Circulating Biomarkers in Diffuse Gliomas: Current Landscape and Future Directions

Gliomas are primary brain cancers of varied histologic grades and molecular alterations. Diffuse gliomas are, for the most part, fatal, and carry a poor prognosis. Recent advancements in the genomic landscape of gliomas have allowed for a shift from morphologic characterization to a more molecular-based diagnosis. The 2021 World Health Organization classification of central nervous system tumors reflects this shift by incorporating molecular information with histology for an integrated diagnosis.¹

Several tissue-based molecular markers have been recognized for accurate diagnosis and prognostication in glioma (Table). These include isocitrate dehydrogenase (IDH) 1/2, 1p/19q-codeletion, methylguanine-DNA methyltransferase (MGMT) promoter methylation status, alpha thalassemia/mental retardation syndrome X-linked (ATRX) expression, telomerase reverse transcriptase (TERT), histone H3 (H3), epidermal growth factor receptor (EGFR), serine/threonine-protein kinase B-Raf (BRAF), and cyclin-dependent kinase inhibitor 2A/B (CDKN2A/B) loss. Of these, IDH sequence variation status is particularly critical as it has prognostic significance in glioma classification. Tumors with an IDH sequence variation and 1p/19q-codeletion are classified as IDH-mutant oligodendroglioma; those without the codeletion are labeled as IDH-mutant astrocytoma.

The presence of an IDH1/2 sequence variation is associated with longer overall survival and progression-free survival. Lower-grade gliomas with IDH sequence variation and 1p/19q-codeletion typically have the most favorable outcome. Tumors without an IDH sequence variation are labeled as IDH–wild-type glioblastoma, with higher grades automatically categorized as such. Other glioma biomarkers can be used to estimate prognosis, such as MGMT promoter methylation, a favorable prognostic marker independent of treatment. In contrast, EGFR amplification and EGFR variant III (EGFRvIII) expression are indicators of poor survival in glioblastoma. Lower-grade IDH–wild-type gliomas with chromosomal loss of 7 and gain of 10, EGFR amplification, and TERT promoter sequence variation are considered molecular glioblastoma and are hallmarks of the disease.

The diagnosis of glioma is typically established using standard MRI and confirmed through examination of tumor tissue obtained by surgical resection or biopsy. These procedures are invasive and carry substantial risk, and the sample taken may not always represent the entire tumor accurately because of spatial heterogeneity. Standard response assessment for glioma is based on serial MRI scans. However, relying solely on imaging during treatment can provide limited information on disease progression and treatment response. This presents a challenge in managing gliomas, because resampling tumor tissue for further evaluation through repeated surgeries may not always be feasible. As a result, there is a growing need for a noninvasive method that can be used repeatedly throughout the course of the disease to profile the tumor. This has led to an active focus on liquid biopsy as a promising potential solution in neuro-oncology.

Liquid biopsy involves identifying and analyzing circulating biomarkers in blood, cerebrospinal fluid (CSF), and other bodily fluids to gather molecular and genomic information for tumor characterization (Figure). Whereas blood-based biomarkers are commonly used to monitor treatment response and disease progression in many solid organ cancers, there are no established circulating biomarkers for assessing response in gliomas.^2,3 We provide an overview of the current state of circulating biomarkers in gliomas and their role in the evolving landscape of glioma management.

Tumor-Associated Circulating Biomarkers

Ideal biomarkers in neuro-oncology would assess the dynamic tumor state objectively and reproducibly during the course of treatment and help determine pharmacologic response to a therapeutic intervention. Tumors release multiple components into the circulation, and the following tumor-derived components have been shown to be promising biomarkers in gliomas.

Circulating Tumor Cells

Circulating tumor cells (CTCs) are cells released from the primary tumor that enter the systemic circulation, potentially leading to distant metastases.² Whereas gliomas, particularly glioblastoma, were once believed to lack CTCs because of their rare extracranial spread and the presence of the blood–brain barrier (BBB), recent evidence suggests otherwise. Although glioma cells do not express epithelial cell adhesion molecule (EpCAM), a cell surface membrane protein commonly used to isolate CTCs in other solid cancers, studies have identified other markers and methods for CTC detection and isolation in gliomas. GFAP-positive CTCs harboring EGFR amplification and gain of chromosome 7 and loss of chromosome 10 have been found in the blood of people with glioblastoma.⁴ CTCs derived from blood from individuals with glioblastoma were noted to have a mesenchymal profile rather than a neural differentiation, suggesting a more aggressive phenotype, with a higher frequency of CTCs in those with progressive disease.⁵ Other methods of CTC enrichment, such as negative selection and microfluidic technology, have been described in the literature.^5,6 CTCs derived from glioma are more abundant in CSF, and their detection in CSF has a higher sensitivity than in peripheral blood. There is some preliminary evidence that CTC count may decrease with adjuvant treatment and increase with progressive disease, but the clinical value of glioma CTCs remains unclear because of limited data from studies with insufficient sample sizes and short follow-up. Challenges in using glioma CTC isolation and characterization in clinical practice include the need for fresh samples, the need for immediate processing, and low detection technology sensitivity. Further studies are required to validate the results of these initial studies and establish the clinical utility of glioma CTCs.

