International Journal of Neurooncology

REVIEW ARTICLE
Year
: 2020  |  Volume : 3  |  Issue : 2  |  Page : 68--74

Gliomas: History of diagnosis and classification: Part 2


Amitava Ray1, Swain Meenakshi2, Lath Rahul1,  
1 Department of Neurosurgery, Apollo Hospitals, Hyderabad, Telangana, India
2 Department of Pathology, Apollo Hospitals, Hyderabad, Telangana, India

Correspondence Address:
Dr. Amitava Ray
Third Floor, Nirvanaz, 8-2-293/82/A/240, Road 36, Jubilee Hills, Hyderabad - 500 033, Telangana
India

Abstract

In the first article of this series, we traced the history of glioma classification from the middle of the 19th century to the start of the current revolution in classification driven by molecular markers. Although molecular markers were used sporadically in the past to enumerate specific cancer dysregulated pathways, it was never an integral part of the classification. This article tries to summarize the changes that have taken place since the recommendations following the Haarlem meeting in 2014. The objective of this article is to highlight how molecular classification impacts diagnosis, prognostication and management of brain tumors, and the integration with traditional histopathology. This review article is an exhaustive review of the literature summarizes the gradual evolution in glioma classification starting with 1p19q as described in the late 20th century, the value of methyl-guanine-DNA methyltransferase promoter methylation leading up to the World Health Organization 2016 classification. It touches upon the current trends of transcriptomic and epigenetic classifications, highlighting the advantages and pitfalls of each testing method. At this time of rapid advancement, this article helps us understand current grading and classification, offering insights into how this may evolve in the future. Understanding the concepts of integrated molecular and histopathological grading holds the key to future research in the field. It will also be invaluable in being able to choose the best therapeutic option for patients, for whom a plethora of diverse management options may be available.



How to cite this article:
Ray A, Meenakshi S, Rahul L. Gliomas: History of diagnosis and classification: Part 2.Int J Neurooncol 2020;3:68-74


How to cite this URL:
Ray A, Meenakshi S, Rahul L. Gliomas: History of diagnosis and classification: Part 2. Int J Neurooncol [serial online] 2020 [cited 2021 Sep 20 ];3:68-74
Available from: https://www.Internationaljneurooncology.com/text.asp?2020/3/2/68/305064


Full Text



 Introduction



The basis of histopathological classification for gliomas has been the result of the work of the seminal treatise by Bailey and Cushing in 1926.[1] This has formed the mainstay of diagnosis and management of gliomas for almost 100 years. The classification was finally codified in 1979 by the World Health Organization (WHO) guidelines for Central Nervous System Classification and has been subsequently nominally updated three times in 1993, 2000, and 2007.[2] Although this system served the clinicians well overall, the system was not without its shortcomings. Concordance among neuropathologists in determining lineage, i.e., astrocytic versus oligodendroglial versus mixed was seen to be as low as 52%,[3] not taking into account the interobserver differences with regard to the grade, especially between Grade 2 and Grade 3 tumors.[4] These inaccuracies were further confounded by nonrepresentative sampling of small amounts of tissue, usually from surgically inaccessible deep-seated lesions or those lesions affecting the eloquent areas of the brain. To further complicate matters, a histological diagnosis did not invariably predict the outcome, even after taking into account clinical factors such as age, performance status, and extent of resection.[5] The idea that gliomas, though morphologically similar, may represent different disease processes driven by separate molecular pathways had been hypothesized since the mid 1990s. The research that started as a trickle, found renewed impetus in the new millennium and molecular profiling to tease out differences in histologically identical brain tumors showed an exponential rise in the last 15 years, especially in gliomas.[6]

