Preanalytical variables and analytes in liquid biopsy approach for brain tumors: A comprehensive review and recommendations from the RANO Group and the Brain Liquid Biopsy Consortium

Abstract

This review explores the pivotal role of preanalytical variables in bringing liquid biopsy approaches into the clinic for brain tumors. Preanalytical variables encompass a range of critical issues, from blood sample collection and handling to the impact of tumor heterogeneity and patient-specific factors. These variables introduce challenges such as false positives, false negatives, and variability in the analysis of tumor signals, which can hinder the diagnostic and prognostic utility of liquid biopsies. Understanding the nuances of preanalytical variables is essential for the successful implementation of liquid biopsy in clinical settings. This paper delves into strategies aimed at mitigating the influence of preanalytical variables by emphasizing the importance of standardized sample collection protocols, optimized sample processing and storage, quality control measures, and the integration of multiple liquid biopsy modalities.

Liquid biopsy has emerged as a promising minimally invasive diagnostic tool for the detection and monitoring of brain tumors, offering a novel approach that can complement traditional tissue biopsies. Various analytes, including cell-free DNA (cfDNA), circulating tumor cells (CTCs), and extracellular vesicles (EVs), as well as microRNA (miRNA) profiling and proteomic/metabolomic approaches, may be employed. Each of these analytes, combined with the right technologies to analyze them, offers distinct advantages for brain tumor detection, but the accuracy and reliability are influenced by physiologic and preanalytical factors that ultimately affect assay performance and clinical translation.

The field of liquid biomarker discovery in neuro-oncology has advanced at an impressive pace (Table 1).^1–6 Parallel efforts are standardizing the preanalytical phase and establishing common, robust pathways for biofluid collection and analysis. However, the unique aspects of brain cancers, such as limited tumor-derived material in circulation and trace DNA quantities in cerebrospinal fluid (CSF), make it imperative to have preanalytical studies dedicated to neuro-oncology. Furthermore, the regulatory and health economic aspects of liquid biopsy remain incompletely understood.^7–9

Table 1.

Benefits, Limitations, and Sources of Analytes in Brain Tumor Liquid Biopsy

Analyte	Biofluid	Technology	Benefits	Limitations
cfDNA	Blood, CSF (higher yield in CSF)	PCR, next-generation sequencing	Blood: Suitable for serial monitoring CSF: High sensitivity; can be used to evaluate brain tumor mutations and evolution	Blood: Lower tumor DNA fraction and lower sensitivity for brain tumors CSF: Yield can be limited for patients with low-grade/deep-seated brain tumors that do not shed sufficient cfDNA
Circulating tumor cells	Blood, CSF (rare in parenchymal tumors)	EpCAM-based immunoflow cytometry, single-cell sequencing	Blood: Effective method to monitor potential systemic metastasis CSF: Effective for leptomeningeal metastasis detection	Blood: Low detection rates for gliomas; dependent on tumor aggression CSF: Rare in primary brain tumors
EVs	Blood (plasma preferred), CSF	Flow cytometry, ultracentrifugation, size exclusion	Blood: Captures diverse tumor-related biomolecules CSF: Direct brain tumor vesicle detection, stable markers	Blood: Susceptible to contamination because EVs from non-tumor sources can interfere with analysis CSF: Sensitive to handling and temperature
miRNA	Blood (serum/plasma), CSF	Digital droplet PCR, exosome isolation	Blood: miRNAs are highly stable in exosomes and provide insight into tumor gene expression CSF: Higher specificity for identifying brain tumors	Blood: Differs between serum/plasma; some miRNAs can exist outside of EVs, complicating analysis CSF: Lower yield compared with blood
Proteins and metabolites	Blood (plasma preferred), CSF	Mass spectrometry	Blood: Real-time monitoring of systemic changes CSF: Reflects local tumor environment, brain-specific marker sensitivity	Blood: Prone to batch effects and contamination CSF: Requires fast processing because both are prone to degradation

cfDNA, cell-free DNA; CSF, cerebrospinal fluid; EV, extracellular vesicles; miRNA, microRNA.

Preanalytical variables, such as sample handling, processing time, and variations in operating procedures, can introduce inconsistency and bias to analyte detection, particularly when multiple labs are involved (Figure 1). Thus, standardized and transparent preanalytical protocols are crucial to minimize variations and ensure reliable, reproducible results in the analysis of liquid biopsies for brain tumor diagnosis. If successfully integrated into brain tumor management, liquid biopsies will be transformative, enabling noninvasive tumor detection and real-time disease tracking that could enhance cancer care through earlier diagnosis and tailored treatments.

Figure 1.

Overview of the major pre-analytical and analytical steps in neuro-oncology–based liquid biopsy. CSF, cerebrospinal fluid; ctDNA, circulating tumor DNA; NGS, next-generation sequencing; NMR, nuclear magnetic resonance.

The Response Assessment in Neuro-Oncology (RANO) Group and the Brain Liquid Biopsy Consortium have developed a common Task Force of experts to critically review the impact of preanalytical variables on the results of liquid biopsy studies in brain tumors and to propose strategies to mitigate their influence.

Overview: Liquid Biopsy Analytes and Technologies

Cell-free DNA.—

Brain tumors can release cfDNA into the bloodstream, mainly during cell death. cfDNA molecules are mostly short, wrapped around nucleosomes, and of hematopoietic origin.¹⁰ They are present in blood and other biofluids, notably CSF and urine.¹¹ The biology and structure of cfDNA released from cancer differs from that in healthy individuals. cfDNA exhibits the genetic and epigenetic alterations from the cell of origin, and these characteristics can be analyzed with a range of molecular methods based on PCR or sequencing. The fraction of tumor-derived plasma cfDNA varies depending on the cancer type, with gliomas being among the most challenging to detect.¹ The fraction of tumor-derived cfDNA in plasma is typically < 1% in glioma patients due to blood-brain barrier (BBB) limitations. In contrast, CSF directly interfaces with the tumor microenvironment and is significantly enriched for tumor-derived DNA, with studies reporting tumor fractions as high as 10%–50% in CSF compared to <1% in matched plasma from the same patients.^12–15 Therefore, CSF-based liquid biopsy has greater sensitivity than serum-based liquid biopsy for identifying copy number alterations and mutations.^13,14 Additionally, cfDNA in CSF exhibits a higher tumor fraction, which enhances sensitivity for detecting key oncogenic alterations such as copy number variations and point mutations.

CSF-derived circulating tumor DNA (ctDNA) more accurately reflects the genomic alterations of brain tumors than does ctDNA in plasma, with significantly higher sensitivity for detecting somatic mutations, including epidermal growth factor receptor (EGFR) amplifications and isocitrate dehydrogenase (IDH) mutations, making it a superior analyte for molecular profiling and disease monitoring in CNS malignancies.¹⁵ Tumor-derived cfDNA in CSF also exhibits a distinct fragmentation pattern, with a peak length of ~145 base pairs (bp) compared with ~167 bp for non-tumor cfDNA, aiding in tumor identification.¹⁶ Notably, similar alterations in cfDNA fragment size have also been observed in plasma samples from patients with extracranial malignancies and glioma, with an enrichment of shorter fragments associated with tumor-derived DNA.^17–19 This tumor-specific fragmentation bias can be leveraged in both CSF and plasma analyses to enhance signal-to-noise discrimination and improve mutation detection sensitivity.

