36.2.1 Trauma

Multiple case reports exist documenting intracranial dural-based,3 ^,​ 4 intraosseus,5 and extracranial cutaneous6 ^,​ 7 meningiomas at the site of prior convexity or skull base injury. Likewise, several case–control studies have demonstrated an increased risk of meningioma genesis after head injury,8 ^,​ 9 ^,​ 10 ^,​ 11 ^,​ s. Literatur although these results have not been universally reproducible.13 ^,​ 14 Recall bias has been suggested as a confounding factor limiting the effectiveness of case–control studies. Further epidemiological studies are necessary to validate and delineate a relationship.

36.2.2 Radiation

Exposure to ionizing radiation is a known etiological factor in the development of primary CNS tumors, with meningioma being the most common.15 Of all radiation-induced tumors, the linear dose–response curve for development of meningioma is second only to that for sarcoma.16 For a meningioma to be defined as radiation-induced, specific criteria have been proposed, including the following: (1) tumor was not present prior to irradiation, (2) tumor arose within the previously irradiated field, (3) a reasonable interval separated radiation therapy and detection of the second tumor, (4) tumor was confirmed histologically, (5) no family history is present of tumor predisposition syndrome such as phakomatosis, (6) histologic features differ from the primary tumor, and (7) tumor must not be recurrent or metastatic.15 ^,​ 17 ^,​ 18 Compared with spontaneous meningiomas, radiation-induced meningiomas (RIMs) more commonly display higher World Health Organization (WHO) grade features and are more biologically aggressive tumors. For example, in two separate series of 10 RIMs, 50% were grade II or III (Galloway and Mack). Similarly, Al-Mefty et al reported a series of 16 RIMs, 38% of which were atypical or anaplastic tumors, with 100% of RIMs recurring at least once, 62% twice, and 17% thrice.15 RIMs are more commonly multicentric rather than solitary19 ^,​ 20 and demonstrate different cytogenetic changes, including less frequent NF-2 mutations and more frequent 1p and 6q mutations than their spontaneous counterparts.15 ^,​ 21

Examples of RIMs from both high- and low-dose exposure abound in the literature. High-dose exposure is loosely defined as therapeutic radiation used to treat another disease, and RIMs have been reported to develop after exposure to doses as low as 10 Gy20 ^,​ 22 up to levels as high as 20 Gy23 or 30 Gy. One of the first studies demonstrating the oncogenic potential of therapeutic radiation was published in 1974 by Modan and colleagues.25 In this series, approximately 11,000 children who received high-dose radiation for treatment of tinea capitis, along with an age-matched cohort, were retrospectively followed for up to 23 years. Those exposed to the cranial radiation showed a relative risk of 9.5 for the development of meningioma. Similarly, children receiving high-dose prophylactic radiation to the neuroaxis for acute lymphoblastic leukemia have been shown to develop RIMs.26

RIMs developing after exposure to lower doses of radiation typically are from occupational, industrial, or environmental sources such as medical imaging (X-rays, CTs), nuclear power plants, or nuclear bombs, but the radiation dose from each of these is highly variable. For example, studies of atomic bomb survivors in Hiroshima and Nagasaki have shown a relative risk of 6.48 compared with non-exposed populations.27 ^,​ 28 In the past few decades, the rising use of cellular telephones and the low-dose radiation levels to which they expose users prompted several large multinational studies evaluating the risk of brain tumor development.29 ^,​ 30 The study results were congruent, demonstrating that cell phone use is not associated with an elevated risk of meningioma development. However, in the French CERENAT study, the heaviest cell phone users were found to have a 2.5-fold increased risk of meningioma development when considering lifelong radiation exposure.29

