Search

Recent studies of the medical literature have revealed numerous and serious problems. Errors in the design, methods and interpretation of studies can frequently be identified. A huge hidden problem is publication bias, the tendency for positive articles to be published, while negative articles are either not written or submitted. This can systematically lead to an overestimation of the value of treatments, of diagnostic or prognostic studies. Even more worrisome is selective reporting: only a subset of a wide array of tested hypotheses are presented (the ones that turned out to be positive with significance testing). This is particularly true for secondary endpoints and subgroup findings, but even the primary endpoints of trials have been modified when publications are compared to protocols. The peer-review process is fallible. Even if it were strengthened, reviewers cannot examine what is not reported. Hence many problems can only be mitigated with better reporting. Numerous initiatives have proposed guidelines to promote transparent reporting, but progress is slow., (Copyright © 2021 Elsevier Masson SAS. All rights reserved.)

Medulloblastoma is a frequent high-grade neoplasm among pediatric brain tumours. Its classical imaging features are a midline tumour growing into the fourth ventricle, hyperdense on CT-scan, displaying a hypersignal when using diffusion-weighted imaging, with a variable contrast enhancement. Nevertheless, atypical imaging features have been widely reported, varying according to the age of the patient, and histopathological subtype. In this study, we review the classical and atypical imaging features of medulloblastomas, with emphasis on advanced MRI techniques, histopathological and molecular subtypes and characteristics, and follow-up modalities., (Copyright © 2017 Elsevier Masson SAS. All rights reserved.)

Prophylaxis or treatment of tumor-associated seizures is adaily concern in neurosurgical practice but is often guided by the surgeon's habits rather than evidence from clinical trials, which is lacking. The present study reviews the literature on the incidence, clinical aspects and treatment of epilepsy and epileptic seizures in patients undergoing surgery for meningioma. Based on the published data, we then performed a French nationwide survey of neurosurgeons' practices regarding perioperative management of meningioma-related epilepsy and epileptic seizures., (Copyright © 2019 Elsevier Masson SAS. All rights reserved.)

A model is a simplified preparation that reproduces only the most critical features of a disease. To be considered as a validated animal model, such an experimental preparation must fulfill three criteria: isomorphism or similarities of the symptoms; predictivity or identical pharmacological reactivity; homology or etiological similarity. In epilepsy, the use of animal models helps our understanding of physiological and pathological networks involved in the genesis, maintenance, and propagation of seizures. The animal models of epilepsy are also useful in designing and testing new surgical therapeutical strategies, in particular using deconnection or neuromodulation in drug-resistant focal epilepsies. Here we describe three animal models of focal epilepsy, adapted to addressing experimental surgery issues. Kindling consists in the regular liminar stimulation of a given brain structure in the rodent to develop a focal discharge that is secondarily generalized. The local application of epileptogenic agents such as cobalt, iron, or penicillin leads to focal discharges that do not generalize in the rodent or the primate. It is a model of focal neocortical epilepsy without secondary generalization. The focal application of kainate, an excitotoxic glutamate agonist, in the dorsal hippocampus of the adult mouse results, after a latent period, in spontaneous and recurrent focal discharges, behavioral interictal troubles, drug resistance, and histological anomalies reminiscent of hippocampal sclerosis. This constitutes a model of mesial-temporal epilepsy. Better knowledge, in these models, of the neural networks generating, propagating, and/or controlling the seizures should make it possible to design innovative surgical approaches for the treatment of drug-resistant epilepsies.

Malformations of cortical development (MCD) correspond to a broad spectrum of cerebral lesions resulting from cortical development abnormalities during embryogenesis. They are frequently associated with drug-resistant epilepsy as well as more or less severe neurological and cognitive deficits. Diagnosis of MCD has greatly improved with the progress in contemporary imaging techniques, and patients with cryptogenic epilepsy are increasingly recognized as having MCD. Current classifications based on the combination of clinical, imaging, genetic and pathological data allow analysis of homogeneous patient series and optimal therapeutic strategies. Successful surgical treatment can be proposed to patients with focal lesions such as focal cortical dysplasia or tumors associated with cortical dysplasia (dysembryoplastic neuroepithelial tumors and gangliogliomas). Favorable outcome can also be obtained in some diffuse cases such as tuberous sclerosis, periventricular heterotopia and polymicrogyria. Invasive monitoring, especially stereoelectroencephalography (SEEG), has proved to be useful in determining the organization of the epileptogenic zone in each MCD type and planning cortical resections. Moreover, establishing correlations between neurophysiological data, imaging and pathological findings has allowed surgery without previous invasive procedures in the majority of focal MCD types; however, intracranial recordings remain necessary in case of more diffuse MCD.

It has been established that 20-30% of epilepsies are not controlled by antiepileptic drugs. Drug resistance is associated with several major problems, including prognosis, cognitive function, behavior, mortality, cost and quality of life. Apart from classic risk factors for drug resistance, such as neurological, psychiatric, imaging, EEG abnormalities, a high frequency of seizures before medical therapy and complex febrile convulsions, the potential role of multidrug transporters as well as their genetic control and the altered sensitivity of neuronal drug receptors has gained growing attention. In the future, pharmaceutical engineering may bypass these factors. To a certain extent, drug resistance may develop progressively in a neurobiological process and the control of this process could limit its development.

The surgical treatment of epilepsy requires careful preparation and presents a certain number of technical specificities. The neurosurgeon must master not only the technical aspects but also the therapeutic and functional trade-off in order to modulate the procedure according to morphological and electrophysiological intraoperative data. A large number of technical variants have been developed to correspond to epileptological or functional anatomical considerations. Until this point, the choice of a particular technique does not seem to have a significant impact on the therapeutic effectiveness of surgery, and differences in results can be related to the presurgical evaluation and surgical indications. On the other hand, technical development promises to play an important role in limiting the long-term neurocognitive consequences of surgery.

Surgical resections for intractable epilepsy are generally associated with a high risk of permanent neurological deficit and a poor rate of seizure control. We present a series of 89 patients operated on from 1992 through 2007 for drug-resistant partial epilepsy, in whom surgery was performed in a functional area of the brain: the central (sensorimotor and supplementary motor areas) region in 48 cases, posterior regions (parietal and occipital) in 27, the insula in eight, and the language areas in six. Epilepsy was cryptogenic in 12 patients, and lesion-related in 77: malformation of cortical development in 43, tumor in 17, perinatal cicatrix in 13, vascular lesion in three, and another prenatal lesion in one. Seventy patients underwent stereoelectroencephalographic (SEEG) exploration. The surgical procedure was resective (lesionectomy or SEEG-guided corticectomy) in 83 patients and multiple stereotactic thermocoagulations in six. Ten patients were reoperated because of early seizure recurrence. A postoperative complication was observed in 12 patients. Postoperative deficits were observed in 54 patients (61%) and resolved completely in 29. In 25, a permanent deficit persisted, minor in 19 and moderate to severe in six, which did not correlate with localization or etiology. With a one-year follow-up in 74 patients (mean, 3.6 years), 53 (72%) were in Engel's class I, including 38 (51%) in class IA. Seizure outcome was significantly associated with etiology: 93% of Taylor-type focal cortical dysplasia, whereas only 40% of cryptogenic epilepsies were in class I (p<0.05). This suggests that resective or disconnective surgery for intractable partial epilepsy in functional areas of the brain may be followed by excellent results on seizures and a moderate risk of permanent neurological sequelae.