Methodologies for cfDNA isolation can affect the quantity and quality of DNA extracted. cfDNA degrades rapidly, with a half-life ranging from 2 minutes to 2 hours, necessitating immediate processing and storage in specialized cfDNA-stabilizing tubes to preserve integrity.²⁰ Studies have shown that silica columns recover only ~20%–40% of cfDNA fragments <200 bp, whereas magnetic bead-based methods can recover >60%–70% of these shorter fragments, substantially improving tumor DNA yield.^20,21 Right-sided size selection using bead-based protocols (eg, 0.6–0.8 × AMPure XP ratios) enriches fragments <500 bp and can increase tumor-derived DNA signal by 1.5–2.5×.¹⁷ Right-sided size selection, which enriches for fragments <500 bp, has been shown to enhance glioma cfDNA detection.²⁰

In addition, the sequencing library preparation protocol can affect the fidelity of the starting DNA molecules that are represented in downstream analyses. Double-stranded library preparation improves sequencing sensitivity, particularly for detecting chromosomal copy number alterations.¹⁶ Low-input protocols enable Shallow whole-genome sequencing (sWGS) from as little as 1–10 ng of cfDNA, making it feasible for small-volume CSF samples.¹⁶ sWGS at low coverage (<0.4×) has been successfully used to detect copy number alterations in CSF cfDNA, sometimes identifying alterations that are undetected in tumor tissue, highlighting the potential of CSF-based liquid biopsy for assessing tumor heterogeneity.¹⁶

Standardized PCR or next-generation–based sequencing assays, essential for molecular characterization and commonly used with other diseases, have exhibited low sensitivity in plasma. In CSF, the relative tumor fraction is higher, but the clinical implementation might be affected by the relative invasiveness of sampling. CSF could be used to aid in the detection, prognostication, and monitoring of both primary and metastatic cancers.^15,22 Detection methodologies vary in sensitivity and applicability. Droplet digital PCR (ddPCR) provides highly sensitive cfDNA detection, requiring only 1–5 ng of input DNA, with a variant allele fraction detection limit as low as 0.001%, making it particularly useful for identifying rare tumor mutations.¹² Indeed, ddPCR has been used to detect the H3K27M mutation in CSF, with a 2-fold higher copy number in CSF taken from lateral ventricle than in CSF from lumbar puncture.²³

Next-generation sequencing (NGS) enables broader genomic profiling but requires a higher cfDNA input, which can be limiting for plasma-based analyses.¹² Most NGS panels require ≥10–50 ng of input DNA, though some low-input protocols function with <5 ng.²⁴ However, nanopore sequencing, which reduces error rates to <0.05%, presents a real-time and cost-effective alternative for cfDNA analysis.¹² Additionally, newer NGS-based technologies can detect brain- and spinal cord-derived tumor DNA (CSF-tDNA) in ≤1 mL of CSF.²⁵

Tumor-guided assays, which use mutations identified in a patient’s tumor tissue to inform targeted plasma-based cfDNA detection, can improve sensitivity for identifying rare circulating tumor DNA and are especially useful in follow-up and recurrence surveillance.^11,26 This is particularly relevant in gliomas, where longitudinal tracking of ctDNA has shown that rising plasma cfDNA levels correlate with tumor burden (ρ = 0.77, P = .003) and precede radiographic progression, highlighting its potential as an early biomarker of recurrence.²⁷ In one study, plasma ctDNA detected recurrence in 64% of patients prior to MRI-confirmed progression.²⁷ These approaches rely on detecting key tumor-specific alterations such as hTERT, IDH, H3F3A, and BRAF mutations, or copy number variations like EGFR amplification and TP53 loss.²⁷ Such alterations not only assist in diagnosis but also inform treatment selection and clinical trial enrollment.

In extracranial solid tumors, ctDNA has been used to detect minimal residual disease and predict tumor progression.²⁸ ctDNA analysis has also shown utility in distinguishing pseudoprogression from true tumor progression, especially in the post-treatment setting.^29–31 Emerging CSF-tDNA-based approaches may also aid in staging metastatic disease and detecting early CNS involvement.

Epigenetic alterations, for example, genome-wide methylation, are among the earliest and most widespread genomic changes observed in cancer development, including in CNS tumors.^27–30,32 Recent research indicates that these epigenetic abnormalities can be detected in liquid biopsy samples, showing remarkable stability in body fluids, and mirror the epigenomic landscape from corresponding tumor tissues, rendering liquid biopsy approaches suitable for detecting tumor-specific methylation biomarkers.^33–46 DNA methylation patterns exhibit cell-type and tumor-type specificity, even in tumors sharing similar cell-of-origin lineages.^47–49 For example, Moss et al. developed a cfDNA methylation atlas that enabled tissue-of-origin determination and could be used to classify CNS tumors non-invasively. This characteristic has led to the development of other methylation-based atlases, defining references for diverse cell types and tissue sources in circulating cfDNA.^50–53 The assessment of these epigenetic markers with liquid biopsy enables detection of low circulating levels of tumor DNA and alterations across multiple regions, as well as multiplexed analysis of methylation markers from a single sample.^{35,52,54–57}

Like other analytes, downstream cfDNA methylation is susceptible to the potential impact of preanalytical variables. Therefore, quality control measures and reference materials are recommended to foster reproducibility.^58–60 cfDNA methylation profiling has shown promise in the noninvasive classification of CNS tumors. For instance, the detection of glioma-specific methylation markers such as MGMT promoter methylation, G-CIMP (glioma CpG island methylator phenotype), and SHH-pathway–associated signatures has aided tumor subtyping from CSF samples. In a recent study of pediatric CNS tumors, cfDNA methylation profiling of CSF correctly classified 7 out of 20 samples, with improved accuracy in cases with higher tumor-derived cfDNA fractions.⁶¹ To ensure consistent interpretation across studies, it is critical to document preanalytical variables and adopt standardized operating procedures (SOPs) for methylation-based assays.

Circulating tumor cells.—

Similar to other systemic cancers, the detection of CTCs emerges as one of the liquid biopsy methods for primary and metastatic brain tumors. CTCs be found either in blood or CSF, although the overall levels remain very low. In patients with metastatic cancers, the number of CTCs in blood is associated with the presence of systemic metastases and overall survival.¹⁷ CTCs in the blood are also detectable in patients with brain metastases, however, their presence does not discriminate between systemic and brain metastases. In addition, current CTC detection methods do not allow for discrimination of CTC derived from brain metastases versus primary brain tumor lesions, like HGG. In HGG patients, CTC detection in blood varies from 20% to 70%, dependent on the detection method and aggressiveness of the tumor.¹⁸ Despite this, CTCs are exceedingly rare in individual patients, often numbering only 1–5 cells per 7.5 mL of blood, which underscores the technical difficulty of reliable detection and the need for highly sensitive assays. This rarity limits their immediate clinical utility despite detectable rates across populations. Interestingly, higher CTC levels were detected in the blood of GBM patients post-surgery and during disease progression, with CTCs showing increased mesenchymal features.¹⁹ Although less studied, CTCs have been occasionally identified in anaplastic astrocytomas, diffuse midline gliomas medulloblastomas, and other brain tumor entities.⁶² The dissemination of CTCs is very limited for benign brain tumors, such as low-grade meningiomas, due to low proliferation and relatively intact BBB. CSF-based detection of CTCs is more sensitive than blood. CTCs in CSF are often found in patients with leptomeningeal metastases of solid tumors but remain relatively rare in parenchymal brain metastases or in HGG.⁶³

