Historical Background

A meningioma is, in many ways, the soul of neurosurgery. The progress in meningioma treatment mirrors advances in neurosurgery, and advancements in neurosurgery are put to maximal use to improve the treatment of meningiomas.¹ A historical account of meningiomas and their surgical management highlights that “meningiomas have left their mark, in the form of hyperostosis, on human skulls as far back as prehistoric times,”² and authors began to document the recognition of these tumors as pathologic entities starting in the 17th century.

In the 18th and 19th centuries meningiomas were diagnosed during a patient’s life only if they caused changes in the overlying skull that could be appreciated through inspection or palpation. Only a few attempts were made to remove these lesions surgically, and few were beneficial to the patient. Of 13 such operations performed between 1743 and 1896 whose outcome was specified, 9 ended in the patient’s death.²

In 1864 John Cleland, professor of anatomy in Glasgow, “set forth with prescience … the view that two tumours which he had found in the dissecting room, one of them arising from the cribriform plate and the other from the right frontal region adjacent to the superior longitudinal sinus, took their origin from the arachnoid rather than the dura. He observed that in structure they resembled the Pacchionian granulations in a number of points.”³ In 1915 Cushing and Weed⁴ reasserted Cleland’s opinion that meningiomas derived from arachnoid cell clusters.

The sequential nomenclature of meningiomas has included fungoid tumor, sarcoma, cylindroma, endothelioma, and fibroma.^2,5 Cushing proposed the term meningothelioma in an effort to describe these tumors according to the tissue involved. He avoided a histiogenic name because the cellular composition of this tumor was still in dispute; he also avoided a place name because of the various sites of distribution of meningiomas. Later, Cushing opted for the term meningioma. In his Cavendish lecture in 1922, he reported on 85 cases of meningioma.³ In 1938 Cushing and Eisenhardt⁶ published Meningiomas: Their Classification, Regional Behaviour, Life History, and Surgical End Results, in which they reported in detail the cases of 313 patients encountered between 1903 and 1932. The interest in meningiomas has not declined since then. In 1922 Cushing wrote: “There is to day nothing in the whole realm of surgery more gratifying than the successful removal of a meningioma with subsequent perfect functional recovery.”³ These words are still true more than 80 years later.

Pathology

Meningioma’s cell of origin is believed to be the arachnoid cap cell. The arachnoid villi protrude into the venous sinuses. The venous endothelium is in contact with all or a portion of the arachnoid villi. In the latter case, these cells are referred to as arachnoid cap cells. The rest of the granulation is covered by a fibrous capsule. Arachnoid villi are most numerous in the area of the superior sagittal sinus, followed by the cavernous sinus, tuberculum sellae, lamina cribrosa, foramen magnum, and torcular Herophili. Arachnoid granulations and pacchionian bodies are larger and more pronounced versions of arachnoid villi (Fig. 131-1).

FIGURE 131-1 Semidiagrammatic representation of the arachnoid granulation shows its relation to the superior sagittal sinus (venous sinus) and the continuity of its layers and spaces with those of the meninges on the surface of the brain. The individual layers or cell clusters are not intended to be proportional but are drawn to emphasize the cytoarchitectural features of the granulation. SAS, subarachnoid space.

(From Haines DE, Frederickson RG. The meninges. In: Al-Mefty O, ed. Meningiomas. New York: Raven Press; 1991:9.)

The 2000 World Health Organization (WHO) classification of tumors of the nervous system lists meningiomas under the heading “tumours of the meninges” and the subheading “tumours of meningothelial cells.”⁷ WHO recognizes three grades based on pathologic criteria and the risk of recurrence and aggressive growth (Table 131-1).

TABLE 131-1 Meningioma Grades

Adapted from Louis DN, Scheithauer BW, Budka H, et al. Meningiomas. In: Kleihues P, Cavenee WK, eds. Pathology and Genetics of Tumours of the Nervous System. Lyon, France: IARC Press; 2000:176-180.

Meningiomas are usually globular, encapsulated tumors. They are attached to the dura and compress the underlying brain without invading it. Even though invasion of the dura and dural sinuses is common, meningiomas are usually easily separated from the pia mater. The cleavage plane, however, may not encompass the whole surface of the tumor.⁸ Meningiomas may also occur as a flattened sheath of tumor, taking the shape of the underlying bone. This so-called meningioma en plaque is more common in the area of the sphenoid bone. Seldom, meningiomas occur in a location where dural attachment cannot be shown (e.g., intraventricular, intraparenchymal). The cut surface of a meningioma is pale and translucent or homogeneous and reddish brown, depending on the degree of vascularity. A gritty consistency is common. After fixation of a specimen, a whorled pattern may be apparent on the cut surface. Intratumoral hemorrhage is rare, and necrosis is generally absent.

