Introduction
Circulating tumor cells (CTCs) are tumor cells that eroded into the vasculature from the primary tumor and have the capacity to form new cancers away from the original site (Ref. Reference Lin, Shen, Luo, Zhang, Li, Yang, Zhu, Zhou, Zheng, Chen and Zhou1). Thomas Ashworth first identified them in 1869, and he described them as having significant metastatic potential (Ref. Reference Lee, Kim, Jeon and Park2). As our understanding of cancer progressed in the later 20th century, the divergent roles of cells within a tumor were also extensively studied. Tumours may be solid, as in the case of carcinomas, lymphomas and sarcomas, or ‘liquid’, as in the case of leukemias or myelomas (Ref. Reference Carbone3). Thus, tumor pathology has lent itself to be a core discipline in understanding the true nature of cancer, and thereby addressing the questions of how and why it may spread in the human body. The earliest studies, conducted in the 1880s by Rudolf Virchow, established malignancy as the fundamental difference between a benign mass and a malignant mass (Ref. Reference David4). These findings, although valid, were delayed in being substantiated due to a lack of technological advancements. However, with improvements to microscopy, cytology and scientific procedure, earlier hypotheses, like those of Virchow’s, were validated, along with new theories emerging as to the true nature of the cells that contribute to cancerous behaviours. A landmark 1955 study by the Danish scientist, H.C. Engell, suggested that circulating tumor cells could be implicated in recurrence and metastasis, thereby having a direct effect on prognosis, of which the latter conclusion was corroborated by Roberts et al., in 1958 (Refs. Reference Engell5, Reference Roberts, Jonasson, Long, Mcgrath, Mcgrew and Cole6). Such studies progressed late into the 20th century, with research on the topic continuing well into the 1970s, when Fidler et. al., established that distant metastasis is directly correlated with the number of CTCs (embolic cells) injected in mice (Ref. Reference Fidler7). While establishing CTC roles in the pathophysiology of cancer is important, the primary focus of related research in the 21st century remains the isolation, detection and capturing of CTCs for prognostic, diagnostic and therapeutic purposes (Ref. Reference Harouaka, Kang, Zheng and Cao8). The concurrent advancements in related technologies, such as microfluidics, liquid biopsies, immunocapture, etc., have lent themselves to developing high-precision techniques that can isolate CTCs (Ref. Reference Harouaka, Kang, Zheng and Cao8). Despite their low concentration in the bloodstream, their potential as a therapeutic target cannot be ignored, owing to their distinct role in distant metastasis and recurrence (Ref. Reference Paoletti, Hayes and Stearns9). These techniques span a wide variety of dependent characteristics, as they can be based off the physical characteristic of CTCs, such as size differences between CTCs and normal blood cells, or biomarker differences, such as the presence of EpCAM (Refs. Reference Joosse, Gorges and Pantel10–Reference Zhong, Zhang, Zhu, Liang, Yuan, Li, Li, Li, Jia, He, Zhu, Sun, Jiang, Zhang, Wang and Ke12). However, the primary question remains – Are CTCs the optimal target for preventing or predicting metastasis and recurrence? To answer this question, we must consider the available targets – in the case of solid tumours, the primary targets for treatment or diagnosis are the primary tumor, immune cells in the TME and CTCs. Resection of the primary tumor, especially in the early stages of cancer, is often the standard of care, although it is not always foolproof (Ref. Reference Hiller, Perry, Poulogiannis, Riedel and Sloan13). However, complete resection is not a viable option at advanced stages or in certain cancers like those of the lung, owing to the risk in damaging vital organs, as outlined in the NCCN treatment guidelines (Ref. Reference Ettinger, Wood, Aisner, Akerley, Bauman, Bharat, Bruno, Chang, Chirieac, D’Amico, DeCamp, Dilling, Dowell, Gettinger, Grotz, Gubens, Hegde, Lackner, Lanuti, Lin, Loo, Lovly, Maldonado, Massarelli, Morgensztern, Ng, Otterson, Pacheco, Patel, Riely, Riess, Schild, Shapiro, Singh, Stevenson, Tam, Tanvetyanon, Yanagawa, Yang, Yau, Gregory and Hughes14). Even in the case of blood cancers, chemotherapy and radiotherapy may drastically improve the patient’s chances of survival, but the QoL is drastically reduced post treatment (Refs. Reference Shadman15,Reference Allart-Vorelli, Porro, Baguet, Michel and Cousson-Gélie16). Such instances prompt extensive research into CTCs, as their presence can be definitively correlated with malignant cancer, and targeting them would reduce the risk of recurrence at a later stage, which was an issue Virchow highlighted as early as 1888, when Kaiser Wilhelm III died from recurrent laryngeal cancer (Ref. Reference Cardesa, Zidar, Alos, Nadal, Gale and Klöppel17).
Thus, in this paper, we examine the modern technologies at the forefront of CTC detection, capture isolation and targeting, with contextual explanations regarding their production. We also discuss their role in altering immune surveillance within the TME, which serves as the basis for highly precise techniques, such as immunocapture, light- and sound-based therapies and nanotechnology.
Fundamental biology of circulating tumour cells
Basic properties and origin of CTCs
During metastasis to a distant site, mostly through circulating blood, tumor cells often disseminate by executing a coordinated cascade of intravasation, translocation through the circulation, extravasation and finally, colonization of distant tissues (Ref. Reference Chen, Bode and Dong18). These events require a coordinated change in adhesion, cytoskeletal organization and stromal signalling that enable vascular entry (Ref. Reference Fife, Mccarroll, Kavallaris, Kavallaris and Program19). Once tumour cells breach the endothelial barrier and enter the vasculature, they are classified as CTCs (Ref. Reference Mayo and Kutys20).
The circulatory environment is highly hostile to these CTCs, as they are exposed to shear stress, oxidative damage and continuous immune surveillance (Ref. Reference Wang, Zhang, Peng, Li, Wang, Bie, Shi, Lin and Zhang21). As a result, only a very small fraction of intravasated cells survives long enough to contribute to metastatic spread. To endure these conditions, CTCs depend on strategies such as modulation of biomechanical and cytoskeletal properties to better tolerate hemodynamic forces during circulation (Ref. Reference Kurma and Alix-Panabières22). They also activate stress–response pathways and undergo metabolic remodelling that enhances their ability to withstand the conditions in the bloodstream (Ref. Reference Senft and Ronai23). To avoid immune surveillance, CTCs frequently engage in transient interactions with platelets and immune cells, which provide shielding and modulate immune detection (Refs. Reference Wang, Zhang, Peng, Li, Wang, Bie, Shi, Lin and Zhang21,Reference Pereira-Veiga, Schneegans, Pantel and Wikman24), this is discussed further in Section “Mechanism to escape immune surveillance”.
CTCs can be found in two main distributions, either as single cells or clusters. The functional and distinctions and clinical relevance of CTC clusters are covered in Section “Biology and Clinical Role of CTC Clusters”. Similarly, the stemness-associated and epithelial-mesenchymal plasticity programs that influence CTC survival and metastatic potential are discussed in Section “Stemness and Epithelial Mesenchymal Transition in CTCs”.
Tumour cells detach from the primary tumour and enter the bloodstream through intravasation. Upon entering circulation, CTCs encounter a highly hostile environment that limits their survival and metastatic potential. Loss of attachment to the ECM can trigger anoikis, which can lead to apoptosis. In parallel, CTCs are also exposed to shear stress in the bloodstream, which can result in cellular fragmentation and death. Immune surveillance further eliminates CTCs through cytokine-mediated cytotoxicity and interactions with immune cells. Despite these challenges, a subset of CTCs survives circulation and metastasizes in distant organs, leading to metastatic tumours.
Stemness and epithelial–mesenchymal transition in CTCs
CTCs comprise a highly heterogenous population, which may include tumour cells, or drivers of proliferating cells, i.e. cancer stem cells (CSCs) (Ref. Reference Papaccio25). These cells possess unique abilities that enable them to separate from the primary tumour, enter the bloodstream and initiate metastasis at distinct locations (Ref. Reference Janjua, Chaudhary, Joshi, Tripathi and Bharti26). Introducing CSC biology in the context of CTCs is therefore essential, as accumulating evidence suggests that only a subset of CTCs have true metastatic capabilities (Refs. Reference Rapanotti, Cenci, Scioli, Cugini, Anzillotti, Savino, Coletta, Di Raimondo, Campione, Roselli, Bernardini, Bianchi, De Luca, Ferlosio and Orlandi27, Reference Tinhofer, Saki, Niehr, Keilholz and Budach28).