36.2.3 Infection

Viruses, and in particular polyomaviruses, have been studied as etiological agents of meningioma development. Simian virus 40 (SV40), a polyomavirus, is capable of transforming cells into those having a neoplastic phenotype.31 The oncogenic and transforming properties of SV40 are related to expression of large tumor antigen (Tag), which is postulated to have a role in inactivating the tumor suppressor functions of p53, pRb, p107, and others.32 Of 10 human meningioma samples analyzed, SV40 Tag was identified in 7 samples, the Tag-p53 complex in 3, and the Tag-pRb complex in 2.33 In 2004, the first report linking SV40 exposure to meningioma was published.34 The authors described a case of a scientist who had direct exposure to SV40 in the lab setting and who later developed an intracranial meningioma having identical SV40 DNA sequences. Although these data implicate SV40 in the pathogenesis of meningioma, other studies have not demonstrated the same results. In a study of 15 meningiomas, Rollison et al were unable to identify SV40 in any of the samples.35 Similarly, in a larger series, Weggen et al found SV40 DNA in only 1 of 131 meningiomas sequenced.36 Furthermore, Polymerase chain reaction (PCR) analysis for the Tag gene demonstrated a rate of 1 gene in every 250 tumor cells. Although the low frequency of SV40 and Tag DNA does not completely rule out a viral role in the pathogenesis of meningiomas, it appears to play a smaller role than was initially thought.

36.3.1 Familial Meningioma Syndromes

The majority of meningiomas are sporadic tumors in patients having no history of brain tumors. However, familial meningiomas have been identified in a number of conditions, including neurofibromatosis (NF) type 2 (NF2, 22q12.2), NF-1 (NF1, 17q11.2), Cowden’s disease (PTEN, 10q23.31), Gorlin’s or nevoid basal cell carcinoma syndrome (PTCH, 9q22.3), Li-Fraumeni syndrome (TP53, 17q11.2; CHEK2, 22q12.1), Gardner’s syndrome (APC, 5q21–22), Rubinstein-Taybi syndrome (CREBBP, 16p13.3; EP300, 22q13), von Hippel-Lindau syndrome (VHL, 3p26–25; CCND1/cyclin D1, 11q13), Werner’s syndrome (LMNA, 1q21.1; RECQL2, 8p12-p11.2), and multiple endocrine neoplasia Type 1 (MEN, 11q13), as well as melanoma/astrocytoma–brain tumor syndrome.37 ^,​ 38

36.3.2 Chromosomal Abnormalities

Meningiomas were one of the first solid tumors to be associated with a characteristic cytogenetic change—loss of heterozygosity (LOH) of chromosome 22. Forty to seventy percent of meningiomas exhibit LOH for markers from the chromosomal region 22q12.2, which encompasses the NF2 gene. NF2 encodes the protein merlin, which links membrane proteins to the cytoskeleton.39 Biallelic inactivation of merlin results in loss of contact inhibition of cell proliferation and tumorigenesis.40 Mutations in the NF2 gene have been reported in 30 to 60% of sporadic meningiomas. The frequency of NF2 mutations is similar among WHO grade I, II, and III meningiomas.37 ^,​ 41 This finding suggests that NF2 gene inactivation may be an important initiation step in the formation of meningiomas but might not play a role in tumor progression.38 Hansson et al used microarray-based comparative genomic hybridization to study 126 sporadic meningioma specimens.42 They found the incidence of biallelic NF2 inactivation to be 52% in fibroblastic variants, compared with 18% in meningothelial histologies, suggesting that NF2 inactivation might not be a critical step in the formation of meningothelial meningiomas.42 Instead, an alternate pathogenetic pathway could be deletion of the terminal segment of the long arm of chromosome 22 (22qter), as suggested by Yilmaz et al after the authors noted this specific genetic alteration in 26 of 36 patients (72%) who had meningothelial meningioma.43

Mutations on chromosome 1 represent the second most common genetic alteration in meningiomas.42 Specific loss of the 1p36 locus has been associated with increased44 ^,​ 45 and early46 recurrence. In a large series of 247 grade I and II meningiomas, loss of 1p36 resulted in almost twice the number of recurrences as for patients who did not have 1p36 loss (33% vs. 18%).44 Loss of chromosome 14q represents the third most common genetic abnormality after mutations in chromosomes 22 and 1.47 As opposed to chromosome 22 loss, which occurs uniformly across all three grades of meningioma, 14q deletions occur in up to half of atypical meningiomas and the majority of anaplastic variants.47 However, as with 1p loss, this genetic finding occurs at much lower rates in grade I meningiomas.48 Not surprisingly, 14q deletions are associated with an increased risk of relapse and poorer prognosis.49

