Kathy Warren, MD, pediatric oncologist, National Cancer Institute

INTRODUCTION: Although ependymomas are the third most common type of brain tumor in children (following astrocytoma and medulloblastoma), they are relatively rare, with approximately 200 cases diagnosed in the US each year in children and adults less than 25 years of age. Ependymomas account for 6-12% of brain tumors in children less than 18 years of age, but 30% of brain tumors in children less than 3 years of age 1. The average age at diagnosis is 4-6 years 2-4. The underlying cause and basic biology of these tumors is not fully understood, and the optimal staging and treatment of ependymomas in children requires more study.

BIOLOGY: Ependymomas are tumors that arise from the cells lining the ventricles (the fluid containing spaces within the brain) and central canal within the spinal cord. According to the World Health Organization (WHO) classification of tumors, there are four types of ependymal tumors: myxopapillary ependymomas, subependymomas, ependymomas, and anaplastic ependymomas 5. The microscopic appearance and clinical behavior of these tumors vary considerably. Myxopapillary ependymomas are well-circumscribed tumors that occur primarily at the base of the spine in an area called the filum terminale. They are relatively uncommon in children. Subependymomas are well-circumscribed tumors that are usually located in or around the ventricles. Most are incidental tumors, but some may be large enough to cause symptoms. Following surgical resection, the long-term prognosis is generally excellent because these tumors rarely recur. Subependymomas are also uncommon in children. Classic ependymomas are fairly well delineated tumors. They can arise anywhere in the central nervous system, but commonly develop in the posterior fossa (the portion of the skull containing the cerebellum and brain stem), usually around the fourth ventricle, the spinal cord or the lateral ventricles (within the cerebral cortex). Anaplastic ependymomas, in general, exhibit a high growth rate and have been associated with a less favorable prognosis. However, there is uncertainty and controversy regarding the microscopic characteristics that distinguish the anaplastic form of ependymoma from the classic (lower grade) ependymoma.

Ependymomas tend to grow relatively slowly and displace, rather than invade adjacent brain or spinal cord tissue. They are described as plastic because they grow outside the ventricle through paths of least resistance within the brain. Some ependymomas, particularly those located in the posterior fossa, spread to other parts of the brain or spinal cord through the cerebrospinal fluid. Cerebrospinal dissemination occurred in approximately 13% of patients in one study. 4 Ependymomas rarely metastasize to sites outside of the central nervous system. When ependymomas recur after treatment, they tend to grow back locally (i.e. at or near the site of the original tumor), rather than spreading to other sites.

PRESENTATION: Symptoms that can be produced by ependymomas depend on the location of the tumor. Ependymomas are typically found in three major locations: the posterior fossa (the area of the brain below the tentorium, containing the cerebellum and the brainstem), the supratentorium (the area of the brain above the tentorium containing the cerebral hemispheres), and the spinal cord. 5 In adults, >75% of ependymomas arise within the spinal canal, but in children, about 90% arise within the brain in the posterior fossa, in or around the fourth ventricle and only 10% arise within the spinal cord.

Tumors in the posterior fossa can lead to obstruction of cerebrospinal fluid flow, and these children can develop signs and symptoms of increased intracranial pressure (headache, vomiting, head tilt, double vision). Specific localized symptoms can also occur. Tumors that arise on the floor of the fourth ventricle are associated with torticollis (wry neck with head tilted and twisted) and ataxia (loss of balance). Those arising to the side of the fourth ventricle are associated with cranial nerve dysfunction including impaired hearing, dysarthria (problems with speech articulation), dysphagia (difficulty swallowing) and dysmetria (unsteady, clumsy movements). Those arising on the roof of the fourth ventricle are associated with ataxia. Spinal cord lesions can be associated with low back pain, sciatica, extremity weakness, leg length discrepancy and scoliosis.

DIAGNOSIS: CT scan or MRI is used in the initial evaluation of children suspected of having a tumor in the brain. MRI is usually preferred because of the added detail to the images and the absence of radiation exposure. Although ependymomas have a suggestive appearance on these scans, biopsy or surgical removal of tumor tissue for microscopic evaluation is necessary to make a definitive diagnosis. Evaluation of cerebrospinal fluid may be performed after surgery to detect tumor spread through the cerebrospinal fluid. Spine MRI is also usually performed.

