Citation, DOI, disclosures and article data
At the time the article was created Frank Gaillard had no recorded disclosures.View Frank Gaillard's current disclosures
Glioblastomas (GBM) are the most common adult primary brain tumor and are unfortunately aggressive, relatively resistant to therapy, and have a corresponding poor prognosis.
They typically appear as heterogeneous masses centered in the white matter with irregular peripheral enhancement, central necrosis, and surrounding vasogenic edema.
Treatment primarily consists of surgery with concurrent radiotherapy and temozolomide.
Since 1926 when the term "glioblastoma multiforme" was coined, the definition of this tumor has substantially changed, particularly over the past decade with an increasing reliance on molecular markers to define these tumors.
In the 5th edition (2021) of the WHO classification of CNS tumors, glioblastomas have been defined as diffuse astrocytic tumors in adults that must be IDH-wildtype, and are now an entirely separate diagnosis from astrocytoma, IDH-mutant grade 2, 3 or 4 5.
Glioblastoma was previously known as glioblastoma multiforme; the multiforme referred to the tumor heterogeneity. In the revised 4th edition (2016) of the WHO classification, the term "multiforme" was dropped, with these tumors referred to merely as glioblastomas. In the revised 4th edition, the abbreviation GBM was kept for disambiguation 16 however it appears to have been deprecated in the 5th edition summary 20.
Primary and secondary
Glioblastomas had traditionally been divided into primary and secondary; the former arising de novo (90%) and the latter developing from a pre-existing lower grade tumor (10%).
These historical terms now correlate closely to IDH-mutation status but should no longer be used.
Primary glioblastomas largely equate to glioblastoma, IDH-wildtype, whereas secondary glioblastomas now equate to astrocytoma, IDH-mutant, WHO CNS grade 4.
In the 5th edition (2021) WHO classification of CNS tumors, three glioblastoma histological variants are recognized (which are discussed separately), as well as a number of histological patterns which are discussed below 16.
The three recognized variants are:
Glioblastomas, now defined as IDH-wildtype tumors, are essentially tumors of adults, usually occurring after the age of 40 years with a peak incidence between 65 and 75 years of age. There is a slight male preponderance with a 3:2 M:F ratio 5. White patients are affected more frequently than other ethnicities: the prevalence in Europe and North America is 3-4 per 100,000, whereas in Asia it is 0.59 per 100,000 16.
The vast majority of glioblastomas are sporadic. Rarely they are related to prior radiation exposure (radiation-induced glioma). They can also occur as part of rare inherited tumor syndromes, such as p53 mutation-related syndromes including neurofibromatosis type 1 (NF1) and Li-Fraumeni syndrome. Other syndromes in which glioblastomas are encountered include Turcot syndrome, Ollier disease, and Maffucci syndrome.
Typically patients present in one of three ways:
focal neurological deficit
symptoms of increased intracranial pressure
Rarely (<2%) intratumoral hemorrhage occurs and patients may present acutely with stroke-like symptoms and signs.
The 5th edition (2021) of the WHO classification of CNS tumors incorporates molecular parameters into the diagnostic criteria. In this classification, to make the diagnosis of a glioblastoma the following are required 20:
diffuse astrocytic tumor
and at least one of the following:
TERT promoter mutation
EGFR gene amplification
combined gain of whole chromosome 7 and loss of chromosome 10 [+7/-10]
In the rare situation where these criteria are not met, it is likely the tumor will be denoted as not elsewhere classified (NEC) although a variety of pediatric-type diffuse gliomas may be worth considering 20.
An IDH wild-type status can be reached without the need for sequencing in patients over the age of 55 years on the basis of negative IDH-1 R132H immunohistochemistry, as the likelihood of finding other IDH mutations in older age is very unlikely 20.
Although glioblastomas can arise anywhere within the brain, they have a predilection for the subcortical white matter and deep grey matter of the cerebral hemispheres, particularly the temporal lobe 16.
Glioblastomas are typically poorly marginated, diffusely infiltrating, necrotic masses localized to the cerebral hemispheres. The supratentorial white matter is the most common location.
These tumors may be firm or gelatinous. Considerable regional variation in appearance is characteristic. Some areas are firm and white, some are soft and yellow (secondary to necrosis), and others are cystic with local hemorrhage. Glioblastomas have significant variability in size from only a few centimeters to lesions that replace a hemisphere. Infiltration beyond the visible tumor margin is always present.
Histologically, pleomorphic astrocytes with marked atypia and numerous mitoses are seen. Necrosis and microvascular proliferation are hallmarks of glioblastomas.
Microvascular proliferation results in an abundance of new vessels with a poorly formed blood-brain barrier (BBB) permitting the leakage of iodinated CT contrast and gadolinium into the adjacent extracellular interstitium resulting in the observed enhancement on CT and MRI respectively 11.
