malignant progression. Integrated, multimodal strategies, supported by artificial intelligence (AI)–driven analyses, may empower timely, sustainable interventions and improve both individual outcomes and public health (Fig. 1; refs. 11–20). This review examines the evolving paradigm from circulating DNA analyses and liquid biopsy technologies to multicancer early detection (MCED) strategies, proposing a combined approach that targets key oncogenic drivers and enablers and their interconnected biological pathways to refine individual cancer risk assessment and allow timely intervention. By addressing genetic, epigenetic, metabolic, immune, and microbiota-related alterations implicated in tumorigenesis and treatment resistance, we explore how integrated early detection and interception strategies, potentially augmented by AI-based analytics, could preempt malignant progression and improve outcomes at both the individual and population levels (19–21).

Figure 1.

Pathways for cancer prevention, which include the interception of cancer drivers, allowing for the stopping of cancer development and supporting early diagnosis and early treatment.

Identifying and Tackling the Mechanisms of Cancer through Early Detection and Interception

The notion of cancer interception is based on the identification of carcinogenesis drivers and enablers, which form the basis of the pathologic mechanisms underlying cancer development (22). Detecting carcinogenesis biomarkers can serve for cancer risk assessment and early diagnosis but also delineate the level of risk of cancer in the context of interception. Cancer is, per se, a complex and multimodal disease in which onset is ascribed to numerous mechanisms. In the classic view, cancer is a genetic disease resulting from the accumulation of genetic mutations and epigenetic modulations. These gene alterations can be identified in circulating tumor DNA (ctDNA) and circulating tumor cells (CTC) in the bloodstream, helping detect cancers early and monitor their progression. It is possible to analyze DNA fragments to detect mutations linked to specific cancers, such as BRCA1/2 for breast cancer or TP53 for various tumors, and several other oncogenes and oncosuppressors (23). Additionally, clonal hematopoiesis of indeterminate potential (CHIP) is frequently observed in patients with solid tumors, in which it is linked to prior cytotoxic therapy and has a negative prognostic role (24, 25). Furthermore, it is known that alterations of the host microenvironment, macroenvironment, and bacterial flora are potent cancer-promoting agents. Over a lifespan, normal tissue cells accumulate genetic and epigenetic changes that lead to increasing genetic instability (26, 27) and mutations in critical genes, often due to impairments in the DNA damage response (28, 29). The immune system plays a crucial role in eliminating mutated cells, including potential cancer cells, through various defense mechanisms, such as DNA repair machinery, immune surveillance, and cytotoxic immune responses. However, some tumor cells evade immune detection by exploiting immune checkpoints, such as PD-L1 (30). The polarization of immune cells, including neutrophils, NK cells, macrophages, and T cells, can shift their role from protective barriers to protumorigenic agents (2). Thus, although genetic instability may act as a marker of hereditary or somatic cancer risk, abnormalities in the immune system may impair the clearance of mutated cells, increasing the likelihood that tumor cells evade immune control systems (31). From the host’s side, chronic inflammation and oxidative stress, which are significantly influenced by resident microorganisms (32, 33), may amplify genetic damage while reducing immune activity (34). Specific bacterial strains are emerging as involved in the development of malignancies, whereas certain microbiome compositions seem to be protective (bioRxiv 2024 2024.08.21.608814; refs. 35–37). Collectively, several factors can contribute to the development of cancers and must be addressed and potentially tackled through a comprehensive, combined approach. Capturing the ecosystem underlying carcinogenesis, along with DNA alterations, for interception and early diagnosis requires integrated diagnostic approaches that utilize a single sample.

CHIP as a marker and driver of cancer

An emerging marker and driver in cancer is CHIP. Hematologic malignancies are a key area of interest for early cancer interception, as they border an expanding list of “premalignant” conditions, such as (i) monoclonal gammopathy of undetermined significance, potentially evolving into multiple myeloma; (ii) monoclonal B-cell lymphocytosis of undetermined significance, possibly evolving into chronic lymphocytic leukemia; and (iii) the newly defined entities of CHIP and CHIP-related clonal cytopenia of undetermined significance (CCUS). CHIP, in particular, is a condition characterized by the presence of somatic mutations in genes associated with myeloid hematologic malignancies at a variant allele fraction of at least 0.02 in the absence of other diagnostic criteria for hematologic malignancy or unexplained cytopenia (38). Approximately 80% of patients with CHIP harbor mutations in epigenetic regulators, such as DNMT3A, TET2, and ASXL1; DNA damage repair genes, such as PPM1D and TP53; the regulatory tyrosine kinase JAK2; or messenger RNA spliceosome components, including SF3B1 and SRSF2 (39–41). CHIP is associated with an increased risk of hematologic malignancy and cardiovascular disease, as well as reduced overall survival (40–44). Research on CHIP may shed light on the mechanisms leading to the accumulation of somatic mutations, epigenetic rewiring, and clonal evolution in the hematopoietic system. Such studies may enhance understanding of how inflammation and aging affect cellular and tissue homeostasis (44, 45).

Inflammatory cytokines as cancer enablers

Inflammation plays a critical role in tumorigenesis, as observed by Virchow’s discovery of inflammatory cells within tumors (46–48). These cells contribute to tumor proliferation and progression in the tumor microenvironment (49–51). Chronic inflammation, often linked to aging, increases cancer risk (52), whereas genetic mutations can trigger cancer-related inflammation (53). Inflammatory mediators promote tumor growth in gastric, liver, and colon cancers (54–56), with cytokines such as TNF-α, IL-1β, IL-6, IL-8, and TGF-β being upregulated in many cancers (57, 58). TNF-α regulates cell death, proliferation, and immune responses through pathways such as MAPK, Akt, and NF-κB, and its interaction with TGF-β enhances metastasis (59, 60). IL-31 influences cancer progression and is linked to pruritus in patients (61, 62), whereas IL-17 and IL-23 are associated with poor outcomes in colorectal cancer (63). IL-15 supports T-cell and NK cell proliferation, mediating exercise-related benefits in cancer (64). IL-6 is often elevated in tumors, promoting cancer cell survival, proliferation, and angiogenesis (65, 66). Finally, IL-23 receptors on regulatory T cells (Treg) modulate tumor size in colon cancer models (67), whereas increased IL-17A levels in pancreatic cancer correlate with poor survival (68). Conversely, IL-10 helps regulate inflammation and maintain cell homeostasis (69).

