Key Ideas

Glioblastoma outcomes with temozolomide depend heavily on MGMT methylation status, a distinction worth revisiting given how often it gets lost in practice. Tumor treating fields improve survival regardless of methylation status, pseudoprogression remains a major diagnostic challenge best predicted by methylation status rather than imaging, and IDH inhibitors have meaningfully changed management of IDH-mutant low-grade gliomas.

A Field in the Middle of Real Change

Neuro-oncology moves quickly enough that it can be hard for general oncologists to keep pace with every subspecialty update. What follows revisits some fundamentals about temozolomide that are easy to lose sight of in practice, reviews the evidence for tumor treating fields, addresses the persistent challenge of pseudoprogression, and covers the growing role of IDH inhibitors in lower-grade, IDH-mutant gliomas.

Temozolomide: Revisiting the Original Trial and Who Actually Benefits

The Stupp trial remains the foundation of glioblastoma standard of care: six weeks of radiation with concurrent temozolomide, four weeks off, followed by six months (not longer) of temozolomide five nights a month. It is worth remembering that glioblastoma is a disease of older adults, yet the Stupp trial's population older than 70 was vanishingly small even though enrollment technically extended to age 79; the trial's true population was largely patients under 70, a group that represents well under half of glioblastoma patients today. Extending temozolomide beyond six months has become common in practice but is not supported by the original trial data.

Temozolomide is generally well tolerated, with fatigue as the main issue, particularly toward the end of radiation in older patients or those with comorbidities. Lymphopenia is a less visible but clinically important risk: it will not necessarily show up on routine blood counts, and steroid use compounds the risk further, so minimizing steroid duration and dose is a priority. Counsel patients to receive vaccinations before starting radiation if possible, and watch for opportunistic infections including shingles, thrush, cold sores, and, rarely, disseminated herpes zoster or HSV encephalitis. Chronic hepatitis B or C can flare significantly, largely driven by concurrent steroid use; screening remains a matter of clinical judgment, but there is little downside to checking given that flares can be managed proactively once identified.

The original temozolomide survival benefit in the overall trial population was modest, a little over two months in median overall survival, but this understates the more important finding: a durable long-term survival tail exists in a subset of responders. That observation led to the recognition that MGMT methylation status defines two genuinely different diseases within what is called glioblastoma. Methylated tumors are inherently slower growing, even without treatment, and this is the group where temozolomide has substantial benefit, with 20% to 30% remaining recurrence-free at two to three years. Unmethylated tumors, roughly two-thirds of glioblastoma, derive only a marginal, barely statistically significant benefit from temozolomide, and the small tail of long-term survivors in this group may partly reflect testing error rather than true benefit. For unmethylated patients experiencing significant treatment side effects, it is reasonable to have a frank conversation about whether continuing temozolomide is worthwhile.

Tumor Treating Fields: A Real Benefit Independent of Methylation Status

Tumor treating fields (marketed as Optune) faced substantial early skepticism in the field, partly because initial trial designs used a non-inferiority approach without an appropriate control. A subsequent trial of nearly 700 patients, with all patients receiving standard surgery, radiation, and temozolomide and randomized by methylation status, showed that tumor treating fields worn after radiation improve survival regardless of MGMT status, generally delaying recurrence and improving survival by three to five months for the average patient. The device is not blinded and carries no sham-device control, a legitimate limitation, but the effect is real. Skin irritation is the main side effect and is manageable with scheduled breaks. The device requires substantial patient commitment, at least 18 hours of daily wear, along with head shaving and skin care, so patient personality and preference matter considerably in this decision. For unmethylated patients in particular, where temozolomide benefit is marginal, tumor treating fields represents a genuinely more effective option worth discussing.

Pseudoprogression: Trust Methylation Status, Not the Radiology Report

Roughly a third of glioblastomas will show worse-appearing imaging within the first six months after chemoradiation, and standard MRI has essentially no reliable ability to distinguish true tumor progression from pseudoprogression, a phenomenon that became far more common once temozolomide was incorporated into standard treatment, likely reflecting added alkylating injury to already leaky, angiogenic tumor vasculature. Radiology reports that confidently favor "tumor" or "treatment effect" on standard MRI should be viewed with real skepticism given the absence of predictive value; MGMT methylation status is a far more reliable predictor, since pseudoprogression is substantially more common in methylated tumors and closer to a coin flip in unmethylated disease. FDG-PET and perfusion imaging offer modest additional discriminative value, at best around 80% accuracy, and standard pathology from a biopsy is not useful for this specific question since both tumor and treatment-related changes coexist in the tissue. A newer amino acid PET tracer (FET-PET) has recently reported roughly 95% accuracy in distinguishing pseudoprogression from true progression and is under FDA review; if it becomes widely available, it should meaningfully improve this diagnostic dilemma.

