• brain stem;
  • cerebellum;
  • differential diagnosis;
  • loss of heterozygosity;
  • oligodendroglioma


  1. Top of page
  2. Abstract
  3. Introduction
  4. Clinical Presentation
  5. Pathological Findings
  6. Discussion
  7. References

With respect to localization, oligodendrogliomas are characterized by a marked preponderance of the cerebral hemispheres. Outside these typical sites, any tumor histopathologically reminiscent of oligodendroglioma a priori is likely to represent one of its morphological mimics, including clear cell ependymoma, neurocytoma, pilocytic astrocytoma or glioneuronal tumors. This is particularly relevant as several of the latter are in principle curable by surgery. Among extrahemispherical sites, bona fide oligodendroglioma – as characterized by loss of heterozygosity (LOH) of chromosome arms 1p and 19q – so far has not been documented to occur in the brain stem. Here, we report the case of a 55-year-old female patient with an anaplastic oligodendroglioma (WHO grade III) of the brain stem and cerebellum diagnosed by stereotactic biopsy and featuring combined LOH of 1p and 19q. A morphological peculiarity was a population of interspersed tumor giant cells, a phenomenon that has been referred to as polymorphous oligodendroglioma. Our findings confirm the notion that – although very infrequently – true oligodendrogliomas do occur in the infratentorial compartment.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Clinical Presentation
  5. Pathological Findings
  6. Discussion
  7. References

The various types of glial tumors show diverse patterns of spatial distribution within the CNS. The biological basis of this phenomenon is not well understood in most of these entities. Among them, oligodendroglioma is characterized by a marked tropism for the cerebral hemispheres.[1] Outside this typical location, any tumor morphologically reminiscent of oligodendroglioma will require particular consideration of tumor types able to histologically mimic oligodendroglioma by virtue of monotonous round nuclei and perivascular haloes. Intriguingly, reports of infratentorial oligodendrogliomas have remained rare[2-13] even though diagnosis of oligodendroglioma arguably has become increasingly more permissive over the past years.[14] To our knowledge, only two cases of cerebellar oligodendrogliomas with loss of heterozygosity (LOH) of 1p and 19q – the genetical hallmark of oligodendrogliomas – have recently been described in the literature,[15, 16] while we are not aware of any report of genetically confirmed oligodendroglioma of the brain stem. Conversely, the vast majority of reports on infratentorial oligodendrogliomas antedate the era of diagnostic testing for LOH 1p19q or show no or non-specific genetic alterations.[2-13]

The differential diagnosis of infratentorial oligodendroglioma includes in particular pilocytic astrocytoma,[17] clear cell ependymoma,[18, 19] central or extraventricular neurocytoma,[20, 21] clear cell meningioma, and glioneuronal tumors. Of the latter, rosette-forming glioneuronal tumor (RGNT) or cerebellar liponeurocytoma are not only prone to imitate oligodendroglioma, but have their sites of predilection in the infratentorial compartment.[22, 23]

Accurate classification is of particular relevance to clinical management as – in contrast to oligodendroglioma – several of its mimics are in principle curable by surgery. Conversely, in the setting of a high-grade neoplasm, certain tumors such as embryonal neoplasms or small cell glioblastoma, when mistaken for anaplastic oligodendroglioma, may receive less aggressive adjuvant therapy than required.

Clinical Presentation

  1. Top of page
  2. Abstract
  3. Introduction
  4. Clinical Presentation
  5. Pathological Findings
  6. Discussion
  7. References

A 55-year-old woman presented with generalized weakness, hypesthesia of all four extremities and dysphagia. MRI showed a contrast-enhancing space-occupying lesion in the medulla oblongata extending to both cerebellar hemispheres (Fig. 1). When the cerebellar component was found to be progressive over the next 3 weeks, a stereotactic biopsy of the right cerebellar part of the mass was performed.


Figure 1. At disease presentation, MRI revealed a mass encompassing the entire diameter of the medulla oblongata and expanding it.

