SEARCH

SEARCH BY CITATION

Keywords:

  • ependymoma;
  • histology;
  • malignancy grade;
  • prognosis;
  • radiotherapy

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

BACKGROUND

Ependymomas account for 3–5% of all intracranial malignancies and occur most often in children and young adults. These neoplasms continue to generate considerable controversy with regard to their rational clinical management. It has been shown that the histologic classification of ependymoma is a significant predictor of clinical outcome in patients with ependymoma.

METHODS

Ependymomas from 258 patients who underwent microsurgery at a single institution were evaluated histologically to elucidate the prognostic utility of a recently proposed grading scheme. Pathologic and clinical data then were compared using univariate and multivariate analyses.

RESULTS

Increasing grade of ependymoma malignancy was found to be associated strongly and independently with worse clinical outcomes in terms of both event-free survival and overall survival. The effect of radiotherapy also was found to be related to histologic grade and was more beneficial for patients who had anaplastic ependymomas and had undergone complete tumor removal.

CONCLUSIONS

The application of a uniform diagnostic criteria for grading ependymomas highlighted the key role of tumor histology in clinical outcome in a cohort of patients who were treated in the microsurgical era. The recently proposed grading scheme is likely to be practically useful, reproducible, and clinically applicable. Cancer 2004. © 2004 American Cancer Society.

Ependymomas account for 3–5% of all intracranial malignancies and occur most often in children and young adults, constituting the third most common brain tumor entity in those populations.1 These neoplasms continue to generate considerable controversy with regard to their rational clinical management.2–4 Patients with spinal ependymomas usually have a favorable prognosis after undergoing macroscopic total tumor resection. Outcomes for patients with intracranial tumors are significantly worse, and the 5-year survival rate does not exceed 70%, regardless of the treatment regimen. The predominant reason for a dismal ependymoma prognosis is disease recurrence, and approximately 95% of deaths in patients with ependymoma are attributed to local tumor progression.

The prognostic significance of various parameters, including patient age, tumor location, volume of resection, and adjuvant therapy protocol, has been examined extensively in patients with ependymomas.5–14 Special attention also has been paid to histologic malignancy in ependymomas, and there remains controversy regarding whether this is a prognostic factor. It was noted fairly by Merchant et al.15 that the influence of tumor grade on the outcome of patients with intracranial ependymomas will remain controversial until uniform diagnostic criteria have been more widely accepted. Those authors recently proposed their own intriguing approach to the histologic grading of intracranial ependymomas that already has shown a significant association with clinical outcome in 50 pediatric tumors that were treated with radiotherapy. The main objective of the current study was to elucidate the real prognostic utility of the grading scheme of Merchant et al. by analyzing a series of 258 intracranial ependymomas in patients who were treated in a single institution within the last decade.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Patient Population and Treatment Modalities

Two hundred fifty-eight patients with newly diagnosed intracranial ependymomas who were treated in our institute from January 1, 1990 to January 1, 2001 were studied retrospectively, including 115 children age < 16 years and 143 adult patients. The mean patient age was 23 years (range, 1–75 years), and there were 139 males and 119 females.

According to the preoperative computed tomography (CT) and/or magnetic resonance imaging (MRI) findings, the tumor location was infratentorial in 164 patients and supratentorial in the remaining 94 patients. Tumor resection using a microneurosurgical technique was performed as completely as possible. There were no perioperative deaths in the current series. Evaluation of the extent of resection was based on data from postoperative contrast CT scan (143 patients) or MRI (115 patients). Macroscopic total tumor resection was achieved in 121 patients, and the remaining 137 patients underwent subtotal resection.