Circulating Tumor DNA or Cell-Free DNA

Circulating tumor DNA (ctDNA), or cell-free DNA (cfDNA), refers to small fragments of tumor-derived DNA that circulate in the blood or CSF and are not coupled to cells.² Various methods, such as digital droplet polymerase chain reaction (PCR), next-generation sequencing (NGS), and methylation profiling, have been used to detect alterations in ctDNA derived from CSF or blood of people with gliomas or other brain tumors. Further studies demonstrated the ability to detect IDH sequence variation; loss of heterozygosity in chromosomes 1p, 19q, and 10q; and MGMT promoter methylation in ctDNA derived from serum or plasma with high specificity but moderate sensitivity.^7,8 Similarly, EGFRvIII deletion, a tumor-specific sequence variation, was detected in circulating DNA of people with glioblastoma using a long-range PCR amplification strategy.⁹ NGS has also been used to detect alterations in the cfDNA in blood samples of people with primary brain tumors, including glioblastoma, astrocytoma, and oligodendroglioma, with the most common alterations detected being in Tp53, JAK2, NF1, EGFR, BRAF, IDH1, NRAS, GNAS, and ATM.¹⁰ ctDNA was found to be more representative in CSF than plasma, with a higher sensitivity.¹¹ ctDNA was detected in CSF collected either through a lumbar puncture (LP) or from the sylvian or interhemispheric fissure using NGS.^12–14 High concordance was seen in the detection of multiple gene sequence variations using NGS in CSF ctDNA and glioma tissue.¹⁴ The detection rate and levels of ctDNA were higher in high-grade tumors compared with lower-grade ones,^10,15,16 and tumors located near a CSF space consistently had more detectable ctDNA than those not directly related to a cisternal space.¹⁵ Despite the potential for higher detection yield of ctDNA in CSF compared with plasma, its feasibility may be limited in people with intracranial masses because of the risk of herniation from an LP. Repeated sampling through an LP is also impractical and carries some risks. Current studies involving ctDNA have used different collection and detection methods and patient populations, and involved few participants.² Nevertheless, these data suggest that ctDNA has the potential to detect tumor-derived genetic alterations.

Extracellular Vesicles

Extracellular vesicles (EVs) are released by normal and tumor cells into the cellular microenvironment and biologic fluids, and carry molecules, including proteins, lipids, and nucleic acids (eg, fragments of DNA, messenger RNA, microRNA), representative of their cells of origin.² These particles are delimited by a lipid bilayer and cannot replicate on their own.¹⁷ EVs are found in the peripheral blood and CSF of people with glioma, indicating that they cross the BBB. In people with glioblastoma, EGFRvIII was identified in serum EVs, and its presence was correlated with poor survival.¹⁸ The same sequence variation can be detected in CSF-derived EVs with a high specificity but low sensitivity. Through BEAMing (beads, emulsions, amplification, magnetics) PCR, copies of variant IDH1 transcripts were detected in CSF using mRNA extracted from EVs and reverse transcribed into cfDNA for copy number analysis.¹⁹ Studies have found that microRNA (miRNA) in serum and CSF EVs could be potential diagnostic and prognostic biomarkers. A combination of 4 miRNAs (miR-182-5p, miR-328-3p, miR-485-3p, and miR-486-5p) has been shown to distinguish people with glioblastoma from healthy controls.²⁰ In addition, the levels of CSF-derived miR-21 were found to be correlated with tumor recurrence in a cohort of individuals with recurrent glioma, suggesting promising prognostic value.²¹ Serum EV miRNA (mir-210) expression levels were higher in individuals with glioma compared with healthy controls, decreased significantly after surgery, and increased upon recurrence.²² Furthermore, glioblastoma stem cell–like cell-derived EVs were found to impede T-cell activation, indicating the role of glioblastoma-derived EVs in immunosuppression.²³ This demonstrates the potential for an improved understanding of treatment resistance mechanisms through EVs. Overall, studies on EVs in gliomas are limited by their small sample sizes, and more standardized procedures are needed.