Positron emission tomography (PET) is currently used in the presurgical workup for drug-resistant partial epilepsies in addition to MRI. Interictal metabolism is studied in clinical practice using (18)fluoro-desoxy-glucose ((18)FDG). In medial temporal lobe epilepsy (MTLE) associated with hippocampal sclerosis, hypometabolism ipsilateral to the epileptogenic focus is found in 70-90% of cases. However, hypometabolism is larger than the structural lesion observed on MRI and includes the epileptogenic zone and ictal discharge spread areas. Hypometabolism is related to surgical outcome and cognitive disturbances in MTLE. Although the usefulness of PET appears less well-established in extratemporal lobe epilepsy and in children, its sensitivity may be improved by coregistration and superimposition of PET on MRI at any age. Focal hypometabolism can be easily detected by visual analysis, allowing detection of minor gyral abnormalities that may correspond to focal cortical dysplasias. Moreover, in cases of negative MRI, focal hypometabolism findings may help invasive monitoring planning and deep electrode placement for SEEG, and finally improve surgical outcome.

Background and Purpose: Verbal memory decline can occur after temporal lobe surgery, especially when the left dominant hemisphere is involved. This potential functional risk must be evaluated before surgery. Among all factors that have been identified by several studies, the side of surgery (left dominant) and high baseline memory performance have been found to be predictive of verbal memory decline. Other factors such as etiology, sex, age at surgery, age at seizure onset, and duration may influence memory decline, but the results are not clear. Our purpose was to identify, in our population of patients and among all risk factors, those that may be predictive of verbal memory decline., Methods: Logistic regression was used to examine the effect of each factor on the postoperative verbal memory index (WMS-R) in 101 patients who underwent a right (n=49) or left (n=52) anterior temporal lobe resection., Results: In the group as a whole, 22 % of the patients demonstrated verbal memory decline of more than one standard deviation. The verbal memory decline was significantly related to surgery on the left side and a high level of verbal memory performance. These factors were significant predictors of decline. The other factors (etiology, sex, age at surgery, age at seizure onset, and duration) were not found to be predictive of this decline., Conclusions: Our analysis demonstrates that the patients who are most at risk of undergoing verbal memory deterioration are those who undergo left-sided temporal resection and have good memory scores preoperatively. The contradictions found in the literature about the other factors could be explained by the diversity of the tests and criteria used to assess memory decline.

Nontumoral epileptogenic lesions account for the major pathological group of surgical specimens obtained from patients with temporal or extratemporal drug-resistant epilepsy. Hippocampal sclerosis remains the predominant etiology, but cerebral cortical dysplasias actually make up the second major cause of nontumoral epilepsy and are increasingly recognized. The percentage of vascular lesions or glial/glio-mesodermal scars remains stable, but the minor or nonspecific lesion group is decreasing because of imaging investigation technique improvement.

No strict relationship has been observed between seizure outcome after surgery and the psychosocial status of patients with refractory epilepsy. The main reason is that seizures are not the sole criteria in evaluating social outcome. The presurgical status in terms of education, the mainly neuropsychological associated handicap, socio-occupational life, and the expectations of the patient must be taken into account. Psychosocial adjustment after surgery needs to be prepared long before surgery, in full cooperation with the patient and his or her family. Standardized tools must be developed to allow for a multidisciplinary approach and an objective evaluation of the psychosocial status of epileptic patients after surgery.

Proposed as an additive symptomatic treatment of refractory epilepsy, vagus nerve stimulation (VNS) has proven to be effective and well-tolerated in patients presenting with refractory epilepsy for whom cortical resection is not indicated. After two years of treatment, the overall reduction of seizure frequency averaged 40%. In 50% of the patients, the frequency of seizures decreased by at least 50%. Moreover, even in absence of a significant reduction of seizures, patients who undergo this treatment reported an improvement in their quality of life. Economic surveys also demonstrate a favorable impact of VNS on the management of refractory epilepsy. Since 1988, 65,000 patients with refractory epilepsy throughout the world have been treated by VNS for this indication (1000 in France). The surgical implantation technique used in our department, the effects of vagus nerve stimulation reported in the literature, and our experience with a cohort of 70 patients with refractory epilepsy who received implants over the last 10 years are described.

Magnetic resonance imaging (MRI) is the prominent imaging modality in the field of epilepsy. MRI plays a major role in describing and recognizing the malformations of cortical development and hippocampal sclerosis and has improved epilepsy patient care. MRI is also accurate in detecting the other major epilepsy etiologies (vascular malformations and cicatricial lesions). Presurgical MRI is relevant since it provides a high-resolution study of the brain.

We report here the results of the first survey on epilepsy surgery activity in France. Data from a questionnaire sent to 17 centers practicing epilepsy surgery were analyzed. All centers responded; however, all items were not completely documented. Over 50 years, more than 5000 patients have been operated on for drug-resistant epilepsy and more than 3000 patients underwent some invasive monitoring, most often SEEG. Currently, nearly 400 patients (including more than 100 children) are operated on yearly for epilepsy in France. Over a study period varying among centers (from two to 20 years; mean, 9.5 years), results from more than 2000 patients including one-third children were analyzed. Important differences between adults and children, respectively, were observed in terms of location (temporal: 72% versus 4.3%; frontal: 12% versus 28%; central: 2% versus 11%), etiology (hippocampal sclerosis: 41% versus 2%; tumors 20% versus 61%); and procedures (cortectomy: 50% versus 23%; lesionectomy: 8% versus 59%), although overall results were identical (seizure-free rates following temporal lobe surgery: 80.6% versus 79%; following extratemporal surgery: 65.9% versus 65%). In adults, the best results were observed following temporomesial (TM) resection associated with hippocampal sclerosis or other lesions (class I: 83% and 79%, respectively), temporal neocortical (TNC) lesional (82%), while resections for cryptogenic temporal resections were followed by 69% (TM) and 63% (TNC) class I outcome. Extratemporal lesional resections were associated with 71% class I outcome and cryptogenic 43%. In children, the best results were obtained in tumor-associated epilepsy regardless of location (class I: 80%). A surgical complication occurred in 8% after resective surgery - with only 2.5% permanent morbidity - and 4.3% after invasive monitoring (mostly hemorrhagic). Overall results obtained by epilepsy surgery centers were in the higher range of those reported in the literature, along with a low rate of major surgical complications. Growing interest for epilepsy surgery is clearly demonstrated in this survey and supports further development to better satisfy the population's needs, particularly children. Activity should be further evaluated, while existing epilepsy surgery centers as well as healthcare networks should be expanded.

Magnetic resonance imaging is an imaging technique allowing morphological and functional study of the brain. Algorithmic and technical advances offer new insights in such brain studies using new approaches to search epileptogenic lesion. Voxel based analysis appears as the dominant methodology to study grey and white matter using the following contrast: T1, T2, T2 relaxometry, magnetization transfert and diffusion weighted imaging. Ongoing development focuses on sulcal morphometry and gyrification index, to improve our understanding of developmental epilepsy. Magnetic resonance spectroscopy appears as a promising tool following availability of high field strength clinical MRI and evidence about its utility in the field of epileptology. Functional MRI benefits from such high field strength and new pulse sequence using diffusion gradients to probe neuronal firing.

