The pathological diagnosis of diffuse gliomas: towards a smart synthesis of microscopic and molecular information in a multidisciplinary context

Open AccessPublished:October 03, 2011DOI:https://doi.org/10.1016/j.mpdhp.2011.08.005

      Abstract

      Gliomas form a heterogeneous group of tumours of the central nervous system. Most gliomas, the so-called ‘diffuse gliomas’, are characterized by diffuse infiltrative growth in the neuropil. Based on histopathological analysis, diffuse gliomas are subdivided in astrocytic, oligodendroglial, and oligoastrocytic tumours and graded as WHO grade II (low grade), WHO grade III (anaplastic), or WHO grade IV (glioblastoma). Accurate distinction between the different diffuse glioma types and malignancy grades has significant prognostic and therapeutic implications. This review describes essential aspects of the microscopic diagnosis of diffuse gliomas. Because of problems with tissue sampling and imprecise criteria for typing and grading, unequivocal histopathological classification of diffuse gliomas can be challenging. The addition of molecular parameters contributes to a more objective pathological diagnosis of diffuse gliomas. A multidisciplinary approach is mandatory for achieving an optimal diagnosis for patients with these tumours.

      Keywords

      Introduction

      The histogenesis and biological behaviour of primary tumours of the central nervous system (CNS) and its coverings is very diverse. Over half of such tumours encountered in the general population are benign, the major categories being meningiomas, schwannomas and pituitary adenomas (Figure 1).

      CBTRUS; Central Brain Tumor Registry of the United States 2010. www.cbtrus.org.

      The most frequent primary malignant tumour of the CNS is glioma. Gliomas are considered to originate from glial (progenitor) cells or stem cells that develop glial characteristics. The group of gliomas encompasses many different histological types and malignancy grades. The most malignant tumour in this group, i.e. glioblastoma (WHO grade IV), is by far the most common glioma. Most gliomas are characterized by diffuse infiltrative growth of tumour cells in the preexistent parenchyma of the CNS and can be typed based on their histopathological features as astrocytic, oligodendroglial, or oligoastrocytic tumours. Additionally, a malignancy grade is attributed to these gliomas based on the presence of esp. the following features: nuclear atypia, mitotic activity, florid microvascular proliferation, and necrosis.
      Figure thumbnail gr1
      Figure 1Relative frequency of primary brain tumours; information based on 158,088 patient diagnoses in 2004–2006 (CBTRUS Statistical Report 2010; see CBTRUS.org).
      The histopathological diagnosis is the gold standard for the classification of gliomas. An adequate microscopic diagnosis carries important prognostic information and forms the basis for further patient management. Even after the current standard of care (surgery, radiotherapy and chemotherapy) most glioblastoma patients die within 1–2 years after diagnosis. In contrast, patients with a WHO grade II glioma may survive for over 10 years, but sooner or later such tumours generally progress to a high-grade malignant lesion. Distant metastases of diffuse gliomas are very rare. Still, because of their locally aggressive behaviour and the fact that they cannot be cured by current therapies, diffuse gliomas are considered as one of the most devastating cancers.
      Unequivocal histopathological classification of diffuse gliomas can be challenging because of e.g. problems with tissue sampling and the fact that the biological diversity in the spectrum of these tumours is difficult to capture in precise histopathological criteria. As it is impossible to cover all such aspects in one review, we will stress the essentials of the microscopic diagnosis as proposed by the most recent WHO edition (2007) on tumours of the CNS.
      For detailed information on the diagnosis of diffuse gliomas, the value of immunohistochemistry, or the intra-operative diagnosis by frozen sections and/or smears, the reader is referred to other recent review articles and textbooks.

      Burger P C, Scheithauer B W. Tumors of the central nervous system; AFIP Atlas of tumor pathology, 4th series, vol. 7; 2007.

      • Perry A.
      • Brat D.J.
      Practical surgical neuropathology.
      • Brat D.J.
      • Prayson R.A.
      • Ryken T.C.
      • Olson J.J.
      Diagnosis of malignant glioma: role of neuropathology.
      • Dunbar E.
      • Yachnis A.T.
      Glioma diagnosis: immunohistochemistry and beyond.
      • Burger P.C.
      Smears and frozen sections in surgical neuropathology: a manual.
      Molecular analysis aids in more robust classification of diffuse gliomas into clinically meaningful subgroups. We briefly summarize the state of the art in this field, this summary is based on the recent, detailed review of Riemenschneider et al. on molecular diagnostics in gliomas.
      • Riemenschneider M.J.
      • Jeuken J.W.
      • Wesseling P.
      • Reifenberger G.
      Molecular diagnostics of gliomas: state of the art.
      Finally, we stress the importance of a multidisciplinary approach for obtaining an optimal diagnosis in patients who suffer from diffuse gliomas.

