Brain, Vol. 122, No. 4, 709-726,
April 1999
© 1999 Oxford University Press
Meningioangiomatosis
A comprehensive analysis of clinical and laboratory features
1 Departments of Clinical Neurological Sciences, 2 Neuropathology and 3 Diagnostic Radiology, University of Western Ontario, London, Ontario, Canada
Correspondence to:
Dr Samuel Wiebe, Department of Clinical Neurological Sciences, London Health Sciences Centre, University Campus, 339 Windermere Road, London, Ontario, Canada N6A 5A5 E-mail: swiebe{at}julian.uwo.ca
 |
Abstract |
|---|
 Top Abstract IntroductionMethodResultsDiscussionReferences |
|---|
Meningioangiomatosis (MA) is a rare, benign, focal lesion of the leptomeninges and underlying cerebral cortex characterized by leptomeningeal and meningovascular proliferation. It may occur sporadically or in association with neurofibromatosis type 2. Previous reports have emphasized histological and imaging features. Data on the management of these patients are sparse, and electrophysiological features of MA lesions have not been published. We assessed the clinical, electrophysiological, histopathological and imaging features as well as the surgical outcome in MA, and compared MA with and without neurofibromatosis. Seven patients with MA at our centre were investigated and their outcome was assessed. A review of the literature is included. MA exhibits a wide range of clinical, imaging, histopathological and electrophysiological features, making the diagnosis difficult. Sporadic MA cases are not associated with neurofibromatosis and the two disorders are genetically distinct. Medically refractory, localization-related epilepsy is the commonest presentation in sporadic cases, but atypical presentations also occur. Unlike sporadic cases, MA with neurofibromatosis is often found incidentally, does not produce seizures, occurs less frequently (ratio of 1 : 4), and is multifocal. MRI findings in MA correspond to the histological picture. However, the appearance on imaging is non-specific and may suggest cystic atrophy, angioma and tumours. Several abnormalities have been found in close proximity to MA lesions, i.e. meningioma, oligodendroglioma, arteriovenous malformation, encephalocoel and orbital erosion. In spite of histopathological diversity, MA lesions are either predominantly cellular or vascular. Immunohistochemical results are inconsistent among cases, add little to the diagnosis, and do not support a meningeal origin. Electrocorticographic recordings from the surface and within MA lesions revealed a spectrum of electrophysiological expressions. Intrinsic epileptogenicity of MA lesions was documented in some cases. Epileptogenicity was confined to the perilesional cortex in some patients and it was complex (extralesional, multifocal, generalized) in others. Only 43% of our patients became seizure-free postoperatively compared with 68% previously reported, and >70% of our patients and those in the literature continued to require antiepileptic drugs. This is in keeping with the diverse electrophysiology of MA and suggests a less optimistic postoperative outcome than previously recognized.
meningioangiomatosis; electrophysiology; histopathology; imaging; seizure outcome
ECoG = electrocorticography; GFAP = glial fibrillary acidic protein; MA = meningioangiomatosis; NF = neurofibromatosis
|
Introduction |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Meningioangiomatosis (MA), a focal lesion of the leptomeninges and underlying cerebral cortex, was originally described in association with von Recklinghausen's disease (Bassoe and Nuzum, 1915
) and has since been known to also occur sporadically. Histopathologically, it is characterized by cortical meningovascular proliferation and leptomeningeal calcification (Halper et al., 1986). MA presents clinically as partial seizures that are difficult to control or as an incidental finding in individuals who are asymptomatic, experience vague neurological complaints (headache, nausea) or have neurofibromatosis (NF) (Halper et al., 1986; Paulus et al., 1989; Partington et al., 1991; Prayson, 1995). The electrophysiological characteristics of seizure-producing MA lesions have not been well described. Furthermore, although epilepsy surgery is briefly described as producing good results in MA (Kasantikul and Brown, 1981; Halper et al., 1986; Sakaki et al., 1987; Kuzniecky et al., 1988; Liu et al., 1989; Ogilvy et al., 1989; Partington et al., 1991; Tien et al., 1992; Gomez-Anson et al., 1995; Prayson, 1995), outcome assessment is not described, and attention has not been given to the diagnostic and therapeutic problems faced by clinicians treating this condition. Using our experience and a systematic review of the literature, we delineate the spectrum of clinical, electrophysiological, imaging and pathological features of MA. In addition, we compare the features of MA with and without phakomatoses and assess the surgical outcome in symptomatic cases. |
Method |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Patients
We reviewed neuropathological records from 1980 to 1997 for cases of MA at the London Health Sciences Centre in London, Ontario, Canada. Clinical features and results of imaging and electrophysiological investigations were obtained. Data on the surgical outcome were obtained through clinicians' interviews or by telephone using standard forms that enquired about seizures, antiepileptic drugs and neurological deficits. All clinical data were independently obtained by two of the authors to ensure reproducibility. Disagreements were resolved at conference. All imaging and EEG studies were reviewed blind.
Subdural electrocorticographic (ECoG) recordings employed strip electrodes consisting of linearly arrayed platinum contacts at 10-mm intervals, led to a seven-pin connector by stainless-steel wires through Silastic-coated Teflon tubing. Intraoperative ECoG used carbon-coated ball electrodes in monopolar fashion. Cortical exposures were determined by clinical, imaging and ictal/interictal EEG findings.
The following immunohistochemical stains were performed on each case to determine the origin of proliferating or lesional cells: cytokeratin (low molecular weight, cam 5.2) to assess epithelial differentiation; vimentin, a non-specific marker of mesenchymal cells; epithelial membrane antigen, a marker of arachnoid cap cells, positive in most meningiomas; S-100 protein, found in cells with neuroectodermal or Schwannian differentiation; glial fibrillary acidic protein (GFAP), an intermediate filament in cells of astroglial differentiation; smooth muscle actin, a marker of smooth muscle in blood vessel walls; neuronal specific enolase, a non-specific marker of neuronal cells; and factor VIII, an endothelial cell marker.