Download figure to PowerPoint

Following the stereotactic biopsy, the patient received combined radiation (48 Gy) and concomitant chemotherapy with temozolomide which had to be discontinued after 4 weeks because of fever. The patient was referred to a palliative care unit.

Pathological Findings

  1. Top of page
  2. Abstract
  3. Introduction
  4. Clinical Presentation
  5. Pathological Findings
  6. Discussion
  7. References

A total of six stereotactic biopsies, including one for intraoperative frozen section, were received for histopathological analysis. Biopsy cores showed dense infiltration by a moderately to highly cellular neoplasm consisting of two distinct populations of tumor cells. The majority of tumor cells were characterized by monotonous round, medium-sized nuclei with perinuclear haloes and no discernible cytoplasm or processes. Diffusely interspersed among these, a second minor population of giant tumor cells with often multiple bizarre, atypical nuclei was present (Fig. 2). Mitotic rate varied between six and 10 mitotic figures in most series of 10 consecutive high-power fields. They were distributed in a diffuse fashion rather than being associated with proliferative nodules. Intratumoral vessels were mostly capillary-sized and of non-specific arrangement rather than forming a typical “chicken-wire” lattice. There was incipient endothelial hyperplasia, but no florid microvascular proliferation. Necrosis was absent. There was no morphological evidence of ganglion cell or neurocytic differentiation, nor desmoplastic reaction or sarcomatous transformation. No histopathological features of ependymal differentiation such as ependymal channels, true rosettes or pseudo-rosettes were identified. The tumor did not feature Rosenthal fibers, eosinophilic granular bodies, calcifications or secondary structures of Scherer.


Figure 2. HE-stained histological slides showed a moderately to highly cellular, diffusely infiltrative neoplasm (A). The main tumor cell population consisted of cells with no discernible individual processes with round, monotonous nuclei and perinuclear haloes. Mitotic figures (arrowheads) were conspicuous (B). Giant cells (arrow) represented a second minor tumor cell population (C).

Download figure to PowerPoint

An immunohistochemical staining for GFAP showed exquisite sparing of the tumor cells while resident reactive astrocytes were GFAP immunoreactive (Fig. 3). No immunoreactivity for epithelial membrane antigen (EMA) was present, in particular no dot- or ring-like perinuclear staining was observed. Immunohistochemical stainings for synaptophysin and neurofilament showed positivity in a pattern attributable to residual parenchyma rather than neoplastic cells. There was no immunoreactivity of tumor cells (including the tumor giant cells) for neuronal nuclear antigen (NeuN). Proliferation rate – as assessed by immunohistochemical staining for MiB-1/Ki67 – ranged between 10% and 20% and marked proliferative activity was also present in the population of bizarre tumor giant cells. Nuclear p53 accumulation was not present. There was no immunoreactivity for R132H-mutated isocitrate dehydrogenase 1.


Figure 3. Immunohistochemical stainings for GFAP, neurofilament protein (NF) revealed sparing of a majority of tumor cells, In particular, the tumor giant cells (arrows) did not show neuronal differentiation. Proliferative activity (MIB-1) was high in both tumor cell populations.

Download figure to PowerPoint

Microsatellite analysis showed LOH of both 1p and 19q (i.e. 1p/19q co-deletion) as compared to genomic DNA isolated from peripheral blood (Fig. 4)[25] for all informative markers. Methylation-specific PCR was negative for O6-methylguanine methyltransferase (MGMT) promoter methylation (not shown).[26]


Figure 4. Microsatellite analysis shows relative loss of one allele each (arrowheads) as compared to the corresponding reference genotype (upper row) for all informative markers. The peaks corresponding to the respective alleles are indicated by the boxes below each graph. The relative reduction in signal (defined as area under the curve; AUC) ranges from 45% (D1S468, D19S219, D19S412) to 52% (D1S214).[24] The analysis was performed in triplicates and signal reductions are statistically significant (D1S468: P = 0.0003; D1S214: P = 0.024; D19S219: P = 0.001; D19S412: P < 0.0001; Student's t-test, two-tailed).