Two hypotheses are used to explain the origin of circulating CSCs (CCSCs). The first hypothesis suggests that CCSCs originate within the primary tumour. The second hypothesis proposes that CCSCs may arise from a state of dormancy after evading the primary tumours and subsequently propagate to form new cancers. These cells must navigate through challenges such as escaping the immune surveillance system and surviving in the hostile host environment. The subset of circulating tumour cells that successfully survives exhibits complex cascades of metastasis (Refs. Reference Oskarsson, Batlle and Massagué29, Reference Yang, Imrali and Heeschen30).
A connection can be assumed between CTCs and CSCs due to the shared properties between the two regarding self-renewal, differentiation, maturation and resilience to apoptosis. CCSCs contribute to tumour heterogeneity, metastasis and therapeutic resistance. The sequential steps involve complex phenotypic plasticity, as well as epigenetic and genetic factors (Refs. Reference Lin, Shen, Luo, Zhang, Li, Yang, Zhu, Zhou, Zheng, Chen and Zhou1, Reference Agnoletto, Corrà, Minotti, Baldassari, Crudele, Cook, Di Leva, D’Adamo, Gasparini and Volinia31). Consistent with this, stem-like CTCs have been detected in multiple cancer types, including breast, lung, glioblastoma and colorectal cancer (Refs. Reference Barrière, Riouallon, Renaudie, Tartary and Rigaud32–Reference Visioli, Giani, Trivieri, Pracella, Miccinilli, Cariglia, Palumbo, Arleo, Dezi, Copetti, Cajola, Restelli, Papa, Sciuto, Latiano, Carella, Amadori, Gallerani, Ricci, Alfieri, Pesole, Vescovi and Binda35), using CSC-associated markers such as CD44+/CD24-, ALDH1+, CD133 (Ref. Reference Keysar and Jimeno36) and pluripotency-related transcription factors including OCT4, NANOG and SOX2 (Ref. Reference Masciale, Banchelli, Grisendi, Samarelli, Raineri, Rossi, Zanoni, Cortesi, Bandini, Ulivi, Martinelli, Stella, Dominici and Aramini37). Clinically, the presence of these stem-like CTCs correlates with increased metastatic potential, therapy resistance, disease recurrence and poor prognosis (Refs. Reference Dianat-Moghadam, Azizi, Eslami-S, Cortés-Hernández, Heidarifard, Nouri and Alix-Panabières38, Reference Papadaki, Stoupis, Theodoropoulos, Mavroudis, Georgoulias and Agelaki39) .
Epithelial-mesenchymal transition (EMT), a progressive process commonly observed across various cancers, has been linked to the properties of CSCs. EMT often leads to the development of mesenchymal-like CTCs, these CTCs exist as a singular unit, while partial EMT is often associated with multicellular CTCs. This suggests that the regulation of stemness may differ between individual CTCs and those that are found in clusters, highlighting the necessity for therapeutic strategies that specifically target the physical state of CTCs in circulation (Refs. Reference Yu40, Reference Barriere, Fici, Gallerani, Fabbri, Zoli and Rigaud41). Importantly, EMT is not a binary process, and many CTCs display hybrid epithelial/mesenchymal (E/M) phenotypes that combine both epithelial survival advantages with mesenchymal invasive traits. Emerging evidence also suggests that these cells can lead collective invasion and coordinate multicellular dissemination, all while retaining their self-renewing capacity and plasticity (Ref. Reference Quan, Wang, Lu, Ma, Wang, Xia, Wang and Yang42). Hybrid E/M CTCs are recognized as the most aggressive and plastic subpopulation, exhibiting strong self-renewal capacity and metastatic competence (Refs. Reference Liao and Yang43, Reference Jolly, Boareto, Huang, Jia, Lu, Onuchic, Levine and Ben-Jacob44).
Biology and clinical role of CTC clusters
Although most CTCs are single cells, a small portion circulate as a group of clustered cells, this has shown to be an adaptive mechanism that improves the survival of CTCs in the circulatory system, while also enhancing their metastatic capabilities (Refs. Reference Schuster, Taftaf, Reduzzi, Albert, Romero-Calvo and Liu45, Reference Liu, Taftaf, Kawaguchi, Chang, Chen, Entenberg, Zhang, Gerratana, Huang, Patel, Tsui, Adorno-Cruz, Chirieleison, Cao, Harney, Patel, Patsialou, Shen, Avril, Gilmore, Lathia, Abbott, Cristofanilli, Condeelis and Liu46). CTCs have traditionally been defined as groups of tumour cells with two nuclei that have the ability to shed as clusters from the primary tumour, formation through cellular aggregation has also been observed. These clusters can be composed of tumour cell-tumour cell homotypic clusters or they can be composed of tumour cell-blood cell heterotypic clusters (Ref. Reference Schuster, Taftaf, Reduzzi, Albert, Romero-Calvo and Liu45).
Homotypic clusters
Homotypic clusters consist exclusively of tumour cells that disseminate as cohesive units from the primary tumour or arise through intracellular aggregation in the bloodstream. Compared to singular CTCs, homotypic clusters exhibit a marked uptick in metastatic efficiency, this has been attributed to cell-cell interactions that promote tumorigenic properties (Refs. Reference Liu, Taftaf, Kawaguchi, Chang, Chen, Entenberg, Zhang, Gerratana, Huang, Patel, Tsui, Adorno-Cruz, Chirieleison, Cao, Harney, Patel, Patsialou, Shen, Avril, Gilmore, Lathia, Abbott, Cristofanilli, Condeelis and Liu46, Reference Aceto, Bardia, Miyamoto, Donaldson, Wittner, Spencer, Yu, Pely, Engstrom, Zhu, Brannigan, Kapur, Stott, Shioda, Ramaswamy, Ting, Lin, Toner, Haber and Maheswaran47). Their formation is driven by tumour cell-cell adhesion mediated by adhesion and junction molecules such as CD44, intracellular adhesion molecule 1 (ICAM1) and plakoglobin, which reinforce cluster strength under stressful conditions, such as intratumour hypoxia (Refs. Reference Schuster, Taftaf, Reduzzi, Albert, Romero-Calvo and Liu45, Reference Donato, Kunz, Castro-Giner, Paasinen-Sohns, Strittmatter, Szczerba, Scherrer, Di Maggio, Heusermann, Biehlmaier, Beisel, Vetter, Rochlitz, Weber, Banfi, Schroeder and Aceto48). At the molecular level, homotypic clustering is associated with transcriptional and epigenetic programs that promote stemness, proliferation and survival. This is especially done through upregulation and DNA hypomethylation of key stemness genes such as OCT4, SOX2 and NANOG (Ref. Reference Gkountela, Castro-Giner, Szczerba, Vetter, Landin, Scherrer, Krol, Scheidmann, Beisel, Stirnimann, Kurzeder, Heinzelmann-Schwarz, Rochlitz, Weber and Aceto49), as well as activation of the EGFR and PAK2/FAK signalling pathways (Refs. Reference Liu, Taftaf, Kawaguchi, Chang, Chen, Entenberg, Zhang, Gerratana, Huang, Patel, Tsui, Adorno-Cruz, Chirieleison, Cao, Harney, Patel, Patsialou, Shen, Avril, Gilmore, Lathia, Abbott, Cristofanilli, Condeelis and Liu46, Reference Liu, Adorno-Cruz, Chang, Jia, Kawaguchi, Dashzeveg, Taftaf, Ramos, Schuster, El-Shennawy, Patel, Zhang, Cristofanilli and Liu50). In addition, homotypic clusters had increased expression of keratin 14, which showed lower major histocompatibility complex 2 (MHC2) gene expression, suggesting an adhesion and immune evasion phenotype for clusters (Refs. Reference Schuster, Taftaf, Reduzzi, Albert, Romero-Calvo and Liu45, Reference Cheung, Padmanaban, Silvestri, Schipper, Cohen, Fairchild, Gorin, Verdone, Pienta, Bader and Ewald51). A promising target for the mediation of homotypic clusters is heparanase, an extracellular matrix-degrading enzyme. Overexpression of heparanase has increased CTC clustering and induced aggregation of ICAM1 and FAK in the bloodstream (Ref. Reference Wei, Sun, Yang, Yan, Zhang, Zheng, Fu, Geng, Huang and Ding52). A phase I trial evaluated the drug digoxin for its ability to disrupt heparanase-associated survival of CTC clusters (Ref. Reference Kurzeder, Nguyen-Sträuli, Krol, Ring, Castro-Giner, Nüesch, Asawa, Zhang, Budinjas, Gvozdenovic, Vogel, Kohler, Grašič Kuhar, Schwab, Heinzelmann-Schwarz, Weber, Rochlitz, Vorburger, Frauchiger-Heuer, Witzel, Wicki, Kuster, Vetter and Aceto53), this is discussed further in Section “Small Molecule Therapeutics”.
