This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Grill, J. Articles by Kalifa, C. Search for Related Content PubMed PubMed Citation Articles by Grill, J. Articles by Kalifa, C.

Postoperative Chemotherapy Without Irradiation for Ependymoma in Children Under 5 Years of Age: A Multicenter Trial of the French Society of Pediatric Oncology

From the Departments of Pediatrics, Biostatistics, Radiology, and Radiotherapy, Institut Gustave Roussy, Villejuif; Department of Neuropathology, University Hospital; Department of Pediatric Hematology/Oncology, La Timone Hospital, Marseille; Department of Pediatric Neurosurgery, Necker Hospital; Department of Pediatrics, Institut Curie, Paris; Department of Pediatrics, Centre Léon Bérard, Lyon; Department of Pediatric Hematology/Oncology, University Hospital, Rennes; and Department of Pediatric Hematology/Oncology, Children’s Hospital, Nancy, France.

Address reprint requests to Chantal Kalifa, MD, Department of Pediatrics, Gustave Roussy Institute, 39 rue Camille Desmoulins, 94805 Villejuif Cedex, France; email: kalifa{at}igr.fr

	ABSTRACT

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
PURPOSE: To evaluate a strategy that avoids radiotherapy in first-line treatment in children under 5 years of age with brain or posterior fossa ependymoma, by exclusively administering 16 months of adjuvant multiagent chemotherapy after surgery.

PATIENTS AND METHODS: Between June 1990 and October 1998, 73 children with ependymoma (82% with high-grade tumors) were enrolled onto this multicenter trial. Children received adjuvant conventional chemotherapy after surgery consisting of seven cycles of three courses alternating two drugs at each course (procarbazine and carboplatin, etoposide and cisplatin, vincristine and cyclophosphamide) over a year and a half. Systematic irradiation was not envisaged at the end of chemotherapy. In the event of relapse or progression, salvage treatment consisted of a second surgical procedure followed by local irradiation with or without second-line chemotherapy.

RESULTS: Conventional chemotherapy was well tolerated and could be administered in outpatient clinics. No radiologically documented response to chemotherapy more than 50% was observed. With a median follow-up of 4.7 years (range, 5 months to 8 years), the 4-year progression-free survival rate in this series was 22% (95% confidence interval [CI], 13% to 43%) and the overall survival rate was 59% (95% CI, 47% to 71%). Overall, 40% (95% CI, 29% to 51%) of the patients were alive having never received radiotherapy 2 years after the initiation of chemotherapy and 23% (95% CI, 14% to 35%) were still alive at 4 years without recourse to this modality. In the multivariate analysis, the two factors associated with a favorable outcome were a supratentorial tumor location (P = .0004) and complete surgery (P = .0009). Overall survival at 4 years was 74% (95% CI, 59% to 86%) for the patients in whom resection was radiologically complete and 35% (95% CI, 18% to 56%) for the patients with incomplete resection.

CONCLUSION: A significant proportion of children with ependymoma can avoid radiotherapy with prolonged adjuvant chemotherapy. Deferring irradiation at the time of relapse did not compromise overall survival of the entire patient population.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
EPENDYMOMAS ACCOUNT for 10% of childhood brain tumors, and approximately 70% are located in the posterior fossa (PF). More than 50% of these tumors arise in children under 5 years of age.¹ Overall survival (OS) of young children with brain tumors is significantly worse than that of older children,² and treatment-related neurotoxicity can be substantial when irradiation is used.³ Finding the most effective adjuvant therapy while limiting treatment-related toxicity is a formidable challenge in this tumor type.

Since the early 1990s, chemotherapy has been widely introduced in the management of brain tumors in young children to defer radiotherapy and thus improve intellectual outcome.^4,5 Children with PF tumors nonetheless sustain neuropsychologic impairment, even when irradiation is limited to the PF.⁶ In 1990, the French Society of Pediatric Oncology (SFOP) initiated a prospective trial of postoperative chemotherapy without irradiation in children under 3 years of age with malignant brain tumors whatever the histology. After surgery, children received multiagent chemotherapy over 16 months but no irradiation if they remained progression-free. The encouraging results obtained during the first years of the study prompted us to extend the age limit for study entry to 5 after 1995. Because outcomes differ considerably according to the histologic diagnosis, the present report is focused on ependymoma.

	PATIENTS AND METHODS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Patient Eligibility
Children with PF or supratentorial ependymoma were eligible for the study provided the following criteria were fulfilled: (1) age below 3 at diagnosis of the tumor (after 1995, the age limit for study entry was extended to 5); (2) histologically proven ependymoma, whatever the grade; (3) no prior exposure to chemotherapy or radiotherapy; (4) if necessary, surgery in more than one step in order to maximize resection before entering the chemotherapy trial; (5) normal auditory, hepatic, and renal function; and (6) a Lansky score exceeding 30. This was the only protocol available for this tumor type during the study period for all the SFOP centers. Thus, all eligible children referred to these centers (ie, all the children receiving adjuvant treatment for ependymoma in France during this period) were enrolled. Informed consent was obtained from the parents or guardians of each child in accordance with institutional guidelines.

