Pediatric infratentorial ependymoma: prognostic ... - Springer Link

1 downloads 0 Views 318KB Size Report
Aug 24, 2011 - Ji Hoon Phi • Kyu-Chang Wang • Sung-Hye Park • Il Han Kim • In-One Kim •. Kyung Duk Park • Hyo Seop Ahn • Ji Yeoun Lee • Young-Je Son •.
J Neurooncol (2012) 106:619–626 DOI 10.1007/s11060-011-0699-x

CLINICAL STUDY – PATIENT STUDY

Pediatric infratentorial ependymoma: prognostic significance of anaplastic histology Ji Hoon Phi • Kyu-Chang Wang • Sung-Hye Park • Il Han Kim • In-One Kim Kyung Duk Park • Hyo Seop Ahn • Ji Yeoun Lee • Young-Je Son • Seung-Ki Kim



Received: 25 March 2011 / Accepted: 12 August 2011 / Published online: 24 August 2011 Ó Springer Science+Business Media, LLC. 2011

Abstract Pediatric infratentorial ependymomas are difficult to cure. Despite the availability of advanced therapeutic modalities for brain tumors, total surgical resection remains the most important prognostic factor. Recently, histological grade emerged as an independent prognostic factor for intracranial ependymoma. We retrospectively reviewed the treatment outcome of 33 pediatric patients with infratentorial ependymoma. Progression-free survival (PFS) and overall survival (OS) rates were calculated and

J. H. Phi  K.-C. Wang  J. Y. Lee  S.-K. Kim (&) Division of Pediatric Neurosurgery, Seoul National University Children’s Hospital, 101 Daehangno, Jongno-gu, Seoul 110-744, Republic of Korea e-mail: [email protected] S.-H. Park Department of Pathology, Seoul National University Hospital, Seoul, Republic of Korea I. H. Kim Department of Radiation Oncology, Seoul National University Hospital, Seoul, Republic of Korea I. H. Kim  H. S. Ahn Cancer Research Institute, Seoul National University College of Medicine, Seoul, Republic of Korea I.-O. Kim Department of Diagnostic Radiology, Seoul National University Hospital, Seoul, Republic of Korea K. D. Park  H. S. Ahn Department of Pediatrics, Seoul National University Children’s Hospital, Seoul, Republic of Korea Y.-J. Son Department of Neurosurgery, Seoul National University Boramae Medical Center, Seoul, Republic of Korea

relevant prognostic factors were analyzed. Fourteen patients (42%) were under the age of 3 at diagnosis. Gross total resection was achieved in 16 patients (49%). Anaplastic histology was found in 13 patients (39%). Adjuvant therapies were delayed until progression in 12 patients (36%). Actuarial PFS rates were 64% in the first year and 29% in the fifth year. Actuarial OS rates were 91% in the first year and 71% in the fifth year. On univariate analysis, brainstem invasion (P = 0.047), anaplastic histology (P = 0.004), higher mitotic count (P = 0.001), and higher Ki-67 index (P = 0.004) were significantly related to a shorter PFS. Gross total resection (P = 0.029) and a greater age at diagnosis (P = 0.033) were significantly related to a longer PFS. On multivariate analysis, anaplastic histology alone was significantly related to a shorter PFS (P = 0.023). Gross total resection (P = 0.039) was significantly related to a longer overall survival (OS) on multivariate analysis. Anaplastic histology and gross total resection were the most important clinical factors affecting PFS and OS, respectively. Anaplastic histology, mitotic count, and Ki-67 index can be used as universal and easily available prognostic parameters in infratentorial ependymomas. Keywords Ependymoma  Children  Prognosis  Survival  Anaplastic histology

Introduction Ependymomas are the third most common brain tumors in children, following medulloblastoma and astrocytic tumors. Although it is largely considered a brain tumor of intermediate grade, the morbidity and mortality associated with these diseases are substantial. Literature reviews show

