Brain Advance Access originally published online on April 19, 2007
Brain 2007 130(5):1338-1349; doi:10.1093/brain/awm046
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Multicentric French study on adult intracranial ependymomas: prognostic factors analysis and therapeutic considerations from a cohort of 152 patients
1Department of Neurosurgery, Timone Hospital, Marseille, France, 2Neuro-oncological Unit, Timone Hospital, Marseille, 3Department of Pathology and Neuropathology, Timone Hospital, Marseille, 4Department of Biostatistics, Timone Hospital, Marseille, 5Department of Pathology and Neuropathology, "Pierre Wertheimer" Neurological Hospital, Lyon and 6Department of Neurosurgery, "Pierre Wertheimer" Neurological Hospital, Lyon, France For the Club de Neuro-Oncologie de la Société Française de Neurochirurgie (SFNC) and the Association des Neuro-Oncologues d’Expression Française (ANOCEF)
Correspondence to: Philippe Metellus, MD, Service de Neurochirurgie du Professeur GRISOLI, Hôpital de la TIMONE, 264, rue Saint-Pierre, 13385 Cedex Marseille, France E-mail: philippe.metellus{at}mail.ap-hm.fr
 |
Summary |
|---|
 Top Summary IntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Ependymomas account for 2% of all intracranial tumours in adults. Considerable controversy continues to exist with regard to their prognostic factors and therapeutic management due to the rarity and the heterogeneity of series reported so far. The authors report a retrospective study of a homogenous population of 152 adult patients harbouring intracranial ependymomas from 24 French Neurosurgical Centres between 1990 and 2004. All clinico-radiological and follow-up data were analysed and a central pathologic review was performed by two confirmed neuropathologists. The 5- and 10-year overall survival rates were 84.8 and 76.5%, respectively; the 5- and 10-year progression-free survival rates were 63.5 and 52.8%, respectively. On multivariate analysis, overall survival rates were associated with histological grade (P < 0.001), extent of surgery (P = 0.006), patient age (P = 0.004) and patient Karnofski performance status (P = 0.03). The multivariate analysis also revealed that the risk of recurrence was associated with high histological grade (P < 0.001), incomplete resection (P < 0.001) and Karnofski performance status
80 (P = 0.04). The impact of radiotherapy was found to be beneficial for incompletely resected low-grade ependymomas and to a lesser extent for completely removed high-grade tumours. In association with Karnofski performance status and extent of surgery, histological grade is a major prognostic factor in adult intracranial ependymomas. The application of a simple and reproducible grading scheme using objective anaplastic criteria seems useful practically and clinically applicable. The role of adjuvant radiotherapy for patients with incompletely resected low-grade ependymomas seems to be beneficial but remains to be addressed for high-grade tumours.
Key Words: ependymoma; adult; prognostic factors; radiotherapy; classification
Abbreviations: CART, classification and regression tree; GTR (+), gross total removal; GTR (–), incomplete resection; KPS, Karnofski performance status; LV, lateral ventricle; Max, maximum; Min, minimum; OS, overall survival; P, parenchymal; PFS, progression-free survival; RPA, recursive partition analysis; RT, radiotherapy; V3, third ventricle; WHO, World Health Organization
Received November 4, 2006. Revised February 10, 2007. Accepted February 15, 2007.
|
Introduction |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Adult intracranial ependymoma is a relatively rare brain tumour entity, accounting for 2–5% of all intracranial neoplasms (Rawlings et al., 1988
; Ernestus et al., 1997; Guyotat et al., 2002; Korshunov et al., 2004; Reni et al., 2004). A considerable number of retrospective studies have been reported in the last decade on ependymomas (Rawlings et al., 1988; Lyons and Kelly, 1991; Schiffer et al., 1991; Vanuytsel et al., 1992; Rezai et al., 1996; Ernestus et al., 1997; Stuben et al., 1997; McLaughlin et al., 1998; Robertson et al., 1998; Schild et al., 1998; Schwartz et al., 1999; Figarella-Branger et al., 2000; Spagnoli et al., 2000; Guyotat et al., 2002; Korshunov et al., 2004; Reni et al., 2004). However, the pertinent prognostic factors as well as the pattern of recurrence remain to be elucidated. According to several recent reports, the prognostic value of Karnofski performance status, tumour location, the extent of surgical removal, histological grade and even post-operative radiotherapy remain controversial. Actually, the small number of patients included in the studies reported, the heterogeneity of the population of patients in terms of age (adult, children) and location (intracranial, spinal) as well as the very long period in which the patients were treated (1960 to 2000) probably explain the lack of significant prognostic parameters validated in this pathological entity. As a consequence, optimal therapeutic management of adult intracranial ependymomas remains an ongoing debate. We report herein the results of the first and largest multi-institutional retrospective analysis of intracranial ependymomas in 152 adult patients diagnosed in 24 French Neurosurgical University Hospital Centres after 1990. This cohort study was conducted by the French Neurosurgical Society (Société Française de Neurochirurgie, SFNC), the French Speaking Association of Neurologist and Neuro-Oncolgists (Association des Neurologues et Neuro-Oncologues de Langue Française, ANOCEF) and the French Neuropathological Society (Société Française de Neuropathologie, SFNP) to determine whether age, pre-operative clinical status, tumour location, extent of surgery, histological features and post-operative radiotherapy affect overall survival and progression-free survival.