      Morphology

       Microscopic recognition of diffuse gliomas

      Macroscopic evaluation of biopsies or resection specimens is generally of little help in reaching a diagnosis of diffuse glioma. In larger specimens a gradual transition of normal appearing grey or white matter into a lesion with greyish discolouration and blurring of the preexistent anatomical structures may be seen, compatible with the presence of a diffuse glioma. In this context necrosis indicates high-grade malignancy, particularly in patients who did not previously receive radio- or chemotherapy. Extensive calcification is more often found in oligodendroglial than in astrocytic tumours. However, microscopic analysis is required for a specific pathological diagnosis of diffuse glioma. Following the criteria of the WHO classification of tumours of the central nervous system,
      diffuse gliomas are microscopically typed as astrocytic, oligodendroglial or oligoastrocytic gliomas and graded as WHO grade II, III or IV.
      It is important to exclude other diagnoses before reaching the diagnosis of diffuse glioma. To achieve this, the pathologist should follow a decision tree (Figure 2). Especially in biopsy specimens that are small or suboptimally preserved the distinction between normal and abnormal tissue can already be problematic. Various non-neoplastic changes such as reactive astrocytosis, inflammatory lesions (e.g. multiple sclerosis), and infarcts may be difficult to distinguish from diffuse glioma. Occasionally, high-grade diffuse gliomas may show extensive areas with epithelial or sarcomatoid parts suggestive of metastatic carcinoma or sarcoma. Immunohistochemistry for glial fibrillary acidic protein (GFAP) and a spectrum of other markers (CD68, keratins, EMA, S100, neuronal markers) are generally helpful in this respect as well as in the differential diagnosis of diffuse gliomas with other, non-glial, primary CNS tumours such as schwannoma, meningioma, neurocytoma, and primitive neuroectodermal tumour (PNET).
      Figure thumbnail gr2
      Figure 2Decision tree for step-by-step exclusion of differential diagnostic options in the microscopic diagnosis of diffuse gliomas.
      Within the group of glial neoplasms, diffuse gliomas should be distinguished from other gliomas. The latter include the so-called ‘circumscribed’ gliomas (e.g. pilocytic astrocytomas, pleiomorphic xanthoastrocytomas), ependymal tumours and neoplasms showing a combination of glial and neuronal differentiation (gangliogliomas, glioneuronal tumours, PNETs). The presence of Rosenthal fibres and eosinophilic granular bodies in a glial tumour contributes to the diagnosis of circumscribed glioma or ganglion cell tumour. Additional features that are helpful to separate ‘other’ gliomas from diffuse gliomas are biphasic growth pattern (pilocytic astrocytoma), the presence of abnormal (clusters of) ganglion cells (ganglioglioma), the formation of true rosettes or perivascular pseudorosettes (ependymal tumours), and immunohistochemical staining for e.g. neuronal markers such as synaptophysin and Neu-N (ganglioglioma) or for EMA (ependymal tumours).
      Recognition of extensive, diffuse infiltrative growth in preexistent brain tissue is of great help in the diagnosis of diffuse glioma. Histopathologically, the tumour cells tend to invade individually or in small groups in the neuropil, i.e. the network of neuronal and glial cell processes in grey and white matter. Only few neoplasms other than diffuse gliomas (lymphomas, occasionally metastases of small cell carcinoma) display such growth in the CNS. Furthermore, the arrangement of diffuse infiltrative glioma cells along pre-existing tissue elements is very helpful in recognizing a tumour as diffuse glioma. Hans-Joachim Scherer, a pioneer in the study of glioma growth patterns, designated such arrangements as “secondary structures”. Examples of secondary structures are perineuronal growth (perineuronal satellitosis), surface/subpial growth, perivascular growth, and intrafascicular growth (Figure 3). Diffuse gliomas tend to invade over large distances along myelinated fibre tracts and in many patients the tumour cells cross the corpus callosum via such intrafascicular growth.
      • Claes A.
      • Idema A.J.
      • Wesseling P.
      Diffuse glioma growth: a guerilla war.
      Figure thumbnail gr3
      Figure 3Diffuse infiltrative glioma growth. Schematic representation (a) and microscopic example (b) showing diffuse infiltration of glioma cells in the preexistent brain parenchyma with secondary structures of Scherer: accumulation of tumour cells around neurons (perineuronal satellitosis, arrow head), around blood vessels (arrow), under the pia (asterisk), and tumour cells migrating along white matter tracts (intrafascicular growth; + in a). b: Haematoxylin-and-eosin staining, orig. magn. ×200.
      Of note, the circumscribed gliomas may show some infiltration in the neuropil as well, but by far not as extensive as encountered in diffuse gliomas and typically without the formation of secondary structures of Scherer. Vice versa, it may be impossible to appreciate diffuse infiltrative growth in highly cellular regions of diffuse gliomas, e.g. in the core of high-grade malignant lesions. Features of high-grade malignancy such as brisk mitotic activity and necrosis strongly favour a diagnosis of diffuse rather than circumscribed glioma because the latter infrequently shows malignant progression.

