Neuro-oncology is the medical specialty focused on cancers of the brain and spinal cord, and it sits at one of the most difficult intersections in medicine. The brain is simultaneously the most protected and the most vulnerable organ when a tumor takes hold: the same barriers that shield it from infection also block drugs from reaching cancer cells, and the same neural circuitry that makes the brain indispensable makes aggressive surgery risky. Over the past decade, the field has been reshaped by molecular diagnostics, targeted drugs, and a surprising discovery that gliomas can wire themselves into functioning brain circuits. Understanding what neuro-oncology looks like today means understanding why these tumors are uniquely stubborn and where the most promising cracks in their defenses have appeared.
A New Way of Naming Brain Tumors
For most of the twentieth century, brain tumors were classified almost entirely by how they looked under a microscope. Two tumors that appeared identical might behave very differently in different patients, and oncologists often couldn’t explain why. The 2021 edition of the World Health Organization’s classification of central nervous system tumors changed this by making molecular markers central to diagnosis rather than optional add-ons. Specific genetic mutations and chromosomal features now determine a tumor’s official name and grade, not just its microscopic appearance.
1PubMed Central. The 2021 WHO Classification of Tumors of the Central Nervous System: a summaryThe practical consequence is significant. A tumor that previously might have been called a “grade II astrocytoma” based on its appearance alone could now be reclassified at a higher grade if it carries certain molecular features, even before it starts behaving aggressively on imaging. For IDH-mutant gliomas in particular, molecular features can affect grading for the first time, meaning two tumors that look identical under the microscope may receive different grades and therefore different treatment plans based on their genetic profile.
2PubMed Central. Updates on the WHO diagnosis of IDH-mutant gliomaThis shift matters for patients because it improves treatment matching. A neurosurgeon considering how urgently and aggressively to intervene now has molecular data pointing toward a tumor’s likely trajectory, rather than relying solely on what a pathologist sees through a microscope. It also standardizes communication: the same tumor name in Tokyo, Toronto, and São Paulo now refers to the same molecular entity, not just the same visual pattern.
3PubMed Central. The WHO 2021 Classification of Central Nervous System tumours: a practical update on what neurosurgeons need to know—a minireviewWhy Brain Tumors Are So Hard to Treat
Three biological features conspire to make brain cancers, especially glioblastoma, extraordinarily resistant to treatment: a physical barrier that blocks drugs, a population of stem-like cells that survive therapy, and an immune environment that actively suppresses the body’s defenses.
The Blood-Brain Barrier
The blood-brain barrier is a tightly sealed layer of cells lining the brain’s blood vessels. It exists to keep toxins, pathogens, and large molecules out of brain tissue, and it does this job well enough to block the vast majority of chemotherapy drugs. In and around brain tumors, this barrier is often partially disrupted, creating what researchers call the blood-tumor barrier, but that disruption is inconsistent: parts of a tumor may be reachable by drugs while other regions remain sealed off. Strategies under investigation to get around this barrier include focused ultrasound to temporarily open the barrier in targeted areas, nanoparticle delivery systems, and drugs redesigned to be small or lipid-soluble enough to cross on their own.
4PubMed Central. Challenges and opportunities to penetrate the blood-brain barrier for brain cancer therapyGlioma Stem Cells
Within a glioblastoma, a small subpopulation of cells behaves like stem cells: they divide slowly, can regenerate the entire tumor, and are resistant to both radiation and chemotherapy. These glioma stem cells tend to sit in a quiescent state that makes them invisible to treatments designed to kill rapidly dividing cells. When the bulk of the tumor is destroyed by surgery, radiation, and temozolomide, these survivors can reawaken and seed a recurrence. The failure to eliminate this specific subpopulation is widely considered a major reason why glioblastoma almost always comes back.
5PubMed Central. The role of glioma stem cells in chemotherapy resistance and glioblastoma multiforme recurrence6PubMed. Hunting glioblastoma recurrence: glioma stem cells as retrospective targets
An Immunosuppressive Microenvironment
Most solid tumors provoke at least some immune response, but glioblastoma is considered immunologically “cold.” The tumor tissue is heavily infiltrated by immune cells called microglia and macrophages, but instead of attacking the tumor, these cells are co-opted by it. They release signals that suppress T cells and promote tumor growth. When immune T cells do manage to enter the tumor, they are typically dysfunctional or exhausted, unable to mount an effective attack.
7PubMed Central. Deciphering the Labyrinthine System of the Immune Microenvironment in Recurrent Glioblastoma: Recent Original Advances and Lessons from Clinical Immunotherapeutic Approaches8Neuro-Oncology. Immunosuppressive tumor-infiltrating myeloid cells mediate adaptive immune resistance via a PD-1/PD-L1 mechanism in glioblastoma
The Standard Treatment Backbone
For newly diagnosed glioblastoma, the standard of care has remained largely the same for two decades: maximum safe surgical removal of the tumor, followed by radiation with concurrent temozolomide chemotherapy, and then additional cycles of temozolomide alone. This regimen extends survival, but the gains depend heavily on a molecular feature of the tumor.