|
Material and methods |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Patient population
A multi-institutional database search on adult intracranial ependymomas including 24 French Neurosurgical University Hospital Centres was conducted by the SFNC, the ANOCEF and the SFNP. Inclusion criteria were: confirmation by two independent neuropathologists (DFB, AJ) of intracranial ependymoma in patients of both sexes, aged 18 years or older, operated after 1990, and with no previous brain irradiation for any intracranial pathology (Table 1). After central pathological review of 258 cases, 152 patients with confirmed diagnosis of ependymoma were eligible for this multicentric retrospective study. The clinical and radiological treatment and follow-up data were collected by a senior neurosurgeon (PM) and a senior neuro-oncologist (MB).
|
Clinico-radiological data
The following clinical data were collected: patient's age at surgery, sex, presenting symptoms and pre- and post-operative Karnofski performance status (KPS) score. Tumour location was divided into two groups: supratentorial and infratentorial ependymomas. Among supratentorial ependymomas, three subgroups were identified: intraparenchymal tumours, tumours arising in the lateral ventricles and tumours of the third ventricle. In the infratentorial tumour group, the invasion of the floor of the IVth ventricle and lateral extension into the foramen of Lushka were always notified. Contrast enhancement as well as associated hydrocephalus was also analysed. Patients were considered metastatic if MRI scans demonstrated intracranial or spinal tumour dissemination and/or if CSF analysis revealed abnormal cells.
Treatment modalities
Except for patients who died in the peri-operative period, the extent of surgery was evaluated by post-operative MRI scans. Peri-operative mortality was defined as death occurring within 3 months post-surgery and included neurosurgical complications and systemic complications such as pulmonary embolism, sepsis or cardiac failure. For these patients, extent of surgery was evaluated by a single neurosurgeon (PM) according to an extensive review of all operative protocols. Surgical resection was classified as complete (patients with no residual tumour on MRI scans or considered with total tumour removal based on operative protocol examination) or incomplete for all patients.
Regarding irradiation, the following data were collected: doses, time to surgery (adjuvant, at recurrence or progression) and location—cranial (focal or panencephalic) or cranio-spinal. For patients who received chemotherapy or radiosurgery treatments, no precision was collected concerning protocol or doses but precise time to surgery was noted.
Pathological examination
For all patients, slides used for diagnosis, coloured by hematein eosin and/or paraffin-embedded blocks, were sent to the same neuropathologist. Pathological examination was conducted centrally by two senior neuropathologists (DFB, AJ). Subependymomas and ependymoblastomas were excluded according to the WHO classification (2000). Ependymomas were first classified as WHO grade II or WHO grade III (anaplastic) (Bouvier et al., 2003). Secondarily, ependymomas were graded according to the Marseille neograding system based on objective anaplastic features. The following criteria were assessed and quantified as follows: necrosis (present versus absent), microvascular proliferation (presence versus absence) and mitotic count in 10 consecutive high-power fields (<5 versus 
5). An ependymoma was considered low grade if 0 or 1 of the criterion was present and high grade if 2 or 3 criteria were present. Ki67-labelling index was performed in 108 cases (paraffin-embedded blocks and specimen fixed in formalin).
In doubtful cases, immunohistochemistry (immunoperoxydase with avidin biotin complex after antigen retrieval on Ventana Devices®) was performed to reach diagnosis. Expression of the following antigens was searched for with appropriate antibodies: GFAP (polyclonal Dakopatts), EMA (clone E29), synaptophysin (polyclonal Dakopatts), keratin (clone AE1/AE3/PCK26 and KL1) and vimentin (clone V9).
Statistical analysis
Categorical variables are expressed as percentage. Survivals were estimated by using the Kaplan–Meier method and curves were compared by using the log-rank test. The effect of potential risk factors on the disease-free and overall survival were evaluated with Cox proportional hazards models. Recursive partitioning analysis (RPA) using classification and regression trees (CART) algorithm was applied to establish prognostic groups (Breinman et al., 1984). This is a method of building decision trees to model predictors. RPA using CART algorithm was performed using the variables significantly correlated to survival in both uni- and multivariate analysis. These variables were examined for the best split in a given population. A restriction was imposed on the tree construction such that terminal subgroups resulting from any given split must contain at least 10 patients. Terminal node populations were tested by log-rank test to determine whether any two groups were similar enough in survival to be merged. All statistical tests were two-sided, and the threshold for statistical significance was P = 0.05. Analyses were performed with SPSS for Windows version 11.5 (SPSS Inc, Chicago, Illinois, USA).