Megavoltage external beam irradiation was applied in 161 patients. For supratentorial tumors, the treatment volume was to encompass the tumor bed only. All infratentorial tumors were treated with application by the extending fields, including the upper cervical segments of the spinal cord. Radiotherapy was initiated 2–3 weeks after surgery. Total radiation doses varied from 52 grays (Gy) to 64 Gy (mean, 58 Gy). Conventional fractioning was used for all irradiated patients, with 1.6–2.0 Gy per fraction. For the remaining 97 patients, postoperative radiotherapy was not performed for various reasons, including age < 3 years, poor neurologic status after surgery, the discretion of the neurosurgeon, and patient or parental refusal of further treatment. Fifty-six patients had received chemotherapy with cisplatin, lomustin, and vincristine. Among them, 49 patients received maintenance chemotherapy that was initiated 6–12 weeks after the end of radiotherapy. The remaining seven patients received postoperative chemotherapy instead of irradiation.

The endpoint of the follow-up analysis was August 1, 2003, and the dates of death, recurrence, or last contact were considered the end of the study. The mean follow-up was 64 months (range, 7–143 months). The time of recurrence was determined as the date when tumor progression was confirmed either by CT scan or MRI. Data on survival are shown as both the median survival and the 5-year survival rates. Progression-free survival (PFS) and overall survival (OS) were estimated separately.

Pathologic Examination

In all patients, a diagnosis of ependymoma was made according to the World Health Organization classification of tumors of the central nervous system.1 In addition, all tumors were graded either as low-grade (well differentiated) or high-grade (anaplastic) according to the criteria presented by Merchant et al..15 Two independent referees (A.K. and R.S.) were unaware of the clinical data. The following criteria were used as patterns of anaplasia: increased cellularity compared with neighboring, less cellular regions (Fig. 1A); atypia (i.e., clusters of cells with increased nuclear-to-cytoplasmic ratio and coarse chromatin); and microvascular proliferation. The latter was determined as multilayered, mitotically active endothelial cells together with proliferating smooth muscle cells and pericytes (Fig. 1B–D). Ependymoma was diagnosed as anaplastic when the tumor met even one of the three histologic criteria described above.

thumbnail image

Figure 1. The histopathologic patterns of ependymoma malignancy. (A) An example of frankly increased cellularity accompanied with atypia, brisk mitotic activity, and small necrotic foci. (B) An example of microvascular proliferation with glomeruloid formation. (C) An endothelial component of the microvascular proliferation (immunohistochemistry with the CD31 antibody; labeled streptavidin-biotin [LSAB]; LSAB® kit; Dako Corporation, Carpinteria, CA). (D) A smooth muscle component of the microvascular proliferation (immunohistochemistry with the smooth muscle alpha actin antibody; LSAB® kit; Dako Corporation, CA). Original magnification × 200 (A); × 400 (B–D).

Download figure to PowerPoint

Statistical Analysis

For categoric data, the chi-square test was used. Survival analyses from the date of surgery were estimated with the Kaplan–Meier method. Comparisons among various patient subgroups were performed using the log-rank test. Multivariate analysis for survival was performed using the Cox proportional hazard models. Probability (P) values < 0.05 were considered significant. A significant correlation between 2 parameters was used as the 95% confidence interval.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Pathologic Findings

All 258 tumors examined had a histopathologic appearance typical for cellular ependymoma, as described elsewhere.1, 16, 17 A group of 127 ependymomas (47%) was identified as high grade (anaplastic). Among those high-grade tumors, 104 tumors contained all histologic patterns of malignancy, 15 tumors exhibited isolated microvascular proliferation, and the 2 remaining tumors demonstrated only increased cellularity and atypia. The other 131 intracranial ependymomas (53%) revealed no histologic patterns of malignancy and subsequently were diagnosed as low grade (well differentiated). The two referees agreed on an estimate of tumor malignancy grade in 255 tumors (99%), and a consensus on the final diagnoses for the 3 remaining tumors was reached after a cooperative evaluation.

High-grade tumors prevailed among pediatric patients (58% compared with 42% for adults; P = 0.03, chi-square test) as well as among patients who had ependymomas located above the tentorium (60% compared with 43% for infratentorial tumors; P = 0.01, chi-square test). Nonetheless, there was no correlation between patient age and tumor location.