 The International Society of Neuropathologists-Haarlem Reporting Guidelines



In 2014, in keeping with the explosion in molecular information and the obvious pitfalls of the histological classification alone, the International Society of Neuropathologists (ISN) convened an expert group of 28 neuropathologists from ten different countries to address these challenges and discuss the possibility of standardization and an integrated diagnosis. The group had received inputs from over 150 experts from around the world, including neurosurgeons, pediatrics, and adult neurooncologists to set up a framework for molecular testing of brain tumors. Their set of recommendations included defining diagnostic entities as narrowly as possible into almost homogenous groups to optimize reproducibility, in diagnosis, clinicopathological predictions, and therapeutic planning. The ISN-Haarlem system's recommended a “layered” report comprising of both histopathological and molecular information. The WHO 2016 classification, built on the basis of these recommendations, specified the use of four layers – the molecular information of the tumor (layer 4), the WHO grade (layer 3), the histological classification (layer 2), and the final integrated diagnosis (layer 1) – allowing for more accurate diagnosis and prognostication,[7] though the WHO still allows the reporting of “not otherwise specified” (NOS) to reflect incomplete molecular information, either due to lack of tissue or testing facilities. In cases where the molecular information obtained did not fit any particular existing diagnostic category, a different qualifier of “not elsewhere classified (NEC)” was applied.[8] These nuances are further elaborated in The Consortium to Inform Molecular and Practical Approaches to CNS Tumor Taxonomy-Not Otherwise WHO (cIMPACT-NOW) updates, a working committee formed to update the specific questions about molecular diagnosis, in between the WHO editions, which have usually been separated by at least 7 years.[9]

 Low-Grade Gliomas



Traditionally, low-grade glial tumors fall into two distinct categories – the well-circumscribed Grade 1 tumors that have a well-defined border that is usually surgically removed with little or no long-term impact. These include pilocytic astrocytomas, pilocytic xanthoastrocytomas, subependymal giant cell astrocytomas, and mixed glial-neuronal tumors such as gangliogliomas.[10] The low-grade infiltrating tumors consist of astrocytic and oligodendroglial tumors separated by their microscopic appearances – astrocytic tumors characterized by their elongated hyperchromatic nuclei with irregular contours and scant cytoplasm and oligodendroglial tumors with uniform round to oval nuclei, sharp nuclear borders, perinuclear clearing (fried egg appearance) with delicate branching vessels (chicken wire patterns).[11] Infiltrating tumors are, however, characterized by “secondary features” with entrapped neurons within the substance of the tumor and accumulation of tumor cells in the subpial space and around blood vessels. The presence of mitosis, microvascular proliferation, or necrosis placed the tumor in a higher grade.[12]

 1P 19Q-Where the Story Begins



The fact that some oligodendroglial tumors with a deletion of the short arm of chromosome 1 (1p) and the long arm of chromosome 19 (19q) had a better prognosis has been known as early as 1998.[13] The initial retrospective series has since been validated by large prospective studies – the RTOG 9402, demonstrating that 1p/19q co-deleted anaplastic oligodendrogliomas and oligoastrocytomas had a longer survival when treated with radiation alone when compared to their nonco-deleted counterparts, and the addition of PCV (procarbazine, lomustine, and vincristine) doubled the overall survival in co-deleted tumors.[14] The EORTC 2691 presented very similar results from an European cohort of co-deleted tumors.[15] Although the marker has been used informally for treatment and prognostication, this was not a part of the formal classification system till the WHO classification of 2016 [Figure 1].[8]{Figure 1}

As a result of the system of reporting introduced by the WHO in 2016, it is now possible to define tumors on the basis of molecular alterations. A clear demarcation has been drawn between oligodendroglial and glial tumors based on the presence or absence of the 1p19q co-deletion. The diagnosis of oligodendroglioma now has strict molecular criteria-a 1p19q co-deletion along with an isocitrate dehydrogenase (IDH) mutation (described in detail later) and strong nuclear staining of alpha thalassemia/mental retardation gene X linked (ATRX), indicating a nonmutated gene [Figure 2]. Oligodendrogliomas are also associated with point mutations in the telomerase reverse transcriptase (TERT) causing reactivation of TERT expression that is usually silenced in nonneoplastic cells. TERT mutations were initially found in melanomas but have now been found in increasing frequency in oligodendrogliomas and glioblastomas (GBMs). Although the term “oligoastrocytoma” has been discontinued in the WHO 2016 classification, “oligoastrocytoma-NOS” still exists for tumors where molecular testing is not available.{Figure 2}