CTC detection of primary brain tumors or non-epithelial systemic cancer (eg, melanoma) is particularly challenging also due to technological limitations of currently available assays, largely due to a lack of unique molecular markers.⁶³ Several studies with the FDA-approved Veridex CellSearch assay^® or Epithelial Cell Adhesion Molecule (EpCAM)-based immune flow cytometry techniques in CSF showed a higher sensitivity compared to cytology in patients with leptomeningeal metastases derived from epithelial tumors.⁶⁴ EpCAM-based assays cannot be used for primary brain tumors due to the non-epithelial nature of the majority of these tumors. New technologies implementing functional enrichment of glial and mesenchymal CTCs in the CD45-negative fraction are needed to heighten sensitivity.⁶⁵ Due to the lack of universal glioma-specific surface markers and tumor heterogeneity, detection of CTCs in gliomas requires the use of panels integrating mesenchymal, neural, and/or stem cell markers (eg, CD44, cell-surface Vimentin, Nestin, Sox2, A2B5, S100), glioma-subtype specific markers (GFAP, Olig2, EGFR, IDH1-mutant) and pan-cancer markers (Ki-67, telomerase, TERT, aneuploidy).^66–69 CTCs associated with glial malignancies are typically GFAP-positive, S100-positive/Nestin-positive, and negative for CD45, EPCAM, and cytokeratins, reflecting their distinct glial lineage and helping to rule out hematopoietic or epithelial origins as well as certain metastases.^65,70 In the future, brain tumor-specific panels will be needed, detecting also other brain tumor entities. Optionally, detections based on genome-wide or targeted genetic or transcriptomic profiling at the single-cell level may provide more sensitive options by revealing genomic and transcriptomic aberrations.⁷¹ Alternative options enriching for CTCs include size-based microfiltration methods or other microfluidic devices allowing for label-free isolation. Multimodal platforms incorporating CTC detection as one of the readouts may increase the sensitivity of liquid biopsies.

EVs and particles.—

Primary and metastatic brain tumor cells and their stroma release a wide spectrum of multimolecular EVs and extracellular particles (EPs) into biofluids. EVs are heterogeneous, membrane-covered structures often grouped into classical (exosomes, microvesicles, and apoptotic bodies) and nonclassical (autophagic EVs, stress EVs, and matrix vesicles) subsets.^72–74 They differ by biogenesis mechanisms, release pathways, sizes, densities, and compositions.^75–77 EPs, including exomeres and supermeres, are membraneless multimolecular structures <50 nm in diameter, distinguished from EVs by their lack of a lipid bilayer and unique biophysical properties. Several methods have been developed for the isolation and characterization of specific EV/EP subtypes.^78,79 Ultracentrifugation is the most commonly used method, but it results in low purity owing to co-isolation of non-EV proteins and contaminants.⁸⁰ Density gradient ultracentrifugation improves purity by separating vesicles based on buoyant density, but it is time-intensive and can lead to EV loss.⁸⁰ Size-exclusion chromatography preserves vesicle integrity and prevents aggregation but has lower scalability.⁸⁰ Ultrafiltration is a rapid method but can lead to EV deformation and loss of smaller vesicles.⁸⁰ Precipitation-based methods provide high EV yield but introduce polymer contamination that may affect downstream analyses. Microfluidic-based isolation offers high specificity and rapid processing, though its standardization remains a challenge.

The unique qualities of EVs/EPs as biomarkers include the natural combination of informative molecular entities (lipids, proteins, nucleic acids), including oncogenic drivers and their targets.^81–83 Studies have shown that tumor-derived EVs, such as glioblastoma-derived exosomal EGFRvIII mRNA, can serve as real-time, noninvasive biomarkers for tumor detection and molecular subtyping.⁸⁴ Additionally, EVs are uniquely stable, and their surrounding membrane protects their cargo from degradation, improving their reliability as biomarkers.^85–88 EV/EP-based liquid biopsy holds promise in early-stage diagnostics, disease monitoring, treatment response evaluation, and minimal residual disease detection.^81,89–93 RNA cargo within EVs includes mRNAs, miRNAs, tRNAs, rRNAs, small nucleolar RNAs, and long noncoding RNAs, all of which can provide insights into tumor activity.⁷⁷ In addition to RNA and protein cargo, EVs can also carry double-stranded DNA that reflects both genetic and epigenetic alterations present in the original tumor, providing another avenue for noninvasive tumor profiling.⁹⁴

In brain cancer, EVs also represent important regulators of disease progression and may act systemically by crossing the BBB.^95–99 Glioblastoma EVs have been found to carry pro-angiogenic factors such as VEGF and HIF-1α, which promote vascular remodeling and facilitate tumor progression.¹⁰⁰ Patients with glioblastoma have been shown to exhibit 5.5-fold more circulating EVs than healthy controls, and levels correlate with tumor burden and recurrence risk.⁹⁵ Postoperatively, EV levels drop and remain low during stable disease, increasing upon recurrence, sometimes before MRI detects relapse.⁹⁵

EV-mediated tumor microenvironment remodeling and immune modulation are key mechanisms in brain metastasis, reinforcing their role as diagnostic and prognostic markers.⁹⁰ Studies have shown that EV-associated integrins, such as ITGβ4 and ITGαV, contribute to brain metastases by facilitating tumor cell adhesion and migration across the BBB. Additionally, glioblastoma-derived EVs can carry programmed death ligand-1 (PD-L1), which suppresses T cell function and has been associated with poor prognosis and resistance to immune checkpoint inhibitors, highlighting their potential role in predicting immunotherapy response.¹⁰¹ EVs/EPs are more numerous and diverse than CTCs, making them better at recapitulating brain tumor heterogeneity.¹⁰² However, the promise of EV/EP-based biomarkers for brain tumor applications faces several analytical challenges.^82,83 Addressing these challenges requires the implementation of adequate preanalytical measures that may differ from those applied to other liquid biopsy analytes.¹⁰³ These measures are designed to consider the unique nature of EVs/EPs, minimize artifacts (aggregation, contamination, platelet degranulation), and preserve salient features of EV/EP populations (abundance, nature, cargo, heterogeneity).⁷⁸ Therefore, among key preanalytical considerations and reporting requirements are the methods of sample collection (tubes, anticoagulants), EV/EP isolation (filtration, immunocapture, ultracentrifugation), quality controls, analyte stability, and assay standardization.^104,105 Given the increasingly significant role of EVs in liquid biopsy and their growing clinical relevance, specific preanalytical protocols are being developed that depend on the sample source (blood, CSF) and assay purpose.