Drug-resistant mesial temporal lobe epilepsy with hippocampal sclerosis is associated with anatomical, ultrastructural and functional changes that facilitate generation and spread of epileptic seizures. Intrahippocampal circuits are modified in their transversal lamellar and longitudinal translamellar organization. Neuronal death and the neuroplasticity of surviving cells contribute to major phenomena: an increased hyperexcitability of the hippocampal formation and an increased synchronization of its principal cells. Selective disruption of the epileptic networks that are involved in mesial temporal lobe epilepsy may have a therapeutic effect. We present here the preliminary results of a selective intrahippocampal transection in a chronic model of mesial temporal lobe epilepsy after focal injection of kainic acid in adult mice. A complete transection of the hippocampal formation (including dentate gyrus and hippocampus proper, sparing the fimbria) results in a blockade of ictal activities spread from the generator, a reduction in their frequency and an increase in their duration. In contrast, after a transection sparing the dentate gyrus and hilus, no modification was noted. In this model of mesial temporal lobe epilepsy, longitudinally projecting axonal circuits of the dentate gyrus and hilus appear to be implicated in generation, propagation and interruption of ictal activities within hippocampal formation.

Background and Purpose: Taylor-type focal cortical dysplasias (TTFCD) represent a particular pathological entity responsible for severe drug-resistant epilepsy of extratemporal location. Epilepsy can be surgically cured if complete removal of the lesion can be performed. However, identification on imaging may be difficult and negative standard MRIs are not rare. The frequent location of TTFCD in the central region restrains the possibilities of complete resection. We report a series of patients operated on for intractable epilepsy associated with TTFCD in the central area., Patients and Methods: Between 2000 and 2006, of 34 consecutive patients with TTFCD, 17 had a lesion located in the central area. MRI was considered normal in eight, although in five a subtle gyral abnormality was disclosed on further analysis. A (18)FDG PET scan performed in 16 cases demonstrated focal hypometabolism in 15 that correlated with abnormalities on MRI when visible. SEEG performed in 13 cases revealed typical abnormalities for TTFCD in 10 cases. At resection, cortical and subcortical stimulations of the dysplastic cortex did not elicit a motor response., Results: Postoperative motor or sensory deficit was observed in 13 patients--severe in four--which subsequently resolved completely in seven. Six patients had a minor permanent, motor or sensory deficit. Four patients were reoperated for seizure recurrence and residual dysplastic tissue was found at reoperation in three cases. Average postoperative follow-up was 3.7 years. Sixteen patients (94%) were in Engel Class I (65% in Class IA)., Conclusion: This study suggests that surgical resection of central region TTFCD may be associated with favorable seizure outcome and no or minor functional permanent disability. In cases of seizure relapse, reoperation can be performed without further permanent deficit and lead to seizure-free outcome. Future techniques for intraoperative detection of these lesions could optimize their complete resection in functional areas.

Background and Purpose: No standard treatment for brainstem cavernoma has been established because of the lack of sufficient data about the natural history of these lesions in a highly functional location with potential difficult surgical accessibility., Methods: We present a series of 82 brainstem cavernomas managed at the Sainte-Anne Hospital. Surgery was undertaken for 25 with stereotactic biopsy for 9 and direct surgery for 19 (3 after biopsy)., Results: Surgical outcome was good or fair for 17 patients. Two patients worsened and one died. Biopsy results were disappointing with high morbidity (4 patients with 2 permanent deficits). Histological diagnostic was possible for all biopsies., Conclusion: In light of these results, an active surgical attitude could be proposed for cavernomas in an accessible locations which have produced at least one previous hemorrhage. Stereotactic biopsies for suspect brainstem cavernoma must be avoided.

Cerebral oligodendrogliomas represent more than 30% of glial tumors in adults. Mean age at diagnosis is 41 for grade A and 45(1/2) for grade B, epilepsy being the main revealing symptom (91.5% of A, 76% of B). Survival at 5, 10 and 15 years is respectively 75.5%, 51% and 22.4% for grade A (median: 136 months), and 45.2%, 31.3% and 0% for grade B (median: 52 months). It is influenced by age at diagnosis: median, before 40 years of age, is 12 years for A and 8(1/2) for B; between 40 and 60, is 12 years for A and 4(1/2) for B; over 60, is 4 years for A and 1(1/2) for B. In children, they represent less than 2.5% of cerebral tumors and include 23% grade A and 77% grade B (48.5% WHO grade II and 51.5% WHO grade III). Mean age at surgical removal is 9.5 +/- 5 years. One third are hemispheric, 2/3 are revealed by epilepsy. A complete surgical removal is performed in 60%. Mean survival is 13 years (154 +/- 20 months), with a mortality at 5 and 10 years of 60% +/- 9%, and a risk of recurrence of 54% +/- 9% at 5 years and 46.5% +/- 10.5% at 5 years. The main differential diagnosis of grade A oligo is dysembryoplastic neuroepithelial tumors (DNT). Inversely, thalamic locations, most often grade B, generally present with a motor deficit; complete removal can be achieved in only 15%. The only efficient treatment is chemotherapy, requiring search for chemosensitivity (1p19q deletion, expression of MGMT gene, analysis by MR-spectroscopy and TEP). Logically, surgical debulking should be as wide as possible, but with preservation of neurological and cognitive functions with the help of modern imaging progresses (fMRI, diffusion and perfusion MRI, tractography, PET) and operative techniques (navigation, cortical stimulation, and even opertaive MRI). Tumors presenting clinical and radiological progression should be removed. Operative mortality varies from 0 to 6%, transient morbidity from 20 to 25%, definit morbidity from 5 to 10%. Radiotherapy does not significantly modify survival whatever the delivered dose. The important point is the patient's radiosensitivity: survival is 148 months for radiosensitive patients versus 12 months for non responsive patients. Studies and development of pretherapeutic assays of tumoral radiosensitivity will be a fundamental research axis (PET, sample studies).

Surgical resection is the first step in the treatment of adult supratentorial oligodendrogliomas (OLG). However, the role of resection on prognosis, the most appropriate time for surgery along the natural history of those tumors, and the best operative strategy remain debated. Survival curves after resection vary greatly among reported series, in particular as a result of a persisting confusion in identification and classification of cerebral OLG. Surgical or stereotactic biopsy is the first surgical procedure which enables confirmation of the diagnosis suggested on imaging, assessment of extension of tumor cell infiltration beyond abnormalities limit described an imaging, and currently available molecular biology studies. Biopsies may be the only surgical procedure in patients having a deep-seated tumor with minimal mass effect, or prior to a surgical resection or a "wait and watch" strategy. Surgical resection is indicated for the other patients. However, it has not been demonstrated that time for resection has an influence on survival, excepted in patients with rapidly growing tumors, with mass effect causing increased intracranial pressure. A wait and watch strategy is therefore warranted in patients with a tumor aspect suggestive of a grade A OLG; surgical resection may be indicated later. There is a current trend for maximal safe resection, preserving functional cerebral areas, since truly complete resection of the tumor including infiltration is exceptional. However, from the contradictory results reported to date, one cannot ascertain whether large or complete resection based on imaging is associated with significantly longer survival. Neuronavigation guidance, intraoperative imaging, and cortical stimulation techniques are helpful neurosurgical techniques enabling maximal safe resection with preservation of functional areas.

MRI has dramatically improved the management of cerebral tumors and consequently oligodendrogliomas. T1 and T2-weighted images and gadolinium enhancement are very useful for tumor detection and characterization. Tumor enhancement is of a great prognostic value because it is highly predictive of high-grade oligodendroglioma. Three-dimensional MR images provide, with high precision, the anatomical location and the relationships with functional structures (motor and language areas). The recent technical progress in MRI and the use of diffusion images in the screening of tumors lead to a better definition of the lesion. Comparative analysis of MRI images is helpful to detect eventual relapse and adverse effects of treatment.

