Section 3 -

Diffuse, Low Grade Gliomas: is it Possible to Distinguish Oligodendroglioma from Astrocytoma? (Adapted from Gupta, et al, Am. J Clin Pathol 2005; 124:1-14) Daniel J. Brat Department of Pathology and Laboratory Medicine Emory University School of Medicine Atlanta , GA

The infiltrative, or diffuse, gliomas are the most frequent primary central nervous system (CNS) tumors and include astrocytomas, oligodendrogliomas, and mixed oligoastrocytomas. Here we discuss the pathologic features of the two main subtypes—the oligodendroglioma and the astrocytoma—with an emphasis on morphologic features and diagnostic markers that can assist the pathologist to more reliably distinguish them. Histopathology The diagnosis of diffuse gliomas still rests primarily on H&E histopathology. A shared property of all of these tumors is the widespread infiltration by individual tumor cells through the CNS parenchyma. Infiltrative properties are suggested by MRI images, which generally show expansion of the involved brain with associated signal abnormalities that are hypointense on T1-weighted images and hyperintense on T2 and FLAIR. The neuroimaging features of grade II and III oligodendrogliomas and astrocytomas overlap significantly and a tissue diagnosis is required to establish tumor identity and grade. Of primary importance is the identification of a neoplastic population interspersed within the CNS neuropil indicating infiltrative properties. Once this pattern has been established, tumors are subclassified based on morphology as oligodendroglioma, astrocytoma, or mixed oligoastrocytoma and a grade is applied. Unfortunately, many studies have shown that this method lacks a high degree of reproducibility, especially for distinguishing oligodendroglial from astrocytic differentiation in non-classical or ambiguous gliomas. Even among experienced neuropathologists, the concordance has varied from 50% to 70%. The variability in the diagnostic criteria for oligodendroglioma, in particular, is unacceptably wide. Proper classification is clinically meaningful, since prognosis and therapy for these lesions are guided by diagnosis. Oligodendrogliomas generally have slower growth rates and are associated with a better prognosis than astrocytomas when compared grade for grade and the presence of an oligodendroglioma component within an infiltrative glioma usually predicts a longer survival. Moreover, specific chemotherapies have shown effectiveness against a subset of oligodendrogliomas that harbor chromosomal losses of 1p and 19q. Oligodendroglioma Oligodendrogliomas account for 4% of primary brain tumors and 10-20% of the infiltrating gliomas. In their original descriptions by Bailey and Cushing (1926) and later in the article "Oligodendrogliomas of the Brain" by Bailey and Bucey (1929), the authors described a tumor with nuclei that "are almost all perfectly round and of a fairly constant size" and are "surrounded by a ring of cytoplasm which stains very feebly", adding that they have a "network of fine capillaries" and "are prone to become calcified". Our current concept retains most of these features, yet there is an impression that the diagnostic criteria have expanded to gradually encompass non-conventional morphologies. Nonetheless, early articles also foreshadowed some of the diagnostic dilemmas that we continue to face: "There are also many cells which appear to be transitions between gigantic oligodendroglia and astrocytes. It is impossible definitely to classify them as belonging in either group" (Baily and Bucy, 1929). Current diagnosis of oligodendroglioma requires the identification of infiltrating glioma cells that have round, regular and monotonous nuclei, with little cell-to-cell variability. There are often well-defined cell membranes and cytoplasmic clearing that give rise to the classic "fried egg" cell (also called "honeycomb" and "woody plant" histology by Bailey and Bucy, 1929). Perinuclear clearing is helpful, but not requisite, for establishing the diagnosis of oligodendroglioma. More important are nuclei that are round, regular, and bland with delicate chromatin as compared to astrocytic nuclei that are hyperchromatic, elongate and irregular. "Oligodendroglia" means "glia with few processes". In tissue sections and especially on smear preparations, there is a paucity of glial processes emerging from oligodendrogliomas as compared to the long, finely