Isolation of the different EV constituents presents various challenges. The estimation of EV protein concentration may increase with a delay in centrifugation owing to leakage of other intracellular contents. Freeze-thaw cycles can further fragment cells, making them indistinguishable from EVs during flow cytometry, or cause vesicle fragmentation that increases non-EV contaminants, particularly lipoproteins, affecting downstream analysis.^93,106 Though it is almost impossible to completely overcome these challenges, a combination of techniques in exosome isolation, including precipitation, size-exclusion and affinity columns, and rapid storage at −80 °C, may largely mitigate contamination.¹⁰⁷ EV transcriptome seems to be stable at −80 °C even after a long storage period (less stable at −20 °C) and better than circulating free RNAs. Indeed, EV RNA can be protected from nucleases, proving that EV/EP are a coherent source of pathological biomarkers. Additionally, RNAs like miRNAs seem to remain consistent between different EV isolation protocols.¹⁰⁸ Although the CSF has low cell counts, EV integrity is remarkably sensitive to temperature, withdrawal method, and storage conditions. Therefore, protocols should be adapted to the EV content to be analyzed and the expected analytical technology to be used (PCR, sequencing, immunodetection, flow cytometry).

The intricacies of EV preanalytical variables are perhaps best demonstrated in the study of metastatic brain disease, where the role of EVs has been documented at every disease stage: immune modulation, microenvironment manipulation, migration, angiogenesis, intra- and extravasation disruption of the BBB, and creation of the metastatic niche for the survival of metastatic cells.¹ Exosome-based liquid biopsy approaches, such as RNA-based prostate cancer tests, are already in clinical use, highlighting their translational potential.⁹⁰ Although malignant cells do secrete EVs, EVs are also secreted from normal cells and tumor-educated platelets, all of which play different roles in the cancer cascade; however, the exact interactions between these different EVs remain unclear. The study of cancer EVs highlights the variability present among the EVs, analytes, cancers, and organs affected.¹⁰⁹ For instance, STAT 3 protein levels are elevated in exosomes of patients with brain metastases from breast cancer but not from cancers originating in the lung or kidney. Similarly, PD-L1 levels in small EVs are higher in melanoma patients with brain metastasis than in those without CNS dissemination; thus they play a role in staging the disease.¹¹⁰ Interestingly, PD-L1 mRNA derived from tumor-educated platelets is a potential immunotherapy biomarker in non-small cell lung cancer.¹¹¹

A major challenge of using exosomal diagnostics is the number of different methods that are already being used for EV isolation. In a first but significant step, the International Society for EVs has published guidelines on the “minimal information for studies of EVs” (MISEV) that highlight advantages, limitations, and a series of recommendations to apply during the collection and pre-processing of samples to minimize errors and increase reproducibility.⁷³ Standardized methods for EV characterization, such as nanoparticle tracking analysis, dynamic light scattering, tunable resistive pulse sensing, flow cytometry, and mass spectrometry, are critical for ensuring reproducibility.⁹³ Although implementing standardization and strict reporting criteria is justified and accepted, the information sought may ultimately become the key consideration in designing preanalytical sample handling protocols for brain tumors.^73,112,113 The biofluid choice, biosample collection process, volume available, access route, anticoagulants, sample stabilizers, tubes, sample processing, EV/EP purification method, and temperature and duration of storage may all influence the results. The effect of anticoagulants on the cargo of platelet-derived EVs in pro-thrombotic tumors (like glioblastoma) is an obvious target of investigation, but the effect of other drugs on exosomes also needs to be studied. In addition to all these variables, the type of EV, the disease being investigated, and the analyte will also have significant effects on the outcome. Although investigations into exosomes at single-vesicle resolution may be a groundbreaking research discovery, its clinical use requires validation.^114,115

MicroRNA.—

Circulating miRNA has been explored as a potential biomarker for brain tumor diagnosis, prognosis, and tumor response.^116,117 Following SOPs during sample preparation, including avoiding heparin in blood samples and ensuring that miRNA is of sufficient quality/quantity, is crucial. In general, having an external spike-in control is preferable for the analysis.¹¹⁸ Cell-free miRNAs in CSF exhibited higher specificity and possibly greater sensitivity than cell-free miRNAs, reflecting brain proximity.¹¹⁹ For example, miR-21, miR-125b, and miR-222 were significantly upregulated in CSF from glioblastoma patients compared with non-tumor controls, suggesting their potential as high-specificity diagnostic markers.¹²⁰ Exosomal miRNAs may offer advantages because of their greater stability and well-defined, standardized separation and analysis procedures.¹²¹ They are protected from RNase degradation and can be reliably detected in plasma and CSF, making them attractive candidates for liquid biopsy applications.¹²¹ Advances in exosome isolation techniques, such as ultracentrifugation, immunoaffinity pulldown, and density gradient centrifugation, have improved the recovery of EV-associated miRNA, though standardization remains a challenge.¹²² Moreover, miRNAs such as miR-21 and miR-26a are significantly overexpressed in glioblastoma, highlighting their potential as brain tumor biomarkers.¹¹⁹

However, it should be noted that expression patterns differ between CSF and blood and even between plasma and serum.^118,119 Also, a significant proportion of miRNA in the blood is nonvesicular.¹¹⁹ Quantification methods such as ddPCR have been shown to improve reproducibility over qPCR-based approaches, mitigating issues related to preanalytical variability.¹²³ Although plasma-derived miRNAs are widely studied, CSF-based miRNAs may offer greater specificity in diagnosing brain tumors.¹¹⁹ Notably, higher levels of miR-10b and miR-200a are associated with glioblastoma progression, increased tumor grade, and poorer survival outcomes.¹²⁴ Other approaches that may increase sensitivity and early diagnostic accuracy include pan-cancer-immune-related miRNA analysis and a blood-based classification system that uses large-scale serum miRNomics combined with machine learning.^125,126

Current challenges for using miRNAs as circulating biomarkers include establishing biologically relevant values of absolute and relative differential expression. Standardization of isolation methods, normalization strategies, and validation across different detection platforms remains critical for clinical implementation.¹¹⁸ Future studies should focus on the sample source, the subtype of brain tumor, and different populations, as well as different miRNA detection methods and threshold values, to provide rapid prognostic and diagnostic panels.

Proteins and metabolites.—

Proteomics and metabolomics, both reliant on mass spectrometry are powerful tools for interrogating the biological underpinnings of diseases, their susceptibilities, and their responses to therapeutic interventions. Liquid biopsies allow real-time monitoring of a patient’s metabolome and proteome over time. In recent years, high-throughput proteomic and metabolomic platforms have facilitated the analysis of thousands of proteins and metabolites from small sample volumes. Unfortunately, these technologies are costly and often semiquantitative, making them susceptible to batch effects.¹²⁷ Therefore, careful attention must be given to sample processing, including run order and case-control balancing. Additionally, rigorous protocols for sample handling, quality control, and storage are essential, particularly in longitudinal studies that collect prediagnosis samples. CSF samples undergo immediate centrifugation (2000–4000 × g) to remove debris, followed by storage at −80 °C to prevent degradation.^128,129 Protein handling and processing methods include polyvinylidene difluoride (PVDF) 96-well plate binding, solid-phase extraction, and Wessel-Flügge precipitation.^128,129 Quantification relies on ELISA, bicinchoninic acid assays, and high-throughput techniques like label-free quantification and aptamer-based proteomics.^128,129 Liquid chromatography/mass spectrometry and Orbitrap-based mass spectrometry enhance detection sensitivity.^128,129 Quality control measures include pooled technical samples for batch correction, false discovery rate thresholds (typically 1%), hierarchical clustering for validation, and monitoring freeze-thaw cycles and preanalytical variables for accuracy.^128,129 The EPIC-Europe cohort exemplifies how long-term biobanking efforts can facilitate the study of protein and metabolite behavior over time, offering valuable insights for disease monitoring. Factors such as patient demographics, fasting status, time in the freezer, and freeze-thaw cycles must be carefully recorded to ensure accurate comparisons and meaningful biological interpretations.¹³⁰

