paediatrics Brussels 17

Bifractionated RT in childhood ependymoma d C. C ONTER et al .

1537

well as endocrine and neurologic function (7) , strategies to delay or avoid RT have been reported (8–10) . The present study focused on strategies combining surgery and HFRT that were investigated prospectively in children >5 years at diagnosis.

The rules to stop the study at 2 years were defined as follows: >30% of relapses among patients with CR and >50% of relapses or progressive disease in patients with incomplete resection.

Evaluation procedures The initial screening included brain MRI with and without con- trast, and three-dimensional measurements of the tumor. Spinal MRI was performed if the tumor was located in the posterior fossa. Examination of the cerebrospinal fluid was mandatory only if le- sions were found on spinal MRI. The quality of the resection was evaluated using early postoperative imaging by the local physician during treatment and was centrally reviewed retrospectively. The clinical examination and MRI were repeated 6 weeks after treatment to assess the response rate. The response to RT corre- sponded to a reduction in the size of the residual tumor surface by >50% on imaging (12) . MRI was repeated every 4 months for 2 years, every 6 months during the third year, and every year for 5 years. Long-term effects ( e.g., audiometric, endocrinal, or psycho- metric impairment and school or professional difficulties) were recorded each year. Statistical analysis The OS rates, estimated using the Kaplan-Meier method, were calculated from the date of surgery to death or the date of the last follow-up visit for patients who were still alive. The progression- free survival (PFS) rates were estimated from the date of surgery to the time of documented failure (date of progression for patients whose disease progressed before achieving complete remission, date of relapse or date of death for others) or to the date of the last follow-up visit for those remaining in their first complete remission. The median follow-up was estimated using the method of Schemper and Smith (13) . Statistically significant differences in OS and PFS were tested using the two-tailed log–rank test. Statistical analyses were performed using Statistical Package for Social Sciences, version 11.5 (SPSS, Chicago, IL). Study population A total of 24 patients from nine Society of Pediatric Oncol- ogy centers were included. Ten patients were registered but not included for the following reasons: previous chemother- apy in 1, relapse in 5, nonbifractionated RT in 1, age <5 years in 2, and wrong histologic type in 1 (medulloepithelioma). Because of the low rate of inclusion, the study was stopped before it reached the expected number for inclusion ( n = 40). The median age was 8.6 years (range, 5–17). Of the 24 pa- tients, 16 were boys and 8 were girls. The tumor location was the posterior fossa in 20 patients and supratentorial in 4. Among the 20 patients with infratentorial EP, 19 underwent spinal MRI; all findings were considered normal. The histo- logic grade could be assessed in 23 of the 24 patients; 13 tu- mors were classified as Grade 2 (classic) and 10 as Grade 3 (anaplastic). Surgery Surgical resection, as reported by the neurosurgeon, was total in 15 patients, subtotal in 8, and partial in 1. One patient underwent second-look surgery for resection of a postopera- tive residual mass, after which the surgery was considered RESULTS

PATIENTS AND METHODS

Patient eligibility Postoperative local HFRT was offered to all children with local- ized intracranial EP seen between November 1996 and December 2002 in centers affiliated with the French Society of Pediatric Oncol- ogy. The criteria for enrollment included age of 5–17 years at diag- nosis, EP of any pathologic grade, and written informed consent. The criteria for exclusion were spinal primary EP, disseminated EP, previous chemotherapy or RT, relapse, associated disease, re- fusal, or impossible follow-up. Patients with ependymoblastoma were not included in this study. The findings were prospectively re- corded. The primary objective was to determine the 5-year overall survival (OS) of the patients after surgery and HFRT. Secondary ob- jectives were to determine the response rate to HFRT when applica- ble, identify the prognostic factors (especially the value of grading) and patterns of relapse, and define the rate of sequelae in a homoge- neous group of patients. Histologic slides of all patients were reviewed by a qualified pe- diatric neuropathologist panel (M.M. Ruchoux, A. Lellouch-Tubi- ana, and A. Jouvet) and classified as Grade 2 (classic) or Grade 3 (anaplastic) according to the absence or presence of signs of ma- lignancy ( i.e., mitoses, necrosis, and neovascularization) according to the World Health Organization criteria (11) . Treatment characteristics The extent of resection was rated by the surgeon as incomplete (bi- opsy or remaining tumor of >1.5 cm 3 ), subtotal (remaining tumor <1.5 cm 3 ), or total (no recognizable residue), as recommended by the Inter- national Society of Pediatric Oncology Brain Tumor Subcommittee for the reporting of trials (12) . The extent of resection was rated by the radiologist as complete (CR), partial, or doubtful as determined from the findings on postoperative craniospinal magnetic resonance imaging (MRI) or computed tomography (CT) performed as soon as possible (24–48 h) after surgery. The ratings were reviewed retrospec- tively by a panel of experts (A. Geoffray and P. Thiesse). In the case of a CR, 60 Gy HFRT in two daily fractions of 1 Gy (sep- arated by $ 6 h) was performed. The target volume was the preopera- tive tumor volume plus a 1-cm safety margin. RT could be conformational or not, at the discretion of the group. The photon energy was $ 8MeV. No more than 44 Gy was delivered to areas in the occip- ital region and no more than 60 Gy if more than one-third of the brain was involved. A customized immobilization device was used for each patient. The clinical target volumewas defined onCT, withMRI fusion for the last patients treated, when this procedure was available. In the case of partial removal, second-look surgery was discussed before RT. If the second surgical resection was complete, the pa- tients were treated as described for those with a CR. Otherwise, an extra boost of 6 Gy was delivered to the residual volume, in ad- dition to the 60-Gy dose already described. No patient received che- motherapy in this first line of treatment, because the place of chemotherapy for this disease is still debated (4) . For patients in relapse, repeat surgery was performed whenever feasible, unirradiated areas were irradiated, and inclusion in a protocol of oral etoposide was encouraged.