fibrillar processes that are in abundance in cytologic and histologic preparations of astrocytic neoplasms. Other common but non-specific features include cortical involvement; a branching capillary network with a "chicken-wire" appearance; perinuclear satellitosis by tumor cells; microcalcifications; and microcysts filled with mucin. Grading systems have typically divided oligodendrogliomas into 2, 3, or 4 grades depending on cellularity, cytological atypia, mitotic activity, vascular proliferation, and necrosis. The current WHO Classification recognizes two grades: oligodendroglioma (grade II) and anaplastic oligodendroglioma (grade III). Grade II tumors vary from low to moderate cellularity, have a tendency to involve the cerebral cortex, and as they progress, they often grow as nodules. Nodular growth is compatible with a grade II lesion, but may represent a transition to a higher grade. Grade II oligodendrogliomas can show occasional mitotic figures and cytological atypia, but marked mitotic activity, microvascular proliferation, or necrosis is consistent with a WHO grade III, anaplastic oligodendroglioma. A recent investigation identified a) endothelial hypertrophy; b) necrosis; and c) >6 mitotic figures per high power field as significant univariate markers of poor outcome, providing a solid framework for establishing the diagnosis of anaplastic oligodendroglioma (grade III). Infiltrating "Diffuse" Astrocytomas The infiltrative or "diffuse" astrocytomas represent a spectrum ranging from low grade to highly malignant. The current World Health Organization (WHO) uses a 3-tiered scheme that spans from grade II to grade IV (GBM). These tumors are more common than oligodendrogliomas and account for one-third of all primary brain tumors and 70-75% of the diffuse gliomas. The diagnosis of infiltrative astrocytoma (WHO grade II) is applied when individual tumor cells showing astrocytic differentiation are seen invading CNS parenchyma with a low cell density. Astrocytic differentiation is best determined morphologically by the presence of nuclei that are elongate, hyperchromatic and irregular, having angulated and distorted contours. While oligodendrogliomas have nuclei like Florida oranges, those of astrocytomas are like Idaho potatoes. Cytologic preparations are extremely helpful for appreciating both the nuclear features and the fibrillarity of astrocytic neoplasms. There are numerous morphologic variants of astrocytoma, including granular cell, giant cell, gliosarcoma, gemistocytic, fibrillary, protoplasmic and small cell types. Grading schemes are the same among the morphologic variants. Mitotic activity has traditionally been used to separate grade II and grade III astrocytoma, with those tumors lacking mitoses classified as grade II tumors. Clinico-pathologic studies have demonstrated that astrocytomas with zero or one mitosis have similar clinical behaviors and therefore, current criteria of grade II astrocytoma allow 0 or 1 mitotic figures, but generally not more. Anaplastic astrocytoma (AA; WHO grade III) has higher cellularity, a greater degree of nuclear pleomorphism and atypia, and increased proliferation. Mitotic activity (greater that one mitosis) should be identified to apply the diagnosis of AA. Glioblastoma (GBM; WHO Grade IV) is the highest grade form of infiltrating astrocytoma. In addition to the histopathologic findings of AA, either microvascular hyperplasia or necrosis, often with pseudopalisading (or both), are required for the diagnosis of GBM. Oligoastrocytomas Oligoastrocytomas contain distinct regions of oligodendroglial and astrocytic differentiation and account for 5-10% of infiltrative gliomas. The two components may be completely separate or intermingled, with the two neoplastic cell types in close proximity. The minimal percentage of each component required for the diagnosis of a mixed glioma has been debated, resulting in highly variable diagnostic criteria and poor inter-observer reproducibility. Perhaps more problematic is the tendency for pathologists to "dump" diagnostically challenging infiltrating gliomas into the mixed oligoastrocytoma category. It should be emphasized that a mixed oligoastrocyoma is not equivalent to morphologically ambiguous glioma, but rather contains tumors cells with two distinct histologies. A recent study suggested that a single 100X field filled with an oligodendroglioma component could be