paediatrics Brussels 17

86

M. Reni et al. / Critical Reviews in Oncology/Hematology 63 (2007) 81–89

trial of childrenwith infratentorial ependymoma, without evi- dence of survival benefit [84] . Currently, there is no proof that the addition of chemotherapy to radiotherapy improves out- come and application of adjuvant chemotherapy should not be recommended as standard treatment and should be restricted to investigational trials [44,63] .

these tumours may be similar to low-grade gliomas, where no untoward impact on overall survival has been observed following a ‘wait and see’ policy following surgery [77] . On the other hand, no data about the impact of delaying RT on tumour control and survival are available for adult patients with totally resected low-grade ependymomas, and thus the best postoperative policy remains controversial. The option of a close observation and delaying irradiation until tumour progression seems appropriate for individual clinical use on type Rbasis for selected subsets of patients, such as thosewith totally resected low-grade tumours. The standard treatment volume should be defined according to modern conforma- tional techniques and is limited to the pre-surgical tumour bed with an added margin of 1–2 cm [73,78] , on a type C basis in all low-grade lesions [34,35,49,55,56] , and on a type R basis in high-grade supratentorial lesions [33,54–56,64,79] . Craniospinal irradiation should be reserved for patients with evidence of cranio-spinal seeding [33,59,61] and is suitable for individual clinical use on type R basis in high-grade infratentorial lesions [33,34,53,56,64,80,81] , based on their high potential of tumour spread into the ventricular sys- tem and into the spinal subarachnoid space] [81] . However, prophylactic spinal irradiation did not seem to modify pat- terns of failure [33,34,39,49,54,65,71,80,81] for high-grade lesions, which disseminated into the spine in 8–9.4% of patients treated with and in 6.6–10% without craniospinal irradiation [35,55] . Moreover, local recurrence is the pre- dominant pattern of failure for both low-grade and high-grade ependymomas [33,47,56,80–82] , and the lack of local con- trol represents the main risk factor for subarachnoid seeding [49] , which rises from 3.3–5% to 8.5–10% based on either the absence or presence of local recurrence [65,80] . Finally, no survival advantage has been demonstrated for cranio- spinal irradiation [49,73,78,80] , and the rate of spinal seeding among high-grade tumours may be overestimated since most authors have not excluded ependymoblastomas, and treatment-related toxicity remains a serious concern, espe- cially in small children [35,80] . Little is known about the optimal radiation schedules dose to be employed. Doses >45 Gy [46,49,56,64,67] or >50 Gy [34,54] have been shown to be superior to lower doses. Other authors did not observe differences in outcome using doses of 40–50 Gy or of 50–55 Gy [35] . A standard dose, on type Rbasis, of 54–60 Gy, should be delivered to the tumour bed. Whenever possible, the dose to the optic chiasm should be limited to 55.8 Gy, to the upper cervical spinal cord to 54 Gy and to the optic nerves to 50.4 Gy [83] . Recently, the issue of treatment duration was addressed in a series of 34 patients collected over e period of 33 years [57] . The authors concluded that treatment duration was the most important prognostic factor.

6.4. Treatment of recurrent disease

A standard salvage therapy for recurrent ependymoma has not been identified. Second surgery and re-irradiation can be suitable for individual clinical use. Patients with ependymoma failure provide an important opportunity for prospective investigation of potentially effective drugs, and their inclusion in investigational multi-centre clinical trials should strongly be encouraged. Cisplatin and carboplatin are the most extensively tested single agents, showing response rates of 31% and 13%, respectively, in a total of about 30 patients [85] . Therefore, chemotherapy including cisplatin may be suitable for individual clinical use. Among the other few agents which have been assessed in 10 cases or more, are ifosfamide, thiotepa, arizidinybenzoquinone, dibromod- ulcitol, idarubicin and PCNU. None of these has achieved a better than 10% response rate [85] . Anecdotal experiences with procarbazine, vincristine and cytarabine have not been encouraging, while etoposide achieved two responses in nine cases [85] . A few combinations of agents have been investi- gated, mainly in infants and young children. There are only a few trials including more than 20 cases, in which 48% of 25 children, under 3 years of age responded to a combination of vincristine and cyclophosphamide [86] , and 55% of 21 cases responded to a combination of procarbazine, ifosfamide, etoposide, high-dose methotrexate, cisplatin and cytarabine [87] . High-dose chemotherapy has also undergone limited clinical investigation in children, without proven benefits and up to 33% treatment-related fatal toxicities [85,88,89] . To date this approach should be considered not recommended. Since the inability to eradicate the primary tumour in both low- and high-grade ependymomas is the single most important factor of treatment failure [55] , more aggres- sive local therapies are being assessed in clinical trials to improve tumour control. Stereotactic radiosurgery has given favourable results in small series as a first-line [90] and as a salvage treatment [91,92] and deserves further evalua- tion. Conformational radiotherapy as hyperfractionated dose schedules have also been explored, and may help decreas- ing local failure rate in subtotally resected patients [33,83] . Among the new drugs, topotecan and paclitaxel have been investigated, however without evidence of activity [93] , while preliminary experience with temozolomide, as reported in abstract form, seems somewhat more encouraging [94] . The use of temozolomide should be further examined in the con- text of a clinical trial. 6.5. New active drugs and therapeutic options

6.3. Chemotherapy

Information concerning the activity of chemotherapy in ependymoma is very limited. The role of postoperative chemotherapy has been assessed in a randomised phase III