Skip to content
Bending the Curve Trial by Trial
?

Brain (central nervous system)

Glioblastoma

How surgery, radiotherapy, temozolomide and molecular classification shaped care for glioblastoma.

All brain and nervous system cancers, not glioblastoma alone · 5-year relative survival · United States, SEER 8 registries

23.51%→35.49%

1975 (observed value; first year in the table) to 2018 (observed value; last observed year in the table). Read these with the caveats under “Then and now”.Source

24 milestones, by decade

  • A 2021 historical review of surgical series, from the first craniotomy for glioma in 1884 to 2020, reports that the earliest reported craniotomy for glioblastoma was followed by the patient's death less than a month after surgery. It describes early postoperative mortality in Harvey Cushing's series falling from 50% to 6% as management of intracranial pressure improved, and gives a mean survival of 10.1 months in the first major surgical series (1912), a figure it says did not improve until radiotherapy was introduced in the 1950s. These figures are as stated in the review's abstract; the original reports are not cited here.
  • In the randomised trial of 573 people with newly diagnosed glioblastoma reported in 2005, adding temozolomide to radiotherapy raised median survival from 12.1 to 14.6 months (unadjusted hazard ratio for death 0.63, 95% CI 0.52 to 0.75) and two-year survival from 10.4% to 26.5%. In a companion analysis of tumour specimens from the same trial, the MGMT promoter was methylated in 45% of 206 assessable cases; within that group median survival was 21.7 months with temozolomide plus radiotherapy against 15.3 months with radiotherapy alone, while the difference in tumours without methylation was smaller and not statistically significant. The MGMT analysis is a subgroup analysis of the trial, not a separately randomised comparison.
  • On 5 May 2009 the FDA granted bevacizumab accelerated approval for glioblastoma with progressive disease following prior therapy and required a confirmatory placebo-controlled phase III trial of bevacizumab with temozolomide and radiotherapy in newly diagnosed glioblastoma. That setting was then tested in AVAglio and RTOG 0825, both published in 2014: neither found a significant overall survival difference (AVAglio hazard ratio for death 0.88, 95% CI 0.76 to 1.02; RTOG 0825 median 15.7 versus 16.1 months, hazard ratio 1.13), although progression-free survival was longer with bevacizumab (in RTOG 0825 the improvement did not reach the prespecified target). EORTC 26101, published in 2017, found no overall survival advantage from adding bevacizumab to lomustine at first progression (hazard ratio 0.95, 95% CI 0.74 to 1.21). On 5 December 2017 the FDA approved a supplement that removed the label statement that effectiveness was based on objective response rate and recorded the postmarketing requirement as fulfilled. FDA's review document for that supplement states that EORTC 26101 showed no statistically significant overall survival difference and that the benefit in recurrent glioblastoma was supported by progression-free survival data from three randomised trials (EORTC 26101 in recurrent disease; AVAglio and RTOG 0825, in newly diagnosed disease, are described as supporting the consistency of the progression-free survival effect) and by corticosteroid-use data.

Evidence current to October 2026 · Checked against primary sources to the extent possible

A guided tour of 12 milestones. You can pause or step through it yourself.

Filter milestonesshowing all 24
Treatment type
Setting

Anatomists & surgeons

  1. 1884 · Origin · SurgeryGlioma surgery since 1884: a historical review of outcomesA 2021 historical review of surgical series in glioblastoma starts from the introduction of craniotomy for glioma in 1884. It reports that the earliest operation was followed by the patient's death within a month, and that early deaths after surgery fell from 50% to 6% in Harvey Cushing's series after better management of pressure inside the skull.McCutcheon and Preul (World Neurosurg 2021), review of major glioblastoma surgical series from 1884 to 2020: the earliest reported craniotomy was followed by death within a month; early postoperative mortality fell from 50% to 6% in Cushing's series after improved intracranial pressure management; mean survival in the first major surgical series (1912) was 10.1 months. These are the reviewers' figures, not checked against the original series.

The age of rays

  1. 1926 · Understanding · UnderstandingBailey and Cushing classify gliomas by cell of originIn 1926 Percival Bailey and Harvey Cushing published a book that sorted brain tumours of the glioma group by the developing brain cells they resembled, and linked each type to how long patients lived. It drew on 414 cases indexed as glioma between 1903 and 1924.Case series from two hospitals (Johns Hopkins, Peter Bent Brigham), January 1903 to January 1924: 414 cases indexed as glioma, of which 254 were re-examined histologically and assigned to 13 categories. Spongioblastoma multiforme, the largest group (77 patients, 'about one-third' of classified cases in the authors' words), had an average survival of 12 months by the authors' table; one of the 77 was alive 29 months after operation. Less differentiated tumours had shorter survival.

Molecular origins

  1. 1978 · Treatment · RadiationRandomised trials of postoperative radiotherapy and nitrosoureasA 1978 report randomised 303 people operated on for anaplastic glioma to best conventional care, carmustine, radiotherapy, or both. Median survival was 14 weeks with best conventional care and about 35 weeks with radiotherapy. A 1980 report in 467 people with malignant glioma suggested that radiotherapy is best used after surgery.Walker 1978 (303 randomised, 222 in the valid study group; anaplastic glioma): median survival 14 weeks with best conventional care, 18.5 BCNU, 35 radiotherapy, 34.5 BCNU plus radiotherapy; all active arms showed some statistical superiority over best conventional care. Walker 1980 (467 randomised, 358 valid study group; malignant glioma): radiotherapy with or without a nitrosourea improved survival over semustine alone; carmustine plus radiotherapy gave the best survival, not significantly different from semustine plus radiotherapy.