       Typing of diffuse gliomas

      Based on the resemblance of the tumour cells with non-neoplastic glial cells (astrocytes, oligodendroglial cells), most diffuse gliomas can be typed as astrocytic, oligodendroglial, or mixed oligoastrocytic tumours. Normal astrocytes are stellate cells with an oval-to-elongate, somewhat vesicular nucleus, little eosinophilic cytoplasm and delicate eosinophilic cell processes. The phenotype of these cells and their processes is hard to identify in the fine fibrillar neuropil network without immunohistochemical (GFAP) staining. Reactive astrocytes often show marked increase in cytoplasm with some enlargement of the nucleus and more stout cellular processes, or may undergo gemistocytic change with a plump, rounded or angular, eosinophilic cell body and an eccentric nucleus. Astrocytic tumour cells may show a mixture of these phenotypes with variable nuclear atypia and mitotic activity. Because of the diffuse infiltration, various reactive glial cells are intermixed with the tumour cells and the neuropil may be represented by relatively normal white or grey matter or may have undergone microcystic or fibrillar change. Several phenotypical variants of diffuse low-grade astrocytoma are described in the WHO classification: fibrillary astrocytoma, the most common phenotype composed of tumour cells with clear fibrillary cell processes; protoplasmic astrocytoma which consists of tumour cells with small cell bodies with few, flaccid processes; and gemistocytic astrocytoma, consisting of at least 20% of cells with the aforementioned plump phenotype. Gemistocytic astrocytomas often harbour dispersed perivascular lymphocytic infiltrates and tend to show more rapid malignant progression than other WHO grade II astrocytomas.
      Normal oligodendroglial cells are characterized by a round nucleus with a relatively dense and delicate chromatin pattern and a perinuclear halo and lack of staining for GFAP. Typical oligodendroglial tumour cells also have a round rather than oblong nucleus and with a perinuclear halo. Oligodendroglial tumours are often highly cellular lesions with compact fields of relatively small and round cells and less cellular areas in which the diffuse infiltrative nature is more easily appreciated, e.g. in the form of perineuronal satellitosis. Additional features indicative of oligodendroglial rather than astrocytic differentiation are the presence of a branching network of delicate capillaries (‘chickenwire pattern’) and extensive calcification in the tumour. Some oligodendroglial tumours show scattered cells with marked polymorphism (‘polymorphic oligodendroglioma’) or cells arranged in a rhythmic/spongioblastic fashion.
      The presence of areas with less prominent oligodendroglial features is acceptable for the diagnosis of pure oligodendroglioma. Esp. in the peripheral, diffuse infiltrative part of oligodendroglial tumours the neoplastic cells may acquire a non-descript or even more astrocytic phenotype. Furthermore, in pure oligodendroglial tumours part of the tumour cells may show a gliofibrillary or minigemistocytic phenotype with strong GFAP staining of the cytoplasm. It is important to realize that the presence of perinuclear halos in oligodendroglial tumours is in fact the result of delayed fixation and can thus be absent in specimens that are more promptly fixed or used for frozen section diagnosis. One should thus exert great caution when asked to precisely type (and grade) diffuse gliomas intra-operatively using frozen sections or smears.
      • Burger P.C.
      Smears and frozen sections in surgical neuropathology: a manual.
      • Plesec T.P.
      • Prayson R.A.
      Frozen section discrepancy in the evaluation of central nervous system tumors.
      • Powell S.Z.
      Intraoperative consultation, cytologic preparations, and frozen section in the central nervous system.
      As indicated by their name, the key criterion for the diagnosis of mixed oligoastrocytomas is the presence of neoplastic glial cells with convincing morphological characteristics of astrocytes and oligodendrocytes. The tumour cells of both lineages may be diffusely mixed or separated, but the latter form is rare. The precise extent of the oligodendroglial versus astrocytic part is often hard to define, also because these glial neoplasms generally show areas in which the tumour cells are not easily recognized as either astrocytic or oligodendroglial. Consequently, exact criteria for distinguishing mixed oligoastrocytic tumours from pure astrocytic or oligodendroglial neoplasms are not available, and substantial inter-observer variability exists in the recognition of these different types of diffuse gliomas.
      • Kros J.M.
      • Gorlia T.
      • Kouwenhoven M.C.
      • et al.
      Panel review of anaplastic oligodendroglioma from European Organization for Research and Treatment of Cancer Trial 26951: assessment of consensus in diagnosis, influence of 1p/19q loss, and correlations with outcome.
      In daily clinical practice, diffuse gliomas lacking prototype oligodendroglial and astrocytic phenotype of the tumour cells are often considered as astrocytic neoplasms rather than diagnosed as glioma ‘not otherwise specified’.
      Within the glioblastoma category several phenotypical variants are recognized, the most frequent of these being giant cell glioblastoma (with a predominance of multinucleated, giant tumour cells), small cell glioblastoma (with a predominance of small, relatively monomorphous tumour cells with little cytoplasm), and gliosarcoma (with extensive presence of a sarcomatoid phenotype). Recognition of these variants is important because of the differential diagnosis with other tumours requiring another treatment (e.g. giant cell glioblastoma versus pleomorphic xanthoastrocytoma; small cell glioblastoma versus anaplastic oligodendroglioma or PNET; gliosarcoma versus meningeal or metastatic sarcoma). In this context the result of immunohistochemistry for GFAP may be of little help. For instance, both the small cell component of glioblastomas and oligodendrogliomas are notorious for the lack of GFAP staining of the tumour cells.

