Discussion

Ependymoma is a glial tumor arising from ependymal cells in the periventricular area. Percival Bailey, in 1924 coined the term ependymoma for a group of tumors arising from ependymal cells.[11] Although ependymoma (like germ cell tumors) have a propensity for CSF dissemination, the exact frequency is debated in the existing literature (9%–20%). This figure was 12.5% in our patient cohort. The WHO has classified ependymoma in three distinct groups. Myxopapillary and subependymal tumors are classified as grade I. Grade II ependymoma are the most common variant of ependymoma followed by anaplastic ependymoma or grade III ependymoma. Grade II ependymoma has four distinct types viz., cellular, papillary, clear cells, tynactic.[1] The present analysis only included grade II and III tumors; 57.5% of patients in our series had grade II histology.

Ependymoma most commonly arise from the posterior fossa (60%–70%). Due to the primary tumor location these tumors often present with acute symptoms secondary to raised intracranial tension. However, supratentorial ependymoma is often associated with mood changes, personality changes, and lobar syndromes. Nearly 65% of patients in our cohort had infratentorial tumor location and 85% of patients had symptoms of raised intracranial tension.

Surgical resection has long been considered the most important treatment for ependymoma. A GTE has been reported to impart survival advantage[6,7] and is often attempted if feasible without increasing morbidity. In general, GTE can be accomplished in 50%–79% of patients. Some authors advocate second look surgery whenever possible to optimize survival. However, the decision is crucial and needs to be weighed against potential complications. In a study of 45 patients, Rogers et al.[12] reported 10 years OS rates of 83% versus 43% in patients treated with GTE plus radiotherapy versus STE plus radiotherapy. GTE was also associated with a lower risk of death from any cause (HR: 0·16, 95% CI: 0·07–0·37, P ≤ 0·0001) compared with near- or subtotal resection in the study by Merchant et al.[6] two years DFS for patients with GTE versus STE in our study was 55% versus 30%, respectively. However, the difference was not statistically significant. Only 42.5% of patients in the study could undergo GTE.

Radiotherapy is considered the most important adjuvant therapy to optimize long-term disease control.[13] The philosophy of radiation treatment has been rapidly evolving over the period. While earlier series[3,4] recommended larger fields (whole brain for supratentorial, whole brain with cervical cord extension for infratentorial and CSI for high-grade ependymoma), more recent series have shown the patterns of failure to be predominantly local.[5] Only 2/9 patients receiving tumor bed plus margin had a recurrence in the study by Paulino et al. and both of them occurred in the tumor bed itself. In a large series[6] of 153 patients, authors used 1 cm margin only to the tumor bed as the target volume and reported all local failures (21 patients) to lie within the 95% isodose volume. Current Children's Oncology Group (Protocol) is using a further reduced volume of tumor bed plus 0.5 cm margin for CTV. We used tumor bed plus 2 cm margin in our study, and most of the failures in our study were also in tumor bed itself (13/17 failures).

The dose response has always remained an area of active research in ependymoma. Conventionally, a dose of 50 Gray is considered adequate for low grade and 60 Gray for high-grade tumors, respectively. Paulino, however, failed to establish a dose response relationship beyond 45 Gray.[5] Merchant et al.[6] reported improved outcome in patients with anaplastic ependymoma when treated with a higher dose of radiation. Review of older series revealed better survival with a radiation dose of 45 Gray or higher. Recently, studies are evaluating a further dose escalation of >54 Gray. Patients receiving a higher dose in our series did significantly better than those treated with a lower dose (P = 0.0229). Two years DFS was 68% versus 21% in favor of the higher dose. In an interesting prospective study, Massimino et al. randomized children to either hyperfractionated radiotherapy (HFRT) 70.4 Gy (1.1 Gy/fraction twice daily) or conventionally fractionated radiotherapy to improve disease control by increasing the total dose of radiation but failed to show improved outcome with HFRT.[14]

Adjuvant chemotherapy has long been hypothesized to improve survival in ependymoma. However, there exists little data to support the use of chemotherapy. In few reports, authors have reported variable rate of chemo sensitivity. The response to a single agent and multi agent chemotherapy has ranged from 0% to 11% and 0%–86%, respectively. White et al.[15] reported 86% response rate in 7 patients (<4 xss=removed>et al.[16] reported 48% response in 25 patients younger than 3 years when treated with 2 cycles of vincristine and cyclophosphamide. The HIT trial[7] reported 3 years OS to be 75.6% with adjuvant chemotherapy administered after completion of radiotherapy. We could not find any effect of adjuvant chemotherapy on survival in our series (P = 0.48).