123

620

that long-term survival rates of pediatric intracranial ependymoma patients are around 50–70% [1–4]. Considering that long-term survival of a highly malignant pediatric brain tumor, medulloblastoma, reaches 60–80% [5–7], the poor treatment outcome of ependymoma is perplexing, even to neurosurgeons and neuro-oncologists. Why are intracranial ependymomas so challenging to cure? Pediatric intracranial ependymomas predominantly arise from the fourth ventricle. Intraventricular ependymomas can involve the dorsal brainstem or extend into the cerebellopontine angle, prepontine or lateral medullary cisterns, encasing important cranial nerves and vessels. This can prevent a safe total removal of the tumor in many patients [8]. Furthermore, ependymomas are relatively insensitive to radiation and chemotherapeutic agents which made medulloblastoma, once a lethal cancer, a potentially curable disease since the 1960s [9, 10]. Therefore, incomplete surgical resection of ependymomas remains the most important prognostic factor in many studies, and the optimal roles for adjuvant therapies have been debated. Anaplastic ependymoma has more aggressive histological features than classic ependymoma. The World Health Organization (WHO) pathological grading system classifies anaplastic ependymoma as a grade III tumor, in contrast to classic ependymoma, which is classified as a grade II tumor [11]. However, the prognostic significance of anaplastic variant has been questioned [12–14]. Molecular and cytogenetic studies revealed that intracranial ependymomas are a heterogeneous group of tumors with variable biological characteristic and prognosis [15, 16]. In this regard, the prognostic value of histological grade has to be re-evaluated [2, 17]. Furthermore, some investigators have tried more intensive adjuvant treatments, such as high-dose chemotherapy with autologous stem cell transplantation for ependymoma, either primary or recurrent disease [18]. To justify applying intensive treatment modalities, we need sufficient data for clinical stratification of patients, some of whom may benefit from the more aggressive treatment. In this study, we reviewed the treatment outcome of 33 pediatric patients with posterior fossa ependymomas. We induced relevant prognostic factors, based on longitudinal data analyses. This study provides up-to-date information on the prognosis of intracranial ependymoma, especially regarding the patient stratification.

Patients and methods Data sources and patient selection The Institutional Review Boards of Seoul National University Hospital and Seoul National University College of

123

J Neurooncol (2012) 106:619–626

Medicine approved the study protocol. We reviewed our electronic databases of inpatients for children younger than 18 years who were diagnosed and treated with intracranial ependymoma. The databases were non-selectively constructed for all inpatients at Seoul National University Children’s Hospital and Boramae Municipal Hospital. We limited our search from January 1999 to August 2009, because the radiological and pathological data of patients enrolled before 1999 were not digitized, and many of them were unavailable for review. Also, we required a minimum 1-year follow-up period after initial pathological diagnosis, unless the patient expired before 1 year. We excluded patients with a diagnosis of subependymoma, supratentorial ependymoma, and ependymoma of the spinal cord because the prognosis for those diseases is different from that of infratentorial ependymoma. Thirty-six patients were retrieved from the databases. Two patients had insufficient follow-up periods (\1 year) after initial diagnosis, and one patient had an uncertain diagnosis. The remaining 33 patients were eligible for this study. We retrospectively reviewed the electronic medical records (EMR) of the selected patients. We also reviewed data from individual departments, including operative reports, radiation therapy records and chemotherapy sheets. The treatment outcome was confirmed with the EMR, and death certificate information was retrieved from the National Statistical Office and the Ministry of Public Administration and Security. Construction of database We developed a structured data extraction form, and clinical data of the patients were input via double entry. Preoperative and postoperative magnetic resonance images (MRIs) were reviewed independently by a neuroradiologist (Kim I.O.) and two other authors, and discrepancies were resolved by discussion of the case. Operative reports were reviewed. The presence of frank brainstem invasion into the fourth ventricular floor or medulla oblongata and remnant tumor that could not be removed was recorded. The degree of tumor resection was classified as gross total resection (GTR) and incomplete resection. GTR was defined as no visible enhancing/non-enhancing tumor on postoperative imaging, with the surgeon recording no remnant tumor on microscopic operative views. Pathological slides were reviewed by a neuropathologist (Park S.H.). If the original slides were unavailable for review, new slides of hematoxylin and eosin staining were prepared from the paraffin-block of the tumor tissue. Diagnosis of ependymoma and histological grading were made according to the criteria of WHO classification of brain tumors revised in 2007 in which high cellularity, nuclear pleomorphism, brisk mitosis, and/or microvascular