       Grading of diffuse gliomas

      According to the WHO-2007 classification diffuse gliomas are graded as WHO grade II (low grade), WHO grade III (anaplastic) or WHO grade IV (glioblastoma) whereas WHO grade I is reserved for the more circumscribed gliomas.
      For grading of diffuse gliomas the histological features nuclear atypia, mitotic activity, necrosis, and florid microvascular proliferation (MVP) are used (Figure 4). A diffuse astrocytic neoplasm without marked mitotic activity, necrosis or florid MVP is diagnosed as low-grade diffuse astrocytoma, irrespective of the degree of nuclear atypia. When the tumour shows marked mitotic activity the diagnosis of anaplastic astrocytoma is rendered. The presence of necrosis and/or florid MVP leads to a diagnosis of glioblastoma. Often, necrosis in these tumours consists of irregular, serpiginous foci surrounded by densely packed, somewhat radially oriented small tumour cells (‘pseudopalisading necrosis’).
      Figure thumbnail gr4
      Figure 4Microscopic criteria used for grading of diffuse gliomas.
      In oligodendroglial and mixed oligoastrocytic tumours the malignancy grade is assessed using the same set of histological criteria but in a somewhat different way. In (pure) oligodendroglial tumours necrosis and florid MVP do not have the same unfavourable connotation as in diffuse astrocytic neoplasms. Oligodendrogliomas with MVP or necrosis are still considered as WHO grade III lesions. In mixed oligoastrocytic tumours the presence of MVP is compatible with WHO grade III, but the presence of necrosis in this category leads to a diagnosis of glioblastoma with oligodendroglial component (GBM-O). Examples of the histology of diffuse low-grade and high-grade astrocytic, oligodendroglial, and oligoastrocytic tumours are given in (Figure 5).
      Figure thumbnail gr5
      Figure 5Examples of microscopic features of diffuse gliomas and non-neoplastic glial cells: (a) GFAP positive reactive astrocytes; (b) oligodendrocytes in normal white matter; (c) diffuse low-grade (fibrillary) astrocytoma; (d) gemistocytic astrocytoma; (e) low-grade oligodendroglioma (arrows indicate calcifications); (f) anaplastic oligodendroglioma; (g) GFAP staining of minigemistocytes and gliofibrillary cells in oligodendroglial tumour; (h) glioblastoma with oligo-component (arrow indicates pseudopalisading necrosis); (i) glioblastoma with florid/glomeruloid microvascular proliferation and pseudopalisading necrosis; (j) small cell variant of glioblastoma with pseudopalisading necrosis; (k) giant cell glioblastoma; (l) gliosarcoma. All images show haematoxylin-and-eosin staining and orig. magn. ×200, except for a, g (GFAP staining, orig. magn. ×400).
      The scheme presented in Figure 4 may give the impression of a robust and reproducible procedure. In daily clinical practice, however, not only typing but also grading of gliomas can be difficult. Firstly, tissue sampling can be incomplete and a whole spectrum of other factors (pre-operative, surgical, pathological) may negatively influence tissue quantity and quality, resulting in a suboptimal diagnosis, e.g. underestimation of the true malignancy in regionally heterogeneous tumours (Figure 6). Secondly, the reproducibility of the WHO criteria for typing and grading of diffuse gliomas is not optimal also due to the fact that it is impossible to capture the biological variation in these tumours in strict criteria. The scoring of nuclear atypia is problematic because no scale for scoring is available. In addition, the degree of cell density and nuclear atypia may not correlate with overall malignancy grade: low-grade oligodendrogliomas are often highly cellular while parts of glioblastomas may show moderate cellularity. Significant nuclear atypia may be encountered in otherwise low-grade gliomas while not be a prominent feature in glioblastomas.
      Figure thumbnail gr6
      Figure 6Important determinants for quality and quantity of tissue used for microscopic and molecular diagnostics of tumours of the central nervous system.
      Evaluation of mitotic activity should be performed in the context of sample size: a single mitosis in a large resection specimen is not sufficient for the diagnosis of anaplastic glioma, whereas the identification of a single mitosis in a small biopsy fragment may well indicate high proliferative activity and thus a WHO grade III neoplasm. The Ki-67/MIB-1 labelling index increases with malignancy grade (roughly in low-grade tumours <5%, in anaplastic gliomas between 5% and 10%, and in glioblastomas over 10%). However, MIB-1 labelling index is not routinely incorporated into a grading system because of substantial overlap of this marker for the different malignancy grades, tumour heterogeneity and differences in staining results between different laboratories.
      In glioblastomas prominent, often ‘glomeruloid’ MVP and/or necrosis emerge. These features are generally spatially and temporally distributed.
      • Claes A.
      • Idema A.J.
      • Wesseling P.
      Diffuse glioma growth: a guerilla war.
      Florid MVP is a term used for the presence of multilayered microvessels with hypertrophy and hyperplasia of endothelial cells and pericytes in the vessel walls. The minimum requirements for recognition of this phenomenon are not clear. Also the recognition of necrosis may be troublesome in biopsy samples that are small or poorly preserved. As a consequence substantial inter-observer variation exists in the classification of diffuse gliomas even among experienced neuropathologists, which may well have undesirable clinical consequences.
      • Kros J.M.
      • Gorlia T.
      • Kouwenhoven M.C.
      • et al.
      Panel review of anaplastic oligodendroglioma from European Organization for Research and Treatment of Cancer Trial 26951: assessment of consensus in diagnosis, influence of 1p/19q loss, and correlations with outcome.
      • van den Bent M.J.
      Interobserver variation of the histopathological diagnosis in clinical trials on glioma: a clinician’s perspective.