Targeted therapy

  1. 2000 · Understanding · UnderstandingMGMT promoter methylation predicts benefit from alkylating agentsIn 2000, researchers linked a silenced repair gene, MGMT, in glioma tissue to better responses to carmustine. In 2005, in samples from a randomised trial, glioblastomas with the silenced gene did better when temozolomide was added to radiotherapy. Without it, the gain was smaller and not statistically significant.Esteller 2000: MGMT promoter methylated in 19 of 47 (40%) gliomas from carmustine-treated patients; associated with tumour regression and longer overall and disease-free survival. Hegi 2005 (analysis of specimens from a randomised trial of radiotherapy with or without temozolomide): methylated in 45% of 206 assessable cases; prognostic irrespective of treatment (HR 0.45, 95% CI 0.32 to 0.61). In methylated tumours, median survival 21.7 vs 15.3 months with vs without temozolomide (P=0.007).
  2. 2001 · Understanding · SurgeryExtent of resection is linked to survival in glioblastomaAnalyses from 2001 onwards found that people whose glioblastoma was removed more completely lived longer. A 2001 analysis of 416 patients reported median survival of 13 months after removal of 98% or more of the tumour and 8.8 months after less. These are associations, not randomised comparisons; a 2016 meta-analysis rates the evidence moderate to low quality.Lacroix 2001 (416 patients, retrospective): resection of 98% or more of tumour volume, median survival 13 months (95% CI 11.4-14.6) vs 8.8 months (7.4-10.2) with less (P<0.0001). Sanai 2011 (500 patients): survival advantage from 78% resection. Brown 2016 meta-analysis (37 studies, 41,117 patients): 1-year mortality RR 0.62 (0.56-0.69) for gross total vs subtotal resection; GRADE evidence moderate to low.
  3. 2005 · Treatment · ChemotherapyTemozolomide added to radiotherapy for newly diagnosed glioblastomaIn 2005 a trial of 573 adults with newly diagnosed glioblastoma compared radiotherapy alone with radiotherapy plus temozolomide, a chemotherapy capsule. Median survival was about 15 months with temozolomide and 12 months without. About one in four people given temozolomide were alive at two years, against about one in ten with radiotherapy alone.EORTC-NCIC trial (573 randomised): radiotherapy 60 Gy with or without daily temozolomide then six adjuvant cycles. Median OS 14.6 vs 12.1 months, unadjusted HR for death 0.63 (95% CI 0.52-0.75; P<0.001); 2-year survival 26.5% vs 10.4%. Five-year analysis: OS 9.8% vs 1.9% at 5 years. Grade 3 or 4 haematological toxicity 7% with concomitant treatment.
  4. 2006 · Treatment · Surgery5-ALA fluorescence-guided surgery increases complete resectionIn 2006 a randomised trial of people with suspected malignant glioma compared surgery guided by a fluorescent dye (5-ALA) with ordinary white-light surgery. Complete removal of the tumour visible on scans was achieved in about two in three people with the dye and about one in three without. Six-month progression-free survival was 41% against 21%.Randomised phase 3 trial (322 randomised; interim full-analysis population 270, study terminated at the interim analysis): contrast-enhancing tumour resected completely in 65% with 5-ALA fluorescence guidance vs 36% with white light (difference 29%, 95% CI 17-40; P<0.0001); 6-month PFS 41.0% vs 21.1% (difference 19.9%, 95% CI 9.1-30.7; P=0.0003). The abstract reports no overall survival result.
  5. 2007 · Treatment · RadiationRadiotherapy improves survival in older people with glioblastomaIn 2007 a randomised trial of 85 people aged 70 or older with glioblastoma or anaplastic astrocytoma compared radiotherapy plus supportive care with supportive care alone. In the 81 people with glioblastoma, median survival was about 29 weeks with radiotherapy and about 17 weeks without. Quality of life and thinking did not differ significantly between the groups.Randomised trial, 85 patients aged 70 or older (Karnofsky score 70 or higher), stopped at the first interim analysis. In 81 with glioblastoma: median survival 29.1 weeks (39 patients) with 50 Gy radiotherapy plus supportive care vs 16.9 weeks (42) with supportive care alone; HR for death 0.47 (95% CI 0.29-0.76; P=0.002). No severe radiotherapy-related adverse events; quality of life and cognition not significantly different.
  6. 2008 · Understanding · UnderstandingIDH1 and IDH2 mutations found in gliomasIn 2008, genome sequencing of glioblastoma found recurrent changes in the IDH1 gene, often in younger people and in most tumours that had grown from lower-grade ones. A 2009 study found IDH1 or IDH2 changes in most lower-grade gliomas, with better outcomes, and a third showed the changed enzyme makes an unusual molecule.Parsons 2008: 20,661 genes sequenced in 22 glioblastoma samples; IDH1 active-site mutations in 12% of patients, frequent in young patients and most secondary glioblastomas, associated with longer overall survival. Yan 2009: IDH1 and IDH2 sequenced in 445 CNS and 494 non-CNS tumours; mutations affecting amino acid 132 of IDH1 in more than 70% of grade II and III astrocytomas and oligodendrogliomas; mutant tumours had better outcome. Dang 2009: mutant IDH1 produces 2-hydroxyglutarate.
  7. 2009 · Treatment · TargetedBevacizumab approved for progressive glioblastoma on response rateIn May 2009 the US regulator gave accelerated approval to bevacizumab for glioblastoma that had progressed after earlier treatment. The decision rested on how many tumours shrank on scans, not on survival. In a randomised phase 2 trial of 167 people, about 28 in 100 on bevacizumab alone had a response.FDA accelerated approval (BLA 125085, supplement 169; 5 May 2009) for glioblastoma with progressive disease after prior therapy. Supporting phase 2 BRAIN (n=167, randomised, noncomparative; bevacizumab 10 mg/kg alone or with irinotecan): independently assessed ORR 28.2% vs 37.8%, 6-month PFS 42.6% vs 50.3%, median OS 9.2 vs 8.7 months. Traditional approval followed on 5 December 2017 (supplement 319).
  8. 2010 · Understanding · UnderstandingExpression subtypes and somatic landscape of glioblastomaIn 2010, building on the Cancer Genome Atlas catalogue, researchers sorted glioblastomas into four subtypes by gene activity and reported that response to aggressive treatment differed between them. A 2013 analysis of more than 500 tumours described further gene changes, including in the TERT promoter.Verhaak 2010 (TCGA data): expression-based classification into proneural, neural, classical and mesenchymal subtypes; EGFR, NF1 and PDGFRA/IDH1 abnormalities defined classical, mesenchymal and proneural; response to aggressive therapy differed by subtype, greatest in classical, none in proneural. Brennan 2013: more than 500 glioblastomas; TERT promoter mutations correlated with mRNA expression; proneural survival advantage attributed to G-CIMP; MGMT methylation may be predictive only in classical subtype.