Temozolomide works by damaging tumor DNA, but tumors can repair that damage using an enzyme called MGMT. When the gene for MGMT is silenced through a chemical modification called promoter methylation, the tumor loses its main repair tool and becomes much more vulnerable to temozolomide. A landmark study found that patients whose tumors had a methylated MGMT promoter survived a median of about 22 months with combined treatment, compared with about 15 months with radiation alone. Without that methylation, the survival difference between adding temozolomide and using radiation alone was smaller and not statistically significant.
9PubMed. MGMT gene silencing and benefit from temozolomide in glioblastomaMGMT status is now routinely tested, but using it to make treatment decisions remains imperfect. Some patients with methylated tumors still don’t respond well to temozolomide, and some with unmethylated tumors do better than expected. Part of the discrepancy comes from the fact that methylation of the gene and actual expression of the repair enzyme don’t always line up cleanly.
10PubMed Central. MGMT Status as a Clinical Biomarker in GlioblastomaSurgical Advances and Intraoperative Tools
Removing as much tumor as safely possible improves outcomes, but the challenge with gliomas is that they don’t have clean borders. Tumor cells infiltrate normal brain tissue in ways that are invisible to the naked eye. Two technologies have meaningfully improved a surgeon’s ability to distinguish tumor from healthy brain during an operation.
5-aminolevulinic acid, or 5-ALA, is an oral agent taken before surgery that causes tumor cells to fluoresce under blue light. This fluorescence lets surgeons see residual tumor tissue that would otherwise look identical to normal brain, significantly increasing the rate of complete resections compared to conventional white-light surgery.
11PubMed Central. What is the Surgical Benefit of Utilizing 5-Aminolevulinic Acid for Fluorescence-Guided Surgery of Malignant Gliomas?Intraoperative MRI lets surgeons scan the brain during the procedure itself, checking whether tumor remains before closing. When 5-ALA and intraoperative MRI are combined, both groups achieved similarly high resection rates, though one study found comparable median overall survival of roughly 17 to 18 months whether surgeons used 5-ALA alone or the combination. The value may lie less in raw survival statistics and more in giving surgeons real-time feedback to maximize resection while minimizing damage to critical structures.
12PLOS ONE. Surgery for Glioblastoma: Impact of the Combined Use of 5-Aminolevulinic Acid and Intraoperative MRI on Extent of Resection and SurvivalTargeted Drugs for Slower-Growing Gliomas
Not all brain tumors are as aggressive as glioblastoma. Low-grade gliomas that carry IDH mutations grow more slowly and have traditionally been managed with watchful waiting or radiation, often for years. A new drug called vorasidenib has changed the calculus for these patients in a way the field hasn’t seen in a long time.
Vorasidenib is an oral inhibitor that blocks the mutant IDH enzyme, which drives tumor growth by producing an abnormal metabolite. In a phase 3 trial, patients taking vorasidenib had a median time without disease progression of about 28 months, compared with about 11 months on placebo. The time before patients needed another intervention like surgery or radiation more than tripled.
13PubMed Central. Vorasidenib in IDH1- or IDH2-Mutant Low-Grade GliomaWith an additional six months of follow-up, these benefits held: tumors in the vorasidenib group were actually shrinking on average, while tumors in the placebo group grew at a rate of about 14% over the same period. Patients on the drug also experienced fewer seizures, which are one of the most disruptive symptoms of low-grade gliomas. Quality of life and thinking ability were not negatively affected.
14The Lancet Oncology. Vorasidenib in grade 2 IDH1- or IDH2-mutant low-grade glioma: a 6-month extension of the INDIGO trialThe drug’s design was deliberate. An earlier phase 1 trial showed that vorasidenib penetrated the brain more consistently than a related drug, ivosidenib, and reduced the tumor-driving metabolite by over 90%. Reduction of this metabolite was associated with reversal of genetic signatures linked to tumor stemness and decreased cell proliferation, which helped explain the clinical results and led to vorasidenib being advanced to phase 3 testing.
15Nature Medicine. Vorasidenib and ivosidenib in IDH1-mutant low-grade glioma: a randomized, perioperative phase 1 trialTumor Treating Fields and Emerging Physical Therapies
One of the more unusual treatments to enter neuro-oncology in recent years involves wearing a device on the head that delivers low-intensity alternating electric fields to the brain. Called tumor treating fields, or TTFields, the therapy targets dividing cells by disrupting the proteins that organize chromosome separation during cell division. Tubulin, the building block of the structures that pull chromosomes apart, has a strong electrical charge asymmetry. When TTFields are applied, they interfere with tubulin’s ability to assemble properly, leading to botched cell division, abnormal chromosome distribution, and cell death.