The increased recurrence rate of 14q deleted tumors was amply demonstrated in a 2017 meta-analysis of 742 patients who had meningioma, with an odds ratio for recurrence of 7.6 (95%CI: 4.3–13.6).50 Similarly, 1q deletions were associated with an increased risk of recurrence (Overall Risk [OR] = 5.4; 95%CI: 3.6–8.1); however, chromosome 22 deleted tumors were far less likely to recur (OR = 1.6; 95%CI: 1.1–2.4).50 With respect to meningioma, a specific locus of interest on chromosome 14q is the tumor suppressor gene maternally expressed gene 3 (MEG3), located at 14q32.51 The antitumoral effect of MEG3 is mediated through activation of other tumor suppressor pathways, including p53 and Rb,51 ^,​ 52 and through downregulation of MDM2, a gene encoding a protein responsible for p53 degradation.52 MEG3 is overexpressed in normal arachnoidal cells but is not expressed in most meningiomas.51 Among meningiomas, loss of MEG3 occurs more frequently in more aggressive recurrent tumors.48 ^,​ 53 Work by Zhang et al in 2010 demonstrated that MEG3 is downregulated by epigenetic mechanisms such as CpG methylation within the promoter and the imprinting control region in meningiomas.51 In addition, MEG3 suppression of DNA synthesis has been demonstrated via in vitro meningioma models, collectively demonstrating the tumor suppressor effect of MEG3.51

Chromosomal losses (1p, 6q, 10q, 14q, and 18q) and gains (1q, 9q, 12q, 15q, 17q, and 20q) have been associated with meningioma progression from low-grade to atypical.54 Not infrequently, meningiomas express multiple genetic mutations, including LOH or co-deletion of 1p and 22q or 1p and 14q. These tumors demonstrate abnormal gene expression patterns on cDNA microarray analysis,55 resulting in biologically more aggressive, higher-grade tumors56 that have earlier recurrence rates.46 Grade III or anaplastic tumors are associated with gains on 17q23 and losses on 9p. They may also demonstrate more frequent losses on 6q, 10, and 14q than are seen with atypical tumors.

36.3.3 Next-Generation Sequencing and NF2 Wild Type Mutations

Historically, meningioma has been considered a surgical disease: it most commonly exhibits slow growth with a low mitotic index, and surgical resection, if complete, is considered curative. Some low-grade tumors, however, behave more aggressively with early recurrence; others are in locations that prevent safe gross total resection; and still others are resistant to all available treatment methods, including surgery, radiation, and chemotherapy. Accordingly, a significant amount of genetic investigation has been undertaken to elucidate possible genetic targets for medical treatment in both nonresectable low-grade skull base meningiomas and recurrent high-grade tumors.57 ^,​ 58 ^,​ 59 With the advent of next-generation sequencing, a rapid and relatively inexpensive whole genome sequencing technique, researchers have been able to identify specific mutation patterns among tumors of similar histologic grades. This has allowed the identification of subgroups of tumors that may be expected to recur earlier or more often, resulting in more specific behavior prediction. Furthermore, meningioma genotyping has identified specific molecular signatures of histologic phenotypes.60 Taken together, next generation sequencing (NGS) is transforming the way meningiomas are categorized, from histologic subtype to specific genotype.61

Beginning in 2013, several oncogenic driver mutations were discovered in meningioma, including smoothened (SMO) and V-AKT murine thymoma viral oncogene homolog 1 (AKT1), both of which activate tumorigenic pathways in multiple cancers, including basal cell carcinoma, medulloblastoma, breast cancer, colorectal cancer, and lung cancer.57 ^,​ 62 This was followed by the discovery of two additional oncogenic mutations: tumor necrosis factor receptor-associated factor 7 (TRAF7), which loses its normal apoptotic function when mutations occur in the WD40 binding domain, and Kruppel-like factor 4 (KLF4), which loses its normal regulatory role of cell differentiation due to a mutation in its zinc finger DNA binding site.58 ^,​ 60 Several years later Abedalthagafi and colleagues discovered that phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha (PIK3CA) mutations occur at a similar frequency as SMO, AKT1, TRAF7, and KLF4.59 PIK3CA mutations exert a tumorigenic effect via activation of the AKT/mTOR pathway for cell growth and proliferation.59