used as a threshold for the diagnosis of mixed oligoastrocytoma, since this criterion identifies a subset with a better prognosis than astrocytoma and results in improved inter-observer concordance. Low grade oligoastrocytomas (WHO grade II) can contain occasional mitotic figures, low to moderate cellularity and mild to moderate cytological atypia, while anaplastic oligoastrocytoma (WHO grade III) should have histological features of anaplasia, which includes nuclear atypia, cellular pleomorphism, high cellularity, high mitotic activity, microvascular proliferation and necrosis. Frozen Section Diagnosis of Diffuse Gliomas The diagnosis of infiltrating gliomas at frozen section is challenging, since this preparation does not optimally reveal the histologic features that are typically used to separate oligodendrogliomas from astrocytomas. Perinuclear halos, delicate chromatin patterns, and nuclear regularity of oligodendrogliomas are not as evident in frozen sections. Ice crystal artifact can further degrade morphology. The diagnosis of "infiltrating glial neoplasm" together with a general degree of histologic differentiation (well-, moderately, or poorly differentiated) or grade (low, intermediate, or high) is the best that can be communicated in many, perhaps most, circumstances. The process of freezing brain tumor tissue during frozen section introduces artifacts that remain in permanent sections. Most notably, nuclei appear more hyperchromatic and atypical in previously frozen tissue; perinuclear halos of oligodendroglioma are not as evident; and the cytologic resolution is lower. These changes give an "astrocytic" appearance to oligodendrogliomas. It is always prudent to submit tissue for permanent sections that has not been previously frozen. Oligodendroglial and Astrocytic Markers Immunohistochemistry The search continues for reliable immunohistochemical and genetic markers of astrocytic and oligodendroglial differentiation. The intermediate filament glial fibrillary acidic protein (GFAP) is a marker of glial differentiation that is consistently expressed by resting astrocytes and greatly overexpressed in reactive astrocytosis. GFAP has been invaluable as a marker of glial differentiation in neoplasms involving the CNS. It is often wrongly assumed that GFAP can be used as a marker for astrocytomas to distinguish them from oligodendrogliomas. However, both immunohistochemical and ultrastructural studies have shown that oligodendrogliomas also express GFAP and therefore GFAP is not currently considered a useful marker for distinguishing among the infiltrating gliomas. Alterations of p53 are more common in astrocytomas than oligodendrogliomas, suggesting that it might be a discriminating marker. Inactivating point mutations of the TP53 gene occur in 50-60% of grade II astrocytomas but only in 5-10% of grade II oligodendrogliomas. TP53 mutations lead to the production of mutant p53 protein, which can be detected immunohistochemically in tumor cell nuclei. Staining for p53 can be diagnostically useful in some instances. For example, the presence of strong, nuclear p53 staining could indicate an underlying TP53 mutation and favor the diagnosis of astrocytoma. However, mechanisms other than TP53 mutation can cause p53 protein overexpression and p53 immunoreactivity does not correlate perfectly with the presence of TP53 mutations. A positive p53 immunostain requires interpretation in the context of other clinical and pathological data. A wide array of proteins expressed by non-neoplastic oligodendrocytes have been investigated for their utility as markers of oligodendrogliomas, including myelin basic protein (MBP), proteolipid protein (PLP), NG2, myelin associated glycoprotein (MAG), galactocerebroside (GC), Leu7, cyclic nucleotide-3'-phosphatase (CNP), and proteins encoded by the oligodendrocyte lineage (Olig) genes, Olig1, Olig2, and Olig3. Thus far these proteins have not shown specificity or diagnostic utility. Genetic Markers We now know that each histologic category of infiltrating glioma contains multiple distinct molecular genetic subsets and these findings have been exploited to assist with diagnosis, prognosis, and directing therapy. The most frequent genetic tests employed on tumor tissue are loss of heterozygosity (LOH), fluorescence in-situ hybridization (FISH) and comparative genomic hybridization (CGH). These tests demonstrate excellent concordance (73-99%) and the choice depends largely on the preferences of the pathologist, department, and institution. FISH has some advantages from a pathologist's perspective: 1) analysis is based on the morphologic identification of genetic alterations within tumor cell nuclei; 2) non-neoplastic cells (positive controls) are almost always present within the tissue sections examined (i.e. normal endothelial cells, neurons, etc); 3) FISH does not require microdissection, and 4) genetic gains and losses in infiltrative tumors with a low ratio of neoplastic/normal cells can be analyzed by FISH, whereas these alterations may not be detected by PCR-based analysis (LOH studies) due to overwhelming amounts of normal DNA. One major disadvantage of FISH is that it can be highly labor intensive and automation has not yet reached all of its applications. Some genetic alterations occur in both astrocytic and oligodendroglial tumors, generally with increasing frequency at higher grades [e.g., loss of 9p21 (p16/CDKN2A) or losses on chromosome 10 (PTEN/DMBT1)] and are therefore not useful markers for discriminating histologic subtypes. 1p/19q One of the best recognized molecular signatures among the diffuse gliomas, which also has prognostic significance, is the combined loss of genetic material from chromosomes 1p and 19q in oligodendrogliomas. Losses on both 1p and 19q almost always coexist within oligodendrogliomas and can be detected in 60-80%. In most, the entire arms of the affected chromosome are deleted, allowing reliable detection by LOH, FISH, or CGH. Combined 1p/19q l9 loss is strong predictor of response to chemotherapy and overall survival in both low and high grade oligodendrogliomas, while those with deletions of p16/CDKN2A (9p21), LOH 10q, or EGFR amplifications have poor survival. Initial studies showed that 1p/19q deletion conferred increased responsiveness to "PCV" chemotherapy (procarbazine, CCNU, vincristine), but it is now evident that this subset is more responsive to other therapies including radiation and temozolamide. The combined loss of 1p/19q occurs much less frequently in astrocytic tumors (only in 1-10%). Additional studies will be needed to demonstrate the clinical utility, if any, for 1p/19q testing in other primary brain tumors. EGFR Amplifications of the epidermal growth factor receptor (EGFR) gene occur in approximately 40% of GBMs and can be detected by FISH, CGH, or PCR-based tests. Amplifications are much less frequent in lower grade astrocytomas and are therefore considered a late genetic event in tumor progression. Either wild type or mutated forms of EGFR can be amplified. The most common EGFR amplification is a mutated form lacking exons 2-7, which results in a truncated cell surface protein with constitutive tyrosine kinase activity (EGFRvIII). Therapies directed at the overexpressed EGFR in GBMs are finding there way into neuro-oncology practice. EGFR amplifications are rare in oligodendroglial tumors and analysis of EGFR status has proven useful for distinguishing high grade astrocytomas from anaplastic oligodendrogliomas in some instances. For the majority of GBMs, a correct diagnosis can be made based on morphology alone and EGFR testing is not necessary. However, the recently described "small cell astrocytoma" is a high grade, biologically aggressive astrocytoma that has a great deal of morphologic similarity to anaplastic oligodendroglioma and may require ancillary tests to correctly diagnose. Small cell astrocytomas are characterized by a high frequency of EGFR amplification and chromosome 10 losses, but have intact chromosome 1p and 19q. In contrast, anaplastic oligodendrogliomas show the opposite genetic alterations, having a high frequency of 1p/19q deletions but only rare EGFR amplifications and chromosome 10 losses. Chromosome 7 Alterations on chromosome 7 are common in astrocytomas and amplification of EGFR at 7p12 is the best known example. While frequent in GBM, EGFR amplifications are less common in grade II and III astrocytomas, limiting its utility as a marker for distinguishing lower grade gliomas. Among grade II-III astrocytomas, gains of chromosome 7 are the most frequent genetic alterations (40-66%), occurring either as entire chromosome gains (trisomy/polysomy 7) or as gains of 7p or 7q alone. These gains are already present in 40-50% of grade II astrocytomas, suggesting that it is an early genetic event, and