Multi-omic profiling of 24 glioblastoma patients revealed an average of 107 neoantigen candidates per patient, but almost no shared targets — requiring a minimum 11-peptide cocktail to cover all patients.
107 neoantigen candidates per patientEach glioblastoma patient had over 100 predicted neoantigen targets on average, but almost none were shared between patients — highlighting the extreme personalization needed for brain cancer vaccines.
What the researchers found
Multi-omic analysis of 24 GBM patients identified an average of 148 mutated genes and 200 mutation sites per patient, with no dominant shared mutations across the cohort. An average of 107 neoantigen candidates were predicted per patient. Very few neoantigens were shared by more than two patients, and no dominant shared neoantigen could be identified.
A minimum of 11 peptides was required to create a bulk vaccine covering all 24 patients, ensuring each patient had at least one targetable neoantigen. The tumor immune microenvironment was dominated by NK cells and Th1 cells. TCR/BCR repertoires showed clustered CDR3 sequences in tumors with reduced diversity compared to peripheral blood, suggesting tumor-specific immune responses were already present but limited.
Why it matters
Glioblastoma has a dismal prognosis with median survival around 15 months. Immunotherapy with checkpoint inhibitors has largely failed in GBM. Personalized neoantigen vaccines represent one of the most promising remaining strategies, but this study reveals the fundamental challenge: extreme heterogeneity means each patient needs a bespoke vaccine. Understanding this landscape is essential for designing realistic vaccine strategies that can actually cover patient cohorts.
How the study worked
Integrative multi-omic profiling of 24 GBM patients, including whole-exome sequencing (mutations), HLA typing, TCR/BCR repertoire sequencing, and immune cell component analysis from both tumor tissue and peripheral blood mononuclear cells (PBMCs). Neoantigen prediction was based on mutation identification and HLA binding affinity algorithms. The study was designed to inform a planned GBM clinical vaccine trial.
What this study cannot tell us
The cohort of 24 patients is relatively small and may not capture the full neoantigen diversity of GBM. The study profiled neoantigen candidates computationally but did not validate their immunogenicity through functional T-cell assays. The minimum 11-peptide coverage assumes each patient needs only one neoantigen, which may be insufficient for effective anti-tumor immunity. The study did not assess whether the identified neoantigens would actually elicit therapeutic immune responses.
How to read the evidence
This is a translational genomics study profiling 24 patients to inform clinical trial vaccine design. The multi-omic approach is comprehensive, but the neoantigens are computationally predicted and not functionally validated. No clinical outcomes are reported.
When this study was published
Published in 2025, this is very current research directly informing an ongoing clinical vaccine trial. The neoantigen vaccine field for brain cancer is rapidly evolving.
The bigger picture
This study confronts the central challenge of personalized cancer vaccines: tumor heterogeneity. While neoantigen vaccines have shown promise in melanoma and other cancers with high shared mutation burdens, GBM's lack of shared targets makes the 'off-the-shelf' vaccine approach nearly impossible. The finding that 11 peptides could cover a 24-patient cohort suggests a semi-personalized approach may be feasible — creating small peptide libraries rather than fully individualized vaccines — which could be more practical and scalable than pure personalization.
Questions still open
- Will the 11-peptide bulk vaccine approach be effective in clinical trials, or does each patient need multiple personalized neoantigens for therapeutic benefit?
- Can the clustered tumor-infiltrating TCR/BCR repertoires be used to reverse-engineer which neoantigens are already being recognized by the patient's immune system?
- How does GBM's neoantigen heterogeneity compare to other cancer types, and does this explain why checkpoint immunotherapy has failed in GBM?
Common questions
Why can't scientists make one vaccine that works for all brain cancer patients?
What is a neoantigen vaccine?
Read the original research
Integrative multi-omic profiling of the neoantigen landscape of glioblastoma for the development of therapeutic vaccines reveals vast heterogeneity in immunogenic signatures.
Frontiers in oncology, 15, 1507632
Citation
Lin, Qingtang; Wei, Yukui; Xu, Geng; Wang, Leiming; Ling, Feng; Chen, Xiaojie; Cheng, Ye; Zhou, Yiming. (2025). Integrative multi-omic profiling of the neoantigen landscape of glioblastoma for the development of therapeutic vaccines reveals vast heterogeneity in immunogenic signatures.. Frontiers in oncology, 15, 1507632. https://doi.org/10.3389/fonc.2025.1507632