Edema and enhancement are however also seen in lower grade tumors that lack endovascular proliferation (such as diffuse astrocytomas, IDH-mutant) and this is thought to be due to disruption of the normal blood-brain barrier by tumor produced factors. Vascular endothelial growth factor (VEGF) for example has been shown to both disrupt tight junctions between endothelial cells and increase the formation of fenestrations 12.
Glioblastomas are capable of demonstrating varied patterns, sometimes within one tumor. In addition to giant cell glioblastoma, gliosarcoma, and epithelioid glioblastoma, other histological features are sometimes encountered which impact imaging appearance and biological behavior. These include 16:
more commonly seen in grade 4 astrocytomas
histologically mimic macrophages and thus can lead to a misdiagnosis of macrophage-rich demyelination
most commonly squamous epithelium
if dominant feature then a diagnosis of gliosarcoma should be considered
multinucleated giant cells
a common feature of glioblastoma
if they are the dominant feature then a diagnosis of giant cell glioblastoma should be considered
primitive neuronal cells
previously known as glioblastoma with PNET-like component
more frequently has CSF spread
MYC or MYCN amplification common
IDH-mutant in 15-20% of cases
small cell glioblastoma
histologically appears similar to oligodendroglioma, but usually demonstrate EGFR amplification
like oligodendrogliomas, they have a predilection for extensive cortical involvement
GFAP: positive but of variable intensity
p53 protein: positive if TP53 mutated
EGFR: positive in 40-98% of cases 16
IDH-1 R132H: negative (by definition, otherwise not an IDH-wildtype glioblastoma, but rather an astrocytoma, IDH-mutant WHO CNS grade 4) 16
H3 K27M mutation: negative (if positive then diffuse midline glioma H3 K27-altered)
combined gain of whole chromosome 7, loss of chromosome 10 [+7/-10]
alterations of the CDK4/6–RB1 cell-cycle pathway: 80% due to deletions of CDKN2A 20
Glioblastomas are typically large tumors at diagnosis. They often have thick, irregularly enhancing margins and a central necrotic core, which may also have a hemorrhagic component. They are surrounded by vasogenic-type edema, which in fact usually contains infiltration by neoplastic cells.
Multifocal disease, which is found in ~20% of cases, is where multiple areas of enhancement are connected to each other by abnormal white matter signal, which represents microscopic spread to tumor cells. Multicentric disease, on the other hand, is where no such connection can be seen.
irregular thick margins: iso- to slightly hyperattenuating (high cellularity)
irregular hypodense center representing necrosis
marked mass effect
surrounding vasogenic edema
hemorrhage is occasionally seen
calcification is uncommon
intense irregular, heterogeneous enhancement of the margins is almost always present
hypo to isointense mass within the white matter
central heterogeneous signal (necrosis, intratumoral hemorrhage)
T1 C+ (Gd)
enhancement is variable but is almost always present
typically peripheral and irregular with nodular components
usually surrounds necrosis
surrounded by vasogenic edema
flow voids are occasionally seen
Lacks T2/FLAIR "mismatch" sign
susceptibility artifact on T2* from blood products (or occasionally calcification)
low-intensity rim from blood products 6
incomplete and irregular in 85% when present
mostly located inside the peripheral enhancing component
absent dual rim sign
elevated signal on DWI is common in solid/enhancing component
diffusion restriction is typically intermediate similar to normal white matter, but significantly elevated compared to surrounding vasogenic edema (which has facilitated diffusion)
ADC values in the solid component tend to be similar to normal white matter 745 ± 135 x 10-6 mm2/s 13
non-enhancing necrotic/cystic component
the vast majority (>90%) have facilitated diffusion (ADC values >1000 x 10-6 mm2/s)
care must be taken in interpreting cavities with blood product
MR perfusion: rCBV elevated compared to lower grade tumors and normal brain
typical spectroscopic characteristics include
PET demonstrates the accumulation of FDG (representing increased glucose metabolism) which typically is greater than or similar to metabolism in grey matter.
A number of features are seen to correlate with molecular marker status, such as MGMT promoter methylation, which typically demonstrates:
high ADC values
limited surrounding edema
This has a sensitivity of 79% (95% CI, 72-85%) and specificity of 78% (95% CI, 71-84%) 19.
When reporting a new diagnosis of a mass that is likely a glioblastoma, it is useful to include:
size in three dimensions
presence and degree of central necrosis
non-enhancing tumor involving cortex, deep grey or white matter: look at ADC for lower values
relationship to/involvement of
major white matter tracts
Treatment and prognosis
Biopsy and tumor debulking with postoperative adjuvant radiotherapy and chemotherapy (temozolomide) are the most common treatment (Stupp protocol). Newer therapies include antiangiogenesis (e.g. bevacizumab) and immunotherapy.