J Neurooncol (2012) 106:619–626

proliferation are indicative of grade III [11]. Mitotic count was calculated in sequential 10 high-power fields from the area of high cellularity. Ki-67 labeling index was measured by Aperio image analyser (Aperio; Vista, CA, USA) using a nuclear algorithm on 9200 scanning fields. The clinical data from the EMR were merged with the death certificate information. After a manual check of inconsistent cells, the database was locked for the analysis. Statistical analysis We calculated the actuarial progression-free survival (PFS) and overall survival (OS) using the Kaplan–Meier method. Disease progression was defined as appearance of a new lesion or growth of a previous tumor ([25% in 2-dimensional analyses) on follow-up MRI. All participants were followed up from the time of initial surgery (initial diagnosis) until the date of death or until September 1, 2010, whichever came first. Then, we compared the PFS and OS according to various clinical factors using a Cox proportional hazards model to define relevant prognostic factors. Patient’s sex, age at diagnosis, age under 3 years, tumor seeding at presentation, brainstem invasion, degree of resection, histological grade, mitotic count, Ki-67 index, and application of postoperative RT were evaluated for prognostic relevance. For multivariate analyses, we constructed a Cox proportional hazards model with clinical variables that had P values \0.1 on univariate analyses. The patient’s sex and age were incorporated into the model as basic variables and we selected variables with more clinical relevance when there was co-linearity between two variables (Pearson coefficient [0.5). For comparison of variables in the Kaplan–Meier curves, a log-rank test was used. For comparison of continuous variable between groups, Student’s t test was applied. P values were twosided and significance was set at 0.05. The SPSS 17.0 software (SPSS, Chicago, USA) was used in the statistical analyses.

621

Lumbar cerebrospinal fluid cytology was obtained from 8 patients and tumor cells were not found in any patient. Radical resection of the primary tumor was attempted in all patients. Brainstem invasion was found in 15 patients (52%; 15/29) based on the surgeon’s operation records. There was no record regarding brainstem invasion in 4 patients. GTR was achieved in 16 patients (49%). Incomplete resection was noted in 17 patients (52%). Pathological examination revealed WHO grade II ependymoma in 20 patients (61%) and WHO grade III (anaplastic) ependymoma in 13 (39%) patients. The median value of mitotic count was 2 (range 0–31) and the mean Ki-67 index was 8.8 (range 0–30). Progression-free and overall survival The treatment policy was different according to the patient’s age at diagnosis (Fig. 1). for patients over the age of 3 (n = 19), if initial surgical resection was complete (n = 8), local radiation therapy (RT) was indicated (n = 6) with the exception of 2 patients. If the surgical resection was incomplete (n = 11), local or craniospinal RT was given to all patients with the exception of 1 patient. Two patients received pre-radiation chemotherapy followed by

Results Clinical profile of the patients Overall, 15 patients were male and 18 patients were female. The mean age at diagnosis was 3.5 years (range 5 months to 12 years). Fourteen patients (42%) were under the age of 3 at diagnosis. All patients had an infratentorial ependymoma (30 patients in the fourth ventricle and 3 patients in the cerebellopontine angle). Perioperative spinal MRI was available for review for 27 patients, and 5 patients had seeding in the spinal subarachnoid space. No intracranial seeding was found on the initial brain MRI.

Fig. 1 a Treatment flow chart of the 19 patients who were over the age of 3 at diagnosis; 13 patients experienced tumor progression. b Treatment flow chart of 14 patients who were under the age of 3 at diagnosis; 11 patients had tumor progression

123

622

RT. For patients under the age of 3, if initial surgical resection was complete (n = 8), observation was indicated for the majority of patients (n = 6). If initial surgical resection was incomplete (n = 6), chemotherapy was indicated for 5 patients, but the parents of 2 patients refused further treatment and 1 patient could not receive chemotherapy because of surgical complications. RT was indicated for 1 patient (2.8 years at diagnosis and RT started at the age of 3 years). The mean follow-up period was 54 months (range 8–131 months). Tumor progression was observed in 24 patients (73%) during the follow-up period. Actuarial progression-free survival (PFS) rates were 64% in the first year, 36% in the second year, and 29% in the fifth year. The median PFS was 17 months [95% confidential interval (CI) 9.8–24.2 months] (Fig. 2). After tumor progression was documented, the treatment was highly individualized and based on previous treatment, respectability of the tumor, and clinical status of the patient. A second surgical resection was performed for 15 patients. Five patients received a third surgical resection. One patient received a fourth surgical resection. Conventional RT (6 patients), radiosurgery (7 patients), proton therapy (3 patients), chemotherapy (9 patients), and high-dose chemotherapy with autologous stem cell transplantation (2 patients) were also given. Overall, 21 patients (64%) received conventional RT during the treatment course. The extent of irradiation was local in 15 patients and craniospinal in 6 patients. The mean radiation dose for primary tumor site was 52.8 Gy (range 48.6–59.4 Gy). Fourteen patients (42%) received chemotherapy of various regimens during the treatment course: the Children’s Cancer Group 9921 regimen (1 patient) [19], eight-drug-in-a-day regimen (9 patients) [20],