Download figure to PowerPoint

Upon these findings a diagnosis of anaplastic oligodendroglioma (WHO grade III) was issued.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Clinical Presentation
  5. Pathological Findings
  6. Discussion
  7. References

In many instances, attempts to define a pathological entity solely based on histomorphology fail to result in a consistent association with clinical features or may not provide relevant predictions regarding disease course or response to therapy. In this setting, the integration of additional information such as clinical features or the presence of recurrent genetic alterations may be required to allow for a clinically meaningful attribution of a given tumor to an entity.

Along this line of thought, the a priori likelihood is high for any tumor consisting of monotonous round cells with prominent perinuclear haloes to represent oligodendroglioma when occurring in an appropriate clinical context, for example in the cerebral hemispheres in a young or middle-aged adult.[1]

An identical or very similar histomorphology in a different clinical setting will rightfully evoke different diagnostic considerations. When observed in an infratentorial location, odds are more in favor of one the oligodendroglioma mimics such as pilocytic astrocytoma, clear cell ependymoma or glioneuronal tumors with oligodendroglioma-like morphology, which are well-documented to regularly occur in – or even a predilection site of which is – the infratentorial compartment.[27-29] Therefore any diagnosis of infratentorial oligodendroglioma may be seriously challenged.

Beyond conventional histology, immunophenotyping provides only limited support in differentiating oligodendroglioma from its morphological mimics.[30] Dot- or ring-like positivity for EMA – when extensive – will strongly speak in favor of clear cell ependymoma.[27] Immunohistochemical detection of R132H-mutated IDH1 protein as a surrogate of the underlying point mutation in the IDH1 gene will allow for attribution to the group of diffusely infiltrating gliomas (see below). Otherwise, immunohistochemistry will rarely discriminate in a reliable fashion between the lines of differentiation of the above-mentioned differential diagnoses. Parenthetically, also electron microscopy will only infrequently contribute in this setting beyond demonstrating evidence of ependymal differentiation.[31]

In contrast, in a given case molecular genetic analysis has the potential to clarify the differential diagnosis between the variations on the theme of “monotonous round cells with perinuclear clearing”. Classically, oligodendroglioma is associated with LOH 1p/19q, which in the appropriate histopathological setting is considered highly specific of oligodendroglial differentiation,[32] even though its sensitivity may be reduced in certain sites such as in temporal lobe tumors and it is not a consistent feature of pediatric oligodendrogliomas.[1]

Furthermore, point mutations in the IDH1 and IDH2 genes encoding isocitrate dehydrogenases (IDH) 1 and 2, respectively, among tumors of the CNS are largely restricted to the group of diffusely infiltrating gliomas, including oligodendrogliomas. Therefore, demonstration of an IDH1 or IDH2 mutation can serve to rule out the above-mentioned differential diagnoses of oligodendroglioma.[33] Conversely, BRAF codon 600 mutations or the presence of a KIAA1549-BRAF fusion product are characteristic of certain non-infiltrative glial or glioneuronal tumor types, including pilocytic astrocytoma and ganglioglioma. Their presence would argue against a diagnosis of oligodendroglioma according to the current understanding.[29] Finally, recurrent PIK3CA mutations have recently been described in RGNT (WHO grade I) of the fourth ventricle. At present, it is unknown if these – among glial and glioneuronal tumors – are specific for RGNT.[28]

However, this issue is complicated by recent evidence that phenotypic boundaries in the above setting may indeed not be always clear-cut. In fact, immunostaining for synaptophysin had been instrumental in the historical emancipation of neurocytoma from oligodendroglioma; however, subsequent studies have established neurocytic differentiation in genotypically unambiguous oligodendrogliomas.[34] To make this theme of overlapping phenotypes come full circle, mention must be made of a rare instance of ependymal differentiation within a neurocytoma, located in the cerebellum.[35]