Pathology Review
Histologic slides concerning all patients were reviewed by a panel of four different neuropathologists and classified according to World Health Organization (WHO) criteria.⁷ For the purposes of the analysis, ependymoma were divided into low-grade or high-grade lesions according to the absence or presence of signs of malignancy (ie, mitoses, necrosis and neovascularization [grade 3 according to WHO criteria]). Patients were not included in the study if the panel did not confirm the diagnosis of ependymoma. Ependymoblastoma were not included in this study.

Surgery and Extent of Resection
All patients were to undergo maximal surgical resection, if possible, before enrollment, with the caveat that the anatomic location of the tumor and the patient’s clinical condition might contraindicate extensive resection. All operative reports were reviewed centrally. Resection was deemed complete when the neurosurgeon confirmed the absence of residual tumor at the end of the operative procedure. Resection was considered subtotal when at least 90% of the tumor had been removed. All other resections were considered as partial.

Computed tomographic (CT) scan or magnetic resonance imaging (MRI) films were centrally reviewed. Early imaging-documented extents of resection (ie, before day 3) were defined according to the guidelines of the International Society of Pediatric Oncology (SIOP).⁸ Patients were divided into two groups according to the absence or the presence of a visible residue before or after contrast injection on CT scan or MRI performed between day 2 or day 3 after surgery for the purposes of the analysis.

Treatment Regimens
Chemotherapy was to begin within a month of surgery and consisted of three courses (A, B, and C) of two different drugs administered in seven cycles. The interval between the first day of each course was 21 days. More than 0.8 x 10⁹ of granulocytes/L and more than 120 x 10⁹ of platelets/L were mandatory hematologic criteria for commencement of a new course. In course A, patients were administered carboplatin 15 mg/kg (450 mg/m²/d) in a 1-hour infusion on day 1, and procarbazine 4 mg/kg/d (120 mg/m²/d) orally on days 1 to 7. In course B, patients were administered etoposide 5 mg/kg/d (150 mg/m²/d) in a 1-hour infusion on days 22 and 23, and cisplatin 1 mg/kg/d (30 mg/m²/d) in a 3-hour infusion with mannitol plus saline on days 22 and 23. In course C, patients were administered vincristine 0.05 mg/kg (1.5 mg/m²) as an intravenous bolus on day 43, and cyclophosphamide 50 mg/kg (1,500 mg/m²) in a 1-hour infusion with uromitexan on day 43. For children over 3, doses were calculated in milligrams per meters squared. Cumulated dose levels were 3,150 mg/m² for carboplatin, 840 mg/m² for procarbazine, 2,100 mg/m² for etoposide, 420 mg/m² for cisplatin, 10.5 mg/m² for vincristine, and 10.5 g/m² for cyclophosphamide.

All patients were fitted with a central line and chemotherapy could be administered in an outpatient clinic. Patients were not to receive granulocyte colony-stimulating factor. The planned duration of chemotherapy was 16 to 18 months. Chemotherapy was discontinued if disease progression or unacceptable toxicity occurred. Doses were reduced by two thirds in case of hematologic toxicity, which postponed the start of the next course for more than 5 weeks. When residual disease remained stable after surgery, chemotherapy was continued. No irradiation was planned after the end of the chemotherapy.

Salvage therapy (including irradiation) was only indicated for disease progression or a relapse, either during chemotherapy or after the end of chemotherapy. The protocol specified that relapses were to be treated with a second operative procedure and involved-field radiotherapy at a dose of 50 Gy. Two opposed lateral beams encompassing the whole PF were used for PF tumors. All children had to be treated with megavoltage machines. High-energy photons (ie, > 15 MV) were also used in most of the centers for part of the treatment of PF tumors in order to minimize the dose received by superficial anatomic structures such as the scalp and middle ear. Conventional fractionation delivering 1.8 Gy in five weekly sessions was recommended. For tumors located below the foramen magnum, the total dose to the spinal cord was maintained below 45 Gy. The conventional technique used for irradiation of supratentorial tumors was at the discretion of the radiotherapist. The quality of treatment planning and delivery improved during the study because of the increased use of CT scan–based three-dimensional treatment planning and thermoplastic masks.

Craniospinal irradiation (CSI) was delivered exclusively in cases of proven distant spread, because we had previously reported that CSI prophylaxis was not mandatory for treatment of localized disease.⁹ Technical requirements included medulloblastoma-like field matching and the possibility of treating the spine with high-energy electrons.