The distribution of intracranial meningiomas is approximately as follows: convexity (35%), parasagittal (20%), sphenoid ridge (20%), intraventricular (5%), tuberculum sellae (3%), infratentorial (13%), and others (4%). In this chapter we briefly mention the characteristics of the meningothelial (syncytial), transitional, and fibroblastic varieties. These meningiomas mimic the appearance of normal arachnoid villi in their ultrastructural features seen on electron microscopy: prominent interdigitation of the plasma membrane, abundant cytoplasmic intermediate filaments (10 nm) immunocytochemically consistent with vimentin, frequent hemidesmosomes to which the intermediate filaments anchor, and focal intercellular deposits of electron-dense granular material. Arachnoid and meningioma cells are connected by epithelial cadherins (E-cadherins), which are Ca²⁺-dependent adhesion molecules, and both express glutathione-independent prostaglandin D₂ synthase.⁹

Histologically, meningothelial (syncytial) meningiomas are characterized by densely packed cells arranged in sheets with no clearly discernible cytoplasmic borders (Fig. 131-2). Microscopically, they mimic normal arachnoid cells. Whorls can be found but are not prominent. Mineralized whorls, containing calcium apatite and collagen, are called psammoma bodies (from the Greek word psammos, meaning “sand”). Distinctive features of meningiomas include intranuclear cytoplasmic pseudoinclusions, in which an invaginated cytoplasmic remnant occupies the interior of the nucleus and displaces the nuclear chromatin. Another useful feature is the presence of so-called Orphan Annie’s eye nuclei—target-like nuclei with central clearing and peripheral margination of the chromatin.

FIGURE 131-2 Hematoxylin-eosin staining shows the histologic pattern of meningothelial meningiomas and polygonal cells with ill-defined cytoplasm.

Microscopically, fibroblastic (fibrous) meningiomas reveal multilaminated sheets of interlacing, elongated spindle cells. The intervening stroma is composed of reticulin fibers and collagen (Fig. 131-3). Transitional meningiomas represent a combination of the meningothelial and fibroblastic types. Characteristically, cellular whorls are seen, separated by elongated spindle cells (Fig. 131-4). Variations in meningioma histology may reflect mutations at separate genetic loci, in that the loss of heterozygosity on chromosome 22 is much more common in fibroblastic than in meningothelial variants.¹⁰

FIGURE 131-3 Hematoxylin-eosin staining shows a fibroblastic meningioma formed of sheets of elongated meningothelial cells.

FIGURE 131-4 Hematoxylin-eosin staining shows a transitional meningioma with cellular whorls and psammoma bodies.

Many other variants of meningioma have been reported, including psammomatous, angiomatous, microcystic (humid), xanthomatous, lipoblastic, myxoid (myxomatous), osteoblastic, chondroblastic, secretory,¹¹ melanotic, lymphofollicular, chordoid,¹² hemangiopericytic, oncocytic,¹³ and papillary meningiomas. Not all these terms for variants are currently in use.

It is beyond the scope of this chapter to discuss all these variants; however, it is important to briefly discuss the so-called hemangiopericytic variety. Sometimes meningiomas are composed partly or totally of small cells that grow focally in a hemangiopericytic pattern. The biologic behavior of this variant has not been well characterized. It is important to differentiate these so-called hemangiopericytic meningiomas (of meningothelial origin) from true hemangiopericytomas, which are mesenchymal tumors of nonmeningothelial origin. True hemangiopericytomas of the meninges are similar to those occurring in other parts of the body. Their behavior is characterized by early recurrence and a tendency to metastasize.

Atypical meningioma is associated with a higher rate of recurrence and aggressive growth. The criteria used to diagnose atypical meningioma are independent of meningioma subtype. Atypical meningioma exhibits increased mitotic activity or three or more of the following features: increased cellularity, small cells with a high nucleus-to-cytoplasm ratio, prominent nucleoli, uninterrupted patternless or sheetlike growth, and foci of spontaneous or geographic necrosis. For this variant, increased mitotic activity has been defined as four or more mitoses per 10 high-power fields (Fig. 131-5).⁷

FIGURE 131-5 Hematoxylin-eosin staining shows an atypical meningioma. A mitotic figure is seen in the middle.

The exact definition of anaplastic and malignant meningiomas is still open to discussion.¹⁴ One feature undoubtedly labels a meningioma as malignant: distant extraneural metastasis.¹⁵ The most common sites for metastasis are the liver, lungs, pleura, and lymph nodes. Frank parenchymal invasion of the underlying brain also carries an ominous prognosis. Anaplastic meningioma is a meningioma exhibiting histologic features of frank malignancy far in excess of the abnormalities present in atypical meningiomas. Such features include obviously malignant cytology (e.g., an appearance similar to sarcoma, carcinoma, or melanoma) or a high mitotic index (≥20 mitoses/10 high-power fields) (Fig. 131-6).

FIGURE 131-6 Hematoxylin-eosin staining shows a meningioma with obvious malignant cytology and a very high mitotic index.