J Neurooncol (2012) 106:619–626

the Korean Society for Pediatric Neuro-Oncology (KSPNO) regimens (5 patients) [21], and a regimen consisting of cyclophosphamide, vincristine, and methotrexate (1 patient). For the 5 patients with tumor seeding at presentation, 3 patients older than 3 years received craniospinal irradiation after surgery and 2 patients younger than 3 years received chemotherapy. Eighteen patients (55%) received a ventriculoperitoneal shunt and 7 patients (21%) received an endoscopic third ventriculostomy for hydrocephalus. Actuarial overall survival (OS) rates were 91% in the first year, 88% in the second year, and 71% in the fifth year (Fig. 2). Factors affecting the survival In univariate analysis with a Cox proportional hazards model for PFS, anaplastic histology [relative risk (RR) = 3.430, 95% CI = 1.476–7.970; P = 0.004], higher mitotic count (RR = 1.086, 95% CI = 1.037–1.139; P = 0.001), and higher Ki-67 index (RR = 1.064, 95% CI = 1.020–1.109; P = 0.004) were all significantly associated with a shorter PFS. Brainstem invasion was also significantly related to a shorter survival (RR = 2.581, 95% CI = 1.012–6.581; P = 0.047). GTR (RR = 0.392, 95% CI = 0.169–0.908; P = 0.029) and a greater age at diagnosis (RR = 0.756, 95% CI = 0.584–0.977; P = 0.033) were significantly related to a longer PFS. Age less than 3 years at diagnosis as a dichotomous variable showed no significant association with a PFS. Figure 3 and 4 illustrate the Kaplan–Meier curves for survival according to the degree of surgical resection and histological grade. Patients who experienced disease progression had a significant higher mitotic counts and Ki-67 indices in the initial tumor than patients without disease

Fig. 2 Kaplan–Meier plots for a progression-free survival (PFS) and b overall survival (OS) of 33 patients. Actuarial PFS rates were 64% in the first year and 29% in the fifth year. Actuarial OS rates were 91% in the first year and 71% in the fifth year

123

J Neurooncol (2012) 106:619–626

623

Fig. 3 Kaplan–Meier plots for progression-free survival (PFS) and b overall survival (OS) according to the degree of resection. A log-rank test showed that gross total resection was related to a PFS (P = 0.022), significantly, and b OS (P = 0.053), marginally

Fig. 4 Kaplan–Meier plots for progression-free survival (PFS) and b overall survival (OS) according to the histological grade. A log-rank test showed that histological grade was significantly associated with a PFS (P = 0.002), but not with b OS (P = 0.271). Patients who later

experienced disease progression had a significantly higher c mitotic count and d Ki-67 index in the initial tumor, compared with data from patients without disease progression (error bars standard deviations)

123

624

J Neurooncol (2012) 106:619–626

Table 1 Relative risks for shorter progression-free survival estimated with a Cox proportional hazards model Clinical factors

Univariate analysis

Multivariate analysis

P value

RR

95% CI

P value

RR

95% CI

Sex (male)

0.850

0.925

0.412–2.075

0.160

2.645

0.682–10.26

Age (continuous variable)

0.033

0.756

0.584–0.977

0.108

0.714

0.474–1.076

Age \3 years

0.325

1.502

0.668–3.375

NA

NA

NA

Seeding at presentation

0.061

2.708

0.954–7.685

0.428

1.963

0.370–10.41

Brainstem invasion

0.047

2.581

1.012–6.581

0.576

2.375

0.115–49.18

Anaplastic histology

0.004

3.430

1.476–7.970

0.023

6.354

1.288–31.35

Mitotic count Ki-67 index

0.001 0.004

1.086 1.064

1.037–1.139 1.020–1.109

NA NA

NA NA

NA NA

Gross total resection

0.029

0.392

0.169–0.908

0.968

1.060

0.060–18.84

Postoperative RT

0.160

0.557

0.246–1.260

NA

NA

NA

RR relative risk, CI confidence interval, RT radiation therapy, NA not available

Table 2 Relative risks for shorter overall survival estimated with a Cox proportional hazards model Clinical factors

Univariate analysis

Multivariate analysis

P value

RR

95% CI

P value

RR

95% CI

Sex (male)