Second-line chemotherapy (consisting of high-dose chemotherapy when feasible) could also be included in the salvage treatment according to center-specific policies. When high-dose chemotherapy was administered, the initiation of radiation therapy had to be delayed for a minimum of 70 days to avoid cumulative toxicity.

Evaluation Procedures
Disease extent at diagnosis was assessed by means of a spinal MRI study and CSF cytology. The neuroradiologic follow-up consisted of a cranial MRI study every 3 months until the second year after the diagnosis, then every 6 months. Audiograms were performed before each course containing cisplatin and every 2 years after the end of chemotherapy. Neurodevelopmental tests were scheduled to be administered by the clinician in charge of the patient postoperatively in neurologically stable children, then every 6 months during treatment. After treatment, a complete neuropsychologic evaluation with the Brunet-Lezine or Wechsler scales adapted to the age of the child was mandatory.

Radiologic postoperative residues and responses were evaluated by the individual investigators during chemotherapy and centrally reviewed retrospectively using the standard SIOP criteria.⁸ The central review analysis included at least imaging studies before chemotherapy, after one cycle of chemotherapy (ie, 3 months after surgery on average), and at the end of chemotherapy for all patients. The best response at any time after the start of chemotherapy was retained for evaluation. Disease progression was radiologically documented, with or without clinical signs.

Statistical Analyses
OS rates were estimated using the Kaplan-Meier method,¹⁰ from the first day of chemotherapy to death or the date of the last follow-up visit for patients who were still alive. Progression-free survival (PFS) rates were estimated from the first day of chemotherapy to the time of documented failure (date of the progression for patients whose disease progressed before achieving complete remission [CR], time of relapse or time of death for the others) or to the date of the last follow-up visit for those remaining in first CR. The radiotherapy-free survival rates were estimated using the Kaplan-Meier method from the first day of chemotherapy to the start of irradiation or to the date of death, both being considered as failures, or to the date of the last follow-up visit for patients who were still alive in first CR. The 95% confidence intervals (CIs) for survival rates were estimated using the Rothman method.¹¹ Follow-up data were updated as of June 1, 1999. Median follow-up was estimated with the Schemper method.¹² Statistical differences in OS and PFS were tested by the two-tailed log-rank test.¹³ Relative risks (RR) were estimated with their 95% CIs using a Cox model both in the univariate and in the multivariate analysis.¹⁴ Adjusted RRs were estimated in the final model including the two variables that attained significance in the multivariate analysis.

	RESULTS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Histology/Inclusion
Between June 1990 and December 1998, 406 infants and young children entered the French Society of Pediatric Oncology Baby Protocol 90. After a central review of all 406 histologic diagnoses by the panel of pathologists, 73 children with confirmed ependymoma were retained for the trial, and 11 patients with an initial diagnosis of ependymoma were excluded from the study because the diagnosis was rejected after review of the slides (six medulloblastoma/primitive neuroectodermal tumors, three astrocytomas, two choroid plexus carcinomas). Among the 73 children with a histologically confirmed ependymoma, seven children had initially been misdiagnosed as having other tumors (medulloblastoma/primitive neuroectodermal tumor in six and astrocytoma in one). This represents a mean accrual rate of 9.7 patients per year in this cooperative trial. Of the 68 cases in whom the histologic grade could be assigned, 88% were high grade and 12% were low grade according to the WHO classification. In five patients, grading of the ependymoma was not possible. In total, 82% of the patients had a high-grade tumor. The percentage of low-grade tumors was similar in both locations: six of 60 PF tumors and two of 13 supratentorial tumors were low-grade tumors.

Demographic Description
The median age at diagnosis was 27 months (range, 5 to 62 months). There were 40 boys and 33 girls.

Tumor Location and Metastases at Diagnosis
The tumor was supratentorial in 13 children and infratentorial (ie, in the PF) in the other 60 patients. Among the 60 patients with PF ependymoma, the tumor had extended to the pontocerebellar angle in 31 and remained strictly intraventricular in 25. In four patients, the absence of extension to the pontocerebellar angle could not be ruled out. No children with spinal ependymoma were entered onto this trial.

At diagnosis, distant spread, found in two patients, was diagnosed on the basis of positive CSF cytology. These two patients had a high-grade PF ependymoma. Both patients also had a postoperative residuum. In a third patient, spinal metastases were suspected on MRI but were not confirmed after the central review of the radiologic files.

Extent of Resection
Surgery, as assessed by the neurosurgeon, was complete in 44 (60%), subtotal in 18 (25%), and partial in 11. The extent of resection was evaluated by CT scan and MRI in 20, by MRI only in 30, and by CT scan only in 23. Residual tumor was present in 27 (37%) and absent in 46 (63%) patients on the early postoperative imaging studies.