Immunotherapy

  1. 2013 · Reversal · ChemotherapyDose-dense temozolomide shows no survival gain (RTOG 0525)A 2013 trial of 833 people with newly diagnosed glioblastoma compared standard temozolomide with a dose-dense schedule after chemoradiation. Median survival was about 17 months with the standard schedule and 15 months with the dose-dense schedule, not a significant difference. Severe side effects were more common with the dose-dense schedule.RTOG 0525 (833 randomised of 1,173 registered): standard vs dose-dense temozolomide, median OS 16.6 vs 14.9 months (HR 1.03; P=.63); median PFS 5.5 vs 6.7 months (HR 0.87; P=.06). Treatment effect did not differ by MGMT methylation status. Across both arms, MGMT methylation was associated with better OS (21.2 vs 14 months; HR 1.74 as printed, between methylation groups; P<.001). Grade 3 or higher toxicity 34% vs 53% (P<.001).
  2. 2014 · Reversal · TargetedBevacizumab added to chemoradiation does not improve survivalIn 2014 two large randomised trials added bevacizumab to radiotherapy and temozolomide for newly diagnosed glioblastoma. People given bevacizumab went longer before their tumour progressed, but overall survival was not significantly different, and more had severe side effects.AVAglio (458 bevacizumab, 463 placebo): median PFS 10.6 vs 6.2 months, stratified HR 0.64 (95% CI 0.55 to 0.74); OS stratified HR 0.88 (95% CI 0.76 to 1.02; P=0.10); grade 3 or higher adverse events 66.8% vs 51.3%. RTOG 0825 (n=637 randomised): median OS 15.7 vs 16.1 months (HR 1.13); PFS 10.7 vs 7.3 months (HR 0.79), below the prespecified target.
  3. 2014 · Reversal · TargetedThree further add-on agents fail in randomised phase 3 trialsThree randomised phase 3 trials added a new drug to radiotherapy and temozolomide in newly diagnosed glioblastoma: cilengitide (reported 2014), a vaccine called rindopepimut (2017) and marizomib (2024). None improved overall survival, and the cilengitide authors said it would not be developed further.CENTRIC (MGMT-methylated, n=545): median OS 26.3 vs 26.3 months, HR 1.02 (95% CI 0.81 to 1.29; P=0.86). ACT IV (EGFRvIII-expressing; 745 enrolled; minimal residual disease group): median OS 20.1 vs 20.0 months, HR 1.01 (95% CI 0.79 to 1.30); terminated for futility at interim. EORTC 1709 marizomib (n=749): median OS 17 months with standard treatment vs 16.5 with marizomib, HR 1.04 (P=.64); more grade 3 or 4 adverse events with marizomib.
  4. 2015 · Understanding · UnderstandingIDH, 1p/19q and TERT define diffuse glioma groupsTwo 2015 studies showed that a few DNA features, IDH mutation, 1p/19q loss and TERT mutation, sort diffuse gliomas into groups that differ in age at diagnosis and outcome. Most lower-grade gliomas without an IDH mutation resembled glioblastoma.TCGA: 293 adult grade II and III gliomas; three prognostic classes captured better by IDH, 1p/19q and TP53 status than by histology; IDH-mutant with 1p/19q codeletion had the most favourable outcomes; most IDH-wildtype lower-grade gliomas resembled primary glioblastoma. Eckel-Passow: 1087 gliomas scored for TERT promoter, IDH and 1p/19q: five groups; among 472 grade IV gliomas 74% had only TERT mutation and 7% only IDH; groups were associated with overall survival in grade II and III but not grade IV.
  5. 2015 · Treatment · TargetedTumour treating fields added to maintenance temozolomide (EF-14)In 2015 FDA approved a device that delivers electric fields through the scalp (tumour treating fields) for newly diagnosed glioblastoma. The open-label EF-14 trial, which enrolled people only after chemoradiation, reported median overall survival of about 21 months with the device plus temozolomide and 16 months with temozolomide alone. Its place in care is debated.EF-14 (695 randomised 2:1 after chemoradiation, open-label): TTFields plus maintenance temozolomide vs temozolomide alone. Final analysis (ITT): median PFS 6.7 vs 4.0 months (HR 0.63; 95% CI 0.52-0.76; P<.001); median OS 20.9 vs 16.0 months (HR 0.63; 0.53-0.76; P<.001). Skin toxicity under the arrays in 52% vs none. Randomised after a median 3.8 months from diagnosis.
  6. 2016 · Understanding · UnderstandingWHO classification becomes molecular: IDH-wildtype glioblastomaThe WHO classification of brain tumours began, in 2016, to define tumours by molecular features as well as by appearance under the microscope. The 2021 edition names the adult tumour glioblastoma, IDH-wildtype, and places IDH-mutant tumours in a separate astrocytoma type.WHO 2016: its authors describe the first CNS classification to use molecular parameters with histology; defined glioblastoma, IDH-wildtype and IDH-mutant. cIMPACT-NOW update 3 (2018): histologically grade II or III IDH-wildtype diffuse astrocytic gliomas with EGFR amplification, +7/-10 or TERT promoter mutation correspond to WHO grade IV. WHO 2021: glioblastoma, IDH-wildtype, grade 4, in adults with IDH-wildtype diffuse astrocytic glioma plus one of microvascular proliferation, necrosis, TERT promoter mutation, EGFR amplification or +7/-10; IDH-mutant astrocytoma graded 2 to 4.
  7. 2017 · Reversal · TargetedAdding bevacizumab to lomustine does not extend survival (EORTC 26101)In 2017 a trial of 437 people whose glioblastoma had progressed after first treatment added bevacizumab to lomustine chemotherapy. The cancer stayed controlled for longer on scans, but median survival was about nine months with the combination and about nine without it.EORTC 26101 (n=437 randomised 2:1; 288 lomustine plus bevacizumab, 149 lomustine alone) at first progression after chemoradiation: median OS 9.1 vs 8.6 months, HR 0.95 (95% CI 0.74 to 1.21; P=0.65). Locally assessed PFS 4.2 vs 1.5 months, HR 0.49 (95% CI 0.39 to 0.61). Grade 3 to 5 adverse events 63.6% vs 38.1%. MGMT status was prognostic.
  8. 2017 · Treatment · RadiationShort-course radiotherapy plus temozolomide in older peopleIn 2017 a randomised trial of 562 people aged 65 or older with glioblastoma compared a short course of radiotherapy alone with the same radiotherapy plus temozolomide. Median survival was 9.3 months with temozolomide and 7.6 months without. The gain looked larger in tumours with methylated MGMT than unmethylated ones, but that difference was not statistically significant.CE.6 (562 randomised, age 65 or older): 40 Gy in 15 fractions with or without temozolomide. Median OS 9.3 vs 7.6 months (HR 0.67; 95% CI 0.56-0.80; P<0.001); median PFS 5.3 vs 3.9 months (HR 0.50; 0.41-0.60). MGMT methylated (n=165): OS 13.5 vs 7.7 months (HR 0.53); unmethylated (n=189): 10.0 vs 7.9 months (HR 0.75; 0.56-1.01; P=0.055; P=0.08 for interaction).
  9. 2019 · Treatment · ChemotherapyLomustine plus temozolomide in MGMT-methylated glioblastoma (CeTeG)A 2019 trial in Germany of 141 people with newly diagnosed glioblastoma and a methylated MGMT promoter compared standard temozolomide chemoradiation with radiotherapy plus lomustine and temozolomide. Median survival was about 31 months with standard treatment and 48 months with the combination, a result the authors call suggestive and interpret with caution because the trial was small.CeTeG/NOA-09 (open-label phase 3; 141 randomised, 129 in modified ITT; MGMT-methylated, age 18-70): median OS 31.4 months (95% CI 27.7-47.1) with temozolomide chemoradiation vs 48.1 months (32.6-not assessable) with lomustine-temozolomide; HR 0.60 (95% CI 0.35-1.03); log-rank p=0.0492. Grade 3 or higher adverse events 51% vs 59%. Authors advise caution owing to small size.

Precision & engineered

  1. 2020 · Reversal · ImmunotherapyNivolumab phase 3 trials in glioblastoma do not meet their endpointsThree randomised phase 3 trials tested the immunotherapy nivolumab in glioblastoma: one at first recurrence and two alongside first-line radiotherapy. In none did people live longer with nivolumab. In one first-line trial, those given temozolomide with radiotherapy lived longer than those given nivolumab with radiotherapy.CheckMate 143 (recurrent, n=369): nivolumab vs bevacizumab, median OS 9.8 vs 10.0 months, HR 1.04 (95% CI 0.83 to 1.30). CheckMate 498 (MGMT-unmethylated, n=560): nivolumab plus radiotherapy vs temozolomide plus radiotherapy, median OS 13.4 vs 14.9 months, HR 1.31 (95% CI 1.09 to 1.58). CheckMate 548 (MGMT-methylated or indeterminate, n=716): median OS 28.9 vs 32.1 months, HR 1.1.
  2. 2021 · Treatment · TargetedDabrafenib plus trametinib for BRAF V600E-mutant gliomaIn a single-arm trial of adults with BRAF V600E-mutant brain tumours that had returned or progressed, about one in three people with high-grade glioma and about seven in ten with low-grade glioma had a tumour response to dabrafenib plus trametinib. US approvals followed in 2022 and 2023.ROAR basket trial, single-arm phase 2, adults with recurrent or progressive BRAF V600E-mutant glioma (interim analysis 16, cutoff 14 September 2020): investigator-assessed ORR 33% (95% CI 20 to 49; 15 of 45) in the high-grade cohort, which included 31 with glioblastoma, and 69% (95% CI 39 to 91; 9 of 13) in the low-grade cohort. Randomised phase 2 in paediatric low-grade glioma: response 47% vs 11% with chemotherapy.
  3. 2023 · Treatment · TargetedVorasidenib delays progression in IDH-mutant grade 2 glioma (INDIGO)In 2023 a trial of 331 people with grade 2 IDH-mutant glioma, who had had only surgery, compared vorasidenib, a daily tablet, with a placebo. Time without tumour growth on scans was about 28 months with vorasidenib and 11 months with placebo. This tumour type is not glioblastoma.INDIGO (n=331: 168 vorasidenib 40 mg daily, 163 placebo): double-blind phase 3 in residual or recurrent grade 2 IDH-mutant glioma after surgery only. Imaging-based PFS by blinded review: median 27.7 vs 11.1 months, HR 0.39 (95% CI 0.27 to 0.56; P<0.001); time to next intervention HR 0.26 (95% CI 0.15 to 0.43). Grade 3 or higher adverse events 22.8% vs 13.5%.
  4. 2024 · Treatment · TargetedDordaviprone gets accelerated approval for H3 K27M midline gliomaONC201 (dordaviprone), a capsule, was studied in 50 people with recurrent H3 K27M-mutant diffuse midline glioma. About one in five had a tumour response on scans. In August 2025 US regulators gave it accelerated approval and required a randomised trial to confirm benefit. This is not glioblastoma.Integrated analysis of 50 patients (46 adult, 4 paediatric) with recurrent H3 K27M-mutant diffuse midline glioma given single-agent ONC201 across trials and an expanded access protocol: ORR by RANO-HGG 20.0% (95% CI 10.0 to 33.7), median DOR 11.2 months, median time to response 8.3 months; combined RANO-HGG/LGG ORR 30.0% (95% CI 17.9 to 44.6). FDA accelerated approval 6 August 2025.