Literature review
We used free-text and MeSH terms to search the following databases: Medline® 1966 to April, 1997; CANCERLIT® 1983–96; Current Contents® 1996–97. In addition, we searched neurological and neuropathological textbooks and their references for additional cases. Cases were included if histopathological descriptions contained at least one of the two classical features of MA, i.e. leptomeningeal proliferation and meningovascular proliferation, and if the authors classified them as MA. The following information was sought: demographic features, association with phakomatoses in patients and their family, clinical features, location of MA, imaging and EEG data, surgical treatment and outcome, and histopathological findings. Literature searches and data abstraction were performed independently by two of the authors to ensure accuracy and reproducibility. Disagreements were resolved at conference.
Analysis
Descriptive statistics were used to analyse the distribution of variables of interest. Ninety-five percent confidence intervals for means and binomial proportions were obtained. Continuous variables were compared with independent-group t tests. For discrete variables, 
2, Fisher's exact or binomial tests of proportions versus standard were used. Association was expressed as the odds ratio where indicated.
|
Results |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Patients
Our records yielded seven cases of MA; all were surgical specimens (Table 1
). Median age at presentation was 17 years (range 10–37 years). All patients had refractory seizures for 1–18 years (median 7 years). Two patients had exclusively simple partial seizures, one with visual (right mesial occipital-parietal lesion) and one with motor phenomena (left Rolandic lesion). All other patients had complex partial seizures, with secondary generalization in two and a tendency to status epilepticus in one. Case descriptions follow.
|
Case 1
An 18-year-old woman presented with increasingly frequent and severe simple partial seizures since age 10 years, consisting of coloured diamonds in the left lower visual field followed by experiential phenomena. Neurological, general and visual field examinations were normal. Scalp EEGs showed minimum posterior head slowing with no epileptiform activity. Head CT showed a 3 x 5 cm, partially calcified, gyriform mass in the right supracalcarine, anterior mesial occipital region, resembling a partially calcified arteriovenous malformation. Cerebral angiogram revealed an avascular mass. Brain CT (Fig. 1A
) and MRI (Fig. 1B) showed a cortical, calcified, gyriform mixed signal pattern. MA was suspected. ECoG revealed perilesional cortical spikes. A partial lesionectomy was performed. Histopathology revealed MA. The patient has a persistent, incomplete, left homonymous hemianopia but remains seizure-free without antiepileptic drugs 7 years later.
|
Case 2
An 8-year-old diabetic girl presented with a violent headache, vomiting, transient dysarthria and right facial weakness. Initial head CT showed an acute, left inferior Rolandic parenchymal haemorrhagic lesion. The cerebral angiogram was normal. One month later, simple partial seizures developed consisting of right facial numbness with hemicorporeal spread, slurred speech, drooling, a peculiar taste, and postictal dysphasia. Neurological and ophthalmic examinations were normal. A cutaneous haemangioma was found at the T6 dermatome. Her mother had von Hippel–Lindau disease. Subsequent MRI revealed a 2-cm haemorrhagic lesion with blood of various ages and well-demarcated surrounding hemosiderin. There was mild oedema and a small amount of posterior gadolinium enhancement that was probably related to reactive changes in the wall of the haematoma (Fig. 2
). This was considered highly suggestive of a cavernous angioma. There were no features on imaging to suggest the diagnosis of MA. An extensive search for systemic and spinal angiomata was negative. Scalp EEG showed left temporal–central–parietal spikes and generalized spike wave. ECoG was normal under general anaesthesia. The location of the lesion near the dominant Rolandic cortex allowed only a semicomplete lesionectomy. At surgery, the lesion gave the impression macroscopically of a subcortical cavernous angioma. Histopathology revealed exclusively cortical MA which had previously bled several times into the brain parenchyma. This explained the impression of a cavernous angioma on imaging. The patient was seizure-free for 1 year. Thereafter seizures recurred and remain refractory to antiepileptic drugs 6 years later.
|
Case 3
A 21-year-old male presented with progressively intractable complex partial seizures with secondary generalization since age 16 years. Seizures consisted of experiential phenomena, loss of awareness, staring, drooling, automatisms and postictal dysphasia. Neurological examination was normal. Head CT scan demonstrated a left inferior lateral temporal enhancing mass interpreted as a glioma. Scalp EEG showed left temporal and left hemisphere spikes and delta-band activity. Clinically typical seizures originated in the left temporal area. ECoG revealed widely synchronous left temporal and perisylvian sharp waves and delta-band activity. Partial lesionectomy and temporal lobectomy were performed sparing the amygdala and hippocampus because of failed ipsilateral carotid amytal memory testing. Post-resection ECoG showed prominent spikes overlying the residual lesion. As seizures continued unabated, a discrete, malignant-appearing lesion was completely removed in a second operation. Electrodes directly on and within this lesion revealed very active epileptogenesis (Fig. 3
). Because of the striking hypercellularity, gliosis, vascular proliferation and leptomeningeal invasion and because the histopathological features of MA were not widely recognized at the time (1978), the erroneous diagnosis of glioblastoma multiforme was made by the pathologist. Ten years after the second surgery intractable seizures continued. Scalp EEGs showed left anterior temporal spikes and seizures. A left amygdalohippocampectomy was performed following a successful left carotid amytal test. ECoG showed very active hippocampal spike activity abolished by the resection. Hippocampal sclerosis was found on histopathology. The patient remains seizure-free 10 years later on a single antiepileptic drug. Re-examination of the histological slides revealed the characteristic features of MA, i.e. a plaque restricted to the cortex and meninges but sparing the white matter, consisting of a dense proliferation of cells and blood vessels, in the absence of pleomorphism and mitotic activity (Fig. 8H).