Immune cell imbalance and polarization

Key components of immunity, such as NK cells, neutrophils, macrophages, B cells, and T cells, undergo functional changes within the tumor microenvironment, influencing tumor growth, immune evasion, and metastatic spread (70, 71). The presence of NK cells in colorectal cancer correlates with better prognosis and increased survival (72). However, in various cancers, NK cells are altered, allowing tumor cells to evade immune surveillance by releasing ligands, proteases, or extracellular vesicles (EVs; ref. 73). In non–small cell lung cancer, tumor-infiltrating NK cells adopt proangiogenic phenotypes (74), whereas in prostate cancer, they release cytokines that support monocyte recruitment and inflammatory angiogenesis (75). NK cells from healthy donors exposed to TGF-β can acquire proangiogenic phenotypes, similar to those seen in cancer (76–78). Neutrophils are crucial for host defense and acute inflammation but also contribute to tumorigenesis (79, 80). In the tumor microenvironment, they can display both antitumor and tumor-promoting properties (81). Tumor-associated neutrophils often differ phenotypically from their nontumor counterparts, with C-X-C chemokine receptors modulating their recruitment and function in cancer (82, 83). They are influenced by factors such as cholesterol and interactions with T cells, Tregs, and dendritic cells, linking them to the cancer risk associated with hypercholesterolemia (84). Tumor-associated macrophages (TAMs) coevolve with tumor cells, affecting T-cell immunosurveillance and supporting tumor growth (58, 85). TAMs display an M2 phenotype and are linked to immune evasion and metastasis, whereas M1 macrophages are associated with antitumor responses (86–88). Reprogramming TAMs toward the M1 phenotype is under investigation as a therapeutic strategy (89, 90). B cells also participate in the tumor microenvironment, displaying pro- or antitumor roles (91). In hepatocellular carcinoma, they aid immune evasion (92), and their somatic hypermutation contributes to malignant transformation (93). B cells and tumor-associated lymphoid structures correlate with immunotherapy outcomes across cancer types (94). In pancreatic adenocarcinoma, for instance, B cells present tumor antigens to T cells, producing antibodies and growth factors that may promote immune escape (95). In cancer immune evasion, T cells often become dysfunctional due to tumor-induced mechanisms such as immune checkpoint activation, suppressive cytokines, and inhibitory cells in the tumor microenvironment, leading to reduced tumor recognition and impaired antitumor activity (96).

Gut microbiome imbalance

Recent research has highlighted the role of microbiota in inflammation, immune function, and cancer risk, emphasizing microbiota-based immune regulation as a potential avenue for cancer prevention (97–99). The microbial composition evolves with tumor development, influencing carcinogenesis, disease progression, and therapy response (100). Gut microbes interact with the host immune system, modulating immunity and antitumor responses, whereas the immune system, in turn, shapes microbiota composition (101). These interactions are mediated by pathogen-associated molecular patterns recognized by pattern recognition receptors, such as Toll-like receptors, on intestinal epithelial cells. This recognition activates antigen-presenting cells, which migrate to lymph nodes and trigger B- and T-cell differentiation, including CD4⁺ T cells (101). Gut microbes also regulate systemic immune responses through immunomodulatory cytokines and influence B-cell maturation within gut-associated lymphoid tissue (101). Microbial metabolites, such as short-chain fatty acids and polyamines, further modulate immune responses and inflammatory pathways (102). Advances in microbiome research have enabled the identification of microbial strains as biomarkers for cancer risk (103). Dysbiosis is linked to increased colorectal cancer risk through the production of toxic metabolites and proinflammatory molecules (104, 105). Microbiota composition also affects immunotherapy and chemotherapy responses, with certain bacterial strains predicting therapeutic outcomes (106–108). For instance, Lactobacillus fermentum and Streptococcus anginosus are enriched in patients with hepatocellular carcinoma, whereas Bacteroides stercoris is associated with therapy resistance (109). Conversely, Bifidobacterium, Firmicutes, and Saccharomyces promote eubiosis and support immunomodulation (110). Akkermansia muciniphila has been proposed as a biomarker for checkpoint inhibitor response in non–small cell lung cancer (35). Table 1 summarizes the role of selected microbial strains in cancer.

Table 1.

Role of gut microbiota species in cancer risk.

Bacteria	Activity	Linked cancers
Components of gut microbiota with protective effects against cancer
A. muciniphila	Maintains the gut barrier, systemic anti-inflammatory effects, and immunomodulation	Colorectal Breast
Firmicutes, Lactobacillus	Enhances immune response	Endometrial
Firmicute Roseburia intestinalis	Produces short-chain fatty acids	Endometrial
Bifidobacterium	Enhances immune response, anti-inflammatory activity, and production of short-chain fatty acids	Colorectal Breast Prostate
Bacteroides, Faecalibacterium	Produce butyrate	Colorectal
Firmicutes, Ruminococcus	Enhance anti-inflammatory activity, short-chain fatty acid production, and immunomodulation	Colorectal Endometrial
Components of gut microbiota with tumor-promoting activity
Firmicutes streptococcus	Promotes inflammation and protumorigenic metabolites	Colorectal
F. nucleatum	Promotes inflammation	Colorectal Breast
Bacteroides fragilis	Promotes inflammation and immunosuppression	Colorectal
Escherichia coli	Promotes inflammation	Endometrial Bladder
Shigella	Promotes inflammation	Endometrial
Enterococcus faecalis	Promotes inflammation	Colorectal Pancreatic
Erysipelotrichaceae	Promotes inflammation	Breast

Additionally, vaginal microbiota dysbiosis has been linked to cervical cancer, especially in human papillomavirus (HPV)–associated cases. Disruptions in the vaginal microbiota reduce hydrogen peroxide and lactic acid levels while increasing proinflammatory cytokines, raising the risk of persistent HPV infection and cervical cancer (111).

Expanding the Horizons of Liquid Biopsy in Cancer toward Interception and Screening

The most promising and advanced technique in the context of interception and screening is liquid biopsy, which tests bodily fluid, especially blood, to detect the presence of cancer cells. Liquid biopsy is minimally or noninvasive and is used to detect and monitor cancer-related products that are released into the bloodstream (112–114). Liquid biopsy can serve to evaluate a spectrum of blood-derived biomarkers (115), including CTCs (116); ctDNA, which is one of the main transformative universal cancer biomarkers discovered to date and is shed by transformed cells (as well as normal cells; refs. 117, 118); cell-free DNA (cfDNA; ref. 119); immune cells; EVs (which contain DNA fragments); and cancer, immune, and microbial metabolites (120). Next-generation sequencing (NGS)–based panel testing allows for the simultaneous interrogation of multiple molecular alterations, enabling the comprehensive identification of cancer gene alterations (121, 122). Of note, recent liquid biopsy opportunities have emerged to include profiling glycoproteins, which are produced by posttranslational modifications of proteins. Advances in chemical and mass spectrometry–based technologies have facilitated the study of changes in glycans and glycoproteins through glycomics and glycoproteomics. These analyses are promising tools for biomarker identification and may further expand the opportunities for liquid biopsy to profile glycoproteins, which could help make personalized therapeutic decisions (123). Additionally, liquid biopsies can help identify cancer-related materials from diverse sources: Exosomes and nano-sized vesicles secreted by cells play a major role in cell-to-cell communication and contribute to activating or inhibiting cellular and molecular pathways. Tissue- or disease-specific exosomal contents, including nucleic acids, proteins, and lipids, have been identified and could be assessed as novel biomarkers for the early detection of cancers (114, 124).