Heterotypic clusters
Heterotypic clusters involve clustering with cell types such as WBCs, specifically neutrophils and myeloid-derived suppressor cells, cancer-associated fibroblasts have also been found to form clusters (Refs. Reference Schuster, Taftaf, Reduzzi, Albert, Romero-Calvo and Liu45, Reference Szczerba, Castro-Giner, Vetter, Krol, Gkountela, Landin, Scheidmann, Donato, Scherrer, Singer, Beisel, Kurzeder, Heinzelmann-Schwarz, Rochlitz, Weber, Beerenwinkel and Aceto54, Reference Sprouse, Welte, Boral, Liu, Yin, Vishnoi, Goswami-Sewell, Li, Pei, Jia, Glitza-Oliva and Marchetti55). Sequencing analysis revealed that CTC-neutrophil clusters are the most common (Ref. Reference Szczerba, Castro-Giner, Vetter, Krol, Gkountela, Landin, Scheidmann, Donato, Scherrer, Singer, Beisel, Kurzeder, Heinzelmann-Schwarz, Rochlitz, Weber, Beerenwinkel and Aceto54). The existence of cytoskeletal bridges between tumour cells and other cells most often influences the formation of clusters. Key cytoskeleton genes, such as TUB, GLU and VLM, were observed to be highly upregulated in metastatic breast cancer (Ref. Reference Kallergi, Aggouraki, Zacharopoulou, Stournaras, Georgoulias and Martin56).
Patients with CTC-WBC clusters have significantly worse overall survival after 6 months of treatment than patients without CTC-WBC clusters in their blood (Ref. Reference Jansson, Bendahl, Larsson, Aaltonen and Rydén57). Heterotypic clusters have been found to use their clustered partner for protection, while also enhancing dissemination and metastatic growth. For example, CTC-neutrophil clusters have been found to have enhanced cell-cycle progression transcriptomes (Ref. Reference Szczerba, Castro-Giner, Vetter, Krol, Gkountela, Landin, Scheidmann, Donato, Scherrer, Singer, Beisel, Kurzeder, Heinzelmann-Schwarz, Rochlitz, Weber, Beerenwinkel and Aceto54). Similarly, cancer-associated fibroblasts have been found to aid in the migration or circulation of the tumour cells (Ref. Reference Hurtado, Martínez-Pena and Piñeiro58). Finally, myeloid-derived suppressor cells were found to contribute to metastatic efficiency through promotion of CTC proliferation and survival through the ROS/Notch/Nodal pathway (Ref. Reference Sprouse, Welte, Boral, Liu, Yin, Vishnoi, Goswami-Sewell, Li, Pei, Jia, Glitza-Oliva and Marchetti55). While certain heterotypic links have been uncovered, further research is required to fully understand their roles in cancer progression.
Mechanism to escape immune surveillance
Escape from the immune system has long been seen as a crucial step in the development of tumours and their progression (Ref. Reference Mohme, Riethdorf and Pantel59). The survival of CTCs in the bloodstream also has clinical applications as it allows for the early detection of tumours in patients. This, in turn, allows clinicians to determine the risk of tumour relapse after surgical removal, and any possible resistance to treatments (Ref. Reference Pantel and Alix-Panabières60).
Cancer cells undergo metastatic dissemination and dormancy (Ref. Reference Gao, Zhang, Tang and Liang61). The communication between inactive cancer cells and extravasation sites involves different immune and stromal cells (Ref. Reference Chernosky and Tamagno62). After exiting the tumour microenvironment (TME), CTCs are exposed to vigorous immuno-surveillance in non-cancerous tissues. Due to the volume disparity between resident immune cells and CTCs outside the immunosuppressive environment of the tumour, CTCs are easily disintegrated by tumour targeting immune cells. Although the circulatory system represents a hostile environment for tumour-derived cells, due to the abundance of peripheral immune cells, a small subset of CTCs can still survive circulation and metastasize to distant sites (Ref. Reference Mohme, Riethdorf and Pantel59).
The expression of major histocompatibility complex 1 (MHC 1) is inhibited by tumour cells to escape cytotoxic T lymphocytes (CTLs), thereby triggering the activation of natural killer (NK) cells (Ref. Reference Waldhauer and Steinle63). CTCs evade immune supervision by attaining a ’pseudo normal’ phenotype over the transfer of platelet-derived MHC I molecules and the intervention of cytokeratins with MHC I interactions (Ref. Reference Placke, Örgel, Schaller, Jung, Rammensee, Kopp and Salih64).
In non-classic MHC molecules, human leukocyte antigen (HLA), like HLA-G, functions by binding to immune cell receptors such as Killer cell immunoglobulin-like receptors (KIRs), CD8 and LIR-1, leading to protection against immune-cytotoxicity mediated by T cells and NK cells (Ref. Reference Mohme, Riethdorf and Pantel59). Soluble HLA-G (sHLA-G) is generated through alternative splicing and acts as a systemic modulator of antitumor responses. High levels of sHLA-G, both free and vesicle-bound, are connected with poor outcomes in clinical settings in breast cancer patients, indicating immune evasion and CTC survival (Ref. Reference König, Kasimir-Bauer, Hoffmann, Bittner, Wagner, Manvailer, Schramm, Bankfalvi, Giebel, Kimmig, Horn and Rebmann65).
The Fas receptor (CD95) is present on many immune cell types. The physiological ligand, FASL binds to the FAS receptor, triggering a signalling cascade within the cell that eventually leads to apoptosis (Refs. Reference Strasser, Jost and Nagata66,Reference S Huang, Hahne, Schroeter, Frei, Fontana, Villunger, Newton, Tschopp and Strasser67). The tumour cells that express FASL can actively induce apoptosis in immune cells that would recognize and destroy them, thereby evading immune surveillance and clearance. Tumour cells undergo downregulation of FAS expression while upregulating FASL expression (Ref. Reference Reichmann68). Loss of FAS decreases the vulnerability of tumour cells to death induced by immune cells, weakening their chances of being eliminated. When FASL is upregulated on tumour cells, it improves their ability to activate apoptosis in immune cells, thereby suppressing immune responses against the tumour and CTCs (Ref. Reference Mohme, Riethdorf and Pantel59).
CD8+ T cells can be kept in an inactive state by CTCs by expressing immune checkpoint proteins. Most research currently focuses on the interactions between these proteins. The ideal example includes the interaction between programmed death protein 1 (PD1) and its associated ligand (PD-L1). The PD-1-PD-L1 interaction maintains cell survival by reducing cytokine production and simultaneously suppressing T-cell proliferation (Ref. Reference Keir, Butte, Freeman and Sharpe69). However, CTCs from various cancers have the ability to express PD-L1, making it a poor biomarker. Additionally, the heterogeneity regarding PD-L1 expression in CTCs when compared to primary tumours might provide an advantage in immune evasion (Ref. Reference Masmoudi, Vialaret, Hirtz and Alix-Panabières70).
Once cancer cells disseminate, CTCs rely on interactions with hemopoietic cells like platelets, neutrophils, monocytes and Treg cells for their survival (Refs. Reference Wang, Zhang, Peng, Li, Wang, Bie, Shi, Lin and Zhang21, Reference Pereira-Veiga, Schneegans, Pantel and Wikman24). Platelets encapsulate CTCs, sheltering them from the body’s immune response, while also protecting them from the excess shear stress exerted by blood flow (Refs. Reference Egan, Cooke and Kenny71, Reference Anvari, Osei and Maftoon72). The induction of platelet aggregation by tumour cells can facilitate the process of extravasation and adhesion, in addition to immune response evasion (Ref. Reference Wang, Sun, Liu, Wang, Yao and Wang73). The release of transforming growth factor-β (TGF-β) by platelets can deactivate NK cells, while the transfer of MHC-I complex from granular platelets to CTCs plays a crucial role in protecting CTCs against the cytotoxic assault of NK cells (Ref. Reference Ju, Chen, Zhang, Xu, Zhang, Li, Chen, Zhou, Ji and Wang74)
Techniques for enriching, detecting and isolating CTCs
Enrichment of CTCs
Numerous methods have been developed with the goal of CTC capture, however due to the rarity of CTCs in blood samples, it still proves challenging to isolate CTCs from the bloodstream accurately (Ref. Reference Lin, Shen, Luo, Zhang, Li, Yang, Zhu, Zhou, Zheng, Chen and Zhou1). CTCs can be separated based on different methods such as immunoaffinity, physical methods and biological methods (Refs. Reference Lin, Shen, Luo, Zhang, Li, Yang, Zhu, Zhou, Zheng, Chen and Zhou1, Reference Bankó, Lee, Nagygyörgy, Zrínyi, Chae, Cho and Telekes75, Reference Edd, Mishra, Smith, Kapur, Maheswaran, Haber and Toner76). Currently, the only Food and Drug Administration (FDA) approved enumeration technique is the immunoaffinity-based CellSearch (Ref. Reference Andree, van Dalum and Terstappen77). However, the Parsortix system received FDA clearance in 2022 for the capture and harvest of CTCs, expanding the range of clinically validated technologies.