Immunohistochemistry helps in the diagnosis of meningiomas. The test for epithelial membrane antigen (EMA) is positive in 80% of meningiomas. The results of S-100 staining are quite variable. Meningiomas also express markers for fibroblasts (vimentin) and epithelial cells (EMA and cytokeratins). Antileu 7, an antibody positive in schwannomas, is uniformly negative in meningiomas. Even though results of glial fibrillary acidic protein (GFAP) stains are negative in meningiomas, a few cases of GFAP-positive meningiomas have been reported in the world literature.¹⁶ Syncytial and transitional meningiomas express E-cadherin.¹⁷ Another use of immunohistochemistry lies in differentiating atypical meningiomas from similar but pathologically distinct tumors, such as secretory meningioma from metastatic carcinoma.¹¹

The nonhistologic diagnosis of meningiomas is being developed. In this technique, biopsy specimens are treated with perchloric acid and analyzed with high-resolution ¹H magnetic resonance spectroscopy and automatic amino acid analysis with ionic exchange chromatography. This method can accurately differentiate between meningiomas and other tumors involving the brain (e.g., high- and low-grade astrocytoma, medulloblastoma, metastasis, neurinoma, and oligodendroglioma).¹⁸

Hyperostosis is a characteristic finding in meningiomas, especially in en plaque meningiomas. In most cases, histologic studies of hyperostotic bone reveal tumor cells in the diploë and haversian canals (Fig. 131-7).¹⁹

FIGURE 131-7 Meningioma cells are invading a portion of hyperostotic bone.

Al-Mefty and coworkers²⁰ documented that meningiomas also go through tumor progression, which is defined by irreversible changes in tumor characteristics reflecting the sequential appearance of a genetically altered subpopulation of cells with the new characteristics.²¹ In some instances, the properties of advanced malignany may be established before the neoplasm reaches macroscopic size; in other cases, well-differentiated and slow-growing tumors may persist for years before shifting to more aggressive behavior. The progression of a tumor is explained by the model of clonal evolution: tumor development is initiated by a single cell carrying a mutation (the mutation model) that gives it a select growth advantage. The distinction between malignant transformation and de novo malignant tumor was addressed with a recent study demonstrating differences in clinical behavior, hormone receptor status, proliferative indices, and cytogenetic profile between these two subgroups of meningioma.²²

Extensive efforts have been made to determine the clinical behavior and malignant potential of meningiomas. These efforts have focused on different aspects of meningioma pathology: histology, labeling, karyotype and genetics, radiology, and hormone receptors. Some histologic features portend aggressive behavior. These features include hypercellularity, loss of architecture, nuclear pleomorphism, increased mitotic index, focal necrosis, hypervascularity, hemosiderin deposition, and small cell formation.¹⁴ Labeling techniques have been devised to quantify the mitotic rate of meningiomas and therefore predict their behavior. Bromodeoxyuridine must be injected intravenously shortly before tumor removal, and the surgical specimen must be fixed in 70% ethanol before being embedded in paraffin. Bromodeoxyuridine labeling allows the examiner to determine the percentage of cells in the S phase of mitosis. An alternative is immunohistochemical staining for proliferating cell nuclear antigen (Fig. 131-8). Fresh specimens can also be labeled for the monoclonal antibody Ki-67²³ or MIB-1, which can be done on any paraffin-embedded specimen. A high labeling index indicated by either method denotes a more aggressive tumor.^24,25 However, an elevated Ki-67 labeling index can paradoxically be seen in tumors that were previously irradiated.²⁶ Immunoreactivity for transforming growth factor-α has also been correlated with an increased level of malignant behavior.²⁴

FIGURE 131-8 Immunohistochemical staining of a meningioma for proliferating cell nuclear antigen. Note the high labeling index, which portends aggressive behavior.

Schwechheimer and coworkers²⁷ studied the expression of E-cadherin in a variety of brain tumors. Only choroid plexus papillomas (five of five) and meningiomas showed E-cadherin expression. In benign meningiomas, positive E-cadherin immunoreactivity was found regardless of the histomorphologic subtype or the tumor’s invasion into dura, bone, brain, or muscle. In contrast, E-cadherin was absent from most morphologically malignant meningiomas. In recurrent meningiomas, E-cadherin expression was identical to that in the primary neoplasm except in cases of malignant progression, in which the malignant recurrent tumor was negative for E-cadherin. In two cases of metastasizing meningiomas, no E-cadherin immunoreactivity was found in the primary tumors or in their metastases.²⁷

SPARC, a secreted, extracellular matrix–associated protein implicated in the modulation of cell adhesion and migration, was investigated as an aid to assess the invasiveness of meningiomas. It was not expressed in 9 benign meningiomas and highly expressed in 20 invasive tumors, regardless of grade. SPARC can be used to assess the invasiveness of histologically benign meningiomas in the absence of a tumor-brain interface.²⁸ Kitange and colleagues²⁹ studied the immunohistochemical expression of Ets-1 transcription factor and the urokinase-type plasminogen activator (u-PA) in meningiomas. The Ets-1 transcription factor is thought to play an important role in the invasive process of tumor cells through the induction of u-PA. A significant difference was observed between Ets-1 and u-PA expression in benign and high-grade meningiomas.²⁹ Complex karyotypes increase progressively from benign (34%) to atypical (45%) to anaplastic (70%) meningiomas.³⁰ The absence of progesterone receptors also correlates with poor outcome.³¹

Multiple meningiomas are defined as two or more meningiomas appearing simultaneously or sequentially in the same patient.³² The reporting of multiple meningiomas has increased with the advent of new imaging techniques such as computed tomography (CT) and magnetic resonance imaging (MRI). The incidence of multiple meningiomas ranges from 1% to 16% of all meningiomas. Between 60% and 90% of patients with multiple meningiomas are women. Multiple meningiomas occur in association with neurofibromatosis type 2. They have also been described in families with no evidence of neurofibromatosis.³³ Multiple meningiomas may be secondary to a recurrence at the edge of the surgical resection or to postoperative seeding through the cerebrospinal fluid. This possibility is supported by the fact that recurrent meningiomas were found to be clonal with respect to the primary tumors.³⁴ Intraosseous meningiomas and extraneuraxial meningiomas are uncommon. All the reported intraosseous meningiomas have been in cranial bones. Extraneuraxial meningiomas can involve the orbit, paranasal sinuses, and nasopharynx. Sixteen percent of reported primary extraneuraxial meningiomas occurred in the skin and subcutis; others have been reported in the lungs,^35,36 mediastinum, and adrenal gland.