Overall, liquid biopsy holds significant promise and is widely implemented although it still faces some technical challenges. Blood is the preferred medium for biomarker analysis due to its richness in ctDNA, CTCs, and miRNAs though it often carries a high background of nontumor DNA. Recovery rates vary and can be improved with techniques such as centrifugation and magnetic separation (125). Broader implementation in screening is hindered by the low abundance of cfDNA in early disease and a limited understanding of mutation significance in healthy individuals (126). Preanalytic variables, such as sample handling and EV enrichment, further complicate reliability (127). Besides this, effective population screening requires low-cost and real-time ctDNA measurements (21).

Although the most common sample used in clinical practice is blood, other fluids may serve to uncover cancers. Urine sampling is used to detect urological cancer biomarkers. It offers the advantage of being minimally invasive; however, biomarker dilution and enzymatic degradation necessitate concentration and preservation techniques for accurate analysis (128). Saliva is gaining importance for head and neck cancer detection due to innovative microfluidics and PCR-based methods, which enable the differentiation of tumor-derived biomarkers from microbial content (129). To a certain extent, urine and saliva may overcome the limitations of blood samples for tumors arising in districts where blood shedding of cfDNA is limited or when tumors are in earlier stages of development. As such, only a small proportion of cfDNA is indeed detectable. Accordingly, a study showed the potential of saliva miRNA analysis as a noninvasive method to determine the risk of progression of oral premalignant lesions (130). In addition, exhaled breath analysis is under development for cancer detection, particularly lung cancer, and a study showed the predictive value of volatile organic compounds in human breath as biomarkers for breast cancer risk (131).

Multimomic integrated approaches

Combining the analysis of several parameters could be a powerful strategy to increase the predictive value of tests by addressing key hallmarks of cancer (114). Multiomics approaches that integrate genomics, transcriptomics, proteomics, and metabolomics are able to enhance the detection of biomarkers in cancer. The combination of techniques, such as ELISA, mass spectrometry, multiplex assays, and NGS, offers deeper insights into molecular networks, facilitating biomarker discovery and personalized treatment and interception strategies (132). Recent studies on integrated multimodal diagnostic approaches in the space of lung and colorectal cancer screening are paradigmatic. A prospective study showed that a combined model of clinical, imaging, and cfDNA methylation biomarkers had better discrimination capacity for pulmonary nodule classification than models based solely on imaging, clinical biomarkers, and tumor DNA methylation alone (133). Additionally, a multimodal ctDNA-based blood assay that integrates genomics, epigenomics, fragmentomics, and proteomics increased the effectiveness of screening for the early diagnosis of colorectal cancer in patients with altered fecal immunochemical tests and for detecting advanced precancerous lesions (134). The multiomics approach has been used in the study by Wang and colleagues, combining proteome- and transcriptome-wide association studies to identify plasma proteins linked to breast cancer risk. This multiomics approach enhances the ability to detect cancer as proteomic and transcriptomic associations strengthen the role of potential biomarkers in cancer screening and as therapeutic targets. In this way, five plasma proteins (PEX14, CTSF, SNUPN, CSK, and PARK7) were found to have strong causal links to breast cancer (135).

Liquid biopsy, using blood or saliva samples, is also used to detect CHIP, and recently, technologies are being developed to discriminate CHIP from other germinal or somatic tumor molecular signatures in circulating DNA using NGS (40–44, 136, 137). Advanced bioinformatics platforms for CHIP filtering are being validated to ensure effective CHIP discrimination (138), including AI-enhanced methods. These methods help differentiate actual pathogenic mutations contributing to CHIP from sequencing artifacts or benign genetic variations that could otherwise be misclassified as CHIP-related mutations. Publicly available CHIP datasets are provided for exome and whole-genome sequencing, including a list of common false positives (139). It has also been shown that prognostication for CHIP and CCUS will require integrating clinical factors (e.g., cytopenias) and molecular features (e.g., clone size, mutation pathogenicity) in a combined strategy. Besides this, leveraging large datasets with paired genomic and demographic information improves CHIP identification for clinical and research applications (140, 141). Meanwhile, tools for malignancy risk stratification based on CHIP or CCUS detection are being progressively implemented (140, 141).

Although we may be used to perceiving liquid biopsy as informing on cancer biology alone, it is noteworthy that such a methodology can retrieve more data on the broader cancer milieu (142). For instance, the Vogelstein group recently discovered that in a cohort of 178 patients with cancer, 83% had cfDNA concentrations significantly higher than those observed in healthy subjects. However, the excess cfDNA was found to be originating from leukocytes rather than from the tumor per se, confirming that cancer may exert a systemic impact on cell turnover or DNA clearance that can be detected with liquid biopsy (143). This study exemplifies the potential of including integrated assessments when testing approaches aimed at sensing early traces of cancer. It showcases that such sensing may not be strictly related to cancer components but to the broader ecosystem in which carcinogenesis is rooted and how it is affected.

Liquid biopsy–based detection of inflammation and immune activation hallmarks to enable early cancer detection and interception

Beyond its traditional role in identifying ctDNA, CTCs, and EVs, liquid biopsy is increasingly recognized for its potential to detect inflammatory cytokines and immune cells, which play a critical role in tumor initiation and progression (144).

Although highly advanced tests for detecting early alterations in DNA and RNA are already available or under development to improve risk assessment and early cancer detection, the evaluation of inflammatory mediators in “healthy” individuals or at-risk populations, such as those with metabolic syndrome or obesity, has not been established as a standard approach for cancer prediction.

The establishment of a reliable panel of cytokines for ELISA or other antibody-based assays correlated to risk conditions could be of invaluable help. On the path to combination early detection, in this study, we delve into reviewing some cytokines that have been associated with tumors and tumor progression and that might be useful in cancer early detection, screening, interception, and treatment (Table 2; refs. 59–67, 69, 145–147). Many of these have been detected in tumor tissue, but the promising avenues are blood-based tests, such as those that can employ liquid biopsy.

Table 2.

