Globally, it was estimated that 308,102 people were diagnosed with a primary central nervous system (CNS) tumour in 2020, with incidence rates projected to rise by 6% between 2014 and 2035 (1, 2). Brain tumours cause more fatalities in children and adults under the age of 40 than any other cancer, reducing the life expectancy by an average of 20 years per patient, the highest of any cancer (3, 4).

Despite advances in surgical resection, chemotherapy and radiotherapy, around only 13.5% of adults survive brain tumours for five or more years after diagnosis (5). Data compiled by the Brain Tumour Charity found that a third of brain tumour patients had visited their GP at least five times before diagnosis (6, 7). Furthermore, over 50% are diagnosed via emergency departments rather than the GP; many of those patients presenting later in the course of the disease with large, inoperable tumours (8, 9).

Presently diagnosis and disease monitoring rely on access to imaging in secondary care, which is costly and overburdened: with 230,000 patients waiting more than a month for test results (10). Imaging can also intermittently produce false positive results due to non-malignant inflammatory changes mimicking tumour recurrence (11). Subsequent to imaging, the patient will undergo neurosurgery, during which a diagnostic tissue biopsy is taken. However, this biopsy provides static information that becomes obsolete as the cancer evolves. Different sub-clones expressing altered targetable biomarkers may emerge within the cancer during the course of the disease, highlighting the limitations of relying solely on static biopsy data (12). A better understanding of intertumoral heterogeneity is required to inform mechanisms of tumour resistance to therapies (13, 14). Consequently, there is an urgent need to utilise innovative methodology to improve patient diagnosis and overall survival.

A potential solution is to utilise a liquid biopsy assay for circulating tumour cells (CTCs) in peripheral blood samples from brain tumour patients. Not only are these simple blood tests low in cost and minimally invasive; they could be implemented in both the primary or secondary care setting (15). This has the potential to expedite diagnosis, monitor tumour genomic changes through serial samples and detect early relapse or resistance to current therapies (16, 17).

The benefits of CTC detection have been widely explored in other malignancies such as breast, colorectal, prostate, gastric, bladder, melanoma and small and non-small cell lung carcinoma cancer (18–26). In 2013, the LANSCAPE trial investigated CTC levels in breast cancer patients, with metastases to the brain, before and after treatment with lapatinib and capecitabine at 21 days, in Her2 positive tumours (27). The trial demonstrated a correlation between CNS metastasis response, outcome, and early CTC clearance under targeted treatment of Her2 positive, metastatic breast cancer (27). CTC count has also been shown to predict progression-free survival and overall survival in non-small cell lung carcinomas after multivariate analysis (23).

Using MTW9 carcinomas, (28) demonstrated that the presence of large numbers of tumour cells in the blood is not, by itself, a sufficient condition for metastatsis to occur. Multiple studies have similarly demonstrated that despite the detection of a high number of cancer cells in the blood, as few as 0.01% of CTCs develop into secondary tumours (29–31). CTC intravasation can occur through active and passive shedding (32). Bockhorn et al. (33), identified the loss of CD44 and α3 integrin in CTCs shed from renal cell carcinoma. Both CD44 and α3 integrin play a role in cell adhesion and a reduction makes it much easier for the cells to pass into the blood stream (34, 35). Blood vessels created by angiogenesis are immature, malformed and leaky with detached endothelial cells and an irregular or missing basement membrane (36, 37). Although it is not fully understood how these abnormalities affect intravasation it most likely helps with this process and could also account for non-viable cells, as well as viable cells, being leaked into the blood stream (38, 39). Proliferating cells have also been shown to compress and collapse intra-tumour blood vessels, which would enable the tumour cells to passively enter the blood stream (38, 39).

Throughout tumour progression there is active cross-talk between the tumour cells and micro-environment (30). This signalling is mediated by cell-to-cell interactions and cytokine/growth factors. Morphological changes which support metastasis are triggered by this signalling (40). Neurons, a crucial component of the glioma microenvironment, have been shown to regulate malignant growth in an activity dependent manner (41, 42). Synaptic communication is suggested to occur through AMPA (α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid) receptors, particularly the glutamate receptor (43). Glutamate, a key neurotransmitter, is considered a potential growth factor for glioma development (44).

Many studies have supported the hypothesis that neoplasms are heterogeneous and there is a distinct sub-population of tumour cells, with differing angiogenic, invasive and metastatic properties (45). This distinct subgroup can use epithelial-mesenchymal transition (EMT) to become increasingly motile, which in turn enables them to migrate to the vascular system through growth factor and nutrient gradients (46, 47). In contrast to the passive model described above these cells have been shown to actively migrate, passing either paracellularly through the endothelial cell junction or transcellularly through the endothelial cell body into the blood stream (48). These highly metastatic cells have been shown to produce matrix metalloproteinases, which actively digest the interstitial matrix and basement membrane, enabling them to pass through the tissue into the blood stream (49).

CTC arrest can be triggered by several obstacles within the bloodstream, including entrapment by capillaries; reduced diameter dimensions and biomechanical constriction forces of the capillary lumen, which have been shown to severely deform the cell cytoplasm and nucleus thus triggering cell death (50, 51). It has been suggested that capillary constriction can reduce the potential for CTCs to enter the vessels by as much as 90% (52). The role of capillary entrapment is less clear because entrapment may also be important for metastatic progression, enabling CTCs to adapt to the new environment, facilitating invasion and colonisation at the metastatic site (53, 54). In order to survive in the bloodstream CTCs must also evade hemodynamic shear forces and the immune system (55, 56). However, increasing evidence is emerging that CTCs are not as mechanically fragile as first thought and in fact can withstand fluid shear stresses encountered through circulation (57). CTCs have been shown to induce platelet activation and aggregation to protect their survival in the blood stream. Mounting evidence has also validated this interaction as a key feature of metastasis (58–61).

Originating from glial cells it is estimated that gliomas account for 75% of all primary malignant brain tumours (62). Glioblastomas (GBMs), the most aggressive and common glioma, are associated with dismal prognosis and rapid recurrence, despite multimodal therapies (62, 63). GBM cells are highly migratory and extensive infiltration of these cells into the brain parenchyma makes remedial surgical resection almost impossible (64). Systemic metastases from GBMs however are incredibly rare, 0.5% metastasise compared to 10–45% of other primary cancers that metastasise to the brain. It is thought that the brain’s distinct microenvironment, containing the blood brain barrier and stem cell niches, significantly influences this rate (65).

The permeability of the blood–brain barrier is associated with GBM progression, heightened intravasation chances, and is suggested to be due to the disruption of endothelial/ astrocytic interaction and impaired vessel formation (66). Davis (67), reported the first ever case of GBM metastasis. Since then this number has increased progressively: this is thought to be due to improvements in imaging and patient survival (68). Metastatic GBM cells can spread through blood and lymphatic vessels (69). Onda et al. (70), undertook autopsies on 51 patients who had died from GBM and found that 14 of the 51 cases had dissemination by cerebral fluid.

Since 2014, there has been substantial progression in CTC isolation and characterisation from high grade glioma patients (Table 1). Sullivan et al. (71), found isolated CTCs had elevated markers, which are associated with the more aggressive mesenchymal subtype. GBMs can be divided into 4 subtypes: proneural, neural, classical and mesenchymal (82). The mesenchymal subtype, characterized by higher migratory capabilities, is associated with worse prognosis and is strongly linked to GBM metastases and recurrences (83, 84). Microglia have been shown to induce mesenchymal status through the tumour necrosis factor alpha (TNF-α)/nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) pathway. Additionally, hypoxia has been shown to induce transition and increase stem cell markers in GBM cells (84, 85). Multiple subtypes coexist within the same tumour, and mesenchymal transition is thought to occur late in GBMs, resulting in a more aggressive, invasive and recurrent tumour (86, 87).

A study by (72), determined that genomic abnormalities not only correlate between isolated CTCs and the tumour of origin, but also revealed the maintenance of epidermal growth factor receptor (EGFR) amplification in CTCs, indicating sustained growth potential. EGFR also promotes stemness in GBM cells (88). Although cohort numbers were small, results suggest that CTC detection could be used to identify GBM patients with a large tumour or those at risk of recurrence (72). In another study, utilising sensitive immunocytochemical detection, with glial fibrillary acidic protein (GFAP) as a marker for CTCs in peripheral blood cytospin preparations, putative CTC cells were detected in 29 out 141 GBM patients (73). Furthermore, these reputed CTCs were more frequently detected in patients with EGFR gene amplification in the corresponding tumour tissues (73).

In 2016 (74), detected 7 different glioma subtypes in peripheral blood samples using an integrated cellular and molecular approach. Clinical data revealed that CTC detection was superior to MRI in monitoring treatment response and differentiating radionecrosis (74). This study identified nonhematogenic aneuploid circulating aneupolid cells in seven diverse subtypes of brain glioma and reported their significance. This has been further supported by Li et al. (89), who detected and characterised aneuploud circulating rare cells in glioma patients and demonstrated their unique clinical significance.

Malara et al. (75), captured CTCs in a 67-year-old GBM patient pre- and 2 months post-surgery. Interestingly, the post-surgery sample showed a higher number of CTCs. Unfortunately, the patient experienced tumour recurrence 9 months after the sample, and succumbed to the disease 5 months later (75).

Zhang et al. (76), found that the positive rate of CTCs in gliomas increased progressively with the advancing stage of glioma. They utilised cell surface marker independent technology based on telomerase specific, replication-selective oncolytic herpes-simplex-virus-1, which identifies viable CTCs from a wide range of malignancies. The first evidence of CTC clusters was confirmed by (77), who noted them in 53.8% of progressive GBM patients. Bang-Christensen et al. (78), successfully isolated CTCs in every blood sample processed with magnetic beads coated with VAR2CSA malarial protein (rVAR2), which detected CTCs through the protein oncofetal chondroitin sulfate. Spiral microfluidic technology was used by (79), to successfully isolate CTCs from GBM patients. The study also demonstrated that patients with CTC counts equal to 0 after surgery had significantly longer recurrence free survival.

A sized based separation protocol with MetaCell® tubes was used by (80), to detect more mutations in CTC samples compared with the paired primary tumour. Qi et al. (81), used biocompatible parylene polymer membranes, with a pore diameter of 8 μm under a high flow rate, to enrich CTCs without requiring tumour cell-specific capture. Qi et al. (81), found CTC numbers to be higher in astrocytoma samples compared to oligodendroglioma samples. A number of CTC-white blood cell clusters, which could be used to monitor recurrence, were also detected in the study. Qi et al. (81), noted no difference in glioma subtype but in contrast found that resection could promote CTCs. It was also found that the level of CTCs was related to p53 mutation, isocitrate dehydrogenase 1 (IDH1) status and poor outcome (81).

CTC isolation techniques can be divided into 2 broad groups: physical and biological (90). Physical properties include separation by size, elasticity and surface charge (91). Methods used include density gradient centrifugation, microfiltration, microfluidics and dielectrophoresis (92). Antibodies with conjugated magnetic or non-magnetic beads are used to separate the CTCs through their biological properties. This can be either through positive selection, CTCs targeted directly, or negative selection, blood cells for example, are targeted and removed through this method (93).

The aim of this review is to assess and compare four commercially available methods for CTC enrichment: OncoQuick®, Screen Cell®, pluriBead® and Cell Search®. By analysing performance metrics and clinical adaptability, the objective of this study is to provide guidance in selecting a suitable method for potential translational application in the clinical setting, with a primary focus on the isolation of CTCs from GBMs. Commercially available methods were intentionally selected in this study to facilitate easier implementation in the clinical setting.

Each method applies a distinct enrichment technique. The Cell Search® system uses anti-epithelial cell adhesion molecule (EpCAM) conjugated with magnetic beads to isolate CTCs. This system is among the most widely used CTC enrichment techniques, as it is the only CTC detection system approved by the Food and Drug Administration in the United States for the enumeration of CTCs in metastatic colorectal, prostate, and breast cancers (94). Pierga et al. (27), used the Cell Search system to demonstrate that CTCs can be used as early predictive markers for poor overall survival and progression free survival in metastatic breast cancer patients. The study also demonstrated the use of CTCs in monitoring treatment benefit. The downside to this system is that it solely relies on EpCAM for detection. Therefore alternative methods have been developed.

The OncoQuick® method, which uses density gradient centrifugation has been shown to yield higher relative tumour enrichment when compared to standard to the standard density gradient centrifugation system Ficoll (95). In addition to CTC detection in gliomas, the OncoQuick® method has successfully isolated CTCs in studies involving colorectal cancer, melanoma and breast cancer patients (18, 96, 97). The isolation of CTCs through pluriBead® involves the use of non-magnetic beads coupled with monoclonal antibodies specific to the CTC surface antigens. Pierzchalski et al. (98), successfully validated this system for simultaneous separation of CD4+ and CD8+ cells from human EDTA-blood samples. The ScreenCell® method, which captures CTCs through size isolation, determined CTCs in patients with a less favourable stage III laryngeal squamous cell carcinoma (99). The ScreenCell® method has also isolated CTCs from urinary bladder, metastatic prostate and colorectal cancer (100–102).

The Cell Search® System was unable to detect U251 MG cells spiked in normal whole blood due to the lack of EpCAM on the surface of these cells. Modifications could be made to the system to include additional markers, but at present only EpCAM positive cells are detected by this system.

Other studies have shown the failure of the Cell Search® System to detect rare CTCs and also detect CTCs in patients with a widely metastatic disease, despite those patients have high numbers of CTCs identified through alternative methods (94, 104, 105).

Tumour progression is associated with a loss of epithelial features and a transition towards a mesenchymal phenotype, a process known as EMT. Tumour cells are known to undergo EMT as a means of entering circulation (106). A loss in epithelial markers, such as EpCAM, would prevent the CTCs from being detected by the Cell Search® System. Konigsberg et al. (107), determined, in metastaic breast cancer patients, that the density gradient centrifugation method OncoQuick® appeared advantageous for CTC isolation compared to MACS HEA MicroBeads® (MACS), which also relied on EpCAM immunomagmetic enrichment technology. This technology also requires the costly purchase of the Cell Search® System, which could prevent many clinical settings from using this system.

The U251 MG cells were successfully enriched using the OncoQuick® method, which correlates with the findings from (72), who successfully isolated CTCs from high grade glioma patients using this method. Although the method was successful, many steps were required to perform the analysis, which is time consuming. This is an important factor to consider when reviewing this technique for clinical application. Numerous steps would also increase the chances of losing ‘rare’ CTCs, particularly when transferring the solution between centrifuge tubes. The manufacturer recommends that the plasma fraction is discarded when platelet contamination is seen following centrifugation. Removing the plasma fraction could result in loss of CTCs due to unwanted contamination of this fraction. Alternatively, CTCs could form non-specific aggregates, which could cause them to move to the bottom gradient, again leading to false negative results. CTCs have been shown to bind with platelets, fibroblasts, and leukocytes to evade blood stream hazards (108, 109). If CTCs are present in these clusters, they could also move to the bottom gradient and subsequently be missed through the OncoQuick® method.

GBMs have a high degree of intratumoral heterogeneity (110). CTC profiles can change during tumour cell dissemination (111–113). Before entering the blood CTCs undergo varying degrees of EMT, which leads to variability in cell markers (114). Although the selected markers for the pluriBead® technique are associated with tumorigenesis and cell migration in GBMs, c-Met acting as an independent predictor for GBMs, they are heterogenous (115–117). The successful capture of PC3 cells compared to U251 MG isolation, suggests that the selected markers were not present on the surface of the U251 MG cells. This also highlights the potential difficulty of attempting to capture CTCs in a clinical setting using the pluriBead® technique.

To isolate GBM CTCs using purely biological properties rather than their physical properties increases the chances of CTCs being missed. Multiple clones of the same primary antibody were selected for the pluriBead® technique. One clone typically binds to just one target molecule presenting a single epitiope. The epitopes present could vary greatly in each CTC, even from the same parent tumour, which would minimise the successful chances of capturing CTCs with pluriBead®. To increase the possibility of capturing CTCs, the pluriBead® technique could be repeated overall several rounds with beads conjugated to different antibodies. This however also increases the chances of losing CTCs, particularly when there could just be one CTC present in the blood sample. This would also increase the time required for the completion of this technique.

Both Screen Cell®- LCD and Screen Cell®-MB kits were able to isolate U251 MG cells, which were then successfully cultured. The Screen Cell® method was easy to use and rapid, taking only 3 min to perform the process. The technique is also sensitive, capturing cells from a blood sample, which had been spiked with just 2 × 10¹ U251 MG cells. Another benefit of the method is the single enrichment step, consisting of blood passing directly through a filter, which may reduce the chance for CTCs to be lost. Both kits offer the advantage of supporting further analysis of markers. The Screen Cell®-MB kit has a particular advantage that the cells can be analysed directly for DNA/RNA, or they can be cultured first and then analysed for DNA/RNA. This quick and cheap method could be undertaken at the patient’s bedside with no requirement for pre-processing. Fast enrichment also minimises disruption to the CTCs, which preserves the cell phenotype.

Cell counts used for the validation of the enrichment techniques used in this study were as recommended by the manufacturers. A clear limitation of this study is that the cell counts are much higher than those expected when if clinical samples were tested with the selected method. Nonetheless this has provided us with a good opportunity to assess the limitations of each method, even with a higher cell count. Bang-Christensen et al. (78), reported enriching between 0.5 and 42 CTCs in 3 ml blood. Therefore 3 mL of normal whole blood spiked with 2 × 10¹ U251 MG cells was used to test the sensitivity of the ScreenCell® technique. The ScreenCell® method was successfully able to isolate the U251 MG cells at this concentration. Another limitation of this study is that commercial cultures rather than patient derived cultures have been used. Additionally, the techniques were not validated on patient plasma samples.

In comparison to the other enrichment methods compared in this study, Screen Cell® appears most favourable to use in a healthcare setting as it is simple, cheap and quick to use (Table 3). Only one step is required before the CTCs are captured on the membrane filter, through size isolation, thus, maximising the chances of CTC capture. This method could be easily introduced into a busy clinical setting, where reliable and quick results are required. This system offers an option for simple cytomorpholical diagnosis after routine staining of CTCs. It also supports a number of other potential characterisation techniques and could enable captured CTCs to be successfully cultured.

By contrast, isolation methods which rely on single CTC biomarkers such as pluriBead® and the Cell Search® System could lead to higher false negative results, due to CTC heterogeneity. EGFR for example, which was used to characterise isolated U251 MG cells, demonstrates heterogeneous expression in GBMs (115). A multi-step process such as the OncoQuick technique could also lead to CTCs being missed, due to them being lost during one of the processing stages.

The potential benefit of using GBM CTCs diagnostically in the healthcare setting is threefold: it could enable earlier diagnosis, disease monitoring and potential reassurance of the worried well. Out of the four commercially available CTC enrichment methods – OncoQuick®, Screen Cell®, pluriBead® and Cell Search®- we found that the Screen Cell® method offered the most potential for translational application in the clinical setting. Alongside being simple, cheap and quick, this CTC enrichment method was not limited to isolating CTCs through one characteristic. It also supports a wide range of downstream analysis options. Further validation of the ScreenCell® technique is now required, which will be completed on GBM patient blood samples in a clinically relevant setting.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

The studies involving humans were approved by The Brain Tumour Bank South West and Brain UK Ethics 15/006. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

HB: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Validation, Visualization, Writing – original draft, Writing – review & editing, Funding acquisition. CP: Supervision, Writing – original draft, Writing – review & editing. KK: Supervision, Writing – original draft, Writing – review & editing, Conceptualization, Funding acquisition.