The story of the classifications for gliomas is related to the development of the techniques used for cytological and histological examination of brain parenchyma. After a review of these techniques and the progressive discovery of the central nervous system cell types, the main classifications are presented. The first classification is due to Bailey and Cushing in 1926. It was based on histoembryogenetic theory. Then Kernohan introduced, in 1938, the concept of anaplasia. The WHO classification was published in 1979, then revised in 1993 and 2000. It took into account some data from both previous systems and introduced gradually the notion of histological criteria of malignancy. More recently; molecular genetics data and clinical evolution were retained. The Sainte-Anne classification for oligodendrogliomas is based on both histological and imaging data. It includes the notion of spatial histological structure of oligodendrogliomas. Contrast enhancement is closely related to endotheliocapillary hyperplasia. Gliomas classifications are changing and confusions can be made because of lack of reproductibility and misinterpretations of samples.

Oligodendrogenesis is a complex and dynamic phenomenon. Knowledge of the underlying molecular control mechanisms advances steadily, especially in rodents. While the parallelism with human oligodendrogenesis is not fully established, the main characteristics are recognized. Neuroepithelial cells of the neural tube participate in both gliogenesis and neurogenesis. Oligodendrogenesis begins after neurogenesis and stops after birth. It is a focal phenomenon under the control of specific morphogenic proteins, and can generate precursors which are able to proliferate and migrate in the same time. Five steps of oligodendrogliogenesis follow one another acquiring and loosing proteinic markers. They lead to intricated maturation steps for generating myelinizing oligodendrocytes, NG2 cells and precursors of quiescient adult oligodendrocytes.

Introduction: Incidence of cerebral oligodendrogliomas is increasing because of better recognition made possible by improved classifications. We studied a homogeneous series using the Sainte-Anne grading scale in order to better understanding the history of these tumors with or without treatment and to assess prognosis and associated factors., Patients and Methods: A retrospective series of 318 adult patients with oligodendroglioma (OLG) treated at Hôpital Sainte-Anne, Paris (SA) and Hôpital Neurologique, Lyons (L) between 1984 and 2003 was analyzed: 182 grade A OLG (SA + L), 136 grade B among which a homogenous series of 98 (SA) were included. For grade A: age at diagnosis ranged from 21 to 70 (mean: 41), sex ratio was 1.28. For grade B: age at diagnosis ranged from 12 to 75 (mean: 45.5), sex-ratio was 1.58. The main first symptoms were: epilepsy (A: 91.5%; B: 76%), intracranial hypertension (A: 7.9%; B: 14.6%), neurological deficit (A: 5.1%; B: 17.7%). The most frequent locations were: frontal, insular and central for both A and B. Mean size was 55 mm for grade A, 62 mm for B. Calcifications were found in 20% of A, 48.5% of B. No tumor was enhanced on imaging (CT/MRI) in grade A, all but 7 in grade B. All patients underwent surgery either for biopsy (A: 47.2%; B: 53%), or removal which was partial (A: 26.4% vs B: 19.4%) or extended (A: 36.3% vs B: 37.8%). Fifty-six patients underwent 2 procedures and 12 three procedures. Radiotherapy was performed in 76.9% of grade A, and 91% of B patients, in the immediate postoperative period for 71% A and 82.7% B. Chemotherapy was delivered for 36% of grade A (in the event of transformation to grade B or failure of radiotherapy) and 67.5% of B patients. Among grade A tumors, 38% transformed into grade B within a mean delay of 51 months with a mean follow-up of 78 months., Results: Median survival was 136 months for grade A and 52 for grade B. Survival at 5, 10 and 15 was 75.5%, 51% and 22.4% for grade A vs 45.2%, 31.3% and 0% for grade B respectively. In univariate and multivariate analysis, grade A survival was associated with age at diagnosis, tumor size, large removal and response to radiotherapy. Grade B survival was associated with age at diagnosis, wide removal and sharply defined limits of the tumor on imaging., Conclusions: Analysis of both published data and this series underlines many prognostic parameters. It shows that OLG are heterogeneous tumors even in each grade (A and B). Treatment should consequently progress towards more targeted procedures for patients mainly with postoperative radiotherapy and chemotherapy.

Purpose: Definition of homogeneous tumor groups of oligodendrogliomas or oligo-astrocytomas is a basic condition for an adequate evaluation and comparison of the results of treatments in patients from various institutions. However, increasing discordances are observed in the histological diagnosis of these tumors. The main goal of this study is to assess whether, for retrospective studies, MRI data may serve as a common basis for encompassing asymmetry in diagnosis established according to the WHO or Ste-Anne (SA) classification., Patients and Methods: This study included 251 adult patients in whom a SA grade A or B oligodendroglioma or oligo-astrocytoma was newly diagnosed at our institution from 1984 to 2003. Routine histological preparations and post-contrast preoperative MRI/CT-scan were simultaneously reviewed in order to assess the impact on survival of the following features: presence or absence of a polymorphous or gemistocytic astrocytic component, of necrosis and of contrast enhancement (CH); endothelial hyperplasia (EH) assessed as absent, present minor (HE+) or (HE++) when conform to the threshold of HE defined in the SA grading system of oligodendrogliomas. The tumors were graded A: no CH and no EH; in B: CH and /or HE++, and A/B: EH + but no CE., Results: 70.1% of the tumors were classified as "pure" oligodendroglioma, 19.5% as "polymorphous oligo-astroastrocytoma" and 10.3% as "gemistocytic oligo-astrocytoma". In grade A, or B tumors, the presence of a polymorphous or a gemistocytic component had no significant influence on survival; however respectively 53% and 65% of these tumours versus 32% of "pure" oligodendrogliomas were grade B at the time of diagnosis. In either histological subtypes, survival was not significantly different when HE was absent or minor (HE+). After regrouping of the histological subtypes and of the tumors with HE+ or absent, the series included 153 oligodendrogliomas grade A and 98 grade B. Survival in patients with grade A versus grade B tumors was respectively 142 versus 52 months (p<0.0001). In grade B tumors, necrosis had no significant influence on survival. Ring-shaped contrast enhancement surrounding large foci of necrosis was observed in only 4 cases. In tumors with or without CE, patient survival was respectively 148 versus 40 months (p<0.0001). On post contrast MRI done in 235 patients, only 7 tumors (3%) were grade A/B (EH++ but no CH)., Conclusions: From these results and our previous observation that, according to the SA classification of gliomas, only oligodendrogliomas or oligo-astrocytomas may not show CE, we propose that for retrospective studies: 1) tumors diagnosed according to the Ste-Anne classification as oligodendroglioma or oligo-astrocytoma be regrouped in a unique category, 2) independent of their histological type and grade according to the WHO, gliomas that do not show CE be regrouped with SA oligodendrogliomas grade A, 3) concerning gliomas that show CE on MRI: oligodendrogliomas or oligo-astrocytomas WHO grade II or III, as well as WHO secondary glioblastomas or glioblastomas with an oligodendroglial component, be regrouped with SA oligodendrogliomas grade B; however tumors that show ring-like CE surrounding large foci of necrosis and finger-like "peritumoral" edema should be excluded or analysed separately.