As these technologies advance, liquid biopsy-based proteomics and metabolomics are transitioning from research applications to clinical practice. Mass spectrometry, bioinformatics, and automated screening tools have been integrated for early disease detection, treatment monitoring, and patient stratification. In neuro-oncology, CSF proteomics is gaining recognition as a minimally invasive alternative to traditional biopsies, with large-scale proteomic analyses demonstrating high accuracy in distinguishing tumor types. However, standardization remains a key challenge, as variability in sample collection methods, cohort differences, and the high cost of comprehensive profiling must be addressed to enable broader clinical adoption.

Recent advances in liquid biopsy-based proteomics and metabolomics have led to the identification of clinically relevant biomarkers for various diseases, particularly in neuro-oncology. CSF proteomics has shown promise in differentiating between brain tumor types, with biomarkers such as GAP43, TFF3, and CACNA2D2 distinguishing glioblastoma multiforme, brain metastases, and CNS lymphoma with high accuracy.¹²⁸ CHI3L1 and GFAP have been validated as glioblastoma biomarkers, correlating with tumor size and survival.¹²⁹ Furthermore, IL-6, galanin, HSPA5, and WNT4 have emerged as potential CSF biomarkers for gliomas, providing insight into tumor progression and therapy response.¹³¹ Proteomic studies have also highlighted the importance of sampling location and post-surgical changes in biomarker interpretation, as CSF proteomes vary significantly based on anatomical source (ventricular, subarachnoid, lumbar) and time since tumor resection.¹³² These findings emphasize the increasing role of liquid biopsy-based proteomics in real-time disease monitoring, treatment stratification, and biomarker discovery.

Emerging technologies and platforms for panel screenings.—

Technical advances, alternative biofluids, and biomolecule exploration mark developments in the field of liquid biopsies. The promise of many such novel developments is an increase in the sensitivity of tumor-derived or -associated biomarkers in patients with brain cancer. Genome-wide genetic and epigenetic approaches allow for the interrogation of DNA in a less biased approach.^25,133 Alternatively, combining various DNA-based analytes, such as detecting somatic aberrations in cfDNA alongside cfDNA fragment size analysis, may also improve current detection limits.¹³⁴ Other circulating nucleic acids, like cell-free mitochondrial DNA, can be leveraged from liquid biopsy to further boost assay sensitivity.¹³⁵ Mitochondrial DNA offers several potential advantages, including higher copy numbers per cell, smaller genome size facilitating analysis, and increased release during cellular stress or apoptosis, making it a sensitive marker in certain tumor contexts.¹³⁵

Apart from expanding or combining existing tools, various novel and innovative tools for glioma diagnostics may also allow for brain cancer treatment monitoring. These include (1) single-cell DNA sequencing to detect rare cancerous cells in the blood or CSF; (2) identification of GFAP-positive monocytes; (3) detection of methylguanine-DNA methyltransferase auto-antibodies; and (4) use of optical and electronic biosensors.^136–139 In addition, it has been shown that circulating blood platelets sequester extracellular-derived EGFRvIII mRNA and that RNA sequencing of blood platelets enables accurate detection of glioblastoma-associated RNA signatures.^140,141 These signatures are distinct from the platelet RNA signatures derived from patients with brain metastasis or multiple sclerosis and may enable glioblastoma progression to be distinguished from pseudoprogression or radionecrosis. Many of these approaches remain experimental and require independent and thorough validation studies.

Another approach is to increase levels of cfDNA by temporarily opening the BBB, which normally prevents the passage of many analytes into the peripheral circulation.¹⁴² Noninvasive techniques, such as focused ultrasound (FUS), may open the BBB in a spatially targeted manner and increase cfDNA, ctDNA, and EV levels.^143,144 Ongoing clinical trial efforts are evaluating the ability of BBB disruption via FUS to increase the quantity of analytes in the peripheral circulation (NCT05383872, NCT05281731). Intratumoral fluid obtained from tumor resection cavities also provides a potential source for cfDNA; levels nearer to the tumor are higher because the BBB is bypassed.^145,146 An ongoing trial is evaluating the use of an Ommaya reservoir for longitudinal sampling (NCT06322602).

Preanalytical Variables in Liquid Biopsy

Blood collection and handling.—

Blood is the most widely investigated biofluid, but blood component and collection tube choice significantly affect yield and analyte quality. Best practices emphasize immediate processing or stabilized collection methods to prevent analyte degradation and contamination. CTCs require specialized preservation tubes to maintain stability for 72–96 hours. EDTA and citrate tubes require immediate processing because they lower cfDNA stability to 2–6 hours.^147–149 Preservative-containing tubes extend cfDNA stability for up to 14 days, reducing transport constraints.^147,149 Plasma is preferred for ctDNA analysis because it minimizes WBC DNA contamination.¹⁴⁷

Plasma reduces platelet-derived EVs compared with serum, and ACD-A/EDTA tubes have the greatest performance for samples processed within ≤8 hours.^150,151 Centrifugation protocols and filtration steps have been optimized to improve EV purity. To ensure sample quality, long-term storage requires freezing at –80 °C with minimal freeze-thaw cycles.^147,149,152 Double centrifugation of cfDNA and EV samples improves sample purity^120,122 before further processing.^73,149,153 For maximal performance, CTCs should be processed immediately. Reporting guidelines now provide standardized workflows for CTC isolation and enumeration.^71,154,155

CSF collection and handling.—

CSF biomarker sensitivity depends on the collection site. Ventricular and cisternal CSF yield more tumor-derived material than lumbar puncture, particularly for brain parenchymal tumors.^22,156,157 Standardized protocols should account for shunt presence, which may alter biomarker composition. CSF should be processed within 30 minutes to 2 hours to prevent cell degradation and should be separated into supernatant and pellet fractions.¹⁵⁸ Although liquid biopsy is valuable in pediatric oncology, the lack of specialized low-volume cfDNA tubes limits feasibility in infants owing to vein collapse and incompatibility with conventional aspiration systems.¹⁵⁹

Whole CSF provides a comprehensive sample but may contain cellular debris that interferes with analysis. The supernatant, obtained after centrifugation, is preferable for ctDNA, proteins, and miRNAs because it provides higher sensitivity.¹⁵⁷ The pellet contains CTCs and tumor-derived components. Advanced ultracentrifugation techniques, such as iodixanol-based density gradient ultracentrifugation, have been shown to enhance the purity of EV preparations by effectively separating EVs from contaminants like lipoproteins and plasma proteins. Additionally, precipitation-based methods utilizing polyethylene glycol (PEG) have emerged as cost-effective alternatives for EV isolation, offering higher yields suitable for downstream analyses, albeit with considerations regarding purity.^160,161 Biobanked CSF should be stored at –80 °C with minimal handling to prevent RNA degradation.