Immunoaffinity-based methods
Immunoaffinity methods were amongst the first isolation techniques developed for CTC enrichment. The basis of this technique is unique as it uses specific antibodies designed to target antigens expressed exclusively on the surfaces of CTCs. Enrichment approaches can be broadly classified as positive and negative selection. Positive enrichment isolates CTCs by directly targeting tumour-associated markers, such as EpCAM, epidermal growth factor receptor (EGFR), HER2, mucin 1 (MUC1) and N-cadherin (Refs. Reference Eslami-S, Cortés-Hernández and Alix-Panabières78, Reference Li, Li, Shao, Li, Zhao and Ye79). In contrast, negative enrichment removes hematopoietic cells, primarily by targeting leukocyte makers such as CD45, leaving unlabelled CTCs behind (Ref. Reference Li, Li, Shao, Li, Zhao and Ye79). An advantage of negative enrichment is that it can retain CTCs that do not express the targeted antigen. (Refs. Reference Bankó, Lee, Nagygyörgy, Zrínyi, Chae, Cho and Telekes75, Reference Baccelli, Schneeweiss, Riethdorf, Stenzinger, Schillert, Vogel, Klein, Saini, Bäuerle, Wallwiener, Holland-Letz, Höfner, Sprick, Scharpff, Marmé, Sinn, Pantel, Weichert and Trumpp80). However, positive enrichment methods generally yield a higher purity.
In practice, immunoaffinity-based enrichment represents a broader methodological framework than a single platform. These approaches differ (i) in selection strategy (positive versus negative enrichment), (ii) antibody presentation, (iii) marker strategy (Ref. Reference Graves and Czerniecki81). While early immunoaffinity methods primarily relied on EpCAM-based positive selection, more recent approaches have introduced combinatorial marker panels to address the heterogeneity of CTCs and epithelial-mesenchymal plasticity (Refs. 81, Reference Lee, Kim, Park, Park, Yoo, Yum, Kang, Lee, Youn and Youn82). Platform-specific implementations of immunoaffinity principles, such as immunomagnetic, microfluidic and hybrid systems, are discussed in later sections.
The most popular example of immunoaffinity-based enrichment is the CellSearch© technology. This system uses whole blood and isolates CTCs based on EpCAM expression through the use of magnetic core particles coated with anti-epithelial cell adhesion molecule (anti-EpCAM) antibodies. Cells that are EpCAM-positive are sorted via magnetic field, and immunostained for CD45 and cytokeratins. Events/samples/cells which express cytokeratins, have a nucleus, do not express CD45, contain a DAPI-stained nucleus and are larger than 4×4μm2 in size are classified as CTCs (Refs. Reference Costa, Dávila-Ibáñez and Piñeiro83, Reference Repollet, Diagnostics and Rao84). Figure 1 details the role of EpCAM in CTC metastasis. (Refs. Reference Costa, Dávila-Ibáñez and Piñeiro83, Reference Repollet, Diagnostics and Rao84)
Biological challenges faced by CTCs during hematogenous metastasis.
EpCAM is highly involved in CTC-driven metastasis, cells that overexpress EpCAM separate from the primary tumour before entering the bloodstream as CTCs. CTCs can either exist as single cells or aggregates, where EpCAM plays an important role in their adhesion. Certain CTCs undergo EMT in the primary tumour leading to the downregulation of EpCAM, allowing for single cells/ loosely aggregated CTCs. Once they reach distant sites, CTCs interact with the stroma, leading to metastasis and the formation of secondary tumours.
Enrichment based on physical characteristics
CTC enrichment through physical separation is dependent on the differences in the density, deformability, electrical properties and size between blood cells and CTCs (Ref. Reference Lin, Shen, Luo, Zhang, Li, Yang, Zhu, Zhou, Zheng, Chen and Zhou1). Size isolation through an 8μm diameter polycarbonate TRACK-ETCH membrane was found to have low efficiency. However, this method has been improved through the introduction of a flexible micro spring array device, which incorporates a pressure-regulating system. This device is able to achieve 90% detection in 76% of samples (Ref. Reference Harouaka, Da, Yeh, Khan, Das, Liu, Christ, Dicker, Baney, Kaifi, Belani, Truica, El-Deiry, Allerton and Zheng85).
Building on this principle, isolation by size of epithelial tumour cells (ISET) was introduced as a standardized, label-free enrichment platform. In the original study by Vona et al., ISET employed track-etched polycarbonate microfilters with 8 μm pores to enrich CTCs from hepatocellular carcinoma patient blood samples; here, only 0.0002% of leukocytes were retained while allowing for detection of cytokeratin-positive CTCs (Ref. Reference Vona, Sabile, Louha, Sitruk, Romana, Schü, Capron, Franco, Pazzagli, Vekemans, Lacour, Bré and Paterlini-Bré86). Subsequent clinical studies highlight substantial discrepancies between EpCAM-dependent and size-based CTC enrichment approaches. In metastatic breast, lung and prostate cancers, comparison of CellSearch and ISET revealed concordant CTC detection in 55%, 20% and 60% of cases, respectively. These results suggest that technologies limited to CTC capture from EpCAM-positive cells might show poor performance in some cancers, such as metastatic lung carcinoma (Ref. Reference Farace, Massard, Vimond, Drusch, Jacques, Billiot, Laplanche, Chauchereau, Lacroix, Planchard, Le Moulec, André, Fizazi, Soria and Vielh87). Similarly, in pancreatic cancer, ISET detected CTCs in a significantly higher percentage of patients in CellSearch (93% vs 40%) and yielded higher CTC counts overall (Ref. Reference Khoja, Backen, Sloane, Menasce, Ryder, Krebs, Board, Clack, Hughes, Blackhall, Valle and Dive88). Other filtration methods include MetaCell and Parsotix, both of which are in use (Refs. Reference Raman, Prabhu, Kumar and Mani89, Reference Zhang, Xie, Zhang and Zhang90). MetaCell is an alternative size-based filtration system to isolate CTCs. It uses a porous polycarbonate membrane with a diameter of 8μm, which allows smaller blood components to pass through the membrane while CTCs are entrapped in the membrane. Due to its label-free nature, MetaCell is especially effective in filtering heterogenous CTC populations (Refs. Reference Edd, Mishra, Smith, Kapur, Maheswaran, Haber and Toner76, Reference Vasantharajan, Barnett, Gray, McCall, Rodger, Eccles, Munro, Pattison and Chatterjee91). In optimization studies using colorectal cancer cell lines spiked in healthy blood samples, MetaCell demonstrated consistently high CTC recovery rates (≥ 85%) across a range of concentrations, comparable to those reported for CellSearch, and simultaneously, it also achieved ≥ 95% leukocyte depletion. Clinical application of MetaCell showed CTC detection in approximately half of the samples, similar to results seen in ISET, with smaller CTCs possibly limiting sensitivity in certain tumours (Ref. Reference Vasantharajan, Barnett, Gray, McCall, Rodger, Eccles, Munro, Pattison and Chatterjee91). In contrast, Parsortix uses a microfluidic cassette with a tapering channel, which traps larger blood components such as CTCs, while smaller, more flexible cells, such as RBCs, are able to flow through (Ref. Reference Miller, Robinson, Wagner and O’Shannessy92). This design reduces clogging, facilitates gentle cell recovery and preserves cell viability. Comparative studies demonstrated that Parsortix captures CTCs at rates comparable to or exceeding EpCAM-based platforms, such as CellSearch, including CTCs with low or absent EpCAM expression. However, partial size overlap between leukocytes and CTCs may lead to smaller CTC loss, and leukocyte contamination (Ref. Reference Xu, Mao, Imrali, Syed, Mutsvangwa, Berney, Cathcart, Hines, Shamash and Lu93).