0.610

1.410

0.377–5.276

0.281

2.284

0.508–10.26

Age (continuous variable)

0.156

0.709

0.440–1.141

0.078

0.602

0.342–1.058

Age \3 years

0.756

1.222

0.327–4.564

NA

NA

NA

Seeding at presentation

0.184

3.241

0.571–18.40

NA

NA

NA

Brainstem invasion

0.105

3.828

0.757–19.36

NA

NA

NA

Anaplastic histology

0.281

2.114

0.542–8.249

NA

NA

NA

Mitotic count

0.176

1.057

0.975–1.147

NA

NA

NA

Ki-67 index

0.431

1.035

0.951–1.126

NA

NA

NA

Gross total resection

0.074

0.235

0.048–1.149

0.039

0.178

0.034–0.918

Postoperative RT

0.292

0.492

0.131–1.840

NA

NA

NA

RR relative risk, CI confidence interval, RT radiation therapy, NA not available

progression (mitotic count: 7.3 ± 9.2 vs. 0.9 ± 1.1, P = 0.004; Ki-67: 13.1 ± 9.4 vs. 1.2 ± 2.4, P \ 0.001; Student’s t test) (Fig 4). Because anaplastic histology of ependymoma incorporates high proliferative activity in definition, anaplastic histology, mitotic count, and Ki-67 index can be regarded as representing slightly different aspects of the same phenomenon. Pearson correlation coefficients between these variables were all [0.5. Therefore, we included only anaplastic histology in the multivariate analysis among the three variables. In the multivariate analysis, the anaplastic histology alone was significantly related to a shorter PFS (RR = 6.354, 95% CI = 1.288–31.35; P = 0.023) (Table 1). In univariate analysis for OS, no single factor was significantly associated with OS. GTR was marginally related to a longer OS. GTR was the single significant prognostic factor for longer OS in multivariate analysis (RR = 0.178, 95% CI = 0.034–0.918; P = 0.039) (Table 2).

123

Discussion In this study, we reviewed the treatment outcome of 33 pediatric patients with infratentorial ependymoma, 42% of whom were under the age of 3 at diagnosis. Thirty-nine percent of patients had anaplastic ependymoma. GTR of the tumor was possible in half the patients at the initial surgery. The 5-year PFS rate was 29% and the 5-year OS rate was 71%. Anaplastic histology and gross total resection were the most important clinical factors affecting PFS and OS, respectively. The expected PFS was shorter with repeated radical resections. Ependymoma remains as one of the most challenging tumors in pediatric neuro-oncology. Despite the advances in surgery and adjuvant treatments, improvement of treatment outcomes has been stagnant for intracranial ependymoma. The 5-year PFS rates after initial surgery ranged from 26% for babies aged \3 years [22] to 58% for mixed

J Neurooncol (2012) 106:619–626

children and adults [23]. The reported 5-year OS rates ranged from 54 [24] to 73% [3]. The great discrepancy in treatment outcome results from different treatment periods and patient heterogeneity as well as from different treatment protocols regarding the application of adjuvant treatments. As for medulloblastoma, the treatment protocol has been highly standardized according to the perioperative risk stratification. However, there is substantial heterogeneity for the treatment of infratentorial ependymoma in terms of adjuvant therapy modalities, radiation doses, chemotherapy regimens, and indication of each modality [25]. This heterogeneity of treatment protocols largely stems from the fact that ependymoma is relatively resistant to adjuvant therapies, although radiation has a modest effect on the tumor [10, 25]. Medulloblastoma is relatively sensitive to adjuvant therapies and sufficiently small residual tumors can be handled by craniospinal radiation with or without chemotherapy. In contrast, residual ependymoma, especially on the dorsal brainstem, cannot be easily eradicated by adjuvant therapies, and the role of complete surgical resection has been emphasized. The relatively low PFS rate in this study may be caused by the high proportion of young patients (age \3 years; 42%) and by the high rate of delayed adjuvant therapies after surgery (36%). On the other hand, the relatively high OS rate may be attributed to active application of local control modalities (surgery and radiosurgery) after progression. However, a recent study from St. Jude Children’s Research Hospital showed that, after surgery and postoperative conformal radiotherapy for pediatric intracranial ependymoma, the 7-year event-free survival and OS rates were 69 and 81%, respectively [26]. In the study, the minimum required patient’s age was 12 months in the early phase, and there was no age limitation for radiation in the later phase of the series. Therefore, the value of aggressive local treatment should be emphasized in treatment of ependymoma, and mandatory postoperative adjuvant RT should be considered, even if the patient is under the age of 3. A substantial proportion of infratentorial ependymomas develops in children under the age of 3 (42% in our study), and proactive adjuvant treatment is highly recommended for this patient group. High precision radiation modalities, such as intensity-modulated radiotherapy, proton therapy, and radiosurgery may also reduce deleterious effects of radiation on immature brains [27, 28]. The prognostic significance of histological grade in intracranial ependymoma has been debated. In the past, several authors reported that there was no or little significant difference in survival between grade II and grade III (anaplastic) ependymoma [12–14]. However, the importance of histological grade of ependymoma is being emphasized in many recent studies [2, 17, 29]. High mitotic count and Ki-67 index are indicative of high proliferative

