Printer Friendly

The Updated World Health Organization Glioma Classification: Cellular and Molecular Origins of Adult Infiltrating Gliomas.

Diffusely infiltrating gliomas (IGs) are a diverse set of primary neoplasms of the central nervous system (CNS) characterized by tumor cells that perniciously invade the surrounding CNS parenchyma from which they arise. In this article we will review the classification of adult IGs in light of the recently updated World Health Organization (WHO) (1,2) text on CNS tumors. We will then review how this classification can be viewed in the context of selected studies concerning gliomagenesis.

Relative to cancers arising in other organs, one of the unique features of IGs is that neoplastic precursor lesions have not been described. That is, by the time they declare themselves either to the patient, to the treating physician, or at the pathologist's microscope, IGs are already universally expected to progress. Precursor dysplastic lesions or in situ neoplastic entities that have not yet become invasive, that may be cured by surgical excision, and that do not inherently represent a fatal threat to the patient, are foreign to the spectrum of IGs.

It is estimated that the prevalence of occult gliomas (ie, those that have not yet come to clinical attention and are not yet radiologically detectable) is approximately 0.03%. (3) In 1 study that screened 1000 asymptomatic individuals with magnetic resonance imaging of the brain, however, an IG was detected in 1 patient (subsequently biopsy proven), and a second was suggested by imaging characteristics, for a prevalence rate of 0.1% to 0.2% for silent lesions, or those that are clinically silent but radiologically detectable. (4) If this reported rate of silent lesions were closer to the de facto rate, one might expect the prevalence of occult lesions to be even higher. Certainly, the existence and prevalence of a set of theoretical "premalignant" glial lesions is not known.

Infiltrating gliomas, when discussed from a histologic and historical frame of reference, consist of 2 broad classes of tumors, originally designated based on their morphologic characteristics alone that recall 2 nonneoplastic glial cell types: oligodendrogliomas and infiltrating astrocytomas. The latter group excludes noninfiltrating astrocytomas, such as pilocytic astrocytoma, subependymal giant cell astrocytoma, and pleomorphic xanthoastrocytoma.

In contrast to other tumors of the CNS, IGs demonstrate a particular propensity to infiltrate along preexisting structures, including white matter tracts, perivascular spaces, and subventricular and subpial compartments, conforming to and exploiting a terrain that is ostensibly familiar to the neoplastic cell type. (5) This infiltrative feature is readily highlighted by immunohistochemical staining using an antibody against a mutated protein characteristic of most lower-grade IGs--isocitrate dehydrogenase 1 (IDH1), to be discussed in greater detail below. This staining demonstrates individual neoplastic cells in close proximity to nonneoplastic cells, such as mature cortical neurons, that appear completely unperturbed by the presence of a nearby tumor clone (Figure 1).

In the following discussion we will highlight those IGs typically arising in the adult and will exclude the category of diffuse midline gliomas that harbor distinct entity-defining alterations from those considered here.

The Updated Classification of IGs

Recent molecular advances have determined that canonic histologically defined categories of IGs correlate well with relatively few recurrent molecular features. (6,7) Indeed, these alterations have become entity-defining features, acquiring gold standard status and trumping more conventional histologic assessment. This has the effect of simplifying our categorization of IGs and will serve to frame our subsequent discussion on the potential origins of these tumors.

Infiltrating gliomas of all histologic grades can be subdivided broadly into 3 groups dependent on only 2 parameters: the deletional status of chromosomal arms 1p and 19q, and the mutational status of IDH1, or its mitochondrial cousin, IDH2. This renders 3 basic groups of IGs (Figure 2): (1) astrocytoma, IDH-mutant; (2) astrocytoma, IDH-wild type; and (3) oligodendroglioma, IDH-mutant and 1p/19q-codeleted.

The fourth possible category, that of 1p/19q-codeleted tumors lacking IDH mutation, represents an exceedingly rare entity, with 1 study of 1087 IGs showing 0% of cases with whole-arm 1p/19q codeletion in the absence of IDH mutation (approximately 0.3% of cases showed partial deletions in 1p/19q without IDH mutation). (7) Indeed, it is no longer recommended by the WHO to render a diagnosis of oligodendroglioma in the setting of definitive negative results for IDH mutation and/or 1p/19q codeletion, even when morphologic characteristics consistent with oligodendroglioma are present. However, the WHO permits tumors designated as astrocytoma or oligodendroglioma to be tagged with the "not otherwise specified" modifier. The purpose of the "not otherwise specified" modifier is not to demarcate a unique entity as such, but to act as a flag to subsequent clinicians or researchers that the case either requires additional molecular workup or that the obtained molecular results are unusual in some way. (2)

Each of the 3 core groups of IGs is further associated with recurrent somatic alterations in other loci. For example, astrocytomas harboring IDH mutations are also frequently mutated for ATRX and TP53, whereas oligodendrogliomas frequently show mutations in FUBP1 or CIC. (6,7) Astrocytomas wild type for IDH, including IDH-wild-type glioblastoma (GBM), frequently harbor amplifications in EGFR, loss of genetic material on chromosome 10, and deletions of CDKN2A/B. (8) In line with these findings and as a practical consideration for the surgical pathologist, immunohistochemical staining for ATRX, in addition to IDH1 and p53, has become routine in the assessment of IGs at many institutions, and its utility has been well documented. (9-13) For example, a tumor with strong staining for IDH1 and p53 and negative staining for ATRX (indicating the presence of a mutation in all 3 cases) would be diagnostic of an astrocytoma, even in the presence of areas with oligodendroglioma-like morphology. Similarly, a tumor that is positive for IDH1 mutation, that is positive for ATRX (indicating intact expression), and that demonstrates a wild-type pattern of p53 expression, should prompt further assessment for 1p/19q deletional status in the presence of any oligodendroglial morphologic characteristics (or potentially even in their absence) to support or deny the designation of oligodendroglioma. It should be mentioned that IDH testing should be conducted via sequencing, particularly for patients younger than 50 to 60 years, in the absence of immunohistochemical evidence for IDH mutation, because the common commercially available antibody recognizes only IDH p.R132H. Indeed, of IDH1/IDH2 mutated IGs, only about 90% harbor the p.R132H mutation, with the next most common alterations comprising p.R132C (eg, Figure 3). (14,15) Finally, the presence of an ATRX mutation is inversely correlated with mutation in the TERT promoter, an alteration characteristic of both oligodendrogliomas and IDH-wild-type infiltrating astrocytomas, but not of IDH-mutant astrocytomas. (6,7,16) This inverse correlation presumably reflects the mutual exclusivity of redundant mechanisms in achieving telomere maintenance, a rapidly advancing area of study in gliomagenesis that has also generated recognition of novel germ line glioma susceptibility loci, to be addressed later. (17,18) The TERT promoter locus is increasingly being interrogated on next-generation sequencing panels for solid tumors, even if it is not yet widely assessed during the standard initial workup of IGs. (7)

Glioblastoma

Glioblastoma assigned a WHO grade of IV is the most common astrocytic IG and retains a dismal prognosis of roughly 10 to 15 months despite decades of research, improvement of surgical techniques, and advancements in clinical care. (19) In the adult setting, the most recent WHO guidelines (2) explicitly distinguish between GBMs that are IDH-mutant and those that are not. The former arise often in clinically evident fashion as a progression from a lower-grade glioma and are traditionally termed "secondary glioblastoma." In contrast, those GBMs that are IDH-wild-type are classically termed "de novo glioblastoma" or "primary glioblastoma" (Figure 2). Primary GBMs develop quickly, have no clinically evident lower-grade precursor, and are in fact molecularly distinct from secondary GBM.

The current gold standard for a diagnosis of GBM in the context of astrocytic IGs and irrespective of IDH status remains the histologic hallmarks of microvascular proliferation and/or necrosis. That said, it is evident that the set of GBMs so defined comprises a markedly heterogeneous group of tumors. Early findings of The Cancer Genome Atlas Research Network revealed recurrently altered core pathways involving receptor tyrosine kinases (RTKs), p53, and RB, (20) and also demonstrated that cluster analysis algorithms based on DNA microarray data yielded robust molecular subtypes of adult GBM. These subtypes were designated as "proneural," "neural," "mesenchymal," and "classical" in one seminal paper. (21) Subsequent work on a set of GBMs from all age groups and incorporating methylation data demonstrated similar clusters, designated as "IDH," "mesenchymal," "RTK-I (PDGFRA)," and "RTK-II (Classic)," in addition to 2 clusters, "K27" and "G34," enriched in pediatric and young adult patients and corresponding to tumors with mutations in histone genes such as H3F3A. (22) What is clear is that GBM can first be separated into 2 main categories based on IDH status. For those GBMs that are wild type for IDH, a set of molecular alterations exists that is generally foreign to the spectrum of IDH-mutant IGs, including EGFR amplification in the "classical" or "RTK-II" subvarieties.