Although in individuals 70 years of age or younger a standard Stupp protocol is usual, in older individuals the optimum treatment regimen is less well established 15,21. This is particularly the case in the very elderly or those with significant comorbidities 21.
In such cases, surgical resection has less marked survival benefit. Radiotherapy is usually administered as a shorter course (e.g. 25-40 Gy in 5-15 daily fractions, rather than 60 Gy over 6 weeks), but even in this setting adding temozolomide significantly increases survival, especially in MGMT methylated (inactive) tumors 15,21.
Despite all of this, even in the best-case scenario, glioblastoma carries a poor prognosis with a median survival of <2 years 15.
Negative prognostic factors include:
increased necrosis 10
greater enhancement 10
deep location (e.g. thalamus)
lower pre-diagnosis functional status (e.g. ECOG performance status)
Glioblastomas should be followed up closely with MRI.
Immediate post-operative imaging
Generally a scan is obtained within 24-48 hours of surgery to assess residual disease. Early scanning is necessary to avoid post-operative enhancement that can make interpretation difficult. It should be noted, however, that some post-operative enhancement can occur within the first day post surgery 22 and therefore it is essential that these scans are interpreted alongside the pre-operative scan.
Although timing and frequency will vary between institutions and treating surgeons/oncologists typically scans are obtained every 8 to 12 weeks. In individuals who have no residual macroscopic disease and remain stable for a protracted time, the frequency of follow-up imaging can be gradually decreased. In contrast, sometimes it is worthwhile performing an earlier scan to problem-solve ambiguous imaging features.
The primary aims of on-going follow-up are:
identify tumor progression and complications
distinguish tumor progression from pseudoprogression
distinguish pseudoresponse from tumor progression
Response assessment criteria
Glioblastomas have been the subject of close trial scrutiny with many new chemotherapeutic agents showing promise. As such a number of criteria have been created over the years to assess response to treatment. Currently, the response assessment in neuro-oncology (RANO) criteria are most widely used. Other historical systems are worth knowing to allow the interpretation of older data. These systems for response criteria for first-line treatment of glioblastomas include 9:
RANO criteria (most commonly used today)
History and etymology
The original term glioblastoma multiforme was coined in 1926 by Percival Bailey and Harvey Cushing; the suffix multiforme was given to describe the various appearances of hemorrhage, necrosis, and cysts.
General imaging differential considerations include:
astrocytoma, IDH-mutant WHO CNS grade 4
may appear very similar/indistinguishable
generally younger patients
T2/FLAIR mismatch sign is common and highly specific23
may look identical
both may appear multifocal
metastases usually are centered on grey-white matter junction and spare the overlying cortex
rCBV in the "edema" will be reduced
should be considered especially in patients with AIDS, as in this setting central necrosis is more common
otherwise usually homogeneously enhancing
central restricted diffusion is helpful, however, if glioblastoma is hemorrhagic then the assessment may be difficult
presence of smooth and complete SWI low-intensity rim 6
presence of dual rim sign 6
can appear similar
often has an open ring pattern of enhancement
usually younger patients
subacute cerebral infarction
history is essential in suggesting the diagnosis
should not have elevated choline
should not have elevated rCBV
especially in patients with AIDS
- 1. Vinay Kumar, Abul K. Abbas, Nelson Fausto. Robbins and Cotran Pathologic Basis of Disease. (2005) ISBN: 9780721601878 - Google Books
- 2. Rees J, Smirniotopoulos J, Jones R, Wong K. Glioblastoma Multiforme: Radiologic-Pathologic Correlation. Radiographics. 1996;16(6):1413-38; quiz 1462. doi:10.1148/radiographics.16.6.8946545 - Pubmed
- 3. Krex D, Klink B, Hartmann C et al. Long-Term Survival with Glioblastoma Multiforme. Brain. 2007;130(Pt 10):2596-606. doi:10.1093/brain/awm204 - Pubmed