Then and now

  • 5-year relative survival · United States, SEER 8 registries

    • 23.51%

      1975 (observed value; first year in the table)

    Brain and other nervous system cancer as a SEER Cancer Stat Facts category (the page does not define which tumour types it includes); all races, both sexes. It is not specific to glioblastoma: it pools malignant brain and other nervous system tumours of many types, most of which are less aggressive than glioblastoma. The SEER 8 five-year relative survival series (1975-2018) on the page; no age range is stated for it.

    Modeled trend value for 1975: 21.83%. The observed series moves from year to year (lowest 22.33% in 1977, highest 38.12% in 2009). Relative survival excludes deaths from other causes. 'Observed' is the table's label for the actual yearly value, as opposed to the Joinpoint modeled-trend value in the adjacent column. The page's column is headed 'Year' and the chart note says only 'SEER 8 5-Year Relative Survival Percent from 1975-2018'; it does not say the year is the year of diagnosis. SEER 8 covers a limited set of registries. The page does not describe changes over time in imaging, grading, histology mix or case ascertainment, does not apportion the change to any cause, and gives no glioblastoma-specific series, so the rise from 1975 to 2018 must not be read as an effect of glioblastoma treatment. Read beside the death rates. Compare only with other values from this SEER 8 series, not with the SEER 21 value or with CBTRUS.

    Source

  • 5-year relative survival · United States, SEER 8 registries

    • 35.49%

      2018 (observed value; last observed year in the table)

    Same category and series as the 1975 value above (brain and other nervous system cancer, all races, both sexes, SEER 8).

    Modeled trend value for 2018: 35.55%. The observed column has no value for 2019 onward (shown as '-'); the modeled column continues to 2023 (34.20%) without observed values. Same limits as the 1975 value: a broad category, not glioblastoma, and the year label is not stated to be year of diagnosis.

    Source

  • 5-year relative survival · United States, SEER 21 registries (excluding IL)

    • 32.9%

      2016-2022

    Brain and other nervous system cancer, all races, both sexes; the page's headline five-year relative survival. A separate series from SEER 8, with a different registry set and period.

    Do not join or subtract it from the SEER 8 values. Not glioblastoma-specific. By summary stage on the page: localized (77% of cases) 35.3%, regional (14%) 20.1%, distant (2%) 28.0%, unknown (7%) 32.2%; the distant value is higher than the regional one, and the page does not explain the stage categories for these tumours, so the splits should not be read as a prognostic ladder.

    Source

  • Age-standardised mortality · United States (U.S. Mortality data, not SEER registries)

    • 4.11per 100,000

      1975 (observed value; first year in the table)

    Brain and other nervous system cancer deaths, all races, both sexes, age-adjusted rate as labelled on the SEER Cancer Stat Facts page (the page does not name the standard population in the text read).

    Modeled trend value for 1975: 4.20. The observed series peaks at 4.95 in 1991. Not glioblastoma-specific: the category pools many tumour types. A death rate also moves with incidence, diagnosis and coding practice, which the page does not discuss. Read beside the survival values, not as a measure of treatment effect.

    Source

  • Age-standardised mortality · United States (U.S. Mortality data, not SEER registries)

    • 4.33per 100,000

      2024 (observed value; last year in the table)

    Same category and series as the 1975 death rate above.

    Modeled trend value for 2024: 4.34. The page text reports the death rate as 4.4 per 100,000 for 2020-2024 and says age-adjusted death rates have been falling on average 0.3% each year over 2015-2024. The change from 1975 is small and, as above, is not glioblastoma-specific.

    Source

  • 5-year relative survival · United States, CBTRUS (NPCR registries)

    • 7%

      Cases diagnosed 2004-2021 (Table 13 of the CBTRUS report for 2018-2022; the table's data-source line gives the NPCR registries for 2004-2021)

    Glioblastoma (malignant behaviour code /3), all ages, 154,365 cases; 95% CI 6.8 to 7.1. For glioblastoma the table's 'All' and 'Malignant' columns are identical. By age at diagnosis: 0-14 years 20.1% (95% CI 17.8 to 22.4), 15-39 years 28.5% (27.4 to 29.6), 40 years and older 5.6% (5.5 to 5.8; 145,230 of the 154,365 cases).

    From the CBTRUS Statistical Report for 2018-2022 (Neuro Oncol 2025), which supersedes the 2017-2021 report; the cohort-based table is the same type as in the previous edition, whose footnote says such estimates may not reflect the long-term outlook of newly diagnosed cases. The period spans several editions of the WHO classification, so the group is not guaranteed to be defined the same way throughout. The ten-year relative survival in the same row is 4.4% (95% CI 4.2 to 4.5). The previous edition (2017-2021 report, Table 18) gave 6.8% (95% CI 6.7 to 7.0; 172,774 malignant cases diagnosed 2001-2020). Not comparable with the SEER values, which pool all brain and CNS tumours, or with trial medians. These values were read from the publisher's text of the report, because the PMC copy sat behind a bot check that was not bypassed; a human check of Table 13 against the PDF is advisable.

    PubMedDOI

Survival figures are affected by earlier detection (lead time) and overdiagnosis, and by changes in who is diagnosed. For screened cancers, read them alongside mortality.

Evidence chains

A lane follows one defined population and setting, where each new trial tests against what was then standard. Compare only within a lane, and read every comparison as illustrative.

Newly diagnosed glioblastoma, randomised at the start of radiotherapy (not selected by MGMT status)

Adults with newly diagnosed glioblastoma randomised before or at the start of radiotherapy, without selection by MGMT promoter status (EORTC-NCIC 26981/22981, RTOG 0825, EORTC 1709 marizomib) · Adjuvant · endpoint: OS

Results are not comparable across trials.

Only the Stupp trial used radiotherapy alone as its control; the later trials used radiotherapy plus temozolomide, so their control medians are not comparable with the Stupp control arm. Enrolment rules differ: EORTC 1709 required a Karnofsky performance status above 70, and RTOG 0825 enrolled centrally confirmed glioblastoma. RTOG 0825 and AVAglio also reported progression-free survival (RTOG 0825: 10.7 vs 7.3 months, below the prespecified target; AVAglio: 10.6 vs 6.2 months, investigator-assessed), which differs from the overall survival drawn here. AVAglio is not drawn because its abstract gives no overall survival medians (hazard ratio 0.88, 95% CI 0.76 to 1.02). EORTC 1709 medians are given as standard versus marizomib arm, without confidence intervals in the abstract. Eras, backbone treatment and follow-up differ; compare only each trial's own two arms. The chart is not a pooled or network estimate.