In 14 children (19%), there was a discrepancy between the operative report and the results of the early postoperative scan. In six cases, surgery was apparently complete but a residue was depicted on the radiologic films. In eight cases, no residue was evidenced on CT scan or MRI, whereas the neurosurgeon had reported subtotal surgery.

There was no significant relationship between the tumor location and the extent of resection assessed by CT scan (P = .69); residual tumor was found in 46% of the supratentorial tumors, in 39% of the tumors located in the PF without involvement of the pontocerebellar angle, and in 32% of the tumors located in the PF involving the pontocerebellar angle.

Postoperative Residuum: Response to Chemotherapy
After a central review of all the radiologic follow-up MRI studies in the 27 children with a postoperative residue, no radiologic response to chemotherapy was found to exceed 50% at any time during chemotherapy (ie, no CR and no partial response). All the children exhibiting a radiologic residuum progressed ( Fig 3). In the two patients with leptomeningeal spread evidenced at CSF analysis only, the CSF was free of tumor cells after one cycle of chemotherapy, but the residual tumor in the PF did not respond to chemotherapy.

View larger version (17K): [in this window] [in a new window]   Fig 3. PFS according to risk factors.

View larger version (19K): [in this window] [in a new window]   Fig 1. OS, PFS, and radiotherapy-free survival of the entire population.

BBSFOP Chemotherapy Failures and Salvage Treatment
Progression or relapses were diagnosed in 53 patients, 2 to 45 months after the initiation of chemotherapy (median, 17 months). The first event was a local failure in 46 patients (87%), local and distant in three (6%), and only distant in four (7%). At the time of our analysis, 15 of these 53 patients were still in a second CR, with a median time after relapse of 45 months (range, 2 to 79 months). Only a complete second operative procedure followed by local irradiation permitted long-term survival. High-dose chemotherapy, administered to 20 patients, did not impact favorably on tumor response or survival as already published in a preliminary report.¹⁵ After a second relapse, none of the patients achieved a CR3 lasting more than 6 months.

At the time of the analysis, 31 patients had died, all of progressive disease, 3 months to 5.8 years after the initiation of chemotherapy (median, 29 months). Among the 42 survivors, 17 patients were in CR1 (40%), 15 in CR2 (36%), five had stable disease or had achieved a good partial response (three initial disease, two after relapse or progression), and five had progressive disease.

Radiation Therapy
At the time of the analysis, 39 patients had received radiation therapy. Only three patients received irradiation as part of their first-line treatment: two because the residuum remained stable at the end of chemotherapy (protocol violation) and one because the child’s parents refused further chemotherapy.

Among the 36 patients treated with radiotherapy as part of the salvage treatment for the first relapse, it was possible to defer irradiation for a median of 15 months (range, 5 to 46 months). Overall, 40% (95% CI, 29% to 51%) of the patients were alive 2 years after the initiation of chemotherapy without recourse to radiotherapy and 23% (95% CI, 14% to 35%) were still alive at 4 years without radiotherapy (Fig 1).

Prognostic Factor Analysis
The results of the univariate and multivariate analysis of OS are listed in Table 1. This protocol included a treatment strategy for relapses because we anticipated the likelihood of a significant number of relapses after first-line treatment without radiotherapy. However, the same prognostic factors emerged when they were evaluated for their correlation with PFS. Age and histologic grading were not associated with the risk of death.

Of note, the conclusion of the operative report regarding resection (partial, subtotal, or complete) was not significantly associated with the risk of death or relapse. This signifies that the radiologist’s conclusion should take precedence over that of the surgeon when findings are inconsistent. In this respect, it is noteworthy that among the eight patients who had no radiologically documented residue but a subtotal resection according to the neurosurgeon, only one had a poor outcome.

The results of the multivariate analysis allowed us to identify a very-high-risk subgroup of patients, namely, those with an incompletely resected PF tumor (29% of the population). Figures 2 and 3 show that this classification was relevant regarding both OS and PFS rates. Among these very-high-risk patients, disease relapsed or progressed in all but one, and 15 of them ultimately died, leading to a PFS rate of 5% (95% CI, 1% to 24%) at 2 years and an OS rate of 54% (95% CI, 34% to 73%) at 2 years and 16% (95% CI, 5% to 41%) at 4 years.

View larger version (16K): [in this window] [in a new window]   Fig 2. OS according to risk factors.

The BBSFOP chemotherapy regimen was stopped prematurely in 38 patients. This was because of progression or relapse in 32 patients, stable disease in three (two patients received local radiation therapy and one received second-line chemotherapy), parental refusal in two (one did not receive any further treatment, the other was treated with local radiation therapy), and one patient was lost to follow-up.

