Which molecular biomarkers must a glioblastoma patient exhibit to be eligible for checkpoint‑inhibitor immunotherapy?

Molecular Biomarkers for Glioblastoma Immunotherapy Eligibility

Glioblastoma patients should be tested for mismatch repair (MMR) deficiency or high tumor mutational burden (TMB-H) to be considered candidates for checkpoint inhibitor immunotherapy, though even with these biomarkers present, treatment should preferably occur within clinical trials or prospective registries due to limited efficacy data. 1

Primary Biomarkers for Immunotherapy Consideration

MMR Deficiency and High TMB

The most recent EANO guidelines (2025) provide specific testing recommendations 1:

• Testing should be considered in:

  • Young adults (<50 years)
  • Tumors with unusual histological or molecular features (e.g., high-grade glioma IDH wild-type with ATRX loss, severe pleomorphism, giant cell features)
  • Tumors with DNA methylation patterns suggestive of MMR deficiency
  • Patients with personal or familial history of germline DNA polymerase or MMR deficiency

• Critical caveat: Even when MMR deficiency or TMB-H is identified in newly diagnosed glioblastoma, prospective and retrospective cohorts have not shown significant response rates to checkpoint blockade (ESCAT IIIA evidence level) 1

Post-Treatment Acquired TMB-H

For recurrent glioblastoma after alkylating agents 1:

• Testing is only relevant in the context of available clinical trials
• Consider specifically for:
  • IDH-mutant gliomas
  • MGMT promoter methylated IDH-wild-type glioblastoma
  • Patients who initially responded to alkylating agents
• Reports of benefit remain anecdotal (ESCAT IIIB)
• Current evidence does not justify biopsy solely for TMB/MMR analysis

Why Standard Checkpoint Inhibitors Have Failed

Negative Phase III Trial Data

The 2021 EANO guidelines explicitly state that checkpoint inhibitor nivolumab showed no overall survival benefit in patients with MGMT promoter-unmethylated glioblastoma 2. Recent meta-analyses confirm this disappointing reality:

• No significant OS benefit (HR = 1.06-1.12) 3, 4
• No PFS benefit or even worse PFS in some analyses 3, 4
• Associated with increased toxicity, particularly headache 3

Frequency of Actionable Biomarkers

Research data reveals the rarity of favorable biomarkers 5:

• High TMB found in only 3.5% of glioblastoma patients (7 of 198)
• High TMB associated with loss of MMR protein expression (MLH1, MSH2, MSH6, PMS2)
• Importantly: Even high/moderate TMB glioblastomas did not show enriched CD8+ T cells, PD-1+ T cells, or tumor PD-L1 expression

Biomarkers Associated with Poor Response

Recent research (2025) identified molecular features predicting worse outcomes with checkpoint inhibitors 6:

• PIK3CA mutation: Associated with significantly shorter OS (p=0.014)
• RB1 mutation: Associated with shorter PFS (p=0.009)
• TERT promoter mutations: Associated with shorter PFS (p=0.053)

Clinical Factors Predicting Poor Response

• Steroid use at baseline: Associated with shorter OS (p=0.0016) 6
• Multifocal disease: Associated with shorter PFS and OS (p=0.0011,0.0015) 6

Practical Clinical Algorithm

For newly diagnosed glioblastoma:

1. Test for MMR deficiency if patient is <50 years OR has unusual histological/molecular features
2. If MMR deficient or TMB-H identified → Enroll in clinical trial (do not use outside trial setting)
3. If standard molecular profile → Checkpoint inhibitors are not recommended outside clinical trials

For recurrent glioblastoma after temozolomide:

1. Consider TMB/MMR testing only if clinical trial available AND patient meets criteria:
   • IDH-mutant glioma, OR
   • MGMT promoter methylated IDH-wild-type, OR
   • Initial responder to alkylating agents
2. Do not perform biopsy solely for this testing
3. Avoid checkpoint inhibitors if patient on steroids or has multifocal disease

Critical Pitfalls to Avoid

• Do not assume PD-L1 expression predicts response: Unlike other solid tumors, PD-L1 expression in glioblastoma does not correlate with checkpoint inhibitor efficacy 5
• Do not use checkpoint inhibitors as standard therapy: Three large phase III trials (CheckMate 143,498,548) failed to show benefit 7
• Do not test for microsatellite instability routinely: While MSI+ cancers respond to anti-PD-1, this is exceedingly rare in glioblastoma
• Recognize the immunosuppressive microenvironment: Glioblastoma has profound myeloid-driven suppression, metabolic barriers (adenosine, arginine depletion, lactate), and T cell paucity that limit checkpoint inhibitor efficacy 7, 8