      Molecular analysis

       Molecular aberrations in diffuse gliomas

      Like other cancers, diffuse gliomas result from (epi)genetic alterations that lead to inactivation of tumour suppressor genes or activation of proto-oncogenes and that accumulate with tumour progression. Low-grade diffuse astrocytomas frequently show mutation of the tumour suppressor gene TP53 (located at 17p13.1), loss of heterozygosity (LOH) on chromosome arm 17 p, and gains on the long arm of chromosome 7 (7q). In contrast, combined and complete loss of the short arm of chromosome 1 (1p) and of the long arm of chromosome 19 (19q) is present in the vast majority of typical oligodendroglial tumours. The complete 1p/19q co-deletion results from a translocation between chromosome 1 and 19. Some gliomas histopathologically diagnosed as oligoastrocytic tumour genetically resemble either pure astrocytic or oligodendroglial tumours. Recently, mutation of the isocitrate dehydrogenase 1 (IDH1) gene (less commonly of the IDH2 gene) was identified as a common and early event in the oncogenesis of low-grade and anaplastic diffuse gliomas.
      • Riemenschneider M.J.
      • Jeuken J.W.
      • Wesseling P.
      • Reifenberger G.
      Molecular diagnostics of gliomas: state of the art.
      Anaplastic diffuse gliomas often carry additional, progression-associated genetic changes such as loss of the tumour suppressor genes CDKN2A (coding for p14ARF and p16INK4A), CDKN2B on 9p21, deletions on chromosomes 6, 11p, 22q, amplification of CDK4 or CDK6, inactivating alterations of RB1. The vast majority of glioblastomas present de novo in elderly patients with a short clinical history (‘primary glioblastomas’). In contrast, ‘secondary glioblastomas’ develop by progression from pre-existing lower grade gliomas. While primary glioblastomas frequently show EGFR amplification and PTEN mutation and lack IDH1 mutation, the secondary glioblastomas are characterized by frequent mutations in the TP53 and IDH1 genes while lacking EGFR amplification. Additionally, at the chromosomal level, primary glioblastomas are distinct from secondary glioblastomas by the frequent occurrence of trisomy of chromosome 7, monosomy of chromosome 10, and gains of chromosome arms 12p, 19q and 20q. Despite these differences and the slightly better prognosis for patients with secondary glioblastomas, most of the genetic alterations in primary and secondary glioblastomas can be assigned to a common set of functional pathways.
      • Riemenschneider M.J.
      • Jeuken J.W.
      • Wesseling P.
      • Reifenberger G.
      Molecular diagnostics of gliomas: state of the art.
      • Ohgaki H.
      • Kleihues P.
      Genetic pathways to primary and secondary glioblastoma.
      • TCGA; The Cancer Genome Atlas Research Network
      Comprehensive genomic characterization defines human glioblastoma genes and core pathways.
      • Jeuken J.W.M.
      • Sijben A.
      • Bleeker F.E.
      • et al.
      The nature and timing of specific copy number changes in the course of molecular progression in diffuse gliomas; further elucidation of their genetic ‘life story’.
      Evidence is accumulating that certain molecular changes can be used as biomarkers that provide additional diagnostic, prognostic and/or predictive information that supplements or even overrules the information provided by microscopic investigation. Thereby, these markers can be used to improve tailored management of glioma patients. The presently most promising molecular markers in diffuse gliomas are combined deletion of the chromosome arms 1 p and 19q, IDH1 and IDH2 mutations, hypermethylation of the MGMT promoter, and EGFR amplification/EGFRvIII expression (Figure 7).
      • Riemenschneider M.J.
      • Jeuken J.W.
      • Wesseling P.
      • Reifenberger G.
      Molecular diagnostics of gliomas: state of the art.
      • Tabatabai G.
      • Stupp R.
      • van den Bent M.J.
      • et al.
      Molecular diagnostics of gliomas: the clinical perspective.
      These aberrations are discussed in somewhat more detail below. The role of molecular diagnostics can be expected to rapidly increase. For instance, effective implementation of targeted therapeutic approaches using ‘small molecules’ will require information on the relevant molecular components in individual tumours.
      Figure thumbnail gr7
      Figure 7Indication of diffuse glioma types and malignancy grades in which particular molecular aberrations are relatively common.

       1 p/19q loss

      Losses of 1p and 19q are detected in up to 80% of low grade oligodendrogliomas and approximately 60% of anaplastic oligodendrogliomas, whereas 30–50% of low grade oligoastrocytomas, 15–20% of anaplastic oligoastrocytomas, and less than 10% of diffuse astrocytic gliomas, including the glioblastomas, carry this aberration. There is a strong association between 1p/19q co-deletion and classical oligodendroglial features on histology. Discussion remains regarding the predictive (response to therapy) versus prognostic (independent of therapy) nature of this marker. Recent evidence suggests that 1p/19q loss characterizes a group of gliomas that are more sensitive to genotoxic therapy (radiotherapy and alkylating chemotherapy) in general and is associated with significantly longer survival. The prognostic relevance of 1p/19q loss may be less pronounced in the presence of other, prognostically unfavourable genetic alterations. Dependent on the exact set of probes used, loss of heterozygosity (LOH) or (fluorescent) in situ hybridization ((F)ISH) analysis (i.e. the most commonly used techniques for 1p/19q testing in daily clinical practice) may not allow for the distinction between tumours with partial, terminal or interstitial 1p or 19q loss and those with complete, combined 1p/19q loss. As several studies now demonstrated that in this context only complete co-deletion of 1p and 19q signifies a more favourable prognosis, it is important to use a technique for molecular diagnosis that allows for robust recognition of complete loss of these chromosome arms. Testing for 1p/19q co-deletion is nowadays often used to stratify patients into different clinical trials.
      • Riemenschneider M.J.
      • Jeuken J.W.
      • Wesseling P.
      • Reifenberger G.
      Molecular diagnostics of gliomas: state of the art.
      • Jeuken J.W.M.
      • Sijben A.
      • Bleeker F.E.
      • et al.
      The nature and timing of specific copy number changes in the course of molecular progression in diffuse gliomas; further elucidation of their genetic ‘life story’.

       IDH1/IDH2 mutations

      Only in 2008 mutations in the gene encoding the human cytosolic NADPH-dependent isocitrate dehydrogenase (IDH1) were identified using large-scale sequencing analysis of 22 glioblastomas.
      • Parsons D.W.
      • Jones S.
      • Zhang X.
      • et al.
      An integrated genomic analysis of human glioblastoma multiforme.
      The mutations were detected preferentially in glioblastomas of young patients and in secondary glioblastomas. Soon after, multiple studies corroborated these findings and additionally revealed that somatic IDH1 (and less frequently IDH2) mutations are present in the vast majority of diffuse, low-grade and anaplastic gliomas and are strongly associated with better outcome. The presence of IDH1 mutations in diffuse gliomas is strongly correlated with TP53 mutation or 1p/19q deletion and seems to represent a very early event affecting a common glial precursor cell. The fact that the vast majority of high-grade gliomas in the paediatric age group lacks IDH1 mutation corroborates the notion that these paediatric neoplasms are fundamentally different from their adult counterparts. It is increasingly clear that the heterozygous IDH1 and IDH2 mutations in gliomas not just result in a loss of function but also cause excessive production of the ‘oncometabolite’ 2-hydroxyglutarate (2HG). Testing for mutations in IDH1 or IDH2 can now effectively be performed using various molecular methods. Additionally, specific monoclonal antibodies were established that allow for immunohistochemical analysis of the product of the most frequently occurring IDH1 mutation (R132H; representing >90% of the IDH mutations in gliomas). IDH1 and IDH2 mutation analysis or immunohistochemistry using the antibody against the IDH1 R132H mutated protein may be of help in the differential diagnosis of diffuse glioma versus pilocytic astrocytoma or ependymoma, for discrimination between primary and secondary glioblastoma, and in the differential diagnosis of diffusely infiltrating glioma versus reactive astrogliosis.
      • Riemenschneider M.J.
      • Jeuken J.W.
      • Wesseling P.
      • Reifenberger G.
      Molecular diagnostics of gliomas: state of the art.