Within the IDH-wild-type set of astrocytomas including GBM, methylation data have also revealed a group distinct from those enriched for tumors previously described as "classical" and "mesenchymal." This new group, termed "LGm6," is enriched for histologic low-grade gliomas but also contains a subset of tumors with GBM-defining histologic criteria. (23) The authors termed the LGm6 tumors with GBM histology "LGm6-GBM," whereas those with lower-grade IG histology were termed "PA-like" for pilocytic astrocytoma-like." Indeed, tumors within the LGm6 group showed a higher frequency of alterations in genes associated with pilocytic astrocytoma (recall, a noninfiltrative astrocytoma), including BRAF, NF1, and NTRAK1/2, and a lower frequency of TERT mutations compared with classical" and mesenchymal" tumors. (23) Further refinements in extrapolating survival ratios between molecular subgroups of GBM are underway, including the combinatorial assessment of particular genetic and epigenetic alterations; for example, TERT and MGMT methylation status in combination yield refined survival curves among IDH-wild-type GBM. (24)

It should be noted that the current aim of the WHO is to molecularly classify the vast majority of IGs while acknowledging that rare cases may fall outside of the predominant groups. (1) For example, the entity "oligoastrocytoma," notorious for its high rate of interobserver diagnostic discordance even among expert neuropathologists, has been virtually eliminated from the updated classification. (25,26) As a second example, the distinct entity of glioblastoma with oligodendroglioma component" found within the 2007 edition of the WHO (27) has also fallen by the wayside. Both decisions were driven by evidence that the vast majority of these histologically defined entities do not represent molecularly distinct tumors, as currently defined, but rather a collection of tumors that are better designated on the basis of 1p/19q and IDH mutational status. (6,28)

As Sahm and von Deimling (29) and Sahm et al (30) noted in support of their early proposal to eliminate the diagnosis oligoastrocytoma," the use of that designation varies widely among different institutions (purportedly 0%-80% of IGs diagnosed at any particular institution at the time of their report), and they suggested that this reflects practice style more than clinical relevance. Importantly, as the authors note, what treating clinicians typically are interested in is whether the tumor demonstrates 1p/19q codeletion and/or IDH mutation, not the subjective interpretation of glial morphology. In particular, close to 0% of lower-grade IGs overall have mixed molecular features with respect to IDH mutation, 1p/19q codeletion, and TP53 mutation, and these molecular features are superior to histopathology in predicting patient outcome. (6) Nevertheless, there are well-documented, albeit rare, cases of apparently truly hybrid tumors. (31,32)

As a final point on the subject of tumors that are histopathologically ambiguous with respect to emerging molecular criteria, it should be recognized that, in the absence of an IDH mutation, the distinction between IDH-wild-type infiltrating astrocytomas and noninfiltrating astrocytomas is of paramount importance. Although the presence of additional molecular features may help resolve some histologically ambiguous or poorly sampled cases (eg, EGFR amplification in GBM or BRAF-KIAA1549 fusion in pilocytic astrocytoma), in other cases there may exist no detected molecular correlate to the light microscopic findings. In these cases, correlation with all available clinical parameters (radiologic features, patient course, etc) is necessary.

The Interface Between Classification and Biology

What is the relationship between these refined diagnostic criteria and what we know about how these tumors arise? Do each of the 3 broad IG categories correspond to distinct cells of origin? If not, what are the underlying biologic mechanisms between tumor types that have pervaded through to the classification scheme, either meriting the labeling of distinct tumor designations or, conversely, requiring histologically diverse tumors to be lumped together into single diagnostic categories? Finally, is it possible that there is no precise relationship between tumor cell of origin and even the most updated classification? That is, just as diverse cells of origin could theoretically give rise to tumors that ultimately share histologic features, so too could a single cell of origin give rise to a diversity of tumors, both histologically and molecularly. An intriguing possibility could be that non-cell autonomous factors acting upon a putative cell of origin could dictate many of the features that we now consider diagnostic of the tumor entity. In this discussion it is important to remember that the primary purpose of a successful diagnostic schema is not necessarily to reflect the biology of the tumor preceding the point of diagnosis per se, but rather to make an effort to predict outcome and offer guidance for treatment going forward. Although the latter on face value seems to be of more practical concern, the clinical relevance of cell of origin nevertheless runs deep. In particular, understanding the cell of origin can lead to insights into the earliest mechanisms of cell-type susceptibility to particular oncogenic lesions, which in turn has the potential to drive further research into cancer prevention and early detection. Moreover, with greater knowledge about a cell of origin, there is increased opportunity to exploit cell-type-specific molecular targets that will inform tumor-specific therapeutic strategies.

Interestingly, a somewhat paradoxic consensus is emerging. At the same time that we are uncovering the molecular heterogeneity within and between tumors, we are also moving toward a paradigm that implies neoplastic purity with respect to astrocytic and oligodendroglial lineages. A testament to this concept is the virtual elimination in the current WHO of "oligoastrocytoma," as discussed earlier. Highlighting the complexity of the issue of tumor purity with respect to glial lineage, however, Venteicher et al (33) recently used single-cell RNA sequencing for thousands of cells from molecularly defined oligodendrogliomas and astrocytomas to reveal a cellular hierarchy in which subclones of tumor cells demonstrate transcriptional profiles associated with more than one distinct glial component, irrespective of their ultimate designation as oligodendroglioma or astrocytoma. Other studies have examined the intratumoral heterogeneity of GBM at cellular resolution, revealing that the transcriptional profile of one tumor cell may cluster differently from that of an adjacent cell. That is, when clustering algorithms developed using tumor homogenates are applied to single cells, each cell may align with different groups in varying proportions (eg, proneural, classical, and mesenchymal). (34) Other studies have demonstrated differences in the molecular classification of distinct tumor foci that correlate with radiologic parameters, such as contrast enhancement. (35) This is not to say that the transcriptional profiles obtained at the population level of resected tumor cells are of no potential value clinically or scientifically, but such findings do highlight the need to think about the idea of heterogeneity and subclones within a tumor, and the relative extent to which each subclone contributes to clinically meaningful and treatable parameters.

Inextricably linked to the question of subclonality and tumor heterogeneity between and within IG patients is the notion of cell of origin and early events in gliomagenesis. As a starting point we will examine candidate cells of origin and the early role of IDH mutation, perhaps the single most important molecular alteration that has impacted the new classification scheme.

ORIGINS OF GLIOMA

One of the most fascinating and persistent questions in glioma biology relates to the cellular origin of IGs. Although the cell of origin can be challenging to define across cancers (see Cedric Blanpain's review of this topic), (36) the diversity of IGs, the lack of recognized premalignant lesions, and the absence of screening methodologies yielding tissue specimens (eg, that are analogous to endoscopic procedures of the gastrointestinal tract) have made such discoveries even more difficult in the human brain. Nevertheless, progress has been made in defining the relationship between neurodevelopmental biology and neuroepithelial tumors, perhaps most remarkably to date in the context of medulloblastoma, (37-41) as well as the relationship between neuroepithelial progenitor cells, differentiated cells, and neoplasia in the adult human brain.

Candidate Precursor Cells

Whether IGs result from dedifferentiation of mature glia or whether they are derived from neuroepithelial stem cells or progenitor cells--and if so, at precisely what stage of differentiation--has been a subject of great interest and study for decades. Once thought to be entirely dormant, it is now well established that the adult human brain harbors several populations of neural stem cells and progenitor cells that might serve as precursors to neoplastic lesions. These populations include cells residing in the subventricular zone of the lateral ventricles, the dentate nucleus of the hippocampus, and the subcortical white matter. (42-45) Presenting an added level of complexity to the origin of gliomas is the notion that a cell acquiring early mutational events may be distinct from the cell that would best be described as a particular tumors "cell of origin." (46) That is, a progenitor cell with acquired mutations may continue to differentiate in the usual fashion along a particular pathway--say, toward an oligodendroglial progenitor cell (OPC)--and only then exhibit an appropriate susceptibility to additional environmental cues and genetic lesions that would steer the cell toward definitive neoplasia. An elegant experimental demonstration of this concept used mosaic analysis with double markers (47) to show that when a small set of neural stem cells are manipulated to inactivate p53 and Nf1, it is the progeny of these cells, and in particular OPCs, that go on to generate tumors. (48) It is possible, of course, that the cell of mutation and cell of origin are distinct for different classes of IGs, and in particular for those that are IDH-mutated and those that are not. We will first examine in more detail the functional importance of IDH mutation, followed by the notion of cell susceptibility to IDH mutation.

