Final Diagnosis -- Glioblastoma multiforme

DIAGNOSIS    Glioblastoma multiforme.

Brain neoplasms are the most common childhood cancer after leukemia, accounting for approximately 25% of all pediatric cancers (5). Both the histological type and location of pediatric brain tumors differ from those occurring in the adult population. Low-grade astrocytomas and medulloblastomas are the most frequently encountered in children, whereas anaplastic astrocytomas and glioblastoma multiforme (GBM) are more common in adults and are extremely rare in children (6,8).

In our case the radiological differential diagnosis included medulloblastoma, ependymoma and low grade astrocytoma. The constellation of CT and MRI findings suggested medulloblastoma as the most likely diagnosis. Atypical teratoid/rhabdoid tumor (AT/RT) should also be included in the differential diagnosis but is usually encountered earlier in life.

Cerebellar GBMs are extremely rare, especially in children, with only occasional reports (1,3,4). From a neuropathological perspective, it is very important that the tumor fulfill the diagnostic criteria for GBM (significant pleomorphism, hypercellularity, mitoses and especially necrosis with palisading and glomeruloid type vascular hyperplasia) (2). On histological examination primitive neuroectodermal tumors (PNETs) and medulloblastomas that are common tumors in the posterior fossa in children, may also exhibit pseudopalisading necrosis. Nevertheless, PNETs and medulloblastomas are usually positive for synaptophysin by immunohistochemistry. In our case there was no expression of neuronal antigens (synaptophysin, Neu-N and neurofilaments). Furthermore, there was no convincing evidence of neuronal or neuroblastic differentiation, such as Homer Wright rosettes. AT/RT, a tumor that also shares the morphologic features of both medulloblastoma and PNET with epithelial, primitive neuroepithelial and mesenchymal differentiation, was included in the differential diagnosis. However, AT/RT tumors lack INI1 immunohistochemical expression in the nucleus and our case showed positive INI1 staining. Furthermore, anaplastic oligodendrogliomas may also exhibit pseudopalisading necrosis but they are relatively uncommon and the increased pleomorphism was also more compatible with glioblastoma.

Regarding the management of cerebellar lesions, a gross-total resection, if feasible, would be the appropriate treatment in all cases. Post-operative radiation in children older than 3 years is no longer in question; given that in few cases treated without radiation survival was less than a month. The question is whether the radiation therapy should be delivered to the tumor bed alone or the entire neuraxis. The radiation of entire neuraxis has now been established as the treatment of choice, given that the incidence of CSF dissemination in patients who did not received craniospinal irradiation was high and with elective neuraxis irradiation there was nearly an elimination of this complication (1,4,5,7). Chemotherapy may hold a role in pediatric cerebellar GBM but its usefulness has not been well studied due to the limited number of patients. However, it definitely offers a reasonable treatment available to very young children in whom the effect of brain radiation would be devastating. The same treatment options are offered for high-grade diffuse pontine gliomas, however in such cases a gross total resection is unfeasible. The prognosis of cerebellar GBMs is poorer even compared to supratentorial GBMs. In a series of 14 patients with cerebellar GBM, 11 died after a mean survival time of 9.9 months whereas the remaining patients were alive with a mean follow-up period of 24 months (7).

In conclusion, GBM should be kept in mind in the differential diagnosis of the lesions in the posterior fossa. A gross total resection should be always attempted, where possible, in order a better outcome to be achieved. Regarding further treatment, despite numerous advances and adjuvant chemotherapy, radiotherapy continues to be the most effective therapy for children with high-grade gliomas.


  1. Campbell JW, Pollack IF (1996) Cerebellar astrocytomas in children. J Neurooncol 28:223-231.
  2. Chamberlain MC, Silver P, Levin VA (1990) Poorly differentiated gliomas of the cerebellum. A study of 18 patients. Cancer 65:337-340.
  3. Georges PM, Noterman J, Flament-Durand J (1983) Glioblastoma of the cerebellum in children and adolescents. Case report and review of the literature. J Neurooncol 1:275-278.
  4. Kopelson G (1982) Cerebellar glioblastoma. Cancer 50:308-311.
  5. Salazar OM (1981) Primary malignant cerebellar astrocytomas in children: a signal for postoperative craniospinal irradiation. Int J Radiat Oncol Biol Phys 7:1661-1665.
  6. Sanders RP, Kocak M, Burger PC, Merchant TE, Gajjar A, Broniscer A (2007) High-grade astrocytoma in very young children. Pediatr Blood Cancer 6: [Epub ahead of print]
  7. Ushio Y, Arita N, Yoshimine T, Ikeda T, Mogami H (1987) Malignant recurrence of childhood cerebellar astrocytoma: case report. Neurosurgery 21:251-255.
  8. Viano JC, Herrera EJ, Suarez JC (2001) Cerebellar astrocytomas: a 24-year experience. Childs Nerv Syst 17:607-610.

Contributed by Evriviadis Mpairamidis, MD, George A Alexiou, MD, Kalliopi Stefanaki, MD, PhD, Manolakos Ilias, MD, Sfakianos George, MD, PhD and Prodromou Neofytos, MD, PhD

Case IndexCME Case StudiesFeedbackHome