The caveat mentioned above may apply to most of the cases in the literature reported as infratentorial oligodendrogliomas. Most published cases date to the era before LOH 1p/19q became a mainstay of diagnostic work-up of oligodendrogliomas.[12, 13] Of the more recent reports, most either do not include molecular studies at all or demonstrate no or non-specific genetic alterations.[2-11] A subset of these cases may represent examples of the emerging entity of leptomeningeal oligodendroglioma-like tumor of childhood.[3, 36, 37] While in a given – especially pediatric – case, lack of LOH 1p/19q does not exclude a diagnosis of oligodendroglioma,[38, 39] the concept of infratentorial oligodendroglioma in general might be questioned, if LOH 1p/19q were not identified in this group at all or in an inadequately low proportion of tumors.

When taking together the very typical cytomorphology of the predominant tumor cell population, the immunophenotype, which is consistent with – even though not specific for – oligodendroglioma, and the molecular genetics findings, in the authors' opinion, the oligodendroglial differentiation of the tumor reported here is established beyond reasonable doubt. Parenthetically, CSF dissemination to the brain stem and cerebellum as opposed to primary infratentorial oligodendroglioma was excluded by MRI.

In addition to the unusual localization, the interspersed bizarre tumor giant cells represent another feature uncommon for oligodendroglioma in the case presented here. Neither morphologically, nor by immunophenotype, these giant tumor cells appeared to represent ganglion cell differentiation, which is a rare but well-documented event in oligodendrogliomas.[40] Rather, this phenomenon may correspond to what has been reported as polymorphous type of oligodendroglioma in the pre-molecular era.[41] More specifically, the term polymorphous oligodendroglioma has been considered synonymous with oligodendroglioma containing bizarre giant cells.[42] The authors of the present report have recently found that occasional tumors otherwise resembling oligodendroglioma but characterized by the presence of bizarre tumor giant cells do indeed feature combined LOH 1p/19q, indicating that they represent a morphological variant of true oligodendrogliomas.[43]

Our finding confirms the perception that – although exceedingly rare – bona fide oligodendrogliomas do occur in the infratentorial compartment. Of note, both the present and one of the other two genetically confirmed cases of infratentorial oligodendroglioma[15] displayed increased mitotic activity, which in the present case qualified for grading as anaplastic oligodendroglioma (WHO grade III). Furthermore, both cases showed an aggressive clinical course, which is in contrast with the relatively favorable prognosis of hemispherical oligodendrogliomas.[39]