16PubMed Central. The Mechanisms of Action of Tumor Treating Fields17British Journal of Cancer. Tumour treating fields therapy for glioblastoma: current advances and future directions
TTFields are approved for use alongside standard temozolomide treatment in glioblastoma. The therapy is non-invasive but demanding: patients need to wear the device for at least 18 hours a day, which affects appearance, comfort, and daily life. The evidence for a survival benefit exists but has been debated, in part because the pivotal trial’s design made it hard to separate the effect of the device from other factors. Still, the approach represents a genuinely different treatment mechanism, one that doesn’t rely on chemicals crossing the blood-brain barrier.
Immunotherapy’s Difficult Road in the Brain
Immunotherapy has transformed treatment for melanoma, lung cancer, and several other tumor types, but its success in brain tumors has been limited. CAR T cell therapy, which engineers a patient’s own immune cells to recognize and attack a specific protein on tumor cells, has been tested against glioblastoma for over a decade. Targets have included EGFR, HER2, and IL13Rα2, and the cells have been delivered both through the bloodstream and directly into the brain.
18PubMed Central. CAR T cell therapy for glioblastoma: A review of the first decade of clinical trialsThe safety profile has been acceptable, but the results have been inconsistent. Most trials used each patient’s own T cells to avoid rejection, yet even so, the engineered cells often failed to expand and persist within the tumor long enough to mount a durable response. The immunosuppressive environment described earlier plays a role: even freshly engineered, highly motivated immune cells can be neutralized once they enter glioblastoma’s territory. Researchers are now working on next-generation constructs that include built-in countermeasures against tumor immune suppression, but a breakthrough remains elusive.
Telling Real Progression from Pseudoprogression
After a glioblastoma patient finishes radiation and chemotherapy, routine MRI scans often show what looks like the tumor growing back. In a meaningful fraction of cases, this apparent worsening is actually pseudoprogression: inflammation and treatment-related changes that mimic tumor regrowth on standard imaging. Mistaking pseudoprogression for true recurrence can lead to unnecessary surgery or premature treatment changes.
Advanced MRI techniques can help distinguish the two. Perfusion MRI, which measures blood flow through tissue, shows that pseudoprogression typically has lower blood volume than true tumor recurrence. One study found that perfusion metrics successfully identified pseudoprogression in patients whose scans looked worse but who didn’t actually need a change in treatment.
19PubMed Central. MRI perfusion in determining pseudoprogression in patients with glioblastomaLiquid biopsy is another emerging monitoring tool. Tumor DNA shed into the cerebrospinal fluid can be sequenced to track what a tumor is doing genetically over time. For brain cancers specifically, this cell-free tumor DNA is more abundant in spinal fluid than in blood, making spinal fluid a better sampling source. Longitudinal samples can detect residual disease after treatment or reveal how a tumor’s genetics are shifting at relapse, potentially allowing doctors to adjust therapy without another brain surgery.
20PubMed Central. ctDNA-Based Liquid Biopsy of Cerebrospinal Fluid in Brain CancerWhen Cancer Spreads to the Brain from Elsewhere
Primary brain tumors like glioblastoma start in the brain itself, but the majority of brain tumors in adults are actually metastases from cancers elsewhere, most commonly lung, breast, and melanoma. These metastatic brain tumors pose their own challenges. Certain cancer subtypes seem to have a specific affinity for brain tissue, driven by a combination of genetic features that allow cancer cells to cross the blood-brain barrier, survive in the brain’s low-glucose and low-oxygen environment, and co-opt the brain’s resident immune cells to create a hospitable niche.
21PubMed Central. Brain Metastasis OrganotropismFor patients with brain metastases receiving whole-brain radiation, cognitive decline is a serious concern. The drug memantine, better known for its use in Alzheimer’s disease, has been studied as a protective agent. A systematic review found that patients receiving memantine had a lower risk of cognitive failure after whole-brain radiation, and that combining memantine with a radiation technique that avoids the brain’s memory centers provided even greater cognitive protection. These interventions did not appear to affect overall survival, but they addressed something patients consistently rank as a top priority: preserving their ability to think clearly.
22PubMed Central. Role of memantine to mitigate radiation-induced cognitive dysfunction in brain metastasis patient receiving whole brain radiotherapy: a systematic reviewProton Therapy Versus Standard Radiation
Proton beam therapy deposits its energy more precisely than conventional X-ray radiation, theoretically sparing more healthy brain tissue. This has led to significant interest in proton therapy for brain tumors, particularly for tumors near critical structures. However, for glioblastoma specifically, a randomized comparison with roughly four years of follow-up found no difference in progression-free survival, overall survival, or time to cognitive decline between proton and standard photon radiation. Proton therapy was associated with lower rates of fatigue but otherwise showed no significant differences in patient-reported outcomes.