|
|
Case 4
A 17-year-old male presented with a 1-year history of progressively frequent simple and complex partial seizures consisting of auditory illusions, visual blurring and formed visual hallucinations (people, animals), loss of awareness, staring, pallor and diaphoresis followed by headaches. General and neurological examinations were normal. Head CT scan showed a subtle right superior temporal, hyperdense, gyriform, mildly enhancing lesion consistent with a low-grade glioma. Brain MRI showed a heterogeneous lesion surrounded by low signal on T2 sequences, suggesting MA (Fig. 4A and B
). Scalp EEGs showed right mid-anterior temporal spikes and subclinical seizures. ECoG revealed frequent spikes directly over and around the lesion and in the hippocampal depth electrodes independently (Fig. 4C). The indurated lesion, anterior temporal lobe, amygdala and hippocampus were resected. Histopathology showed MA only. Seven years after surgery medically refractory seizures persist with modest improvement in frequency and severity.
|
Case 5
A 36-year-old woman presented with progressively severe complex partial and frequent secondarily generalized seizures starting at age 19 years, consisting of deja vu, loss of awareness, vacant staring, drooling, right hand dystonic posturing and bimanual automatisms without postictal deficit. General and neurological examinations were normal. Head CT, previously reported as normal, was not available for review. Brain MRI demonstrated a 2.5 cm (axial plane) right temporal pole cystic lesion (Fig. 5A
) with increased cortical signal on proton density images (Fig. 5B), interpreted as cystic encephalomalacia or tumour (Fig. 5A–C). There were no MRI changes suggestive of mesial temporal sclerosis. Scalp EEG revealed right temporal-frontal spikes and delta-band activity. Seizures had an ambiguous onset in the right hemisphere. Subdural EEG recordings showed independent right mesial temporal, inferior anterior temporal and orbitofrontal spikes. Seizures started in the right temporal pole and hippocampus. A right anterior temporal lobectomy including 1.5 cm of the anterior hippocampus was performed. Histopathology showed MA and mesial temporal sclerosis. She continues to have generalized motor and partial seizures 5 years after surgery, with moderate improvement in frequency and severity.
|
Case 6
A 33-year-old woman had refractory simple and complex partial seizures starting at age 26 years, consisting of a cephalic sensation, intrusive thoughts, inability to understand or speak, loss of awareness, right hand automatisms and postictal dysphasia. General and neurological examinations were normal except for otosclerosis. Head CT was normal. Brain MRI demonstrated a left amygdalohippocampal mass whose signal was hyperintense on T2-weighted images (Fig. 6A
) and homogeneously isointense with white matter on T1-weighted images (Fig. 6B), consistent with a low-grade tumour. EEG revealed left anterior temporal spikes, delta-band activity and seizures. ECoG revealed prominent spike activity in electrodes within and directly over the lesion (Fig. 6C). Left anterior temporal lobectomy and amygdalohippocampectomy were performed. Histopathology revealed MA entirely replacing the amygdala and hippocampus. The patient has sporadic seizures 3 years after surgery on single antiepileptic drug therapy.
|
Case 7
A developmentally normal 12-year-old boy experienced typical febrile convulsions at age 2 years. At age 6 years, refractory complex partial seizures developed consisting of a blank stare and bimanual and oroalimentary automatisms without postictal deficit. General and neurological examinations were normal. Head CT disclosed a non-enhancing left anterior mesial occipital-cingulate lesion. On MRI, the cortical cystic lesion was hypointense on T1-weighted images and hyperintense on T2-weighted images with surrounding increased T2 signal (Fig. 7A and B
). A tentative diagnosis of dysembryoplastic neuroepithelial tumour was given. Scalp EEGs showed generalized spike wave and independently occurring frontal and occipital epileptiform discharges (Fig. 7C). Seizures had an ambiguous onset without occipital predominance. Subdural electrode recordings showed spikes principally in the left mesial temporal and convexity regions, but also in the left mesial frontal and posterior cingulate area. Seizures originated in the perilesional cortex with late propagation to the ipsilateral posterior temporal convexity, hippocampus and mesial frontal region. The lesion and adjacent cortex were resected. Histopathology showed MA. He remains seizure-free 1 year after surgery on two antiepileptic drugs.
|
Histopathology
All our patients' MA lesions were confined to the cortex, with variable involvement of the overlying leptomeninges. Although all lesions shared unifying features, i.e. cortical vascular proliferation and perivascular cell proliferation, each was unique. Cases were easily classified into those with predominantly cellular (patients 3–7) and those with predominantly vascular (patients 1 and 2) lesions.
Predominantly cellular cases demonstrated moderate to high cellularity (Fig. 8A and B
). Varying architecture was noted, consisting of focal areas of storiform, rhythmic palisading and fascicular patterns (Fig. 8C). All cellular cases had lesional cells that in areas appeared to emerge from the perivascular location and infiltrate the cortex. This occurred centrally within the lesions, where cellularity was most dense. Peripherally, the perivascular relationship of the cells became evident (Fig. 8D). The blood vessels in these cases had a similar appearance, i.e. they were thin-walled, slit-like and increased in number. Predominantly vascular cases contained thick-walled, hyalinized and calcified blood vessels with minimal perivascular cell proliferation. (Fig. 8E and F). In one case (patient 2) the MA lesion showed evidence of haemorrhage. Neither case demonstrated cortical invasion by proliferating cells.