       MGMT

      The O6-methylguanine–DNA methyltransferase (MGMT) gene is located at chromosome 10q26. In diffuse gliomas this gene is frequently silenced by promoter hypermethylation. An association has been reported between this epigenetic event and the response of malignant gliomas to alkylating chemotherapy. The predictive significance of MGMT promoter status can be explained by the fact that MGMT encodes a DNA repair protein that removes alkyl groups from the O6 position of guanine, which are introduced by alkylating chemotherapeutic agents (like temozolomide) that are frequently used in glioma patients. In other studies, however, a prognostic favourable effect of MGMT promoter methylation was found in patients receiving radiotherapy only, indicative of a prognostic effect besides the putative predictive effect. Interestingly, MGMT hypermethylation in diffuse gliomas is strongly associated with prognostically favourable molecular features such as 1p/19q co-deletion and IDH1 mutation. MGMT promoter methylation may thus indeed reflect a global molecular constellation that is associated with higher sensitivity to cytotoxic therapy and more favourable outcome. Testing for MGMT promoter methylation is usually performed by methylation-specific polymerase chain reaction (MSP or MS-PCR), methylation-specific pyrosequencing or methylation-specific multiplex ligation-dependent probe amplification (MS-MLPA). Alternative tests such as simple immunohistochemical analysis of MGMT protein expression or expression analyses based on Western blotting, reverse transcription-PCR, or biochemical assays detecting enzymatic activity have proven to be difficult to interpret because of the presence of contaminating non-neoplastic cells in the investigated tissue specimens. It is important to realize that the MGMT promoter methylation pattern is very heterogeneous from tumour to tumour and that the relevant (combinations of) CpG sites in the MGMT promoter that need to be methylated to silence transcription are not yet defined.
      • Riemenschneider M.J.
      • Jeuken J.W.
      • Wesseling P.
      • Reifenberger G.
      Molecular diagnostics of gliomas: state of the art.
      • Weller M.
      • Stupp R.
      • Reifenberger G.
      • et al.
      MGMT promoter methylation in malignant gliomas: ready for personalized medicine?.

       EGFR amplification/EGFRvIII

      The epidermal growth factor receptor gene (EGFR) at chromosome 7p12 is the most frequently amplified and overexpressed gene in primary glioblastomas, affecting approximately 40% of these tumours. EGFR rearrangements are also frequently found. The most common EGFR variant is variant III (EGFRvIII), consisting of an 801-bp in-frame deletion of exons 2–7 that results in a constitutively activated truncated receptor protein lacking the ligand binding domain. Identification of EGFR amplification and rearrangements (like EGFRvIII) is suggestive of high-grade malignancy and therefore may provide diagnostic as well as prognostic information. In fact, detection of EGFR amplification/EGFRvIII in anaplastic or low-grade gliomas strongly suggests that these tumours are more malignant than indicated by their histopathology. Detection of EGFR aberrations also may be relevant from a therapeutic point of view. Up till now, the clinical benefit of the use of EGFR inhibitors in glioblastomas has been rather disappointing. However, the EGFRvIII mutant may serve as an attractive target for immunotherapy, and recent studies reported that the anti-EGFRvIII peptide vaccine CDX-110 increased progression free and overall survival in EGFRvIII-positive glioblastoma patients. Existing patents currently prohibit the use of an antibody specifically recognizing EGFRvIII for immunohistochemistry in the clinic. As an alternative, EGFRvIII analysis can be performed by reverse transcription-PCR analysis using primers located in exon 1 and 9 or by MLPA analysis.
      • Riemenschneider M.J.
      • Jeuken J.W.
      • Wesseling P.
      • Reifenberger G.
      Molecular diagnostics of gliomas: state of the art.
      • Jeuken J.W.M.
      • Sijben A.
      • Bleeker F.E.
      • et al.
      The nature and timing of specific copy number changes in the course of molecular progression in diffuse gliomas; further elucidation of their genetic ‘life story’.