Cytokines involved in cancer development and progression, which are of potential utility in cancer early detection, screening, interception, and treatment.

Cytokine	Function	Role in cancer
TNF-α	Regulates cell death and proliferation and activates immune responses via MAPK, Akt, and NF-κB pathways	Plays a dual role: promotes tumor survival and induces apoptosis (59)
IL-31	Induces inflammation	Plays a dual role: contributing to either the promotion or suppression of cancer (61) Induces pruritus associated with cancer (61) Elevated levels of endometrial carcinoma (62)
IL-1β	Acts as a key mediator of inflammation and immune responses	Induces inflammation-related carcinogenesis, particularly in lung cancer (145)
IL-6	Contributes to host immune defense and acute stress response	Elevated in patients with cancer (65) Supports the malignant phenotype (66)
IL-8 (chemokine)	Involves in neutrophil recruitment	Elevated levels linked to lung cancer risk and tumor progression (146)
IL-10 (anti-inflammatory cytokine)	Modulates immune homeostasis	Contributes to immune suppression in tumors (69)
TGF-β	Regulates cell proliferation, differentiation, and immune responses	Promotes tumor progression by inducing extracellular matrix remodeling and metastasis (60)
IL-17	Induces inflammation	Associated with poor prognosis in colorectal cancer (63)
IL-23	Induces inflammation	Elevated in colorectal cancer, influences Treg-mediated immune suppression (67)
IL-15	Stimulates T-cell and NK cell proliferation	Contributes to cancer immunity benefits (64)
MCP-1	Recruits monocytes and macrophages	Promotes tumor-associated inflammation and immune suppression (147)

Proinflammatory cytokines have been implicated as promoters of carcinogenesis in individuals exposed to environmental particulate matter measuring ≤2.5 μm, which promotes lung cancer by acting on cells that harbor preexisting oncogenic mutations in healthy lung tissue. Notably, IL-1β has been identified as a key mediator of lung inflammation induced by particulate matter measuring ≤2.5 μm, promoting a progenitor-like cell state in EGFR-mutant lung epithelial cells and driving pulmonary tumorigenesis (145). Liquid biopsy techniques enable the identification and quantification of specific inflammatory mediators associated with cancer progression, offering opportunities for tailored anti-inflammatory strategies aimed at high-risk conditions (148). Evaluating IL-8 and IL-6 levels is particularly useful for lung cancer interception, with IL-8 levels linked to cancer risk years before diagnosis and IL-6 elevations indicating immediate risk or disease presence (146). In gastric cancer, the combination of cytokines (TNF-α, IL-6, IL-8) with carcinoembryonic antigen and carbohydrate antigen 72-4 enhances screening accuracy through a multiparametric approach (149). For ovarian cancer, simultaneous measurement of IL-6, IL-8, MCP-1, VEGF, EGF, and CA-125 improves diagnostic precision, enabling better differentiation between benign and malignant conditions (147). A emerging marker is the neutrophil-to-lymphocyte ratio, which serves as a simple, cost-effective marker of systemic inflammation and immune dysregulation in cancer (150). Another interesting strategy has been recently pioneered by Dyikanov and colleagues (151). Using a multiplatform phenotypic and transcriptional approach, they validated a flow cytometric tool capable of identifying five immunotypes, each characterized by unique distributions of cell types and gene expression profiles in healthy individuals and patients with cancer. These immunotype signature scores correlated prognostically and predictively with systemic immunity and patient responses to various cancer treatments, including immunotherapy.

Cancer per se is associated with a remodeling of the immune system architecture and changes in immune function, resulting in a reduced immune response that facilitates tumor cell escape. T cells play a relevant role in antitumor immunity with a crucial contribution from T-cell receptors, which are produced in a vast diversity of repertoires. The analysis of T-cell receptor repertoires can be used to monitor the emergence and progression of cancers (152). T-cell exhaustion is one of the major causes leading to immune escape in cancer, creating an environment that supports tumor development and metastatic spread. In this state, T cells are impaired due to prolonged stimulation, followed by increased expression of immune inhibitory receptors, altered epigenetics, and transcriptional programs (96). It has been shown that dynamic changes in T-cell exhaustion occur during the progression of oral squamous cell carcinoma (OSCC). Using a murine OSCC model, the study highlights how T-cell exhaustion emerges early in premalignant lesions and progressively worsens in malignant stages. T cells in dysplastic lesions show early signs of exhaustion, including upregulation of inhibitory receptors (PD-1, TIM-3, LAG-3). Despite the presence of antigen-specific T cells, they fail to eliminate dysplastic epithelial cells. A shift toward immune suppression is observed, characterized by an increase in Tregs, a decline in Th17 cells, and progressively reduced T-cell function with decreased cytokine secretion (IL-2, TNF-α, IFN-γ). In malignant lesions, T-cell exhaustion is more pronounced, with significantly higher expression of PD-1, and tumor-associated immune suppression is established, likely contributing to immune evasion. The proportion of exhausted CD4⁺ T cells increases earlier and more markedly than that of CD8⁺ T cells, suggesting that CD4⁺ T cell exhaustion may play a key role in early immune dysfunction. Administering anti–PD-1 antibodies in the premalignant phase significantly reduced progression to OSCC by restoring T-cell function, enhancing cytokine production, and reducing exhaustion markers. These findings support the potential prophylactic use of immune checkpoint inhibitors for cancer prevention in patients with high-risk oral premalignant lesions. Current phase II clinical trials are exploring this strategy to prevent malignant transformation (153). Research on tumorigenesis also focuses on the expression and function of MHC class I (MHC I) and MHC II receptors. Tumor cells downregulate MHC I, thereby escaping recognition by cytotoxic CD8⁺ T lymphocytes (154). Immunotherapy with checkpoint inhibitors is now the gold standard for metastatic melanoma (155) and is widely used for lung cancer (156). Tregs, a subset of CD4⁺ immunosuppressive T cells, are believed to promote cancer development. They are recruited into the tumor microenvironment by chemokines and metabolic factors, typically impairing the antitumor immune response. However, their presence has been observed to improve survival rates in some cases. Their function varies depending on tumor type, location, stage, and inflammation (67).