Despite cellularity, the proliferating cells in all cases had bland cytological features, without significant atypia, mitoses or necrosis. Two of the cellular (patients 1 and 2) and both vascular cases contained a meningeal component. Of these, three showed calcification in areas of meningeal proliferation and within the cortex (patients 2, 5 and 6). Extensive pericellular reticulin deposition occurred in two of the five cellular cases (patients 2 and 4) (Fig. 8G). In all other cases, reticulin was confined to blood vessels.
Bielschowsky staining and tau immunostaining demonstrated neurofibrillary tangles in the cortex adjacent to the lesion only in patient 4. One case (patient 3) demonstrated cortical dysplastic neurons adjacent to the MA focus (Fig. 8H). All cases showed gliosis within and adjacent to the lesion.
Correspondence between imaging and pathological findings was good, For example, in patient 5 a cyst identified on MRI corresponded with a white matter cyst underlying the cortical MA lesion.
On immunostaining, the proliferating cell population expressed vimentin uniformly. Results for other markers varied, i.e. three cases (patients 3, 5 and 6) showed focal GFAP positivity, two were focally positive for neuronal specific enolase (patients 3 and 5) and one was positive for S-100 (patient 3). All were negative for cytokeratin, factor VIII and actin. Our results illustrate the variability of immunostaining and suggest that proliferating MA cells do not correspond to a known, normally occurring cell type. Notably, epithelial membrane antigen, a known meningothelial cell marker, was focally positive in only two cases (patients 5 and 6).
Literature review
Fifty-six cases with histopathological confirmation of MA were found in the literature. Only 13 (23%) occurred in association with NF (Table 2). Forty-three cases without NF have been described in 18 series prior to our report, starting with the initial description by Kasantikul and Brown (1981) (Table 3). The following sections refer to our series and the literature cases combined.
|
|
Demographic features
The mean age at the time of diagnosis was 28 and 21 years for MA with and without NF, respectively. Because MA with NF was diagnosed incidentally at the time of autopsy in most cases, the age difference reflects only the average survival of patients with NF in earlier series. MA without NF presents at an early age; 12, 25 and 60% of patients were <5, <10 and <20 years old, respectively, and only 10% were
40 years old. The youngest patient in this group was 9 months old (Blumenthal et al., 1993). MA occurred twice as frequently in males as in females in both groups (P = 0.05). The duration of seizures prior to diagnosis of MA was 
1 year in 30% of patients, from 2 to 9 years in 46% and 10 years in 24%. A family history of phakomatosis (NF, type unknown) was found in one patient with NF, and in one of our patients without NF whose mother had von Hippel–Lindau disease (Tables 1–3).
Location of MA
The chance of having multifocal MA lesions was almost 10 times higher (odds ratio = 9.78) in the presence of NF than in its absence. Five cases (38%) in the NF group had multiple MA foci compared with only one (2%) of those without NF (P = 0.017). MA involved the cortex in 90% of all patients. However, extracortical lesions occurred three times more frequently (odds ratio = 2.6) with NF (23% compared with 6% for cases without NF). Unusual MA locations included the third ventricle (Beck, 1938) and the pulvinar-cerebral peduncle (Rubinstein, 1963) in those with NF, and the corpus callosum (Jun and Burdick, 1984), trigeminal ganglion (Garen et al., 1989), and the medulla intra-axially (Kollias et al., 1994) in those without NF. Seventy percent of all lesions occurred in the frontal temporal region, and the single most frequent location was the temporal lobe (40%). In the posterior head, the distribution was similar between the occipital and parietal lobes (four cases each). The mesial aspect of the occipital, frontal and temporal lobes was involved in three, four and two patients, respectively. Although the numbers were small, there was a statistically significant association between occipital lobe and mesial surface involvement (odds ratio = 7.7). In 46 patients with unilateral lesions, the right hemisphere was affected twice as frequently as the left [32 and 14 cases, respectively (P = 0.005)].
Clinical presentation
All but one of the cases with NF were found incidentally at autopsy (Table 2). One patient presented with headaches, drowsiness and a right frontal mass on CT. After partial resection and radiotherapy the patient developed seizures but survived for many years. This prompted review of the specimen and reclassification from astrocytoma to MA (Huson et al., 1988).
In MA without NF, seizures were the only or predominant problem in 42 patients (85%) (Tables 1 and 3). Seizures were the exclusive clinical problem in all patients with temporal and opercular lesions. Information on seizure type was available for 29 of 39 patients (74%). Eighteen of the 29 patients (62%) had only partial seizures without secondary generalization. Localization of seizure onset by clinical features corresponded with MA location. There was no association between location and tendency to generalized seizures. Seizures were refractory to antiepileptic drugs in 33 of 39 (85%) patients. In the remaining six patients, early imaging and surgery were done because of focal neurological deficits in four patients (Matias-Guiu et al., 1988; Wilson et al., 1991; Tien et al., 1992; Blumenthal et al., 1993) and following single seizures in two patients (Partington et al., 1991; Lopez et al., 1996). Febrile convulsions occurred in four patients, two with mesial occipital (Tien et al., 1992) (patient 7) and two with opercular and temporal MA, respectively (Worster-Drought et al., 1937; Halper et al., 1986). Atypical febrile convulsions occurred in one patient (Tien et al., 1992).
Eight patients (16%) did not have seizures. They presented with headache (three cases with frontal or anterior callosal MA), facial pain (two patients with trigeminal and orbital MA) or lower cranial nerve palsies (one case with intramedullary MA), or were asymptomatic (two patients with incidental parietal MA). Atypical clinical features included fatal subarachnoid haemorrhage from right frontal-mesial MA (Auer et al., 1982), and facial pain and papilloedema from right frontal-temporal MA eroding the orbital roof (Louw et al., 1990).