IDH Mutation as an Early Event

A major advance in our understanding of IGs was afforded by the discovery of recurrent somatic mutations in IDH1, initially found in a subset of GBMs noted to be enriched in younger patients and those with secondary GBMs. (8) This soon led to the characterization of IDH mutations in most lower-grade IGs, (49-51) a finding that was confirmed in more recent comprehensive molecular analyses of lower-grade IGs. (6,7) IDH1 is ubiquitously expressed in all cells as a basic metabolic enzyme that is part of the citric acid cycle, catalyzing the conversion between isocitrate and a-ketoglutarate (a-KG). Mutations of codon 132 from arginine to, most commonly, histidine alter the catalytic properties of the enzyme such that a-KG is converted into a new product, 2-hydroxyglutarate. (52,53) This latter metabolite is thought to promote oncogenesis, in part by altering the activity of proteins that use a-KG as a cofactor, including histone demethylases, with the potential to broadly alter the epigenomic landscape. (54-57) Other potential downstream effects of 2-hydroxyglutarate include alteration of HIF-1[alpha], alteration of extracellular matrix metabolism, and downregulation of DNA damage pathways involving ATM. (58) Because [NADP.sup.+] is normally converted to NADPH during conversion of isocitrate to a-KG, it is thought that an imbalance of the [NADP.sup.+]:NADPH ratio as a result of mutated IDH may also contribute to an oncogenic metabolic milieu. (54)

Mutations in IDH1 have been shown to be sufficient for inducing genome-wide changes in methylation patterns, including the CpG island methylator phenotype seen in a subset of gliomas (G-CIMP) that correlates with diverse transcriptional changes as a result of promoter methylation alterations. (59) Recent work has demonstrated that among IDH-mutant astrocytomas, methylation profile clustering can further subdivide these tumors into G-CIMP-low and G-CIMP-high, with a significant difference in survival and with evidence that G-CIMP-high tumors may in fact progress to those that are G-CIMP-low. (23)

Cell Susceptibility to IDH Mutation

There is evidence to suggest that IDH mutation is a very early event in the genesis of gliomas that will ultimately harbor these mutations and precedes subsequent alterations in either TP53 or 1p/19q. (6,49) Intriguingly, it also has been suggested that for IDH and TP53 comutated tumors, even if both genes are mutated nearly concurrently, IDH mutant protein is likely to become available to the cell earlier on as a consequence of whether or not the initial mutational events occur on the template or coding strand. (60) In particular, the authors show evidence that IDH-mutated tumors are enriched for TP53 mutations whose initial mutation occurs on the coding strand. The consequence of this is that although IDH p.R132H protein is immediately available to the mutated cell because of a mutation on the template strand, p53-mutated protein would not be available until the cell has undergone at least 1 cycle of replication. Depending on the cell cycle duration of the putative cell of mutation, there could be substantial time for the consequences of IDH mutation to take effect even before mutated p53 contributes to tumor progression. (60)

Assuming for a moment that IDH is indeed the first critical genetic alteration to occur in the genesis of IDH-mutated tumors, an interesting question is why heterozygous IDH mutations confer survival advantage to specific cell types such that they give rise to stereotyped tumors. Given the ubiquity of IDH in cells as a basic metabolic enzyme, the cell-type specificity of IDH alterations in cancer may be related to the metabolites to which a particular cell type is exposed. As highlighted in prior reports, glutamate is particularly abundant within the brain, (61) and large pools of this amino acid would be necessary to maintain a pool of a-KG that could then go on to be converted into 2-hydroxyglutarate. (52) Indeed, it is thought that loss of heterozygosity of the IDH1 locus in the mutated setting is rarely seen in cancer because the wild-type allele is additionally necessary to maintain this pool of a-KG. (54)

Interestingly, IDH mutations are also seen in several neoplastic lesions apart from glioma. These include acute myelogenous leukemia, (62) angioimmunoblastic T-cell lymphoma, and hematopoietic precursor lesions, such as myelodysplastic syndrome. IDH mutations are also relatively common in enchondromas, chondrosarcoma, intrahepatic cholangiocarcinoma, and sinonasal undifferentiated carcinoma. (63,64) Why IDH mutations should arise in this set of neoplasms is unclear. Similar to glioma cells, myeloid cells are also rich in glutamine availability, perhaps a prerequisite to exploiting IDH alterations. (52) Ultimately, the link between cell type and IDH mutations remains to be elucidated, but initial insights are being sought using mouse models of IDH alteration.

Mouse Models of IDH Mutation

A conditional knock-in murine model of IDH alteration has been employed to generate enchondromas when driven by the Col2a1 promoter (63) and has been shown to generate hyperproliferative marrow along with epigenetic alterations seen in myeloid neoplasms when driven by the LysM promoter. In an initial study of gliomagenesis, (65) the same mouse employed to drive expression of mutant IDH via the Nestin promoter exhibited a perinatal lethal phenotype characterized by cerebral hemorrhage. In the same study, expression of mutant IDH driven by the Gfap promoter yielded some mice surviving to adulthood; however, they did not show evidence of glioma formation. Although one potential conclusion would be that IDH mutations are insufficient to form gliomas, a major limitation of the study was that mice were not permitted to achieve later stages of development prior to near ubiquitous expression of mutant IDH in relatively large pools of neural stem cells within the embryo. To circumvent this problem, more recent work by Bardella et al (66) used a similar paradigm but with tamoxifen-inducible Nestin-Cre-mediated induction of IDH mutation performed at 5 to 6 weeks of age, targeting Cre to neural progenitor cells within the subventricular zone of the adult animal (as well as the subgranular zone of the dentate gyrus). These experiments demonstrated that IDH mutation alone increases the number of proliferating cells in the subventricular zone and rostral migratory stream, that these cells infiltrate" into adjacent areas, and that subventricular nodules of proliferating cells are generated. Although these findings are interesting and recapitulate some of the features of IG, the question remains as to whether the presence of IDH alteration alone in Nestin-positive cells is ultimately sufficient to initiate a subsequent cascade of genetic lesions that lead to frank malignancy.

Insights From Additional Mouse Models of Glioma

In addressing the question of which of the diverse cell types in the mammalian brain could be specifically targeted to induce glioma formation prior to the discovery of IDH, a major advance within the field came with the generation of mouse models wherein oncogenes and tumor suppressors were selectively manipulated in specific cell types. A model in which expression of a receptor (tv-a) for the replication-competent ALV splice-acceptor viral vectors (RCAS) is placed under the direction of a cell-type-specific promoter enabled this kind of cell-type specificity. (67) When virus carrying a payload of oncogenic material is introduced to the animal via the RCAS vector, the cell type of interest alone, as defined by expression of a relatively cell-type-specific gene whose promoter is used to drive tv-a, will receive the material and drive tumor growth. This system has been successfully applied using a number of promoters driving tv-a expression and a variety of oncogenes delivered (either directly or via Cre-mediated inactivation of a floxed tumor suppressor) by the viral vector. The original demonstration of this involved, again, Nestin-expressing neural progenitor cells driving expression of tv-a and receiving vectors carrying Kras in combination with Akt alterations. (67)

Studies have also demonstrated the tumorigenic capacity of cells in which the Gfap promoter is used to drive oncogenic expression. (68-73) In one such study, a mouse model in which transgenic Gfap promoter-driven expression of PDGFB was superimposed upon a Tp53-null background not only generated GBM-like lesions in mice, but also intriguingly demonstrated hyperplastic astroglial lesions in both subependymal and subpial locations, possibly representing precursor lesions. (72) It is interesting to note that in autopsy specimens, subependymal glial nodules (apart from those seen in patients with tuberous sclerosis) are sometimes attributed to evidence of remote ventriculitis or other injury to the ependymal lining. (74) Whether or not a subset of these lesions may also represent early precursor lesions to glioma is unknown. Finally, it should be noted that the lack of specificity of Gfap expression itself relative to mature astrocytes and their committed precursors should influence the interpretation of Gfap promoter-driven tumor models. For example, GFAP is known to be expressed in radial glia, subventricular zone precursor cells (neural stem cells) that may give rise to neurons, tanycytes, and Schwann cells. (75,76) Also of note, some studies have demonstrated an imperfect correlation between Gfap promoter-driven transgene expression and endogenous expression of GFAP protein. (77) Moreover, GFAP expression can be demonstrated in a wide variety of tumors, including astrocytomas, oligodendrogliomas, ependymomas, and CNS embryonal tumors. (27)

Oligodendroglial Progenitor Cells

Because the vast majority of IGs arise in adults, with many of these showing radiologically identifiable epicenters within the subcortical white matter, (78-80) several groups have sought to create models of glioma formation in which tumorigenic molecular lesions are induced within the subcortical anatomic compartment. In one such elegant study, building on prior work demonstrating successful PDGF-driven glioma induction by targeting white matter progenitor cells in rats, (81) Lei et al (82) developed a mouse model of GBM recapitulating the proneural (21) molecular subtype of human GBM. Here, mice were injected with a retroviral vector carrying PDGF, Cre recombinase, and a reporter. Using the properties of the retrovirus along with stereotactic anatomic precision of the injection site (into the rostral subcortical white matter), the investigators were able to selectively infect the endogenous proliferative pool of OPCs in the adult mammal, showing that these cells give rise to tumors. This procedure, performed on mice with either a background floxed Pten allele or a combination of floxed Pten and floxed Tp53 alleles, generated GBMs with 100% penetrance, albeit with different intervals to progression in each model. In particular, the mice harboring only a floxed Pten allele demonstrated relatively slow tumor onset. (82) In a subsequent study using this model, this feature of slow progression was exploited to study early time points of disease and revealed recurrent genetic alterations that occur as a tumor progresses in a particular genotype/ phenotype context (83). Finally, this same tumor model has also been used in a more recent study that intriguingly demonstrates a potentially central role of the gene Olig2 in maintaining the proneural phenotype in OPC-induced gliomas, such that abrogation of Olig2 causes a phenotypic shift toward an EGFR-driven astrocytic lesion. (84) These data highlight the central role of cell-of-origin-derived transcriptional networks in regulating and/or maintaining a tumor's characteristic molecular trajectory throughout the life of the tumor.

