The Different Facets of Liquid Biopsy: A Kaleidoscopic View

Abstract

The current limitations of cancer diagnosis and molecular profiling based on invasive tissue biopsies or clinical imaging have led to the development of the liquid biopsy field. Liquid biopsy includes the isolation of circulating tumor cells (CTCs), circulating free or tumor DNA (cfDNA or ctDNA), extracellular vesicles (EVs), and tumor-educated platelets (TEPs) from body fluid samples and their molecular characterization to identify biomarkers for early cancer diagnosis, prognosis, therapeutic prediction, and follow-up. These innovative biosources show similar features as the primary tumor from where they originated or interacted. This review describes the different technologies and methods used for processing these biosources as well as their main clinical applications with their advantages and limitations.

Cancer diagnosis faces several technological and clinical problems. In most cancer subtypes, tissue biopsy represents the “gold standard” for diagnosis and biomarker profiling. However, this is an invasive procedure. Moreover, the natural cancer course is characterized by the formation of distant metastases and the development of resistance against chemotherapy. These issues challenge the current approach because tissue biopsy only reflects the cancer features at the specific time of the intervention, and obtaining biopsies of metastatic tumors is humanly unviable, although theoretically possible. Medical imaging is a less invasive alternative for real-time follow-up, but it does not allow molecular profiling.

“Liquid biopsy” is becoming a viable alternative for real-time cancer follow-up in patients, and for assessing biomarkers that are usually tested only in tissue biopsies. Liquid biopsy is a minimally invasive procedure based on sampling blood, cerebrospinal fluid, urine, sputum, ascites, and theoretically any other body fluid. The term liquid biopsy was first used for circulating tumor cells (CTCs) in the bloodstream (Pantel and Alix-Panabières 2010), but it has broadened to include circulating free DNA (cfDNA), circulating tumor DNA (ctDNA), extracellular vesicles (EVs), and more recently, tumor-educated platelets (TEPs).

In this review, we describe the different circulating cells that can be analyzed with the liquid biopsy procedure, and their technical, biological, and clinical advantages and limitations.

CIRCULATING TUMOR CELLS

Cancer cell dissemination through the body and the metastatic cascade begin with the active release of the most aggressive tumor cells in the bloodstream and/or lymphatic vessels. These cells are called CTCs (Pantel and Alix-Panabières 2010). CTC enumeration and analysis give important information on the tumor molecular profile, and can lead to the identification of metastasis initiating cells (MICs). There is a direct correlation between CTC number in blood, the expression of specific biomarkers (e.g., cancer stem cell markers), and the formation of distant metastases (Kang and Pantel 2013; Zhang et al. 2013; Alix-Panabières et al. 2017). Therefore, CTC detection/analysis contributes to the early discovery of metastatic lesions and to precision oncology (e.g., prognostic information, patient stratification for targeted therapies, real-time monitoring of treatment efficacy, identification of therapeutic targets, and resistance mechanisms).

CTCs represent the ideal biosource for solid cancer characterization and monitoring because they can be used for genome, proteome, transcriptome, and secretome analyses. Although CTCs are rare in the bloodstream (Pantel and Alix-Panabières 2010), recent high-technology methods allow the detection and characterization of single CTCs (Pantel and Alix-Panabières 2019).