Bevacizumab deserves a specific mention here: it makes MRI findings look dramatically better by reducing vascular leakiness, but this improvement is largely cosmetic rather than a true anti-tumor effect, so it should be framed to patients as a tool for symptom control, essentially a more powerful steroid, rather than disease control. About a third of patients with bulky, symptomatic disease will see a noticeable symptom improvement within two to three doses; if no clear benefit is seen by the third dose, further doses are unlikely to help. Randomized trials have not shown a survival or tumor control benefit despite the imaging improvement.

IDH-Mutant Glioma: A Genuinely New Chapter

The molecular reclassification of gliomas around IDH status has substantially clarified an old area of confusion: grade 3 astrocytomas that are IDH-mutant behave much more like grade 2 tumors, while IDH-wild-type grade 3 tumors behave molecularly like glioblastoma. The INDIGO trial established vorasidenib, an oral IDH1/2 inhibitor, as an option to delay radiation in patients with IDH-mutant, grade 2 diffuse glioma following surgery; the most recent follow-up data show a median progression-free survival of 44.1 months with vorasidenib, and time to next intervention was not yet reached with extended follow-up, both substantial improvements over placebo. The drug is generally well tolerated, though roughly 10% of patients develop elevated liver enzymes requiring monitoring, which is reversible with drug discontinuation. Because this population skews toward women under 40, the drug's teratogenicity in animal studies is an important discussion point.

Germline Testing Deserves Broader Consideration

Glioblastoma has traditionally not been considered a hereditary cancer warranting routine genetic testing referral in the absence of a strong family history. More recent, broader genomic surveys of glioma patients suggest that 4% to 7% of glioma patients, independent of family history, carry a germline pathogenic variant (including Lynch syndrome and BRCA-family or CHEK2 mutations). Given that this is not a negligible rate, referral for genetic counseling deserves more routine consideration than has historically been standard practice, particularly for patients with living biological relatives who could benefit from cascade screening.

For Patients

Brain tumor treatment has become considerably more individualized. For glioblastoma, a molecular test called MGMT methylation status has turned out to be extremely important: "methylated" tumors tend to grow more slowly and respond meaningfully better to chemotherapy, while "unmethylated" tumors respond only marginally, worth discussing with your oncology team when weighing side effects against expected benefit. A wearable device called tumor treating fields (Optune) has also been shown to extend survival regardless of methylation status, though it requires significant daily commitment and is not the right fit for everyone. If an MRI looks worse after treatment, this does not necessarily mean the tumor is growing back; it is frequently a treatment-related effect called pseudoprogression that resolves on its own, particularly in methylated tumors, so a change in scan appearance alone should prompt a careful conversation rather than immediate alarm. Finally, for gliomas that carry an IDH gene mutation, typically found in younger patients with lower-grade tumors, a new class of oral medications can meaningfully delay the need for radiation. Given emerging data on inherited genetic risk, it is reasonable to ask your care team whether genetic counseling is appropriate for you or your family.

Key Takeaways

  • MGMT methylation status defines two biologically distinct glioblastoma populations: methylated tumors derive substantial benefit from temozolomide, while unmethylated tumors (two-thirds of cases) derive only marginal benefit.

  • Tumor treating fields (Optune) improve survival by three to five months regardless of MGMT status, though the device requires at least 18 hours of daily wear and is not appropriate for every patient.

  • Pseudoprogression affects roughly a third of glioblastoma patients after chemoradiation; standard MRI cannot reliably distinguish it from true progression, but MGMT methylation status can help predict its likelihood, and a new amino acid PET tracer under FDA review may soon offer a more definitive answer.

  • Bevacizumab improves MRI appearance and can meaningfully improve symptoms in bulky, symptomatic disease, but has not been shown to improve survival or true tumor control.

  • Vorasidenib (INDIGO trial) delays the need for radiation in IDH-mutant grade 2 glioma, with median progression-free survival of 44.1 months in extended follow-up.

  • Roughly 4% to 7% of glioma patients carry a germline pathogenic variant regardless of family history, supporting broader consideration of genetic counseling referral.

References

  1. Stupp R, et al. Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma. N Engl J Med. 2005 (original trial); long-term methylation-status analyses, Hegi ME, et al.

  2. Stupp R, et al. Effect of tumor-treating fields plus maintenance temozolomide vs maintenance temozolomide alone on survival in patients with glioblastoma: a randomized clinical trial (EF-14). JAMA. 2017.

  3. Mellinghoff IK, van den Bent MJ, et al. Vorasidenib in IDH1- or IDH2-mutant low-grade glioma (INDIGO). N Engl J Med. 2023; extended follow-up presented at ASCO 2026.

  4. Vorasidenib in IDH1-mutant or IDH2-mutant low-grade glioma (INDIGO): secondary and exploratory endpoints. Lancet Oncol. 2025.