|
Results |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Clinical data
Demographic data are summarized in Table 2. Mean age was 45.2 years (±16.3 yr) and the age range was 16 to 82 years. Impairment of cranial nerves as well as cerebello-vestibular symptoms were predominantly seen in infratentorial tumours, intracranial hypertension was more frequent in tumours of the third ventricle (V3) but also in tumours located infratentorially and in the lateral ventricle (LV). Hydrocephalus requiring shunting or endoscopic ventriculostomy was present in the intraventricular location only. Supratentorial parenchymal tumours were more likely to be associated with epilepsy, motor deficit and behavioural changes. No correlation was found between age, sex, clinical status and location, grade or extent of surgery (not shown).
|
Pathological data
Of the 258 cases sent for central pathological review, only 152 fulfilled diagnostic criteria of ependymomas. On both pathological and immunohistochemical features, the following diagnoses were excluded: glioblastomas, oligodendrogliomas and mixed oligoastrocytomas, pilocytic astrocytomas, subependymomas, papillary tumours of the pineal gland, central neurocytomas, papillary neuroglial tumours, oligodendroglioma with neurocytic differentiation, medulloblastomas, metastases and papillary meningiomas (Table 3).
|
Tumours were graded in two different manners according to WHO classification and the Marseille grading system. In WHO classification, 109 patients (88 infratentorial and 21 supratentorial tumours) harboured grade II ependymomas (71.7%) and 43 patients (18 infratentorial and 25 supratentorial tumours) grade III ependymomas (28.3%). In the Marseille grading system, 112 patients (90 infratentorial and 22 supratentorial) harboured low grade (0 or 1 anaplastic criterion) ependymomas (73.7%) and 40 patients (16 infratentorial and 24 supratentorial) high grade (2 or 3 anaplastic criterion) ependymomas (26.3%).
Location, grade and extent of surgery
These data are summarized in Table 4. Tumour location was supratentorial in 46 patients (30.3%) and infratentorial in 106 (69.7%). Among supratentorial tumours, parenchymal location was found in 22 patients (14.5%), V3 location in 16 (10.5%) and LV location in 8 (5.3%). Among patients with infratentorial tumours (106 patients), 30 patients (28.2%) presented with an extension into the floor of the IVth ventricle, 20 (18.9%) with a lateral extension through the foramen of Lushka, 31 (29.2%) with both extension and 25 (23.6%) with no extension. Supratentorial location was significantly associated with both grade III (WHO) and high grade (Marseille grading system) tumours compared to infratentorial tumours (P < 0.001). Furthermore, within supratentorial tumours, parenchymal location was also significantly associated with grade III (WHO) tumours (P = 0.02) and high-grade (Marseille grading system) tumours (P = 0.01). Post-operative MRI scans were available for all patients alive 3 months post-operatively and constituted the reference method to evaluate the extent of surgery. There was neither significant correlation between location and extent of surgery nor between grade (WHO and Marseille grading system) and extent of surgery.
|
Follow-up data and patterns of failure
At the endpoint of the follow-up analysis, 114 patients (79 disease-free) (75%) were still alive after a median duration follow-up of 73 months (range, 14–159 months). Thirty-eight patients (25%) died during the follow-up period. Of these patients, causes of death included ependymoma progression in 25 patients (65.8%), complications within 3 months after surgery (operative mortality) in 12 patients (31.6% which represents 7.9% of the entire population) and unrelated cause during disease-free period (lung cancer) in 1 patient (2.6%). Fifty-one patients presented recurrence or progressive disease (33.6% of patients). Disease progressed as an isolated local recurrence in 46 patients (30% of patients; 82% of failures). Disseminated disease within the central nervous system was found in 10 patients (6.6% of patients; 18% of failures). Among these patients, five developed distant intracranial metastasis (3.3% of patients), four presented with spinal metastasis (2.6% of patients) and one patient combined distant intracranial and spinal metastasis (0.7% of patients). All of these patients presented associated focal progressive or recurrent disease. Of the 10 patients with central nervous system metastasis, eight were histological grade III and four had initially incomplete resection as assessed by post-operative MRI scans. Histology was significantly correlated to metastasis occurrence (P = 0.03) but not the extent of surgery (P > 0.05).