Electrophysiology
Interictal scalp EEG data, available for 24 of 39 patients with seizures, showed congruence in location of epileptiform discharges and MA in 18 (75%) patients. Extralesional multifocal spikes or generalized spike-wave were seen in six patients (25%), three with frontal, two with temporal and one with occipital lobe lesions. Ictal scalp EEG was reported in five patients, all with MA and seizures confined to the temporal lobe.
Our patients' surface and depth ECoG disclosed a wide array of findings. In three patients with temporal lobe MA (patients 3, 4 and 6), spikes and sharp waves were recorded from the surface and within MA lesions (Figs 3, 4C and 6C). MA and hippocampal spikes occurred independently and differed in morphology. In patient 6, what appeared to be hippocampal spikes and seizures proved to originate from MA which entirely displaced the normal mesial temporal structures. In other patients (patients 1, 2 and 7), MA was electrographically quiescent and the surrounding cortex gave rise to spikes and seizures. In patient 2, parenchymal haemorrhage and destruction could explain the lack of lesional epileptogenicity. Three of our patients had multifocal extralesional spikes; two had generalized spike-wave without secondary bilateral synchrony, one with occipital and one with opercular MA. Mesial temporal ECoGs, obtained in five of our patients, recorded spikes in each instance.
Imaging
Patients with and without NF had similar findings on the various imaging modalities and are reported as a single group. Angiography was described in 24 patients. It was normal in 15 (63%) patients, showed an avascular mass in seven (29%) and suggested an arteriovenous malformation in two (8%) (Rhodes and Davis, 1978; Kasantikul and Brown, 1981).
The commonest finding on CT scan was a calcified, enhancing lesion with surrounding low density, occurring in 21 (55%) of 38 patients.
MRI was reported in 25 cases. Low or mixed central signal on T1- and T2-weighted images and surrounding high signal on T2-weighted sequences were found in 21 (84%) of 25 patients. Gadolinium enhancement occurred in six of seven reported patients. No MRI abnormality was produced directly by MA in four cases (16%), viz. one patient each with meningioma (Blumenthal et al., 1993), temporal encephalocoel (Whiting et al., 1990), low-grade tumour (oligodendroglioma) (Lopez et al., 1996) and normal MRI (Prayson, 1995). Although the white matter appeared to be involved by MA in eight (32%) cases, histopathology showed only secondary changes, e.g. gliosis or haemorrhage into the white matter. MA resembled various diseases on MRI, including meningioma (Partington et al., 1991; Fujimoto et al., 1993; Prayson, 1995), acute haemorrhage (Auer et al., 1982; Goates et al., 1991) (2 cases), calcified arteriovenous malformation (1 case) and encephalomalacia (6 cases). Review of our MRIs suggested MA in only two patients (patients 1 and 4), both with partly calcified, gyriform cortical lesions.
Associated abnormalities
Cranial abnormalities occurred in two reported cases, one with a temporal lobe encephalocoel (Whiting et al., 1990) and one with orbital `erosion' (Louw et al., 1990). MA was associated with adjacent meningioma in four cases (Kasantikul and Brown, 1981; Louw et al., 1990; Wilson et al., 1991; Blumenthal et al., 1993). A definitive vascular malformation adjacent to the area of MA occurred in three reported cases (Kasantikul and Brown, 1981; Halper et al., 1986; Sakaki et al., 1987). MA was diagnosed incidentally in proximity to an oligodendroglioma in one case (Lopez et al., 1996).
Seizure outcome following surgery
Seizure-free rates in our series and in the literature were 43 and 68%, respectively. Improvement in seizures occurred in 30% of patients in our series and in the literature. Seizures did not improve in two (28%) of our patients, compared with only one (5%) in the literature. Some of our unimproved patients had temporal (limbic and neocortical) and opercular seizures. One of our seizure free patients (patient 3) required three operations. Seizures stopped only after resecting sclerosed amygdala and hippocampus. A similar proportion of patients in our series and in the literature continued to require antiepileptic drugs (71 and 79%, respectively). Median follow-up was longer in our patients (6 years) than in the literature (2 years). MA was partly removed in two of our patients. One is seizure-free (patient 1) while refractory seizures persist in the other (patient 2). Information on completeness of lesion removal and seizure outcome was available for 30 patients in the literature. Fifteen of 23 (65%) and four of seven (57%) patients who underwent complete or partial resections became seizure-free, respectively. These proportions were not significantly different (odds ratio = 1.4; 95% confidence interval, 0.25–7.9).
Age, MA location and size, duration of illness and interictal EEG findings did not correlate with seizure outcome. Therefore, our data suggest that seizure outcome after surgery is variable and that resection of the lesion and of epileptogenic cortex may be required.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
MA is diagnosed with a frequency of approximately one every 1.5 years at our centre, where 40–50 patients undergo epilepsy surgery each year. With few exceptions, MA involves cortical grey matter only, although imaging studies, including MRI, may give the impression of white matter involvement. The literature suggests higher occurrence in males and in the right hemisphere. The commonest presentation is progressive difficulty in controlling partial seizures.