Tumors of the central nervous sysytem may metastasize to a primary intracranial tumor. Three fourths of these metastases target meningiomas, even though meningiomas represent only 20% of intracranial tumors. There are probably many reasons for this propensity, including the fact that patients with meningiomas, which are slow-growing tumors, are at greater risk for metastasis than patients with other brain tumors. Other factors may be the increased vascularity of meningiomas and the peculiar microenvironment of these tumors. Abscesses may also form within meningiomas.^37,38

Epidemiology

When discussing the epidemiology of meningiomas, a distinction must be made between studies dealing with only a limited population (hospital based) and those dealing with the population at large. A distinction must also be made between series including only operated cases and those including autopsies, and between series including metastases and those dealing exclusively with primary intracranial tumors. In their 1938 review Cushing and Eisenhardt⁶ found that meningiomas constituted 13.4% of Cushing’s series of intracranial tumors. In a population-based clinical study performed in Manitoba from 1980 through 1985, 22% of primary intracranial tumors were meningiomas.³⁹ The occurrence of meningiomas in the general population varies from 2.3 cases per 100,000 people during their life span to 5.5 per 100,000 if autopsy data are included.^39,40 In the Manitoba study, the incidence rates for meningioma were 1.5 and 3.1 per 100,000 for males and females, respectively.⁴⁰

A study performed in Rochester, Minnesota, covering the years 1935 through 1977 revealed the following distribution of primary brain tumors: gliomas 35%, meningiomas 40%, and pituitary adenomas 13% when data from postmortem examinations were included; gliomas 43%, meningiomas 21%, and pituitary adenomas 17% when postmortem data were excluded.³⁷ These data emphasize the fact that many meningiomas remain clinically silent and do not require surgery. Staneczek and Jänisch⁴¹ reported on 8119 new cases of meningioma diagnosed in the former German Democratic Republic between 1961 and 1986. Approximately one half of all these tumors were first discovered at autopsy. The crude annual incidence of meningiomas in this study was 1.85 per 100,000 in the general population. In 1985 Walker and coworkers⁴² reported on the epidemiology of brain tumors in the United States by surveying 166 hospitals. Among 13,720 patients with pathologically confirmed primary intracranial tumors seen during 1973 and 1974, 58% had gliomas, 20% had meningiomas, and 14% had pituitary adenomas.

The incidence of meningiomas increases with age. A slight dip after the eighth decade may be explained by several factors, including a less aggressive surgical approach in the elderly. Another factor is the failure to realize that even though the total number of cases in the ninth decade is less than in earlier decades, the incidence of meningiomas in this age group is actually higher. A meningioma may be found incidentally, and the surgeon must then weigh the risks and benefits of surgery against the natural history of the disease.^43,44 As Rengachary and Suskind⁴⁵ have written, “Some die from meningiomas, others with them. A neurosurgeon’s role is to recognize these two sets of populations and give the benefit of surgery to those who need it and spare those who do not.”

Most studies show a predominance of meningiomas in women. The female-to-male ratio is about 2 : 1. In the Manitoba study, the female predominance was noticeable in patients only after the fifth decade.³⁹ This pattern reflects meningnioma’s lack of predominance in female children.

Meningiomas constitute 1% to 4% of all brain tumors in children (younger than 18 years). They are extremely uncommon in infancy. The average age at presentation is 11.6 years, compared with 6.3 years for all other pediatric brain tumors. Several features distinguish meningiomas in children from their adult counterparts. There is an equal incidence of the tumor between girls and boys, but a male predominance (71%) has been reported among infants. Meningiomas in children may arise in unusual sites. Eleven percent of meningiomas in children are intraventricular, compared with 3.9% in adults. They may be multiple (23%) or have a cystic component (23%). They are often associated with neurofibromatosis (23% to 41%)⁴⁶ and have no dural attachments in up to 13% of cases.⁴⁷