In other work targeting OPCs as a glioma cell of origin, Galvao et al (85) used tamoxifen injections to conditionally delete both Tp53 and Nf1 in OPCs expressing creER driven by the Ng2 promoter. These investigators were careful to rule out active Cre recombination in neural stem cells within the subventricular zone, and were able to obtain tumor formation with 100% penetrance. (85) Moreover, they detected an interesting temporal dynamic of tumor formation relative to OPC proliferative rates following injection. In particular, after an initial reactivated, hyperproliferative phase following the acute genetic lesional event, a dormant phase similar to the quiescent state seen in normal animals was seen in OPCs, albeit with coincident likely repression of differentiation programs. After several months, this dormant phase was followed by a rapid blooming of the de facto malignant tumor. One conclusion from this temporal dynamic is that there may indeed be an indolent preneoplastic phase triggered by an initial genetic event insufficient for complete neoplastic transformation. This would lead to a potentially detectable and targetable phase of disease, prior to the onset of frank malignancy. The authors demonstrate, in their murine model, that mTOR inhibition is one component to an approach that might abrogate neoplastic progression at this early phase in disease, an interesting prospect in the event that such a phase could be detected in a clinically relevant way in humans. (85)

Certainly, the data supporting a crucial role for OPCs in tumorigenesis continue to accumulate. The question of what leads to the proposed heterogeneity of that set of IGs derived from OPCs, setting aside for the moment other possible cells of origin, is in itself complex. Ostensibly the most likely candidate tumor to originate from OPCs, the oligodendroglioma, could be thought of as a fairly stereotyped tumor, with IDH1 mutations and 1p/19q codeletions dominating its defining molecular hallmarks. Even within this set, however, recent data have divided oligodendroglioma into 3 groups with distinct molecular profiles that correlate with demographic features and tumor aggression. (86) Of these 3 groups, 1 group is more OPC-like, whereas the others exhibit more differentiated transcriptional profiles. The OPC-like subgroup demonstrates a proclivity toward MYC amplifications as the tumor progresses, in addition to other features that are seen less often in the other subgroups. What emerges, then, is a continually nesting pattern of subclassification that, because it is recurrent across individuals, likely reflects stereotyped cell-of-origin-derived factors (including the relative degree of differentiation) in combination with stereotyped microenvironmental interactions.

Another series of elegant studies has further tackled the relationship between distinct cells of origin and a diversity of molecular phenotypes in resultant gliomas using conditional tamoxifen-induced constructs driven by promoters specific for distinct stages of neural stem cell and progenitor differentiation. Alcantara Llaguno et al (87) demonstrated induction of high-grade astrocytomas by crossing tamoxifen-inducible Nestin-creER lines with floxable tumor suppressor alleles, including Pten, Nf1, and Tp53. Moreover, even in the absence of a Nestin driver, the group was able to induce tumor formation by stereotactically targeting the subventricular zone with a viral vector carrying Cre. In more recent work, Alcantara Llaguno et al (88) used the Ascl1 promoter and the Cspg4 (encoding NG2) promoter as drivers for cre-ER to further specifically target oncogenic alterations to both committed neural progenitor cells and adult OPCs, and adult OPCs only, respectively. Not only were gliomas produced in both models, but their important findings reveal that clinically and molecularly distinct tumor types were predictable based on the targeted cell of origin. That is, among 2 tumor types described, type 2 tumors were only produced by OPCs in the NG2 model, whereas type 1 and type 2 tumors were produced by the Ascl1 model, despite the fact that the induced oncogenic alterations were identical in both cell types.

Having discussed a range of possible neuroepithelial cells of origin for IGs and several examples of pertinent mouse models, we will now turn our attention to other factors affecting our current knowledge about early gliomagenesis.

Glioma Stem Cells

The possible presence of a minority population of tumor cells with stemlike properties that shows resistance to treatment and provides the substrate for tumor recurrence and progression has been explored for several decades, with some of the initial descriptions reported in acute myeloid leukemia and breast cancer. (89-91) Since that time, many groups have studied the presence of such a population within gliomas, and in particular GBM, with some of the initial characterization of these glioma stem cells" (GSC) reported by Ignatova et al (92) and Singh et al. (93,94) In 2004, Singh et al (94) described a subpopulation of tumor cells identified by expression of CD133 that was capable of regenerating tumors using very low numbers of cells in serially transplanted xenografts, in contrast with larger numbers of [CD133.sup.-] cells that failed to induce tumor growth. Whether or not the notion of discrete tumor cell populations comprising either stemlike cells or more differentiated cells represents an accurate model of the in vivo reality is debatable. Single-cell sequencing data from human GBM specimens reveal cells with transcriptional profiles that correlate with stemness" as a continuous gradient. (34) The authors suggest that such a gradient implies that, in vivo, human tumors may display a range of differentiation phenotypes and that the process of cell culture may select for the extremes of this range. (34) As reviewed elsewhere, including by Rahman et al (95) and Lathia et al, (96) there are several challenges associated with the study of putative GSCs that make their potential clinical relevance unclear. One critical issue is how to define GSCs. Borrowing from the concept of somatic noncancerous stem cells, the definition has been dependent primarily on functional criteria. In particular, cancer stem cells should show the capacity to proliferate, self-renew, and differentiate into multiple lineages. In the context of cancer, importantly, they should also be able to generate a tumor that recapitulates the original, particularly via serial transplantations demonstrating in vivo tumor growth. Of course, the process of expeditiously evaluating any subpopulation of cells to assess for any or all of these criteria demands significant ex vivo manipulation, which may call into question a particular result's relevance to the actual biologic system in question.

In practice, most studies have searched for protein-level markers such that cell subpopulations enriched for these markers might then meet the functional criteria required for the label of GSC (eg, CD133 or CD15). (94,97) A potential problem with this approach is that it may rely on the relative stability of a particular marker in a particular cell type despite the fact that the system being interrogated is ripe with opportunities for cellular plasticity, not the least of which potentially includes active differentiation and dedifferentiation of the cells in question both in the tumor environment and in vitro. Suffice it to say here that the precise relationship between a putative glioma cell of origin, putative GSCs, the cells comprising the bulk of the tumor mass either at initial presentation or at recurrence, and the clinical relevance of any of these cell types as measured under experimental conditions continues to be an area of active study.

Selected Environmental Factors

The role of inherited genetic factors, environmental factors extrinsic to the organism, and microenvironmental players, such as neurons and other cell types in glioma formation and propagation, are also critical issues to address and may provide insight into early events in neoplastic transformation.

Epidemiologic studies have thus far failed to identify many risk factors for glioma with the exception of genetic predisposition, ionizing radiation, and a protective effect of allergy and atopy. (98) The relative rarity of any individual neoplastic entity within the brain in part accounts for the challenge in identifying etiologically relevant environmental agents. Although much attention has been focused on the use of cellular telephones, the extent to which these devices might contribute to gliomagenesis remains controversial. (99) Environmental factors also have been implicated in tumor progression via immunomodulatory effects. For example, mice housed in an enriched environment" show an increase in interleukin 15 and natural killer-cell-mediated immune response and a concomitant increase in survival when receiving transplants of several distinct glioma cell lines. (100) The extent to which physical and sensory stimuli might affect the tumor microenvironment holds implications for holistic approaches to cancer care.

Several groups have reported that cytomegalovirus (CMV) may play a role in gliomagenesis, (101-104) and the use of antiviral agents, such as valganciclovir, has been investigated in the treatment of glioma. (105) However, the role of this virus is controversial, and studies using next-generation sequencing to detect viral DNA or RNA in glioma samples have failed to produce an association between glioma and CMV. (106) A consensus meeting in 2012 determined that although CMV antigenic or genetic material may be present within and may modulate the biology of glioma, whether or not CMV plays any role in initiating tumor formation had not been established. (102) Some recent studies have continued to address this question of tumor modulation, in particular with regard to the effect of CMV-derived determinants on the maintenance or progression of GSCs. (104,107)

Recent evidence intriguingly points to the role of neuronal activity in promoting glial oncogenesis. (108,109) Using optogenetic neurostimulation, neuroligin-3 was identified as a secreted synaptic protein that drives the PI3K-mTOR pathway in xenografted high-grade glioma. (109) Moreover, expression of this protein correlates negatively with survival over the Cancer Genome Atlas Research Network GBM cohort. Whether genetic and neurodevelopmental susceptibility to imbalances in excitatory and inhibitory activity in the brain correlate with IG formation deserves further study. Indeed, for those tumors arising in the setting of epilepsy, or those that present with new-onset seizures, it would be interesting to determine the precise cause-and-effect relationship of tumor formation, excitatory overdrive, and tumor progression.