Recurrences happen in approximately 50% of patients of ependymoma. The median time to recurrence is 13–25 months.[17,18] Recurrences are predominantly local and only 20% have isolated distant recurrence. The treatment options are limited after a recurrence. Surgical excision of recurrent disease should be tried in all cases followed by postoperative radiotherapy or due to consideration given to re-irradiation.[19,20] Reirradiation has been found to yield durable response at times. Bouffet et al.[19] treated 18 of 47 recurrent ependymoma patients with 54 Gray reirradiation and reported 3 years OS of 81%. Although there were no severe acute toxicities, 2/18 patients had endocrine dysfunction later on and 1 patient required special educational support. Stereotactic radiotherapy is often considered as an effective modality to minimize toxicity during re-irradiation.[21] In the recent years, proton therapy has also emerged as an effective option to treat recurrent ependymoma without increasing morbidity.[22,23] However, there are limited prospective data on these promising modalities.

Salvage chemotherapy for recurrent or metastatic patients has been found to be of questionable benefit. The response rates for single or combination chemotherapy has been found to be 12.9% and 17.4%, respectively.[24] Concerns of toxicities have been keeping physicians away from reirradiation, however in view of encouraging results reported by Bouffet et al.[19] and Merchant et al.,[20] it appears imperative to use this modality to achieve a durable response. Only 1 patient received reirradiation in our cohort and 2 patients received postoperative radiotherapy.

Great enthusiasm has been witnessed in classifying ependymoma in different distinct molecular subgroups to optimize treatment and outcome. Two landmark articles by Witt et al.[25] and Pajtler et al.[26] classified ependymoma in distinct molecular and prognostic subgroup. This classification makes it is possible to deliver treatment directed to the distinct molecular characteristic.[27] However, the cost of such classification is a major limitation to include in regular patient care and needs prospective validation.

Retrospective nature of the present study is a limitation in itself. Our patients also had less rates of GTE and outcomes are not comparable to those of developed world series. Relatively, short follow up and lost to follow up are other limitations of our present study. The effectiveness of this combined modality approach in an unselected patient population outside a clinical trial (simulating a real world scenario) in a resource constrained setting can be considered as the strength of the data.

Several challenges exist in the optimal management of brain tumors including ependymoma in our country. Most of the patients whom we treat are referred from low volume centers where neurosurgical facilities are less developed and a second look surgery is not done routinely because of long waiting times or resource constraints. Radiation facilities (particularly conformal radiation and facilities of radiation for pediatric patients, many of them requiring general anesthesia) are not universally available and accessible to all patients, and few centers include ours provide these treatments. This “double trouble” leads ultimately in patients receiving either sub-optimal surgery or radiation or both which leads to an inferior outcome. These tumors also are predominant in the pediatric population and require long-term follow up not only for surveillance of disease recurrence but also for neurocognitive outcomes. Since patients come to our center from far off places, many of them are lost to follow up immediately after or within 2–3 years of completion of treatment. This further precludes a better analysis of prognostic factors as well as survival.

Ours is the largest series published till date from India and could serve as a benchmark and comparator for other studies too. The challenges and the factors identified in our study leading to inferior outcomes may help other centers in overcoming these hurdles prospectively and thus leading to improvement in the outcome of patients with ICE in our country.

Conclusion

ICE is a rare group of malignancies. Maximal safe resection followed by post-operative radiotherapy to a higher dose of 60 Gray (particularly for anaplastic ependymoma) should be used. Higher MIB labeling index confers poor prognosis. Role of chemotherapy remains controversial, and radiation therapy should not be delayed even in young patients. Reirradiation and surgical excision with postoperative radiation therapy remain valid treatment options at recurrence and may yield durable response in some patients.