Among the 1,131 courses administered, 170 resulted in complications (15%). No toxicity-related death occurred. Neutropenia less than 500/mm³ occurred in 83 courses (7%): 36 after cyclophosphamide and vincristine, 34 after etoposide and cisplatin, and 13 after carboplatin and procarbazine. Infections occurred after 70 courses (21 without neutropenia). There were eight cases of septicemia (four in the absence of neutropenia). Only two children required a platelet transfusion and 35 courses were complicated by anemia necessitating RBC transfusion (mostly during the first three cycles). Five patients had an allergic reaction caused by procarbazine (skin rash) that did not necessitate withdrawal of the drug. No allergy to carboplatin was observed. Only one child lost more than 10% of body weight.

To date, no second tumors have occurred in this cohort. Neurodevelopment, as reported by the clinician, is normal in 65% and subnormal in 35% of the 26 long-term survivors evaluated. Complete neuropsychologic evaluation is ongoing in the different participating centers and a detailed analysis will soon be available.

	DISCUSSION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
Ependymoma arises predominantly in children under 5 years of age and represents a therapeutic challenge for neurosurgeons, radiotherapists, and oncologists. Radical surgery, when feasible, and adjuvant radiotherapy for incompletely resected tumors generally exert a favorable impact on outcome, but the role of chemotherapy has yet to be defined.¹⁶ We investigated the treatment of ependymomas without irradiation to avoid the deleterious neuropsychologic effects that are particularly pronounced in infants and young children.¹⁷

The finding of paramount importance in this study is that a significant proportion of children with ependymoma can be cured without radiotherapy. Indeed, 23% (95% CI, 14% to 35%) of the children in whom surgery was deemed complete on early imaging studies were alive without receiving radiotherapy 4 years after the diagnosis. Moreover, deferring radiotherapy at the time of relapse or progression did not compromise OS of the whole patient population. The analysis of available prognostic factors indicated that both the supratentorial location and the radiologically confirmed extent of resection were independently predictive of cure with this protocol. Thus, a subgroup of children in whom irradiation can be avoided in first-line treatment can be defined as those with a complete resection and no metastasis whatever the tumor site and whatever the age.

The 2-year PFS and OS rates are comparable to those reported for ependymoma in the baby Pediatric Oncology Group (POG) protocol,⁴ in which irradiation was delivered at the end of chemotherapy. Table 2 compares the 2-year results of the main brain tumor protocols for infants and young children.^4-5,18-21 Outcome analysis was limited to 2 years because follow-up did not exceed this duration in most of the reports. However, it should be emphasized that late relapses are often observed (ie, after 2 years) when updated results are provided with a corresponding decrease in PFS and OS. For example, the 5-year OS rate in the first POG study was only 40%, whereas it was 74% at 2 years²²; and the 5-year OS rate in the study of Mason was only around 40%, whereas it was 60% at 2 years.¹⁸ In our study, the 5-year OS rate of 52% (38% to 65%) is at least comparable with these two studies.

Responses of residual disease to conventional chemotherapy were disappointing, as were responses of relapsed tumors to high-dose chemotherapy.¹⁵ After central review of the radiologic file, no radiologic response was found to exceed 50%. With respect to response of ependymoma to adjuvant chemotherapy, published reports are conflicting. Some authors failed to obtain a significant response rate,^5,19,24 whereas the reverse was true for others.^4,20,25 This probably reflects how difficult it is to assess tumor response after surgery in these patients. However, intensification with high-dose multiagent chemotherapy seems pointless in these tumors.^15,26

The strategy applied in this trial allowed irradiation to be deferred for a median of 15 months (range, 5 to 46 months) in the 36 patients who ultimately had to receive salvage therapy. Postponing radiotherapy to this extent did not compromise the general outcome of the pediatric patient population treated in this protocol. It is also noteworthy that it was possible to avoid CSI in practically all the patients who relapsed, and this did not jeopardize the survival of these patients. Our trial therefore confirms that CSI is no longer mandatory in young children with localized ependymomas, whatever the histology and whatever the location, and this is in accordance with what is already shown in the literature.^9,21 We have recently shown that therapeutic strategies avoiding CSI have a favorable impact on the neuropsychologic outcome of children with PF tumors.⁶