23PubMed Central. Controversies in neuro-oncology: Focal proton versus photon radiation therapy for adult brain tumors – Section: GlioblastomaThis doesn’t mean proton therapy has no role in neuro-oncology. For pediatric brain tumors, where minimizing radiation exposure to developing brain tissue is critical, the theoretical advantages of proton therapy may matter more over a lifetime. But for adult glioblastoma patients weighing the often-substantial cost difference, the current evidence doesn’t support a survival benefit.
Artificial Intelligence and Imaging
One of the most active areas of neuro-oncology research involves using machine learning to extract molecular information from standard MRI scans. The idea is that a tumor’s genetic makeup leaves a fingerprint in how it looks on imaging, even if no human radiologist can see it. Machine learning models trained on imaging features have predicted IDH mutation status and chromosomal deletions in gliomas with accuracies above 87%, and in some training datasets above 96%.
24Clinical Cancer Research. Machine Learning–Based Radiomics for Molecular Subtyping of GliomasDeep learning models, which analyze the raw images rather than hand-crafted features, have generally performed even better. Comparisons suggest that deep learning outperforms traditional radiomics in most classification tasks for diffuse gliomas.
25PubMed. Molecular subtyping of diffuse gliomas using magnetic resonance imaging: comparison and correlation between radiomics and deep learningThe practical promise is speed and access. Molecular testing of tumor tissue requires surgery or biopsy and days of laboratory processing. If an MRI scan could reliably predict the same information before a patient even enters the operating room, it could guide surgical planning, identify patients who might benefit from targeted therapy earlier, and extend molecular-level diagnostics to hospitals that lack advanced genetic testing facilities. The technology isn’t there yet for standalone clinical use, and external validation datasets show lower accuracy than training datasets, but the gap is narrowing.
26PubMed. Molecular subtype classification of low-grade gliomas using magnetic resonance imaging-based radiomics and machine learningPediatric Brain Tumors Are a Different Disease
Children’s brain tumors are biologically distinct from adults’. The most common malignant brain tumor in children, medulloblastoma, has been subdivided into molecular groups with very different prognoses and different vulnerabilities to targeted therapy. One subgroup, driven by abnormal activation of the Sonic Hedgehog signaling pathway, is already the focus of targeted drug development. These molecular subgroups are increasingly guiding both prognosis and treatment selection, with the long-term goal of sparing children from the harshest treatments when their tumor biology suggests a favorable outcome, and intensifying therapy when it doesn’t.
27PubMed Central. Pediatric medulloblastoma – update on molecular classification driving targeted therapiesGliomas That Wire Themselves into Brain Circuits
Perhaps the most striking discovery in neuro-oncology in recent years is that glioma cells don’t just sit passively in the brain. They form functional connections with neurons, essentially plugging themselves into the brain’s electrical circuitry. Research published in Nature demonstrated that glioma cells receive genuine synaptic input from neurons through receptors normally used for fast neural communication. When researchers used optogenetics to stimulate neural activity near a glioma in mice, the tumor grew faster. When they blocked the electrical signaling pharmacologically or genetically, tumor growth slowed and the mice survived longer.
28PubMed Central. Electrical and synaptic integration of glioma into neural circuitsThis reframes gliomas not as inert masses but as active participants in brain function, hijacking neural activity to fuel their own expansion. The implications are both unsettling and potentially useful: if synaptic integration is a driver of tumor progression, then disrupting the electrical conversation between neurons and glioma cells could become a new therapeutic strategy. Ongoing work is exploring whether existing drugs that modulate neural signaling might have anti-tumor effects as a secondary benefit.
29PubMed Central. Neuron-Glioma Synapses in Tumor ProgressionThe Financial Weight of Treatment
Brain tumor treatment is expensive, and the financial burden on patients is often underappreciated. A survey of patients with primary brain tumors found that roughly two-thirds reported their provider never discussed the cost of oral chemotherapy before starting it, or didn’t recall such a conversation. About one in six patients on standard-label therapy reported that chemotherapy costs negatively affected their quality of life, and the proportion was even higher among the smaller group receiving off-label oral agents.
30PubMed. Financial toxicity of oral chemotherapy in patients with primary brain tumorsThis financial toxicity compounds the cognitive and physical toll of disease. Patients dealing with seizures, fatigue, and cognitive changes from both the tumor and its treatment are simultaneously navigating insurance approvals, copays, and the indirect costs of being unable to work. As newer and more expensive therapies enter the field, from targeted inhibitors to cell-based immunotherapies, the gap between what is medically possible and what is financially accessible for patients is likely to widen before it narrows.