Immune senescence

Capturing the immune status may inform the individual risk of cancer, whereas developing cancer immune profiles under certain risk factor conditions may help dissect the individual risk of cancer and orient risk-reducing strategies. Aging contributes to changes in the phenotype and activity of macrophages in the continuum of immune dysfunction, age, and cancer risk. Senescent macrophages, as well as senescent fibroblasts, interact with other cells in the tumor microenvironment and secrete senescence-associated secretory phenotype factors. These molecules may promote tumor cell proliferation, invasion, and metastasis or, conversely, exert antitumor effects (157). Immune senescence can also be involved in cancer and is influenced by nutrition. Interventions, such as the intake of bioactive nutrients and supplements (e.g., vitamin D), may regulate the immune response and inflammatory status, possibly delaying tissue aging and reducing cancer risk (158–159). Additional options for modulating the immune system include natural products (160). Given the relevant and diverse role of immune cells within the tumor microenvironment, evaluating immune status is an essential component of cancer risk assessment and should be integrated with genetic testing and inflammation analysis. For example, it has been shown that higher circulating B-cell proportions are associated with an increased risk of breast cancer, particularly in premenopausal women and in diagnoses occurring four or more years after blood collection. In contrast, higher monocyte proportions are linked to a reduced near-term risk, especially within 1 year of blood collection (161). The epigenetic approach utilizes immune cell–specific methylation markers to profile immune composition from stored DNA. One such method, the immunoCRIT assay, quantifies the ratio of FOXP3+ Tregs to CD3⁺ T cells using qPCR-based analysis of cell type-specific demethylated loci. A study showed that a higher immunocrit level could represent an independent risk factor for lung, colorectal, and estrogen receptor-negative breast cancer. These findings provided strong prospective evidence supporting the role of elevated peripheral immune tolerance in cancer development (162). The complementary value of cytokine analysis alongside ctDNA within a multianalyte liquid biopsy framework has been assessed in the context of lung cancer, proving the potential for assessing disease progression (163). A review by Cui and colleagues (164) discussed how immune cell populations, particularly Tregs and myeloid-derived suppressor cells (MDSC), evolve during the transition from premalignant colorectal adenomas to colorectal cancer. These findings suggest that an immunosuppressive microenvironment is progressively established, facilitating immune evasion and tumor progression. Previous studies have shown that FoxP3+ Tregs and their expression of immunosuppressive cytokines IL-10 and TGF-β are significantly increased in colorectal adenomas compared with normal tissues, with an even greater increase in colorectal cancer tissues, suggesting an early role in dampening antitumor immunity. In early adenomas, DCs exhibit increased expression of IL-12, which supports antitumor immunity, whereas in colorectal cancer, IL-12 expression declines, indicating a shift toward an immunosuppressive phenotype. The expansion of MDSCs begins in premalignant lesions and persists in malignancy. CTL function becomes progressively inhibited as Tregs and MDSCs accumulate, limiting immune-mediated tumor suppression. Given that Tregs and MDSCs contribute to immune suppression from the adenoma stage, targeting these cells could be a promising preventive strategy. Additionally, anti–PD-1 and anti–PD-L1 therapies could restore CD8⁺ T-cell function in high-risk adenomas, potentially delaying or preventing colorectal cancer progression (164). Besides this, the analysis of epigenetic features of immune cell–derived cfDNA can shed light on immune cell turnover dynamics in healthy people and inform the study and diagnosis of cancer, local inflammation, infectious or autoimmune diseases, as well as responses to vaccination.

Hence, in the future, integrating the analysis of inflammatory cytokines, immune cells, and other tumor markers into liquid biopsy–based screening and prevention strategies holds great potential for advancing precision oncology by enabling interception or earlier, more accurate cancer detection and personalized risk stratification. Furthermore, real-time monitoring of these markers may support treatment response assessment and disease progression tracking.

Cancer Interception Strategies: From Vaccination to Microbiome Modulation

Cancer interception encompasses proactive strategies aimed at halting tumorigenesis at its earliest stages. Vaccination against oncogenic viruses and tumor-specific antigens has demonstrated the feasibility of priming durable immune responses before malignant progression. Complementing this, immune modulation through anti-inflammatory agents, such as aspirin and NSAIDs and immune checkpoint inhibitors targets key tumor-promoting pathways. Also, interventions aimed at restoring microbial balance and reducing cancer risk seem promising. Together, these approaches offer a multifaceted framework for intercepting cancer before clinical onset.

Cancer interception via vaccination aims to halt tumorigenesis at premalignant or early stages by priming the immune system against tumor-associated or mutation-derived antigens. Approved vaccines for oncogenic viruses, such as HPV and hepatitis B virus, demonstrate the power of immunoprevention. Vaccines targeting antigens, such as MUC1 or KRAS mutations, have shown safety and immunogenicity in early trials. In high-risk individuals, such as those with Lynch syndrome or BRCA mutations, tailored vaccines targeting neoantigens or telomerase (hTERT) are being evaluated. These efforts reflect a shift from treatment to prevention, leveraging immune memory to intercept cancer before it becomes invasive (165). Inflammation plays a central role in carcinogenesis, and its modulation offers promising avenues for cancer interception. Anti-inflammatory agents, particularly aspirin and NSAIDs, have shown protective effects against several cancers, including colorectal, esophageal, and gastric cancers. Aspirin, in particular, has demonstrated delayed but significant mortality reduction in colorectal cancer after long-term use, suggesting efficacy in early-stage disease interception. Other agents such as iloprost, green tea catechins, and metformin are being evaluated in phase II trials for their ability to modulate inflammatory pathways and prevent progression from precancerous lesions to invasive tumors (166).

Cancer immune interception is an emerging strategy that aims to prevent invasive cancers by targeting premalignant lesions before immune evasion fully develops. This approach makes use of immune-based interventions, such as prophylactic vaccines, immune checkpoint inhibitors, and immunomodulators, to eliminate early neoplastic changes. It is particularly effective in high-risk individuals such as those with hereditary syndromes or identifiable premalignancies. By acting during the window when antitumor immunity remains functional, immune interception offers a proactive approach to reducing cancer burden through early immune engagement and durable immune memory (165).

Emerging evidence highlights the role of gut microbiota in colorectal cancer interception, beyond early detection. Microbiome profiling has demonstrated potential in identifying precancerous lesions and improving noninvasive diagnostic tools, complementing traditional methods such as fecal occult blood testing (163). Targeted microbial markers, such as Fusobacterium nucleatum, can signal risk even when fecal occult blood testing fails, and their modulation through probiotics, antibiotics, and anti-inflammatory supplementation, such as berberine, modulates tumor risk (167). Also, strategies using prebiotics and postbiotics restore gut balance and inhibit tumor-promoting inflammation (168). Postbiotics, nonviable microbial products, including microbial metabolites such as short-chain fatty acids, exert direct anti-inflammatory and anticancer properties by inducing apoptosis and improving intestinal barrier function. Together, these interventions restore microbial balance, enhance chemotherapy and immunotherapy responses, and mitigate treatment side effects. Microbiota-based approaches, including fecal microbiota transplantation and bacteriophage therapy, show promise in improving colorectal cancer outcomes, offering safe, personalized alternatives to conventional therapies. Such microbiota-based strategies hold promise as complementary approaches for colorectal cancer prevention and interception, especially when integrated with conventional therapies (168). In pancreatic cancer, microbiome-derived metabolites such as TMAO and 3-IAA support immune modulation and treatment response, offering interception opportunities through personalized microbiota-targeted interventions (169). Diet and fecal microbiota transplantation also show promise in reducing tumorigenic pathogen colonization and mitigating cancer risk (170, 171).