MA has histological features of meningiomas and angiomas. Solitary MA lesions measuring up to 5 cm are accompanied by meningeal thickening and calcification, producing tan-yellow, gritty tissue. The histological spectrum can be broadly classified into predominantly cellular and predominantly vascular lesions. Although each lesion is unique, increased cortical vascularity and perivascular cellular proliferation are constant findings. The main histopathological features are leptomeningeal meningothelial proliferation and meningovascular proliferation. The literature includes a few cases in which the second feature was absent. Except for bone formation, our seven cases demonstrate the full range of recognized histological morphologies, i.e. calcification, gliosis, perivascular connective tissue proliferation, dysplastic neurons, white matter cysts and large-vessel hyalinization. In many cases, proliferating perivascular cells infiltrate the cortex in association with marked cellularity and reactive gliosis. Unless the pathologist is familiar with the histological features of MA, these features may lead to the erroneous diagnosis of malignancy, as illustrated by our case 3 and other cases in the literature (Huson et al., 1988). Nuclear pseudoinclusions, typical of meningiomas, are not a major feature of MA. Neurofibrillary tangles, unassociated with amyloid plaques or granulovacuolar degeneration, may be a reactive phenomenon rather than an intrinsic MA component. In the past, histopathological pleomorphism has created diagnostic difficulties. Original diagnoses other than MA were given for seven (14%) and two (15%) cases without and with NF, respectively. Pathological differential diagnoses include malignant meningioma and vascular lesions such as Sturge-Weber syndrome and arteriovenous malformation.
Immunohistochemistry has limited diagnostic value, as staining patterns vary among MA cases. Some immunostaining was done in 24 published cases, although often not in a panel. Results of our immunostaining panel parallel those in the literature, i.e. only vimentin, an intermediate filament protein of fibroblasts and mesenchymal cells, is consistently positive. Epithelial membrane antigen, a marker for arachnoid cap cells, and cam 5.2, co-expressed in 10% of meningiomas, are often negative in MA. GFAP, S-100 and neuronal specific enolase show inconsistent staining and factor VIII was not expressed by the lesional cells (Table 4). These findings do not support a meningothelial origin for the perivascular cells. Instead, it is possible that a pluripotent cell line undergoes differentiation towards various cell types.
|
Results of electron microscopy are sparse and inconsistent in the literature. Some cases suggest a meningothelial derivation, i.e. interdigitating cell membranes, cell junctions and intermediate filaments, while others lack such features (Halper et al., 1986
; Kunishio et al., 1987; Goates et al., 1991; Kim et al., 1993; Kollias et al., 1994; Gomez-Anson et al., 1995). Atypical neuronal inclusions resembling Pick bodies have been recently described in sporadic MA (Mokhtari et al., 1998).
The pathogenesis of MA remains unclear. Proposed hypotheses (Goates et al., 1991; Kollias et al., 1994; Prayson, 1995) suggest that: (i) MA is a hamartoma that undergoes degenerative changes, and association with NF in some cases supports this theory; (ii) MA results from invasion of brain tissue by a leptomeningeal meningioma, though not all cases have a meningeal component and features of malignancy are typically absent; (iii) a cortical vascular malformation induces perivascular meningothelial proliferation of cells from vessel walls or from pluripotent arachnoid cap cells in Virchow–Robin spaces. Leptomeninges and arachnoid cap cells normally surround blood vessels as they penetrate the cortex. Conceivably, chronic leptomeningeal stimulation by the underlying cortical lesion could result in MA histopathological changes.
Presurgical diagnosis remains difficult because diagnostic tools lack specificity. For example, MA does not have a typical CT or MRI appearance. Experienced neuroradiologists at our centre suspected MA from the MRI only in two of the seven cases (cases 1 and 4), based on gyriform cortical patterns with calcification or old haemorrhage. MRI erroneously suggested low-grade tumour, vascular malformations and cystic encephalomalacia in the other five cases. CT and MRI enhancement occurs with sufficient frequency to blur the distinction between MA and other lesions. Furthermore, the electrophysiology of MA is complex. Interictally, it encompasses circumscribed background slowing and epileptogenicity, multifocal extralesional spikes and sharp waves, and generalized spike-waves without definitive secondary bilateral synchrony. Ictal origin may be circumscribed to the region of MA or involve broader areas.
Our ECoG recordings showed intrinsic epileptogenicity of some MA lesions, a previously unrecognized phenomenon. In addition, MA exhibits a range of electrophysiological abnormalities. At one end of the spectrum, they resemble focal, non-dysplastic lesions whose epileptogenicity is produced by adjacent cortex. At the other end, the epileptogenicity of MA is similar to that of cortical dysplasias in which the lesion itself produces spikes, in addition to multifocal, independent epileptiform activity. The variable occurrence of epilepsy in MA remains unexplained, as is the case with other cortical lesions. Although we provide definitive evidence of epileptogenicity of sporadic MA lesions in some of our patients, other factors (e.g. genetic and biochemical) must be at play in determining the expression of epilepsy in MA. Genetic differences exist between sporadic MA and MA with NF-2 (Stemmer-Rachamimov et al., 1997). This suggests that genetic make up may play a role in determining the dissimilar clinical expression of histopathologically identical lesions.
The concept of prompt seizure control following partial or complete MA resection (Kasantikul and Brown, 1981; Halper et al., 1986; Sakaki et al., 1987; Kuzniecky et al., 1988; Liu et al., 1989; Ogilvy et al., 1989; Partington et al., 1991; Tien et al., 1992; Gomez-Anson et al., 1995; Prayson, 1995) requires revision. Seizures persist in a significant proportion of patients despite removal of the lesion and apparent confinement of epileptogenicity to one focus. In our series, 43% of patients became seizure-free compared with 68% in the literature. Systematic outcome assessment and longer follow-up in our cases may account for this discrepancy. Furthermore, antiepileptic drug requirements continue in >70% of patients, several years after MA removal. This corresponds with our findings of complex epileptogenicity, even with solitary MA lesions. No single factor emerged as determinant of seizure outcome following resection of MA lesions.
Cranial abnormalities occurred in two reported patients without NF. In contrast to the osseous malformations seen in NF (sphenoid wing and posterior orbit agenesis and long-bone pseudoarthrosis), all abnormalities in MA overlay the lesions. Although MA and meningiomas may co-exist, their relationship remains unclear. Some suggest that meningiomas may arise from MA (Wilson et al., 1991) while others (Blumenthal et al., 1993; Gomez-Anson et al., 1995) question whether cases of `sclerosing meningioma' represent MA (Davidson and Hope, 1989). Other similarities between MA and meningiomas include an association with NF, multiplicity of lesions, and multifaceted histopathological expression.
Non-sporadic cases are associated with type 2 NF (Stemmer-Rachamimov et al., 1997). Sporadic MA is four times more common than MA with NF. All symptomatic cases occur sporadically and do not exhibit NF-2 gene mutations. Therefore, they probably do not represent formes frustes of NF-2 (Stemmer-Rachamimov et al., 1997). Our analysis supports the preliminary notion that MA associated with NF differs from sporadic MA in that, in the former, lesions tend to be multiple and extracortical and are rarely associated with seizures.
Conclusions
The wide spectrum of EEG and imaging expressions of MA often impedes the clinical diagnosis. Although histopathological diversity is common, MA can be classified into cases with predominantly cellular features and those with predominantly vascular features. Little diagnostic gain accrues from immunostaining because of its variability. Our data do not support a meningeal origin of MA, but suggest that pluripotent cells may differentiate into various cell types found in MA. Sporadic MA commonly presents as refractory localization-related epilepsy, but other clinical presentations are recognized. Although MA occurs infrequently, it is important to establish the correct diagnosis. For example, its histologically benign and non-recurrent nature is prognostically reassuring. On the other hand, extralesional epileptogenesis and variable seizure outcome must be considered when planning surgical treatment. Finally, the association of symptomatic MA with NF is extremely unusual and the two are probably genetically distinct.
|
Acknowledgments |
|---|
We wish to thank Drs W. T. Blume, R. S. McLachlan, J. P. Girvin, A. Parrent and R. Sahjpaul for giving us access to their patients, Dr W. T. Blume for comments on an earlier manuscript, Dr Alberto Díaz for assistance with the literature review and Mrs Dorota Ociepa for preparing the electrophysiological data. We are especially grateful to the patients for their co-operation and timely reply to our enquiries. Part of this work was presented in abstract form at the Canadian Congress of Neurological Sciences in June, 1996.
|
References |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionReferences |
|---|
Aizpuru RN, Quencer RM, Norenberg M, Altman N, Smirniotopoulos J. Meningioangiomatosis: a clinical, radiologic, and histopathologic correlation. Radiology 1991; 179: 819–21.
Auer RN, Budny J, Drake CG, Ball MJ. Frontal lobe perivascular schwannoma. Case report. J Neurosurg 1982; 56: 154–7.[Web of Science][Medline]
Bassoe P, Nuzum F. Report of a case of central and peripheral neurofibromatosis. J Nerv Ment Dis 1915; 42: 785–96.[Web of Science]
Beck E. Ein Beitrag zur Kenntnis der Neurofibromatose (Recklinghausensche Krankheit). Ztschr f d ges Neurol u Psychiat 1938; 162: 426–42.
Beck E. Zwei Falle von Neurofibromatose mit Befallensein des Zentralnervensystems. Ztchr f d Ges Neurol u Psychiat 1939; 164: 748–89
Blumenthal D, Berho M, Bloomfield SS, Schochet SS Jr, Kaufman HH. Childhood meningioma associated with meningioangiomatotis. J Neurosurg 1993; 78: 287–9.[Web of Science][Medline]
Davidson GS, Hope JK. Meningeal tumors of childhood. Cancer 1989; 63: 1205–10.[Web of Science][Medline]
Feigin I. The nerve sheath tumor, solitary and in von Recklinghausen's disease; a unitary mesenchymal concept. Acta Neuropathol (Berl) 1971; 17: 188–200.[Medline]
Foerster O, Gagel O. Ein Fall von Recklinghausenscher Krankheit mit fünf nebeneinander bestehenden verschiedenartigen Tumorbildungen. Ztschr f d ges Neurol u Psychiat 1932; 138: 339–60.
Fujimoto K, Nikaidoh Y, Yuasa T, Nagata K, Ida Y, Fujioka M, et al. Meningioangiomatosis not associated with von Recklinghausen's disease—case report. Neurol Med Chir (Tokyo) 1993; 33: 651–5.[Medline]
Garen PD, Powers JM, King JS, Perot PL Jr. Intracranial fibro-osseous lesion. J Neurosurg 1989; 70: 475–7.[Web of Science][Medline]
Goates JJ, Dickson DW, Horoupian DS. Meningioangiomatosis: an immunocytochemical study. Acta Neuropathol (Berl) 1991; 82: 527–32.[Medline]
Gomez-Anson B, Munoz A, Blasco A, Madero S, Esparza J, Cordobes F, et al. Meningioangiomatosis: advanced imaging and pathological study of two cases. Neuroradiology 1995; 37: 120–3.[Web of Science][Medline]
Halper J, Scheithauer BW, Okazaki H, Laws ER Jr. Meningio-angiomatosis: a report of six cases with special reference to the occurrence of neurofibrillary tangles. J Neuropathol Exp Neurol 1986; 45: 426–46.[Web of Science][Medline]
Harada K, Inagawa T, Nagasako R. A case of meningioangiomatosis without von Recklinghausen's disease. Report of a case and review of 13 cases. [Review]. Childs Nerv Syst 1994; 10: 126–30.[Web of Science][Medline]
Hozay J. Une angioneuromatose meningo-encephalique diffuse. Rev Neurol (Paris) 1953; 89: 222–36.[Medline]
Huson SM, Harper PS, Compston DA. Von Recklinghausen neurofibromatosis. A clinical and population study in south-east Wales. Brain 1988; 111: 1355–81.
Johnson PC, Nielsen SL. Localized neurofibrillary degeneration in vascular malformations [Abstract]. J Neuropathol Exp Neurol 1976; 35: 300.