Two main classification systems are used in France for the histological typing and grading of oligodendrogliomas: the WHO and Sainte-Anne Hospital (SA) classifications. According to the WHO, the typing of diffuse gliomas is based on the predominant cell type, oligodendroglial versus astrocytic. In contrast, the SA classification is based on the distinction of two patterns of tumor growth, solid tumor tissue versus isolated tumor cells and also relies on imaging and clinical features. According to this approach, the SA classification includes in the category of oligodendrogliomas, the fibrillary or gemistocytic diffuse astrocytomas (WHO grade II) as well as a substantial proportion of astrocytomas WHO grade III, 2) the WHO uses multiple histological criteria for the grading of oligodendrogliomas (grade II versus grade III), including the degree of differentiation, cellular atypia, mitotic activity and necrosis. In contrast, the SA grading of these tumors (grade A versus B) only uses two criteria: the presence or absence of endothelial hyperplasia, and the presence or absence of contrast enhancement. This last criterion allows overcoming the problems related to the representativeness of surgical samples. Difficulties and discrepancies regarding the diagnosis of oligodendrogliomas are in part due to the lack of immunomarker for the identification of tumoral oligodendrocytes. The potential interest of new markers of oligodendroglial precursors for the diagnosis of these tumors will further be discussed.

Image-guided surgery is the central element of therapeutic management of low grade gliomas and consequently, a precise preoperative definition of their spatial extension is necessary. The question of the present work is: do the imaging abnormalities delineate the real spatial development of low grade oligodendrogliomas? A review of the literature showed that MRI on T2-weighted and FLAIR sequences are used to delineate the spatial developement of these tumours and that spectroscopic magnetic resonance imaging is more sensible to appreciate it. Moreover, mathematical models and histological studies suggest that MRI does not indicate the actual spatial extension of low grade oligodendrogliomas. This study focused on histological analysis of biopsy samples performed outside MRI imaging abnormalities in patients who harboured a low grade oligodendroglioma. It showed that isolated tumour cells were identified beyond imaging abnormalities in all of the 17 patients studied. In 15 of those 17 patients, isolated tumour cells were identified in the most distant biopsy samples taken outside imaging abnormalities. Thus, conventional imaging findings, including MRI on T2-weighted and FLAIR sequences, are not able to provide the real spatial development and boundaries of low grade oligodendrogliomas.

Purpose: To review, from a retrospective series of 48 patients presenting with seizures associated with one or more supratentorial cavernoma(s), the natural history of the seizures and outcome according to medical and surgical treatment., Methods: Patients were divided into two groups: group A included patients presenting with a single seizure or rare seizures (n=21), and group B patients having intractable epilepsy (n=27). All received antiepileptic drugs and 35 were operated on (12 in the group A and 23 in the group B). Stereo-EEG was performed in 8 patients in group B. Surgery included lesionectomy alone (n=16), resection of the cavernoma and perilesional tissue (n=7) or tailored corticectomy including the cavernoma (n=12)., Results: The natural history of seizures was different in the two groups: mean age at seizure onset was 25 years in group B and 33 years in group A (p<0.05), seizures were partial in all patients in group B and 8 patients in group A (p<0.05). Seizure frequency and periodicity also varied. Prolonged seizure-free periods were observed. The cavernoma was temporal in 17 patients in group B and 4 patients in group A (p<0.01). In group A, seizure outcome was favorable following surgery or with antiepileptic medication only (7 patients out of 12 operated were seizure-free, as were 5 out of 7 non-operated). In group B, seizure outcome was better after surgery than with medication only (17 patients out of 23 operated were in Engel's Class I, while 3 patients of 4 non-operated patients had persisting seizures despite antiepileptic polytherapy)., Conclusion: Variations in seizure severity in patients harboring cavernomas suggest different therapeutic approaches. In case of unique or rare seizures, surgical resection of the cavernoma is appropriate, but benefits of surgery over antiepileptic medication in terms of seizure control remains unclear. Intractable epilepsy associated with cavernomas is better controlled after surgery rather than with medication only. In these patients, a detailed preoperative work-up is necessary and should be followed by wide resection associated or not with corticectomy, especially in the temporal lobe. Evaluation of outcome after surgery should consider the surgical strategy, antiepileptic medications and the patient's seizure history.

Purpose: Cerebral arteriovenous malformations surrounded by cerebrospinal fluid seemed to exhibit worse response to radiosurgery than others. We searched to verify if this is was true and to find causes., Material: and methods. From our series of 705 patients with cerebral arteriovenous malformations treated by radiosurgery, 3,8% (27/705) had choroidal or cisternal arteriovenous malformations. Revelation mode was hemorrhage in 86% of cases but sometimes headaches occurred before hemorrhage; thus overall hemorrhage rate was 96%. Mean age of revelation was 24. Mean size was 20 mm, mean volume was 4 cc. Spetzler & Martin's grades were 35% in grade II, 43% in grade III and 22% in grade IV. Location was ventricular in 63% of cases and cisternal in 37%. Mean dose at reference isodose was 24 Gy with a higher rate of monoisocenters. We looked for differences between this population of arteriovenous malformations and the rest of the series for patients, treatments prior radiosurgery, cerebral arteriovenous malformations's characteristics, dosimetric parameters and complications. Statistical analysis was done with a Pearson chi2 test and Spearman non parametric correlation test., Results: Obiteration rate was 47.6% with a mean delay of 26 months. Differential characteristics of choroidal or cisternal arteriovenous malformations were: younger age of revelation, higher frequency of hemorrhage, of intra or paranidal aneurysms, of deep unique drainage, a higher Spetzler grade and a smaller rate of complete recovering. Mortality and clinical morbidity due to radiosurgery were 0%. Actuarial rate of hemorrhage after radiosurgery was 4,34% per year and per patient or per hemorrhage. This rate was higher than in the global series. Parenchymal changes seen on MRI were less frequent (26,6%) and less serious (no grade 4 radionecrosis-like parenchymal changes)., Conclusion: Choroidal or cisternal arteriovenous malformations seem to respond less to radiosurgery than others. One potential explanation is the higher frequency of multiafferences of these arteriovenous malformations with anastomoses of anterior and posterior choroidal arteries. However, radiosurgery still stay a treatment of choice for these arteriovenous malformations with little adaptations of the irradiation strategy.

Background and Purpose: Obliteration is progressive after radiosurgery of cerebral arteriovenous malformations. Thus the hemorrhage risk still remains until obliteration. Purposes of this study are to appreciate severity of post-radiosurgery hemorrhages, actuarial risk of hemorrhage and parameters associated with it., Patients: and method. Over 705 patients treated, 46 (6.5%) had one or several hemorrhages. Clinical, anatomic, dosimetric parameters and obliteration rates before hemorrhage were studied. Then, actuarial risks per patient and per hemorrhage were calculated. Correlations between parameters and risk were searched by uni and multivariate analysis by drawing hemorrhage-free survival curves (limit-product Kaplan-Meier) and Cox model., Results: Except one pure ventricular hemorrhage causing death of one patient, only parenchymal hemorrhages were associated with morbidity (80% of cases with 45% of permanent deficits). Overall mortality rate by hemorrhage was 6.5%. Overall morbidity rate was 34.8% and 13.6% for permanent deficit. Mean obliteration rate before hemorrhage was 25%. Actuarial hemorrhage rate were 2.98% per patient and 3.24% per hemorrhage. Actuarial rate per patient increased from 1.46% first year to 5.95% 4(th) year after radiosurgery. Parameters correlated with hemorrhage risk were in univariate analysis size (p=0.01), Spetzler and Martin's grade (p<0.001), dose to reference isodose (p=0.03), Dmin (p=0.08), intra or paranidal aneurysms (p<0.001), and recoverage (p<0.001). After multivariate analysis, only intra or paranidal aneurysms, recovering and Dmin were significantly associated with hemorrhage-free survival after RS., Conclusions: Post-radiosurgery hemorrhages are often sum of hemorrhage risk factors of the cerebral arteriovenous malformation and factors predicting low rate of obliteration. They can be in some cases foreseen but rarely avoided.