Traditionally, fluorescent in situ hybridization (FISH) has determined the presence of the 1p/19q co-deletion, using molecular probes that are aligned to the distal ends of chromosome 1p and 19q, and then calculating the ratio of 1p to 1q and 19q to 19p to determine the deletion. Recent evidence, however, has demonstrated that GBMs or other malignant brain tumors can harbor focal deletions and give false-positive results on FISH analysis.[4],[16] Thus in 2016, the WHO recommended molecular testing by a method that tests the whole arm of the chromosome, using SNP microarray, next-generation sequencing with copy number analysis.

 Methyl-Guanine-DNA Methyl Transferase Promoter Methylation



Prognosis continued to define subgroups in gliomas well into the new millennium. In a landmark study published in the New England Journal in 2005, Stupp et al found that tumors with 6 methyl-guanine-DNA methyltransferase (MGMT) promoter methylation had an improved survival of 21.7 versus 12.7 months in tumors where the MGMT promoter was not methylated when treated with radiotherapy and temozolomide.[17] Temozolomide methylates purine bases at O6 and N7 in guanine and N3 on adenine, with O6-methylguanine induction believed to be the primary cytotoxic event. This was later substantiated in longer term studies, where the survival at 2, 3, and 5 years was 48.9%, 27.6%, and 13.8% compared to 14.8%, 11.1%, and 8.3%, respectively.[18] It is interesting to note that TERT promoter mutations in the presence of MGMT promoter methylation seem to increase the sensitivity of the tumor to temozolomide and improve survival, while in the presence of a non-MGMT methylated tumor leads to poorer survival. Hence, an oligodendroglioma that by definition is IDH mutant (usually associated with MGMT hypermethylation), 1p 19q co-deleted has a better prognosis with TERT mutation than otherwise, while in MGMT nonmethyltated GBM, the prognosis with a TERT mutation is worse. MGMT promoter methylation is seen in approximately 40% of gliomas and can be measured in various ways such as methylation microarrays or bi-sulfite sequencing. Although this did constitute an advance in the use of molecular testing, the discovery did not change glioma classification per-sé as it did not identify a distinct glioma subtype.

 Isocitrate Dehydrogenase Mutations



The real breakthrough in glioma classification came in 2008, when next-generation sequencing identified recurrent missense mutations in isocitrate dehydrogenase 1 in a subset of gliomas at position arginine 132, encoded by IDH1 and detected on immunohistochemistry (IHC) by a very specific antibody, or at position arginine 172 of IDH2 [Figure 3]. This was commonly seen in younger patients with a previous history of low-grade gliomas. The mutant forms of IDH acquire neomorphic activity and convert isocitrate to D-2-hydroxyglutarate an “oncometabolite” that builds in tumor cells resulting in changes in DNA and histone methylation patterns and in altered gene expressions. This correlates to a better outcome not only in patients with lower grade tumors but even in IDH mutant GBMs when compared to IDH-wildtype lower grade tumors.[6] IDH mutations are now thought to be associated with early carcinogenesis and may lead to tumors down the oligodendroglial lineage, with 1p19q co-deletions, followed by mutations of CIC and FUBP, or the astrocytic lineage, with mutations in ATRX and TP53. Overcoming replicative senescence and maintaining telomere length in astrocytomas occur through the ATRX along with death domain associated protein. ATRX immunostaining can be patchy, hence positive and negative controls within the same sample must be identified. In addition, not all ATRX mutations are associated with loss of nuclear staining so a classic astrocytoma histology with an IDH mutation, nuclear staining of ATRX does not exclude the diagnosis of astrocytoma.[6]{Figure 3}

