Myxopapillary Ependymoma

Ependymomas are the most common primary spinal cord tumors. They have been classified by the World Health Organization into grades I, II, and III. Grade I tumors include myxopapillary ependymomas and subependymomas. Grade II tumors are just called ependymomas and can be further divided as cellular, papillary, clear cell, or tanycytic. Grade III tumors are called anaplastic ependymomas, tend to occur in the brain, and are rare in the spine.

Myxopapillary ependymomas are thought to arise from ependymal glia of the filum terminale and occur almost exclusively in the conus medullaris and filum terminale. As such, they are intradural and usually extramedullary (in contrast to upper cord lesions that are typically intramedullary). Extradural myxopapillary ependymomas, though rare, can also occur and are thought to arise from extradural remnants of the filum terminale or the coccygeal medullary vestige.

Myxopapillary ependymomas are often encapsulated and may have foci of hemorrhagic or mucinous degeneration. They are usually centered within the filum terminale but can extend into the conus medullaris. Myxopapillary ependymomas usually displace the lumbosacral nerve roots, but may encase them if the tumor gets large enough.

Myxopapillary ependymomas are slow-growing and can get large enough to expand the spinal canal, leading to interpedicular widening on frontal radiographs and widening of the spinal canal and scalloped vertebral bodies on the lateral view. Bone destruction may also be appreciated in advanced cases.

Non-contrast CT images reveal myxopapillary ependymomas to be isoattenuating compared to the spinal cord. The spinal canal may be widened and there may be scalloping of the vertebral bodies with neural foraminal enlargement. Myxopapillary ependymomas typically have intense homogeneous enhancement.

On MRI, myxopapillary ependymomas are typically iso- to hyperintense compared to the spinal cord on T1-weighted images, depending on the amount of proteinaceous mucoid matrix. If present, T1-hyperintensity helps differentiate myxopapillary ependymomas from other ependymoma subtypes, which tend to be hypo- or isointense on T1-weighted images. Myxopapillary ependymomas are T2-hyperintense compared to the spinal cord. As with other ependymomas, myxopapillary ependymomas demonstrate intense enhancement. Heterogeneous enhancement signifies hemorrhage or necrosis.

Most ependymomas of the cauda equina region are slow-growing, and patients present with indolent low back pain. Mass effect from the tumor may cause sciatica or other, less common symptoms, such as sensorimotor disturbance and bowel and bladder dysfunction.

Differential considerations include:

• Other ependymoma subtypes:
• Schwannoma: Can have findings almost identical to those of a small myxopapillary ependymoma.
• Subependymoma: Can have findings almost identical to those of a small myxopapillary ependymoma.
• Astrocytoma:
• Hemangioblastoma:
• Ganglioma:
• Paraganglioma:

When the mass is large and causes bone destruction, aneurysmal bone cyst, chordoma, plasmacytoma, metastasis, and giant cell tumor may also be considered.

Our patient was a young man who presented with several years of low back and flank pain. A renal tumor protocol CT was performed, which showed a mass in the spinal canal. On MR, the mass is well encapsulated, isointense to the spinal cord on T1-weighted images and heterogeneously hyperintense on T2-weighted images. Post-contrast images reveal intense heterogeneous enhancement. On the axial T2-weighted image, the mass splays the nerve roots of the cauda equina and widens the spinal canal.

Case courtesy of Dr. Sibin Thachet.

References

• Shors SM, Jones TA, Jhaveri MD, Huckman MS. Best cases from the AFIP: myxopapillary ependymoma of the sacrum. Radiographics. 2006 Oct;26 Suppl 1:S111-6.
• Wippold FJ 2nd, Smirniotopoulos JG, Moran CJ, Suojanen JN, Vollmer DG. MR imaging of myxopapillary ependymoma: findings and value to determine extent of tumor and its relation to intraspinal structures. AJR Am J Roentgenol. 1995 Nov;165(5):1263-7.