Beyond filtration-based approaches, other physical property-based methods have also been developed. The Apostream© system uses a dielectrophoresis-based microfluidic system to isolate CTCs by exploiting the difference in the dielectric properties of blood cells and CTCs. Using frequency-dependent electric fields, CTCs experience positive dielectric forces and are deflected toward collection regions, while most blood cells are repelled, which allows for effective separation without antibody labelling (Refs. Reference Xu, Mao, Imrali, Syed, Mutsvangwa, Berney, Cathcart, Hines, Shamash and Lu93, Reference Gupta, Jafferji, Garza, Melnikova, Hasegawa, Pethig and Davis94). Apostream demonstrated reproducible recovery percentages between 70 and 75% in spike-in experiments, with there being a strong correlation between input and recovered CTCs. Importantly, the separation method had a post-isolation cell viability of > 95%, allowing recovered cells to be analysed downstream (Ref. Reference Xu, Mao, Imrali, Syed, Mutsvangwa, Berney, Cathcart, Hines, Shamash and Lu93). However, Apostream requires a large volume of blood, which may limit its clinical use (Ref. Reference Xu, Mao, Imrali, Syed, Mutsvangwa, Berney, Cathcart, Hines, Shamash and Lu93). Oncoquick© is dependent on differences in density, which allows RBCs and WBCs to be filtered out (Ref. Reference Zhang, Chen and Fan95). The platform combines density-gradient separation with a porous barrier that stabilizes the layer formation, thereby enriching CTCs together with mononuclear cells, such as leukocytes, marking it as an improvement over Ficoll-based methods. Early clinical and spiking studies showed improved CTC recovery using Oncoquick, with recovery rates in the range of 80-87%, supporting its utility for CTC enumeration. However, the similar density gradients of CTCs and leukocytes often result in substantial co-enrichment, limiting its suitability as the sole enrichment strategy for downstream analysis of CTCs (Ref. Reference Rosenberg, Gertler, Friederichs, Fuehrer, Dahm, Phelps, Thorban, Nekarda and Siewert96).
Enrichment based on biological characteristics
Detection approaches based on biological characteristics are primarily dependent on the presence of EpCAM, mesenchymal markers and antigen-antibody interactions. Among these, EpCAM-based positive selection techniques, such as CellSearch are commonly used by researchers (Ref. Reference Lin, Shen, Luo, Zhang, Li, Yang, Zhu, Zhou, Zheng, Chen and Zhou1). Magnetic nanoparticles covered with a layer of antifouling hydrogel that inhibits nonspecific cell adhesion can be used for CTC capture. EpCAM antibodies are grafted on the hydrogel layer surface, providing greater specificity for CTCs. When tested on mimic clinical blood, results showed that 96% of CTCs were captured (Ref. Reference Wang, Wu, Sun, Cao, Cai, Yuan, Zou, Xing and Pei97). An aptamer-triggered-clamped hybridization chain reaction (atcHCR), has been used for cloaking/uncloaking and in situ hybridization of CTCs through the use of porous hydrogels. CTCs are captured by the EpCAM and are unharmed due to the gentle release of the hydrogel (Ref. Reference Song, Ye, Zuo, Li, Wang, Liu, Hwang, Chao, Su, Wang, Shi, Wang, Huang, Lal and Fan98). Currently, there is a shortage of universal markers that can be used. For example, CTCs that undergo EMT are often devoid of EpCAM, in which case negative selection methods such as CD45 depletion are more optimal (Ref. Reference Sharma, Zhuang, Long, Pavlovic, Kang, Ilyas and Asghar99). RosetteSep is a widely used negative selection-based enrichment method that enables the isolation of both singular CTCs and CTC clusters, while preserving their biological properties. The method relies on the removal of CD45+ cells using tetrameric antibody complexes, followed by density gradient centrifugation using Ficoll-Paque. It is a popular negative selection method, as it obtains both CTC clusters and single-cell CTCs. It does so through the removal of CD45-positive cells bound to tetrameric antibodies, before using centrifugation to precipitate them in a Ficoll-Paque density gradient (Ref. Reference Tretyakova, Menyailo, Schegoleva, Bokova, Larionova and Denisov100). The RosetteSep method has been successfully used across multiple cancer types, such as breast, prostate, small-cell lung and gastroesophageal cancer, to isolate intact CTCs and CTC clusters for downstream molecular and functional analyses, including transcriptomic profiling and in vivo metastasis assays (Refs. Reference Baccelli, Schneeweiss, Riethdorf, Stenzinger, Schillert, Vogel, Klein, Saini, Bäuerle, Wallwiener, Holland-Letz, Höfner, Sprick, Scharpff, Marmé, Sinn, Pantel, Weichert and Trumpp80, Reference Rossi, Valgiusti, Puccetti, Miserocchi, Zanoni, Angeli, Arienti, Pace, Bassi, Vannini, Melloni, Bandini, Urbini, Negrini, Bonafè, Ferracin and Gallerani101–Reference Gao, Vela, Sboner, Iaquinta, Karthaus, Gopalan, Dowling, Wanjala, Undvall, Arora, Wongvipat, Kossai, Ramazanoglu, Barboza, Di, Cao, Zhang, Sirota, Ran, Macdonald, Beltran, Mosquera, Touijer, Scardino, Laudone, Curtis, Rathkopf, Morris, Danila, Slovin, Solomon, Eastham, Chi, Carver, Rubin, Scher, Clevers, Sawyers and Chen103). However, the method requires a significant amount of time and can prove to be labour-intensive, due to which the reproducibility of results becomes unproductively laborious (Ref. Reference Tretyakova, Menyailo, Schegoleva, Bokova, Larionova and Denisov100).
Detection methods
In vivo detection methods
In vivo detection methods have enhanced sensitivity and are mostly used for early detection of CTCs. Presently, it is used only in basic research as it is lengthy, tedious, invasive and expensive. Liquid biopsy and CTC chip implantation methods are used to implement in vivo procedures. CTCs break away from the primary tumour and enter extracellular vesicles and the bloodstream, making it an ideal target for in vivo detection methods (Ref. Reference Wu, Zhu, Wei, Zhang, Zhang, You, Zhang, Song, Liu and Yang104).
The proposed method for capturing CTCs in vivo involves modifying vein indwelling needles with EpCAM antibodies (Ref. Reference Zhang, Jia, Wu, Zang, Yang, Chen and Tang105) and the structured medical wires that capture EpCAM+ CTCs (Ref. Reference Saucedo-Zeni, Mewes, Niestroj, Gasiorowski, Murawa, Nowaczyk, Tomasi, Weber, Dworacki, Morgenthaler, Jansen, Propping, Sterzynska, Dyszkiewicz, Zabel, Kiechle, Reuning, Schmitt and Lücke106). Recently used methods include in vivo flow fluorescence Cytometry (IVFC), Photoacoustic Flow Cytometry (PAFC) and the use of the GILUPI Cell Collector© (Ref. Reference Zhang, Xie, Zhang and Zhang90).
Microfluidic-based detection methods
Microfluidic devices have gained popularity for CTC isolation due to the small sample requirements, low cost and continuous processing (Ref. Reference Anwar, Reis, Zhang, Khan, Ali Kalhoro, Ur Rehman, Zhang and Liu107) and their ability to alter the properties of both fluids and particles at the microscale level. Microscale separation allows for the selective CTC isolation without the need for external/moving forces, and this can lead to shortened processing times and cost reduction (Ref. Reference Mach, Adeyiga and Di Carlo108). These systems can be based on several principles, including inertial focusing, size-based filtration and surfaces coated with EpCAM antibodies (Ref. Reference Myung and Hong109). The CTC-iChip, the ClearCell® FX system and the herringbone microfluidic chip are a few examples of such devices (Ref. Reference Cai, Deng, Wang, Zhu, Huang, Du, Wang, Yu, Liu, Yang, Wang, Wang, Ma, Huang, Zhou, Zou, Zhang, Huang, Li, Qin, Xu, Guo and Yu110).
Immunocapture detection methods
Immunocapture detection in CTCs makes use of unusual protein expression and genes that are not present in other components of blood to identify CTCs in a more specific and sensitive alternative to traditional CTC detection methods (Ref. Reference Liu, Li, Zhang and Guo111). The immunocapture method of isolating CTC can be done via methods such as immunomagnetic and microfluidic-based immunocapture. In microfluidic-based immunocapture, the geometry of the microfluidic device serves to maximize interactions between CTCs and the associated ligands that have been functionalized on the surface of the chip. Thege et al. built a microfluidic device with the purpose of capturing pancreatic CTCs using immunocapture, where a combination of anti- MUC1 and anti-EPCAM antibodies was used. Results showed efficient capture of the CTCs while also extending immunocapture past the use of single markers (Ref. Reference Thege, Lannin, Saha, Tsai, Kochman, Hollingsworth, Rhim and Kirby112). Immunomagnetic capture is carried out by negative or positive selection (Ref. Reference Rushton, Nteliopoulos, Shaw and Coombes113).