With a refined understanding of the role of microbes in tumorigenesis, cancer interception strategies will increasingly focus on effective diagnostic and therapeutic options, mirroring past successes, such as those targeting HPV-related cervical cancer with a preventive vaccine. Moreover, the combination of microbiome modulation and immune-based interventions would offer additional opportunities to halt or even reverse premalignant changes.

Building an Oncological Risk Score: Translating Combination Detection from Cardiology

Cardiovascular diseases, together with cancer, remain a major cause of morbidity and mortality worldwide. The European Society of Cardiology guidelines encourage the use of risk prediction models to enhance the effective management of cardiovascular risk factors and the promotion of healthy behaviors. The European Society of Cardiology has developed and regularly updates the Systematic Coronary Risk Evaluation risk estimation system. In addition to conventional risk factors, emerging ones should be considered and incorporated into a tailored approach to risk stratification (4, 5). In our view, the same approach should be applied to cancer risk assessment and early detection. Evaluating cancer risk in seemingly healthy individuals, as well as in elderly individuals, patients with metabolic syndrome, and those with diabetes, could provide valuable data for personalized interventions. A step-by-step strategy can be used to customize treatment objectives. The combination of early detection carcinogenesis enablers–based risk models should complement the most commonly used parameters of risk factors, as suggested in Fig. 2. It includes liquid biopsy parameters (e.g., DNA alterations, metabolites, inflammatory cells and immune imbalance) and microbiome signatures.

Figure 2.

Host-related factors and environmental exposures that may contribute to cancer risk and progression. (Created with BioRender.com. https://app.biorender.com/illustrations/684a69369ed7ecf3d531cde2).

Interception activities can be incremental with risk, based initially on lifestyle changes: smoking cessation; healthy eating habits; dietary interventions; implementation of physical activity; repurposed drugs, such as aspirin, metformin, and antidiabetic drugs; and therapeutic prophylaxis, as shown in Table 3.

Table 3.

Potential prevention and interception strategies.

Prevention	Interception
Do not smoke	Quit smoking
Avoid alcohol consumption	Quit alcohol consumption
Avoid carcinogens	Improve environment, decrease exposure
Have a healthy diet	Change dietary habits
Keep a healthy microbiome (167–171)	Get probiotics (167)
Exert physical activity	Increase physical activity
Do regular screening	Repeat prescribed exams
Carcinogenesis enablers detection for risk assessment	Dietary supplements Aspirin and anti-inflammatory, metformin and antidiabetics (165) Chemoprevention with natural, synthetic, or biological agents (165, 170)

Advances in Multicancer Detection Tests: Integrating Molecular Hallmarks for Early, Noninvasive Cancer Screening

Many of the limitations of the current approaches to screening and many of the liquid biopsy–based assays in development are that they are intended to detect a single cancer type. Multicancer detection tests (MCDT) are now being developed to identify, through a single assay, the presence of a cancer signal that may originate from one of several possible cancer types (172). Although such an approach holds the promise of improving the detection of cancers with a one-step, one-sample approach, the uncertainties around the results, especially in terms of cumulative false-positive rates, remain a challenge (173). Many cancer hallmarks have pleiotropic and context-specific roles, and recent technological advances, such as MCDTs, provide a molecular and integrative framework to assess these markers simultaneously in a noninvasive, systemic manner. In particular, markers such as ctDNA methylation patterns, cfDNA fragmentomics, and epigenetic signatures reflect hallmark-associated changes across different tumor types and stages and can thus serve as practical proxies for underlying biological processes, including inflammatory responses, within the multicancer screening context (174). Indeed, although cytokine profiles are variable across tumor types and influenced by host factors, they may still contribute to composite biomarker panels when combined with more specific molecular signatures, particularly in tumor-informed screening (175, 176).

The complementary use of MCDTs in screening programs could facilitate the simultaneous evaluation of the very initial phases of many different cancers and potentially enable earlier detection of clinically significant tumors in individuals without known risk factors or overt symptoms (112). By performing broad molecular profiling and ctDNA sequencing in a minimally invasive manner (177), NGS provides a wide identification of oncogenic drivers. Indeed, NGS is a cost-effective method that may be integrated into multitarget screening tests and in the optimization of targeted treatment selection in patients with cancer (174, 178–181). Advancements in NGS applications have allowed the evaluation of multiple cancer-related features, including fragmentomics and methylomics. Fragmentomics has already been validated for the early detection of several cancers, including esophageal squamous cell, colorectal, lung, liver cancers, and peripheral nerve tumors (182–187). Also, epigenomics is broadly used for MCDT applications (112). Tests detecting transformative markers shared by multiple tumor types at early stages have been proposed and are either available or under development (Table 4; refs. 188–190). Among the available MCDTs, the Galleri test by Grail detects cancer-specific DNA methylation patterns in cfDNA and screens for potentially any cancer type. The test was developed and validated in the Circulating Cell-free Genome Atlas study (191). It detected a signal for more than 50 distinct cancer types from a single blood draw and predicted the tumor origin in 1,273 out of 1,435 (89%) participants (188). Its feasibility was further evaluated in PATHFINDER, a prospective cohort study of 6,662 oncology and primary care outpatients aged over 50 years without signs or symptoms of cancer (192, 193). Preliminary findings indicate high specificity for malignancy (>99.0%). However, sensitivity is low in patients with stage I disease and varies depending on the cancer type (194).

Table 4.

Selected available MCDTs.