The co-occurrence of glioma with other neuropathologies that feature astrogliosis leads to the question of whether reactive glial conditions may lead to neoplastic transformation. To date there is little evidence for this. In a series of 8 patients with multiple sclerosis who also developed gliomas, including 7 IGs (4 of which were GBM), molecular characterization of the tumors was similar to previously reported common alterations in glioma, and those that were GBM exhibited features of de novo, IDH-wild-type GBM, suggesting to the authors that they did not arise from transformation of a low-grade process. (110) Cases of GBM occurring at the site of a prior traumatic brain injury have raised the hypothesis that a reactive astroglial proliferation in conjunction with inflammatory processes may provide the soil for neoplastic transformation, but epidemiologic data are lacking. (111,112)

Inherited Factors

Genome-wide association studies have identified several genetic loci associated with increased risk of glioma, including TERT, CDKN2A/2B, RTEL1, PHLDB1, EGFR, TP53, TERC, DDX6, and PIWI-interacting RNAs. (113-115) Of these genes, TERT, CDKN2A/2B, EGFR, and TP53 are well-characterized recurrently altered genes in sporadic IGs. In addition to well-known inherited tumor syndromes, such as Li Fraumeni, the study of familial cases of IG remains an important avenue toward identifying genetic candidates that drive glioma formation. For example, a study of families with multiple cases of oligodendroglioma has revealed that POT1, a gene associated with telomere maintenance, is a recurrently altered gene in at least 3 families. (116) This raises interesting questions about the relationship between POT1, IDH1, and/or 1p/19q codeletion as coconspirators in an oligodendroglioma-specific precursor cell. Interestingly, POT1 has also been implicated in a subset of familial melanoma cases, (117) a tumor that has been shown in 1 study to harbor IDH mutations in approximately 10% of cases. (118) Further study of the possible relationship between POT1, TERT, and IDH1 -induced alterations in all 3 at the epigenetic and transcriptional levels, including their possible functional redundancy or mutual exclusivity, would be of potential interest. An ongoing, carefully designed case control study, the Glioma International Case-Control Study, will incorporate systematic collection of peripheral blood and will hopefully yield additional insight into both germ line genetic as well as environmental risk factors contributing to glioma development. (98)

The Influence of Chromatin Topology

Why is it that cancers arising from different primary sites have recurrent molecular alterations? Why is it that certain cell types and the tumors arising from them are able to exploit mutations in IDH1, whereas others are not? Why is it that the 1p/19q unbalanced translocation is so specific to oligodendroglioma, and presumably to oligodendroglial progenitor cells? Why is it that intergenic fusions and certain patterns of somatic copy number alterations are stereotypical of particular tumors? Core insights into oncogenesis likely lie at the interface of (1) tumor cell of origin and the nonneoplastic biology associated with these cells of origin, including chromatin configuration and transcriptional regulatory networks; and (2) cell-type-specific susceptibility and/or vulnerability to particular genetic lesions. Increasingly, the 3-dimensional architecture of chromatin is being explored in the context of transcriptional regulation. Chromosome conformation capture (119) and techniques derived from it have been used in the context of both developmental biology and cancer to explore intrachromosomal and interchromosomal interactions, and in defining the concept of topologic-associated domains (120,121) and other higher-order chromatin architectural features that mediate transcriptional regulation. (122-128) In particular, the concept that the loss of topologic-associated domain integrity by early oncogenic events leads to secondary tumor-promoting alterations has become a provocative area of research, as discussed by Valton and Dekker. (129)

Studies have shown that early mutational events in cancer, such as transcription factor activation in the setting of translocation (eg, TMPRSS-ERG in prostate cancer), correlates with and likely drives chromatin topologic reorganization in such a way that predisposes the cell to subsequent oncogenic alterations. (130) A particular cell's chromatin topology and nonneoplastic transcription factor network may be similarly determining which particular initiating oncogenic genetic alterations are much more likely to occur, and to have functional significance once they do occur, in a particular cell type. These alterations would then become characteristic of a tumor type relative to its cell of origin.

With respect to IGs, for example, mechanistic insight into the relationship between chromatin topology and hypermethylation in IDH-mutant glioma, and how these features interact to drive further oncogenic alterations are shown in intriguing work by Flavahan et al. (131) This study shows that in IDH-mutant glioma, chromatin topology is disrupted in such a way that genes ordinarily insulated from enhancers in distinct regulatory domains are brought into close contact. As an example, the promoter for PDGFRA, an oncogene frequently upregulated in IDH-mutant gliomas in the absence of gene amplification, is shown in chromatin conformation capture experiments to closely and aberrantly associate with the enhancer for FIP1L1, a gene with constitutive expression in oligodendroglial progenitors. This interaction is permitted by hypermethylation of binding sites for CTCF, a protein with a role in maintaining boundaries between chromatin regulatory domains, such that CTCF fails to bind and oncogenes are exposed to distant, strong enhancers. (131)

CONCLUSIONS

With multidimensional molecular data sets spanning increasingly larger sets of patients with IGs, our understanding of the disease at the point of surgical resection has increased exponentially during the last 5 to 10 years. This has revealed a previously unappreciated molecular heterogeneity both across and within tumors, as well as contexts in which the morphologic diversity seen within a given IG in fact belies what is molecularly proven to be a more uniformly clonal process. Animal models have demonstrated the ability to generate tumors from a diversity of cell types and states of differentiation, including neural stem cells, neural progenitor cells, and OPCs. The presence of GSCs, their relationship to neural stem and progenitor cells, and their in vivo role in tumor formation versus tumor propagation are all active areas of continuing research. The fundamental question of why stereotyped initiating oncogenic events occur in some cell types but not others, both within the context of IGs as well as across cancers, remains a holy grail of sorts that would shed light on both basic developmental processes as well as oncogenesis.

As reflected in the updated WHO classification, adult IGs can now increasingly be considered as a dichotomy of IDH-mutated and IDH-wild-type tumors, each with its own set of diverse subclasses and each subclass with recurrent, entity-defining molecular features. Precise mapping of these entities to the available set of potential cells of origin is still in its relative infancy. As mouse models continue to develop that are increasingly able to conditionally target small sets of specific cell types and trace their resultant progeny, it is exciting that some of the key questions in this issue have begun to be answered. Through the careful study of chromatin topology and transcriptional networks that are ultimately crucially linked to innate differences in early developmental programs as well as neuroglial progenitor programs found in the adult, the stage is set to study and finally understand the environmental and/or microenvironmental agents that work with these factors in tandem with inherited genetic susceptibilities to spark the ignition of early glioma formation (Figure 4).

References

(1.) Louis DN, Perry A, Burger P, et al. International Society Of Neuropathology-Haarlem consensus guidelines for nervous system tumor classification and grading. Brain Pathol. 2014;24(5):429-435.

(2.) Louis DN, 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):803-820.

(3.) Lima GL, Zanello M, Mandonnet E, Taillandier L, Pallud J, Duffau H. Incidental diffuse low-grade gliomas: from early detection to preventive neurooncological surgery. Neurosurg Rev. 2016;39(3):377-384.

(4.) Katzman GL, Dagher AP, Patronas NJ. Incidental findings on brain magnetic resonance imaging from 1000 asymptomatic volunteers. JAMA. 1999; 282(1):36-39.

(5.) Cuddapah VA, Robel S, Watkins S, Sontheimer H. A neurocentric perspective on glioma invasion. Nat Rev Neurosci. 2014;15(7):455-465.

(6.) Brat DJ, Verhaak RG, Aldape KD, et al. Comprehensive, integrative genomic analysis of diffuse lower-grade gliomas. N Engl j Med. 2015;372(26): 2481-2498.

(7.) Eckel-Passow JE, Lachance DH, Molinaro AM, et al. Glioma groups based on 1p/19q, IDH, and TERT promoter mutations in tumors. N Engl j Med. 2015; 372(26):2499-2508.

(8.) Parsons DW, Jones S, Zhang X, et al. An integrated genomic analysis of human glioblastoma multiforme. Science. 2008;321(5897):1807-1812.

(9.) Cai J, Zhu P, Zhang C, et al. Detection of ATRX and IDH1-R132H immunohistochemistry in the progression of 211 paired gliomas. Oncotarget. 2016;7(13):16384-16395.

(10.) Ebrahimi A, Skardelly M, Bonzheim I, et al. ATRX immunostaining predicts IDH and H3F3A status in gliomas. Acta Neuropathol Commun. 2016; 4(1):60.

(11.) Ikemura M, ShibaharaJ, Mukasa A, et al. Utility of ATRX immunohistochemistry in diagnosis of adult diffuse gliomas. Histopathology. 2016;69(2):260-267.

(12.) Takano S, Ishikawa E, Sakamoto N, et al. Immunohistochemistry on IDH 1/2, ATRX, p53 and Ki-67 substitute molecular genetic testing and predict patient prognosis in grade III adult diffuse gliomas. Brain Tumor Pathol. 2016;33(2):107-116.

(13.) Leeper HE, Caron AA, Decker PA, Jenkins RB, Lachance DH, Giannini C. IDH mutation, 1p19q codeletion and ATRX loss in WHO grade II gliomas. Oncotarget. 2015;6(30):30295-30305.

(14.) Hartmann C, Meyer J, Balss J, et al. Type and frequency of IDH1 and IDH2 mutations are related to astrocytic and oligodendroglial differentiation and age: a study of 1,010 diffuse gliomas. Acta Neuropathol. 2009;118(4):469-474.

(15.) Chen L, Voronovich Z, Clark K, et al. Predicting the likelihood of an isocitrate dehydrogenase 1 or 2 mutation in diagnoses of infiltrative glioma. Neuro Oncol. 2014;16(11):1478-1483.

(16.) Nonoguchi N, Ohta T, Oh JE, Kim YH, Kleihues P, Ohgaki H. TERT promoter mutations in primary and secondary glioblastomas. Acta Neuropathol. 2013;126(6):931-937.

(17.) Walsh KM, Codd V, Smirnov IV, et al. Variants near TERT and TERC influencing telomere length are associated with high-grade glioma risk. Nat Genet. 2014;46(7):731-735.