Sample contamination and quality control.—

Hemolysis compromises blood-derived analytes by introducing variability in concentration and stability, primarily by introducing contaminants like genomic DNA from ruptured WBCs, which can degrade sample quality and hinder data analysis.^162–166 Preanalytical workflows now minimize hemolysis through gentle handling, controlled centrifugation speeds, and reduced sample agitation. Collection tube choice, transport conditions, and processing techniques also influence sample stability and quality.¹⁶⁷ The collection tubes appear to directly affect the quality of the sample, including circulation of miRNA, cfDNA, CTCs, and EVs.^65,168,169 The use of standardized anticoagulant tubes and immediate processing has been shown to preserve sample integrity, particularly for cfDNA analysis. Long-term storage blood collection tubes, such as Streck or PAXgene, contain stabilizing agents that prevent white blood cell lysis and preserve cfDNA integrity for several days at room temperature. In contrast, standard EDTA tubes lack these preservatives and require rapid processing to avoid contamination with genomic DNA from lysed cells.¹⁷⁰ Automated hemolysis detection and removal strategies further improve liquid biopsy reliability. Quality control measures must include standardized sample collection, processing, and detection metrics, validated by clinical trials.^6,18,171

Tumor heterogeneity and genetic variability.—

Brain tumors, particularly diffuse gliomas and metastases, exhibit extensive intratumoral heterogeneity at the genomic and transcriptional levels.^172–176 Recent advances in liquid biopsy have demonstrated the potential to capture this heterogeneity more effectively than traditional biopsy approaches, but its success depends on adequate analyte recovery.¹⁷⁷ Moreover, single-cell transcriptional approaches exhibit notable heterogeneity at the level of gene expression programs, dictating complex developmental cell state and transitional plasticity in response to therapeutic intervention.^178–181 New sequencing methodologies, including single-cell RNA sequencing and epigenetic profiling, are being integrated into liquid biopsy research to improve the resolution of tumor heterogeneity.¹⁸² However, the effectiveness of these methods will depend heavily on the extent to which relevant biomaterial (eg, DNA, RNA, tumor cells) can be effectively recovered from the liquid biopsy substrate. In this respect, CSF-based profiling outperforms plasma in assay sensitivity, making it the preferred biofluid for detecting tumor-specific alterations.^{25,134,156,183,184}

Sample collection timing.—

Liquid biopsy timing should align with the clinical objective, whether for early tumor detection, minimal residual disease assessment, or monitoring therapy resistance. The diverse biology of CNS tumors and complexities in biomarker detection challenge standardizing liquid biopsy collection timepoints. Emerging evidence suggests that serial sampling enhances biomarker-based disease monitoring, necessitating study designs tailored to each tumor’s behavior and methodology.⁷ Different treatment modalities may require distinct sampling schedules—immunotherapy trials, for instance, may necessitate different timepoints than those monitoring radiation or chemotherapy response. Sampling before surgery, chemotherapy, or radiation enhances the identification of actionable mutations. Preoperative liquid biopsies for diagnostic accuracy and molecular profiling should ideally be collected at neuroimaging visits, clinic appointments, or on the day of surgery.⁴

Routine blood draws facilitate peripheral inflammatory biomarker integration (eg, Systemic Immune Inflammation Index, C-reactive protein, neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio), offering insights into tumor biology, grade, and prognosis with the minimal logistical burden.^4,184–188 Such longitudinal sampling should be integrated with clinical follow-ups and imaging to minimize patient burden.²⁹ Virtual visits complicate sample collection, requiring logistical adaptations. Proper staff training will ensure adherence to protocols and prevent processing delays, while factors like staff availability and transport logistics will impact timepoint feasibility.

The clinical relevance of immediate postoperative liquid biopsies remains uncertain Indeed, Nørøxe et al. reported fluctuating cfDNA in glioblastoma, with peaks before surgery and at progression, and nadirs post-surgery.²⁹ Moreover, surgical manipulation can affect the microenvironment of the residual tumor and in some instances promote tumor recurrence.¹⁸⁹ After gross total resection, some glioblastoma patients experience rapid early progression (REP), which is associated with poor outcomes.¹⁹⁰ No current imaging or molecular biomarkers reliably predict REP, making it an area of active investigation. Immediate postoperative liquid biopsies, followed by a second collection before adjuvant therapy, may help identify patients at higher risk for REP.

Patient-specific factors.—

Tumor characteristics, treatment history, and patient demographics influence liquid biopsy yield. Higher-grade tumors, leptomeningeal disease, and lesions near ventricular surfaces release more ctDNA into CSF, making CSF-based liquid biopsy more effective for these cases.¹⁹¹ In contrast, lower-grade gliomas and deeply located tumors may yield insufficient ctDNA for detection. Additionally, patient comorbidities, lifestyle factors, and medications must be accounted for to mitigate confounding variables in high-throughput analyses.^192,193

Treatment effects also play a role. Radiotherapy can increase ctDNA release by disrupting the blood-tumor barrier, particularly in the initial months post-treatment, whereas anti-angiogenic therapies like bevacizumab and corticosteroids may reduce permeability, leading to lower biomarker yield.

Systemic factors, including inflammation and immune suppression, are now being actively studied in relation to their impact on biomarker levels and liquid biopsy performance. Patients with glioblastoma multiforme often exhibit profound peripheral immunosuppression, with decreased circulating T-cell counts and altered immune cell phenotypes.^194–197 Additionally, cfDNA levels were shown to increase in both glioma patients and mouse models, but the functional significance of this finding remains unclear.^27,198,199 Additional research is needed to clarify the interactions between systemic immune modulation and liquid biopsy performance in brain tumors.

Strategies to Mitigate Preanalytical Variables

Errors in the preanalytical phase of liquid biopsy, from collection to storage and transport, can significantly affect data accuracy, feasibility, and clinical decisions.^200,201 The RANO group and various international consortia (eg, UKE-ELBS, CancerID, SPIDIA, BLOODPAC, ILSA) emphasize the need for standardized, reproducible workflows with clear SOPs to ensure high-quality sample processing.^202–208 These efforts have culminated in preanalytical standards provided by the International Standards Organization (ISO) and Technical Specifications from the European Committee for Standardization (CEN/TS). SO 20186-3:2019 provides a standard for isolating cfDNA from plasma, whereas 1CEN/TS 17390-3:2020 focuses on specifications for analytical staining of CTCs.^209–211 These standards govern the entire workflow, documentation, collection, processing, and storage and should be updated periodically to reflect advancements in liquid biopsy.

Standardized blood and CSF collection protocols.—

In routine clinical practice, biosample collection is prone to various preanalytical errors, such as inadequate collection tubes, underfilling of collection tubes, improper tube inversion, extended transport times, and suboptimal temperature conditions during transport. Standardization efforts, such as those outlined in Table 2, provide a structured framework for ensuring sample integrity at every step. The ISO and CEN/Ts standards aim to address preanalytical variation (Table 2); however, investment in personnel and resources is essential to ensure implementation.²¹²

Table 2.