Each pair shows one trial: its control arm and its experimental arm. Yesterday's experimental arm is often today's control. Populations, definitions and follow-up differ.

05101520OS, median (months)200520142014eortc-26981-22981RTOG 0825MIRAGE
  • experimental arm
  • control arm
  • different endpoint from the lane's
  • step line through experimental values
OS, median, across 3 trials in this lane. eortc-26981-22981 (2005): experimental 14.6 months vs control 12.1 months; RTOG 0825 (2014): experimental 15.7 months vs control 16.1 months; MIRAGE (2014): experimental 16.5 months vs control 17 months.
Data table
TrialYearEndpointControl (months)Experimental (months)
eortc-26981-229812005OS12.1 months14.6 months
RTOG 08252014OS16.1 months15.7 months
MIRAGE2014OS17 months16.5 months

Newly diagnosed glioblastoma, MGMT-methylated (or methylated/indeterminate) tumours

Adults with newly diagnosed glioblastoma whose trial eligibility was limited to MGMT promoter-methylated tumours (CENTRIC) or to methylated or indeterminate tumours (CheckMate 548) · Adjuvant · endpoint: OS

Results are not comparable across trials.

This lane is kept separate from unselected and MGMT-unmethylated trials. CENTRIC enrolled methylated tumours only; CheckMate 548 enrolled methylated or indeterminate tumours, so its population is not strictly methylated. Both control arms received radiotherapy plus temozolomide (CENTRIC without add-on; CheckMate 548 with placebo). CheckMate 548 also reported progression-free survival (10.6 vs 10.3 months); only overall survival is drawn. CeTeG (lomustine-temozolomide, MGMT-methylated, age 18 to 70) is not drawn here because its milestone is tagged only to the maintenance setting; see the milestone for its result. Eras and follow-up differ; compare only each trial's own two arms.

Each pair shows one trial: its control arm and its experimental arm. Yesterday's experimental arm is often today's control. Populations, definitions and follow-up differ.

010203040OS, median (months)20142020CENTRICCheckMate 548
  • experimental arm
  • control arm
  • different endpoint from the lane's
  • step line through experimental values
OS, median, across 2 trials in this lane. CENTRIC (2014): experimental 26.3 months vs control 26.3 months; CheckMate 548 (2020): experimental 28.9 months vs control 32.1 months.
Data table
TrialYearEndpointControl (months)Experimental (months)
CENTRIC2014OS26.3 months26.3 months
CheckMate 5482020OS32.1 months28.9 months

Newly diagnosed glioblastoma, MGMT-unmethylated tumours

Adults with newly diagnosed glioblastoma with unmethylated MGMT promoter: the CheckMate 498 population and the MGMT-unmethylated analysis of EORTC 1709 marizomib · Adjuvant · endpoint: OS

Results are not comparable across trials.

CheckMate 498 enrolled unmethylated tumours only and compared nivolumab plus radiotherapy with temozolomide plus radiotherapy, so the experimental arm did not receive temozolomide. The EORTC 1709 value is a subgroup analysis of an all-patient trial; its abstract gives no confidence intervals and does not state the direction of the hazard ratio. A 2024 reanalysis of older-adult trials questioned benefit of temozolomide in truly unmethylated tumours (see the CE.6 and Stupp records). Eras differ; compare only each trial's own two arms.

Each pair shows one trial: its control arm and its experimental arm. Yesterday's experimental arm is often today's control. Populations, definitions and follow-up differ.

05101520OS, median (months)20142020MIRAGECheckMate 498
  • experimental arm
  • control arm
  • different endpoint from the lane's
  • step line through experimental values
OS, median, across 2 trials in this lane. MIRAGE (2014): experimental 15.1 months vs control 14.5 months; CheckMate 498 (2020): experimental 13.4 months vs control 14.9 months.
Data table
TrialYearEndpointControl (months)Experimental (months)
MIRAGE2014OS14.5 months15.1 months
CheckMate 4982020OS14.9 months13.4 months

Newly diagnosed glioblastoma, maintenance-phase comparisons

Adults with newly diagnosed glioblastoma in trials that tested a change to the maintenance phase of temozolomide treatment (RTOG 0525; EF-14), not selected by MGMT status · Maintenance · endpoint: OS

Results are not comparable across trials.

EF-14 randomised patients only after they had completed concomitant chemoradiation (median 3.8 months after diagnosis), was open-label, and used a device; its plotted figure is the 2017 final intention-to-treat analysis (the 2015 interim overall survival was per-protocol). RTOG 0525 compared standard with dose-dense temozolomide, and the records do not state when its randomisation occurred relative to chemoradiation. RTOG 0525 and EF-14 also reported progression-free survival (RTOG 0525: 5.5 vs 6.7 months; EF-14: 6.7 vs 4.0 months), which is not the endpoint drawn. Neither trial's start point matches the trials in the chemoradiation lane. Compare only each trial's own two arms.

Each pair shows one trial: its control arm and its experimental arm. Yesterday's experimental arm is often today's control. Populations, definitions and follow-up differ.

0510152025OS, median (months)20132015RTOG 0525EF-14
  • experimental arm
  • control arm
  • different endpoint from the lane's
  • step line through experimental values
OS, median, across 2 trials in this lane. RTOG 0525 (2013): experimental 14.9 months vs control 16.6 months; EF-14 (2015): experimental 20.9 months vs control 16 months.
Data table
TrialYearEndpointControl (months)Experimental (months)
RTOG 05252013OS16.6 months14.9 months
EF-142015OS16 months20.9 months

Older adults with newly diagnosed glioblastoma

Older adults with newly diagnosed glioblastoma (CE.6: age 65 or older; Keime-Guibert: age 70 or older with Karnofsky performance score 70 or higher) · Adjuvant · endpoint: OS

Results are not comparable across trials.

Age limits, radiotherapy schedules and comparators differ: Keime-Guibert compared 50 Gy plus supportive care with supportive care alone, CE.6 compared 40 Gy in 15 fractions with or without temozolomide. The Keime-Guibert abstract gives median survival in weeks (29.1 vs 16.9), which is not stored in months, so only CE.6 is drawn. CE.6 MGMT-methylated (13.5 vs 7.7 months) and unmethylated (10.0 vs 7.9 months) results are exploratory subgroups with P=0.08 for interaction and are not drawn, to keep MGMT strata separate. CE.6 also reported progression-free survival (5.3 vs 3.9 months). These trials do not share a control arm; compare only each trial's own two arms.

Each pair shows one trial: its control arm and its experimental arm. Yesterday's experimental arm is often today's control. Populations, definitions and follow-up differ.

02468101214OS, median (months)2017CE.6
  • experimental arm
  • control arm
  • different endpoint from the lane's
  • step line through experimental values
OS, median, across 1 trial in this lane. CE.6 (2017): experimental 9.3 months vs control 7.6 months.
Data table
TrialYearEndpointControl (months)Experimental (months)
CE.62017OS7.6 months9.3 months

Glioblastoma at first progression after radiochemotherapy

Adults with glioblastoma progressing after standard radiation and temozolomide (EORTC 26101: first progression after chemoradiation; CheckMate 143: first recurrence; BRAIN: recurrent glioblastoma) · Relapsed / refractory · endpoint: OS

Results are not comparable across trials.