The positive impact of complete surgery on OS that has been reported extensively in the literature was here confirmed by an analysis of risk factors for survival in a prospective trial.¹⁶ As the impact of surgery is of major import, we analyzed other factors possibly interfering with the extent of resection. Neither the supratentorial location nor involvement of the pontocerebellar angle was associated with the extent of resection. However, the supratentorial location was associated with a better outcome, whatever the extent of resection. Histologic features were not associated with outcome in our study. In the literature, standard grading criteria for ependymomas are controversial and their prognostic significance is debated.^27-29 Recently, in a multicentric compilation of clinicopathologic data from 11 centers, Horn et al found that WHO grade 3 ependymomas had a worse outcome even after adjustment of other risk factors.²⁷ As the number of children with low-grade ependymoma (WHO grade 2) is low, our prospective study is not sufficiently powered to address this question. Age was not found to have an influence on survival, supporting the hypothesis that tumor behavior is not related to age. Duffner et al reported that a young age had an adverse influence on survival in the first baby POG study.²² They attributed this phenomenon to variable treatment policies according to age, and particularly the longer deferral of irradiation in younger children. Our data are consistent with the latter statement. When we applied the same treatment policy to all the children, regardless of their age, we lost the prognostic weight of age in this population of children below 5 years of age. It is possible, however, that children over 5 years of age have a better prognosis, as suggested by the study by the Italian Pediatric Neuro-Oncology Group.²⁸

As expected, the chemotherapy regimen was well tolerated, because dose-intensity was low. Apart from concern for immediate tolerance, clinicians are increasingly preoccupied by the risk of second cancers in trials combining chemotherapy and irradiation.³⁰ To date, no second cancers have developed in this series. The lower cumulative doses of cyclophosphamide (10 g/m² v 30 g/m² in the baby POG protocol) and etoposide (2 g/m² v 3 g/m² in the baby POG protocol) used in this study could explain the limited incidence of second cancers. However, a longer follow-up in our study is necessary to confirm the lower risk of second malignancies with this protocol.

The preliminary data on the cognitive development of the patients in this study suggest that the strategy may impact favorably on the intellectual outcome of these young children compared with those treated with local irradiation alone in earlier studies.⁶ However, because accurate neuropsychologic evaluations are still awaited for most of the patients, this favorable impression will need to be confirmed in the future.

In conclusion, this study confirms that some patients with PF and supratentorial ependymomas can achieve long-term disease-free survival without irradiation. The children with PF ependymoma who are more likely to benefit from the therapeutic policy described in this protocol are those with a radiologically documented complete resection. Children with supratentorial ependymoma may benefit from the same policy even if the resection is not complete according to the imaging findings. Moreover, deferring radiotherapy until relapse does not compromise the survival of these young children. A benefit is obtained with prolonged but low-dose-intensity chemotherapy that nonetheless was incapable of eradicating residual disease. This raises the question of the need for adjuvant therapy in completely resected ependymomas advocated in single-institution studies.^31-33 Moreover, because two thirds of the children with standard-risk tumors relapsed, efforts should be directed at a better definition of children at a higher risk of relapse on the basis of tumor biologic features. We advocate a second surgical procedure for children with a postoperative residue, to achieve complete resection, given the major impact of this prognostic factor on survival. The systematic use of standard irradiation as part of first-line treatment for completely resected childhood ependymomas is no longer recommended, at least not for young children at high risk of radiation-induced neuropsychologic deficits.

APPENDIX
The following also participated in the study: F. Mechinaud (Nantes), M.C. Baranzelli (Lille), X. Rialland (Angers), A. Deville (Nice), C. Behar (Reims), P. Lutz (Strasbourg), J.P. Vannier (Rouen), J.F. Demeocq (Clermont-Ferrand), E. Plouvier (Besançon), H. Rubie (Toulouse), Y. Perel (Bordeaux), P. De Lumley (Limoges), E. Sariban (Brussels), and H. McDowell (Liverpool).

The members of the Neuropathology Subcommittee are as follows: M.M. Ruchoux, University Hospital, Lille; A. Lellouch-Tubiana, Necker Hospital, Paris; A. Jouvet, Pierre Wertheimer Hospital, Lyon; and D. Gambarelli, La Timone Hospital, Marseille, France.

	ACKNOWLEDGMENTS

 
Supported by grants from the Association Pour la Recherche sur le Cancer and the Lyons Club.

We thank Lorna Saint-Ange for editing the manuscript.

	REFERENCES

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION REFERENCES

 
1. Pollack IF: Current concepts: Brain tumors in children. N Engl J Med 331: 1500-1507, 1994[Free Full Text]

2. Duffner PK, Cohen ME, Myers MH, et al: Survival of children with brain tumors: SEER program, 1973-1980. Neurology 36: 597-601, 1986[Abstract/Free Full Text]

3. Suc E, Kalifa C, Brauner R, et al: Brain tumors under the age of three: The price of survival—A retrospective study of 20 long-term survivors. Acta Neurochir (Wien) 106: 93-98, 1990[Medline]

4. Duffner PK, Horowitz ME, Krischer JP, et al: Postoperative chemotherapy and delayed radiation in children less-than three years of age with malignant brain tumors. N Engl J Med 328: 1725-1731, 1993[Abstract/Free Full Text]