Survival analysis
Overall survival
The 5- and 10-year overall survival rates for the entire cohort were 84.8 and 76.5%, respectively. On univariate analysis (Table 5), infratentorial location (P < 0.001), pre-operative KPS score >80 (P = 0.002), GTR (+) (P = 0.036), low grade (Marseille grading system) (P < 0.001), WHO grade II (P < 0.001), the delivery of adjuvant RT (P = 0.009) and a Ki-67-labelling index <10% (P = 0.001) were found to be associated with a longer survival (Fig. 1). On multivariate analysis (Table 5), age <55 years (P = 0.004), GTR (P = 0.006), pre-operative KPS score >80 (P = 0.03) and low (0 or 1 criterion/3) Marseille grade (P < 0.001) were confirmed as prognostic indicators while infratentorial location (P = 0.06) had borderline significance. Age, sex and adjuvant treatment exhibited no independent association with OS. Ki-67-labelling index was not considered a candidate variable because data was available for only 108/152 patients.
|
|
Progression-free survival
The 5- and 10-year progression-free survival rates for the entire cohort were of 65.3 and 52.8%, respectively. On univariate analysis (Table 6), infratentorial location (P < 0.001), pre-operative KPS score >80 (P = 0.03), GTR (P = 0.004), low grade (Marseille grading system) (P < 0.001), WHO grade II (P < 0.001) and a Ki-67-labelling index <10% (P < 0.001) were found to be associated with a longer progression-free survival (Fig. 2). On multivariate analysis (Table 6), only GTR (P < 0.001), pre-operative KPS >80 (P = 0.04) and low grade (Marseille grading system) (P < 0.001) were confirmed as significant independent prognostic indicators. For the same reason aforementioned, Ki-67-labelling index was not considered a candidate variable.
|
|
Adjuvant RT impact on OS and PFS
Among low-grade (Marseille grading system) ependymomas, GTR was achieved in 66 patients. Of these patients, 59 did not receive RT and 7 benefited from local irradiation. No significant difference was found in terms of PFS or OS in these subgroups. Among the 46 remaining patients in whom GTR was not achieved 24 benefited from post-operative RT. A significant statistical difference was found in PFS (P = 0.05) between patients who received post-operative RT compared to those without adjuvant treatment (Fig. 3) and a trend towards a better OS was also found (data not shown).
|
Among patients with high-grade (Marseille grading system) ependymomas, GTR was achieved in 23 patients. Of these patients, 11 did not receive adjuvant RT and 12 benefited from cranial or cranio-spinal RT. No significant difference was found in PFS or OS in these subgroups; however, a trend toward a better PFS (P = 0.09) was found in irradiated patients (data not shown). Among the 17 remaining patients with incomplete resection, 16 benefited from post-operative RT. No significant statistical difference was found in PFS or OS for patients who benefited from post-operative irradiation. However, due to the limited number of patients harbouring high-grade (Marseille grading system) ependymomas, statistical significance of the different tests remained weak.
Chemotherapy
In the entire population, 21 patients had chemotherapy (13.8%). All patients treated with chemotherapy have had concomitant or previous radiotherapy. According to Marseille grading system, 16 patients had high-grade and 5 had low-grade tumours. Among the 16 patients with high-grade tumours, 8 had gross total resection. Of these eight patients, all seven treated for recurrent disease died but one had adjuvant chemotherapy and is still alive after 36 months follow-up. Of the eight patients with incomplete resection, seven were treated at recurrence and five of them died. One patient benefited from adjuvant chemotherapy and is still alive.
Recursive partition analysis using CART algorithm
The most significant split for the first node was the histological grade evaluated by the Marseille grading system. Among the 40 patients with high-grade tumours, the most significant split was by the extent of surgery. Among the 112 patients with low-grade tumours the next significant split occurred between patients with KPS score 80 versus >80.
The classification tree resulting from the CART modelling process exhibited three final groups of ependymoma patients with distinctly different survival (Fig. 4). According to the CART tree, the patients with the worst survival (group III) are those with high-grade tumours and incomplete resection. The best survival was in patients with low-grade tumours and KPS score >80 (group I). All other patients had minor differences in survival and form a middle stage (group II). Median survival time was not estimable for groups I and II and was of 25.7 months for group III (95% CI, 7.2–44.1%) (Table 7).
|
|
|
Discussion |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Adult intracranial ependymomas are rare CNS tumours that continue to generate considerable controversy with regard to their clinical management. The lack of widely accepted and recognized prognostic factors leads to the absence of standardized therapeutic guidelines. In this study we analysed potential clinical and pathological prognostic factors in the most important and homogenous population of adult patients harbouring intracranial ependymomas treated in the microsurgical era.
The extent of surgery has emerged as one of the most significant predictors of outcome in patients with intracranial ependymomas (Ernestus et al., 1997; Paulino and Wen, 2000; Oya et al., 2002; Paulino et al., 2002; Korshunov et al., 2004; Reni et al., 2004; Kawabata et al., 2005; Rogers et al., 2005). Note, however, that some authors have found no correlation between extent of surgery and prognosis (Vanuytsel et al., 1992; McLaughlin et al., 1998; Robertson et al., 1998; Guyotat et al., 2002). In our series, the extent of resection was a major prognostic factor in terms of survival and recurrence. Actually, gross total removal was significantly associated with better OS and PFS in our population both in univariate and multivariate analysis. These results are strengthened by the fact that almost all of our patient population benefited from post-operative MRI examination. Thus, we recommend performing a second surgery in patients with incomplete resection when technically feasible. In some reports, extent of surgery also correlated with CSF dissemination and metastatic rate (Rezai et al., 1996; Korshunov et al., 2004; Kawabata et al., 2005). This was not the case in our study but histological grade as reported in other studies was strongly correlated with CSF dissemination and metastasis (Rezai et al., 1996; Korshunov et al., 2004; Kawabata et al., 2005).