Positive selection involves coating antibodies that bind to the target cells with magnetic particles, leading to greater sensitivity and specificity. The target cells attach to the antibodies on the magnetic beads when a sample is combined with these particles. The target cells that are bead-bound are then magnetically separated from the remaining sample. Now that the target cells are adhered to the magnetic beads, they can be gathered and subjected to more analysis (Ref. Reference Li, Li, Shao, Li, Zhao and Ye79). Techniques such as CellSearch©, MACS, Magsweeper and Immunomagnetic separation (IMS) are all used in positive selection (Ref. Reference Rushton, Nteliopoulos, Shaw and Coombes113). Figure 2 provides a graphical summary of the positive selection techniques.
Role of EpCAM in CTC-driven metastasis (Created with BioRender.com).
In negative selection, magnetic particles are coated with antibodies that bind to undesirable cell types while sparing the target cell. The undesired cells attach to the magnetic beads when the sample is combined with these particles. The target cells are then left in the supernatant after the undesirable cells that were linked to the beads are removed using a magnet. It is possible to gather the unaltered target cells for later application (Ref. Reference Edd, Mishra, Smith, Kapur, Maheswaran, Haber and Toner76), techniques such as EasySep and Rosette Sep have been applied here (Refs. Reference Rushton, Nteliopoulos, Shaw and Coombes113, Reference Paterlini-Bréchot114). Multiple studies have shown that around 50-70% of patients with metastatic breast, prostate and colon cancers display high levels of CTCs when analysed using immunogenetic assay-based CellSearch® (Ref. Reference Hayes and Smerage115). Figure 3 summarizes the detection and isolation methods of CTCs. Table 1 describes popular devices employed in the detection/isolation of CTCs alongside their main findings.
Process overview of different immunomagnetic positive selection techniques for CTC capture. Note: ab, antibodies (Created with BioRender.com)
Overview of selected devices and techniques employed in CTC detection and isolation
Single-cell detection methods
Despite their great versatility, these cell enrichment-based methods still exhibit compromised efficiency in CTC isolation, limiting their detection sensitivity. Moreover, these methods typically rely on dedicated chip fabrication, tedious operations, or costly equipment, restricting their use in a routine diagnostic test. New strategies with high time- or cost-efficiency and single-cell sensitivity are critical for the implementation of CTC detection in the clinic, particularly for cancer screening (Ref. Reference Chen, Chen, Deng, Li and Jiang116).
The DEPArray technology is a microchip-based digital sorter, combining both microfluidic and microelectronic image-based isolation of single CTCs. It operates on the principle of dielectrophoresis, whereby non-uniform alternating electric fields generated by an array of over 300,000 programmable micro-electrodes create stable “DEP cages”, which are capable of trapping and levitating individual cells without physical contact. Cells loaded into the device’s chamber are randomly distributed in these cages and visualized using high-resolution bright field and multi-channel fluorescence microscopy, which enables single-cell level morphological and phenotypic analysis (Ref. Reference Di Trapani, Manaresi and Medoro117).
Micromanipulation involves the manual picking of stained CTCs using a microinjector, which is controlled by a micromanipulator, and the whole process is visualized using an inverted microscope. As it is a manual process, it can be time consuming; however it allows for the disassociation of CTC clusters and CTC-WBC clusters into single cells to allow for subsequent single-cell analysis of intra-cluster heterogeneity. Automated adaptations of micromanipulation, such as the CellCollector system, employ high-precision glass micro-capillaries mounted on robotic arms to improve both efficiency and reproductivity, providing a more standardized alternative to fully manual approaches (Ref. Reference Rushton, Nteliopoulos, Shaw and Coombes113).
Magnetic beads conjugated with CTC-targeting antibodies are added to blood/plasma samples that contain different tumour cell types. Bound CTCs are then separated using methods such as CellSearch, where anti-EpCAM antibodies are conjugated to magnetic beads, which are then stained with DAPI for easier visualization. In MACS, cells are magnetically labelled and a magnetic sheath is used to filter out the non-residual CTC population. In MagSweeper, CTCs are labelled with magnetic beads, which are then captured by washing to remove loosely bound cells. In immunomagnetic separation, immunomagnetic beads are formed, which bind to CTCs, magnetic separation helps differentiate the CTCs, before elution of purified CTCs using buffers.
Therapeutic strategies targeting CTCs
CTCs are linked with poor prognosis in multiple cancers, due to their ability to evade immune surveillance and promote metastasis (Ref. Reference Gostomczyk, Marsool, Tayyab, Pandey, Borowczak, Macome, Chacon, Dave, Maniewski and Szylberg153). Early diagnosis of CTCs, along with their timely elimination, may increase the overall survival (OS) of patients (Ref. Reference Kim, Yoo, Jeong and Choi154).
Conventional approaches
Surgical resection remains the cornerstone for early-stage cancers and can reduce both the possibility of CTC formation and the progression of metastasis. However, in certain cancers, such as hepatocellular carcinoma (HCC), it was observed that surgical resection caused an uptick in CTC populations (Refs. Reference Tohme, Simmons and Tsung155–Reference Yu, Xiao, Dong, Liang, Zhang, Zhang, Huang, Chen, Zhang, Luo, Chen and Chen157). Both CTC detection and incidence showed increases postoperatively, with CTC counts increasing in 41.7% of patients, decreasing in 25.2% and remaining unchanged in 33.1% of patients’ post-surgery. Evidence suggests that in early-stage HCC, tumour cells tend to spread through the portal venous system and can be driven into the bloodstream via hepatic vein tumour thrombi during surgery. Refinements, such as the “no-touch” technique, which aims to minimize tumour handling, showed no post-operative CTC increase in 5 patients; however, further investigation is required (Ref. Reference Yu, Xiao, Dong, Liang, Zhang, Zhang, Huang, Chen, Zhang, Luo, Chen and Chen157). Similarly, Tamminga et al. observed that surgical resection led to paradoxical metastasis; however increase in CTC volume was not observed. Suggesting that surgery may alter CTC biology in ways other than increasing CTC count (Ref. Reference Tamminga, de Wit, van de Wauwer, van den Bos, Swennenhuis, Klinkenberg, Jeroen Hiltermann, Andree, Spierings, Lansdorp, van den Berg, Timens, Terstappen and Groen158).
Chemotherapeutics have been found to reduce the total volume of CTCs, along with preventing early metastasis (Ref. Reference Gemenetzis, Groot, Yu, Ding, Teinor, Javed, Wood, Burkhart, Cameron, Makary, Weiss, He and Wolfgang159). Chemotherapy was found to reduce CTC counts in 15/30 of the patients treated. However, chemotherapy was also found to activate the CTCs’ phenotypic switch through cytotoxic stress, activating partial-EMT and rendering surviving CTCs more resistant to conventional therapies while simultaneously making them more responsive to tumour environmental stimuli, increasing the chances of distant organ metastasis (Refs. Reference Gostomczyk, Marsool, Tayyab, Pandey, Borowczak, Macome, Chacon, Dave, Maniewski and Szylberg153, Reference D’Alterio, Scala, Sozzi, Roz and Bertolini160, Reference Menyailo, Tretyakova and Denisov161). Since partial-EMT CTCs show resistance towards conventional chemotherapy, this method was found to result in worse prognoses and lowered treatment efficiency over time (Ref. Reference Gostomczyk, Marsool, Tayyab, Pandey, Borowczak, Macome, Chacon, Dave, Maniewski and Szylberg153).
Immunotherapeutic approaches
Many CTCs and other distant tumour cells express surface markers such as PD-L1 and CD47, which assist in escaping immune surveillance, as stated earlier (Ref. Reference Lian, Xie, Ye, Lu, Cheng, Xie, Li and Jia162). These protein-ligand interactions are responsible for the suppression of tumours along with the impairment of an antitumor response. The blocking of PD-L1 makes CTCs more susceptible to immunotherapy, making PD-L1 a favourable therapeutic target (Ref. Reference Strati, Koutsodontis, Papaxoinis, Angelidis, Zavridou, Economopoulou, Kotsantis, Avgeris, Mazel, Perisanidis, Sasaki, Alix-Panabières, Lianidou and Psyrri163).