The aim of this paper is to analyze retrospectively the reasons for the failure in cerebral arteriovenous malformations radiosurgery. Several factors are evoked and discussed mainly: inaccurate target, intentional partial irradiation, repermeabilization of a previously embolized cerebral arteriovenous malformation. The results suggest the necessity of a complete irradiation of the nidus. The strategy of partial volume irradiation should be avoided, even if it necessitates lowering of the doses in large cerebral arteriovenous malformations. Accuracy in the target determination is required and a complete stereoangiography is necessary.

Background and Purpose: Potential severity of hemorrhage often leads to treat a cerebral arteriovenous malformation. Consequences can be very various and serious. Our first purpose is to define the different types of hemorrhage. Our second purpose is to appreciate more precisely individual hemorragic risk of a cerebral arteriovenous malformation with the study of his angioarchitecture. We performed a prospective study in order to validate a logistic model and a classification previously described., Patients: and method. From the whole series of 705 patients, 57% (n=394) suffered a parenchymal, subarachnoid or ventricular hemorrhage. Logistic model and classification of the hemorrhagic risk were prospectively tested on a consecutive population of 78 patients. Comparisons of theorical (calculated hemorrhagic risk) and real (hemorrhage or not) were performed by non parametric tests., Results: Characteristics and clinical consequences of the hemorrhage were analyzed. Results of the prospective study confirmed data of the classification and showed a hemorrhage risk increasing with the grade: grade Ia 0%, grade Ib 30%, grade II 44%, grade III 57%, grade IV 88%., Conclusion: The study of the angioarchitecture of a cerebral arterio-venous malformation allowed to assess with accuracy his individual hemorrhagic risk. However, this precision may be improved by the study of other parameters of intracranial arteriovenous malformation.

Background and Purpose: Treatment of large cerebral arteriovenous malformations is a challenge for embolization, surgery and radiosurgery too. The purposes of this study is to review results of RS for these large cerebral arteriovenous malformations and to study recent methods which can allow to improve results., Patients: and method. From our series, 112 patients had cerebral arteriovenous malformations higher than 30 mm or 10 cc at time of the radiosurgery. Revelation mode was hemorrhage in 47%, seizures in 42%, headaches in 16%, and progressive neurological deficits in 7%. Discovery was fortuitous in 2.7% of cases. Before radiosurgery, 57% were embolized, 5% were operated, 4.5% were reirradiated and 1.8% were operated and embolized. Distribution according to Spetzler & Martin's grade was following: 0% in grade I, 12% in grade II, 41% in grade III and 46% in grade IV. No patient was in grade V., Results: Global obliteration rate was 39% for large cerebral arteriovenous malformations. In multivariate analysis, only recovering (positive correlation), and previous embolization (negative correlation) were correlated with obliteration rate., Conclusion: Current procedures of radiosurgery allow obliteration in a limited number of large cerebral arteriovenous malformations treated. Other strategies as fractioned irradiation, multileaf microcollimators with single isocenter, radiosurgery before surgery or radiosurgery with protons should improve obliteration rate.

Background and Purpose: Description and critical review of our procedure of definition of the target of irradiation of a cerebral arteriovenous malformations., Patients: and method. Fixation of the stereotactic frame, performing of stereotactic angiography and CT scanner were described, and the target including nidus and venous collectors of the a cerebral arteriovenous malformations defined., Conclusion: The use of several three-dimensional imaging techniques can improve and simplify the method of delineating target with a reduction of radio-induced complications.

The purpose of this study is to evaluate the response of headaches associated with cerebral arteriovenous malformations following radiosurgery and determine the factors that are associated with a positive outcome. From our series of 700 patients whose cAVM was treated by radiosurgery, this study includes 109 patients (48 males, 61 females) with a mean age of 32 when radiosurgery was performed. Headaches resolved in 37 patients (53.7%), decreased in 14 patients (20.3%), persisted in 16 patients (23.2%), and even increased in frequency in 2 (2.9%). Headaches disappeared 2 years following radiosurgery in the majority of patients (31 out of 37: 83.3%). The disappearance rate of headaches was better when it was the only sign of the malformation (64%) or preceded an epileptic seizure (60%) with respect to the case when it preceded cerebral hemorrhage (34.8%). Headaches having the best response were those with 1 to 2 episodes per month (64.9%) and whose episode duration was less than 3 hours (67.6%). Unilateral headaches accounted for 86.5% of resolved headaches whereas bilateral headaches were only 8.1% and even less were the diffuse ones (5.4%). The best evolution was obtained for those associated with a parietal location of the cerebral arteriovenous malformation, followed by rolandic (respectively 55.5% and 50%). Occipital malformation was associated with headache disappearance in 38.1% and temporal with 34.3%, whereas frontal location was associated with headache disappearance in only 25% of the cases. Headache disappearance was often correlated with cerebral arteriovenous malformation outcome after radiosurgery (64.9% of disappearance in case of complete obliteration). However, in 24.8%, headaches disappeared prior to the malformation obliteration and were associated with a decrease of the flow, of the nidus volume or with the nidus disappearance with the persistence of a precocious venous drainage. Headaches disappeared in spite of the malformation persistence in 5.4% of the cases. We found de novo headaches in one patient one year following RS. Another patient presented an episode of headache 12 hours following RS. The course of headaches associated with cerebral arteriovenous malformations after radiosurgery is positive in the majority of cases and correlated to the malformation outcome. This correlation suggests a link between cerebral arteriovenous malformations and headaches and the equal benefit of radiosurgery on both of them.

Progressive neurological deficits are uncommonly associated with cerebral arteriovenous malformations. We present 25 cases (3.54%) from our series of 705 patients treated by radiosurgery. Common characteristics were preferentially: large cerebral arteriovenous malformations, arterial steal, arterial recruitment, venous reflux and deep location. After multivariate analysis, only arterial steal was significantly associated with neurological deficits. Course after treatment was variable and correlated with cerebral arteriovenous malformations regression. Different pathophysiologic hypotheses and their therapeutic consequences are discussed.

The goals of the imaging work-up of cerebral arteriovenous malformations are not only the diagnosis of malformations but also their radioanatomic characterization in order to define an appropriate treatment plan for a given lesion, to evaluate and to compare results of treatment modalities, to look for parameters correlated with a high risk of bleeding and to identify indications of radiosurgery. In this review, the diagnostic value of imaging techniques is discussed: CT scan, MR imaging, vascular explorations. Radioanatomic parameters most useful for therapeutic discussion are defined: feeding arteries, associated aneurysms, size and topography of nidus, draining veins.