TP53 mutations are routinely tested by IHC and manifest as strong nuclear staining. This procedure does however come with certain limitations. While common TP53 missense mutations may be detected by this method, truncating or splice mutations that lead to a loss of TP53 function will lead to a loss of TP53 function and immunoreactivity. In addition, nonneoplastic demyelinating conditions like progressive multifocal leukoencephalopathy can have significant astrocytic atypia and nuclear p53 staining [Figure 4].[19]{Figure 4}

Histologically low-grade glial tumors are now classified into three groups with distinct prognostic groups defined by their molecular signatures. The first group of tumors are those that are IDH1/IDH2 mutated, DNA hypermethylated, and 1p/19q co-deleted – these by definition now are oligodendrogliomas and have a good prognosis with a median survival of 7 years regardless of the histological grade.[20] The second group is also IDH mutant, but 1p 19q intact with mutations in ATRX and TP53 and astrocytic histology – these have intermediate prognosis with a median survival of 5 years. The third group of low-grade tumors are IDH-wild type and 1p19q intact – these tumors usually have the worst prognosis and survival is limited to <2 years, worse than the IDH mutant GBMs, in some cases.[21],[22],[23] However, a group of low-grade tumors with IDH-wt that are closer in association with pilocytic astrocytomas do exist.[20] Acknowledging the dichotomy in outcomes, cIMPACT-NOW in its third update has recommended further molecular testing to identify histologically low-grade tumors IDH wild type with aggressive behavior. Recommendations now include testing epidermal growth factor receptor (EGFR) amplification or TERT mutation or the gain of chromosome 7 and loss of chromosome 10,[24] all of which indicate a poorer prognosis.

 Midline Tumors



Sequencing of gliomas in adults and children revealed additional mutations that are uniquely present in midline structures – mainly the thalamus and the brain stem.[25] These are mutations in the histone family – the protein/DNA complex that maintains the chromatin structure, regulates gene expression, and mediates posttranslational histone modifications including methylation and acetylation. These are usually associated with mutations in histone 3.1 and 3.3 proteins that are encoded by HIST1H3B and H3F3A. Both these mutations lead to a H3K27M substitution, which can be detected by a mutation-specific antibody on IHC, though it does not distinguish between HIST1H3B and H3F3A, as the K27M mutation occurs in a highly conserved region of the histone H3 family. Diffuse pontine gliomas with K27M mutations in H3F3A have a worse prognosis than those with the same mutation in HIST1H3B; all H3K27M mutations irrespective of the histone are associated with a poorer survival. However, the WHO 2016 classification still requires the tumors to be located in the midline with diffuse infiltration of the surrounding structures on histology. Interestingly, H3K27M can also rarely co-occur with BRAF V600E mutations in well-circumscribed glial and glioneuronal tumors.[26],[27]

Another H3F3A mutation occurs at guanine position 34, substituting arginine or valine (G34R/V), which is associated with ATRX and TP53 mutations. These tumors are usually seen in adolescents and young adults and occur in tumors of the hemispheres. On histology, these tumors may show primitive neuronal features with high nuclear-cytoplasmic ratio. Although these tumors are IDH-wild type, they often are MGMT hypermethylated, that leads to a better outcome[28] and have a distinct disease defining molecular signature that separates it from other IDH-wild type tumors. As this is still not defined by the WHO as a separate diagnostic entity, the “not elsewhere classified” modifier is used.[29]