Real-World Application of Essential Criteria for Assigning Liquid Biopsy Specimens as ISO or CEN/Ts Compliant^a

Step	Requirements
Collection criteria at the clinical collection site
Clinical data	Cancer diagnosis OS/disease stage/treatment at the time of collection
De-identifying process	Patient ID to study ID
Collecting tubes	Dedicated tubes for specific liquid biopsy specimen collection with consideration for downstream analysis and time from collection to processing
Collection process	Identify personnel collecting specimen Use standardized labeling procedure (temperature/chemical resistant) Consider which anatomical area to sample (eg, CSF proximal to the tumor or lumbar puncture). Adjuncts required (eg, focused ultrasound). Follow the manufacturer’s instructions for specimen collection/tube-filling Follow instructions for mixing (gentle to avoid cell destruction)
Documentation	Date and time of sample collection and person responsible for collection Evidence of tampering with specimen documented
Temporary storage at the clinical collection site
Storage process	Follow the manufacturer’s instructions for specimen storage Ensure the specimen is not frozen or shaken strongly during storage
Documentation	Storage conditions, duration, and any deviation/evidence of tampering
Transport from the collection site to the laboratory
Transport process	Follow the manufacturer’s instructions for transport conditions Ensure the specimen is not frozen or shaken strongly during transport
Documentation	Storage conditions, duration, and any deviation/evidence of tampering
Specimen reception in the laboratory
Check	Correct identity of specimen
Documentation	Date, time, and name of personnel receiving the specimen Any deviation (transport issues, broken tubes, tampering)
Specimen storage in the laboratory
Documentation	Storage temperature between specimen reception and sample processing Time interval between receipt and sample processing Total storage duration; ensure maximal storage duration is not exceeded Use a monitored freezer, minimize freeze/thaw cycles, and document number Ensure traceability of all samples

^aAdapted from ([Bonstingl et al., 2023]; ISO 20186-3:2019 [Dagher et al., 2019] and CEN/TS 17390-3:2020 [140, 141]). CSF, cerebrospinal fluid; ID, identification; OS, Overall Survival.

Sample processing and storage optimization.—

Blood collection in Streck tubes enhances cfDNA stability, particularly when processing is delayed. Streck tubes contain a proprietary, formaldehyde-free preservative that stabilizes nucleated blood cells and prevents cell lysis, thereby minimizing genomic DNA contamination and preserving cfDNA integrity for up to 7 days at room temperature. For samples stored at room temperature beyond 24 hours, leukocyte stabilizing tubes are recommended.¹⁵⁹ Plasma is preferred over serum for ctDNA analysis because it has less contamination from cellular debris.¹⁵⁹

Processing protocols are variable but are characterized by 2 centrifugation steps, the first at a lower speed, and the second at a higher speed either at room temperature or at 4 °C.^{23,159,213,214} During the first centrifugation step, a smooth braking profile is used to prevent disruption of the buffy coat. Plasma should be aliquoted and stored at –80 °C. After the second centrifugation step, serum can be used if it is cooled during transport and processed within 24–72 hours.¹⁵⁹

For genomic cellular tumor DNA (>0.1% tumor cells) and phenotypic analysis, EDTA or cfDNA BCT CE tubes (Streck, La Vista, NE) are recommended. Genomic DNA isolated from leukocytes can be used for germline studies and as reference DNA for NGS approaches with samples stored for up to 4 hours at room temperature or a maximum of 24 hours at 4 °C (recommended).¹⁵⁹ For CSF cfDNA analysis, samples should be stored on ice and centrifuged within 30 minutes at 1000 × g for 10 minutes at 4 °C. The supernatant can be stored at –80 °C, and the pellet can be used for immune cell characterization.^168,215

Quality control measures.—

All samples should be coupled with clinical data but should be de-identified. Plasma and circulating biomarkers should be stored at –80 °C with minimal freeze-thaw cycles in secure, temperature-controlled freezers with tracking software to maintain sample integrity and support research collaborations.¹⁵⁷ Collaboration with neurosurgeons, pathology, and other critical teams helps to ensure optimal sample collection and processing.

The quality and results from liquid biopsy can be influenced by patient and tumor factors. cfDNA concentration varies between patients and is typically higher in cancer patients than in healthy individuals. However, trauma, strenuous exercise, auto-immune disease, and pregnancy can also elevate cfDNA levels. Significant variation in blood cfDNA level is also seen between different cancer types and between patients with the same disease burden. Typically, the fraction of cfDNA fragments in total cfDNA varies from <0.01% to 10%, according to factors such as tumor metabolism or cancer burden.^216,217

Future Prospects for Reducing Preanalytical Impact

Currently, there are no optimized SOPs for liquid biopsy specimen collection and analysis in brain tumor patients. The preanalytical phase is the most vulnerable to error. Variability in sample collection, processing steps, sample freezing and thawing, and tube selection with associated anticoagulants affect results and need to be optimized in dedicated studies. Delays to sample processing, the impact of long-term storage, and optimal processing methods also need to be defined.

Dedicated studies are needed to test the relative benefits and shortcomings of each biofluid and technology and to identify which analyte(s) is most relevant for disease diagnosis, tracking of residual disease, treatment monitoring, and prognosis. Placing these analytes in the broader context of patient- and tumor-related factors will be essential. Significant progress is needed to minimize preanalytical variables and ensure that liquid biopsies from patients yield valid and reproducible results. Once the most critical preanalytical variables are known for each analyte, then optimal handling and analysis SOPs can be developed. Currently, prospective longitudinal patient cohorts are lacking for liquid biopsy validation. Opportunities may exist to establish multicenter patient cohorts that can be interrogated for sample quality and analyte detection suitability and used to solve existing challenges in preanalytical variable optimization. Although academic groups will likely innovate novel biomarkers and technologies for analyte detections, partnerships with industry are needed to accelerate the translation of liquid biopsies to clinical practice. Industry partners can improve and automate processes, reduce costs at scale, provide valuable analytical validation, and financially support larger clinical studies. Without industry partnerships these promising technologies will never reach patients at scale.

Consideration of Preanalytical Variables in Clinical Study Design

Effective design of liquid biopsy clinical studies requires defining the clinical application (eg, screening, diagnosis, minimal residual disease, or treatment monitoring), ensuring an adequate sample size for generalizability, and selecting an optimal study population.¹¹³ Properly matched controls, including age- and sex-matched healthy individuals and patients with non-neoplastic conditions, are essential to minimize bias and account for the chronic inflammatory state of tumors, which may introduce genetic and epigenetic alterations in the tissue of interest.²¹⁸ To date, most biomarker studies on gliomas have lacked any type of control and sensitivity values have been obtained by a simple comparison with tumor tissue.⁶ Additionally, assays designed to detect somatic mutations in the blood must exclude clonal hematopoiesis of indeterminate potential, which can mimic tumor-derived mutations.²¹⁹

Preanalytical variables must be documented and accounted for, as they significantly influence biomarker yield. HGG and brain metastases release more ctDNA into CSF, especially when tumors are near CSF reservoirs, have greater volume, or exhibit subarachnoid dissemination.¹⁹¹ In contrast, lower-grade gliomas and tumors at diagnosis yield less ctDNA.¹ Biomarker levels also vary on the basis of physiologic and environmental factors such as metabolic conditions, smoking, pregnancy, diurnal variation, and ongoing treatments, all of which must be cataloged in clinical studies. cfDNA levels, for example, peak during nighttime, physical activity, and pregnancy, and inflammation, anemia, and heart disease can further contribute to cfDNA release.