With respect to mutation location and frequency, AKT1 and KLF4 are both frameshift mutations that occur at a single location—c.49G > A (p.E17K) and c.1225A > C (p.K409Q), respectively41 ^,​ 57 ^,​ 58—whereas SMO mutations most commonly involve p.L412F and p.W535 L changes. The most common mutation pattern in PIK3CA mutant tumors is p.H1047 R, which is found in a little less than half of tumors.41 TRAF7 mutant tumors demonstrate variability in mutation locus, but 90% occur within the WD40 zinc finger DNA binding domain41 and frequently co-occur with KLF4 and AKT1 mutations.

In general, these somatic mutations occur in about 5 to 10% of all meningiomas. They are mutually exclusive of NF2 mutant tumors, occurring only in NF2 wild type meningiomas57 ^,​ 58 ^,​ 59 ^,​ 60 that are most commonly, but not always, low-grade. Interestingly, genotype seems to predict tumor location within the skull, where NF2 mutant meningiomas occur in the posterior fossa and tumors having TRAF7, AKT1, KLF4, SMO, and PIK3CA mutations most frequently occur in the anterior and middle fossa compartments along the midline.57 ^,​ 59 ^,​ 60 ^,​ 61 In a retrospective review of 62 anterior skull base meningiomas by Strickland et al, 7 tumors demonstrated SMO mutations, of which 6 (82%) arose from the olfactory groove and 2 were grade II.63 The authors concluded that olfactory groove meningiomas should be screened for SMO mutations.63 Multiple recent meningioma sequencing studies have demonstrated that genotype predicts histologic phenotype.

In a 2013 study of 30 secretory meningiomas, Reuss et al found that all 30 tumors demonstrated the KLF4 K409Q mutation and that 29 exhibited the TRAF7 mutation within the WD40 binding domain.60 The authors sequenced 267 other brain tumors, including other meningiomas, gliomas, metastases, glioneuronal tumors, and pituitary adenomas, along with 14 intraductal papillary mucinous neoplasms, 34 ductal adenocarcinoma, and 4 serous cystadenomas of the pancreas, without detecting the KLF4 mutation in any of the tumors; they detected TRAF7 in 8% of meningiomas. Taken together, these data support the conclusion that the KLF4/TRAF7 mutation defines the genetic signature of secretory meningioma. Similarly, germline loss of SMARCE1 has been demonstrated as a cause of both spinal and clear cell meningioma64; SMO mutations frequently result in meningothelial phenotype, and PIK3CA/AKT1 mutant tumors are most often of transitional or meningothelial histology.57 ^,​ 59

36.3.4 Gene Sequencing for Prognosis

Genetic sequencing experiments have demonstrated multiple prognostic implications of these NF2 WT mutations as well as other novel mutations that can help guide patient counseling and treatment and eventually inform the organization of a novel molecular grading system for meningioma. In a series of 79 olfactory groove meningiomas, those having SMO mutations occurred earlier and more than twice as often as AKT1 mutant tumors or wild type tumors, a finding that was independent of tumor grade.65 In another series of 93 grade I and II skull base meningiomas, patients who had AKT1 mutant tumors had a shorter recurrence-free survival than their wild type counterparts.66 Conversely, tumors that had the KLF4 K409Q mutation recur much later than KLF4 WT tumors and might represent a more benign subtype66 TERT promoter mutational status has recently been identified as an extremely ominous genetic abnormality that can occur across all three histologic grades of meningioma and that results in shorter time to recurrence than anaplastic WT tumors.67 In this study, the authors sequenced 267 tumors of grade I, II, and III histology, noting 16 tumors (6%) that had TERT promoter mutations: 2 grade I, 5 grade II, and 9 grade III. TERT promoter mutants demonstrated a significantly shorter recurrence-free survival (10 mo versus 179 mo, p = 0.001) than TERT WT tumors, regardless of grade. Furthermore, both grade I and 4/5 grade II mutant tumors demonstrated higher histologic grade features at recurrence. In view of the significance of this finding, the authors have suggested that TERT promoter mutational status be incorporated in the next meningioma grading system.67