Kept separate from newly diagnosed disease. Controls differ: lomustine alone (EORTC 26101) and bevacizumab (CheckMate 143), and the experimental regimens differ. Several of these trials report endpoints that point in different directions: EORTC 26101 locally assessed progression-free survival was 4.2 vs 1.5 months (hazard ratio 0.49) while overall survival, drawn here, was 9.1 vs 8.6 months (hazard ratio 0.95); CheckMate 143 overall survival was comparable between arms while objective response rate was 23.1% with bevacizumab and 7.8% with nivolumab. The 2009 approval rested on response rate in the BRAIN trial, which was randomised but noncomparative (both arms were bevacizumab-containing), so it has no control arm and is not drawn. Prior treatment and follow-up differ; compare only each trial's own two arms.

Each pair shows one trial: its control arm and its experimental arm. Yesterday's experimental arm is often today's control. Populations, definitions and follow-up differ.

02468101214OS, median (months)20172020EORTC 26101CheckMate 143
  • experimental arm
  • control arm
  • different endpoint from the lane's
  • step line through experimental values
OS, median, across 2 trials in this lane. EORTC 26101 (2017): experimental 9.1 months vs control 8.6 months; CheckMate 143 (2020): experimental 9.8 months vs control 10 months.
Data table
TrialYearEndpointControl (months)Experimental (months)
EORTC 261012017OS8.6 months9.1 months
CheckMate 1432020OS10 months9.8 months

Standard of care, then and now

How the usual sequence of treatment changed for one setting. Choose a point in time; each step links to the milestone behind it.

Newly diagnosed glioblastoma: treatment after surgery

Adults with newly diagnosed glioblastoma after resection or biopsy (1978: patients with anaplastic glioma) · Adjuvant

  1. Radiation1Radiotherapy with daily temozolomideTemozolomide added to radiotherapy for newly diagnosed glioblastoma (2005)
  2. Chemotherapy2Six cycles of adjuvant temozolomideTemozolomide added to radiotherapy for newly diagnosed glioblastoma (2005)
  3. Targeted3Tumour treating fields with temozolomideTumour treating fields added to maintenance temozolomide (EF-14) (2015)
Today. Based only on the trials cited in each step; practice varies. Later add-on agents in the cited trials (bevacizumab, cilengitide, marizomib) did not improve overall survival; the place of tumour treating fields is debated.
Text version
  • 1978: Postoperative radiotherapy
  • 2005: Radiotherapy with daily temozolomide → Six cycles of adjuvant temozolomide
  • 2015: Radiotherapy with daily temozolomide → Six cycles of adjuvant temozolomide → Tumour treating fields with temozolomide
  • Today: Radiotherapy with daily temozolomide → Six cycles of adjuvant temozolomide → Tumour treating fields with temozolomide

Reversals and dead ends

Approaches that were abandoned or withdrawn, and trials whose results were later questioned, are part of the story.

  • 2013

    Dose-dense temozolomide shows no survival gain (RTOG 0525)

    A 2013 trial of 833 people with newly diagnosed glioblastoma compared standard temozolomide with a dose-dense schedule after chemoradiation. Median survival was about 17 months with the standard schedule and 15 months with the dose-dense schedule, not a significant difference. Severe side effects were more common with the dose-dense schedule.

    RTOG 0525 (833 randomised of 1,173 registered): standard vs dose-dense temozolomide, median OS 16.6 vs 14.9 months (HR 1.03; P=.63); median PFS 5.5 vs 6.7 months (HR 0.87; P=.06). Treatment effect did not differ by MGMT methylation status. Across both arms, MGMT methylation was associated with better OS (21.2 vs 14 months; HR 1.74 as printed, between methylation groups; P<.001). Grade 3 or higher toxicity 34% vs 53% (P<.001).

  • 2014

    Bevacizumab added to chemoradiation does not improve survival

    In 2014 two large randomised trials added bevacizumab to radiotherapy and temozolomide for newly diagnosed glioblastoma. People given bevacizumab went longer before their tumour progressed, but overall survival was not significantly different, and more had severe side effects.

    AVAglio (458 bevacizumab, 463 placebo): median PFS 10.6 vs 6.2 months, stratified HR 0.64 (95% CI 0.55 to 0.74); OS stratified HR 0.88 (95% CI 0.76 to 1.02; P=0.10); grade 3 or higher adverse events 66.8% vs 51.3%. RTOG 0825 (n=637 randomised): median OS 15.7 vs 16.1 months (HR 1.13); PFS 10.7 vs 7.3 months (HR 0.79), below the prespecified target.

  • 2014

    Three further add-on agents fail in randomised phase 3 trials

    Three randomised phase 3 trials added a new drug to radiotherapy and temozolomide in newly diagnosed glioblastoma: cilengitide (reported 2014), a vaccine called rindopepimut (2017) and marizomib (2024). None improved overall survival, and the cilengitide authors said it would not be developed further.

    CENTRIC (MGMT-methylated, n=545): median OS 26.3 vs 26.3 months, HR 1.02 (95% CI 0.81 to 1.29; P=0.86). ACT IV (EGFRvIII-expressing; 745 enrolled; minimal residual disease group): median OS 20.1 vs 20.0 months, HR 1.01 (95% CI 0.79 to 1.30); terminated for futility at interim. EORTC 1709 marizomib (n=749): median OS 17 months with standard treatment vs 16.5 with marizomib, HR 1.04 (P=.64); more grade 3 or 4 adverse events with marizomib.

  • 2017

    Adding bevacizumab to lomustine does not extend survival (EORTC 26101)

    In 2017 a trial of 437 people whose glioblastoma had progressed after first treatment added bevacizumab to lomustine chemotherapy. The cancer stayed controlled for longer on scans, but median survival was about nine months with the combination and about nine without it.

    EORTC 26101 (n=437 randomised 2:1; 288 lomustine plus bevacizumab, 149 lomustine alone) at first progression after chemoradiation: median OS 9.1 vs 8.6 months, HR 0.95 (95% CI 0.74 to 1.21; P=0.65). Locally assessed PFS 4.2 vs 1.5 months, HR 0.49 (95% CI 0.39 to 0.61). Grade 3 to 5 adverse events 63.6% vs 38.1%. MGMT status was prognostic.

  • 2020

    Nivolumab phase 3 trials in glioblastoma do not meet their endpoints

    Three randomised phase 3 trials tested the immunotherapy nivolumab in glioblastoma: one at first recurrence and two alongside first-line radiotherapy. In none did people live longer with nivolumab. In one first-line trial, those given temozolomide with radiotherapy lived longer than those given nivolumab with radiotherapy.

    CheckMate 143 (recurrent, n=369): nivolumab vs bevacizumab, median OS 9.8 vs 10.0 months, HR 1.04 (95% CI 0.83 to 1.30). CheckMate 498 (MGMT-unmethylated, n=560): nivolumab plus radiotherapy vs temozolomide plus radiotherapy, median OS 13.4 vs 14.9 months, HR 1.31 (95% CI 1.09 to 1.58). CheckMate 548 (MGMT-methylated or indeterminate, n=716): median OS 28.9 vs 32.1 months, HR 1.1.