For CSF-based liquid biopsy, Ommaya reservoirs facilitate longitudinal sampling and allow biomarker validation alongside neuroimaging and clinical assessments.¹²⁸ Incorporating standardized protocols and recognizing these preanalytical influences will improve study reproducibility and ensure meaningful clinical translation of liquid biopsy approaches.

Conclusion

The lack of biologically relevant biomarkers is a critical unmet need in neuro-oncology as a whole. Liquid biopsy approaches provide an exciting opportunity across a myriad of technologies, biofluids, and analytes to address these needs. To bring these approaches reliably into the clinical realm, a clear understanding of preanalytical variables and their impact on technology and associated analytes is essential. Detailed metrics to monitor, study, and eventually standardize preanalytical approaches will be critical for reducing variability and increasing the performance of liquid biopsy approaches. Incorporating rigorous documentation and testing of preanalytical variables will be required in future prospective studies to harness the full potential of liquid biopsies in patients with tumors of the CNS.

Funding

None.

Authorship statement

The review was conceptualized by C.B., H.N., S.S., and R.S.. The other coauthors wrote single paragraphs. All coauthors reviewed and approved the final manuscript.

Conflict of interest statement

Chetan Bettegowda: CB is a consultant for Bionaut Labs, Haystack Oncology, and Privo Technologies and is a co-founder of Belay Diagnostics and OrisDx. Houtan Noushmehr: No conflict of interest. Alessandra Affinito: No conflict of interest. Manmeet S. Ahluwalia: MSA is a consultant for Allovir, Anheart Therapeutics, Apollomics, Autem Therapeutics, Bayer, Cairn Therapeutics, EquilliumBio, GT Medical Technologies, Incyte, Menarini Ricerche, QV Bioelectronics, Recordati, Servier Pharmaceuticals, Sumitomo Pharma Oncology, Theraguix, Viewray, and Xoft; is on the data safety and monitoring committee for VBI Vaccines; is on the Scientific Advisory Board for Modifibiosciences and Bugworks; and is a shareholder of Cytodyn, MedInnovateAdvisors LLC, Mimivax, and Trisalus Lifesciences. MSA has also received the following funding: 1R01CA277728-01A1, 1R01 CA264017-01A1. Olaf Ansorge: Co-Founder and consultant for EviCure, Ltd. Katayoun Ayasoufi: No conflict of interest. Stephen Bagley: SB has consulted for Bayer, Modifi Bio, Servier, and Telix and has received research grants (to institution) from GSK, Incyte, Kite, Lilly, and Novocure. Jill Barnholtz-Sloan: No conflict of interest. Myron Best: No conflict of interest. Dieta Brandsma: No conflict of interest. Chaya Brodie: No conflict of interest. Anke Brüning-Richardson: No conflict of interest. Ana valeria Castro: No conflict of interest. Gerolama Condorelli: No conflict of interest. Ahmad Daher: No conflict of interest. Vineet Datta: No conflict of interest. John De Groot: JDG is on the advisory boards of Alpha Pharmaceuticals, CapitalOne, Carthera, DSP Pharma, Insightec, Kazia, Kintara Pharmaceuticals, Monteris, Mundipharma, Nerviano, Samus, Sapience, Servier, and Telix; is on the data safety and monitoring board for Chimerix and VBI; and consults for Carthera, Deciphera, Insightec, Kazia, Kintara, MundiPharma, and Nerviano. Pim French: No conflict of interest. Evanthia Galanis: EG is a past advisory board member of Boston Scientific (DMC, compensation to employer <$10 000), Karyopharm Therapeutics, Inc (DSMB, compensation to employer <$10 000), and Kiyatec, Inc (personal compensation <$10 000); is a current advisory board member of Boehringer Ingelheim (compensation to employer <$10 000), Modifi Biosciences (compensation to employer <$10 000), and Servier Pharmaceuticals (compensation to employer <$10 000); is on the Education Steering Committee of Servier Pharmaceuticals (compensation to employer <$10 000/yr), and has received Grant/Research/Clinical Trial Funding (to Mayo) from Servier Pharmaceuticals LLC and Denovo Biopharma. Anna Golebiewska: No conflict of interest. Petra Hamerlik: No conflict of interest. C. Oliver Hanemann: No conflict of interest. Matthias Holdoff: No conflict of interest. Jason Huse: No conflict of interest. Mustafa Khasraw: MK has received grants paid to his institution or contracts from AbbVie, BioNTech, BMS, CNS Pharmaceuticals, Daiichi Sankyo Inc, Immorna Therapeutics, Immvira Therapeutics, JAX lab for genomic research, and Personalis Inc. MK also received consulting fees from AnHeart Therapeutics, George Clinical, Manarini Stemline, and Servier and is on a data safety monitoring board for BPG Bio. Suzanne LeBlang: No conflict of interest. Beatrice Melin: No conflict of interest. Florent Mouliere: FM is co-inventor on patents related to cell-free DNA sequencing; has consulted for Roche Dx; and has received material support from Biomodal. Claire O’Leary: No conflict of interest. Janusz Rak: No conflict of interest. Amitava Ray: AR is an employee of Exsegen Genomics, a for-profit organization that is currently working on liquid biopsy for brain tumor diagnostics. Stephen Robinson: No conflict of interest. Ola Rominiyi: No conflict of interest. Federico Roncaroli: No conflict of interest. Roberta Rudà: No conflict of interest. Joan Seoane: No conflict of interest. Nik Sol: No conflict of interest. Martin J.van den Bent: No conflict of interest. Michael A. Vogelbaum: MAV has received clinical trial funding for his institution from DeNovo, Infuseon, and Oncosynergy and has received honoraria from Alexion, Biodexa, and Servier. Tobias Walbert: No conflict of interest. Colin Watts: No conflict of interest. Tobias Weiss: TW received honoraria from Philogen S.p.A. Michael Weller: MW has received research grants from Novartis, Quercis, and Versameb and honoraria for lectures, advisory board participation, or consulting from Anheart, Bayer, Curevac, Medac, Neurosense, Novartis, Novocure, Orbus, Pfizer, Philogen, Roche, and Servier. Patrick Y. Wen: PYW has received research support from Astra Zeneca, Black Diamond, Bristol Meyers Squibb, Chimerix, Eli Lily, Erasca, Global Coalition For Adaptive Research, Kazia, MediciNova, Merck, Nerviano, Novartis, Philogen, Quadriga, Servier, and VBI Vaccines and has served on the advisory board or as a consultant to Anheart, Alexion/Astra Zeneca, Black Diamond, Chimerix, Day One Bio, Fore Biotherapeutics, Genenta, Glaxo Smith Kline, Kintara, Merck, Mundipharma, Novartis, Novocure, Nuvation Bio, Prelude Therapeutics, Sapience, Servier, Symbio, Tango, Telix, and VBI Vaccines. Victoria Wykes: No conflict of interest. Stephen Yip: SY is an advisory board member and has received honoraria from Amgen, AstraZeneca, Bayer, Pfizer, Servier, and Roche. Susan Short: No conflict of interest. Riccardo Soffietti: No conflict of interest.