5. Geyer JR, Zeltzer PM, Boyett JM, et al: Survival of infants with primitive neurectodermal tumors or malignant ependymomas of the CNS treated with eight drugs in 1 day: A report from the Childrens’ Cancer Group. J Clin Oncol 12: 1607-1615, 1994[Abstract/Free Full Text]

6. Grill J, Kieffer-Renaux V, Bulteau C, et al: Long-term intellectual outcome in children with posterior fossa tumors according to radiation doses and volumes. Int J Radiat Oncol Biol Phys 45: 137-145, 1999[Medline]

7. Kleihues P, Cavenee WK. Pathology and Genetics of Tumours of the Nervous System. Lyon, France, International Agency for Research on Cancer, 1997

8. Gnekow A, on behalf of the SIOP Brain Tumor Subcommittee: Recommendations of the Brain Tumor Subcommittee for the reporting of trials. Med Pediatr Oncol 24: 104-108, 1995[Medline]

9. Rousseau P, Habrand JL, Sarrazin D, et al: Treatment of intracranial ependymomas of children: Review of a 15-year experience. Int J Radiat Oncol Biol Phys 28: 381-386, 1993

10. Kaplan EL, Meier P: Nonparametric estimation from incomplete observations. J Am Stat Assoc 53: 457-481, 1958

11. Rothman KJ: Estimation of confidence limits for the cumulative probability of survival in life table analysis. J Chronic Dis 31: 557-560, 1978[Medline]

12. Schemper M, Smith T: A note on quantifying follow-up in studies of failure time. Control Clin Trials 17: 343-346, 1996[Medline]

13. Mantel N: Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep 50: 163-170, 1966[Medline]

14. Cox DR: Regression models and life-tables. J R Stat Soc (B) 34: 187-220, 1972

15. Grill J, Kalifa C, Doz F, et al: A high-dose busulfan-thiotepa combination followed by autologous bone marrow transplantation in childhood recurrent ependymoma: A phase II study. Pediatr Neurosurg 25: 7-12, 1996[Medline]

16. Bouffet E, Perilongo G, Canete A, Massimino M: Intracranial ependymomas in children: A critical review of prognostic factors and a plea for cooperation. Med Pediatr Oncol 30: 319-331, 1998[Medline]

17. Ris MD, Noll RB: Long-term neurobehavioral outcome in pediatric brain tumor patients: Review and methodological critique. J Clin Exp Neuropsychol 16: 21-42, 1994[Medline]

18. Mason WP, Grovas A, Halpern S, et al: Intensive chemotherapy and bone marrow rescue for young children with newly diagnosed malignant brain tumors. J Clin Oncol 16: 210-221, 1998[Abstract/Free Full Text]

19. White L, Kellie S, Gray E, et al: Postoperative chemotherapy in children less than 4 years of age with malignant brain tumors: Promising initial response to a VETOPEC-based regimen. J Pediatr Hematol Oncol 20: 125-130, 1998[Medline]

20. Kühl J, Müller HL, Berthold F, et al: Preradiation chemotherapy of children and young adults with malignant brain tumors: Results of the German pilot trial HIT’88/’89. Klin Padiatr 210: 227-233, 1998[Medline]

21. Timmerman B, Kortmann RD, Kühl J, et al: Combined post-operative irradiation and chemotherapy for anaplastic ependymomas in childhood: Results of the German prospective trials HIT 88/89 and HIT 91. Int J Radiat Oncol Biol Phys 46: 287-295, 2000[Medline]

22. Duffner PK, Krischer JP, Sanford RA, et al: Prognostic factors in infants and very young children with intracranial ependymomas. Pediatr Neurosurg 28: 215-222, 1998[Medline]

23. Fisher PG, Needle MN, Cnaan A, et al: Salvage therapy after postoperative chemotherapy for primary brain tumors in infants and very young children. Cancer 83: 566-574, 1998[Medline]

24. Evans AE, Anderson JR, Lefkovitz-Boudreaux HB, et al: Adjuvant chemotherapy of childhood posterior fossa ependymoma: Cranio-spinal irradiation with or without adjuvant CCNU, vincristine and prednisone—A Childrens Cancer Group study. Med Pediatr Oncol 27: 8-14, 1996[Medline]

25. Needle MN, Goldwein JW, Grass J, et al: Adjuvant chemotherapy for the treatment of intracranial ependymoma of childhood. Cancer 80: 341-347, 1997[Medline]

26. Mason WP, Goldman S, Yates AJ, et al: Survival following intensive chemotherapy with bone marrow reconstitution for children with recurrent intracranial ependymoma. J Neurooncol 37: 135-143, 1998[Medline]

27. Horn B, Heideman R, Geyer R, et al: A multi-institutional retrospective study of intracranial ependymoma in children: Identification of risk factors. J Pediatr Hematol Oncol 21: 203-211, 1999[Medline]