The prognostic value of histological findings remains a controversial issue probably attributable to sample size, variability in the definition of anaplasia, discrepancies in histological diagnoses and the inclusion in some series of ependymoblastomas and subependymomas which exhibit different biological behaviour and should be analysed separately (Lyons and Kelly, 1991; Schiffer et al., 1991; Ernestus et al., 1997; Schild et al., 1998; Spagnoli et al., 2000; Korshunov et al., 2004; Reni et al., 2004; Kawabata et al., 2005). As emphasized by this study, in 106/258 (41.1%) cases an initial incorrect diagnosis of ependymoma was made. Four classes of pathological misdiagnoses can be distinguished. The first concerns erroneous diagnosis which would be avoided by a trained pathologist aware of the peculiar histological and immunohistochemical features. This category includes central neurocytomas, medulloblastomas, metastatic carcinomas and papillary meningiomas. The second concerns confusion between subependymomas and ependymomas. The third class results from the emergence of new pathological entities such as papillary tumours of the pineal region, glioneuronal tumours and oligodendrogliomas with neurocytic differentiation. The fourth and last class corresponds to gliomas which are frequently misdiagnosed as ependymomas. In our study, except for pilocytic ependymomas which were located in the posterior fossa, the other gliomas were all in supratentorial parenchymal location. Therefore, in the authors’ opinion, in the case of supratentorial parenchymal location, the pathologist should carefully exclude oligodendrogliomas, mixed oligoastrocytomas or glioblastomas before assessing the diagnosis of ependymoma. Because olig2 staining is usually lacking in ependymomas, it might be useful to differentiate true ependymomas from others gliomas based on this molecular marker expression (Bouvier et al., 2003). In a recent series, Reni et al. (2004) did not find histological grade to be correlated to survival. However, in their study, no central pathological review was planned and the role of histology in predicting outcome may have been masked by the varying definitions of anaplasia used by pathologists involved in the diagnostic process. The lack of impact of tumour grade on survival in their report may be also attributable to the relatively small number of anaplastic ependymomas. Some other reports have denied the prognostic value of histology (Schiffer et al., 1991; Gerszten et al., 1996; Robertson et al., 1998). However, most recent series have demonstrated a significant increase in overall and progression-free survival in low-grade ependymomas (Ernestus et al., 1997; McLaughlin et al., 1998; Figarella-Branger et al., 2000; Korshunov et al., 2004; Wolfsberger et al., 2004; Kawabata et al., 2005; Kurt et al., 2006). Korshunov et al. (2004), in a recent single institution study of 258 intracranial ependymomas including 143 adults treated in the microsurgical era, found the grade of tumour malignancy to be a cornerstone for the prognosis. The lack of consensus regarding ependymomas anaplastic criteria may explain the conflicting data concerning histological grade and prognosis. For this reason we have elucidated the Marseille grading system, a simple and more reproducible grading scheme to classify ependymomas. Only three criteria were taken into account, necrosis, microvascular proliferation (present/absent) and mitotic count (threshold 5). We observed that ependymomas exhibiting 0 or 1 criterion had a significantly better prognosis than ependymomas showing 2 or 3 criterion. Although WHO grade was also significantly correlated to a better OS and PFS in univariate analysis, in our study, only the Marseille grading system was significantly correlated to OS and PFS in multivariate analysis. These criteria were already used for prognostic significance in paediatric ependymomas but in this particular population only ependymomas showing the presence of three criteria differed statistically from others. Our results are consistent with those of Korshunov et al. (2004). In the present study, histological grade was significantly and strongly correlated with OS and PFS both on univariate and multivariate analysis. These results are strengthened by the fact that a central pathological review was planned and conducted by two confirmed neuropathologists (D.F.B and A.J.). The results of the RPA using CART modelling process reinforces the importance of the prognostic value of histological grading which correspond to the first split in the entire population. The fact that tumour grade influences outcome for patients with ependymoma, independently of others factors, should be considered in the design and analysis of future prospective trials involving adult patients. In the present analysis, the Ki-67 immunolabelling index was available for 108 patients. On univariate analysis, a Ki-67 index <10 was significantly correlated to a better prognosis in PFS and OS. Due to lack of data for this prognostic factor, we did not use it in the Cox proportional hazards models construct. A review of the literature on tumour cells proliferation index in ependymomas shows a considerable variation of Ki-67 immunolabelling fractions (Rezai et al., 1996; Prayson, 1998; Figarella-Branger et al., 2000; Suzuki et al., 2001; Verstegen et al., 2002; Wolfsberger et al., 2004). The variation is most likely attributable to tissue fixation, staining protocols and mode of quantification. For these reasons, in this multicentric study, Ki-67 immunolabelling index was measured from paraffin-embedded block of tumour in the same laboratory by the same neuropathologist (D.F.B). Our results are consistent with those of Wolfsberger et al. and others and underline the potential interest of assessing the Ki-67 index in adult intracranial ependymomas for outcome prediction in the routine diagnostic setting (Rezai et al., 1996; Prayson, 1998; Figarella-Branger et al., 2000; Wolfsberger et al., 2004; Kurt et al., 2006).