Today and tomorrow

Standard of care today

Adjuvant

Adults with newly diagnosed glioblastoma after surgery or biopsy; regulatory documents read. FDA, 15 March 2005 (NDA 21-029/S-008):

  • temozolomide for newly diagnosed glioblastoma multiforme concomitantly with radiotherapy and then as maintenance treatment. The letter states that approval of the supplement fulfils commitments made under 21 CFR 314.510, the accelerated approval regulations, for the application
  • FDA's verified-clinical-benefit list records the earlier accelerated approval (11 August 1999, adults with refractory anaplastic astrocytoma) with a traditional approval date of 15 March 2005. FDA, 5 October 2015 (device premarket approval supplement P100034/S013, a device and not a drug approval): Optune with temozolomide for adults with newly diagnosed supratentorial glioblastoma following maximal debulking surgery and completion of radiation therapy together with concomitant standard of care chemotherapy
  • the order's general statement covers adults 22 years or older with histologically confirmed glioblastoma. EU: the EMA Temodal EPAR page (last updated 18 August 2026) lists, among its indications, adults with newly diagnosed glioblastoma multiforme concomitantly with radiotherapy and subsequently as monotherapy
  • the page states a marketing authorisation date of 26 January 1999, and the date on which this indication was added was not read. The trials behind these actions are in the milestone timeline. The EANO guidelines on the diagnosis and treatment of diffuse gliomas of adulthood (Nat Rev Clin Oncol 2021, PMID 33293629) are cited by identifier only
  • its text was not read, so no recommendation is attributed to it (a 2025 SNO and EANO consensus review of glioblastoma management, Neuro Oncol, PMID 40827022, exists
  • only its abstract was read and it is not cited, so the 2021 EANO guideline may not be the most recent EANO-linked document for glioblastoma). NCCN guideline text was not opened
US Food and Drug Administration, Center for Drug Evaluation and Research 2005US Food and Drug Administration 2026US Food and Drug Administration 2026US Food and Drug Administration, Center for Devices and Radiological Health 2015European Medicines Agency 2026Guideline: Weller M, van den Bent M, Preusser M, Le Rhun E, Tonn JC, Minniti G, et al 2021
Perioperative
Adults with suspected high-grade glioma having resection; regulatory documents read. FDA, 6 June 2017 (NDA 208630): Gleolan (aminolevulinic acid hydrochloride) for oral solution as an optical imaging agent in patients with glioma (suspected WHO grades III or IV on preoperative imaging) as an adjunct for the visualization of malignant tissue during surgery; the letter contains no accelerated-approval wording. EU: the EMA Gliolan EPAR page (last updated 30 October 2024) lists visualisation of malignant tissue during surgery for malignant glioma (WHO grade III and IV) in adult patients and states that the European Commission granted the marketing authorisation on 7 September 2007. The randomised trial of fluorescence-guided resection is in the milestone timeline. Labelling of the carmustine implant, and guideline text for surgery, were not read.
US Food and Drug Administration, Center for Drug Evaluation and Research 2017European Medicines Agency 2026
Relapsed / refractory

Adults with recurrent or progressive glioblastoma; regulatory documents read. FDA, 5 May 2009 (BLA 125085/S-169):

  • bevacizumab for glioblastoma with progressive disease following prior therapy, under the accelerated approval regulations (21 CFR 601.40-46), with a required confirmatory study in newly diagnosed glioblastoma. FDA, 5 December 2017 (S-319): supplement approval fulfilling that postmarketing requirement, which removed the label statement that effectiveness was based on improvement in objective response rate and that no data demonstrated improvement in disease-related symptoms or survival
  • FDA's verified-clinical-benefit list (content current as of 16 October 2026) gives 5 December 2017 as the traditional approval date for this indication. FDA's review document states that the benefit was supported by progression-free survival data from three randomised trials (EORTC 26101 in recurrent disease
  • AVAglio and RTOG 0825, in newly diagnosed disease, are described as supporting the consistency of the effect) and by corticosteroid-use data, and that EORTC 26101 did not show a statistically significant overall survival difference. FDA, 8 April 2011 (device premarket approval P100034): the NovoTTF-100A system, later named Optune, as monotherapy for adults 22 years or older with histologically confirmed glioblastoma after recurrence following chemotherapy, intended as an alternative to standard medical therapy after surgical and radiation options have been exhausted (indication wording from the FDA PMA database record, and from the 2015 order, which describes the 2011 indication). EU: the EMA Temodal EPAR page lists temozolomide for children from 3 years, adolescents and adults with malignant glioma, such as glioblastoma multiforme or anaplastic astrocytoma, showing recurrence or progression after standard therapy. The indications listed on the EMA Avastin EPAR page (last updated 27 January 2026) do not include glioma. Lomustine labelling was not researched, and no guideline text was read for this setting
US Food and Drug Administration, Center for Drug Evaluation and Research 2009US Food and Drug Administration, Center for Drug Evaluation and Research 2017US Food and Drug Administration, Center for Drug Evaluation and Research 2017US Food and Drug Administration 2026US Food and Drug Administration 2026US Food and Drug Administration, Center for Devices and Radiological Health 2015European Medicines Agency 2026EMA 2026
Adjuvant
IDH-mutant grade 2 astrocytoma or oligodendroglioma after surgery; this is not a glioblastoma indication, and is listed because IDH-mutant gliomas are classified separately from glioblastoma, IDH-wildtype. FDA, 6 August 2024 (NDA 218784): vorasidenib for adult and paediatric patients 12 years and older with grade 2 astrocytoma or oligodendroglioma with a susceptible IDH1 or IDH2 mutation, following surgery including biopsy, sub-total resection, or gross total resection; the letter contains no accelerated-approval wording. EU: the EMA Voranigo EPAR page (last updated 27 August 2026) states a marketing authorisation of 17 September 2025, and its overview describes use in adults and adolescents from 12 years who weigh at least 40 kg, have had surgery as their only treatment and do not need other treatments such as radiotherapy or chemotherapy immediately. Guideline text was not read.
US Food and Drug Administration, Center for Drug Evaluation and Research 2024European Medicines Agency 2026
Relapsed / refractory
BRAF V600E-mutant glioma, paediatric indications and a tumour-agnostic indication; regulatory documents read. FDA, 22 June 2022 (NDA 202806/S-022): dabrafenib with trametinib for adult and paediatric patients 6 years and older with unresectable or metastatic solid tumours with a BRAF V600E mutation who have progressed following prior treatment and have no satisfactory alternative treatment options, under accelerated approval based on overall response rate and duration of response; FDA's list of ongoing cancer accelerated approvals (content current as of 19 October 2026) still includes it. FDA, 16 March 2023 (S-025): dabrafenib for paediatric patients 1 year and older with low-grade glioma with a BRAF V600E mutation who require systemic therapy. EU: the EMA Finlee EPAR page (last updated 30 March 2026) states a marketing authorisation of 15 November 2023 and lists dabrafenib with trametinib for paediatric patients 1 year and older with BRAF V600E low-grade glioma requiring systemic therapy, and with BRAF V600E high-grade glioma after at least one prior radiation and/or chemotherapy treatment. The Tafinlar EPAR page text read does not mention glioma, and no adult glioma indication in the EU was found. Guideline text was not read.
US Food and Drug Administration, Center for Drug Evaluation and Research 2022FDA 2026US Food and Drug Administration, Center for Drug Evaluation and Research 2023European Medicines Agency 2026European Medicines Agency 2026
Relapsed / refractory
H3 K27M-mutant diffuse midline glioma; regulatory document read. FDA, 6 August 2025 (NDA 219876): dordaviprone (Modeyso) for adult and paediatric patients 1 year of age and older with diffuse midline glioma harbouring an H3 K27M mutation with progressive disease following prior therapy, under accelerated approval (section 506(c) of the Federal Food, Drug, and Cosmetic Act and 21 CFR 314.510). FDA's list of ongoing cancer accelerated approvals (content current as of 19 October 2026) includes it and lists a postmarketing requirement for a multiregional randomised trial with overall survival as the primary endpoint; the list does not name the trial. No EU document and no guideline text were read for this indication.
US Food and Drug Administration, Center for Drug Evaluation and Research 2025FDA 2026
Any setting
Classification and response assessment (no regulatory action); cited by identifier, with only the abstracts read, so no recommendation is attributed to any of them. The summary of the fifth edition of the WHO classification of CNS tumours (2021; Neuro Oncol 2021, PMID 34185076) describes major changes that increase the role of molecular diagnostics. cIMPACT-NOW update 12 (Neuro Oncol, 1 September 2026, PMID 42152226) addresses grading of IDH-mutant gliomas, and its introduction states that IDH-mutant diffuse gliomas were first recognised by the WHO in 2016, when they were separated from glioblastoma, IDH-wildtype. RANO 2.0 (J Clin Oncol 2023, PMID 37774317) is the update to the response criteria for high- and low-grade gliomas in adults; its abstract states that, for IDH wild-type glioblastoma, non-enhancing disease will no longer be evaluated except when assessing response to antiangiogenic agents.
Louis DN, Perry A, Wesseling P, Brat DJ, Cree IA, Figarella-Branger D, et al 2021Brat DJ, Aldape K, French PJ, Louis DN, Nasrallah MP, Reifenberger G, et al 2026Wen PY, van den Bent M, Youssef G, Cloughesy TF, Ellingson BM, Weller M, et al 2023