(18.) Walsh KM, WienckeJK, Lachance DH, et al. Telomere maintenance and the etiology of adult glioma. Neuro Oncol. 2015;17(11):1445-1452.

(19.) Bondy ML, Scheurer ME, Malmer B, et al. Brain tumor epidemiology: consensus from the Brain Tumor Epidemiology Consortium. Cancer. 2008;113(7 suppl):1953-1968.

(20.) Comprehensive genomic characterization defines human glioblastoma genes and core pathways. Nature. 2008;455(7216):1061-1068.

(21.) Verhaak RG, Hoadley KA, Purdom E, et al. Integrated genomic analysis identifies clinically relevant subtypes of glioblastoma characterized by abnormalities in PDGFRA, IDH1, EGFR, and NF1. Cancer Cell. 2010;17(1):98-110.

(22.) Sturm D, Witt H, Hovestadt V, et al. Hotspot mutations in H3F3A and IDH1 define distinct epigenetic and biological subgroups of glioblastoma. Cancer Cell. 2012;22(4):425-437.

(23.) Ceccarelli M, Barthel FP, Malta TM, et al. Molecular profiling reveals biologically discrete subsets and pathways of progression in diffuse glioma. Cell. 2016;164(3):550-563.

(24.) Arita H, Yamasaki K, Matsushita Y, et al. A combination of TERT promoter mutation and MGMT methylation status predicts clinically relevant subgroups of newly diagnosed glioblastomas. Acta Neuropathol Commun. 2016;4(1):79.

(25.) Castillo MS, Davis FG, Surawicz T, et al. Consistency of primary brain tumor diagnoses and codes in cancer surveillance systems. Neuroepidemiology. 2004;23(1-2):85-93.

(26.) Aldape K, Simmons ML, Davis RL, et al. Discrepancies in diagnoses of neuroepithelial neoplasms: the San Francisco Bay Area Adult Glioma Study. Cancer. 2000;88(10):2342-2349.

(27.) Louis DN, Oghaki H, Wiestler OD, Cavenee WK, eds. WHO Classification of Tumours of the Central Nervous System. Lyon, France: IARC; 2007.

(28.) Hinrichs BH, Newman S, Appin CL, et al. Farewell to GBM-O: genomic and transcriptomic profiling of glioblastoma with oligodendroglioma component reveals distinct molecular subgroups. Acta Neuropathol Commun. 2016;4:4.

(29.) Sahm F, von Deimling A. Farewell to oligoastrocytoma: response to letters. Acta Neuropathol. 2015;129(1):155.

(30.) Sahm F, Reuss D, Koelsche C, et al. Farewell to oligoastrocytoma: in situ molecular genetics favor classification as either oligodendroglioma or astrocytoma. Acta Neuropathol. 2014;128(4):551-559.

(31.) Wilcox P, Li CC, Lee M, et al. Oligoastrocytomas: throwing the baby out with the bathwater? Acta Neuropathol. 2015;129(1):147-149.

(32.) Huse JT, Diamond EL, Wang L, Rosenblum MK. Mixed glioma with molecular features of composite oligodendroglioma and astrocytoma: a true "oligoastrocytoma"? Acta Neuropathol. 2015;129(1):151-153.

(33.) Venteicher AS, Tirosh I, Hebert C, et al. 142 Genetic and nongenetic determinants of cellular architecture in IDH1-mutant oligodendrogliomas and astrocytomas using single-cell transcriptome analysis. Neurosurgery. 2016; 63(suppl 1):158.

(34.) Patel AP, Tirosh I, Trombetta JJ, et al. Single-cell RNA-seq highlights intratumoral heterogeneity in primary glioblastoma. Science. 2014;344(6190): 1396-1401.

(35.) Gill BJ, Pisapia DJ, Malone HR, et al. MRI-localized biopsies reveal subtype-specific differences in molecular and cellular composition at the margins of glioblastoma. Proc Natl Acad Sci U S A. 2014;111(34):12550-12555.

(36.) Blanpain C. Tracing the cellular origin of cancer. Nat Cell Biol. 2013; 15(2):126-134.

(37.) Lin CY, Erkek S, Tong Y, et al. Active medulloblastoma enhancers reveal subgroup-specific cellular origins. Nature. 2016;530(7588):57-62.

(38.) Manoranjan B, Venugopal C, McFarlane N, et al. Medulloblastoma stem cells: where development and cancer cross pathways. Pediatr Res. 2012;71(4, pt 2):516-522.

(39.) Hovestadt V, Jones DT, Picelli S, et al. Decoding the regulatory landscape of medulloblastoma using DNA methylation sequencing. Nature. 2014; 510(7506):537-541.

(40.) Jones DT, Jager N, Kool M, et al. Dissecting the genomic complexity underlying medulloblastoma. Nature. 2012;488(7409):100-105.

(41.) Pugh TJ, Weeraratne SD, Archer TC, et al. Medulloblastoma exome sequencing uncovers subtype-specific somatic mutations. Nature. 2012; 488(7409):106-110.

(42.) Eriksson PS, Perfilieva E, Bjork-Eriksson T, et al. Neurogenesis in the adult human hippocampus. Nat Med. 1998;4(11):1313-1317.

(43.) Sanai N, Tramontin AD, Quinones-Hinojosa A, et al. Unique astrocyte ribbon in adult human brain contains neural stem cells but lacks chain migration. Nature. 2004;427(6976):740-744.

(44.) Nunes MC, Roy NS, Keyoung HM, et al. Identification and isolation of multipotential neural progenitor cells from the subcortical white matter of the adult human brain. Nat Med. 2003;9(4):439-447.

(45.) Roy NS, Wang S, Harrison-Restelli C, et al. Identification, isolation, and promoter-defined separation of mitotic oligodendrocyte progenitor cells from the adult human subcortical white matter. j Neurosci. 1999;19(22):9986-9995.

(46.) Visvader JE. Cells of origin in cancer. Nature. 2011;469(7330):314-322.

(47.) Zong H, Espinosa JS, Su HH, Muzumdar MD, Luo L. Mosaic analysis with double markers in mice. Cell. 2005;121(3):479-492.

(48.) Liu C, Sage JC, Miller MR, et al. Mosaic analysis with double markers reveals tumor cell of origin in glioma. Cell. 2011;146(2):209-221.

(49.) Watanabe T, Nobusawa S, Kleihues P, Ohgaki H. IDH1 mutations are early events in the development of astrocytomas and oligodendrogliomas. Am J Pathol. 2009;174(4):1149-1153.

(50.) Yan H, Parsons DW, Jin G, et al. IDH1 and IDH2 mutations in gliomas. N Engl J Med. 2009;360(8):765-773.

(51.) Balss J, Meyer J, Mueller W, Korshunov A, Hartmann C, von Deimling A. Analysis of the IDH1 codon 132 mutation in brain tumors. Acta Neuropathol. 2008;116(6):597-602.

(52.) Dang L, White DW, Gross S, et al. Cancer-associated IDH1 mutations produce 2-hydroxyglutarate. Nature. 2009;462(7274):739-744.

(53.) Ward PS, Patel J, Wise DR, et al. The common feature of leukemia-associated IDH1 and IDH2 mutations is a neomorphic enzyme activity converting alpha-ketoglutarate to 2-hydroxyglutarate. Cancer Cell. 2010;17(3): 225-234.

(54.) Clark O, Yen K, Mellinghoff IK. Molecular pathways: isocitrate dehydrogenase mutations in cancer. Clin Cancer Res. 2016;22(8):1837-1842.

(55.) Losman JA, Looper RE, Koivunen P, et al. (R)-2-hydroxyglutarate is sufficientto promote leukemogenesis and its effects are reversible. Science. 2013; 339(6127):1621-1625.

(56.) Lu C, Ward PS, Kapoor GS, et al. IDH mutation impairs histone demethylation and results in a block to cell differentiation. Nature. 2012; 483(7390):474-478.

(57.) Xu W, Yang H, Liu Y, et al. Oncometabolite 2-hydroxyglutarate is a competitive inhibitor of alpha-ketoglutarate-dependent dioxygenases. Cancer Cell. 2011;19(1):17-30.

(58.) Inoue S, Li WY, Tseng A, et al. Mutant IDH1 downregulates ATM and alters DNA repair and sensitivity to DNA damage independent of TET2. Cancer Cell. 2016;30(2):337-348.

(59.) Turcan S, Rohle D, Goenka A, et al. IDH1 mutation is sufficient to establish the glioma hypermethylator phenotype. Nature. 2012;483(7390):479-483.

(60.) Lai A, Kharbanda S, Pope WB, et al. Evidence for sequenced molecular evolution of IDH1 mutant glioblastoma from a distinct cell of origin. J Clin Oncol. 2011;29(34):4482-4490.

(61.) Danbolt NC. Glutamate uptake. Prog Neurobiol. 2001;65(1):1-105.

(62.) Mardis ER, Ding L, Dooling DJ, et al. Recurring mutations found by sequencing an acute myeloid leukemia genome. N Engl J Med. 2009;361(11):1058-1066.

(63.) Hirata M, Sasaki M, Cairns RA, et al. Mutant IDH is sufficientto initiate enchondromatosis in mice. Proc Natl Acad Sci USA. 2015;112(9):2829-2834.