Etiology

Irradiation

In 1953 Mann and colleagues⁵⁴ were the first to report a radiation-induced meningioma. The patient, a 6-year old girl, received 6500 rad after resection of an optic nerve glioma. A meningioma was diagnosed 4 years later within the radiotherapy field. There is no doubt that radiation injury is a factor in the development of meningiomas. In 1909 Adamson described a protocol for irradiation of the scalp to treat tinea capitis (ringworm). The method, referred to as the Kienböck-Adamson technique, delivers 450 to 850 rad to the scalp and between 70 and 175 rad to the surface of the brain. The protocol was widely used from 1900 until 1960, when griseofulvin was introduced. Modan and coworkers⁵⁵ carried out a statistical analysis of 11,000 children and found that meningiomas were four times more common in irradiated patients than in the control group. Although the mean age at diagnosis in the general population is 58 years, it is 45 years in the low-dose radiation group and 31 years in the high-dose radiation group. Even though the female predominance of intracranial meningiomas in the general population is less apparent (and may even be reversed) in the irradiated group, this pattern may be caused by a bias inherent in the population of patients irradiated for tinea capitis. Radiation-induced meningiomas have also been reported in patients who underwent high-dose radiotherapy (Fig. 131-9).⁵⁶ The age of patients at presentation with meningioma and the latency period of radiation-induced meningiomas are dose related. These tumors are more aggressive and are certain to recur, have a higher histopathologic grade, and are associated with complex cytogenetic aberrations, particulary involving 1p and 6q.⁵⁷

FIGURE 131-9 MRI of a radiation-induced meningioma in the left cavernous sinus after 20 years of radiation therapy for craniopharyngioma.

Several studies are being carried out to determine whether the use of mobile phones increases the risk of developing brain neoplasms. A study published in 2001 failed to find such a correlation, but the matter is not settled.^58,59

Genetic Aspects

The genetic aspects of meningiomas have been studied intensively during the past decade, but the final chapter of this research remains to be written. The detailed genetics of meningiomas are beyond the scope of this chapter, but it is possible to succinctly summarize the findings that most researchers agree on. Genetic alterations in the long arm of chromosome 22 play an essential role in the development of meningiomas. Monosomy of chromosome 22 has been observed in up to 50% of patients with meningiomas (Fig. 131-10). The meningioma chromosomal region has been localized to the center of the long arm of chromosome 22 in bands 22q12.3-qter. This is the same area that harbors the neurofibromatosis type 2 (NF2) tumor suppressor gene. This region also harbors SIS, the platelet-derived growth factor (PDGF)-β locus that is homologous to the SIS oncogene. Meningiomas may result from mutation in a tumor suppressor gene or from the SIS oncogene, both located on the long arm of chromosome 22. Whether the tumor suppressor gene responsible for meningiomas is the same NF2 tumor suppressor gene or distinct from it is still subject to debate. Evidence shows, however, that other gene alterations on chromosome 22 may give rise to meningiomas. With the addition of the rhabdoid variety of meningiomas to the WHO classification, it is important to note that a gene involved in the pathogenesis of rhabdoid tumors (the INI1 [SMARCB1/hSNF5] gene) is also mutated in a subset of meningiomas. The INI1 gene may be a second tumor suppressor gene on chromosome 22 important for the genesis of meningiomas.⁶⁴

FIGURE 131-10 Spectral karyotyping demonstrates monosomy of chromosome 22 in a meningioma.

Other chromosomal abnormalities have been detected in meningiomas. Loss of heterozygosity for loci on chromosome arm 1p is relatively common in meningiomas. However, different genes are involved in different tumors, raising the possibility of several tumor suppressor genes on 1p, the inactivation of which may be important in the pathogenesis of meningiomas.⁶⁵ Reviewing cases of meningiomas after radiation treatment, Shoshan and associates⁶⁶ found that 22q deletions are far less frequent in these tumors and that other chromosomal lesions, especially the loss of 1p, possibly induced by irradiation, may be more important in the development of these tumors. Muller and colleagues⁶⁷ found a net progression of chromosome 1 abnormalities in meningiomas according to their pathologic grade; 27% of the common type, 70% of atypical, and 100% of anaplastic meningiomas had a deletion of 1p36.

Lamszus and coworkers⁶⁸ studied gene alterations in five aggressively recurring meningiomas and four malignant nonmeningothelial meningeal tumors (three undifferentiated meningeal sarcomas and one hemangiopericytoma). They stated that “a total of 40 specimens from primary tumors and multiple recurrences of the nine patients were analyzed. … Loss of hetero-zygosity (LOH) at 22q was observed in all meningioma cases at the earliest time point. … While allelic loss at 22q appears to be an early event in aggressive meningioma disease, there is a clear correlation of further deletions on chromosome arms 1p, 9q, 10q, and 14q with histopathological and clinical progression” (Fig. 131-11).⁶⁸ None of these genetic findings was present in the nonmeningiomatous meningeal tumors, indicating that “meningothelial cells have their own lineage-specific genetic pathways towards clinical malignancy.”⁶⁸ Menon and associates⁶⁹ studied 58 meningiomas and found that a loss of chromosome 22 was seen in about half of all these tumors, regardless of their malignancy; however, the most frequent chromosomal losses observed in the malignant and atypical tumors were on the long arm of chromosome 14. Common secondary aberrations include losses or deletions of chromosomes 1p, 14q, and 10q and unstable chromosome aberrations, including rings, dicentrics, and telomeric associations. Despite the analysis of several hundred tumors with cytogenetic and molecular techniques, the mechanisms involved in the progression of chromosome aberrations in meningiomas are poorly understood. Sawyer and colleagues⁷⁰ concluded from their series that the progression of chromosome aberrations in meningiomas is mediated in some respects by telomeric and centromeric instability.