625

activity of tumors and have been associated with poor prognosis in intracranial ependymoma [30]. Some authors have reported that anaplastic histology in intracranial ependymoma was associated with specific chromosomal variation and also with poor prognosis [15, 31]. Our study also showed that anaplastic histology is a strong predictor of tumor progression regardless of the degree of resection. Although there are enormous interests on the molecular staging of intracranial ependymoma, currently no specific tumor marker has been validated for a reliable predictor of prognosis and therapeutic response, such as hypermethylation of O6-methyl-guanine-DNA methyltransferase in glioblastoma or 1p/19q loss in oligodendroglioma [32, 33]. Therefore, anaplastic histology along with mitotic count and Ki-67 index can be used as universal and easily available prognostic parameters in infratentorial ependymoma. Acknowledgment This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MEST) (2008–0061821).

References 1. Tamburrini G, D’Ercole M, Pettorini BL, Caldarelli M, Massimi L, Di Rocco C (2009) Survival following treatment for intracranial ependymoma: a review. Childs Nerv Syst 25:1303–1312. doi:10.1007/s00381-009-0874-y 2. Shu HK, Sall WF, Maity A, Tochner ZA, Janss AJ, Belasco JB, Rorke-Adams LB, Phillips PC, Sutton LN, Fisher MJ (2007) Childhood intracranial ependymoma: twenty-year experience from a single institution. Cancer 110:432–441. doi:10.1002/cncr. 22782 3. van Veelen-Vincent ML, Pierre-Kahn A, Kalifa C, Sainte-Rose C, Zerah M, Thorne J, Renier D (2002) Ependymoma in childhood: prognostic factors, extent of surgery, and adjuvant therapy. J Neurosurg 97:827–835. doi:10.3171/jns.2002.97.4.0827 4. McGuire CS, Sainani KL, Fisher PG (2009) Both location and age predict survival in ependymoma: a SEER study. Pediatr Blood Cancer 52:65–69. doi:10.1002/pbc.21806 5. Krischer JP, Ragab AH, Kun L, Kim TH, Laurent JP, Boyett JM, Cornell CJ, Link M, Luthy AR, Camitta B (1991) Nitrogen mustard, vincristine, procarbazine, and prednisone as adjuvant chemotherapy in the treatment of medulloblastoma. A Pediatric Oncology Group study. J Neurosurg 74:905–909 6. Evans AE, Jenkin RD, Sposto R, Ortega JA, Wilson CB, Wara W, Ertel IJ, Kramer S, Chang CH, Leikin SL et al (1990) The treatment of medulloblastoma. Results of a prospective randomized trial of radiation therapy with and without CCNU, vincristine, and prednisone. J Neurosurg 72:572–582 7. Gandola L, Massimino M, Cefalo G, Solero C, Spreafico F, Pecori E, Riva D, Collini P, Pignoli E, Giangaspero F, Luksch R, Berretta S, Poggi G, Biassoni V, Ferrari A, Pollo B, Favre C, Sardi I, Terenziani M, Fossati-Bellani F (2009) Hyperfractionated accelerated radiotherapy in the Milan strategy for metastatic medulloblastoma. J Clin Oncol 27:566–571. doi:10.1200/jco. 2008.18.4176 8. Ikezaki K, Matsushima T, Inoue T, Yokoyama N, Kaneko Y, Fukui M (1993) Correlation of microanatomical localization with postoperative survival in posterior fossa ependymomas. Neurosurgery 32:38–44