Test	Technology	Screened cancers
Adela (Adela Bio)	cfDNA	Lung, colon–rectum, breast, pancreas, bladder, kidney
Tr(ACE) (Biological Dynamics)	Exosome proteins	Pancreas, ovary, bladder
VPAC receptor TP4303 (Thomas Jefferson University/Intermountain Health)	Near-IR optical microscopy	Lung, breast, prostate, head–neck, uterus
Delphi (Delphi Diagnostics)	cfDNA fragmentomics	Lung, colorectal, pancreas, stomach, ovary
ARTEMIS	Genome-wide approach for analyzing repeat landscapes	​
Galleri (Grail)	ctDNA methylation pattern	Lung, colon–rectum, pancreas, liver, esophagus, stomach, ovary, bladder, head–neck
CancerSEEK	cfDNA NGS	Lung, colorectal, breast, pancreas, liver, esophagus, stomach, ovary
Bluestar MCED	cfDNA 5-hydroxymethylcytosine sequencing; fragmentomics	Lung, colorectal, breast, pancreas, liver, esophagus, stomach, ovary
OverC	ELSA sequencing	Lung, colorectal, pancreas, liver, esophagus, ovary
MI GPSai	cfDNA fragmentomics	Lung, colorectal, breast, stomach, kidney
MIRAM	Ultrahigh-performance LC/MS glycosaminoglycans/Elypta’s SKY software	Breast, stomach, prostate, bladder, kidney, lymphoma/leukemia, brain
FMBT	Multiomics/AI	Lung, colorectal, breast, pancreas, liver, esophagus, stomach, ovary, prostate, bladder, kidney, uterus
LungLB	CTC FISH; imaging AI	Lung, breast, liver, kidney
Signatera	cfDNA NGS; protein markers	Lung, colorectal, breast, pancreas, liver, ovary, bladder
Sentinel-10	CpG-cfDNA qPCR	Lung, colorectal, breast, pancreas, liver, ovary
OneTest	Circulating cancer antigens by AI	Lung, colorectal, breast, pancreas, liver, ovary, prostate
Quantum sensor/OncoProfiler	CTC surface-enhanced Raman scattering/machine learning	Lung, colorectal, breast, prostate
PanSeer	ctDNA methylation	Stomach, esophageal, colorectal, lung, liver cancers

Moreover, there is an ongoing debate about the appropriateness of this test as an MCDT because one of the main concerns about the Galleri test is its limited sensitivity for early-stage cancers (27.5% overall, 52.8% in a subset of 12 cancers with high unmet needs). This may compromise its value as a screening tool aimed at early detection, in which curative interventions are most effective. Also, despite its high specificity and low reported false-positive rate (0.5%), the low positive predictive value (PPV) and downstream consequences of false positives, such as costly and potentially harmful diagnostic investigations, pose a significant burden on the healthcare system. The absence of robust longitudinal data and reliance on surrogate endpoints such as stage shift rather than mortality reduction in the evaluation of the test also represents a concern, as this approach diverges from the established UK National Screening Committee framework for adopting screening programs. These issues, combined with political and commercial influences and the lack of clear evidence for cancer-specific mortality benefits, likely contributed to the suspension of widespread rollout (195). For instance, an early analysis of the UK-based screening clinical trial based on the Galleri test did not provide compelling evidence to justify an immediate large-scale rollout within the national health system. Although initial findings showed promise, they did not meet the high thresholds required for accelerated implementation, and final data are awaited. New trials to refine the Galleri approach are running. Other MCDTs are available, and we are reviewing some of the most interesting.

Based on the DETECT-A study, CancerSEEK, a blood-based MCED test, demonstrated the ability to identify early-stage cancers, including types lacking standard-of-care screening options. Of the 26 participants with cancers first detected by CancerSEEK, 50% were alive and cancer-free at a median follow-up of 4.4 years. All patients with stage I or II cancers who received treatment remained cancer-free, highlighting the potential clinical benefit of early detection. Notably, seven of the cancer-free survivors had malignancies for which no screening programs currently exist. These results suggest that CancerSEEK may support curative interventions through earlier diagnosis although larger randomized trials are required to confirm its population-level impact and long-term benefits (196).

Adela by Adela Bio detects cfDNA and screens for lung, colorectal, breast, pancreas, bladder, and kidney cancers, as well as certain types of leukemia. Tr(ACE) by Biological Dynamics targets exosome proteins, exploiting AI, and screens for pancreas, ovarian, and bladder cancers. The VPAC receptor TP4303 by Thomas Jefferson University/Intermountain Health employs near-IR optical microscopy and screens for lung, breast, prostate, head and neck, and uterine cancers. ARTEMIS is a genome-wide approach for analyzing repeat landscapes. Delphi (Delphi Diagnostics) utilizes a learning machine algorithm to target cfDNA. Changes in repeat landscapes are detected to monitor cancer development and identify the tissue of origin of tumors. Individual repeat landscapes can also be used in machine learning to generate a disease-predicting and characterizing score (Table 4; ref. 197). The SPOT-MAS test showed promising results for early cancer detection in asymptomatic individuals using a noninvasive, ctDNA-based liquid biopsy approach. Among 9,024 participants, the test detected 17 confirmed cancer cases (true positives) and achieved a sensitivity of 70.83% and specificity of 99.71%, with a PPV of 39.53% and a negative predictive value of 99.92%. Importantly, 12 of the confirmed cancers were at early stages and potentially curable. The test also predicted the tissue of origin with 52.94% accuracy and was able to detect cancers that currently lack standard screening methods, such as gastric, liver, and endometrial cancers. These results underscore SPOT-MAS’s potential as a complementary screening tool, especially in low- and middle-income countries where conventional cancer screening is limited. However, limitations include a small number of confirmed cancers and challenges in detecting some types, such as breast cancer (198). The Trucheck Intelli test is an MCED assay capable of identifying approximately 70 different solid tumors, including cancers of the head and neck, thyroid, and skin, by detecting circulating ensembles of tumor-associated cells. In a proof-of-concept study, detection rates were 91.7% for head and neck cancers, 88.3% for thyroid cancers, and 84.4% for skin cancers, with no significant differences based on treatment status or metastatic stage. The test demonstrated high sensitivity and organ-specific accuracy exceeding 98% for most cancer types. For example, thyroid cancers achieved 100% sensitivity and specificity, and sarcomas reached 95.1% sensitivity and 100% accuracy. Results are categorized as positive, negative, or indeterminate based on antigen detection and tissue of origin. Although not FDA-approved in the USA, Trucheck Intelli is commercially available in the UK; in any case, medical consultation is advised (199).

PanSeer is a noninvasive blood test that detects cancer based on ctDNA methylation. The Taizhou Longitudinal Study identified cancer in 95% of asymptomatic individuals who were later diagnosed within 4 years and in 88% of postdiagnosis patients, with a specificity of 96%. These results suggest that PanSeer can detect stomach, esophageal, colorectal, lung, and liver cancers up to 4 years before standard diagnosis, highlighting its potential for early cancer screening. However, further longitudinal studies are needed to confirm its clinical utility (200).