Technological Aspects

As CTC number in blood is limited, CTC detection is always combined with a first enrichment step based on the biological (e.g., expression of surface proteins/receptors) or physical (e.g., size, deformability, density, and electric charge) properties that allow distinguishing CTCs from the other cells in the blood (Pantel and Alix-Panabières 2019). Approaches based on the “biological features” rely on the different expression of membrane proteins between blood cells and CTCs for positive (expression of epithelial markers, such as epithelial cell adhesion molecule—EpCAM—in carcinoma cells) or negative enrichment (Allard et al. 2004; Schulze et al. 2013; Pantel and Alix-Panabières 2019). The CellSearch system (Menarini Silicon Biosystems) is the only method cleared by the United States Food and Drug Administration (FDA) for CTC analysis. In this system, CTCs are enriched based on EpCAM expression and then identified using anti-CK8, 18, and 19, and -CD45 antibodies and DAPI staining. Many other technologies are available (e.g., microfluidic devices [Dong et al. 2013] and intravascular wires [Saucedo-Zeni et al. 2012]) that use antibodies against EpCAM and/or other epithelial surface markers. Very recently, a temporary indwelling intravascular device to capture CTCs from bloodstream and then return erythrocytes and leukocytes to the normal circulation has been developed. This apparatus allows the continuous capture of CTCs in dogs for ∼2 h. It must now be tested in patients with cancer (Kim et al. 2019). Nevertheless, these methods could miss CTCs that undergo epithelial-to-mesenchymal transition (EMT) and that do not express EpCAM. This can be avoided by using negative enrichment methods in which the blood sample is depleted of blood cells, such as leukocytes, by immuno-targeting the CD45 marker that is expressed by all leucocytes but not by solid cancer cells (Iinuma et al. 2000; Bilkenroth et al. 2001).

Numerous marker-independent techniques based on CTC “physical properties” (density, size, deformability, and electric charge) have been developed as well. For example, CTCs, which are epithelial cells, should be bigger than leukocytes and erythrocytes, which have well-defined sizes. Therefore, cell size is used for CTC enrichment by different filtration systems (Hao et al. 2018). Label-free enrichment approaches avoid the biological bias linked to the variability of cell biomarker expression associated with tumor heterogeneity.

After positive/negative enrichment, CTCs need to be clearly differentiated from the remaining leukocytes and endothelial cells. This is usually performed using immunofluorescence methods, similar to the identification step in the CellSearch system. Nucleic acid-based methods can be an alternative to immunological assays. These methods use epithelial gene-specific primers, but their low specificity could lead to false positive results (Pantel et al. 2008; Fehm et al. 2009). Another CTC identification approach focuses on their viability by detecting proteins secreted, released or shed from cancer cells cultured for 24–48 h. This unique functional assay is called Epithelial ImmunoSPOT (EPISPOT) (Soler et al. 2017). A completely new microdroplet technology to detect viable CTCs at the single-cell level is currently under development. This optimized EPISPOT assay is called EPIDROP (Pantel and Alix-Panabières 2019).

After enrichment and identification, CTCs can be characterized at different level: (1) expression of specific surface markers, such as protein death ligand 1 (PD-L1; an immune checkpoint regulator and therapeutic target) (Mazel et al. 2015) and HER2 in breast and gastric carcinomas (Riethdorf et al. 2010; Jaeger et al. 2017); (2) whole-genome or transcriptome analysis in single cells; and (3) identification of metastasis-competent CTCs by in vitro expansion or injection of CTCs in mouse models (Pantel and Alix-Panabières 2019). Single CTCs can be analyzed manually by micromanipulation, or automatically by trapping and moving cells in dielectrophoretic cages (e.g., DEPArray) (Abonnenc et al. 2013).

Biology

Extensive research on the biology of cancer cell dissemination and metastasis formation has brought insights into how CTCs survive in bloodstream. First, EMT is a complex process characterized by down-regulation of epithelial proteins and up-regulation of mesenchymal proteins. This process supports migration of epithelial tumor cells and is thought to play a crucial role in promoting metastasis formation (Nieto et al. 2016). More recent studies have focused on the epithelial-to-mesenchymal plasticity of cells with stem cell characteristics. They showed that cancer epithelial cells may acquire some mesenchymal features and reverse this state in the colonized organs, under the influence of stemness effectors (Alonso-Alconada et al. 2014). Moreover, in normal tissue, loss of adhesion to the extracellular matrix, induces cell death in anchorage-dependent cells, a process called “anoikis” (Kim et al. 2012; Alix-Panabières and Pantel 2014). On the other hand, tumor cells that have acquired anoikis resistance can survive detachment from their primary site and can travel through the circulatory and lymphatic systems and reach ectopic locations (Simpson et al. 2008). Once in the circulation, CTCs face many stresses inherent to this new compartment, particularly the immune system attack. Much research has focused on understanding the immune-suppressive mechanisms that allow CTC escape from the immune system (Mohme et al. 2017). These works led to the discovery of some biomarkers, such as CD47, CTLA4, PD-L1, and PD1 (Steinert et al. 2014), and to the development of immune checkpoint inhibitor-based therapies, with remarkable clinical response in different malignancies (Ohaegbulam et al. 2015). In addition, cancer cells can associate with other cells and form microemboli to survive in the bloodstream environment. Different cell association mechanisms have been described: (1) CTC clusters that show higher metastatic potential through increased cell survival and reduced apoptosis (Aceto et al. 2014; Gkountela et al. 2019; Szczerba et al. 2019); (2) partnership with reactive platelets, used as a shield or camouflage against the immune system assault (Lou et al. 2015); and (3) clustering with white blood cells, mainly neutrophils, that might promote cell cycle progression, leading to more efficient metastasis formation (Szczerba et al. 2019).

Clinical Relevance

The clinical relevance of CTC detection for prognosis and outcome prediction was validated for metastatic breast (Cristofanilli et al. 2019), prostate (Scher et al. 2015), and colorectal cancer (Huang et al. 2015). However, despite the many clinical validation studies, CTCs have not been included in clinical guidelines yet. Although CTC enumeration can improve tumor staging and contribute to the early assessment of therapy effects, their clinical usage needs to be proved (i.e., their usefulness in decision-making concerning the adoption or rejection of a therapeutic action). Currently, two ongoing French interventional studies investigate the clinical usage of CTCs for stratification of patients with metastatic breast and prostate cancer: (1) STIC CTC METABREAST (NCT01710605), and (2) TACTIK (NCT03101046), respectively.

In the clinic, CTCs could be used also for (1) minimal residual disease (MRD) detection and characterization (Pantel and Alix-Panabières 2019), (2) early cancer diagnosis (Ilie et al. 2014; Castro et al. 2018), and (3) prediction of response/resistance to treatment (Mathew et al. 2015).

Moreover, the possibility of real-time monitoring of cancer progression by capturing CTCs in the bloodstream at the exact moment of metastasis initiation gives to this biosource a great potential as liquid biopsy. Finally, CTCs include also MICs that are at the origin of cancer relapse. The main goal is now to identify and specifically eradicate them (Cayrefourcq et al. 2015; Soler et al. 2018).

CIRCULATING TUMOR DNA

CtDNA is DNA that is actively secreted or/and originates from apoptotic or necrotic cancer cells and released directly in the bloodstream (or another biological fluid) (Elazezy and Joosse 2018; Jeppesen et al. 2019; Pantel and Alix-Panabières 2019). It represents only a small fraction of all circulating free DNA (cfDNA) because most cfDNA comes from normal cells in physiological conditions. Mandel and Metais (1948) were the first to describe nucleotide acids in blood in 1948, and Stroun and colleagues identified ctDNA in 1989 (Stroun et al. 1989). Tumor mutations can be detected by sequencing ctDNA, and this allows differentiating ctDNA from cfDNA. Thus, ctDNA is a suitable biosource to analyze the cancer genome (DNA mutations) for diagnosis, prognosis, and prediction of the therapeutic response. Furthermore, as ctDNA can be released from all metastatic sites, it may represent not only the genomic landscape of the primary tumor, but also the intratumor clonal heterogeneity.

Technological Aspects

The preanalytical step (e.g., sample collection) is crucial for ctDNA processing/analysis. Plasma samples should be processed within 6 h after collection to avoid leukocyte degradation that might increase the amount of nontumor cfDNA (Kang et al. 2016; Nikolaev et al. 2018). Leukocyte-stabilizing collection tubes are now commercially available and allow sample storage for up to 48 h (Kang et al. 2016).

There are several cfDNA extraction methods based on centrifugation, immunomagnetic beads, silica column-based enrichment, polymer-mediated enrichment, phenol–chloroform-based extraction, and vacuum generation (Sherwood et al. 2016; Sorber et al. 2017; Lee et al. 2018; Pandoh et al. 2019). However, these methods still require further standardization and harmonization. The choice will depend on the desired DNA purity and the required amount of automatization. As the current DNA extraction methods do not discriminate between ctDNA and cfDNA, DNA extraction must be followed by detection of genomic variations in ctDNA. To this aim, several approaches can be used: next-generation sequencing (NGS), digital polymerase chain reaction (PCR) platforms, real-time PCR, mass-spectrometry, and hypermethylation analysis (Bracht et al. 2018; Elazezy and Joosse 2018; Pantel and Alix-Panabières 2019). The choice is based on the number of genes to be analyzed, the amount of ctDNA and cfDNA in a sample, the nature of the genetic or genomic alteration, and the cost.

Real-time PCR is the most widely used method for liquid biopsy analysis, particularly allele-specific PCR (AS-PCR). These techniques are fast, simple, and cost-effective. However, they display low sensitivity and can detect only few mutations at the same time (Cabel et al. 2018; Pantel and Alix-Panabières 2019). Therefore, there has been a keen interest in improving their sensitivity, and the number of DNA alterations that can be concomitantly tested. NGS allows identifying all point mutations, but is limited by the ctDNA fragment quality, high cost and, and interpretation of complex results. Conversely, digital PCR methods, such as digital-droplet PCR and BEAMing (beads, emulsions, amplification, and magnetics), display high sensitivity and specificity, but detect only a small number of variants and are expensive (Elazezy and Joosse 2018). Recently, mass spectrometry-based methods have been developed: surface-enhanced Raman spectroscopy PCR and UltraSEEK. Although they still require further validation, they show high sensitivity and specificity, requiring low amount of DNA (Mosko et al. 2016; Wee et al. 2016). Finally, the ctDNA methylation status can be assessed by methylation-specific PCR that requires large amounts of ctDNA (Lissa and Robles 2016).

Biology

In normal conditions, cfDNA is mainly released by hematopoietic cells (Sun et al. 2015; Tug et al. 2015), possibly following cell apoptosis and necrosis, but also by active secretion (Diaz and Bardelli 2014; Jeppesen et al. 2019). Most cfDNA fragments are around 143–180 bp in length, which corresponds to the DNA length in a nucleosome (Lo et al. 2010; Thierry et al. 2010), and have a half-life in the bloodstream of 16 min to 2.5 h (Wan et al. 2017). The function of cfDNA (or ctDNA in cancer) is still not clear. It might have a role in the accumulation of macrophages and inflammation after cell necrosis. It might also represent a horizontal gene transfer mechanism between cancer cells (Wan et al. 2017). However, the higher amount of cfDNA found during pregnancy or exercise suggests that its functions might not be only limited to pathological conditions (Lo et al. 2007; Tug et al. 2015).

Clinical Relevance

The only ctDNA-based tests cleared by FDA are the Septin-9 (SEPT9) gene methylation assay for colorectal cancer screening (Song et al. 2017), and the AS-PCR–based assay for the detection of epidermal growth factor receptor (EGFR) mutations in non-small-cell lung cancer (NSCLC) to determine whether a patient is candidate for EGFR tyrosine-kinase inhibitor therapy (Douillard et al. 2014; Weber et al. 2014; Merker et al. 2018). The FDA has also granted the status of “breakthrough device” to CancerSEEK, a method based on the assessment of circulating proteins and ctDNA by mass spectrometry for the detection of early-stage cancer in asymptomatic patients older than 65 years (Cohen et al. 2018). Moreover, ctDNA is used for MRD detection, and a relationship between the tumor stage and amount of ctDNA in blood has been highlighted (Pantel and Alix-Panabières 2019). However, today, ctDNA clinical usage for MRD monitoring has not been established yet, because of the high variability among patients, cancer types, and methods used. Several studies showed the use of ctDNA in different cancer types and at all stages. However, no robust evidence exists on the clinical validity and usage of ctDNA, outside clinical trials (Merker et al. 2018).