Follow-Up Dta and Patterns of Failure

At the endpoint of the follow-up analysis, 145 patients (56%) were free of disease within 32–143 months after surgery (median, 89 months). Disease progression had developed in the other 113 patients (44%), and PFS varied from 4 months to 97 months (median, 38 months). At the end of the study, 54 patients (21%) with recurrent tumors had died within 7–103 months after surgery (median OS, 56 months). The 5-year PFS (± standard error) was 61% ± 9.3%, and the 5-year OS was 82% ± 14.4%.

Disease progressed as an isolated local recurrence in 90 patients (35% of patients; 80% of failures). Disease disseminated within the central nervous system in 18 patients (7% of patients; 16% of failures). Three patients developed isolated distant lesions, and the remaining 15 patients developed combined distant and local failures. In addition, five supratentorial tumors progressed as both local regrowth and extraneural metastases either in scalp soft tissues (two patients) or cervical lymph nodes (three patients). All 23 metastatic ependymomas were high grade and developed in patients who had undergone incomplete resection.

Survival Analysis

Table 1 summarizes the correlations between the estimated variables and clinical outcomes. In addition, Tables 2 and 3 show findings for ependymomas with various histologic grades examined separately. For the entire cohort of patients, the 5-year PFS was reduced significantly for pediatric patients, patients with supratentorial tumor location, patients who underwent incomplete resection, and patients who had tumors with high-grade histology (Fig. 2A). The 5-year OS rate was associated closely with patient age, volume of tumor removal, and ependymoma grade (Fig. 2B). Patients who were treated with irradiation showed no differences in outcome compared with patients who were not treated with irradiation. In addition, chemotherapy did not alter the survival rates. Nevertheless, when the groups with tumors of different histologic grades were examined separately, the receipt of irradiation resulted in a lengthening of both PFS and OS for patients with anaplastic ependymomas. Furthermore, patients who had undergone complete tumor resection followed by radiotherapy demonstrated 5-year OS rates that were comparable to the rates calculated for their counterparts with low-grade tumors (Fig. 3). The receipt of chemotherapy also was associated with prolonged OS for patients with malignant tumors.

Table 1. Factors that Influenced Survival in a Cohort of 258 Patients with Ependymomas
Parameter estimateNo. of patientsNo. of recurrences (%)Five-yr PFS (%)P valueNo. of deaths (%)5-yr OS (%)P value
  • PFS: progression-free survival; OS: overall survival; NS: not significant.

  • a

    The difference was significant only for the comparison between children (age < 16 years) and adults.

Age (yrs)       
 ≤ 314 7 (50)51 4 (29)68
 > 3 to < 1610152 (51)5228 (28)70
 ≥ 1614354 (38)640.01a22 (15)870.01a
Tumor location       
 Infratentorial16461 (37)6727 (17)83
 Supratentorial9452 (55)440.00327 (29)77NS
Tumor grade       
 Low-grade13122 (17)9010 (8)93
 High-grade12791 (72)27< 0.000144 (35)61<0.0001
Type of resection       
 Macroscopic total12139 (32)7019 (16)86
 Subtotal13774 (54)450.000135 (26)750.03
Radiotherapy results       
 Positive16177 (48)6238 (23)80
 Negative9736 (38)59NS16 (17)80NS
Chemotherapy results       
 Positive5621 (38)5711 (20)78
 Negative20292 (46)63NS43 (21)81NS
Table 2. Factors that Influenced Survival in 131 Patients with Low-Grade Ependymomas
Parameter estimateNo. of patientsNo. of recurrences (%)Five-yr PFS (%)P valueNo. of deaths (%)5-yr OS (%)P value
  1. PFS: progression-free survival; OS: overall survival; NS: not significant.