TREATMENT: The initial standard treatment for ependymoma is surgery. In general, the neurosurgeon will attempt to remove as much of the tumor as possible without causing damage to the normal brain, which can result in permanent neurological deficits. A complete resection, confirmed by post-operative MRI or CT scan, correlates highly with a favorable prognosis. 6-8 Although total resection is optimal, it is only possible in approximately 30-40% of cases because vital structures are frequently involved by the tumor. 8,9 Ependymomas arising above the tentorium, in the floor or roof of the fourth ventricle, or in the spinal canal are most amenable to complete resection. 10

Even with a complete tumor resection, tumor cells usually remain in the brain tissue surrounding the tumor. In an attempt to eradicate these residual tumor cells, radiation therapy is usually administered to children >3 years of age after recovery from surgery. 11. Several clinical studies have demonstrated that patients with ependymoma that are treated with post-operative radiation have a better outcome than patients who do not receive radiation. 12,13. Focal radiation to the site of the tumor lowers the risk of tumor recurrence at the site of origin 14,15. Craniospinal radiation (radiation to the entire brain and spinal cord) has also been used in the past, but clinical studies have demonstrated that the most common site of recurrence is at the site of the original tumor. Therefore, children are now treated with focal radiation to the tumor bed with doses of at least 4500-5000 cGy. 11 Children who present with or develop tumor spread within the cerebrospinal fluid may benefit from craniospinal radiation therapy, but there are questions regarding the role of prophylactic craniospinal radiation for children with localized tumors. There is currently no clinical evidence that prophylactic craniospinal radiation reduces the risk of developing tumors at other sites along the craniospinal axis.

The role of chemotherapy after surgery or radiation is unclear. Although chemotherapy has been used extensively in children with ependymomas, there is little clinical evidence that chemotherapy improves survival of children with this type of tumor. Chemotherapy is reserved for patients with residual tumor on scans after surgery (incomplete surgical resection) and for children younger than 3 years of age in an attempt to delay radiation therapy. Responses (tumor shrinkage) have been observed with several chemotherapeutic agents, but this has not always translated in to improved survival rates. Cisplatin is one chemotherapeutic agent that appears to have some activity in ependymomas, but there are significant differences in response rates of this drug across different studies. 16 Because cisplatin is associated with hearing loss and kidney damage, carboplatin has been substituted for cisplatin. However, the activity of carboplatin appears to be less than that of cisplatin. It is not entirely clear why there is not an improved survival with chemotherapy. It is known that drug resistance is common in ependymoma. In one study, 95% of ependymal tumors expressed the gene (MDR-1) that confers resistance to a variety of commonly used chemotherapeutic agents. 17

PROGNOSIS: Extent of tumor resection and age of the patient are generally agreed upon as being predictive of outcome for children with ependymoma, but it is unclear whether tumor location or microscopic characteristics of the tumor are also prognostic indicators. The extent of surgical resection is the primary factor influencing survival, with a significantly better prognosis for patients with completely resected tumors. These patients with completely resected tumors have an estimated 5-year survival rate of 67-80%, compared to an estimated 5-year survival rate of 22-47% in patients with incompletely resected tumors. 18 Most patients with incomplete tumor removal experience a tumor recurrence. Patients with residual tumor are also more than five times more likely to suffer from tumor spread through the cerebrospinal fluid.

Younger patients (< 4 years) have a worse prognosis, possibly due to a higher incidence of high-grade ependymomas, a more frequent lateral posterior fossa location of the tumor (which is more difficult to completely resect), or delays in initiating radiation therapy. Microscopic examination is unreliable in determining the long-term prognosis. In a study of almost 300 cases, the survival of patients with classic ependymoma was not markedly different from those patients with the anaplastic (high-grade) variant. 19

CONCLUSION: Several controversies still exist in the management of children with ependymoma. These include the significance of microscopic characteristics and grading of the tumor, the usefulness of chemotherapy after complete surgical resection of the tumor, and the significance of tumor location and microscopic appearance as prognostic factors. Because of the relative rarity of this tumor type, it is important that children with ependymomas be treated in centers with ongoing clinical studies in these areas so these questions can be answered and the most up-to-date treatment can be delivered.