This brief study tends to point out that each clinical symptom (hemorrhage, seizures or epilepsy, headaches and progressive neurological deficits) related to cerebral arteriovenous malformations cerebral arteriovenous malformations can be correlated with the anatomy and the angioarchitecture of the cerebral arteriovenous malformations. The most dramatic data were the bipolar-like correlations between hemorrhage and seizures: each parameter correlated positively with hemorrhage was correlated negatively with seizures and vice versa. These data seem to show that cerebral arteriovenous malformations causing seizures have a low risk of hemorrhage, and cerebral arteriovenous malformations with a high hemorrhage risk have a low risk of seizures (excepting with an hematoma). For headaches and progressive neurological deficits, anatomoclinical correlations can improve knowledge of their pathophysiology.

Embolization is an important therapeutic tool for the treatment of cerebral arteriovenous malformations and is a part of the therapeutic strategy. In selected cases, it may lead to a total and permanent cure per se, but in most cases it will be an adjunctive therapy to radiosurgery. The authors review techniques of embolization, perioperative care, complications and applications of the treatment.

Background: and purposes. The purpose of this study was to identify the specific characteristics of headaches associated with cerebral arteriovenous malformations in order to differentiate them from other known entities of headaches such as migraine, cluster headache, and trigeminal neuralgia. This differentiation allows an early diagnosis of cAVM and a treatment to be administrated before any cerebral hemorrhage., Patients and Methods: The study included 700 patients with cAVM and treated by radiosurgery. Out of this series, only 109 (48 males, 61 females, mean age of 33) presented with headaches. Headaches were studied as a possible revelation mode of a cAVM, either as an isolated sign, preceding an epileptic seizure, a cerebral hemorrhage, or associated with a neurological deficit. Analysis concerned 13 clinical parameters and 30 anatomic parameters based on angiography., Results: Headaches were found in 15.6%; they were isolated in 6%. They preceded a cerebral hemorrhage in 12.6%, constituting an early alarm signal when increasing in intensity, frequency and duration. They were associated with seizures or a neurological deficit in 9.6%. We found a predominant female sex-ratio (0.78) and occurrence at a young age (72.3% between 10 and 40 years). Headaches were non-pulsating in 95.3%; nausea, vomiting, light or sound phobia were only found in 4.7%. Headaches were unilateral and homolateral to the malformation in 80%, corresponding to the malformation topography in 97.4% in posterior location and 80% in anterior location. Associated neurological symptoms existed in 20.2%; related to the malformation and lasting 5 to 30 minutes. Duration of pain episodes was less than 3 hours in 77% with a frequency of 1 to 2 per month in 82.5%. Pain was mild and responded to simple analgesics. A family migraine was found in only 3 patients. The angiographic characteristics of the malformations were meningeal afferences, superficial venous drainage and posterior location., Conclusions: Headaches associated with cerebral arterio-venous malformations form a distinct category that can be determined from specific characteristics; this should help an early diagnosis of cerebral arterio-venous malformations in order to start a treatment before the occurrence of cerebral hemorrhage.

The goal of this study is to evaluate the response of epileptic seizures associated with cerebral arteriovenous malformations following radiosurgery and determine the factors associated with a positive outcome. The series included 210 patients (123 men, 87 women) with a mean age of 33. One hundred seventy three patients out of 210 were controlled after radiosurgery. Seizures completely disappeared in 58% of cases and decreased in 18%. Anti-epileptic treatment was stopped in 44%. Sex and age upon seizure appearance were not found as determining factors in patients whose seizures disappeared after radiosurgery. Generalized seizures disappeared following radiosurgery in 90.7%, partial secondary generalized seizures in 53.5%, and partial seizures in only 39.6%. Frontal or temporal malformations are associated with a better response after radiosurgery than sylvian or rolandic location. Seizure disappearance was maximal in case of isolated seizure prior to the radiosurgery (84.4%), was reduced to 61.2% when seizures were limited to 2 to 20, and disappeared in 21.4% in severe seizures. Epilepsy present less than 3 years prior to radiosurgery disappeared in 71.9%; was reduced to 45% when the delay was over 3 years. Seizures disappeared in 64.4% when the malformation was obliterated, they were reduced to 46.9% when malformations persisted. The persistence of severe seizures following radiosurgery was linked to persistence of the malformation. On the other hand early seizures occurred in 3.3% during the first 48 hours following radiosurgery; they disappeared in 72.2% after that period. De novo seizures also occurred in 1.7% and disappeared in all these cases. In conclusion, seizures associated with cerebral arteriovenous malformations respond well to radiosurgery; the malformation obliteration evokes its association with the seizure; the seizure disappearance in spite of persistence of the malformation evokes the positive effect of radiotherapy on epileptic seizures.

Background and Purpose: Purposes of this study are to describe different parenchymal changes seen after radiosurgery of cerebral arteriovenous malformations and the clinical symptoms which can be associated, and risk factors correlated with them., Patients: and method. From the whole population of 705 patients with a cerebral arteriovenous malformations treated by radiosurgery between 1984 and 1998, clinical from 615 patients and post radiosurgery MRI data from 367 patients were reviewed. Neurological deficit occurred in 5.37% of cases and was still persistant in 1.46% of cases. Delay of occurrence of deficits ranged from 6 to 83 months (mean: 27 months, median: 15 months). Parenchymal changes seen in MRI were classified in 4 grades: 1 without parenchymal changes, 2 hypersignal in sp T2, 3=2 with homogenous enhancement with gadolinium, 4 with hyposignal in spT1 and annular irregular enhancement. Several parameters (size, volume, angioarchitecture of the cerebral arteriovenous malformation, dosimetric parameters) were studied and correlations were searched by uni and multivariate analysis with occurrence and delay of occurrence of deficits or parenchymal changes., Results: In multivariate analysis, only size was significantly correlated with occurrence of parenchymal changes (p=0.0016); only size of the malformation was significantly correlated with delay of occurrence of parenchymal changes (p=0.0082); only grade 4 was correlated with occurrence of neurological deficit (p<0.00001). However, when only "a priori" parameters (known before radiosurgery) are introducted in logistic model, size taille (p=0.02) and hypoplasy of a sinus (p=0.0049) are significantly correlated with occurrence of neurological deficit. Only parenchymal changes grade 4 was significantly correlated with delay of occurrence of a neurological deficit (p<0.00001). However, when only a priori parameters (known before radiosurgery) are introducted in logistic model, only arterial steal (p=0.054) was significantly correlated with delay of occurrence of a neurological deficit., Conclusion: Parenchymal changes are various in expression, signification and clinical symptoms associated with them. They must be known and recognized for better prevention and symptomatic treatment as well.

Background and Purpose: Radiosurgery has been used in a few cases of cerebral arteriovenous malformations in children. From data of our series and from literature, we tried to determine the place of this therapeutic modality for treatment of pediatric cAVMs., Patients and Method: From the whole series of 705 patients treated by radiosurgery between January 84 and December 1998, 55 were under 15 years of age. Revelation mode was hemorrhage (70.9%), seizures (12.7%), headaches (10.9%) and neurological progressive deficits (1.8%). Size ranged from 5 to 50 mm (mean: 22 mm) and volume ranged from 1 to 16 cc (mean: 4 cc). Treatments prior radiosurgery were embolization alone (28%), surgery alone (16.7%), embolization plus surgery (5.5%). Distribution of cerebral arteriovenous malformations according to Spetzler & Martin's grading were 10% in grade I, 35% in grade II, 40% in grade III, 15% in grade IV and 0% in grade V., Results: Overall obliteration rate was 59% with an exaustibility rate of 80%. Obliteration rate was dependent on size and volume (100% for<1 cc cerebral arteriovenous malformations sized, 73% from 1-4 cc, 40% from 4-10 cc and 100% (but n=1)>10 cc; p=0.019), sex (84.2% for boys and 40% for girls), of angioarchitecture and Dmin (p=0.01). After multivariate analysis, only Dmin was nearly correlated with the obliteration rate (b=0.462 7; SE=0.244 0; p=0.057). Actuarial risk of hemorrhage was 2.12% per year per patient. Patients with seizures were seizure-free without any drugs after radiosurgery in 80% of cases. Parenchymal changes were mainly hypersignals spT2 in MRI., Conclusion: Radiosurgery seems to be an efficient treatment with a low morbidity for cerebral arteriovenous malformations in children.