CTCs use platelets to withstand shear forces, shield themselves from the immune system and limit the presentation of antigens on CTC surfaces (Ref. Reference Egan, Cooke and Kenny71). Platelets stick to sites of injury sustained by vascular endothelial cells. Here, the platelets release microparticles that promote anti PD-L1 binding to CTCs, effectively blocking PD-L1 on tumour cells and inhibiting metastasis (Ref. Reference Zhong, Zhang, Zhu, Liang, Yuan, Li, Li, Li, Jia, He, Zhu, Sun, Jiang, Zhang, Wang and Ke12). Activation of platelets can induce the recruitment of other immune cells, which mount a strong anti-tumour response after the initial PD-L1 blockade in preclinical models (Ref. Reference Wang, Sun, Ye, Hu, Bomba and Gu164). However, despite promise in preclinical models, clinical trials involving anti-PDL1 have demonstrated a large range of response rates (10-30%) in tumours with no selective biomarker. Patients who have PD-L1 high CTCs before treatment had better rates of response to anti-PD-L1 mAbs as well as longer progression-free survival (PFS) and overall survival (OS) rates. Suggesting that CTC profiling may serve as a predictive biomarker for immunotherapy benefit (Ref. Reference Strati, Economopoulou, Lianidou and Psyrri165). Clinical adoption does face challenges, however, with PD-L1 expression and other biomarkers on CTCs being highly heterogenous, and off-target toxicities caused by checkpoint inhibition (Ref. Reference Goodman, Jung, Balko and Johnson166).
Monoclonal antibodies have become an important therapeutic option in the treatment of multiple cancers (Ref. Reference Gostomczyk, Marsool, Tayyab, Pandey, Borowczak, Macome, Chacon, Dave, Maniewski and Szylberg153). Intracellular adhesion molecule-1 (ICAM1) levels have been found to be higher in clusters of CTCs when compared to single CTCs. Although clusters of CTCs are rare, they are associated with poor clinical outcomes and offer 20 to 100 times greater metastatic potential than single CTCs. The overexpression of ICAM1 has been found to promote metastasis in the lungs and is correlated with poor survival in breast cancer patients. In orthotopic models, it was observed that anti-ICAM1 antibodies were able to inhibit the aggregation of CTCs, while reducing the spontaneous metastasis in the lungs, prompting further investigation (Ref. Reference Taftaf, Liu, Singh, Jia, Dashzeveg, Hoffmann, El-Shennawy, Ramos, Adorno-Cruz, Schuster, Scholten, Patel, Zhang, Davis, Reduzzi, Cao, D’Amico, Shen, Cristofanilli, Muller, Varadan and Liu167). However, ICAM1 targeting largely remains preclinical, with no major clinical study addressing this approach. Additionally, the heterogenous expression of ICAM1 across tumour subtypes (only present in 0.3% of HCCs) presents a significant challenge (Ref. Reference Agnoletto, Corrà, Minotti, Baldassari, Crudele, Cook, Di Leva, D’Adamo, Gasparini and Volinia31).
Nanotechnology-based approaches
Nanotechnology has become an essential biomedical science, presenting creative solutions for tumour treatment, specifically in the realm of targeted drug delivery systems (Ref. Reference Patra, Das, Fraceto, Campos, Rodriguez-Torres, Acosta-Torres, Diaz-Torres, Grillo, Swamy, Sharma, Habtemariam and Shin168). The high tumour vasculature retention and permeability in the TME allows nanoparticles and macromolecules to have a relatively long half-life (Ref. Reference Zhang and King169). Through the manipulation of parameters like charge, molecular mass and particle size, nanocarriers allow therapeutic drugs to be delivered precisely to tumour areas with minimal damage to healthy tissues (Ref. Reference Batool, Sohail, Din, Alamri, Alqahtani, Alshahrani, Alshehri and Choi170). Breakthroughs in tumour-targeted systems, from overcoming physiological obstacles to bringing innovative diagnostic and therapeutic approaches to extend life expectancy, are made possible with the use of nanomaterials and techniques (Ref. Reference Prasad, Buragohain, Ghosh, Kumar and Chakraborti171).
Dual-targeting nanoparticles were designed to neutralize CTCs in the bloodstream and prevent metastasis (Ref. Reference Biswas, Sahu, Das, Das and Pathak172). Nanoparticles have displayed the ability to capture CTCs under shear stress from blood vessels and were effective in targeting and neutralizing CTCs in a lung metastasis mouse model (Ref. Reference Li, Sharkey, Huang and King173). EpCAM-targeting aptamer and tumour neovessel-targetable ligands were combined to create drug-loaded nanoplatforms for nanoparticle creation (Ref. Reference Fan, Tao, Zhai, Zhu, Li, Chen, Dong, Yang and Lv174). The anti-EpCAM ligands, Ep23 aptamer and K237 peptide, were loaded onto a single nanoplatform through the use of the drug Paclitaxel, allowing for the neutralization of CTCs while simultaneously damaging the primary tumour (Ref. Reference Bhattacharjee175). Dual targeting nanoparticles showed enhanced cellular uptake, elevated levels of apoptosis induction and higher levels of cell viability inhibition in vitro, demonstrating their effectiveness. From a translational perspective, dual-targeting NPs are a step towards precision therapy and allow for customization based on CTC profiles, allowing for patient-specific ligand combinations (Ref. Reference Velpula and Buddolla176). However, large-scale synthesis and reproducibility of multi-ligand NPs pose significant technical barriers, with off-target binding and systemic toxicity remaining concerns (Ref. Reference Minh Hoang, Nguyen and Tran177).
Tumour-specific apoptosis-inducing ligand (TRAIL) was coupled with activated platelet membrane-coated silica nanoparticles. Silicon (Si) nanoparticles coated in platelet membranes express essential platelet surface glycans and proteins. This allows for efficient targeting of CTCs and avoidance of phagocytosis (Refs. Reference Gay and Felding-Habermann178, Reference Li, Ai, Wang, Bu, Sharkey, Wu, Wun, Roy, Shen and King179). When conjugated with TRAIL, these particles selectively induce apoptosis in CTCs and reduce metastatic lesions in murine models (Ref. Reference Huang and Wang180).
Building on this concept, synthesized platelet-membrane-coated nanovesicles (PM-NVs) were loaded with Doxorubicin (DOX) and TRAIL. These nanovesicles effectively accumulated at the tumour site, enabling TRAIL to be delivered to cancer cells (Ref. Reference Hu, Sun, Qian, Wang, Bomba and Gu181). Additionally, they developed platelet membrane-coated glucan nanoparticles (NPs) loaded with bortezomib and tissue plasminogen activator (tPA), which targeted haematological malignancies. This was achieved by taking advantage of the unique affinity between P-selectin, which is present on the membrane surface of platelets and CD44, highly expressed on malignant cells and the bone marrow niche (Ref. Reference Hu, Qian, Sun, Wang, Chen, Bomba, Xin, Shen and Gu182). This nanomedicine prevented the formation of multiple myelomas, decreased off-target effects, increased the availability of the drug at the myeloma site, and simultaneously eliminated thrombotic repercussions in preclinical models (Ref. Reference Huang and Wang180).
While these biomimetic platforms demonstrate strong potential for CTC treatment, challenges to translating preclinical results to clinical trials remain. The anti-cancer efficacy and safety have still not been fully established in humans. Additionally, the extrusion method for biomimetic NP synthesis is still hard to scale and quite time-consuming, while reproducibility still remains a challenge (Ref. Reference Hu, Huang, Liu, Shen, Wu and He183).
Small molecule therapeutics
Small-molecule inhibitors represent a major class of targeted anticancer therapeutics designed to modulate specific molecular dependencies that drive the initiation, progression and metastasis of tumours. Many cancers exhibit a reliance on dysregulated pathways governing proliferation, survival, metabolic adaptation, immune evasion and apoptosis (Ref. Reference Fouad and Aanei184), which renders them susceptible to therapeutic treatments. Based on the nature of their molecular targets, selective small-molecule inhibitors can broadly be categorized into kinase and non-kinase inhibitors (Ref. Reference Liu, Chen, Zhang, Ma and Shi185).
Selective small-molecule kinase inhibitors have been explored for directly targeting CTC survival and metastatic ability. Recent work using lung cancer CTC cluster models demonstrated that CTC clusters exhibit increased resistance to epidermal growth factor receptor tyrosine kinase inhibitors (EGFR-TKIs) when compared to single or adherent tumour cells. Transcriptomic and functional analysis revealed that activation of the HGF/c-Met signalling axis is an important mediator for the survival of CTCs and their therapeutic resistance. The inhibition of c-Met using the small-molecule kinase inhibitor tivantinib suppressed the survival of lung cancer CTC clusters, while simultaneously reducing the CTC cluster-driven metastasis in vivo (Ref. Reference Zhang, Lu, Tian, Li, Liu, Liu, Luo, Yao, Liu, Wang, Yao, Yang, Shangguan and Que186).
