Radiation Oncology/CNS/Ependymoma

Epidemiology

 * 5-10% of pediatric brain tumors
 * Primarily in infants and children <5 years
 * Approximately 200 pediatric cases per year in USA

Location

 * Glial tumors that arise from ependymal lining of the ventricular system
 * May occur anywhere in the ventricular system or spinal canal
 * 90% are in the brain
 * 1/3 supratentorial
 * 2/3 infratentorial, many arising from the 4th ventricle
 * Up to 50% may infiltrate into the brainstem
 * 10% in the spine
 * Myxopapillary ependymomas are found almost exclusively in the conus and filum terminale and are the most common spinal tumors in this location.
 * Location differs with age:
 * In young children: 90% intracranial, typically roof of 4th ventricle
 * In adolescents/adults: 75% spinal canal

Patterns of spread

 * Spread is primarily local
 * In 50% of pts with posterior fossa tumors, a "tongue" of tumor may extend through the foramen magnum well into the cervical region
 * Risk of leptomeningeal seeding is 5-10%

Presentation

 * Since majority occur in 4th ventricle, symptoms are related to obstruction and subsequently increased intracranial pressure
 * In spinal ependymomas, presentation is typically with sensory deficits (compared with astrocytomas, which tend to present with pain and motor symptoms)

Workup

 * MRI of brain and spine: typically well-circumscribed, with displacement rather than invasion of brain parenchyma
 * CSF cytology

Pathology

 * Myxopapillary ependymoma (WHO grade I)
 * Subependymoma (WHO grade I)
 * Ependymoma (WHO grade II)
 * Cellular
 * Papillary
 * Clear cell
 * Tanycytic
 * Anaplastic ependymoma (WHO grade III)

Treatment Overview

 * No randomized studies
 * Surgery is primary management
 * Adjuvant RT is considered standard of care, although highly selected series report good outcomes with surgery alone
 * Because recurrences tend to be local, RT is typically GTV plus 1-1.5 cm margin (per Perez, 4th edition)
 * 5-year OS 50-65%, PFS 25-45%
 * Recurrences typically local; median time to failure 1-2 years. About 20% isolated distant recurrences

Surgery

 * Extent of surgery most important predictor of outcome
 * 5-year OS with GTR: 70-80%
 * 5-year OS with STR: 25-45%
 * Gross total resection (GTR) achievable depending on location, overall 30-50%:
 * Supratentorial lesions: GTR ~80%
 * Infratentorial lesions: GTR "lower" rate (per Perez, 4th edition), BNI rate 70%
 * Spinal lesions: GTR close to 100%


 * Barrow Neurological Institute; 2005 (1983-2002) PMID 15871504 -- "Is gross-total resection sufficient treatment for posterior fossa ependymomas?" (Rogers L, J Neurosurg. 2005 Apr;102(4):629-36.)
 * Retrospective. Posterior fossas ependymomas, no subependymomas or CNS dissemination. GTR in 71%. RT in 13/32 GTR and 12/13 STR. RT fields CSI + boost (n=6), rest posterior fossa or localized fields. Median F/U 5.5 years
 * Local control: Median LC: GTR alone 6.6 years vs. GTR+RT not yet reached; 10-year LC: GTR alone 50% vs. GTR+RT 100% (SS) vs. STR+RT 36% (SS), GTR alone comparable to STR+RT
 * 10-year OS: GTR alone 67% vs. GTR+RT 83% vs. STR+RT 43% (all NS)
 * Conclusion: Adjuvant RT significantly improves control after GTR


 * BI Medical Center, NYC; 1998 PMID 9755311 -- "Treatment of intracranial ependymoma by surgery alone." (Hukin J, Pediatr Neurosurg. 1998 Jul;29(1):40-5.)
 * Prospective. 10 patients. Children >3 years, s/p GTR
 * Outcome: 7 free of disease, 3 recurred (2 salvaged with surgery and RT). No deaths. Overall median PFS 4 years, and OS 4.2 years
 * Conclusion: Deferral of RT after GTR of supratentorial ependymoma safe; recurrence salvageable

Radiation
Role of post-operative RT:
 * Historically, ependymoma was believed to disseminate throughout the neuraxis, and CSI became the standard management approach
 * However, more recent retrospective studies have shown that failure is primarily local in >90%
 * Benefit for post-operative RT is based on retrospective studies. 5-yr EFS 45% vs 0% and 51-70% vs 13% PFS.