Age (yrs)       
 < 164812 (25)826 (13)90
 ≥ 168310 (12)930.044 (5)98NS
Tumor location       
 Infratentorial9317 (18)909 (10)94
 Supratentorial38 5 (13)90NS1 (3)95NS
Type or resection       
 Macroscopic total67 7 (10)953 (6)96
 Subtotal6415 (23)810.037 (11)93NS
Radiotherapy results       
 Positive6611 (17)894 (7)95
 Negative6511 (17)91NS6 (9)95NS
Chemotherapy results       
 Positive12 3 (25)871 (8)93
 Negative11919 (16)92NS9 (8)96NS
Table 3. Factors that Influenced Survival in 127 Patients with High-Grade Ependymomas
Parameter estimateNo. of patientsNo. of recurrences (%)Five-yr PFS (%)P valueNo. of deaths (%)5-yr OS (%)P value
  1. PFS: progression-free survival; OS: overall survival; NS: not significant.

Age (yrs)       
 < 166747 (70)1826 (39)52
 ≥ 166044 (73)340.0218 (30)740.02
Tumor location       
 Infratentorial7143 (68)4324 (34)75
 Supratentorial5648 (77)80.000120 (36)510.02
Type of resection       
 Macroscopic total5434 (62)3419(35)67
 Subtotal7357 (78)150.0225 (34)560.04
Radiotherapy results       
 Positive9565 (68)3829 (30)70
 Negative3226 (81)120.0115 (47)320.01
Chemotherapy results       
 Positive4424 (54)2212 (27)78
 Negative8367 (81)22NS32 (39)480.01
thumbnail image

Figure 2. Kaplan–Meier survival curves demonstrating the influence of histologic tumor grade on the (A) progression-free survival rate and (B) the overall survival rate in patients with ependymomas in the study population. Solid line: high-grade ependymomas; dashed line: low-grade ependymomas.

Download figure to PowerPoint

thumbnail image

Figure 3. Kaplan–Meier survival curves demonstrating the influence of various combinations of treatment on the overall survival rate of patients with high-grade ependymomas. Solid line: macroscopic total resection and radiotherapy; dashed line: all other treatment combinations.

Download figure to PowerPoint

The multivariate analysis revealed that the risk of recurrence was associated with supratentorial tumor location (hazard ratio, 3.01; P = 0.002), incomplete resection (hazard ratio, 3.82; P = 0.01), and high-grade histology (hazard ratio, 8.53; P = 0.000001). In contrast, the risk of death was found to be associated only with anaplastic histology (hazard ratio, 6.01; P = 0.00001).

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

After patients underwent wide resection that employed a microsurgical technique, perioperative deaths were diminished. Therefore, the biologic behavior of brain tumors plays an important role in determining the optimal therapeutic regimen and predicting clinical outcome in patients with ependymomas. Nevertheless, compared with astrocytomas, oligodendrogliomas, and medulloblastomas, to our knowledge to date, the treatment strategy for ependymal neoplasms has not been standardized. This may be explained as follows: 1) the universally applied diagnostic criteria for ependymoma histologic grading have not been yet determined; or 2) the molecular events that take place at the genomic level and that underlay either ependymoma tumorigenesis or its malignant progression remain poorly understood.

In the current study, we analyzed the prognostic significance of various clinical and pathologic parameters for a cohort of 258 intracranial ependymomas. The 5-year PFS and OS rates (61% and 82%, respectively) were similar to those reported recently. The recurrence rate was 44%, and isolated tumor regrowth at the primary site was the main cause of failure. In the current series, we found that 18 ependymomas (7%) developed cerebrospinal seeding, and 5 tumors (2%) metastasized extraneurally. All distant implants manifested after an incomplete resection of the malignant primary tumor, which is in keeping with data published previously.1, 16, 17

The prognostic significance of ependymoma histology is a controversial issue, as discussed earlier. A few reports have denied the prognostic utility of histologic malignancy in ependymomas.8, 11–14 Those studies usually included heterogeneously treated cohorts of patients, some of whom exhibited evidence of tumor dissemination at presentation. Most of the recent series demonstrated a significant decrease in survival for patients with anaplastic ependymomas, although the rate of high-grade tumors varied broadly in different reports.5, 6, 9, 10, 15, 16 Obvious variations in treatment regimens and the lack of a consensus regarding ependymoma classification may explain the conflicting data concerning tumor grade and prognosis.