Another tumour-related prognostic factor is the tumour location whose predictive role is probably confounded essentially by patient age and histological grade (Figarella-Branger et al., 2000; Korshunov et al., 2004; Reni et al., 2004; Kawabata et al., 2005; Rogers et al., 2005). The location of intracranial ependymomas has been found to be associated with clinical outcome, although conflicting results have been reported (Hamilton and Pollack, 1997). Since very few intracranial ependymoma studies have been conducted in adult patients, the role of tumour location in this population is not well-known. In the current series, tumour location was significantly correlated to OS and PFS in univariate analysis but not in multivariate analysis although infratentorial location was associated with a trend towards a more favourable prognosis in OS. This is consistent with earlier reports (McLaughlin et al., 1998; Korshunov et al., 2004; Reni et al., 2004). A likely explanation is the prevalence of high-grade ependymoma in supratentorial location as found in our study. A recent report on intracranial ependymoma demonstrated site-related differences in the molecular biology of these neoplasms raising the question of whether infratentorial and supratentorial ependymomas represent a molecularly distinct entity (Korshunov et al., 2003).
Among patient-related prognostic factors, age <55 years and KPS >80 had positive impact on survival. Age <55 years was associated with a better prognosis in multivariate analysis in terms of OS but not for PFS. These results are consistent with those of Reni et al. but differ from Guyotat et al. (Guyotat et al., 2002; Reni et al., 2004). These are the only two reports in the literature analysing survival in adult intracranial ependymomas with regard to different age groups. According to the median age of our population two cut-off values were chosen: 45 and 55 years. Only the latter exhibited a difference in survival analysis. KPS >80 was associated with a better prognosis in multivariate analysis both for OS and PFS. Few reports evaluated the influence of KPS in ependymomas. All these reports led to conflicting results (Rawlings et al., 1988; Stuben et al., 1997; Spagnoli et al., 2000; Guyotat et al., 2002).
There is a widespread opinion that post-operative irradiation should be included in standard care for patients with high-grade ependymomas (McLaughlin et al., 1998; Oya et al., 2002; Mansur et al., 2005; Rogers et al., 2005). This was the most usual attitude observed in our retrospective study. However, for patients with low-grade ependymomas, especially when complete tumour excision could be achieved, the role of RT remains controversial and this was observed in the current series. Recently, Rogers et al. reported the impact of RT in a series of 45 patients (25 irradiated) harbouring essentially grade II (96%) posterior fossa ependymomas (Rogers et al., 2005). They concluded that adjuvant RT significantly improves tumour control but not overall survival and thus recommend the use of post-operative RT regardless of the extent of surgical resection. In our study we were unable to answer this question since almost all patients with low-grade ependymomas in whom complete resection of the tumour was achieved did not undergo RT. In the authors’ opinion, there is not yet enough strong evidence that supports the benefit of RT in completely resected low-grade tumours to recommend adjuvant RT in this situation. A ‘wait and see’ policy could also be discussed for these cases reserving RT for recurrent disease. Only a prospective randomized study will be able to resolve this issue. In other respects, concerning incompletely resected low-grade ependymomas, in the current series adjuvant RT significantly improves the PFS rate and shows a trend towards a better OS. Although these data must be interpreted with caution, adjuvant RT in incompletely resected low-grade ependymomas might be of interest in this situation. Concerning high-grade tumours, we did not find significant differences in PFS and OS in completely resected tumours but a trend towards a better PFS and OS in patients with post-operative RT. However, the population subgroups involved were too small to draw any categorical conclusion. Finally, in patients with incompletely resected high-grade tumours, no significant difference could be found in terms of PFS and OS in patients who benefited from adjuvant RT or not but of these 17 patients 16 had adjuvant RT precluding any conclusion about impact of RT in this subgroup of patients. In these patients, 11 had chemotherapy, 10 at recurrence and 1 adjuvant to surgery associated with radiotherapy. Among these patients seven died during follow-up. Among the six patients who did not undergo chemotherapy regimen four died during follow-up. Again the small size of each sample precludes any conclusion on the optimal therapeutic proposal in this grave outcome subgroup. It should be kept in mind that negative selection bias in the ependymomas series reported, which are almost all retrospective, cannot be ruled out and that prospective trials are warranted to delineate strong therapeutic guide lines.