Among non-kinase inhibitors, digoxin has emerged as a drug that targets metastasis-relevant mechanisms. Digoxin is a cardiac glycoside that inhibits the Na+/K+-ATPase, a membrane ion transporter that has been found to be involved in cell-cell adhesion and collective tumour cell behaviour. Preclinical studies demonstrated that inhibition of Na+/K+-ATPase disrupts CTC clusters by weakening their intracellular adhesion, leading to cluster disassociation and reduced metastatic seeding. These findings were translated into a first-in-human proof-of-concept study in patients with metastatic breast cancer, where digoxin treatment at 0.125-0.25 mg per day resulted in a significant reduction in mean CTC cluster size (Ref. Reference Kurzeder, Nguyen-Sträuli, Krol, Ring, Castro-Giner, Nüesch, Asawa, Zhang, Budinjas, Gvozdenovic, Vogel, Kohler, Grašič Kuhar, Schwab, Heinzelmann-Schwarz, Weber, Rochlitz, Vorburger, Frauchiger-Heuer, Witzel, Wicki, Kuster, Vetter and Aceto53).
In vitro CTC-derived models for drug testing and mechanistic studies
An important aspect of CTC research is to study the mechanistic basis behind their properties and explore new CTC-based biomarkers and targeting strategies. The generation of in vitro models, such as CTC-derived cell lines, at relevant time periods of disease progression is crucial to both the functional and longitudinal characterization of these cells. Additionally, the demand for large-scale testing has promoted the development of tumour organoid-on-chips (Refs. Reference He, Zhang, Ozkan, Feng, Duan, Wang, Yang, Xie and Liu187, Reference Tayoun, Faugeroux, Oulhen, Aberlenc, Pawlikowska and Farace188).
CTC in vitro-derived models have been established across multiple tumour types, such as breast, prostate, lung, colorectal and pancreatic cancer. These models primarily consist of short-term CTC cultures and CTC-derived cell lines, which are generated under defined culture conditions. While the fragility and rarity of CTCs make this technically challenging, successful in vitro expansion enables direct functional analysis of these cells. Additionally, CTC-derived cell lines often preserve key biological characteristics of the originating tumours, such as mutations and phenotypic plasticity (Ref. Reference Kahounová, Pícková, Drápela, Bouchal, Szczyrbová, Navrátil and Souček189).
Beyond conventional CTC-derived cell lines, three-dimensional CTC-derived organoids have emerged as more advanced in vitro systems that accurately capture the structural, phenotypic and functional heterogeneity of metastasis. CTC organoids preserve key features such as epithelial-mesenchymal plasticity, stem-like states and therapy-resistant subpopulations that cannot be accurately portrayed in 2D cultures. These organoids also allow for functional analysis of metastatic ability, drug response and signalling responses in an environment more similar to its true biological state (Ref. Reference Li, Deng, Li, Wu, Lu and Liang190). Building on these advances, microfluidic chip-based organoid cultures can stimulate the complex in vivo pathophysiological environment by providing precise control over the biochemical, physical and nutritional microenvironment. This enables more accurate drug testing and tumour modelling. Supported by microfluidic technology, CTC-derived organoids become highly dynamic and reproducible cancer models (Refs. Reference Li, Deng, Li, Wu, Lu and Liang190, Reference Huang, Xu, Wang, Yi, Xi, Si, Zhang, Xiang, Rong, Yuan and Wang191).
The use of CTCs in precision medicine has also been gaining momentum in recent years, with its use in adjuvant/ neoadjuvant therapies, the detection of recurrent diseases and its role as a possible pharmacodynamic biomarker. Advancements in this field could lead to the development of new personalized therapies.
Challenges and future scope
Challenges
CTCs have demonstrated great potential in the realm of cancer prognosis, diagnosis, monitoring and treatment. Identifying and examining these cells released from primary tumours provides valuable insights into the disease dynamics of cancer, thereby offering directions for tailored treatment plans. However, there are numerous challenges involved in detecting and isolating CTCs.
The role of CTCs may differ in different clinical conditions, due to which its roles in various cancers must be carefully studied. There is an increasing number of inconsistent results being reported in relation to the links between the observed levels in CTCs and the survival rate (Ref. Reference Nagaiah and Abraham192). A lack of a standard protocol is also lacking with CTC-cantered prognostic or treatment modules, and currently, the only FDA protocol is the CellSearch© system (Ref. Reference Hong and Zu193).
Another challenge regarding CTC detection is its scarcity in the bloodstream, as approximately only 1-100 CTCs, along with 106-108 RBCs, can be found per mL of blood (Ref. Reference Pantel and Alix-Panabières194).
Biological questions, such as the nature of the factors determining the tendency of CTC to metastasize still haven’t been answered, along with which pathways could be targeted for inhibition to slow CTC metastasis. Regulation of CTCs by the circadian rhythm, due to which their distribution may not be uniform, is also a cause for concern (Ref. Reference Ju, Chen, Zhang, Xu, Zhang, Li, Chen, Zhou, Ji and Wang74).
Many current studies focused on CTCs as a biomarker usually only investigate advanced or metastatic cancers, while rarely discussing early-stage cancers. It is important to understand if CTCs can reliably be used as a possible biomarker for early-stage cancers as well, as this can improve the prognosis of patients diagnosed with highly metastatic cancers, such as glioblastoma (Ref. Reference Ju, Chen, Zhang, Xu, Zhang, Li, Chen, Zhou, Ji and Wang74).
Future directions
Due to the scarcity of CTCs in early-stage cancers, maximizing the sensitivity of CTC detecting devices proves to be the most favourable option. Emerging approaches that utilize microfluidic devices, nanostructured capture surfaces and multiplex isolation strategies can aid in detection/isolation, provided that cell loss is accounted for. Correcting this issue or minimizing its effect could lead to the standardization of CTC detection technologies, allowing for more reproducible results and data comparability, which remains a major hurdle to clinical adoption (Ref. Reference Lawrence, Watters, Davies, Pantel and Lu195).
Enhancing the molecular characterization of CTCs may assist in their detection/isolation. Distinguishing between dormant and proliferative CTCs or single cells versus clusters could prove to provide major insights into CTC metastatic risk. This can be achieved through the identification of specific markers for CTC clusters or RNA sequencing of clusters of CTCs to identify the immune cells that interact with specific CTC clusters, and thereby quantify the effect they have on the cellular pathways, which are potential therapeutic targets (Ref. Reference Brechbuhl, Vinod-Paul, Gillen, Kopin, Gibney, Elias, Hayashi, Sartorius and Kabos196).
Future clinical utility will also depend on multi-analyte liquid biopsies, where CTCs and their associated biomarkers, such as circulating proteins, microRNAs and cell free tumour DNA (ctDNA), are analysed and targeted. Such assays have already shown preliminary success in increasing CTC sensitivity in both triple-negative breast cancer and prostate cancer (Refs. Reference Lawrence, Watters, Davies, Pantel and Lu195, Reference Yang, Bai, Hu, Kong, Li, Zhao, Feng, Cheng, Shou, Zhang and Zhang197).
These advances suggest that the future of CTC research does not rely on a single platform, but on integrated, standardized approaches that enable both sensitive detection and accurate biological interpretation.
Conclusion
The increased diversity of tumour-associated cell populations complicates processes associated with disease diagnosis and treatment for various cancers. Key populations, such as CTCs and CSCs, remain areas of interest. CTCs are particularly important, owing to their association with poor patient prognosis, increased recurrence rates, increased likelihood of metastasis and drivers of immune evasion, amongst other functions. Techniques for CTC detection are hampered by their lack of wide applicability in clinical settings. However, advancements in this area, such as the decreased scale of operations and lowered cost due to the implementation of microfluidics and nanotechnology, have shown promising results. Isolation methods suffer from similar issues such as a lack of specificity and the sparse nature of CTCs, with size-based filtration and immunocapture being the primary technologies used in this area. Modifications that show improved clinical applicability, efficacy and scalability are purported to be essential for further improvements. Therapeutics targeting CTCs offer scope to improve overall survival, reduce chances of recurrence and metastasis, as well as improving patient quality of life post-treatment, as they can act as supplementary treatment modalities to existing chemotherapy modules, whose cytotoxicity can prove to be detrimental in cancers where the median age of diagnosis is higher. Further understanding of CTCs, related cellular mechanisms and effects on the tumour landscape could allow for their use in diagnosing early-stage cancers, enhanced molecular characterization of tumours. This could also aid in improving attempts at detection and isolation, while also improving treatment efficacy by utilizing them to identify and target other effective biomarkers in various cancers.
Author contribution
Conceptualization – Praveen Kumar; Literature search and organization – Parth Agarwal, Rachana Raman and Manasa Bhagwat; Figures – Parth Agarwal and Rachana Raman; Writing original draft – Parth Agarwal, Rachana Raman and Manasa Bhagwat; Writing revisions and editing- Parth Agarwal and Rachana Raman, Final review and editing – Prasoon Agarwal and Praveen Kumar
Funding statement
This research received no external funding.
Competing interests
The authors declare no conflict of interest.
Open access