Trials to watch

  • NRG-BN011 NCT05095376

    In newly diagnosed glioblastoma with a methylated MGMT promoter, does adding lomustine to temozolomide with radiotherapy improve overall survival compared with standard temozolomide with radiotherapy? Registry record: randomised, open-label phase 3 (NRG Oncology); status active, not recruiting; enrolment 265 (actual); the registry lists overall survival as the primary outcome; primary completion 12 February 2026 (actual); no results posted as of the registry record last updated 17 August 2026.

  • EF-41 / KEYNOTE D58 NCT06556563

    In newly diagnosed glioblastoma, does adding pembrolizumab to Optune (tumour treating fields) with maintenance temozolomide improve overall survival compared with placebo? Registry record: randomised, double-blind, placebo-controlled phase 3 (NovoCure GmbH); status recruiting; enrolment 741 (estimated); primary outcome overall survival at 24 months; estimated primary completion April 2029; no results posted.

  • Niraparib versus temozolomide NCT06388733

    In adults with newly diagnosed glioblastoma and an unmethylated MGMT promoter, does niraparib differ from temozolomide in overall survival? Registry record: randomised, open-label phase 3 (Ivy Brain Tumor Center); status active, not recruiting; enrolment 450 (estimated); primary outcome overall survival at 24 months; estimated primary completion December 2027; no results posted.

  • LEGATO NCT05904119

    At first progression of glioblastoma after standard chemoradiotherapy, does adding reirradiation to lomustine improve overall survival compared with lomustine alone? Registry record: randomised, open-label phase 3 (EORTC); status recruiting; enrolment 411 (estimated); primary outcome overall survival; estimated primary completion September 2027; no results posted. A design paper (Trials 2024, PMID 38849943) is published; no efficacy report was found.

  • SONOBIRD NCT05902169

    In IDH-wildtype glioblastoma at first recurrence with planned resection, does carboplatin delivered with the implanted SonoCloud-9 ultrasound device improve overall survival compared with single-agent lomustine or temozolomide? Registry record: randomised, open-label phase 3 (CarThera); status recruiting; enrolment 560 (estimated); primary outcome overall survival up to 24 months; estimated primary completion January 2028; no results posted.

  • ACTION NCT05580562

    In newly diagnosed H3 K27M-mutant diffuse glioma after radiotherapy, does dordaviprone (ONC201) improve overall survival compared with placebo? Registry record: randomised, double-blind, placebo-controlled phase 3 (Jazz Pharmaceuticals) with once-weekly, twice-weekly and placebo groups; status recruiting; enrolment 510 (estimated); primary outcome overall survival; estimated primary completion June 2028; no results posted. A design paper (Neuro Oncol 2024, PMID 38445964) is published. This is a glioma subtype other than glioblastoma.

References

Show all 113 references

Loading references… Open the list as a page

Every figure was checked against its primary source to the extent possible. Items that could not be fully confirmed are marked “needs review”. Educational history, not medical advice.

Report an error

Links open the primary source. A flag marked ⚑ needs review means a value or claim has not yet been independently confirmed.

2015TreatmentTargeted

majorMaintenance

Tumour treating fields added to maintenance temozolomide (EF-14)

In 2015 FDA approved a device that delivers electric fields through the scalp (tumour treating fields) for newly diagnosed glioblastoma. The open-label EF-14 trial, which enrolled people only after chemoradiation, reported median overall survival of about 21 months with the device plus temozolomide and 16 months with temozolomide alone. Its place in care is debated.

EF-14 (695 randomised 2:1 after chemoradiation, open-label): TTFields plus maintenance temozolomide vs temozolomide alone. Final analysis (ITT): median PFS 6.7 vs 4.0 months (HR 0.63; 95% CI 0.52-0.76; P<.001); median OS 20.9 vs 16.0 months (HR 0.63; 0.53-0.76; P<.001). Skin toxicity under the arrays in 52% vs none. Randomised after a median 3.8 months from diagnosis.

Standard before

Radiotherapy with concomitant temozolomide followed by maintenance temozolomide; EF-14 randomised patients after they had completed concomitant chemoradiation.

What changed

FDA approved Optune with temozolomide for newly diagnosed supratentorial glioblastoma following maximal debulking surgery and completion of radiotherapy with concomitant standard-of-care chemotherapy (device PMA supplement P100034/S013, 5 October 2015). Use remains debated: in a 2020 survey of 30 respondents, 60% were convinced that TTFields prolong survival and 30% viewed it as definitely part of standard care.

Why it mattered

A randomised trial in which a non-drug device added to maintenance temozolomide was followed by longer progression-free and overall survival; the open-label design without a sham device and the selection of trial participants are among the critiques summarised in a 2020 report on usage, which also records limited uptake.

Caveats

  • The trial was open-label (final-analysis abstract); neither abstract describes a sham device, and a 2020 report on TTFields usage lists the absence of a sham device among the critiques of the trial design.
  • Patients were randomised after completing concomitant chemoradiation (median 3.8 months after diagnosis), so the comparison starts at maintenance therapy and is not comparable with trials that randomise at the start of chemoradiation. Not plottable in the newly diagnosed chemoradiation lane.
  • The 2015 interim analysis reported overall survival in the per-protocol population (n=280), not the intention-to-treat population; the 2017 final analysis used the intention-to-treat population. Interim hazard ratios carry 98.7% and 99.4% confidence intervals.
  • Sample size: the paper randomised 695 of 700 planned; ClinicalTrials.gov NCT00916409 lists 700 as an estimated enrolment.
  • Usage survey (Lassman 2020, abstract): 3% to 12% of newly diagnosed and 0% to 16% of recurrent patients at two academic programmes (New York City and Heidelberg), and 30 respondents at one ASCO session; it does not represent practice elsewhere. A 2017 critical review of the roundtable discussion is cited; only its abstract was read.
  • The approval is a device premarket approval (supplement S013, Panel Track), not a drug approval; FDA originally approved the device for recurrent glioblastoma on 8 April 2011 (openFDA PMA record). The 5 October 2015 date and indication are confirmed by the FDA PMA database record and the S013 approval order; the summary of safety and effectiveness was not opened.