The utility of MCDTs in clinical practice is supported by modeling studies (201). This research predicts that adding multicancer tests to standard care could prevent 39% of all cancer-related deaths within 5 years of diagnosis, which would otherwise be expected to occur in individuals with a positive test result (201). However, findings suggest that such tests may be less effective in healthy individuals, who arguably represent the most relevant target group (202), raising concerns about their implementation in screening practice at the population level. The clinical utility of these tests in asymptomatic populations, their performance, and their impact on real-world population screening when added to the standard of care are currently being evaluated in prospective randomized trials (194, 203–205). Nevertheless, ongoing trials have faced criticism for employing criteria considered unsuitable to justify the adoption of a new national screening program aimed at reducing cancer burden and saving lives (206). Validating these tests requires not only the demonstration of sensitivity and specificity but also the evaluation of survival benefits, risks associated with invasive over-diagnostic assessments, feasibility, and cost-effectiveness. The relevance of many published studies on the efficacy of MCDTs is limited by their case–control design and the fact that they are conducted in patients already diagnosed with malignancy. Consequently, the sensitivities, specificities, and PPVs reported in these studies might be higher than if the tests were used to screen an asymptomatic population. At present, there are no established criteria and trial methodologies to assess the benefits and harms of positive or negative results, which may lead to invasive diagnostic procedures and unnecessary treatments. For instance, an early analysis of the UK-based screening clinical trial based on the Galleri test did not provide compelling evidence to justify an immediate large-scale rollout within the national health system. Although initial findings showed promise, they did not meet the high thresholds required for accelerated implementation, and final data are expected in 2026. Additionally, the survival benefit associated with early diagnosis should be measured for each cancer type rather than reported only as a global measure (207). A significant unsolved issue is that the organ site responsible for a positive test result is not always easily identified. Despite the promise of determining the source of altered cells, a broad diagnostic workup may still be required to confirm the location and type of underlying cancer (202). Beyond scientific reliability, these tests must be socially acceptable and feasible for healthcare providers (208).

AI and Computational Advances in Cancer Detection

AI plays a crucial role in cancer risk prediction, early detection, and diagnostic accuracy by integrating routinely collected and analyzed patient clinical, multiomics, and imaging data (209). AI-driven approaches enhance biomarker discovery, optimize screening strategies, and support personalized interventions, ultimately improving early cancer diagnosis and prevention (210). Machine learning models refine early cancer detection sought by liquid biopsy–based ctDNA analysis and radiomics, improving tumor classification and reducing false positives (211, 212). Lung Cancer Likelihood in Plasma (CLiP) is an example of an AI-driven tool that applies ensemble-based machine learning to define lung cancer ctDNA in plasma samples. By combining classification algorithms such as nearest neighbor classifiers, naïve Bayes models, logistic regression, and decision trees, Lung-CLiP enhances the robustness and reliability of detection (213). Natural language processing helps analyze electronic medical records to extract valuable diagnostic information, support a more accurate assessment of individual cancer risks, and enhance guidance on the timing and need for screenings (214, 215). AI also aids in drug discovery and repurposing (216). However, challenges such as data biases, overdiagnosis, and standardization must be addressed to ensure clinical reliability (217). Although AI enhances clinical decision-making, it should complement, not replace, physician expertise, particularly in rare cancers and ambiguous clinical diagnoses (218).

AI is set to transform cancer risk prediction, early detection, and diagnostics by integrating key driver and enablers, including multiomics and imaging data, thereby enhancing biomarker discovery, screening and personalized care. Tools such as Lung-CLiP and natural language processing improve ctDNA analysis and electronic health record data mining though challenges such as bias, overdiagnosis, and model standardization remain, requiring AI to complement, not replace, clinical expertise. Refining AI models to balance sensitivity and specificity is essential for meaningful cancer screening while minimizing unnecessary interventions. Future research should explore AI’s potential in integrating multiomics, proteomics, inflammation, and microbiome data; making our suggestion of combination early detection more feasible; refining early cancer interception strategies; and optimizing risk assessment algorithms to improve patient outcomes.

Cancer-Associated Polygenic Risk Scores

Several novel strategies are currently being implemented to identify asymptomatic, apparently healthy people at risk who are more likely to benefit from MCED tools. A modeling analysis recently investigated the utility of current, future, and optimized polygenic risk scores (PRS), which are based on the evaluation of thousands of SNPs in UK cancer screening programs (219). In summary, the PRS-defined high-risk quintile of the population was estimated to capture 37% of breast cancer cases, 46% of prostate cancer cases, 34% of colorectal cancer cases, 29% of pancreatic cancer cases, 26% of ovarian cancer cases, 22% of kidney cancer cases, 26% of lung cancer cases, and 47% of testicular cancer cases. However, these estimates may be significantly attenuated due to incomplete uptake of PRS profiling in non-European ancestry populations. Taken together, these models indicate that at present, there might be only modest efficiency gains in cancer case detection through a PRS-stratified screening program. Nonetheless, they pave the way for future cluster-specific studies in selected populations.

The next step will involve leveraging these signatures, AI, and machine learning to identify changes associated with the escape of nascent cancer cells from the immune system’s coordination in high-risk, healthy individuals.

Conclusions

Assessing the complexity of genetic and nongenetic factors in cancer development can be effectively done through blood-based methodologies, particularly liquid biopsy, and through metagenome analysis to determine microbiota imbalances. Monitoring these factors may help develop advanced risk assessment algorithms and treatment targets. Each impaired mechanism could serve as both a diagnostic tool and a target for specific interventions.

Cancer can no longer be viewed as an inevitable event but as a preventable consequence of identifiable and interceptable biological processes. Liquid biopsy provides a noninvasive means to detect not only CTCs and tumor-derived nucleic acids but also circulating immune cells, EVs, proteins, and metabolites. Beyond genomic and proteomic alterations, evidence suggests the advantage of evaluating systemic inflammation and immune dysregulation. Gut microbiota imbalances are increasingly considered to be implicated in cancer progression and treatment resistance. Advances in liquid biopsy, multiomics, and combined detection of drivers and enablers, with the help of AI-driven analytics, offer unprecedented opportunities to detect transformative signals early, stratify risk, and intervene before malignancy emerges. Multicancer detection assays expand this approach beyond the one-site/one-modality paradigm, identifying cancer-related signals from a broader spectrum of malignancies from a single sample of blood. The future of oncology lies not only in early diagnosis but also in prediagnostic interception, transforming population screening from tumor detection to risk prediction (220, 221). As cardiology has taught us, integrated, noninvasive, multiparameter risk scores could redefine cancer prevention for the 21st century (188–190). It is time for oncology to shift from reactive vigilance to anticipatory medicine, reclaiming health before disease declares itself.

Data availability

Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.

Authors’ Disclosures

F. Bertolini reports grants from Pfizer and Menarini outside the submitted work. No disclosures were reported by the other authors.