Similarly, DNA methylation patterns identifying genetic groups at risk of earlier recurrence have been demonstrated.68 BAP1 (BRCA1 associated protein-1) is a tumor suppressor gene that mediates its effects through chromatin modulation and transcription regulation. BAP1 germline mutations have been identified in a cancer predisposition syndrome that results in multiple tumors, including meningiomas, uveal melanoma, lung adenocarcinoma, and other cancers.69 The frequency of BAP1 loss was studied in 57 rhabdoid meningiomas.69 The authors identified 5 tumors (9%) with BAP1 loss, all of which were associated with a higher percentage of rhabdoid cells, higher grade (II or III), higher mitotic rate (> 5 mitoses/10 high-power fields [HPF]) and shorter time to recurrence than rhabdoid tumors with normal BAP1 expression.

Olar et al optimized a recurrence predictor using a training/validation approach and a support vector machine classification method with radial-basis smoothing kernel.70 Three publicly available Affymetrix gene expression data sets (GSE9438, GSE16581, GSE43290) combining 127 newly diagnosed meningioma samples served as the training set. Unsupervised variable selection was used to identify an 18-gene gene expression profile (18-GEP) model that separated recurrences with a negligible root mean square error of 0.17. The characteristics of the training data set were as follows: WHO grade (I: 92 [73%], II: 32 [25%], III: 2 [2%]), median follow-up = 5.53 years (range: 0.05–25.42), recurrences = 18. This model was tested on 62 cases from their institution (validation data set [VD]) having similar demographics but enriched for cases featuring either long clinical follow-up or known recurrence. When applied to the VD, the 18-GEP separated recurrences with a misclassification error rate of 0.25 (log-rank p = 0.0003). 18-GEP was significantly predictive of tumor recurrence, independently (p = 0.0008, hazard ration [HR] = 4.61, 95%CI = 1.89–11.23]), and was predictive after adjustment for WHO grade, mitotic index, sex, tumor location, and Simpson grade (p = 0.0311, HR = 9.28, 95%CI = [1.22–70.29]). The expression signature included genes encoding proteins involved in normal embryonic development, cell proliferation, tumor growth and invasion (FGF9, SEMA3C, EDNRA), angiogenesis (angiopoietin-2), cell cycle regulation (CDKN1A), membrane signaling (tetraspanin-7, caveolin-2), WNT-pathway inhibitors (DKK3), complement system (C1QA), and neurotransmitter regulation (SLC1A3, secretogranin-II).70

36.3.5 Genetic Underpinnings of Meningioma Development

Additionally, gene sequencing of meningiomas has begun to elucidate novel mechanisms of tumorigenesis. A 2017 meningioma sequencing experiment identified a subset of meningiomas with mutations in the RNA polymerase II gene POLR2A. RNA polymerase II is a key enzyme for transcription of protein coding genes in eukaryotes.61 In this study, 775 tumors were sequenced and 23 tumors (3%) demonstrated the mutation. POLR2A mutants express elevated levels of WNT6 and significant downregulation of ZIC1/ZIC4, two proteins involved in meningeal cell differentiation. Specifically, WNT 6 is a protein secreted by nonneural ectoderm cells to cause the induction of neural crest cells, whereas ZIC1 is expressed by neural crest cells and results in meningeal cell differentiation. Thus the combination of elevated WNT6 and reduced ZIC1/ZIC4 would presumably result in higher numbers of dedifferentiated meningeal progenitor cells that could result in tumor formation.61