(64.) Dogan S, Chute DJ, Xu B, et al. Frequent IDH2 R172 mutations in undifferentiated and poorly-differentiated sinonasal carcinomas [published online ahead of print May 11, 2017]. J Pathol. doi:10.1002/path.4915.

(65.) Sasaki M, Knobbe CB, Itsumi M, et al. D-2-hydroxyglutarate produced by mutant IDH1 perturbs collagen maturation and basement membrane function. Genes Dev. 2012;26(18):2038-2049.

(66.) Bardella C, Al-Dalahmah O, Krell D, et al. Expression of Idh1R132H in the murine subventricular zone stem cell niche recapitulates features of early gliomagenesis. Cancer Cell. 2016;30(4):578-594.

(67.) Holland EC, Celestino J, Dai C, Schaefer L, Sawaya RE, Fuller GN. Combined activation of Ras and Akt in neural progenitors induces glioblastoma formation in mice. Nat Genet. 2000;25(1):55-57.

(68.) Dai C, Celestino JC, Okada Y, Louis DN, Fuller GN, Holland EC. PDGF autocrine stimulation dedifferentiates cultured astrocytes and induces oligodendrogliomas and oligoastrocytomas from neural progenitors and astrocytes in vivo. Genes Dev. 2001;15(15):1913-1925.

(69.) Ding H, Roncari L, Shannon P, et al. Astrocyte-specific expression of activated p21-ras results in malignant astrocytoma formation in a transgenic mouse model of human gliomas. Cancer Res. 2001;61(9):3826-3836.

(70.) Xiao A, Wu H, Pandolfi PP, Louis DN, Van Dyke T. Astrocyte inactivation of the pRb pathway predisposes mice to malignant astrocytoma development that is accelerated by PTEN mutation. Cancer Cell. 2002;1(2):157-168.

(71.) Holland EC, Li Y, Celestino J, et al. Astrocytes give rise to oligodendrogliomas and astrocytomas after genetransfer of polyoma virus middle Tantigen in vivo. Am I Pathol. 2000;157(3):1031-1037.

(72.) Hede SM, Hansson I, Afink GB, et al. GFAP promoter driven transgenic expression of PDGFB in the mouse brain leads to glioblastoma in a Trp53 null background. Glia. 2009;57(11):1143-1153.

(73.) Abel TW, Clark C, Bierie B, et al. GFAP-Cre-mediated activation of oncogenic K-ras results in expansion of the subventricular zone and infiltrating glioma. Mol Cancer Res. 2009;7(5):645-653.

(74.) Ellison D, LoveS, Chimelli LMC, et al. Neuropathology: A Reference Text of CNS Pathology, 3rd ed. Edinburgh, UK: Mosby; 2013.

(75.) Doetsch F, Caille I, Lim DA, Garcia-Verdugo JM, Alvarez-Buylla A. Subventricular zone astrocytes are neural stem cells in the adult mammalian brain. Cell. 1999;97(6):703-716.

(76.) Chojnacki AK, Mak GK, Weiss S. Identity crisis for adult periventricular neural stem cells: subventricular zone astrocytes, ependymal cells or both? Nat Rev Neurosci. 2009;10(2):153-163.

(77.) Su M, Hu H, Lee Y, d'Azzo A, Messing A, Brenner M. Expression specificity of GFAP transgenes. Neurochem Res. 2004;29(11):2075-2093.

(78.) Larjavaara S, Mantyla R, Salminen T, et al. Incidence of gliomas by anatomic location. Neuro Oncol. 2007;9(3):319-325.

(79.) Duffau H, Capelle L. Preferential brain locations of low-grade gliomas. Cancer. 2004;100(12):2622-2626.

(80.) Persson AI, Petritsch C, Swartling FJ, et al. Non-stem cell origin for oligodendroglioma. Cancer Cell. 2010;18(6):669-682.

(81.) Assanah M, Lochhead R, Ogden A, Bruce J, Goldman J, Canoll P. Glial progenitors in adult white matter are driven to form malignant gliomas by platelet-derived growth factor-expressing retroviruses. J Neurosci. 2006;26(25): 6781-6790.

(82.) Lei L, Sonabend AM, Guarnieri P, et al. Glioblastoma models reveal the connection between adult glial progenitors and the proneural phenotype. PLoS One. 2011;6(5):e20041.

(83.) Sonabend AM, Bansal M, Guarnieri P, et al. The transcriptional regulatory network of proneural glioma determines the genetic alterations selected during tumor progression. Cancer Res. 2014;74(5):1440-1451.

(84.) Lu F, Chen Y, Zhao C, et al. Olig2-dependent reciprocal shift in PDGF and EGF receptor signaling regulates tumor phenotype and mitotic growth in malignant glioma. Cancer Cell. 2016;29(5):669-683.

(85.) Galvao RP, Kasina A, McNeill RS, et al. Transformation of quiescent adult oligodendrocyte precursor cells into malignant glioma through a multistep reactivation process. Proc Natl Acad Sci USA. 2014;111(40):E4214-E4223.

(86.) Kamoun A, Idbaih A, Dehais C, et al. Integrated multi-omics analysis of oligodendroglial tumours identifies three subgroups of 1p/19q co-deleted gliomas. Nat Commun. 2016;7:11263.

(87.) Alcantara Llaguno S, Chen J, Kwon CH, et al. Malignant astrocytomas originate from neural stem/progenitor cells in a somatic tumor suppressor mouse model. Cancer Cell. 2009;15(1):45-56.

(88.) Alcantara Llaguno SR, Wang Z, Sun D, et al. Adult lineage-restricted CNS progenitors specify distinct glioblastoma subtypes. Cancer Cell. 2015;28(4):429-440.

(89.) Bonnet D, Dick JE. Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell. Nat Med. 1997; 3(7):730-737.

(90.) Lapidot T, Sirard C, Vormoor J, et al. A cell initiating human acute myeloid leukaemia after transplantation into SCID mice. Nature. 1994; 367(6464):645-648.

(91.) Al-Hajj M, Wicha MS, Benito-Hernandez A, Morrison SJ, Clarke MF. Prospective identification of tumorigenic breast cancer cells. Proc Natl Acad Sci USA. 2003;100(7):3983-3988.

(92.) Ignatova TN, Kukekov VG, Laywell ED, Suslov ON, Vrionis FD, Steindler DA. Human cortical glial tumors contain neural stem-like cells expressing astroglial and neuronal markers in vitro. Glia. 2002;39(3):193-206.

(93.) Singh SK, Clarke ID, Terasaki M, et al. Identification of a cancer stem cell in human brain tumors. Cancer Res. 2003;63(18):5821-5828.

(94.) Singh SK, Hawkins C, Clarke ID, et al. Identification of human brain tumour initiating cells. Nature. 2004;432(7015):396-401.

(95.) Rahman M, Deleyrolle L, Vedam-Mai V, Azari H, Abd-El-Barr M, Reynolds BA. The cancer stem cell hypothesis: failures and pitfalls. Neurosurgery. 2011;68(2):531-545; discussion 545.

(96.) LathiaJD, Mack SC, Mulkearns-Hubert EE, Valentim CL, Rich JN. Cancer stem cells in glioblastoma. Genes Dev. 2015;29(12):1203-1217.

(97.) Son MJ, Woolard K, Nam DH, Lee J, Fine HA. SSEA-1 is an enrichment marker for tumor-initiating cells in human glioblastoma. Cell Stem Cell. 2009;4(5):440-452.

(98.) Amirian ES, Armstrong GN, Zhou R, et al. The Glioma International case-control study: a report from the Genetic Epidemiology of Glioma International Consortium. Am J Epidemiol. 2016;183(2):85-91.

(99.) Morgan LL, Miller AB, Sasco A, Davis DL. Mobile phone radiation causes brain tumors and should be classified as a probable human carcinogen (2A) [review]. Int I Oncol. 2015;46(5):1865-1871.

(100.) Garofalo S, D'Alessandro G, Chece G, et al. Enriched environment reduces glioma growth through immune and non-immune mechanisms in mice. Nat Commun. 2015;6:6623.

(101.) Cobbs CS, Harkins L, Samanta M, et al. Human cytomegalovirus infection and expression in human malignant glioma. Cancer Res. 2002;62(12):3347-3350.

(102.) Dziurzynski K, Chang SM, Heimberger AB, et al. Consensus on the role of human cytomegalovirus in glioblastoma. Neuro Oncol. 2012;14(3):246-255.

(103.) Fiallos E, Judkins J, Matlaf L, et al. Human cytomegalovirus gene expression in long-term infected glioma stem cells. PLoS One. 2014;9(12): e116178.

(104.) Soroceanu L, Matlaf L, Khan S, et al. Cytomegalovirus immediate-early proteins promote stemness properties in glioblastoma. Cancer Res. 2015;75(15): 3065-3076.

(105.) Soderberg-Naucler C, Rahbar A, Stragliotto G. Survival in patients with glioblastoma receiving valganciclovir. N Engl J Med. 2013;369(10):985-986.

(106.) Strong MJ, Blanchard Et, Lin Z, et al. A comprehensive next generation sequencing-based virome assessment in brain tissue suggests no major virus-tumor association. Acta Neuropathol Commun. 2016;4(1):71.