  1. Top of page
  2. Abstract
  3. Introduction
  4. Clinical Presentation
  5. Pathological Findings
  6. Discussion
  7. References
  • 1
    Aldape K, Burger PC, Perry A. Clinicopathologic aspects of 1p/19q loss and the diagnosis of oligodendroglioma. Arch Pathol Lab Med 2007; 131: 242251.
  • 2
    Mohindra S, Savardekar A, Bal A. Pediatric brainstem oligodendroglioma. J Neurosci Rural Pract 2012; 3: 5254.
  • 3
    Mittelbronn M, Wolff M, Bültmann E et al. Disseminating anaplastic brainstem oligodendroglioma associated with allelic loss in the tumor suppressor candidate region D19S246 of chromosome 19 mimicking an inflammatory central nervous system disease in a 9-year-old boy. Hum Pathol 2005; 36: 854857.
  • 4
    Chitkara N, Chanda R, Thakur AK, Chanda S, Sharma NK. Posterior fossa oligodendroglioma. Indian J Pediatr 2002; 69: 10991100.
  • 5
    Nam DH, Cho BK, Kim YM, Chi JG, Wang KC. Intramedullary anaplastic oligodendroglioma in a child. Childs Nerv Syst 1998; 14: 127130.
  • 6
    Alvarez JA, Cohen ML, Hlavin ML. Primary intrinsic brainstem oligodendroglioma in an adult. Case report and review of the literature. J Neurosurg 1996; 85: 11651169.
  • 7
    de Chadarevian JP, Legido A, Halligan GE et al. Cerebellar gliomatosis in a toddler: case report of a challenging condition and review of the literature. J Child Neurol 2011; 27: 511520.
  • 8
    Nadkarni TD, Menon RK, Desai KI, Goel AH. Cerebellar oligodendroglioma in a young adult. J Clin Neurosci 2005; 12: 837838.
  • 9
    Shah S, Schelper RL, Krishnamurthy S, Chang HT. An 11-year-old boy with an incidental mass in the left lateral cerebellum. Neuropathology 2007; 27: 9597.
  • 10
    El Ouni F, Gaha M, Moulahi H, Daadoucha A, Krifa H, Tlili K. Infratentorial oligodendroglioma in a child: case report and review of the literature. Turk Neurosurg 2010; 22: 461464.
  • 11
    Lee IH, Kim ST, Suh Y-L et al. Infratentorial oligodendrogliomas: imaging findings in six patients. Acta Radiol 2010; 51: 213217.
  • 12
    Koeppen AH, Cassidy RJ. Oligodendroglioma of the medulla oblongata in a neonate. Arch Neurol 1981; 38: 520523.
  • 13
    Packer RJ, Sutton LN, Rorke LB et al. Oligodendroglioma of the posterior fossa in childhood. Cancer 1985; 56: 195199.
  • 14
    Burger PC. What is an oligodendroglioma? Brain Pathol 2002; 12: 257259.
  • 15
    Gru AA, Fulling K, Perry A. A 39 year-old man with a cerebellar mass and pancytopenia. Brain Pathol 2012; 22: 251254.
  • 16
    Furtado SV, Venkatesh PK, Ghosal N, Murthy GK, Hegde AS. Clinical and radiological features of pediatric cerebellar anaplastic oligodendrogliomas. Indian J Pediatr 2011; 78: 880883.
  • 17
    Utsuki S, Oka H, Kijima C, Yasui Y, Fujii K, Kawano N. Pilocytic astrocytoma with abundant oligodendroglioma-like component. Brain Tumor Pathol 2012; 29: 103106.
  • 18
    Katoh M, Satoh T, Nishiya M et al. Clear cell ependymoma of the fourth ventricle. Neuropathology 2004; 24: 330335.
  • 19
    Kim JH, Cho BK, Kim IO, Park SH. Clear-cell ependymoma of the cerebellum: a case report. Ultrastruct Pathol 2007; 31: 241247.
  • 20
    Jain D, Roy S, Chopra S. Atypical central neurocytoma of fourth ventricle with hemorrhagic complication during surgery in a child. Clin Neurol Neurosurg 2012; 114: 182184.
  • 21
    Makhdoomi R, Malik NK, Wani A, Bhat S, Baba K. Extraventricular neurocytoma of the vermis in a child. J Clin Neurosci 2010; 17: 14691471.
  • 22
    Vajtai I, Arnold M, Kappeler A et al. Rosette-forming glioneuronal tumor of the fourth ventricle: report of two cases with a differential diagnostic overview. Pathol Res Pract 2007; 203: 613619.
  • 23
    George DH, Scheithauer BW. Central liponeurocytoma. Am J Surg Pathol 2001; 25: 15511555.
  • 24