Kathy Warren, M.D. is a Pediatric Oncologist in the recently formed Neuro-Oncology Branch of the National Cancer Institute for Neurologic Disorders and Stroke. This program is under the direction of Howard Fine, M.D. Its mission is to develop an integrated clinical, translation and basic research program for the purpose of developing novel experimental therapeutics for children and adults with tumors of the Central Nervous System. Investigators in the branch conduct both laboratory and clinical research aimed at improving the prognosis and management of patients with brain tumors.

  1. Duncan J, Hoffman H. Intracranial Ependymomas. In: Kaye A, Lows E, Jr, eds. Brain Tumors. Edinburgh: Churchill Livingstone, 1995:493-504.
  2. Foreman NK, Love S, R T. Intracranial ependymomas: Analysis of pognostic factors in a population-based series. Pediatr Neurosurg 1996;24:119-125.
  3. Perilongo G, Massimino M, Sotti G, al e. Analyses of prognostic factors in a retrospective review of 92 children with ependymoma: Italian Pediatric Neuro-Oncology Group. Med Pediatr Oncol 1997;29:79-85.
  4. Horn B, Heideman R, Geyer R, al e. A multi-institutional retrospective study of intracranial ependymoma in children: Identification of risk factors. J Pediatr Hematol Oncol 1999;21:203-211.
  5. Wiestler O, Schiffer D, Coons S, Prayson R, Rosenblum M. Ependymal Tumors. In: Kleihues P, Cavenee W, eds. Tumors of the Central Nervous System. Lyon (France): IARCPress, 2000:72-82.
  6. Healey E, Barnes P, Kupsky W, al e. The prognostic significance of post-operative residual tumor in ependymoma. Neurosurgery 1991;28:666-672.
  7. Hoppe-Hirsch E, Hirsch J, Lellouch-Tubian A, al e. Malignant hemispheres tumors in children. Childs Nerv Syst 1993;9:131-135.
  8. Nazar G, Hoffman H, Becker L, al e. Infratentorial ependymomas in childhood; prognostic factors and treatment. J Neurosurg 1990;72:408-417.
  9. Duffner P. Postoperative chemotherapy and delayed radiation in children less than three years of age with malignant brain tumors. New England Journal of Medicine 1993;328:1725-1731.
  10. Allen J, Siffert J, Hukin J. Clinical Manifestations of Childhood Ependymoma: A Multitude of Syndromes. Pediatric Neurosurgery 1998;28:49-55.
  11. Paulino A. Radiotherapeutic Management of Intracranial Ependymoma. Pediatric Hematology & Oncology 2002;19:295-308.
  12. Mork SJ, AC L. Ependymoma: A follow-up study of 101 cases. Cancer 1977;40:907-915.
  13. Shuman RM, Alvord EC, RW L. The biology of childhood ependymomas. Arch Neurol 1975;32:731-739.
  14. Goldwein J, Corn B, Finlay J, Packer R, Rorke L, Schut L. Is craniospinal radiation required to cure children with malignant 9anaplastic) intracranial ependymomas? Cancer 1991;67:2766-71.
  15. Chiu J, Woo S, Ater J, Connelly J, Bruner J, Maor M, al e. Intracranial ependymoma in children: analysis of prognostic factors. J Neuroonc 1992;13:283-90.
  16. Bouffet E, Foreman N. Chemotherapy for Intracranial Ependymomas. Child's Nerv System 1999;15:563-570.
  17. Chou P, Barquin N, Gonzalez-Crussi F, Tomita T, Reyes-Mugica M. Ependymomas in children express the multidrug-resistance gene: immunohistochemical and molecular biologic study. Pediatr Pathol Lab Med 1996;16:551-561.
  18. Merchant T. Current Management of Childhood Ependymoma. Oncology 2002;16:629-644.
  19. Schiffer D, Chio A, Giordana M, Migheli A, Palma L, Pollo B, Soffietti R, Tribolo A. Histologic Prognostic Factors in Ependymoma. Childs Nerv Syst 1991;7:177-182.


This article was written for The Childhood Brain Tumor Foundation,Germantown, MD, www.childhoodbraintumor.org

Thanks to Our Sponsors