its presence is associated with shorter survivals. Chromosome 7 gains are less common in oligodendrogliomas, occurring as the whole chromosome in roughly 10% and as gains of either 7p or 7q in 20%. When present, they occur in oligodendrogliomas that also have chromosome 10 losses, but not 1p/19q losses, indicating that these occur in a biologically distinct class of tumors, perhaps more related to astrocytoma. The detection of chromosome 7 gains may occasionally be helpful in distinguishing low grade astrocytomas from oligodendrogliomas, especially in the absence of 1p/19 losses. Summary Even with ancillary tests, the diagnosis of diffuse gliomas remains a challenge, and the search for more specific and diagnostically useful markers continues. Emerging technologies, like DNA, cDNA or protein microarrays, are demonstrating potential for brain tumor classification and for marker discovery. For example, cDNA microarray analysis coupled with computational algorithms have been used to classify high grade gliomas into histologic groups. Moreover, this technique was found to be superior to histologic classification in predicting the prognosis of morphologically ambiguous tumors. Diagnostic microarrays containing a limited number of relevant genes could be designed to classify tumors, predict their clinical behavior, and rank therapeutic options. Such approaches have already uncovered new markers with potential for diagnostic use that await validation. The development of therapies targeted at the molecular underpinnings of disease will ultimately require a molecular component to the diagnosis. Suggested Readings Bailey P, Cushing H (1926) Tumors of the glioma group. Philadelphia: Lippincott. Bailey P, Bucy PC (1929) Oligodendrogliomas of the brain. J Pathol Bacteriol 32: 735-751. Burger PC, Scheithauer BW, Vogel FS. Surgical Pathology of the Nervous System and its Coverings, 4th ed., New York: Churchill Livingston, 2002. Coons SW, Johnson PC, Scheithauer BW, Yates AJ, Pearl DK. Improving diagnostic accuracy and interobserver concordance in the classification and grading of primary gliomas. Cancer. 1997;79:1381-93. Fuller CE, Schmidt RE, Roth KA, Burger PC, Scheithauer BW, Banerjee R, Trinkaus K, Lytle R, Perry A. Clinical utility of fluorescence in situ hybridization (FISH) in morphologically ambiguous gliomas with hybrid oligodendroglial/astrocytic features. J Neuropathol Exp Neurol. 2003;62:1118-28. Giannini C, Scheithauer BW, Weaver AL, Burger PC, Kros JM, Mork S, Graeber MB, Bauserman S, Buckner JC, Burton J, Riepe R, Tazelaar HD, Nascimento AG, Crotty T, Keeney GL, Pernicone P, Altermatt H. Oligodendrogliomas: reproducibility and prognostic value of histologic diagnosis and grading. J Neuropathol Exp Neurol. 2001;60:248-62. Gupta M, Djalilvand A, Brat DJ. Clarifying the diffuse gliomas: an update on morphology and markers that discriminate oligodendroglioma from astrocytoma. Am. J Clin Pathol 2005; 124:1-14. Hartmann C, Mueller W, von Deimling A. Pathology and molecular genetics of oligodendroglial tumors. J Mol Med. 2004;82:638-55. Kleihuis P, Cavenee, WK. Pathology and Genetics of Tumours of the Nervous System. 2nd ed., Lyon : Intl. Agency for Research, 2000. Nutt CL, Mani DR, Betensky RA, Tamayo P, Cairncross JG, Ladd C, Pohl U, Hartmann C, McLaughlin ME, Batchelor TT, Black PM, von Deimling A, Pomeroy SL, Golub TR, Louis DN. Gene expression-based classification of malignant gliomas correlates better with survival than histological classification. Cancer Res. 2003;63:1602-7. Perry A, Aldape KD, George DH, Burger PC. Small cell astrocytoma: an aggressive variant that is clinicopathologically and genetically distinct from anaplastic oligodendroglioma. Cancer. 2004;101:2318-26. Reifenberger G, Collins VP. Pathology and molecular genetics of astrocytic gliomas.J Mol Med. 2004;82:656-70. Smith JS, Perry A, Borell TJ, Lee HK, O'Fallon J, Hosek SM, Kimmel D, Yates A, Burger PC, Scheithauer BW, Jenkins RB. Alterations of chromosome arms 1p and 19q as predictors of survival in oligodendrogliomas, astrocytomas, and mixed oligoastrocytomas.J Clin Oncol. 2000;18:636-45. Smith JS, Alderete B, Minn Y, Borell TJ, Perry A, Mohapatra G, Hosek SM, Kimmel D, O'Fallon J, Yates A, Feuerstein BG, Burger PC, Scheithauer BW, Jenkins RB. Localization of common deletion regions on 1p and 19q in human gliomas and their association with histological subtype. Oncogene. 1999;18:4144-52.