Given their rarity, it is hard to accumulate sufficient numbers of ependymomas in a single institution to establish reliable grading criteria based on follow-up data. Therefore, the main objective of the current study was to evaluate the real clinical significance of histologic malignancy in ependymomas. The recent grading scheme proposed by Merchant et al.15 was applied as a more appropriate tool to reach this goal. Consequently, we found that the grade of tumor malignancy was a cornerstone pattern for the prognosis of patients with intracranial ependymomas treated in the microneurosurgical era. Moreover, an agreement on the definition of histopathologic grade of malignancy was reached in most of the patients tested. Therefore, this grading scheme is likely to be practical, useful, reproducible, and applicable clinically. Nevertheless, the prediction of an individual clinical course for a singular patient with ependymoma based exclusively on the pathologic findings remains difficult. The current study showed that clinical outcomes for patients with low-grade ependymomas are exceptionally good, but late failures beyond the 5-year survival borderline are common. Conversely, long-term survivors with frankly anaplastic-appearing ependymomas have been reported.18 Perhaps this phenomenon, which was believed to originate from tumor molecular heterogeneity, takes place among ependymomas that have a similar grade of malignancy.

The vast majority of previous studies reported worse prognoses for pediatric patients who had ependymomas compared with adults. Also, younger children usually had lower survival rates. The current study revealed poorer outcomes for the pediatric group overall, although differences between younger children and older children were not found. Adverse outcomes for pediatric patients with ependymomas have been attributed to the age-dependent differences in tumor location, resectability, and adjuvant treatment.2–4 High-grade ependymomas usually prevail among the pediatric population, as also demonstrated in the current study. In addition, a few recent reports have suggested that pediatric and adult ependymomas can represent molecularly distinct diseases. Various cytogenetic findings in tumors from adults were different from those found in children.19 Pediatric ependymomas frequently demonstrated prognostically unfavorable chromosomal aberrations, such as gain of 1q.20, 21 A recent cDNA microarray study revealed age-related differences in ependymoma gene expression profiles.22 The abundant expression of gene transcripts involved in oncogenesis and cell cycle activation was identified frequently among a pediatric population. Taken together, these molecular events may explain the more aggressive clinical behavior of childhood ependymomas.

The location of intracranial ependymomas has been found to be associated with clinical outcome, although conflicting results have been reported (for review, see Hamilton and Pollack3). Adverse outcomes have been noted for patients with ependymomas located in the posterior fossa in some series, although those studies also noted a higher perioperative mortality rate for patients who had infratentorial tumors. More recent studies have indicated that patients with supratentorial ependymomas had worse outcomes; however, the majority of reports found no association between tumor location and patient survival.

In the current series, patients who had supratentorial tumors had poorer outcomes, and those patients had a prevalence of high-grade ependymomas. It is interesting to note that the 5-year event-free survival and OS rates for patients who had high-grade infratentorial ependymomas were greater compared with the rates for patients who had supratentorial ependymomas. Site-related differences in the molecular biology of intracranial anaplastic ependymomas have been observed recently.22 Tumors from infratentorial sites showed gene expression patterns similar to those found in their low-grade counterparts, suggesting the likelihood of a “secondary” multistep malignant transformation. In contrast, the genetic signatures for both low-grade and high-grade supratentorially located tumors clearly differed, raising the question of whether these neoplasms represent molecularly distinct ependymoma entities.

The completeness of surgical resection has emerged as one of the most significant predictors of outcome in patients with ependymoma as the number of totally removed tumors has increased over time. Approximately 50% of patients in the current series underwent neuroradiologically confirmed macroscopic total tumor extirpation. Complete removal was associated independently with increased event-free survival, but not with OS. This observation indicates that both adjuvant treatment and tumor biologic behavior may contribute to the final clinical outcome of patients with ependymoma.