123

626 9. Ginguene C, Champier J, Maallem S, Strazielle N, Jouvet A, Fevre-Montange M, Ghersi-Egea JF (2010) P-glycoprotein (ABCB1) and breast cancer resistance protein (ABCG2) localize in the microvessels forming the blood-tumor barrier in ependymomas. Brain Pathol 20:926–935. doi:10.1111/j.1750-3639.2010. 00389.x 10. Kim SK, Lim SY, Wang KC, Kim YY, Chi JG, Choi YL, Shin HJ, Cho BK (2004) Overexpression of cyclooxygenase-2 in childhood ependymomas: role of COX-2 inhibitor in growth and multi-drug resistance in vitro. Oncol Rep 12:403–409 11. Louis DN, Ohgaki H, Wiestler OD, Cavenee WK (2007) WHO Classification of Tumours of the Central Nervous System. International Agency for Research on Cancer, Lyon 12. Perilongo G, Massimino M, Sotti G, Belfontali T, Masiero L, Rigobello L, Garre L, Carli M, Lombardi F, Solero C, Sainati L, Canale V, del Prever AB, Giangaspero F, Andreussi L, Mazza C, Madon E (1997) Analyses of prognostic factors in a retrospective review of 92 children with ependymoma: Italian Pediatric Neurooncology Group. Med Pediatr Oncol 29:79–85 13. Pollack IF, Gerszten PC, Martinez AJ, Lo KH, Shultz B, Albright AL, Janosky J, Deutsch M (1995) Intracranial ependymomas of childhood: long-term outcome and prognostic factors. Neurosurgery 37:655–666 14. Robertson PL, Zeltzer PM, Boyett JM, Rorke LB, Allen JC, Geyer JR, Stanley P, Li H, Albright AL, McGuire-Cullen P, Finlay JL, Stevens KR Jr, Milstein JM, Packer RJ, Wisoff J (1998) Survival and prognostic factors following radiation therapy and chemotherapy for ependymomas in children: a report of the Children’s Cancer Group. J Neurosurg 88:695–703. doi: 10.3171/jns.1998.88.4.0695 15. Korshunov A, Witt H, Hielscher T, Benner A, Remke M, Ryzhova M, Milde T, Bender S, Wittmann A, Schottler A, Kulozik AE, Witt O, von Deimling A, Lichter P, Pfister S (2010) Molecular staging of intracranial ependymoma in children and adults. J Clin Oncol 28:3182–3190. doi:10.1200/jco.2009.27.3359 16. Modena P, Lualdi E, Facchinetti F, Veltman J, Reid JF, Minardi S, Janssen I, Giangaspero F, Forni M, Finocchiaro G, Genitori L, Giordano F, Riccardi R, Schoenmakers EF, Massimino M, Sozzi G (2006) Identification of tumor-specific molecular signatures in intracranial ependymoma and association with clinical characteristics. J Clin Oncol 24:5223–5233. doi:10.1200/jco.2006.06. 3701 17. Korshunov A, Golanov A, Sycheva R, Timirgaz V (2004) The histologic grade is a main prognostic factor for patients with intracranial ependymomas treated in the microneurosurgical era: an analysis of 258 patients. Cancer 100:1230–1237. doi:10.1002/ cncr.20075 18. Thorarinsdottir HK, Rood B, Kamani N, Lafond D, Perez-Albuerne E, Loechelt B, Packer RJ, MacDonald TJ (2007) Outcome for children \4 years of age with malignant central nervous system tumors treated with high-dose chemotherapy and autologous stem cell rescue. Pediatr Blood Cancer 48:278–284. doi: 10.1002/pbc.20781 19. Geyer JR, Sposto R, Jennings M, Boyett JM, Axtell RA, Breiger D, Broxson E, Donahue B, Finlay JL, Goldwein JW, Heier LA, Johnson D, Mazewski C, Miller DC, Packer R, Puccetti D, Radcliffe J, Tao ML, Shiminski-Maher T (2005) Multiagent chemotherapy and deferred radiotherapy in infants with malignant brain tumors: a report from the Children’s Cancer Group. J Clin Oncol 23:7621–7631 20. Geyer JR, Zeltzer PM, Boyett JM, Rorke LB, Stanley P, Albright AL, Wisoff JH, Milstein JM, Allen JC, Finlay JL et al (1994) Survival of infants with primitive neuroectodermal 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