Purposes: The purpose of this study is to analyze the characteristics of epileptic seizures associated with cerebral arteriovenous malformations treated by radiosurgery., Patients and Methods: Seven hundred and two patients are studied (406 men, 296 women) with a mean age of 33 when radiosurgery was performed. 210 patients had epileptic seizures; those seizures either uniquely revealed the arteriovenous malformation or were associated with other clinical signs. Seizures occurring after an hemorrhage were excluded from the study. We studied the age category upon seizures occurrence, the malformation localization, type of seizure (partial, generalized, or partial secondary generalized), the history related to the epileptic seizures, the clinical topographical correlation, the number of seizures, their duration prior to the radiosurgery, antiepileptic treatment provided, electroence-phalographic characteristics, angiographic characteristics, prior endovascular or surgical treatment, and factors associated with severe seizure., Results: Epileptic seizures revealed the arteriovenous malformations in 30% of the cases; they predominated between 10 and 40 years (80.5%) with a peak between 20 and 30 years. The malformation localization most frequently associated with seizures were temporal and rolandic for partial seizures, whereas frontal or sylvian localization were predominantly associated with generalized seizures. The type of seizure was, in all cases, related to the malformation localization because of the initial signs seizure. 47.6% of the patients showed a limited number of seizures (between 2 and 20), 27.6% showed only a single seizure, and 24.7% suffered from severe epilepsy. Generalized seizures are unique in the majority of cases (61.8%) whereas partial seizures were severe in 47.6% of the cases. The majority of patients are under a single-drug therapy (79%). The EEG patterns showed a small increase of slow and sharp localized waves in patients presenting seizures compared with the global population (27% and 15% respectively). Severe seizures predominated between 10 and 19 years of age; seizures were exclusively partial in 59.6% of the cases and secondary generalized in 40.4%, their frequencies were between 1 to 4 per month in the majority of cases (59.1%), the delay was often more than 10 years (57.7%), status epilepticus was rare (5.8%), temporal and rolandic localization were predominant, social and professional insertion of the patient was preserved in 87.5% of the cases, and seizures did not provoke any death. The angiographic characteristics associated with seizures were: superficial localization, volume, superficial venous drainage, recruitment and venous divergence., Conclusion: Anatomic and topographic characteristics of cerebral arterio-venous malformations may provide information on the features of seizures associated with them.

Background and Purpose: After a review of the main radiosurgical published series, to evaluate our own series of 705 patients with cerebral arteriovenous malformations treated by radiosurgery alone or in combination with embolization or surgery., Patients: and method. From January 1984 to December 1998, 705 patients were treated by a multidisciplinary team including neurosurgeons, neuroradiologists, radiophysicians and radiotherapists. Age of revelation of the cerebral arteriovenous malformations ranged between birth to 73 years (mean 27, median 25). Age at time of radiosurgery ranged between 7 and 75 years (mean 33, median 31). There were 410 males for 295 females (sex- ratio 1.4). Symptoms of revelation were hemorrhage for 59%, seizures for 23%, headaches for 14% and progressive deficits for 4%. Discovery of cerebral arteriovenous malformation was fortuitous in 4% of cases. Repartition following Spetzler's grading was 12% in grade I, 36% in grade II, 40% in grade III, 12% in grade IV and 0% in grade V. Maximal size ranged between 4 and 60 mm (mean 23, median 20). Volume ranged between 0.2 and 24.3 cc (mean 3.8, median 2.8). Majority of cerebral arteriovenous malformations were large size (42% with size higher than 25 mm) and large volume (54% higher than 10 cc. 54% of patients had treatment prior radiosurgery: 38% had embolization, 10% were operated, 4% were treated by radiosurgery (reirradiation) and 3% were operated and embolized., Results: Overall complete obliteration rate was 55%. The obliteration rate was correlated with size (77% for cAVMs lower than 15 mm, 62% for cerebral arteriovenous malformations between 15 and 25 mm, and 44% for cerebral arteriovenous malformations higher than 25 mm), with volume (94% for cerebral arteriovenous malformations lower than 1 cc, 64% between 1 and 4 cc, 48% between 4 and 10 cc, and 62% for cerebral arteriovenous malformations higher than 10 cc), dose at reference isodose, minimal dose, morphological parameters (presence of dural components, arteriolovenous fistula, plexiform angioarchitecture, arterial steal, arterial recruitment, deep exclusive drainage, venous plicature, venous confluence, venous ectasia, venous reflux), sectional topography and good recovery of the target. Embolization was a confusion factor not associated with obliteration rate. After multivariate analysis, only Dmin and complete coverage of the cerebral arteriovenous malformations were correlated with obliteration rate. Delay of obliteration was significantly correlated after multivariate analysis with Dmin, complete coverage, arteriolovenulary angioarchitecture (positive correlation) and venous ectasia (negative correlation)., Conclusion: Overall complete obliteration rate is unreliable data to assess efficacy of radiosurgical method in the tretment of cerebral arteriovenous malformations. The obliteration rate must be interpretated after stratification on several morphological and dosimetric parameters.

Risks of bleeding from partially or unobliterated cerebral arteriovenous malformations remain unchanged. A complementary treatment should be indicated after radiosurgery. In this brief review the following data are discussed: the reasons of second treatment, imaging follow-up of cerebral arteriovenous malformations after radiosurgery, types of cerebral arteriovenous malformations to be treated, timing and therapeutic modalities.

We report a multicentric outcome study of cerebral metastases in 174 patients collected consecutively in three neurosurgical centers in France. Age and condition of patients allowed surgical operation; the sex ratio was 1.67. The revelation mode was usual. However tumoral hemorrhage frequency was 3.4%. Primitive cancers were usual cancers as in oncological series with a higher proportion of radioresistant metastases. Mean dimension was 30 mm with mean of 1. 25 metastases per patient. Supratentorial localization was more frequent and the metastasis revealed cancer in 40% of cases. Global median survival was 12.1 months. Factors correlated with the survival in our series of patients with cerebral metastases were: solitary cerebral metastasis, extracerebral extension of cancer, treatment of primitive cancer, complete excision and post operative cerebral radiotherapy.

Background and Purpose: Complete resection of a pituitary adenoma is feasible in almost 50 to 70% of cases. Dealing with the remnant is still controversial. Three positions can be discussed: second stage operation, radiation therapy, clinical and radiological follow-up., Methods: Among 108 pituitary macroadenomas operated on during the last 10 years, 54 patients (50%) with partially removed tumors were studied. Three cases best illustrating the evolution of these remnants are reported., Results: Late MRI performed 4 to 6 months postoperatively is more valuable than early MRI in the evaluation of these remnants, which can remain stable for a long period of time without any adjuvant treatment., Conclusion: MRI should be performed 4 to 6 months postoperatively. Rather than performing radiation therapy systematically for asymptomatic patients, these patients should be followed up at regular intervals and treated only in case of regrowth of the tumor.