According to widely spread opinion, ependymomas are among the most radiosensitive of the neuroectodermal tumors, and irradiation is accepted as an inalienable part of its combined treatment.23 However, the total radiation dose to the bed of the primary tumor and the extent of the treatment fields to our knowledge have been discussed. Currently, novel treatment modalities such as three-dimensional conformal radiotherapy, intensity-modulated radiotherapy, and stereotactic radiosurgery are under consideration.23 In the current series, the standard protocol for local irradiation was applied. A better effect of this treatment modality was established for patients with malignant ependymomas who previously underwent complete tumor removal.

Disagreement persists regarding the benefit of chemotherapy for patients with ependymoma, and to our knowledge most reports have suggested that this modality does not improve clinical outcome.24, 25 In addition, abundant expression of the chemoresistance-related factors at both the gene expression level and the protein level has been identified in ependymomas that hints at intrinsic drug resistance.22, 26–28 The current study disclosed some positive effects of polychemotherapy on the survival of patients with anaplastic ependymomas, although its real value is shadowed by radiotherapy applied previously.

Application of the uniform diagnostic criteria for ependymoma grading highlighted the key role of histologic tumor malignancy played in clinical outcome in a cohort of patients who underwent microneurosurgery. The prognostic significance of clinical factors and treatment modalities appears to be related to some degree with ependymoma histologic grade; however, support for such an assertion will require further proof in prospective trials. The prediction of individual clinical outcomes for tumors within a separate grading category remains an open issue. New insight into the molecular pathology of ependymoma has led us toward an awareness that high-throughput molecular analysis may be worthwhile for subdividing these tumors further into prognostically relevant subsets.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES
  • 1
    Wiestler OD, Schiffer D, Coons SW, Prayson RA, Rosenblum MK. Ependymal tumors. In: KleihuesP, CaveneeWK, editors. Tumors of the nervous system. Pathology and genetics: World Health Organization international classification of tumours. Lyon: IARC Press, 2000: 7281.
  • 2
    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. 1998; 30: 319329.
  • 3
    Hamilton RL, Pollack IF. The molecular biology of ependymomas. Brain Pathol. 1997; 7: 807822.
  • 4
    Smyth MD, Horn BN, Russo C, Berger MS. Intracranial ependymomas of childhood: current management strategies. Pediatr Neurosurg. 2000; 33: 138150.
  • 5
    Ernestus RI, Schröder R, Stützer H, Klug N. The clinical and prognostic relevance of grading in intracranial ependymomas. Br J Neurosurg. 1997; 11: 421428.
  • 6
    Figarella-Branger D, Civatte M, Bouvier-Labit C, et al. Prognostic factors in intracranial ependymomas in children. J Neurosurg. 2000; 93: 605613.
  • 7
    Foreman NK, Love S, Thorne R. Intracranial ependymomas: analysis of prognostic factors in a population-based series. Pediatr Neurosurg. 1996; 24: 119125.
  • 8
    Gerszten PC, Pollack IF, Martínez AJ, Lo KH, Janosky J, Albright AL. Intracranial ependymomas of childhood. Lack of correlation of histopathology and clinical outcome. Pathol Res Pract. 1996; 192: 515522.
  • 9
    Horn B, Heiderman R, Geyer R, et al. A multiinstitutional retrospective study of intracranial ependymoma in children: identification of risk factors. J Pediatr Hematol Oncol. 1999; 21: 203211.