Dose response: Evidence of a dose response above 45-50 Gy.

RT volume: No evidence of need for extended-field or craniospinal radiation. CTV margin around resection cavity is generally 1 cm.

Hyperfractionation: In one study (POG 9132), HF RT produced better event-free survival for pts with incomplete resection. Dose was 69.6 Gy at 1.2 Gy BID.

Subependymoma

 * Uncommon variant of ependymal tumors
 * Histologically clusters of ependymal cells in an astrocyte-based matrix
 * Clinically slow growing, with extremely low Ki-67
 * Many discovered incidentally at autopsy
 * Clinically symptomatic tumors present with CSF obstruction
 * Gross total resection alone frequently leads to cure, but perioperative mortality can be high
 * Role of adjuvant RT is not clear

Anaplastic Ependymoma

 * HIT-SKK87 and HIT-SKK92 (1987-1992, 1992-1997)
 * HIT-SKK87: Low-risk (complete resection, M0) and age <2.5 received maintenance chemo only until RT at 3 years or progression. High-risk (subtotal resction, M+) and age 2.5-3 induction chemo / maintence chemo until RT at 3 years or progression
 * HIT-SKK92: Post-op chemo x3 cycles first, then if CR no RT. If PR and <18 months, then more chemo, if PR and >18 months, then RT.
 * RT given as CSI 35.2/22 Gy + 20/10 Gy boost; if no residual disease, reduced to CSI 24 Gy + PF boost to 55 Gy
 * 2005 PMID 16300848 -- "Role of radiotherapy in anaplastic ependymoma in children under age of 3 years: results of the prospective German brain tumor trials HIT-SKK 87 and 92." (Timmermann B, Radiother Oncol. 2005 Dec;77(3):278-85.)
 * Total 34 children enrolled. No RT given in 13 children, preventive RT in 9, salvage RT in 12
 * 3-year outcomes: PFS 27%, OS 56%. Positive predictive factors: higher age, M0, GTR, RT. If no RT, only 3/13 survived
 * Conclusion: Delaying RT jeopardized survival. Predominant failure primary tumor site; RT of neuraxis should be omitted in localized disease

Spine Ependymoma

 * Rare tumors
 * Gross total resection (~50%) leads to long-term control in >90%
 * Subtotal resection can recur in 50-70% without adjuvant therapy; with adjuvant RT local control 60-100%
 * RT dose seems to show a threshold effect, with worse control <45 Gy. However, radiation myelopathy becomes concerning >55 Gy, and with pediatric patients possibly as low as 40 Gy
 * Field size is unclear, but commonly is +/- 2 vertebral bodies. In sacrum, need to cover nerve roots, so inferior edge at S4/S5 and lateral edges to SI joints


 * Kaiser Permanente; 2007 (USA) PMID 17689025 -- "Outcomes in treatment for intradural spinal cord ependymomas." (Volpp PB, Int J Radiat Oncol Biol Phys. 2007 Nov 15;69(4):1199-204. Epub 2007 Aug 6.)
 * Retrospective. 23 patients (17 surgery alone, 5 surgery + RT, surgery + RT + chemo). Mean RT dose 45 Gy
 * Outcome: 9-year OS 64%, LC 96%
 * Recurrences: surgery alone: 12% recurred (1/17 locally, 1/17 distally). CMT 33% recurrence (both out-of-field)
 * Conclusion: En block total resection primary treatment, RT reserved for unfavorable characteristics (residual tumor, anaplastic features, piecemeal resection)

Reviews

 * PMID 15049020, 2004 &mdash; "Review of radiotherapy dose and volume for intracranial ependymoma." Taylor RE. Pediatr Blood Cancer. 2004 May;42(5):457-60.