 High-Grade Gliomas



The histological classification of high-grade (WHO Grades 3 and 4) tumors has remained unchanged for a number of decades – predominantly astrocytic in morphology, with an infiltrative growth pattern, vascular proliferation, and/or necrosis.[30] The WHO 2016 classification additionally subdivided GBMs into IDH-mutant and IDH-wildtype tumors – which have significant differences in survival. Although there are no accepted histological characteristics to differentiate these two subclasses, recent reports seem to suggest microthrombi are predictive of wildtype status.[4],[31] The problems of prognostication based on histology do not end there-IDH-wildtype lower grade tumors may have a longer survival than IDH-mutant GBMs, and IDH-wildtype tumors are sometimes defined by other defining mutations – H3K27M or H3.3G34R/V as has been described earlier, which have a better prognosis. In addition, newer IDH-wildtype molecular entities with MYB/MYBL fusions, FGFR1 and FGFR3 alterations, or BRAF mutations have a more indolent course with a better prognosis when compared to traditional GBMs.[32]

In addition to classifications based on specific genes, genetic, epigenetic, and transcriptomic changes have also been used to stratify high-grade gliomas. One of the first such publications was by Phillips et al. in 2006. Using gene expression microarrays, they identified three distinct subsets of GBMs – proneural, proliferative, and mesenchymal/angiogenic.[33] Mesenchymal was the most frequent in occurrence and had the worst prognosis, while proneural had the least frequency of occurrence and the best prognosis. It is interesting to note that all the Grade 3 tumors fell into the proneural category. Recurrence had a tendency to shift toward the mesenchymal subclass, but only in a select few cases. This effort was followed by the cancer genome atlas (TCGA) classification of GBMs which used multidimensional gene expression and genomic clustering to identify 4 gene expression subtypes – classic, proneural, mesenchymal, and neural.[34] The classic subtype is characterized by EGFR amplifications or EGFR mutations, the mesenchymal subtype by a higher frequency of NF1 mutations and the proneural associated with PDGFRA amplifications and IDH mutations, while the neural subtype expressed mainly neuron-specific markers.[34] DNA methylation has also been used as a classifier for both high- and low-grade gliomas. Sturm et al. in 2012 used global methylation clustering data to biologically classify adult and pediatric GBMs and identified six different subgroups – IDH (associated with IDH mutations), K27 (associated with histone H3.3 K27 mutations), G34 (associated with histone H3.3 G34 mutations), RTK1 (associated with PDGFR amplification), RTK2 (associated with EGFR amplifications and mutations), and the mesenchymal subgroup.[35] The neural subgroup was not identified as a separate entity, as was the case in the Phillips study, making the existence of the group controversial.

Besides these well-defined groups, there are rare high-grade gliomas – gliosarcomas constitute about 2% of all high-grade tumors and histologically comprise of both sarcomatous and malignant glial components – with identical mutations usually in TP53, PTEN, and CDKN2A of both components – supporting a monoclonal origin for both histologic components.[36] Giant cell GBMs occur in the relatively young, presenting with a median age of 44 years. However, distinct from the other GBMs in this age group, they are not IDH or ATRX mutant. An ultra-mutated subset of these tumors harbors somatic mutations in the exonuclease domain of the polymerase epsilon gene and has a better survival when compared to the other subtype (26.9 vs. 6.9 months).[37] Epithelioid GBMs have been added as a distinct clinical entity in the new 2016 WHO classification, histologically classified by epithelioid cells with abundant cytoplasm, prominent nuclei in rhabdoid cells. BRAF V600e mutations along with ODZ3 and TERT mutations separate it from other GBMs.[38],[39]

