Glioblastoma

Clinical Overview

Glioblastoma (GBM) is the most aggressive primary brain tumor with median survival of 15 months despite maximal therapy including surgical resection, radiation, and temozolomide chemotherapy. ctDNA detection in GBM presents a unique challenge: the blood-brain barrier (BBB) significantly restricts tumor DNA shedding into peripheral circulation, resulting in markedly lower plasma detection rates compared to other solid tumors.

Cerebrospinal fluid (CSF) ctDNA offers superior sensitivity due to direct contact with the central nervous system, achieving detection rates of 83.8-100% compared to plasma detection of 37.5-93.8%. The 41-fold higher variant allele frequency in CSF (23.96%) versus plasma (0.58%) reflects the BBB's profound impact on peripheral blood-based monitoring.

CSF vs Plasma ctDNA Detection: Critical Comparison

The blood-brain barrier fundamentally limits plasma ctDNA detection in glioblastoma. Multiple studies demonstrate CSF's superior sensitivity for detecting tumor-derived mutations, with significantly higher variant allele frequencies enabling more reliable monitoring.

*Note: 93.8% plasma detection achieved with targeted sequencing in Jones et al. 2024 study; lower rates observed with standard methods.

ctDNA Testing Methodology

Tumor-Informed Approach

Uses baseline sample (surgical tissue or baseline blood/CSF) to identify patient-specific mutations, then tracks those mutations at monitoring timepoints. For GBM, baseline tissue from surgical resection typically provides the mutation profile, which is subsequently tracked in CSF or plasma during follow-up.

Advantages in GBM: Knowing which mutations to track improves detection sensitivity, particularly important given the BBB limitation. Tumor-informed approaches can detect variants at variant allele frequencies as low as 0.01%.

Tumor-Agnostic Approach

Tests directly at monitoring timepoint without prior baseline profiling, using fixed gene panels to detect common GBM mutations (TERT promoter, IDH1/2, EGFR, TP53).

Limitations in GBM: Without baseline profiling, detection sensitivity is lower. Given the already-reduced plasma ctDNA shedding due to the BBB, tumor-agnostic approaches may miss clinically significant disease in up to 62.5% of plasma samples.

Sample Type Selection

CSF sampling requires lumbar puncture but provides 41-fold higher VAF and superior detection rates (83.8-100%). Plasma sampling is less invasive but significantly limited by the BBB, with detection rates ranging from 37.5% (standard methods) to 93.8% (advanced targeted sequencing, Jones et al. 2024).

MRD Detection Clinical Utility

Plasma ctDNA Detection Rates

Detection rates in plasma vary significantly based on methodology and disease burden:

Reference: Jones JJ et al. demonstrated 93.8% plasma ctDNA detection in newly diagnosed GBM using tumor-informed targeted sequencing. Despite this improvement, median plasma VAF remained low at 0.58% compared to 23.96% in CSF (Neuro-Oncol Adv 2024;6(1):vdae041).

CSF ctDNA Detection Rates

CSF consistently demonstrates superior detection across multiple studies:

Treatment Response Monitoring

ctDNA changes correlate with radiographic response in GBM patients undergoing therapy. Declining ctDNA levels indicate treatment response, while rising levels precede radiographic progression. However, the BBB limitation means plasma ctDNA may remain undetectable even in active disease, particularly with low tumor burden or intact BBB.

Clinical utility limited by: Pseudoprogression on imaging can complicate interpretation, and plasma ctDNA negativity does not exclude active disease due to BBB restrictions.

Genotyping Clinical Utility

IDH1/2 Mutations

IDH1/2 mutations occur in approximately 5% of primary glioblastomas but are more common in secondary GBM and lower-grade gliomas that progress to GBM. These mutations are associated with improved prognosis and represent actionable therapeutic targets.

Clinical significance: While vorasidenib demonstrates strong efficacy in lower-grade IDH-mutant gliomas, its role in IDH-mutant glioblastoma requires further study. ctDNA can identify IDH mutations non-invasively, but treatment decisions should be based on tissue-confirmed diagnosis and grade.

EGFR and EGFRvIII

EGFR amplification occurs in approximately 40% of GBM, and the EGFRvIII deletion variant is present in 25-30% of EGFR-amplified tumors. Despite being common and targetable mutations, EGFR inhibitors have not demonstrated clinical efficacy in GBM.

Clinical limitation: While ctDNA can detect EGFR mutations, no effective EGFR-targeted therapies currently exist for GBM due to BBB penetration issues. The BBB that limits ctDNA release also prevents most small molecule inhibitors from reaching therapeutic concentrations in brain tumors.

MGMT Promoter Methylation

MGMT promoter methylation status is the strongest predictive biomarker for temozolomide chemotherapy response in GBM. Methylation silences the MGMT DNA repair enzyme, rendering tumors more sensitive to alkylating chemotherapy.

TERT Promoter Mutations

TERT promoter mutations occur in approximately 80% of primary glioblastomas and are associated with telomerase activation and aggressive biology.

Clinical Summary

Bottom Line: The blood-brain barrier fundamentally limits plasma ctDNA detection in glioblastoma, with CSF offering superior sensitivity (83.8-100% vs 37.5-93.8% plasma detection). While advanced targeted sequencing methods improve plasma detection to 93.8%, the 41-fold lower VAF in plasma compared to CSF means negative plasma results cannot exclude active disease. Molecular profiling via ctDNA can identify actionable targets such as IDH mutations and MGMT methylation status, but therapeutic options remain limited by the same BBB that restricts ctDNA shedding.