To our knowledge, the present series represents the largest report on intracranial ependymomas in adults in the microneurosurgical era. This study and analysis of the literature further highlights that complete tumour removal is a main prognostic factor and the treatment of choice in adult intracranial ependymomas. Application of reproducible diagnostic criteria for ependymoma grading has highlighted the key role of histology in clinical outcome. These two issues are strengthened by the fact that in this multicentric study a central pathological review was conducted by two confirmed neuropathologists and that the post-surgical residual disease was evaluated on MRI in almost all cases. The three groups exhibited by the RPA using CART modelling process, with significantly different survival, could be considered in the design and analysis of future prospective trials involving adult ependymoma patients. Although no meaningful conclusion can be reached on the magnitude of impact of RT on clinical outcome, our results demonstrate a trend toward better PFS and OS in incompletely resected low-grade ependymomas and in completely resected high-grade ependymomas.
|
Acknowledgements |
|---|
The authors wish to thank the French Society of Neurosurgery (SFNC) and the ANOCEF group (Association of Neurologists and Neurooncologists of French Expression) who supported this study with two grants awarded to P. Metellus and M. Barrie to collect data from the 24 French neurosurgical centres. We also thank all neurosurgeons and pathologists who participated in this multi-institutional study. We thank particularly Dr A. Maues de Paula, Dr C. Fernandez, Dr C. Bouvier and C. Cazeaux for their help in the management of central pathological review. We also thank Maryna Blankenstein-Gabert for help with the English-language editing.
|
References |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Bouvier C, Bartoli C, Aguirre-Cruz L, Virard I, Colin C, Fernandez C, et al. Shared oligodendrocyte lineage gene expression in gliomas and oligodendrocyte progenitor cells. J Neurosurg (2003) 99:344–50.[Web of Science][Medline]
Breiman L, Friedman J, Olshen R, Stone C. Classification and regression trees. (1984) New York: Chapman & Hall.
Ernestus RI, Schroder R, Stutzer H, Klug N. The clinical and prognostic relevance of grading in intracranial ependymomas. Br J Neurosurg (1997) 11:421–8.[CrossRef][Web of Science][Medline]
Figarella-Branger D, Civatte M, Bouvier-Labit C, Gouvernet J, Gambarelli D, Gentet JC, et al. Prognostic factors in intracranial ependymomas in children. J Neurosurg (2000) 93:605–13.[Web of Science][Medline]
Gerszten PC, Pollack IF, Martinez AJ, Lo KH, Janosky J, Albright AL. Intracranial ependymomas of childhood. Lack of correlation of histopathology and clinical outcome. Pathol Res Pract (1996) 192:515–22.[Web of Science][Medline]
Guyotat J, Signorelli F, Desme S, Frappaz D, Madarassy G, Montange MF, et al. Intracranial ependymomas in adult patients: analyses of prognostic factors. J Neurooncol (2002) 60:255–68.[CrossRef][Medline]
Hamilton RL, Pollack IF. The molecular biology of ependymomas. Brain Pathol (1997) 7:807–22.[Web of Science][Medline]
Kawabata Y, Takahashi JA, Arakawa Y, Hashimoto N. Long-term outcome in patients harboring intracranial ependymoma. J Neurosurg (2005) 103:31–7.[Web of Science][Medline]
Korshunov A, Golanov A, Sycheva R, Timirgaz V. The histologic grade is a main prognostic factor for patients with intracranial ependymomas treated in the microneurosurgical era: an analysis of 258 patients. Cancer (2004) 100:1230–7.[CrossRef][Web of Science][Medline]
Korshunov A, Neben K, Wrobel G, Tews B, Benner A, Hahn M, et al. Gene expression patterns in ependymomas correlate with tumor location, grade, and patient age. Am J Pathol (2003) 163:1721–7.