 Conclusion



Over the last few years, there has been an increasing push to recruit patients to clinical trials, not based on the histology but based on the presence of targetable mutations. The National Cancer Institute sponsored Molecular Analysis of Choice trial and the American Society of Clinical Oncology-sponsored Testing the Use of Food and Drug Administration Approved Drugs that Target a Specific Abnormality in a tumor gene is people with Advanced Stage Cancer are in phase II umbrella trials that are enrolling patients based on genetic events, rather than tumor histology.[40] Hence, GBM patients with mutations involving neurofibromin, PI3α, phosphatase, and tensin homologs of the mammalian target of rapamycin will qualify for some of these trials. Although the trial of anti-EGFRIII vaccine, rindopepimut, failed to improve survival in the recent phase III ACTIV trial,[41] there are several other trials that not only target EGFR but other such mutations. Monoclonal antibodies against EGFRvIII mutant receptor are already in phase 1 trials, and vaccines are being developed against IDH mutant and H3F3A-K27M tumors. Dabrafenib and vemurafenib are already approved for the use against tumors harboring the BRAF-V600e mutation. Immune checkpoint inhibitors are also being tried in several trials in tumors with TP53 mutations.[42]

As the inexorable slide toward personalized medicine continues, targeted therapies will become commonplace. These will be based, in the majority, on molecular targets individually or on defined “molecular clusters” based genome or exome wide somatic testing. Expression arrays, methylation profiles, next-generation sequencing accurately predict outcomes and possible targeted therapies. While such tests have already become commonplace in a few hospitals in the developed world, they still remain out of reach for the majority of the world's population. Artificial intelligence and machine learning tools are being used increasingly to interpret histopathology slides, to bridge the knowledge divide, and to minimize interobserver variability. As diagnosis and prognostication get more accurate, the challenge lies in making these state-of-art diagnostic facilities affordable for the majority of the world's population. In the long term, however, only therapeutic breakthroughs based on molecular testing that makes a difference to survival will be considered a success.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.