FIGURE 131-11 Representative Giemsa-band karyotype of a malignant meningioma. The most common aberration observed in meningiomas is monosomy for chromosome 22. Loss of chromosome 14, structural aberrations resulting in the loss of 1p, and deletion of 6q are found in higher grade tumors.

The role of the NF2 gene in the genesis of meningiomas deserves further analysis. The NF2 gene codes for a protein called merlin (moesin-ezrin-radixin-like protein). Merlin, also called schwannomin, has striking similarity to several proteins involved in the linkage of cytoskeletal components and proteins in the cell membrane. The alterations in this gene were found in four primary cultures of meningiomas (three of the four patients had a positive family history of neurofibromatosis).⁷¹ A low level of merlin seems to be associated with the development of a subset of meningiomas, usually caused by an alteration in the NF2 gene. Kimura and coworkers⁷² advanced an alternative explanation of the role of merlin in the pathogenesis of meningiomas. They found that even though the loss of merlin leads to the development of meningioma, low levels of this protein may result from a mutation in the NF2 gene or to a high turnover of merlin. The latter mechanism may be mediated by calpain, a calcium-dependent neutral cysteine protease, which leads to the degradation of merlin in these tumors.⁷² As its name implies, merlin is a member of the protein 4.1 family of membrane-associated proteins, which also includes ezrin, radixin, and moesin. Another protein 4.1 gene, DAL1, located on chromosome 18p11.3, was identified by Gutmann and associates.⁷³ These investigators showed a loss of DAL1 in 60% of sporadic meningiomas. They concluded that “analogous to merlin, we show that DAL-1 loss is an early event in meningioma tumorigenesis, suggesting that these two protein 4.1 family members are critical growth regulators in the pathogenesis of meningiomas.”⁷³ The same investigators demonstrated that merlin expression was also lost in some schwannomas and ependymomas.⁷⁴

The clonality of multiple and recurrent meningiomas has been studied. Several studies point to the possibility that recurrent and, more surprisingly, multiple concurrent meningiomas may be monoclonal in origin.^34,75

Meningiomas and Receptors

Meningiomas may become symptomatic during pregnancy, with the symptoms abating after parturition only to reappear with the next pregnancy. Symptoms may also be exacerbated during the proliferative phase of the menstrual cycle. It is unclear whether these exacerbations result from vascular engorgement or hormonal changes. In 1979 Donnell and colleagues⁷⁶ were the first to describe the role of estrogen receptors (ERs) in meningiomas. Since then, progesterone receptors (PRs) have been identified much more consistently than ERs in meningiomas. PRs are generally thought to be present in the cytoplasm of meningiomas, but they rarely occur in the nucleus. Maxwell and coworkers⁷⁷ were unable to detect ER messenger RNA (mRNA) but found PR mRNA in 88% and androgen receptor mRNA in 66% of the meningiomas they tested. Other investigators were able to show the presence of ER-β and ER-α mRNA in meningiomas.⁷⁸ Luteinizing hormone-releasing hormone was found to increase the proliferation of meningioma cells in vitro.⁷⁹ The preponderance of PRs and the scarcity of ERs in meningiomas are well known. The expression of PR alone in meningioma signals a favorable clinical and biologic outcome. A lack of receptors or the presence of ERs in meningiomas correlates with an accumulation of qualitative and quantitative karyotype abnormalities, a higher proportional involvement of chromosomes 14 and 22 in de novo tumors, and an increasing potential for aggressive clinical behavior, progression, and recurrence of these lesions. Sex hormone receptor status should routinely be studied for its prognostic value, especially in female patients, and should be taken into account in tumor grading. The initial receptor status of a tumor may change with progression or recurrence of tumor.⁸⁰

Meningiomas exhibit a very high density of somatostatin receptors. Preferential immunoreactive staining for the sst2A subtype somatostatin receptor has been shown in meningiomas.⁸¹

Androgen and glucocorticoid receptors have been found in meningiomas. The dopamine D₁ (but not D₂) receptor has also been demonstrated in meningiomas, and there are some indications that dopamine may play a role in the proliferation of these tumors. Meningiomas are also positive for prostaglandins, most notably prostaglandin E₂.⁸² Several growth factors have been shown to stimulate meningiomas, including epidermal growth factor, fibroblast growth factor, and PDGF.

Some meningiomas are associated with high systemic levels of substances such as carcinoembryonic antigen⁸³ or prolactin. In one study, prolactin receptors were detected in 61.7% of meningiomas, whereas no prolactin binding was found in samples of normal arachnoid tissue.⁸⁴ Meningiomas may interfere with glucose metabolism by increasing insulin levels. These disturbances of the endocrine system may be explained by mechanical pressure on a regulating intracranial structure or by the secretion of specific substances that interfere with hormonal homeostasis.