Jun C, Burdick B. An unusual fibro-osseous lesion of the brain. J Neurosurg 1984; 60: 1308–11.[Web of Science][Medline]
Kasantikul V, Brown WJ. Meningioangiomatosis in the absence of von Recklinghausen's disease. Surg Neurol 1981; 15: 71–75.[Web of Science][Medline]
Kim YW, Choi WS, Lee J, Yang MH. Meningioangiomatosis—a case report. J Korean Med Sci 1993; 8: 308–11.[Medline]
Kollias SS, Crone KR, Ball WS Jr, Prenger EC, Ballard ET. Meningioangiomatosis of the brain stem. Case report. J Neurosurg 1994; 80: 732–5.[Web of Science][Medline]
Kunishio K, Yamamoto Y, Sunami N, Satoh T, Asari S, Yoshino T, et al. Histopathologic investigation of a case of meningioangiomatosis not associated with von Recklinghausen's disease. Surg Neurol 1987; 27: 575–9.[Web of Science][Medline]
Kuzniecky R, Melanson D, Robitaille Y, Olivier A. Magnetic resonance imaging of meningio-angiomatosis. [Review]. Can J Neurol Sci 1988; 15: 161–4.[Web of Science][Medline]
Liu SS, Johnson PC, Sonntag VK. Meningioangiomatosis: a case report. Surg Neurol 1989; 31: 376–80.[Web of Science][Medline]
Lopez JI, Ereno C, Oleaga L, Areitio E. Meningioangiomatosis and oligodendroglioma in a 15-year-old boy. [Review]. Arch Pathol Lab Med 1996; 120: 587–90.[Web of Science][Medline]
Louw D, Sutherland G, Halliday W, Kaufmann J. Meningiomas mimicking cerebral schwannoma. J Neurosurg 1990; 73: 715–9.[Web of Science][Medline]
Matias-Guiu X, Molet J, Ferrer I, Prat J. Meningio-angiomatosis: a case report. Br J Neurosurg 1988; 2: 97–100.[Medline]
Mokhtari K, Uchihara T, Clemenceau S, Baulac M, Duyckaerts C, Hauw JJ. Atypical neuronal inclusion bodies in meningioangiomatosis. Acta Neuropathol (Berl) 1998; 96: 91–6.[Medline]
Ogilvy CS, Chapman PH, Gray M, de la Monte SM. Meningioangiomatosis in a patient without von Recklinghausen's disease. J Neurosurg 1989; 70: 483–5.[Web of Science][Medline]
Oka Y, Sakaki S, Yamashita M, Nakagawa K, Matsuoka K, Matsui H. A case of meningioangiomatosis in an infant. [Review]. [Japanese]. No Shinkei Geka 1991; 19: 761–5.[Medline]
Partington CR, Graves VB, Hegstrand LR. Meningioangiomatosis. [Review]. ANJR Am J Neuroradiol 1991; 12: 549–52.[Web of Science][Medline]
Paulus W, Peiffer J, Roggendorf W, Schuppan D. Meningio-angiomatosis. [Review]. Pathol Res Pract 1989; 184: 446–54.[Web of Science][Medline]
Prayson RA. Meningioangiomatosis. A clinicopathologic study including MIB1 immunoreactivity. [Review]. Arch Pathol Lab Med 1995; 119: 1061–4.[Web of Science][Medline]
Rhodes RH, Davis RL. An unusual fibro-osseous component in intracranial lesions. Hum Pathol 1978; 9: 309–19.[Web of Science][Medline]
Rubinstein LJ. Tumeurs et hamartomes dans la neurofibromatose centrale. In: Michaux L, Feld M, editors. Les phakomatoses cerebrales. Paris: S.P.E.I.; 1963. p. 426–51.
Sakaki S, Nakagawa K, Nakamura K, Takeda S. Meningioangiomatosis not associated with von Recklinghausen's disease. Neurosurgery 1987; 20: 797–801.[Web of Science][Medline]
Stemmer-Rachamimov AO, Horgan MA, Taratuto AL, Munoz DG, Smith TW, Frosch MP, et al. Meningioangiomatosis is associated with neurofibromatosis 2 but not with somatic alterations of the NF2 gene. [Review]. J Neuropathol Exp Neurol 1997; 56: 485–9.[Web of Science][Medline]
Tien RD, Osumi A, Oakes JW, Madden JF, Burger PC. Meningioangiomatosis: CT and MR findings. J Comput Assist Tomogr 1992; 16: 361–5.[Web of Science][Medline]
Whiting DM, Awad IA, Miles J, Chou SS, Luders H. Intractable complex partial seizures associated with occult temporal lobe encephalocele and meningioangiomatosis: a case report. [Review]. Surg Neurol 1990; 34: 318–22.[Web of Science][Medline]
Willson N, Kaufman MA, Bodansky SM. An unusual intracerebral connective tissue mass. J Neuropathol Exp Neurol 1977; 36: 373–8.[Web of Science][Medline]
Wilson D, Dempsey RJ, Clark DB. Meningioma developing from underlying meningioangiomatosis [Abstract]. J Neuropathol Exp Neurol 1991; 50: 371.
Worster-Drought C, Dickson WEC, McMenemey WH. Multiple meningeal and perineural tumours with analogous changes in the glia and ependyma (neurofibroblastomatosis). Brain 1937; 60: 85–117.
Received August 26, 1998. Revised November 3, 1998. Accepted November 23, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M FEDI, R M KALNINS, N SHUEY, G J FITT, M NEWTON, and L A MITCHELL Cystic meningioangiomatosis in neurofibromatosis type 2: an MRI-pathological study Br. J. Radiol., July 1, 2009; 82(979): e129 - e132. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||