References

1Bailey P, Cushing H. A Classification of the Tumors of the Glioma Group on a Histogenetic Basis with a Correlated Study of Prognosis. Philadelphia: Lippincott, 1926.
2Coons SW, Johnson PC, Scheithauer BW, Yates AJ, Pearl DK. Improving diagnostic accuracy and interobserver concordance in the classification and grading of primary gliomas. Cancer 1997;79:1381-93.
3Perry A, Wesseling P. Histologic classification of gliomas. In: Handbook of Clinical Neurology. Amsterdam: Elsevier; 2016. p. 71-95. Available from: https://linkinghub.elsevier.com/retrieve/pii/B9780128029978000050. [Last accessed on 2019 Dec 28].
4Louis DN, Perry A, Burger P, Ellison DW, Reifenberger G, von Deimling A, et al. international society of neuropathology-Haarlem consensus guidelines for nervous system tumor classification and grading. Brain Pathol 2014;24:429-35.
5Wesseling P, Capper D. WHO 2016 Classification of gliomas. Neuropathol Appl Neurobiol 2018;44:139-50.
6Louis DN, Aldape K, Brat DJ, Capper D, Ellison DW, Hawkins C, et al. Announcing cIMPACT-NOW: The consortium to inform molecular and practical approaches to CNS tumor taxonomy. Acta Neuropathol 2017;133:1-3.
7Earnest F, Kernohan JW, Craig WM. Oligodendrogliomas; A review of 200 cases. Arch Neurol Psychiatry 1950;63:964-76.
8Kleihues P, Burger PC, Scheithauer BW. The new WHO classification of brain tumours. Brain Pathol 1993;3:255-68.
9Cairncross JG, Ueki K, Zlatescu MC, Lisle DK, Finkelstein DM, Hammond RR, et al. Specific genetic predictors of chemotherapeutic response and survival in patients with anaplastic oligodendrogliomas. J Natl Cancer Inst 1998;90:1473-9.
10Cairncross G, Wang M, Shaw E, Jenkins R, Brachman D, Buckner J, et al. Phase III trial of chemoradiotherapy for anaplastic oligodendroglioma: Long-term results of RTOG 9402. J Clin Oncol 2013;31:337-43.
11van den Bent MJ, Brandes AA, Taphoorn MJ, Kros JM, Kouwenhoven MC, Delattre JY, et al. Adjuvant procarbazine, lomustine, and vincristine chemotherapy in newly diagnosed anaplastic oligodendroglioma: Long-term follow-up of EORTC brain tumor group study 26951. J Clin Oncol 2013;31:344-50.
12Ballester LY, Huse JT, Tang G, Fuller GN. Molecular classification of adult diffuse gliomas: Conflicting IDH1/IDH2, ATRX, and 1p/19q results. Hum Pathol 2017;69:15-22.
13Wood MD, Halfpenny AM, Moore SR. Applications of molecular neuro-oncology - a review of diffuse glioma integrated diagnosis and emerging molecular entities. Diagn Pathol 2019;14:29.
14Dunbar E, Yachnis AT. Glioma diagnosis: Immunohistochemistry and beyond. Adv Anat Pathol 2010;17:187-201.
15Reuss DE, Kratz A, Sahm F, Capper D, Schrimpf D, Koelsche C, et al. Adult IDH wild type astrocytomas biologically and clinically resolve into other tumor entities. Acta Neuropathol 2015;130:407-17.
16Aibaidula A, Chan AK, Shi Z, Li Y, Zhang R, Yang R, et al. Adult IDH wild-type lower-grade gliomas should be further stratified. Neuro Oncol 2017;19:1327-37.
17Hartmann C, Hentschel B, Wick W, Capper D, Felsberg J, Simon M, et al. Patients with IDH1 wild type anaplastic astrocytomas exhibit worse prognosis than IDH1-mutated glioblastomas, and IDH1 mutation status accounts for the unfavorable prognostic effect of higher age: Implications for classification of gliomas. Acta Neuropathol 2010;120:707-18.
18Brat DJ, Aldape K, Colman H, Holland EC, Louis DN, Jenkins RB, et al. CIMPACT-NOW update 3: Recommended diagnostic criteria for “Diffuse astrocytic glioma, IDH-wildtype, with molecular features of glioblastoma, WHO grade IV”. Acta Neuropathol 2018;136:805-10.
19Louis DN, Giannini C, Capper D, Paulus W, Figarella-Branger D, Lopes MB, et al. cIMPACT-NOW update 2: Diagnostic clarifications for diffuse midline glioma, H3 K27M-mutant and diffuse astrocytoma/anaplastic astrocytoma, IDH-mutant. Acta Neuropathol 2018;135:639-42.
20Chiang JC, Ellison DW. Molecular pathology of paediatric central nervous system tumours. J Pathol 2017;241:159-72.