      Multidisciplinary approach

       Radiological presentation

      Clinical information (esp. patient age, duration of symptoms, previous treatment) and radiological findings (including location and growth pattern of the tumour, contrast-enhancement) provide important clues for narrowing down the differential diagnosis in patients with a tumour of the CNS.
      • Perry A.
      • Brat D.J.
      Practical surgical neuropathology.
      Magnetic resonance imaging (MRI) is now the gold standard for radiological assessment of these tumours and can in fact be used by pathologists as a surrogate for the macroscopy of these neoplasms.
      Low-grade diffuse gliomas show hyperintensity on T2 weighted MRI scans but generally do not enhance in T1 weighted images when using the contrast-agent Gadolinium-DTPA. The absence of contrast-enhancement in these tumours can be explained by incorporation (‘coöption’) of preexistent microvessels with only limited changes to the blood–brain barrier (BBB) and lack of neovascularization. Contrast-enhancement on MRI scans of high-grade gliomas indicates disruption of the BBB of preexistent or newly formed microvessels. Many glioblastomas present radiologically with a non-enhancing, necrotic core surrounded by a contrast-enhancing ring of viable, highly cellular and angiogenic tumour tissue (ring-enhancement). However, the invasive front of these tumours is difficult to visualize by radiology, also because the periphery of high-grade malignant gliomas often lacks contrast-enhancement. Correlation of histological sections of glioblastomas with radiology revealed that tumour cells of diffuse gliomas are often present several centimetres outside the enhancing area and even outside the hyperintense areas on T2 weighted MR images.
      • Claes A.
      • Idema A.J.
      • Wesseling P.
      Diffuse glioma growth: a guerilla war.
      Radiological information may be very helpful in cases where the differential diagnosis of e.g. diffuse versus circumscribed glioma is difficult for the pathologist. Occasionally, in patients with multiple contrast-enhancing lesions in the brain and a clinical differential diagnosis of metastatic or inflammatory/infectious disease, histopathological analysis reveals a diffuse glioma. This latter presentation can often be explained by multifocal dedifferentiation in a widespread diffuse glioma of lower malignancy grade.

       Multidisciplinary approach: mandatory!

      Even when clinical and radiological information is not available, in some cases the pathologist may be able to make an unequivocal and correct diagnosis on a brain tumour biopsy alone (e.g. glioblastoma). However, diffuse gliomas often show marked phenotypical heterogeneity with spatial differences in cellular phenotype and malignancy grade. When only small biopsies are available the malignancy grade of diffuse gliomas may therefore be underestimated (Figure 8). Also, it is impossible to make a diagnosis of gliomatosis cerebri (i.e. the most extreme example of diffuse infiltrative glioma) on histopathological analysis of a biopsy alone, as the required involvement of at least three cerebral lobes (and usually bilateral growth) can only be appreciated by evaluation of radiological images or in post-mortem material. Both radiological and pathological assessment of response to different treatment strategies for diffuse gliomas can be challenging. For example, distinguishing pure radiation necrosis from radiation necrosis with recurrent glioma in biopsy material is often virtually impossible (Figure 9).
      • Wen P.Y.
      • Macdonald D.R.
      • Reardon D.A.
      • et al.
      Updated response assessment criteria for high-grade gliomas: response assessment in neuro-oncology working group.
      Figure thumbnail gr8
      Figure 8Case illustrating the importance of a multidisciplinary approach in the pathological diagnosis of brain tumours. A 58-year old woman showed a small, ring-enhancing lesion deep in the left cerebral hemisphere on MRI scans after Gadolinium (a); with a neuro-navigation guided biopsy procedure small tissue fragments were obtained for pathological analysis (fragments in container with physiologic salt solution (b), top of container (c) and paraffin block (d)); the vast majority of the haematoxylin-and-eosin stained slides of these fragments (e, f) revealed a low-grade diffuse astrocytoma; only in the periphery of one fragment (rectangle in f) one spot was found with florid microvascular proliferation (f); combining the pathology (diffuse astrocytic tumour, largely low grade, but with focal florid microvascular proliferation consistent with glioblastoma) with the radiology (ring-enhancing lesion highly suggestive of a necrotic core in the tumour) led to a final ‘multidisciplinary’ diagnosis of glioblastoma; the biopsy material was considered as a peripheral, not fully representative part of the lesion.
      Figure thumbnail gr9
      Figure 9Case illustrating problematic differential diagnosis between therapy effect and recurrent glioma. A 28-year old male was diagnosed with diffuse, low-grade astrocytoma (a) in the right frontal lobe for which he received chemotherapy (1×) and irradiation (2×) in the course of multiple years; 5 years after his initial biopsy MRI scans with Gadolinium revealed a rapidly growing lesion suspect for glioblastoma but with a differential diagnosis of radionecrosis (b); histopathologically, in a biopsy of this lesion fibrinoid necrosis was present consistent with radionecrosis and with a gradual transition to paucicellular tumour tissue lacking (other) features of high-grade malignancy (c); MRI scans 5 months later revealed spontaneous regression of the lesion (d), corroborating the notion that this patient suffered from radionecrosis rather than from malignant progression of the diffuse glioma.
      In this context we advocate that in daily clinical practice the pathologist starts formulating a diagnosis based on the morphological features in brain tumour specimens and subsequently puts the morphological diagnosis in perspective of the relevant clinical and radiological findings. In our view, the pathologist should actively participate in multidisciplinary neuro-oncology meetings in order to be able to ‘fine-tune’ the pathological diagnosis and to check if additional molecular testing is warranted. Such a multidisciplinary approach in the diagnosis of diffuse gliomas is a key for optimal management of patients with these tumours.