Potential Clinical Applications of Circulating Biomarkers in Gliomas

Diagnosis

Conventional biopsy remains the most reliable method of histologic and molecular characterization of gliomas. Liquid biopsy is an appealing alternative in the setting of inoperable, difficult-to-reach tumors. For example, the histone H3 gene sequence variation was found to be detectable in CSF and plasma of people with diffuse midline glioma.²⁴ Besides the setting of inoperable tumors, circulating biomarkers may be important in tissue biopsies with insufficient samples for molecular testing. Liquid biopsy is a potentially useful complementary test in glioma diagnosis.

Therapy Selection

The mainstay treatment of gliomas includes maximal safe resection, radiation, and chemotherapy. There has been growing interest in precision neuro-oncology, wherein genetic alterations that are potentially targetable are identified in an individual and matched to the relevant drug. Several clinical trials have shown promising results with targeted therapies, such as dabrafenib plus trametinib for BRAFV600E sequence variation–positive recurrent or refractory high- and low-grade gliomas, and vorasidenib for IDH-mutant low-grade gliomas.^25,26 Liquid biopsy has the potential to identify potentially targetable genetic alterations not captured in the initial tissue sample. The detection of resistance alterations may also promote improved therapy selection.

Monitoring

The standard approach for monitoring gliomas is through serial MRI scans. Treatment response assessed using imaging is challenging, because MRI scans cannot reliably distinguish between pseudoprogression or treatment-related effects and disease progression, either of which can occur during conventional treatment. A few studies have elucidated the potential application of circulating biomarkers in longitudinal monitoring, given the fluctuation in levels associated with response and recurrence. For example, the assessment of CSF ctDNA levels over time reflects the treatment course and can be used to monitor brain tumor progression.¹¹ Liquid biopsy can also aid in tracking tumor evolution, as exemplified in identifying temozolomide-induced sequence variations in cfDNA.^13,27

Challenges and Future Directions in Liquid Biopsy as Applied to Glioma

The application of liquid biopsy in the real-world setting is limited because of several challenges. Among these challenges is the BBB, which may act as a barrier to the release of brain tumor–derived biomarkers into the circulation. In addition, the rate of shedding throughout the disease course and the influence of tumor characteristics, such as location, size, histology, grade, and amount of enhancing versus nonenhancing disease, remain to be addressed.² To tackle these challenges, other technologies such as MRI-guided focused ultrasound are being explored to induce transient BBB opening and enrich the signal of circulating biomarkers.²⁸ Methodologic challenges, including cost, complexity, and prolonged processing time, also hinder the implementation of liquid biopsy in practice. Clinicians may not avail themselves of real-time results when these biomarkers are used for monitoring. Given the initial data showing that CSF is more enriched than plasma for circulating tumor components, the best site for CSF collection remains to be determined, because LP may not be the most practical monitoring method, and cisternal CSF collection may be more difficult. Peripheral blood draws remain the simpler, less invasive, and more accessible method for liquid biopsy if it to be used for serial monitoring.

Establishment of the Brain–Liquid Biopsy Consortium in 2020 as a global collaborative effort was a step in the right direction to advance biomarker research in neuro-oncology.³ Further efforts are required to develop more sensitive and convenient detection and analysis methods before the role of circulating biomarkers in glioma management can be defined. The validation of its clinical utility in a large prospective clinical trial setting is necessary. Once liquid biopsy has improved sensitivity, its incorporation into the management workflow should be considered. Considering the challenges posed by liquid biopsy, complementary methods, such as advanced imaging techniques, may be necessary, as well as the generation of a composite of clinically reliable biomarkers with improved accuracy to better characterize gliomas and their evolution throughout the course of treatment. These efforts will pave the way for a more effective and personalized treatment approach for people with glioma.

Conclusions and Summary

Despite recent advances in NGS and molecular profiling, no validated circulating biomarkers can be used to diagnose glioma, distinguish glioma recurrence from pseudoprogression, monitor or predict treatment effectiveness and resistance, or prognosticate outcomes. Most studies have limited sample sizes and short follow-up and have used various technologies. Improved sensitivity and specificity of the analysis methods used are needed. Liquid biopsy has the potential to support clinical decision-making and can lead to improved quality of life for people with glioma; further research and development are required to ensure that it can be incorporated effectively into clinical care.