21Pagès M, Beccaria K, Boddaert N, Saffroy R, Besnard A, Castel D, et al. Co-occurrence of histone H3 K27M and BRAF V600E mutations in paediatric midline grade I ganglioglioma. Brain Pathol 2018;28:103-11.
22Hegi ME, Diserens AC, Gorlia T, Hamou MF, de Tribolet N, Weller M, et al. MGMT gene silencing and benefit from temozolomide in glioblastoma. N Engl J Med 2005;352:997-1003.
23Louis DN, Wesseling P, Paulus W, Giannini C, Batchelor TT, Cairncross JG, et al. cIMPACT-NOW update 1: Not Otherwise specified (NOS) and not elsewhere classified (NEC). Acta Neuropathol (Berl) 2018;135:481-4.
24Daumas-Duport C, Scheithauer B, O'Fallon JA, Kelly P. Grading of astrocytomas. A simple and reproducible method. Cancer 1988;62:2152-65.
25Unruh D, Schwarze SR, Khoury L, Thomas C, Wu M, Chen L, et al. Mutant IDH1 and thrombosis in gliomas. Acta Neuropathol 2016;132:917-30.
26Ellison DW, Hawkins C, Jones DTW, Onar-Thomas A, Pfister SM, Reifenberger G, et al. cIMPACT-NOW update 4: Diffuse gliomas characterized by MYB, MYBL1, or FGFR1 alterations or BRAFV600E mutation. Acta Neuropathol 2019;137:683-7.
27Phillips HS, Kharbanda S, Chen R, Forrest WF, Soriano RH, Wu TD, et al. Molecular subclasses of high-grade glioma predict prognosis, delineate a pattern of disease progression, and resemble stages in neurogenesis. Cancer Cell 2006;9:157-73.
28Verhaak RG, Hoadley KA, Purdom E, Wang V, Qi Y, Wilkerson MD, et al. Integrated genomic analysis identifies clinically relevant subtypes of glioblastoma characterized by abnormalities in PDGFRA, IDH1, EGFR, and NF1. Cancer Cell 2010;17:98-110.
29Sturm D, Witt H, Hovestadt V, Khuong-Quang DA, Jones DT, Konermann C, et al. Hotspot mutations in H3F3A and IDH1 define distinct epigenetic and biological subgroups of glioblastoma. Cancer Cell 2012;22: 425-37.
30Reis RM, Lebleblicioglu KD, Lopes JM, Kleihues P, Ohgaki H. Genetic profile of gliosarcomas. Am J Pathol 2000;156:425-32.
31Erson-Omay EZ, Çağlayan AO, Schultz N, Weinhold N, Omay SB, Özduman K, et al. Somatic POLE mutations cause an ultramutated giant cell high-grade glioma subtype with better prognosis. Neuro Oncol 2015;17:1356-64.
32Alexandrescu S, Korshunov A, Lai SH, Dabiri S, Patil S, Li R, et al. Epithelioid glioblastomas and anaplastic epithelioid pleomorphic xanthoastrocytomas—same entity or first cousins? Brain Pathol 2016;26:215-23.
33Matsumura N, Nakajima N, Yamazaki T, Nagano T, Kagoshima K, Nobusawa S, et al. Concurrent TERT promoter and BRAF V600E mutation in epithelioid glioblastoma and concomitant low-grade astrocytoma. Neuropathology 2017;37:58-63.
34Chen R, Cohn SM, Cohen AL, Colman H. Glioma subclassifications and their clinical significance. Neurotherapeutics 2017;14:284-97.
35Malkki H. Trial Watch: Glioblastoma vaccine therapy disappointment in Phase III trial. Nat Rev Neurol 2016;12:190.
36Darvin P, Toor SM, Sasidharan Nair V, Elkord E. Immune checkpoint inhibitors: Recent progress and potential biomarkers. Exp Mol Med 2018;50:1-1.
37Erson-Omay EZ, Çağlayan AO, Schultz N, Weinhold N, Omay SB, Özduman K, et al. Somatic POLE mutations cause an ultramutated giant cell high-grade glioma subtype with better prognosis. Neuro-Oncol 2015;17:1356-64.
38Alexandrescu S, Korshunov A, Lai SH, Dabiri S, Patil S, Li R, et al. Epithelioid Glioblastomas and Anaplastic Epithelioid Pleomorphic Xanthoastrocytomas--Same Entity or First Cousins? Brain Pathol Zurich Switz 2016;26:215-23.
39Matsumura N, Nakajima N, Yamazaki T, Nagano T, Kagoshima K, Nobusawa S, et al. Concurrent TERT promoter and BRAF V600E mutation in epithelioid glioblastoma and concomitant low-grade astrocytoma. Neuropathol Off J Jpn Soc Neuropathol 2017;37:58-63.
40Chen R, Smith-Cohn M, Cohen AL, Colman H. Glioma Subclassifications and Their Clinical Significance. Neurother J Am Soc Exp Neurother 2017;14:284-97.
41Malkki H. Trial Watch: Glioblastoma vaccine therapy disappointment in Phase III trial. Nat Rev Neurol 2016;12:190.
42Darvin P, Toor SM, Sasidharan Nair V, Elkord E. Immune checkpoint inhibitors: recent progress and potential biomarkers. Exp Mol Med 2018;50:1-11.