Kurt E, Zheng PP, Hop WC, van der Weiden M, Bol M, van den Bent MJ, et al. Identification of relevant prognostic histopathologic features in 69 intracranial ependymomas, excluding myxopapillary ependymomas and subependymomas. Cancer (2006) 106:388–95.[CrossRef][Web of Science][Medline]
Lyons MK, Kelly PJ. Posterior fossa ependymomas: report of 30 cases and review of the literature. Neurosurgery (1991) 28:659–64.; discussion 664–5.[CrossRef][Web of Science][Medline]
Mansur DB, Perry A, Rajaram V, Michalski JM, Park TS, Leonard JR, et al. Postoperative radiation therapy for grade II and III intracranial ependymoma. Int J Radiat Oncol Biol Phys (2005) 61:387–91.[CrossRef][Web of Science][Medline]
McLaughlin MP, Marcus RB Jr, Buatti JM, McCollough WM, Mickle JP, Kedar A, et al. Ependymoma: results, prognostic factors and treatment recommendations. Int J Radiat Oncol Biol Phys (1998) 40:845–50.[CrossRef][Web of Science][Medline]
Oya N, Shibamoto Y, Nagata Y, Negoro Y, Hiraoka M. Postoperative radiotherapy for intracranial ependymoma: analysis of prognostic factors and patterns of failure. J Neurooncol (2002) 56:87–94.[CrossRef][Medline]
Paulino AC, Wen BC. The significance of radiotherapy treatment duration in intracranial ependymoma. Int J Radiat Oncol Biol Phys (2000) 47:585–9.[CrossRef][Web of Science][Medline]
Paulino AC, Wen BC, Buatti JM, Hussey DH, Zhen WK, Mayr NA, et al. Intracranial ependymomas: an analysis of prognostic factors and patterns of failure. Am J Clin Oncol (2002) 25:117–22.[CrossRef][Web of Science][Medline]
Prayson RA. Cyclin D1 and MIB-1 immunohistochemistry in ependymomas: a study of 41 cases. Am J Clin Pathol (1998) 110:629–34.[Web of Science][Medline]
Rawlings CE 3rd, Giangaspero F, Burger PC, Bullard DE. Ependymomas: a clinicopathologic study. Surg Neurol (1988) 29:271–81.[CrossRef][Web of Science][Medline]
Reni M, Brandes AA, Vavassori V, Cavallo G, Casagrande F, Vastola F, et al. A multicenter study of the prognosis and treatment of adult brain ependymal tumors. Cancer (2004) 100:1221–9.[CrossRef][Web of Science][Medline]
Rezai AR, Woo HH, Lee M, Cohen H, Zagzag D, Epstein FJ. Disseminated ependymomas of the central nervous system. J Neurosurg (1996) 85:618–24.[Web of Science][Medline]
Robertson PL, Zeltzer PM, Boyett JM, Rorke LB, Allen JC, Geyer JR, 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) 88:695–703.[Web of Science][Medline]
Rogers L, Pueschel J, Spetzler R, Shapiro W, Coons S, Thomas T, et al. Is gross-total resection sufficient treatment for posterior fossa ependymomas? J Neurosurg (2005) 102:629–36.[Web of Science][Medline]
Schiffer D, Chio A, Cravioto H, Giordana MT, Migheli A, Soffietti R, et al. Ependymoma: internal correlations among pathological signs: the anaplastic variant. Neurosurgery (1991) 29:206–10.[Web of Science][Medline]
Schild SE, Nisi K, Scheithauer BW, Wong WW, Lyons MK, Schomberg PJ, et al. The results of radiotherapy for ependymomas: the Mayo Clinic experience. Int J Radiat Oncol Biol Phys (1998) 42:953–8.[CrossRef][Web of Science][Medline]
Schwartz TH, Kim S, Glick RS, Bagiella E, Balmaceda C, Fetell MR, et al. Supratentorial ependymomas in adult patients. Neurosurgery (1999) 44:721–31.[CrossRef][Web of Science][Medline]
Spagnoli D, Tomei G, Ceccarelli G, Grimoldi N, Lanterna A, Bello L, et al. Combined treatment of fourth ventricle ependymomas: report of 26 cases. Surg Neurol (2000) 54:19–26.; discussion 26.[CrossRef][Web of Science][Medline]
Stuben G, Stuschke M, Kroll M, Havers W, Sack H. Postoperative radiotherapy of spinal and intracranial ependymomas: analysis of prognostic factors. Radiother Oncol (1997) 45:3–10.[CrossRef][Web of Science][Medline]
Suzuki S, Oka H, Kawano N, Tanaka S, Utsuki S, Fujii K. Prognostic value of Ki-67 (MIB-1) and p53 in ependymomas. Brain Tumor Pathol (2001) 18:151–4.[CrossRef][Medline]
Vanuytsel LJ, Bessell EM, Ashley SE, Bloom HJ, Brada M. Intracranial ependymoma: long-term results of a policy of surgery and radiotherapy. Int J Radiat Oncol Biol Phys (1992) 23:313–9.[Web of Science][Medline]
Verstegen MJ, Leenstra DT, Ijlst-Keizers H, Bosch DA. Proliferation- and apoptosis-related proteins in intracranial ependymomas: an immunohistochemical analysis. J Neurooncol (2002) 56:21–8.[CrossRef][Medline]
Wolfsberger S, Fischer I, Hoftberger R, Birner P, Slavc I, Dieckmann K, et al. Ki-67 immunolabeling index is an accurate predictor of outcome in patients with intracranial ependymoma. Am J Surg Pathol (2004) 28:914–20.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. O. Hanemann Magic but treatable? Tumours due to loss of Merlin Brain, March 1, 2008; 131(3): 606 - 615. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||