      Concluding remarks and future perspectives

      The unique diffuse infiltrative growth of gliomas in the brain parenchyma has important diagnostic, prognostic, and therapeutic implications. In a clinical setting combination of clinical, radiological, and pathological information is warranted to avoid diagnostic inaccuracy, particularly in cases where only small biopsy specimens are available for pathological diagnosis. So far, histopathology is the gold standard for typing and grading of gliomas. However, histological classification of gliomas is associated with significant inter-observer variability.
      Molecular diagnostics is a promising tool for more robust classification of diffuse gliomas. The challenge is then to combine the classic morphological approach with molecular diagnostics in order to provide the best possible information for the individual patient. The result of this combined approach may well be that parts of the morphological system for typing and grading of diffuse gliomas will become less relevant or even overruled by certain molecular findings. Meanwhile, parameters like tissue quality and quantity are of utmost importance also for molecular diagnostics and the pathologist as ‘tissue professional’ is instrumental in the assessment and selection of the material that is used for subsequent molecular testing.
      Our knowledge on the genes and pathways involved in the oncogenesis of diffuse gliomas can be expected to substantially expand further in the near future. Hopefully, this will reveal novel clinically useful markers and pave the way to the development of more effective drugs, thereby providing better opportunities for individualized therapy of patients with these so far incurable cancers.
      • The group of gliomas encompasses many different histological types and malignancy grades.
      • Most gliomas, the so-called ‘diffuse gliomas’, are characterized by diffuse infiltrative growth in the preexistent parenchyma of the CNS.
      • Diffuse gliomas are subtyped as astrocytic, oligodendroglial, or oligoastrocytic based on the resemblance of tumour cells with non-neoplastic astrocytes and/or oligodendrocytes.
      • Diffuse gliomas are graded as low grade (WHO grade II), anaplastic (WHO grade III), or glioblastoma (WHO grade IV) based the presence of nuclear atypia, mitotic activity, florid microvascular proliferation, and necrosis.
      • Accurate distinction between the different diffuse glioma types and malignancy grades has significant prognostic and therapeutic implications.
      • Histopathological typing and grading of diffuse gliomas can be challenging because of sampling effect and lack of unequivocal criteria.
      • Molecular diagnostics is a promising tool for more robust classification of diffuse gliomas.
      • The presently most relevant molecular markers in diffuse gliomas are complete 1 p/19q co-deletion, IDH1/IDH2 mutation, MGMT promoter methylation, and EGFR amplification/EGFRvIII expression.
      • The pathological diagnosis of diffuse gliomas should be evaluated in the context of clinical and radiological findings.

      References

      1. CBTRUS; Central Brain Tumor Registry of the United States 2010. www.cbtrus.org.

      2. Louis D.N. Ohgaki H. Wiestler O.D. Cavenee W.K. WHO classification of tumours of the central nervous system. 3rd edn. IARC Press, Lyon2007
      3. Burger P C, Scheithauer B W. Tumors of the central nervous system; AFIP Atlas of tumor pathology, 4th series, vol. 7; 2007.

        • Perry A.
        • Brat D.J.
        Practical surgical neuropathology.
        Churchill Livingstone, 2010
        • Brat D.J.
        • Prayson R.A.
        • Ryken T.C.
        • Olson J.J.
        Diagnosis of malignant glioma: role of neuropathology.
        J Neurooncol. 2008; 89: 287-311
        • Dunbar E.
        • Yachnis A.T.
        Glioma diagnosis: immunohistochemistry and beyond.
        Adv Anat Pathol. 2010; 17: 187-201
        • Burger P.C.
        Smears and frozen sections in surgical neuropathology: a manual.
        PB Medical Publishing LLC, 2009
        • Riemenschneider M.J.
        • Jeuken J.W.
        • Wesseling P.
        • Reifenberger G.
        Molecular diagnostics of gliomas: state of the art.
        Acta Neuropathol. 2010; 120: 567-584
        • Claes A.
        • Idema A.J.
        • Wesseling P.
        Diffuse glioma growth: a guerilla war.
        Acta Neuropathol. 2007; 114: 443-458
        • Plesec T.P.
        • Prayson R.A.
        Frozen section discrepancy in the evaluation of central nervous system tumors.
        Arch Pathol Lab Med. 2007; 131: 1532-1540
        • Powell S.Z.
        Intraoperative consultation, cytologic preparations, and frozen section in the central nervous system.
        Arch Pathol Lab Med. 2005; 129: 1635-1652
        • Kros J.M.
        • Gorlia T.
        • Kouwenhoven M.C.
        • et al.
        Panel review of anaplastic oligodendroglioma from European Organization for Research and Treatment of Cancer Trial 26951: assessment of consensus in diagnosis, influence of 1p/19q loss, and correlations with outcome.
        J Neuropathol Exp Neurol. 2007; 66: 545-551
        • van den Bent M.J.
        Interobserver variation of the histopathological diagnosis in clinical trials on glioma: a clinician’s perspective.
        Acta Neuropathol. 2010; 120: 297-304
        • Ohgaki H.
        • Kleihues P.
        Genetic pathways to primary and secondary glioblastoma.
        Am J Pathol. 2007; 170: 1445-1453
        • TCGA; The Cancer Genome Atlas Research Network
        Comprehensive genomic characterization defines human glioblastoma genes and core pathways.
        Nature. 2008; 455: 1061-1068
        • Jeuken J.W.M.
        • Sijben A.
        • Bleeker F.E.
        • et al.
        The nature and timing of specific copy number changes in the course of molecular progression in diffuse gliomas; further elucidation of their genetic ‘life story’.
        Brain Pathol. 2011; 21: 308-320
        • Tabatabai G.
        • Stupp R.
        • van den Bent M.J.
        • et al.
        Molecular diagnostics of gliomas: the clinical perspective.
        Acta Neuropathol. 2010; 120: 585-592
        • Parsons D.W.
        • Jones S.
        • Zhang X.
        • et al.
        An integrated genomic analysis of human glioblastoma multiforme.
        Science. 2008; 321: 1807-1812
        • Weller M.
        • Stupp R.
        • Reifenberger G.
        • et al.
        MGMT promoter methylation in malignant gliomas: ready for personalized medicine?.
        Nat Rev Neurol. 2010; 6: 39-51
        • Wen P.Y.
        • Macdonald D.R.
        • Reardon D.A.
        • et al.
        Updated response assessment criteria for high-grade gliomas: response assessment in neuro-oncology working group.
        J Clin Oncol. 2010; 28: 1963-1972