Currently, the number of academic and privately funded research projects/studies to find new applications for ctDNA in cancer is increasing worldwide. Nevertheless, cfDNA applications are not limited to cancer, as indicated by its use as a noninvasive prenatal test (Allyse et al. 2015). Therefore, ctDNA might become a common tool in clinical and pathological laboratories worldwide.

EXTRACELLULAR VESICLES

The International Society for Extracellular Vesicles defined extracellular vesicles (EVs) as particles that are naturally released from the cell, delimited by a lipid bilayer and not replicating (Thery et al. 2018). EVs include exosomes, microvesicles, microparticles, oncosomes, apoptotic bodies, lipoprotein particles, and any other nonviral vesicle secreted or shed from cells. They all have specific physiological and pathological functions. Exosomes are the most studied EVs. They are an EV subgroup that has been linked to numerous processes associated with cell-to-cell communications, such as cell proliferation, cell migration, cancer metastasis, and immunomodulatory activity. Exosomes, like other EVs, can be found in blood, urine, saliva, and any other biological fluid. It has been proposed that exosomes can be used as carry-on particles to deliver therapeutic molecules to specific tissues.

In the field of liquid biopsy, the terms EVs and exosomes are often used interchangeably; however, exosomes have an endosomal origin and belong to the small EV (sEV) subgroup, with a size smaller than 100 or 200 nm (Thery et al. 2018). As most EV isolation protocols are based on size and density and not on subcellular origin, it is important to note that exosome is not synonymous of sEV or EV. Nevertheless, it is technically challenging to isolate pure exosomes, and for a practical clinical “liquid biopsy”-based test, many groups call exosomes all particles with a size between 30 and 150 nm (Couto et al. 2018) and with a density between 1.15 and 1.19 g/mL in a gradient (Théry et al. 2006; Chiou and Ansel 2016). In this review, we used the term “exosomes” only for particles of endosomal origin or isolated with high purity methods, whereas we used EVs in all other cases.

Technological Aspects

The gold standard method for EV isolation is ultracentrifugation (Li et al. 2017) that consists in several centrifugation steps (of >100,000 g) of fluids containing EVs. Density gradient ultracentrifugation can be divided in isopycnic and rate-zonal centrifugation. These techniques use a density gradient inside the ultracentrifugation tube to trap specific EVs according to their density (e.g., exosomes) by directly recovering them from the corresponding gradient area (Li et al. 2017). These methods offer the highest EV purity, but a low yield. In addition, they are time-consuming and require the acquisition of an ultracentrifugation equipment (Shao et al. 2016). To improve EV isolation and their clinical applications, other methods have been developed: some are based on size, such as size exclusion chromatography (Böing et al. 2014) and ultrafiltration (Vergauwen et al. 2017), whereas others use immunomagnetic beads against EV surface markers, for instance tetraspanins in exosomes (Oksvold et al. 2015). These methods offer a higher EV yield, can be implemented also in the clinic without the need of special equipment, and are less time-consuming. However, the purity of exosomes and EVs isolated with these methods can be variable (Tauro et al. 2012; Shao et al. 2016; An et al. 2018). Recently, other methods are developed based on microchips or acoustic waves (Wu et al. 2017) to improve the capture of pure EVs. Promising results have been obtained also by combining these different methods (Lobb et al. 2015; Benedikter et al. 2017; Li and Nabet 2019).

Biology

EVs have different functions in cancer and other diseases. Hoshino and colleagues showed that primary tumors secrete exosomes with specific integrin expression profiles at their surface that are associated with formation of metastases in lung or liver. Moreover, their uptake by cells in such organs establishes a premetastatic niche (Hoshino et al. 2015). Exosomes secreted by cultured melanoma or colorectal cancer cells induce higher endothelial permeability (Peinado et al. 2012; Schillaci et al. 2017), and those secreted by fibrosarcoma cells have chemotactic properties (Sung and Weaver 2017). Similarly, exosomes released by the amoeba Dictyostelium discoideum form a kind of pathway that can be directly followed by other cells in vitro (Kriebel et al. 2018).

EV functions are not limited to cancer; indeed, most exosomes in the bloodstream originate from normal physiological events (Johnstone et al. 1987; Italiano et al. 2010; Tao et al. 2017). EVs are also implicated in several inflammatory and chronic diseases (Console et al. 2019) and neurodegenerative diseases (Soria et al. 2017), just to mention a few examples.

EVs and exosomes can contain proteins (Greening et al. 2015), lipids (Skotland et al. 2019), RNA (Janas et al. 2015), and DNA (Thakur et al. 2014). In exosomes, the membrane phospholipid composition/structure reflects that of the membrane of the cell from which they originated (Johnstone et al. 1987; Skotland et al. 2019). Consequently, treatment outcome and prognosis could be predicted by assays based on the detection and analysis of exosomes. Moreover, they may be used to deliver therapeutic agents directly into diseased cells by taking advantage of their specific cell–cell interactions (Luan et al. 2017; Bunggulawa et al. 2018).

Clinical Relevance

The use of EVs in the clinical practice is not established yet; however, EVs and exosomes can bring important information for cancer prognosis and diagnosis and are a candidate predictive biomarker. For instance, EVs in combination with other biomarkers (e.g., ctDNA) have been used to improve the detection of EGFR mutations in lung cancer (Castellanos-Rizaldos et al. 2018; Krug et al. 2018). The National Comprehensive Cancer Network has included EVs (exosomes) in their guidelines (Carroll and Mohler 2018) for early prostate cancer detection (gene expression analysis of EVs from urine) (McKiernan et al. 2016, 2018). The identification of specific biomarkers in exosomes can provide prognostic or diagnostic information. The detection of glypican 1 and macrophage migration inhibitory factor (MIF) in exosomes is associated with colorectal and pancreatic cancer (Costa-Silva et al. 2015; Melo et al. 2015). Also, exosomal miR-19a can predict colorectal cancer recurrence after surgery (Matsumura et al. 2015), whereas the presence of CA125, EpCAM, and CD24 in EVs might be useful for the diagnosis of ovarian cancer (Runz et al. 2007; Zhao et al. 2016). On the other hand, detection of glutathione S-transferase P-1 (GSPT1)-containing exosomes is associated with poor prognosis in breast cancer (Yang et al. 2017). Most of the possible clinical applications of EVs still require validation and standardization, but they reflect the wide range of biomarkers that can be found in such particles.

Besides the identification of biomarkers in exosomes, it has been suggested that other EVs should be also considered. For example, oncosomes and larger EVs contain a higher amount of DNA. As exosome cargos lack double-stranded DNA (dsDNA) and histone proteins (Jeppesen et al. 2019), larger EVs might be a better biosource to identify DNA mutations (Vagner et al. 2018) for clinical use. Recently, exomeres have been described as non-EV nanoparticles. They are smaller than exosomes (diameter of ≤35–50 nm) and are associated with unique cargo profiles (Zhang et al. 2018). Although, more research is necessary to clearly define EV function, in the liquid biopsy field, it would make sense to isolate different EV types in function of the targeted biomarker or disease type, thus using the right EVs for the right application.

TUMOR-EDUCATED PLATELETS

Platelets are anucleated cell fragments that originate from megakaryocytes. They are central players in hemostasis, thrombosis, immunity, inflammation, and metastasis (Rodvien and Mielke 1976; Gros et al. 2014; Ali et al. 2015; Hou et al. 2015; Thomas and Storey 2015; Brass et al. 2016; Leblanc and Peyruchaud 2016; Buettner 2018). In 1865, Trousseau described for the first time, platelet involvement in cancer. He observed that patients with cancer often presented thrombophlebitis and blood clotting far from the tumor site (Trousseau 1865; Menter et al. 2014). Then, Billroth described cancer cell-containing thrombi in the circulation, showing the direct interaction between platelets and tumor cells (Billroth 1877). After a century, Gasic et al. (1973) showed that tumors can induce platelet aggregation, and correlated this with metastases in mice. They suggested that activated/aggregated platelets interact with tumor cells, and increase tumor cell extravasation to the metastatic niche.

High-throughput technology allowed studying the complex interaction between platelets and cancer, leading to the term of TEPs (Nilsson et al. 2011; Sol and Wurdinger 2017; Best et al. 2018). TEPs are a valuable biosource for noninvasive assays. Although platelets are easy to purify, it is important to prevent their activation during handling because it affects their molecular and morphological features. Therefore, strong mechanical or biochemical forces should be avoided (Mustard et al. 1989; Hoffman et al. 1992; Cazenave et al. 2004). Nevertheless, the optimization of a fast and highly efficient method for platelet isolation from a small blood amount is vital. For instance, to minimize platelet activation during blood sampling, blood should be gently collected by using a 1.2 mm intravenous cannula (Amisten 2012). As many drugs interfere with platelet function (Schrör 1997; Scharf 2012), it is important to precisely record all patient's treatments. TEPs can be isolated up to 48 h after blood sampling and allows the obtention of high-quality RNA for molecular tests (Best et al. 2015, 2019).

Platelets contain many growth factors and cytokines that could be involved in the metastatic process through multiple mechanisms (Karpatkin et al. 1988; Trikha and Nakada 2002; Boucharaba et al. 2004). During their lifespan, platelet directly interact with cancer cells via receptors, and indirectly through signaling molecules (Morimoto et al. 2008; Konstantopoulos and Thomas 2009; Labelle et al. 2011; Golebiewska and Poole 2015; Schlesinger 2018; Ward et al. 2018). This interaction leads to platelet activation, resulting in the release different molecules (e.g., PDGF, TGFβ, VEGF-A/C BFGF, EGF, and IGF1) (Erpenbeck and Schön 2010; Radziwon-Balicka et al. 2012; Menter et al. 2014) that can provide a protumor metastatic niche (McAllister and Weinberg 2014) and may induce EMT. Direct platelet–cancer cell interaction and contact-dependent signaling by platelet-derived TGFβ can activate the TGFβ-SMAD and NF-κB pathways, resulting in the acquisition of mesenchymal features by cancer cells and increased metastasis potential (Labelle et al. 2011). TANK-binding kinase 1 (TBK1) is another platelet-induced EMT and metastasis mediator, and a potential therapeutic target to prevent metastasis formation (Zhang et al. 2019). Moreover, platelets can protect tumor cells by forming a protective cloak against natural killer (NK) cell-mediated lysis (Nieswandt et al. 1999) and tumor necrosis factor (TNF)-α (Philippe et al. 1993). Platelets can also disturb the “missing self” recognition by NK cells by conferring a “pseudonormal” phenotype, through transfer of major histocompatibility complex (MHC) class I molecules to the cancer cell surface during aggregation (Placke et al. 2012). Platelet–cancer cell signaling also stimulates matrix metalloproteinases-9 (MMP-9) expression, promoting cancer cell invasion (Alonso-Escolano et al. 2006; Radziwon-Balicka et al. 2014).

Several studies on platelet role in cancer progression consider TEPs as high-potential liquid biopsy biosource. Based on mRNA sequencing of 283 TEP samples, Best and colleagues could differentiate between healthy controls and patients with cancer (96% accuracy). Based on TEP mutational status, they could obtain some information to predict the tumor type. They also developed a highly sensitive algorithm that can classify early-stage (localized) and advanced (metastatic) cancer (Best et al. 2015). TEPs can also be used to predict the treatment response to targeted therapies. The algorithm also predicted the presence of MET amplification, EGFR mutations, and KRAS mutations in TEP samples (Best et al. 2015). The detection of tumor-driving mutations reflects TEP potential use in future clinical trials and to predict the therapeutic response (Joosse and Pantel 2015). In agreement, resistance to chemotherapy was identified by monitoring EM4-ALK rearrangements in TEPs (Nilsson et al. 2016). In addition to transcriptome studies, platelet proteomic analysis allowed differentiating between benign adnexal lesions and late-stage ovarian cancers (Lomnytska et al. 2018). Also, platelets proteome of patients with early stage cancers differs noticeably compared with healthy samples (Sabrkhany et al. 2018).

It has been suggested that CTCs and platelets interact, but the underlying mechanism remains unknown. Jiang et al. (2017) showed the existence of platelet-coated CTCs in patients with metastatic cancer, but they did not molecularly characterize CTCs and platelets to address the interplay between EMT and CTC-TEP interaction. Combinatorial analysis of TEPs with complementary biosources, such as CTCs, EVs, and ctDNA, and protein biomarkers could be the next-generation tools for early cancer diagnosis (In ‘t Veld and Wurdinger 2019).

CONCLUSIONS

In the next decade, cancer diagnosis will focus on early detection and personalized management. Real-time liquid biopsy will play a fundamental role in this progress. CTCs, ctDNA, EVs, and TEPs are complementary and they may become routine clinical laboratory tests in the near future. An index or an algorithm could be developed to combine all these data to obtain a more precise tumor profile. Moreover, more emphasis on technical validation is required, and projects, such as the European CANCER-ID, European Liquid Biopsy Academy (ELBA), European Liquid Biopsy Society (ELBS) networks or the U.S.-based BloodPAC, have been initiated to meet this challenge.

This field is in constant expansion, and it is estimated that the liquid biopsy market will increase from $200 million to $1 billion in the next 5 yr (Budel et al. 2019). The spread of these relatively new biosources will reduce the economic cost of invasive procedures or inefficient therapies, thus improving the cost-efficiency of cancer management with direct implications for the overall survival and quality of life of patients.

Liquid biopsy should not be seen as a substitute of the histopathological diagnosis in tissue biopsies, but as a complementary tool for diagnosis/characterization, and as part of an innovative tumor management strategy (e.g., CTCs might be used to identify new therapeutic targets to stop metastatic initiator cells; EVs as delivery particles for target therapies in specific tissues or tumors) (Fig. 1). The presence and biological functions of these biosources are not limited to cancer, and future work will open a wide range of applications that might not even be directly related to diseases or human well-being.

Figure 1.

Liquid biopsy potential for cancer monitoring. (A) Complementarity of tissue and liquid biopsies. (B) Overview of liquid biopsy biomarkers that represent the tumor molecular heterogeneity and immunologic phenotype. The primary tumor and metastatic sites release different cells and biomolecules in the bloodstream. Circulating tumor cells (CTCs; as single cells or clusters), circulating tumor DNA (ctDNA), extracellular vesicles (EVs), and tumor-educated platelets (TEPs) can be isolated from whole blood to obtain cancer genome, transcriptome, proteome, and secretome data in real time. Moreover, white blood cells (WBC) can give information on the immunity of the patient with cancer, mainly on their potential to eradicate cancer cells. All these biomarkers are complementary and can be used to build a precise index or an algorithm based on qualitative and quantitative data at different times during the disease course.

“Liquid biopsy” will become a crucial tool for oncologists and physicians in general, and is one example of how cancer management will drastically change in the future years.

COMPETING INTEREST STATEMENT

The authors declare no potential conflicts of interest.