    Campbell BA, Horsman DE, Maguire J et al. Chromosomal alterations in oligodendroglial tumours over multiple surgeries: is tumour progression associated with change in 1p/19q status? J Neurooncol 2008; 89: 3745.
  • 25
    Vajtai I, Vassella E, Hewer E, Kappeler A, Reinert MM. Sarcomatous evolution of oligodendroglioma (“oligosarcoma”): confirmatory report of an uncommon pattern of malignant progression in oligodendroglial tumors. Pathol Res Pract 2012; 208: 750755.
  • 26
    Vassella E, Vajtai I, Bandi N, Arnold M, Kocher V, Mariani L. Primer extension based quantitative polymerase chain reaction reveals consistent differences in the methylation status of the MGMT promoter in diffusely infiltrating gliomas (WHO grade II–IV) of adults. J Neurooncol 2010; 104: 293303.
  • 27
    Fouladi M, Helton K, Dalton J et al. Clear cell ependymoma: a clinicopathologic and radiographic analysis of 10 patients. Cancer 2003; 98: 22322244.
  • 28
    Ellezam B, Theeler BJ, Luthra R, Adesina AM, Aldape KD, Gilbert MR. Recurrent PIK3CA mutations in rosette-forming glioneuronal tumor. Acta Neuropathol 2011; 123: 285287.
  • 29
    Korshunov A, Meyer J, Capper D et al. Combined molecular analysis of BRAF and IDH1 distinguishes pilocytic astrocytoma from diffuse astrocytoma. Acta Neuropathol 2009; 118: 401405.
  • 30
    Ligon KL, Alberta JA, Kho AT et al. The oligodendroglial lineage marker OLIG2 is universally expressed in diffuse gliomas. J Neuropathol Exp Neurol 2004; 63: 499509.
  • 31
    Mrak RE. The Big Eye in the 21st century: the role of electron microscopy in modern diagnostic neuropathology. J Neuropathol Exp Neurol 2002; 61: 10271039.
  • 32
    Capper D, Reuss D, Schittenhelm J et al. Mutation-specific IDH1 antibody differentiates oligodendrogliomas and oligoastrocytomas from other brain tumors with oligodendroglioma-like morphology. Acta Neuropathol 2010; 121: 241252.
  • 33
    Capper D, Weissert S, Balss J et al. Characterization of R132H mutation-specific IDH1 antibody binding in brain tumors. Brain Pathol 2010; 20: 245254.
  • 34
    Vajtai I, Arnold M, Vassella E. Oligodendroglioma with neurocytic differentiation and characteristic loss of heterozygosity on chromosomes 1p and 19q. Acta Neuropathol 2005; 110: 520522.
  • 35
    Jouvet A, Lellouch-Tubiana A, Boddaert N, Zerah M, Champier J, Fevre-Montange M. Fourth ventricle neurocytoma with lipomatous and ependymal differentiation. Acta Neuropathol 2005; 109: 346351.
  • 36
    Rodriguez FJ, Perry A, Rosenblum MK et al. Disseminated oligodendroglial-like leptomeningeal tumor of childhood: a distinctive clinicopathologic entity. Acta Neuropathol 2012; 124: 627641.
  • 37
    Agamanolis DP, Katsetos CD, Klonk CJ et al. An unusual form of superficially disseminated glioma in children: report of 3 cases. J Child Neurol 2012; 27: 727733.
  • 38
    Kreiger PA, Okada Y, Simon S, Rorke LB, Louis DN, Golden JA. Losses of chromosomes 1p and 19q are rare in pediatric oligodendrogliomas. Acta Neuropathol 2005; 109: 387392.
  • 39
    McDonald JM, See SJ, Tremont IW et al. The prognostic impact of histology and 1p/19q status in anaplastic oligodendroglial tumors. Cancer 2005; 104: 14681477.
  • 40
    Perry A, Burton SS, Fuller GN et al. Oligodendroglial neoplasms with ganglioglioma-like maturation: a diagnostic pitfall. Acta Neuropathol 2010; 120: 237252.
  • 41
    Zülch KJ. On the definition of the polymorphous oligodendroglioma. Acta Neurochir (Wien) 1964; 11 (Suppl 10): S166172.
  • 42
    Kros JM, Lie ST, Stefanko SZ. Familial occurrence of polymorphous oligodendroglioma. Neurosurgery 1994; 34: 732736.
  • 43
    Hewer E, Beck J, Murek M, Vassella E, Vajtai I. Polymorphous oligodendroglioma of Zülch revisited – a genetically diverse group of anaplastic gliomas including tumors of bona fide oligodendroglial differentiaton. Clin Neuropathol 2012; 31: P017. (Abstract).