28. Perilongo G, Massimino M, Sotti G, et al: Analyses of prognostic factors in a retrospective review of 92 children with ependymoma: Italian Pediatric Neuro-Oncology Group. Med Pediatr Oncol 29: 79-85, 1997[Medline]

29. Hamilton RL, Pollack IF: The molecular biology of ependymomas. Brain Pathol 7: 807-822, 1997[Medline]

30. Duffner PK, Krischer JP, Horowitz ME, et al: Second malignancies in young children with primary brain tumors following treatment with prolonged postoperative chemotherapy and delayed irradiation: A Pediatric Oncology Group study. Ann Neurol 44: 313-316, 1998[Medline]

31. Palma L, Celli P, Cantore G: Supratentorial ependymomas of the first two decades of life: Long-term follow-up of 20 cases. Neurosurgery 32: 169-175, 1993[Medline]

32. Awad YM, Allen JC, Miller DC, et al: Deferring adjuvant therapy for totally resected intracranial ependymoma. Pediatr Neurol 14: 216-219, 1996[Medline]

33. Hukin J, Epstein F, Lefton D, et al: Treatment of intracranial ependymoma by surgery alone. Pediatr Neurosurg 29: 40-45, 1998[Medline]

Submitted June 5, 2000; accepted November 14, 2000.

This article has been cited by other articles:

				L. Ridley, R. Rahman, M.-A. Brundler, D. Ellison, J. Lowe, K. Robson, E. Prebble, I. Luckett, R. J. Gilbertson, S. Parkes, et al. Multifactorial analysis of predictors of outcome in pediatric intracranial ependymoma Neuro-oncol, January 1, 2008; 10(5): 675 - 689. [Abstract] [Full Text] [PDF]

				U. Tabori, J. Ma, M. Carter, M. Zielenska, J. Rutka, E. Bouffet, U. Bartels, D. Malkin, and C. Hawkins Human Telomere Reverse Transcriptase Expression Predicts Progression and Survival in Pediatric Intracranial Ependymoma J. Clin. Oncol., April 1, 2006; 24(10): 1522 - 1528. [Abstract] [Full Text] [PDF]

				T. E. Merchant, R. K. Mulhern, M. J. Krasin, L. E. Kun, T. Williams, C. Li, X. Xiong, R. B. Khan, R. H. Lustig, F. A. Boop, et al. Preliminary Results From a Phase II Trial of Conformal Radiation Therapy and Evaluation of Radiation-Related CNS Effects for Pediatric Patients With Localized Ependymoma J. Clin. Oncol., August 1, 2004; 22(15): 3156 - 3162. [Abstract] [Full Text] [PDF]

				V. Laithier, J. Grill, M.-C. Le Deley, M.-M. Ruchoux, D. Couanet, F. Doz, F. Pichon, H. Rubie, D. Frappaz, J.-P. Vannier, et al. Progression-Free Survival in Children With Optic Pathway Tumors: Dependence on Age and the Quality of the Response to Chemotherapy--Results of the First French Prospective Study for the French Society of Pediatric Oncology J. Clin. Oncol., December 15, 2003; 21(24): 4572 - 4578. [Abstract] [Full Text] [PDF]

				S. Dyer, E. Prebble, V. Davison, P. Davies, P. Ramani, D. Ellison, and R. Grundy Genomic Imbalances in Pediatric Intracranial Ependymomas Define Clinically Relevant Groups Am. J. Pathol., December 1, 2002; 161(6): 2133 - 2141. [Abstract] [Full Text] [PDF]

				R. J. Gilbertson, L. Bentley, R. Hernan, T. T. Junttila, A. J. Frank, H. Haapasalo, M. Connelly, C. Wetmore, T. Curran, K. Elenius, et al. ERBB Receptor Signaling Promotes Ependymoma Cell Proliferation and Represents a Potential Novel Therapeutic Target for This Disease Clin. Cancer Res., October 1, 2002; 8(10): 3054 - 3064. [Abstract] [Full Text] [PDF]

				V. Kieffer-Renaux, C. Bulteau, J. Grill, C. Levy-Piebois, D. Couanet, A. Pierre-Kahn, O. Hartmann, and C. Kalifa Visual Agnosia After Treatment of a Posterior Fossa Ependymoma in a 16-Month-Old Girl J Child Neurol, September 1, 2001; 16(9): 698 - 704. [Abstract] [PDF]

				A. C. Paulino, J. M. Buatti, J. Grill, and C. Kalifa Adjuvant Chemotherapy for Ependymoma: Is It Necessary for All Children Under Five Years of Age? J. Clin. Oncol., August 1, 2001; 19(15): 3588 - 3589. [Full Text]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Grill, J. Articles by Kalifa, C. Search for Related Content PubMed PubMed Citation Articles by Grill, J. Articles by Kalifa, C.