123

J Neurooncol (2012) 106:619–626 21. Kim SY, Sung KW, Hah JO, Yoo KH, Koo HH, Kang HJ, Park KD, Shin HY, Ahn HS, Im HJ, Seo JJ, Lim YJ, Lee YH, Shin HJ, Lim DH, Cho BK, Ra YS, Choi JU (2010) Reduced-dose craniospinal radiotherapy followed by high-dose chemotherapy and autologous stem cell rescue for children with newly diagnosed high-risk medulloblastoma or supratentorial primitive neuroectodermal tumor. Korean J Hematol 45:120–126 22. Massimino M, Gandola L, Barra S, Giangaspero F, Casali C, Potepan P, Di Rocco C, Nozza P, Collini P, Viscardi E, Bertin D, Biassoni V, Cama A, Milanaccio C, Modena P, Balter R, Tamburrini G, Peretta P, Mascarin M, Scarzello G, Fidani P, Milano GM, Sardi I, Genitori L, Garre ML (2010) Infant ependymoma in a 10 year AIEOP (Associazione Italiana Ematologia Oncologia Pediatrica) experience with omitted or deferred radiotherapy. Int J Radiat Oncol Biol Phys. doi:10.1016/j.ijrobp.2010.02.048 23. Mansur DB, Perry A, Rajaram V, Michalski JM, Park TS, Leonard JR, Luchtman-Jones L, Rich KM, Grigsby PW, Lockett MA, Wahab SH, Simpson JR (2005) Postoperative radiation therapy for grade II and III intracranial ependymoma. Int J Radiat Oncol Biol Phys 61:387–391. doi:10.1016/j.ijrobp.2004.06.002 24. Jaing TH, Wang HS, Tsay PK, Tseng CK, Jung SM, Lin KL, Lui TN (2004) Multivariate analysis of clinical prognostic factors in children with intracranial ependymomas. J Neurooncol 68:255–261 25. Zacharoulis S, Moreno L (2009) Ependymoma: an update. J Child Neurol 24:1431–1438. doi:10.1177/0883073809339212 26. Merchant TE, Li C, Xiong X, Kun LE, Boop FA, Sanford RA (2009) Conformal radiotherapy after surgery for paediatric ependymoma: a prospective study. Lancet Oncol 10:258–266 27. MacDonald SM, Safai S, Trofimov A, Wolfgang J, Fullerton B, Yeap BY, Bortfeld T, Tarbell NJ, Yock T (2008) Proton radiotherapy for childhood ependymoma: initial clinical outcomes and dose comparisons. Int J Radiat Oncol Biol Phys 71:979–986. doi: 10.1016/j.ijrobp.2007.11.065 28. Kano H, Niranjan A, Kondziolka D, Flickinger JC, Lunsford LD (2009) Outcome predictors for intracranial ependymoma radiosurgery. Neurosurgery 64:279–287. doi:10.1227/01.neu.0000338257. 16220.f7 29. Tihan T, Zhou T, Holmes E, Burger PC, Ozuysal S, Rushing EJ (2008) The prognostic value of histological grading of posterior fossa ependymomas in children: a Children’s Oncology Group study and a review of prognostic factors. Mod Pathol 21: 165–177. doi:10.1038/modpathol.3800999 30. Zamecnik J, Snuderl M, Eckschlager T, Chanova M, Hladikova M, Tichy M, Kodet R (2003) Pediatric intracranial ependymomas: prognostic relevance of histological, immunohistochemical, and flow cytometric factors. Mod Pathol 16:980–991. doi: 10.1097/01.mp.0000087420.34166.b6 31. Mendrzyk F, Korshunov A, Benner A, Toedt G, Pfister S, Radlwimmer B, Lichter P (2006) Identification of gains on 1q and epidermal growth factor receptor overexpression as independent prognostic markers in intracranial ependymoma. Clin Cancer Res 12:2070–2079. doi:10.1158/1078-0432.ccr-05-2363 32. Hegi ME, Diserens AC, Gorlia T, Hamou MF, de Tribolet N, Weller M, Kros JM, Hainfellner JA, Mason W, Mariani L, Bromberg JE, Hau P, Mirimanoff RO, Cairncross JG, Janzer RC, Stupp R (2005) MGMT gene silencing and benefit from temozolomide in glioblastoma. N Engl J Med 352:997–1003. doi: 10.1056/NEJMoa043331 33. Ramirez C, Bowman C, Maurage CA, Dubois F, Blond S, Porchet N, Escande F (2010) Loss of 1p, 19q, and 10q heterozygosity prospectively predicts prognosis of oligodendroglial tumors— towards individualized tumor treatment? Neuro Oncol 12:490–499. doi:10.1093/neuonc/nop071