Radiology

Plain radiographs reveal three characteristic findings in meningioma patients: hyperostosis, increased vascular markings, and calcification. On non–contrast-enhanced CT, meningiomas are typically isodense to slightly hyperdense compared with contiguous brain parenchyma. Calcification may be seen. Meningiomas usually enhance homogeneously and intensely. The tumor is sharply marginated and is usually broadly based against a bony structure or dural margin. A common bony manifestation is hyperostosis. About 15% of benign meningiomas have a noncharacteristic appearance, including the presence of central lucency denoting necrosis or the presence of a cystic cavity (cystic meningioma). The amount of edema surrounding a meningioma is variable. The dura mater adjacent to the attachment of a meningioma may enhance on CT or MRI after the administration of a contrast agent. These so-called dural tails were studied histologically. Although only connective tissue and vascular tissue proliferation were seen in some cases, meningioma cell nests were identified in other cases.⁸⁵

On T1-weighted MRI, 60% of meningiomas are isointense and 30% are mildly hypointense compared with gray matter. On T2-weighted images, the tumors are isointense (50%) or mildly to moderately hyperintense (40%). Hyperintensity on T2-weighted images suggests a higher water content, denoting a meningothelial meningioma, a vascular meningioma, or an aggressive meningioma. However, it is suggestive of an easily suckable tumor during surgery. Meningiomas usually enhance intensely and uniformly after the injection of gadolinium, with typical dural tail enhancement.

Angiography may be an adjunct in the preoperative assessment of some meningiomas. It enables the surgeon to assess the vascularity and vascular supply of the tumor, the feasibility of embolization, and the presence of tumor encroachment on vascular structures. Meningiomas parasitize the adjacent blood supply (Table 131-2); knowledge of the vascular supply pattern allows the surgeon to gain early control of arterial feeders during surgery. Although there are no pathognomonic angiographic features of meningiomas, some typical findings have been delineated (Table 131-3). Preoperative embolization has not been of great benefit in the surgical treatment of meningiomas.⁸⁶

TABLE 131-3 Typical Angiographic Features of Meningioma

Somatostatin receptor scintigraphy using ¹¹¹In octreotide is an extremely sensitive test for meningiomas. It can be used preoperatively when the diagnosis is not straightforward. In the postoperative period, somatostatin receptor scintigraphy can be helpful in differentiating contrast uptake from residual tumor and that from nonspecific hyperperfusion.⁸⁷

Imaging techniques are being used to determine the histologic subtype or biologic behavior of meningiomas. Results are still preliminary, but it seems that aggressive meningiomas have higher metabolic rates than benign meningiomas, with the metabolic rate gauged by positron emission tomography using ¹⁸F fluorodeoxyglucose. Other investigative groups have focused their attention on different tumoral substrates using magnetic resonance spectroscopy.

Several lesions may mimic meningiomas on radiologic studies. Superficial metastasis and a variety of neoplasms may appear similar to meningiomas on routine radiologic work-up. Two entities that merit mention are neurosarcoidosis and solitary fibrous tumors. Sarcoidosis is a granulomatous disorder of unknown origin. It may be associated with systemic manifestations and is more common in African Americans. Rarely, neurosarcoidosis is the first or sole manifestation of this disease. The intracranial disease responds well to corticosteroids, although this improvement may not be apparent preoperatively. If the neurosurgeon embarks on surgery aimed at meningioma and is faced with an unexpected tumor appearance and texture, a frozen section unveils the real pathology. Although total resection is the surgery of choice for meningiomas, safe debulking followed by corticotherapy is recommended for sarcoidosis.⁸⁸ Solitary fibrous tumors of the meninges are tumors of mesenchymal origin and not of meningothelial lineage, as meningiomas are. They can be differentiated from meningiomas on immunohistochemical and ultrastructural grounds. They are positive for CD34 and vimentin but negative for EMA and S-100.⁸⁹ Ultrastructural features of meningiomas, such as complex interdigitation of cell processes and intercellular specialized junctions, are absent. The cells show the typical appearance of fibroblasts, with proximity of banded collagen and precollagen as well as cytoplasmic, rough-surfaced endoplasmic reticulum.⁹⁰

Surgical Therapy and Tumor Recurrence

The only definitive cure for meningioma is complete surgical resection. The more complete the resection, the less chance there is of recurrence.

In 1957 Simpson⁹¹ introduced a five-grade classification of the surgical removal of meningiomas (Table 131-4). Recurrence rates for grade I are about 10%; those for grade II are twice as high. The rates of recurrence are understandably higher in the higher Simpson grades. The inclusion of an additional 2-cm dural margin has been denoted grade 0 removal.⁹² In one study, no recurrences were seen in patients with convexity meningiomas in whom grade 0 resection was achieved; the mean follow-up period was 5 years and 8 months.⁹² In 1992 Kobayashi and associates⁹³ revised the Simpson grading system from a microsurgical perspective by introducing a classification system based on the extent of microscopic resection (Table 131-5).

TABLE 131-4 Simpson Grading System

Adapted from Simpson D. The recurrence of intracranial meningiomas after surgical treatment. J Neurol Neurosurg Psychiatry. 1957;20:22.

TABLE 131-5 Modified Shinshu Grade or Okudera-Kobayashi Grade

Adapted from Kobayashi K, Okudera H, Tanaka Y. Surgical considerations on skull base meningioma. Paper presented at the First International Skull Base Congress, June 18, 1992, Hanover, Germany.