(107.) Ulasov IV, Kaverina NV, Ghosh D, et al. CMV70-3P miRNA contributes to the CMV mediated glioma stemness and represents a target for glioma experimental therapy. Oncotarget. 2017;8(16):25989-25999.

(108.) Gibson EM, Purger D, Mount CW, et al. Neuronal activity promotes oligodendrogenesis and adaptive myelination in the mammalian brain. Science. 2014;344(6183):1252304.

(109.) Venkatesh HS, Johung TB, Caretti V, et al. Neuronal activity promotes glioma growth through neuroligin-3 secretion. Cell. 2015;161(4):803-816.

(110.) Khalil A, Serracino H, Damek DM, Ney D, Lillehei KO, Kleinschmidt-DeMasters BK. Genetic characterization of gliomas arising in patients with multiple sclerosis. J Neurooncol. 2012;109(2):261-272.

(111.) Tyagi V, Theobald J, Barger J, et al. Traumatic brain injury and subsequent glioblastoma development: review of the literature and case reports. Surg Neurol Int. 2016;7:78.

(112.) Zhou B, Liu W. Post-traumatic glioma: report of one case and review of the literature. Int J Med Sci. 2010;7(5):248-250.

(113.) Shete S, Hosking FJ, Robertson LB, et al. Genome-wide association study identifies five susceptibility loci for glioma. Nat Genet. 2009;41(8):899-904.

(114.) Jacobs DI, Qin Q, Lerro MC, et al. PIWI-interacting RNAs in gliomagenesis: evidence from post-GWAS and functional analyses. Cancer Epidemiol Biomarkers Prev. 2016;25(7):1073-1080.

(115.) Baskin R, Woods NT, Mendoza-Fandino G, Forsyth P, Egan KM, Monteiro AN. Functional analysis of the 11q23.3 glioma susceptibility locus implicates PHLDB1 and DDX6 in glioma susceptibility. Sci Rep. 2015;5:17367.

(116.) Bainbridge MN, Armstrong GN, Gramatges MM, et al. Germline mutations in shelterin complex genes are associated with familial glioma. J Natl Cancer Inst. 2014;107(1):384.

(117.) Robles-Espinoza CD, Harland M, Ramsay AJ, et al. POT1 loss-of-function variants predispose to familial melanoma. Nat Genet. 2014;46(5):478-481.

(118.) Shibata T, Kokubu A, Miyamoto M, Sasajima Y, Yamazaki N. Mutant IDH1 confers an in vivo growth in a melanoma cell line with BRAF mutation. Am J Pathol. 2011;178(3):1395-1402.

(119.) Dekker J, Rippe K, Dekker M, Kleckner N. Capturing chromosome conformation. Science. 2002;295(5558):1306-1311.

(120.) Gibcus JH, Dekker J. The hierarchy of the 3D genome. Mol Cell. 2013; 49(5):773-782.

(121.) Dixon JR, Selvaraj S, Yue F, et al. Topological domains in mammalian genomes identified by analysis of chromatin interactions. Nature. 2012; 485(7398):376-380.

(122.) Lomvardas S, Barnea G, Pisapia DJ, Mendelsohn M, Kirkland J, Axel R. Interchromosomal interactions and olfactory receptor choice. Cell. 2006;126(2): 403-413.

(123.) Markenscoff-Papadimitriou E, Allen WE, Colquitt BM, et al. Enhancer interaction networks as a means for singular olfactory receptor expression. Cell. 2014;159(3):543-557.

(124.) Barutcu AR, Hong D, Lajoie BR, et al. RUNX1 contributes to higher-order chromatin organization and gene regulation in breast cancer cells. Biochim BiophysActa. 2016;1859(11):1389-1397.

(125.) Barutcu AR, Lajoie BR, Fritz AJ, et al. SMARCA4 regulates gene expression and higher-order chromatin structure in proliferating mammary epithelial cells. Genome Res. 2016;26(9):1188-1201.

(126.) Ji X, Dadon DB, Powell BE, et al. 3D chromosome regulatory landscape of human pluripotent cells. Cell Stem Cell. 2016;18(2):262-275.

(127.) Katainen R, Dave K, Pitkanen E, et al. CTCF/cohesin-binding sites are frequently mutated in cancer. Nat Genet. 2015;47(7):818-821.

(128.) Babaei S, Akhtar W, de Jong J, Reinders M, de Ridder J. 3D hotspots of recurrent retroviral insertions reveal long-range interactions with cancer genes. Nat Commun. 2015;6:6381.

(129.) Valton AL, Dekker J. TAD disruption as oncogenic driver. Curr Opin Genet Dev. 2016;36:34-40.

(130.) Rickman DS, Soong TD, Moss B, et al. Oncogene-mediated alterations in chromatin conformation. Proc Natl Acad Sci USA. 2012;109(23):9083-9088.

(131.) Flavahan WA, Drier Y, Liau BB, et al. Insulator dysfunction and oncogene activation in IDH mutant gliomas. Nature. 2016;529(7584):110-114.

David J. Pisapia, MD

Accepted for publication June 1 5, 2017.

From the Department of Pathology and Laboratory Medicine, Weill Cornell Medical College, New York, New York.

The author has no relevant financial interest in the products or companies described in this article.

Reprints: David J. Pisapia, MD, Department of Pathology and Laboratory Medicine, Weill Cornell Medical College, 1300 York Avenue, New York, NY 10065 (email: djp2002@med.cornell.edu).

Caption: Figure 1. Isocitrate Dehydrogenase (IDH) R132H immunohistochemical staining demonstrates infiltrative pattern of growth. A, Representative hematoxylin-eosin staining of an infiltrating astrocytoma. B, Immunohistochemical staining using an antibody recognizing the p.R132H mutated form of IDH1 demonstrates individual neoplastic cells (white arrowheads) infiltrating neocortex. Adjacent mature neocortical neurons appear unperturbed (black arrowheads) (original magnification X400).

Caption: Figure 2. World Health Organization classification of infiltrating gliomas. The diagram depicts the 3 major categories of adult diffusely infiltrating gliomas based on Isocitrate Dehydrogenase 1/2 (IDH1/2) mutational status and 1p/19q codeletional status, along with representative additional recurrent somatically altered genes for each group, and histologic images stained with hematoxylineosin. Histologic grade would be superimposed on these general categories. The category of "astrocytoma, IDH-wildtype" includes glioblastoma, IDH-wild type, also known as primary glioblastoma. The category of "astrocytoma, IDH-mutant" includes glioblastoma, IDH-mutant, also known as secondary glioblastoma. The schema does not allow for a designation of oligoastrocytoma for those entities that are molecularly well defined. Abbreviations: ATRX, alpha-thalassemia/mental retardation syndrome X-linked; CDKN2A, cyclin-dependent kinase inhibitor 2A; CIC, capicua transcriptional repressor; EGFR, epidermal growth factor receptor; FUBP1, far upstream element-binding protein 1; GBM, glioblastoma; PTEN, phosphatase and tensin homolog (original magnification X200).

Caption: Figure 3. Immunohistochemical workup of infiltrating glioma. A, In an exemplary case, hematoxylin-eosin staining reveals a diffusely infiltrating glioma with predominantly astrocytic morphology. B, Isocitrate Dehydrogenase 1 (IDH1) p.R132H immunostaining was negative. The tumor was positive for IDH1 p.R132G by sequencing (data not shown). C, p53 staining was strongly positive in a subset of cells. D, Alpha-thalassemia/mental retardation syndrome X-linked (ATRX) staining was preserved in nonneoplastic cells, and loss of expression was observed in neoplastic cells. 1p/19q analysis (data not shown) by fluorescent in situ hybridization was negative for codeletion. The diagnosis rendered for this tumor was diffuse astrocytoma, IDH-mutant, World Health Organization grade II. Both ATRX and TP53 mutations were additionally confirmed by sequencing (data not shown) (original magnifications X400 [A, C, and D] and X200 [B]).

Caption: Figure 4. Initiation and propagation of infiltrating glioma. Cell-specific susceptibility factors, including the endogenous metabolic milieu, germ line genetic alterations, and environmental insults, conspire to trigger early mutational events, such as Isocitrate Dehydrogenase (IDH) mutation. The metabolic milieu is altered, including aberrant generation of 2-hydroxyglutarate. Transcriptional networks and cell-of-origin-specific chromatin topology dictates susceptibility to particular additional genetic lesions, including Tumor Protein P53 (TP53) mutation or 1p/19q codeletion. Multiple tumor subclones are generated, some of which may be more critical to tumor maintenance (eg, glioma stem cells) and some of which may be less crucial to tumor sustenance but nevertheless impart remarkable heterogeneity to the tumor population. Abbreviation: OPC, oligoden-droglial progenitor cell.

[Please Note: Illustration(s) are not available due to copyright restrictions.]
COPYRIGHT 2017 College of American Pathologists
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2017 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Pisapia, David J.
Publication:Archives of Pathology & Laboratory Medicine
Article Type:Report
Date:Dec 1, 2017
Words:11727
Previous Article:Survey on Transfusion-Transmitted Cytomegalovirus and Cytomegalovirus Disease Mitigation.
Next Article:A Systematic Analysis of Discordant Diagnoses in Digital Pathology Compared With Light Microscopy.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters