The aim of this review is to summarize recent advances regarding circulating tumor cells (CTCs), with a focus on their phenotype and plasticity. Moreover, we aim to shed light on the diagnostic properties of circulating tumor cells and cancer stem cells.

Cancer is one of the deadliest diseases affecting the human population, causing millions of deaths each year. In 2022 alone, there were 20 million newly reported cancer cases worldwide, with 9.7 million deaths. Breast carcinoma is the most common type of cancer in women (2.3 million new cases each year), while lung (1.5 million new cases each year) and prostate (1.4 million new cases each year) cancers are the most prevalent malignancies in men. Among both sexes, lung cancer is the most frequently diagnosed carcinoma with 2.5 million new cases each year [1–3].

Cancer is a disease caused by multiple mutations in a cell, leading to an altered cellular state. It is characterized by abnormal growth, spread, resource consumption, tissue disruption, and impairment of normal bodily functions. Environmental factors, viruses, bacteria, chemical agents, or radiation exposure can all contribute to cancer development [4, 5].

To fight an effective battle against cancer, understanding the disease, its progression, and developing new progression targeting therapeutic techniques is of utmost importance.

Our workgroup has previously conducted examinations of circulating tumor cells (CTCs) and CTC clusters. Using magnetic cell separation, we successfully detected cytokeratin (CK)-positive CTCs and CTC clusters in the blood of colorectal cancer patients. Additionally, our workgroup found cytokeratin positive cells in interaction with cytokeratin negative cells when investigating CTC clusters. This was the first time this was observed in colorectal carcinoma (CRC) patients. Moreover, we also observed that chemotherapy reduces the number of CTCs and clusters in the blood but does not eliminate them [6]. In another of our studies, we found that the higher the number of single CTCs in the circulation, the higher the number of epithelial cells in CTC clusters [5]. In the same study, we concluded that the number of CTC singlets, doublets, and clusters correlates with cytokeratin 20 (CK20) qPCR results from the blood of CRC patients [7]. Moreover, we have performed several liquid biopsy-based analyses on the blood of colorectal cancer patients to investigate the potential diagnostic and therapeutic implications of cell-free nucleic acids. We found that the level of cfDNA was higher in patients with non-metastatic CRC and metastatic CRC compared to individuals with remission or stable disease [8, 9].

In this review, we aim to gather the most recent information on CTCs. Furthermore, we seek to explore their unique plasticity and highlight the significance of CK + epithelial CTC clusters in circulation. Additionally, we provide an overview of the most up-to-date techniques for CTC detection, analysis, and their relation to therapy decisions.

The most lethal feature of cancer is metastasis—a process involving the invasion of distant parts of the body by cancer cells that “break away” from the primary tumor and enter the circulation. These cells are referred to as circulating tumor cells (CTCs). CTCs can travel through the bloodstream either as single cells or in clusters. CTC clusters are defined as groups of two or more CTCs with stable cell–cell junctions. Although clusters represent only a minority of CTCs found in circulation, they possess a higher metastatic potential than single CTCs. Moreover, in several cancer types, the presence of CTC clusters indicates a worse clinical outcome compared to single CTCs [10, 11]. It has been shown that in non-small cell lung cancer (NSCLC), the prevalence of CTC clusters increases with advanced cancer stages. However, no correlation was observed between the number of CTC clusters and the tumor type or stage in lung cancer indicating that cluster number may not distinguish between the most advanced disease stages. However, correlation between CTC number and prognosis was found in a meta-analysis which considered the presence of CTC but not their number or phenotype characterization [12–14].

Other than the blood stream, CTCs can also enter into the lymphatic circulation, where they can reach local lymph nodes and differentiate leading to metastases. Lymph-specific CTCs are usually non-immunogenic so they can avoid detection by the immune system, especially by cytotoxic T cells which helps them in their metastasis initiation [15–17].

Additionally, CTCs are also capable of perineural invasion (PNI), which is defined as an invasion in, around, and through the nerves. PNI is usually associated with poor clinical outcome and decreased survival in different cancer types including ductal adenocarcinoma, prostate cancer, gastric cancer, breast cancer, pancreatic ductal adenocarcinoma and colorectal carcinoma [18–20].

Since these cells are shed into the circulation, peripheral blood serves as an excellent source for the selection and analysis of CTCs. Over the past decade, multiple liquid biopsy techniques have been developed for CTC isolation and analysis. These methods can be categorized as either label-dependent or label-independent techniques. Label-dependent techniques rely on interactions between cell surface markers expressed on CTCs and specific antibodies. These antibodies can be fixed to the surface of magnetic particles or microfluidic chips to enable positive selection of CTCs from blood or negative depletion of white blood cells. These approaches typically target EpCAM, a surface protein commonly expressed on CTCs (Table 1). Amongst these techniques, currently the CellSearch system by Janssen Diagnostics is the only FDA approved method which utilizes EpCAM-coated ferrofluidic nanoparticles for CTC detection. Other commercially available label dependent methods are AdnaTest by Adnagen and MagSweeper by Illumina both of which are based on immunomagnetic capture of CTCs [21–23].

Label-independent detection methods, on the other hand, are based on the physical properties of CTCs, such as size. Using filters with defined pore sizes, the typically larger CTCs can be separated from smaller blood cells. Gradient centrifugation can also be employed, where lower-density cells such as erythrocytes and polymorphonuclear leukocytes settle at the bottom, while higher-density mononuclear leukocytes and CTCs remain in the upper layers. Overall, methods based on physical properties are cost-effective and preserve cell viability well. However, these techniques are often inefficient, yield low purity, and lack specificity. ISET by Rarecells diagnostics and Parylene filter by Circulogix are both based on filter based isolation and enrichment platforms available for label-free detection. Other techniques are also on the market such as RosetteSep by STEMCELL technologies and OncoQuick by Greiner BioOne which are based on density gradient separation [21–23].

CTCs carry information about the originating tumor, making them highly valuable for clinical applications. CTC analysis can be used for early tumor detection, enabling treatment at an earlier, more manageable stage. Usually, the number of CTCs in early disease are low roughly ∼1/10⁸ peripheral blood mononuclear cells (PBMC), while in metastatic cancers their number is much higher at 1/10⁵–10⁷ PBMCs. Moreover, the presence of CTCs in the circulation provides prognostic information, aids in predicting disease outcomes, and helps guide treatment decisions. Furthermore, molecular characterization and genome sequencing of CTCs can provide valuable insights for the development of personalized treatments [23–26].

As a few examples, Baek et al. used the fluid-assisted separation technique (FAST) to enrich CTCs from the blood of healthy donors and CRC patients. They found that CTC counts were significantly higher in CRC patients compared to healthy volunteers. Notably, all patients with stage 4 CRC were positive for CTCs [27]. Dalum et al. utilized the CellSearch system to analyze the blood of CRC patients before surgery, and reported that the presence of CTCs prior to the operation was associated with a significant decrease in recurrence-free survival [28].

Cristofanilli and colleagues, in their study of metastatic breast cancer patients, reported that the presence of more than 5 CTCs per 7.5 mL of blood was associated with shorter median progression-free survival and overall survival compared to patients with fewer than 5 CTCs [29]. According to a metaanalysis by Jin et al., the detection of CTCs in circulation is associated with poor prognosis in small cell lung cancer (SCLC) patients compared to those with non-small cell lung cancer. Moreover, they found that epithelial CTCs predict worse outcomes than mesenchymal CTCs in lung cancer patients [30].

EMT is a complex process involving many molecular and cellular changes, such as the downregulation of epithelial markers (e.g., cytokeratins, E-cadherin, and claudins) and the upregulation of mesenchymal proteins (e.g., vimentin, N-cadherin, and fibronectin), which increase the mobility and invasiveness of the cell. The changes observed during EMT are regulated by transcription factors known as EMT-inducing transcription factors (EMT-TFs), such as Snail-1, Snail-2 (Slug), ZEB1, and Twist (Figure 2) [40]. It is widely accepted that the process of EMT generates multiple hybrid phenotypes along the epithelial-mesenchymal axis, contributing to tumor heterogeneity. Both epithelial and mesenchymal states are believed to harbor limited metastatic potential; however, certain hybrid phenotypes can possess a higher degree of EMT plasticity, enabling them to survive and adapt to different microenvironments encountered during metastatic spread [40].

The process leading to metastasis is complex, involving several biological steps. First, metastatic cells must undergo EMT, detach from the primary tumor, invade the bloodstream, survive in circulation, disseminate into distant organs, extravasate, undergo MET, colonize, and form micrometastasis. Only a fraction of CTCs are capable of undergoing metastatic transformation; these cells are referred to in the literature as circulating cancer stem cells [41].

Balcik-Ercin et al. found in their colorectal carcinoma-derived CTC cell line, that the expression of SIX1, an EMT marker important for the mesenchymal profile, was downregulated. This suggests that tumor cells can utilize alternative pathways to activate genes that promote their plasticity and invasiveness. Furthermore, they found that the MET transcription factor GRHL2 was overexpressed in their CTC lines. GRHL2 may stabilize the epithelial-mesenchymal hybrid phenotype and support cell migration [38].

Seo et al. investigated the phenotypic heterogeneity of CTCs in SCLC using assays to characterize rare cells. In an EpCAM-targeted assay, utilizing a variety of biomarkers, they observed a wide range of CK and EpCAM expression in the CTC population. Their single-cell sequencing results reinforce the presence of tumor cell plasticity by indicating that a phenotypically heterogeneous population of cells can be genomically stable. Recent evidence suggests that cancer cells exhibit a hybrid mesenchymal and epithelial character, and this plasticity is associated with their metastatic ability and poor patient prognosis [42].

It has been described that the activation of the EMT program does not always result in a fully mesenchymal phenotype. It is likely that a partial EMT status is achieved in both non-transformed and cancer cells, the resulting hybrid cells carry both epithelial and mesenchymal markers. Moreover, these hybrid cells are more likely to acquire stemness [33]. Indeed, it has been shown that with EMT induction, breast cancer cells can acquire cancer stem cell markers, such as CD44 [43, 44].

The fact that stemness markers can be expressed by cells undergoing EMT opens the possibility that differentiated mesenchymal cells can also acquire stemness characteristics, leading to the formation of new mesenchymal stem cells.

There are multiple therapies that can be implemented to treat cancer, such as radiotherapy, surgery, and chemotherapy. However, cancer cells can develop resistance to chemotherapy, which is a major factor in therapy failure and poor patient survival [45, 46].

Due to the stress generated by the changing environment and therapy, genetic mutations occur in cancer cells, leading to cancer heterogeneity and, in turn, therapy resistance. Heterogeneity among patients due to environmental, somatic, and germline factors is called intertumoral, while uneven distributions of genetically diverse subpopulations of cancer cells in the same tumor are referred as intratumoral heterogeneity. Moreover, the differences between a primary tumor and the metastasis in a patient are also called intertumoral heterogeneity [47–49].

One of the factors contributing to intratumoral heterogeneity is the presence of cancer stem cells (CSCs), a subset of cancer cells possessing stem cell characteristics such as self-renewal and the ability to differentiate [50, 51]. CSCs were first identified in acute myeloid leukemia (AML) after transplanting isolated CD34+/CD38− cancer cells into non-obese diabetic/severe combined immunodeficient (NOD/SCID) mice. Since then, CSCs have been described in a variety of hematological and solid tumors, such as pancreatic, breast, and colon malignancies [52]. The origin of CSCs is highly debated, with multiple hypotheses suggesting that they arise from either adult stem cells, mutated adult progenitor cells, or cancer cells that gain stem-like properties through dedifferentiation. CSCs can be separated from normal stem cells via the expression of specific cell surface markers such as CD133, CD24, CD44, epithelial cell adhesion molecule (EpCAM), and CD200. Moreover, intracellular proteins have also been used as markers of CSCs, such as aldehyde dehydrogenase 1 (ALDH1) (Figure 2) [53–55].

There’s a connection between the previously mentioned EMT and cancer stemness. The expression of EMT inducing transcription factors such as ZEB1, SNAIL1 and 2 by cancer cells initiates the expression of stem cell markers SOX2, BMI1 and OCT4. It is described that mesenchymal and stemness traits, characterise cancer stem cells within the tumor mass. This indicates that CSCs have specific abilities similar to embryonic stem cells [56].

Chemo- and radiotherapy induce DNA damage in differentiated cancer cells, which in turn leads to therapy-induced senescence (TIS). Senescence is a cell state characterized by prolonged cell-cycle arrest, enhanced secretory capacity, macromolecular damage, and altered metabolism. The main defining characteristic of senescence is stable growth arrest, which ensures that damaged or transformed cells do not preserve and perpetuate their genomes. During this process, specific molecular markers are activated, such as p16INK4a/Rb and p53/p21CIP1. Senescence also has physiological roles; the process is triggered in response to damage and allows the suppression of potentially dysfunctional, transformed, or aged cells. However, the aberrant accumulation of senescent cells during aging has potential detrimental effects, such as contributing to renal dysfunction and type II diabetes (Figure 2) [57].

The senescent state of cancer cells can be beneficial as these cells induce inflammation and attract immune cells, which clear the senescent cancer cells. One of the main characteristics of senescent cells is the senescence-associated secretory phenotype (SASP). They secrete interleukins and other ligands that can negatively affect cancer initiation and progression.

However, the previously mentioned TIS also induces cancer remodeling and promotes CSC generation. Moreover, senescent tumor cells can cause changes in the tumor microenvironment, further promoting cancer development. SASP can also provide a positive environment for tumor progression. It was shown that SASP components can promote cancer cell growth, invasion, metastasis, and tumor vascularization [54, 58, 59].

Cancer cells can escape the senescent state through the acquisition of genetic and epigenetic features, which make them plastic. Additionally, via the paracrine action of SASP, cells in close proximity to tumor cells can be imparted with tumorigenic capacities. Senescence escape and cellular reprogramming via SASP are essential components of epithelial tumor progression. Tumor cells achieve the previously mentioned plasticity through the initiation of EMT [60].

Liquid biopsy-based monitoring of cancer is a promising, non-invasive method which usually involves blood or urine collection, followed by the analysis of extracellular vesicles, circulating tumor cells (CTCs), or circulating tumor DNA (ctDNA) [66].

ctDNA is a form of nucleic acid released mainly from apoptotic or necrotic tumor cells into the circulation. In the peripheral blood, ctDNA circulates in the form of nucleosomes, which can be isolated, and their genetic and epigenetic properties can be analyzed to provide information about the originating tumor (Table 1) [67].

A few examples are listed below for the utilization and shortcomings of CTCs and ctDNA in the diagnosis of different epithelial cancers. Both CTCs and ctDNA can be used in the early detection of colorectal cancer and can be used in prognosis and treatment response monitoring, as well [68]. However, the level of CTCs is usually low in CRC patients, especially in the early phase of the disease. On the other hand, ctDNA can be detected more easily and provide real-time molecular information to monitor treatment response and relapse [69].

In early stage breast cancer, CTCs are present in low numbers and difficult to analyze, while ctDNA is more readily detectable and useful for monitoring tumor response, drug resistance, and mutations. In metastatic breast cancer, ctDNA efficiently tracks treatment response and tumor heterogeneity, whereas elevated CTC levels serve as prognostic markers [70].

CTCs are more common in small cell lung cancer (SCLC) than in non-small cell lung cancer (NSCLC) [71]. Despite this fact, in NSCLC, CTCs provide more informative mutation detection than ctDNA because of more sensitive genotyping [72]. However, as mentioned before CTC counts are highest in SCLC due to rapid tumor growth and early spread, making them better prognostic markers than ctDNA in this subtype [73]. In NSCLC, both CTCs and ctDNA can serve as diagnostic, prognostic, and therapeutic monitoring tools [74].

Our workgroup previously carried out experiments where with high sensitivity we detected the septin 9 gene (SEPT9) from circulation which is an excellent marker of CRC [75]. Moreover in a separate study we also detected that compared to healthy tissue, SEPT9 is hypermethylated in adenoma and CRC cells. Our results indicated that changes in the SEPT9 methylation reflects the cellular progression towards malignancy in the colon mucosa [76]. A list of biomarkers which can be detected with ctDNA analysis are shown on Table 2 with relevant mutations and associated cancers.

In a study, Kong et al. found mutations in CTCs and ctDNA that matched those in the primary tumor. They also discovered that the top mutated genes in CTCs and ctDNA had prognostic value when applied to existing cohorts of cancer [121].

Koyanagi et al. also found in their research that, in stage IV melanoma patients, the number of CTCs correlated with the methylation of ctDNA molecules [122]. Additionally, in the peripheral blood of breast cancer patients, the level of ctDNA correlated with the presence of CTCs. This potentially suggests that CTCs are a major source of ctDNA, or that high numbers of CTCs and ctDNA are both features of a more aggressive tumor [123]. This correlation between ctDNA and CTCs was also observed in another study. Furthermore, methylated ctDNA and CTCs correlated with aggressive tumor biology and advanced disease [124].

CTCs are a pivotal and critical point in the progression and understanding of cancer especially metastasis. Their presence in the circulation of the patient either as single cells or clusters highlights their important role in cancer dissemination. Due to their unique ability to mirror the genetic characteristics of the originating tumor CTCs provide valuable, minimally invasive means of accessing real-time information about the biology of the tumors. However, due to changes like EMT, CTCs can diverge phenotypically from the original tumor. As EMT usually activated in tumor cell sub-populations during dissemination CTCs carry the phenotypic information of the originating cell population. Despite, it has been demonstrated that CTCs carry prognostic and diagnostic utility in detecting MRD, guiding therapeutic decisions and monitoring relapse especially when utilized alongside cfDNA.

Looking forward, advances in CTC isolation and characterization techniques may provide a way for more precise and personalized therapy. With the integration of multi-omics approaches like single-cell sequencing and artificial intelligence researchers could further enhance the ability to profiling these rare cells and also offer deeper insights into the evolution of the tumor and its resistance to therapy.

Finally, leveraging CTCs in clinical practice holds promise for early detection and better monitoring and also for targeted treatment development which could transform and improve cancer care and ultimately patient outcomes. When used in conjunction with ctDNA CTCs can provide a more comprehensive view of tumor dynamics as ctDNA offers insight into genetic alterations while CTCs allow phenotypic and functional analyses. However, there are still challenges remain before CTC-based approaches can be fully utilized in routine clinical use. These include the extremely low abundance of CTCs in early-stage disease compared to ctDNA, limitations in current isolation and detection technologies, and the lack of standardized protocols across platforms. Furthermore, heterogeneity among CTCs, including variable expression of surface markers due to processes like EMT, can lead to false negatives and complicate interpretation. Addressing these technical and biological hurdles through continued innovation and validation in large clinical studies will be important for fully realizing the potential of CTCs in precision oncology.