The anatomic location of a meningioma influences its rate of recurrence. Tumors that are more difficult to remove totally, such as meningiomas of the sphenoid wing, recur more often. Meningiomas that invade a dural sinus, such as parasagittal meningiomas, have a high rate of recurrence. The recurrence rates of meningiomas differ from one series to another; the highest recurrence rates (>20%) are found in patients with sphenoid wing meningiomas, followed by those with parasagittal meningiomas (8% to 24%). The recurrence rate for convexity and suprasellar meningiomas is 5% to 10%.

Although well-delineated meningiomas can be totally removed, meningiomas with flat extensions into the subdural space (10% of meningiomas) are difficult to resect completely, as are en plaque meningiomas. The risk of recurrence is increased for meningiomas with aggressive pathologic features, such as invasion of the dura or brain infiltration. Other aggressive features include a papillary or hemangiopericytic pattern. Cellular criteria portending aggressive behavior include the presence of mitoses, increased cellularity, nuclear polymorphism, and focal necrosis. A high mitotic index is also an ominous sign. In an interesting article published by Yamasaki and colleagues,⁹⁴ 54 patients with supratentorial convexity meningiomas were examined at least 3 years after surgery or until tumor recurrence. Patients with multiple meningiomas, neurofibromatosis, and atypical and anaplastic meningiomas were excluded. All patients had undergone Simpson grade I resection. The correlation between recurrence and the following factors were statistically analyzed: age, gender, tumor volume, tumor shape, bone changes, brain edema, vascular supply, histologic subtype, MIB-1 labeling index, and vascular endothelial growth factor (VEGF). High levels of expression of VEGF constituted the most useful predictor of recurrence, followed by high MIB-1 labeling index.⁹⁴ Moller and Braendstrup⁹⁵ found, however, that proliferating cell nuclear antigen and Ki-67 are of minor value as predictors of the recurrence of benign meningiomas.

Nakasu and coworkers⁹⁶ studied 101 patients who underwent macroscopically complete removal of meningiomas. The patients were followed postoperatively for at least 5 years (maximal duration, 18 years) or until tumor recurrence. Fifteen meningiomas recurred during the follow-up period. Multivariate analysis revealed that only the shape of the tumor was significant; “mushrooming” and lobulated meningiomas were more likely to recur than round ones.⁹⁶

Nonsurgical Therapy

Nonsurgical therapies are used for recurrent or incompletely resected meningiomas. Standard or stereotactic irradiation has been used.^97,98 A comprehensive discussion of radiation therapy, both conformal fractionated radiotherapy and radiosurgery, is beyond the scope of this chapter, although the past decade has seen an increasing application of these two techniques both as intitial therapy and as adjuvants to surgical resection.^99–106 In terms of long-term control, however, the clinical recurrence rate after 15 years with subtotal resection and radiation therapy is 75%, and the rate of complications is 56%.¹⁰⁶ Radiation-induced complications and the possibility of radiation-induced malignancy or tumor progression are concerning issues.^107–112 Also concerning is the pattern of aggressive growth after failed radiosurgery.¹¹³

Acknowledging that the potential benefits of radiotherapy should always be balanced against its known side effects and complications, Guthrie and associates⁹⁷ concluded that “while surgical excision is the treatment of choice, radiation therapy should be considered: (1) after surgery for a malignant meningioma, (2) following incomplete resection of a meningioma for which the risk of resection of an eventual recurrence is judged to be excessive, (3) for patients with multiple recurrent tumors for whom the surgeon judges repeat surgery to be too risky, and (4) as a sole therapy of a progressively symptomatic patient with a meningioma judged by the surgeon to be inoperable.” Although the adjuvant role of stereotactic radiotherapy is gaining ground among neurosurgeons, its use as primary treatment for poorly accessible meningiomas (e.g., cavernous sinus or clival) is the subject of ongoing debate.^114,115 In particular, data reported by Rowe and coworkers¹¹² indicate only a 53% actuarial 15-year survival rate for patients with meningiomas treated with the Gamma Knife, and 67% of these patients died from their meningiomas.

The presence of hormonal receptors in meningiomas has prompted research into hormonal manipulation as treatment. Tamoxifen, an estrogen antagonist, was shown to stimulate meningioma cells in culture, perhaps because of its partial estrogen-agonistic activity. Oura and colleagues¹¹⁶ reported an anecdotal case of a patient with gastric carcinoma treated with mepitiostane, an antiestrogen agent. The patient also suffered from a presumed falcine meningioma discovered incidentally during the metastatic work-up. The patient was treated with mepitiostane for 2 years, and the meningioma had decreased by 73% at the end of this period.¹¹⁶ Bromocriptine, a dopamine antagonist, inhibits meningioma cells significantly in vitro. Other studies are assessing the efficacy of mifepristone (RU 486), an antiprogesterone agent, in the treatment of meningiomas. Mifepristone has progesterone- and cortisol-blocking activities; most patients initially complain of nausea, vomiting, or tiredness. Although preliminary results were encouraging, it is still too early to assess the efficacy of RU 486 in the treatment of meningiomas. Other antiprogesterone agents such as gestrinone have been used in the medical treatment of meningiomas.¹¹⁷ Other antineoplastic drugs, including hydroxyurea^118,119 and interferon alpha-2B,¹²⁰ have also been used.