- 4. Jung C, Foerch C, Schänzer A et al. Serum GFAP is a Diagnostic Marker for Glioblastoma Multiforme. Brain. 2007;130(Pt 12):3336-41. doi:10.1093/brain/awm263 - Pubmed
- 5. Wolfgang Dähnert. Radiology Review Manual. (2003) ISBN: 9780781738958 - Google Books
- 6. Toh C, Wei K, Chang C et al. Differentiation of Pyogenic Brain Abscesses from Necrotic Glioblastomas with Use of Susceptibility-Weighted Imaging. AJNR Am J Neuroradiol. 2012;33(8):1534-8. doi:10.3174/ajnr.A2986 - Pubmed
- 7. Ohgaki H & Kleihues P. The Definition of Primary and Secondary Glioblastoma. Clin Cancer Res. 2013;19(4):764-72. doi:10.1158/1078-0432.CCR-12-3002 - Pubmed
- 8. Anne G. Osborn. Osborn's Brain. (2012) ISBN: 9781931884211 - Google Books
- 9. Chinot O, Macdonald D, Abrey L, Zahlmann G, Kerloëguen Y, Cloughesy T. Response Assessment Criteria for Glioblastoma: Practical Adaptation and Implementation in Clinical Trials of Antiangiogenic Therapy. Curr Neurol Neurosci Rep. 2013;13(5):347. doi:10.1007/s11910-013-0347-2 - Pubmed
- 10. Hammoud M, Sawaya R, Shi W, Thall P, Leeds N. Prognostic Significance of Preoperative MRI Scans in Glioblastoma Multiforme. J Neurooncol. 1996;27(1):65-73. doi:10.1007/BF00146086 - Pubmed
- 11. Zagzag D, Goldenberg M, Brem S. Angiogenesis and Blood-Brain Barrier Breakdown Modulate CT Contrast Enhancement: An Experimental Study in a Rabbit Brain-Tumor Model. AJR Am J Roentgenol. 1989;153(1):141-6. doi:10.2214/ajr.153.1.141 - Pubmed
- 12. Zhao L, Yang Z, Liu Y, Ying H, Zhang H, Xue Y. Vascular Endothelial Growth Factor Increases Permeability of the Blood-Tumor Barrier via Caveolae-Mediated Transcellular Pathway. J Mol Neurosci. 2011;44(2):122-9. doi:10.1007/s12031-010-9487-x - Pubmed
- 13. Hilario A, Ramos A, Perez-Nuñez A et al. The Added Value of Apparent Diffusion Coefficient to Cerebral Blood Volume in the Preoperative Grading of Diffuse Gliomas. AJNR Am J Neuroradiol. 2012;33(4):701-7. doi:10.3174/ajnr.A2846 - Pubmed
- 14. Eckel-Passow J, Lachance D, Molinaro A et al. Glioma Groups Based on 1p/19q, IDH, and TERT Promoter Mutations in Tumors. N Engl J Med. 2015;372(26):2499-508. doi:10.1056/NEJMoa1407279 - Pubmed
- 15. Perry J, Laperriere N, O'Callaghan C et al. Short-Course Radiation Plus Temozolomide in Elderly Patients with Glioblastoma. N Engl J Med. 2017;376(11):1027-37. doi:10.1056/NEJMoa1611977 - Pubmed
- 16. Louis D, Ohgaki H, Wiestler O et al. The 2007 WHO Classification of Tumours of the Central Nervous System. Acta Neuropathol. 2007;114(2):97-109. doi:10.1007/s00401-007-0243-4 - Pubmed
- 17. Mulholland S, Pearson D, Hamoudi R et al. MGMT CpG Island is Invariably Methylated in Adult Astrocytic and Oligodendroglial Tumors with IDH1 or IDH2 Mutations. Int J Cancer. 2012;131(5):1104-13. doi:10.1002/ijc.26499 - Pubmed
- 18. Louis D, Perry A, Reifenberger G et al. The 2016 World Health Organization Classification of Tumors of the Central Nervous System: A Summary. Acta Neuropathol. 2016;131(6):803-20. doi:10.1007/s00401-016-1545-1 - Pubmed
- 19. Suh C, Kim H, Jung S, Choi C, Kim S. Clinically Relevant Imaging Features for Promoter Methylation in Multiple Glioblastoma Studies: A Systematic Review and Meta-Analysis. AJNR Am J Neuroradiol. 2018;39(8):1439-45. doi:10.3174/ajnr.A5711 - Pubmed
- 20. Louis D, Giannini C, Perry A, Reifenberger G, et al. Glioblastoma. In: WHO Classification of Tumours Editorial Board. Central nervous system tumours. Lyon (France): International Agency for Research on Cancer; 2021. (WHO classification of tumours series, 5th ed.; vol. 6). https://publications.iarc.fr/601
- 21. Minniti G, Lombardi G, Paolini S. Glioblastoma in Elderly Patients: Current Management and Future Perspectives. Cancers (Basel). 2019;11(3):336. doi:10.3390/cancers11030336 - Pubmed
- 22. Lescher S, Schniewindt S, Jurcoane A, Senft C, Hattingen E. Time Window for Postoperative Reactive Enhancement After Resection of Brain Tumors: Less Than 72 Hours. FOC. 2014;37(6):E3. doi:10.3171/2014.9.focus14479 - Pubmed
- 23. Park S, Suh C, Guenette J, Huang R, Kim H. The T2-FLAIR Mismatch Sign as a Predictor of IDH-Mutant, 1p/19q-Noncodeleted Lower-Grade Gliomas: A Systematic Review and Diagnostic Meta-Analysis. Eur Radiol. 2021;31(7):5289-99. doi:10.1007/s00330-020-07467-4 - Pubmed