  • 10
    McLaughlin MP, Marcus RBJ, Buatti JM, et al. Ependymoma: results, prognostic factors and treatment recommendations. Int J Radiat Oncol Biol Phys. 1998; 40: 845850.
  • 11
    Perilongo G, Massimino M, Sotti G, et al. Analyses of prognostic factors in a retrospective review of 92 children with ependymoma: Italian Pediatric Neurooncology Group. Med Pediatr Oncol. 1997; 29: 7985.
  • 12
    Robertson PL, Zeltzer PM, Boyett JM, et al. Survival and prognostic factors following radiation therapy and chemotherapy for ependymomas in children: a report of the Children's Cancer Group. J Neurosurg. 1998; 88: 695703.
  • 13
    Schiffer D, Chio A, Cravioto H, Giordana MT, et al. Ependymoma: internal correlations among pathological signs: the anaplastic variant. Neurosurgery. 1991; 29: 206210.
  • 14
    Velen-Vincent MLC, Pierre-Kahn A, Kalifa C, et al. Ependymoma in childhood: prognostic factors, extent of surgery, and adjuvant therapy. J Neurosurg. 2002; 97: 827835.
  • 15
    Merchant TE, Jenkins JJ, Burger PC, et al. Influence of tumor grade on time to progression after irradiation for localized ependymoma in children. Int J Radiat Oncol Biol Phys. 2002; 53: 5257.
  • 16
    McLendon RE, Enterline DS, Tien RD, Thorstad WL, Bruner JM. Tumors of central neuroepithelial origin. In: BignerDD, McLendonRE, BrunerJM, editors. Russel and Rubinstein's pathology of tumors of the nervous system, 6th edition, volume 1. London: Arnold, 1998; 387417.
  • 17
    Ironside JW, Moss TH, Louis DN, Lowe JS, Weller RO. Diagnostic pathology of nervous system tumors. London: Churchill Livingstone, 2002.
  • 18
    Ross GW, Rubinstein LJ. Lack of histopathological correlation of malignant ependymomas with postoperative survival. J Neurosurg. 1989; 70: 3136.
  • 19
    Hirose Y, Aldape K, Bollen A, et al. Chromosomal abnormalities subdivided ependymal tumors into clinically relevant groups. Am J Pathol. 2001; 159: 11371143.
  • 20
    Dyer S, Prebble E, Davison V, et al. Genomic imbalances in pediatric intracranial ependymomas define clinically relevant groups. Am J Pathol. 2002; 161: 21332141.
  • 21
    Carter M, Nicholson J, Ross F, et al. Genetic abnormalities detected in ependymomas by comparative genome hybridization. Br J Cancer. 2002; 86: 929939.
  • 22
    Korshunov A, Neben K, Wrobel G, et al. Gene expression patterns in ependymomas correlate with tumor location, grade and patient age. Am J Pathol. 2003; 163: 17211727.
  • 23
    Paulino AC. Radiotherapeutic management of intracranial ependymoma. Pediatr Hematol Oncol. 2002; 19: 295308.
  • 24
    Souweidane MM, Bouffet E, Finlay J. The role of chemotherapy in newly diagnosed ependymoma in childhood. Pediatr Neurosurg. 1998; 28: 273278.
  • 25
    Timmermann B, Kortmann RD, Kuhl J, et al. Combined postoperative irradiation and chemotherapy for anaplastic ependymomas in childhood: results of the German prospective trials HIT88/89 and HIT 91. Int J Radiat Oncol Biol Phys. 2000; 46: 287295.
  • 26
    Chou PM, Barquin N, Gonzales Crussi F, Ridaura Sanz C, Tomita T, Reyes Mugica M. Ependymomas in children express the multidrug resistance gene: immuno-histochemical and molecular biology study. Pediatr Pathol Lab Med. 1996; 16: 551561.
  • 27
    Geddes JF, Vowle GH, Ashmore SM, Cockburn HA, Darling JL. Detection of multidrug resistance gene product (P-glycoprotein) expression in ependymomas. Neuropathol Appl Neurobiol. 1994; 20: 118121.
  • 28
    Korshunov A, Sycheva R, Timirgaz V, Golanov A. Prognostic value of immunoexpression of the chemoresistance-related proteins in ependymomas. An analysis of 76 cases. J Neuro-Oncol. 1999; 45: 219227.