36.4.1 Growth Factors

Meningiomas express multiple growth factors and their receptors, including epidermal growth factor receptor (EGFR), basic fibroblast growth factor receptor (BFGFR), platelet-derived growth factor receptor (PDGFR), and vascular endothelial growth factor-A (VEGF-A). This suggests that autocrine growth factor secretion and autocrine loops may play a role in the growth of meningiomas.71 This has been supported by experiments demonstrating meningioma cell growth and proliferation in the presence of growth factors.72 An investigation by Smith et al demonstrated PDGFR-β expression in all 84 meningioma samples studied.73 In addition, expression of BFGFR was found in 89% of benign meningiomas, whereas EGFR immunoreactivity was detected in 47% of benign meningiomas. In this study EGFR immunoreactivity was found to be a strong predictor of prolonged survival in patients who had atypical meningioma.73 VEGF-A, which is also known as vascular permeability factor, is considered to be a key factor in angiogenesis and edema formation for meningiomas. Several studies have demonstrated VEGF-A levels in meningiomas to be associated with the extent of peritumoral edema,74 ^,​ 75 and some smaller studies have postulated that VEGF-A mRNA expression may correlate with meningioma vascularity.75 ^,​ 76 Furthermore, loss of the tumor suppressor gene MEG3, a finding commonly seen in meningiomas, has been correlated with increased levels of VEGF in MEG3 knockout mice, suggesting a possible underlying genetic role in the development of elevated vascularity seen in some meningiomas (Gordon). Other studies, however, have found no association between VEGF-A protein levels and microvessel density.77 Multiple studies have assessed the effect of growth factor receptor inhibitors on meningioma growth. Unfortunately, the results from these studies have been largely underwhelming; the details are discussed in the chemotherapy treatment section of this chapter.

Transforming growth factor beta (TGF-β beta) 1, 2, and 3 along, with their receptors, are secreted and expressed by the leptomeninges and by meningioma tumor cells.78 In vitro studies of WHO I meningiomas have demonstrated that tumor cell secretion of TGF-β reduces basal cell proliferation, thereby providing constant autocrine reduction and prevention of tumor growth.79 In higher-grade meningiomas, however, TGF- β receptor levels are significantly reduced compared with grade I tumors, thereby compromising the inhibitory effect.80

36.4.2 Sex Steroid Receptors

The increased incidence of meningioma in women, along with the discovery of sex steroid receptors on meningioma tumor cells, has led to a significant amount of work examining whether endogenous or exogenous exposure to elevated levels of sex steroid increases risk of meningioma formation or tumor recurrence. Unfortunately, studies examining sex steroid sources, such as hormone replacement therapy and hormonal contraceptive use,81 ^,​ 82 ^,​ 83 ^,​ 84 ^,​ 85 ^,​ s. Literatur pregnancy,87 ^,​ 88 ^,​ 89 ^,​ 90 age at menarche or menopause,85 and breast cancer,91 ^,​ 92 have been largely incongruous. Although the vast majority of grade I meningiomas express progesterone receptors (PRs), scarcely any are found on grade II or III tumors.93 ^,​ 94 ^,​ 95 ^,​ 96 Most studies have found meningiomas to lack estrogen receptors (ERs), though some have found low concentrations of ERs in 5 to 33% of tumors.97 ^,​ 98 Although some published data suggest that PR-positive tumors are less aggressive and are less likely to recur than those that exhibit ER-positivity, these studies are offset by data suggesting that there is no such relationship.82 ^,​ 86 ^,​ 99

Studies examining the risk of estrogen-only or estrogen/progesterone combination contraceptives on meningioma recurrence have demonstrated variable results.100 ^,​ 101 ^,​ 102 Conversely, the few studies examining “ever-use” of exogenous, progesterone-only sources such as long-acting implantable and injectable contraceptives consistently demonstrated an elevated risk of meningioma genesis and recurrence.82 ^,​ 86 ^,​ 103 In response to a rise in use of pure progesterone contraceptives, the risk of meningioma recurrence was reexamined in a recent 2017 study.104 The authors studied 67 premenopausal women taking estrogen, estrogen/progesterone, or progesterone-only contraception who had surgically resected meningiomas and found that recurrence was higher in those taking pure progesterone than those taking estrogen only or a estrogen/progesterone combination (33% vs. 19%).104 The progesterone-only group also demonstrated a significantly shorter time to recurrence (18 vs. 32 mo). These data suggest that progesterone-only contraception should be avoided in premenopausal women who have a known history of meningioma.

36.4.3 Somatostatin Receptors

Somatostatin receptors, and specifically somatostatin receptor subtype 2A (SSTR2A), are present at variable concentrations on most meningioma tumor cells.105 Their presence can be confirmed in the clinical preoperative setting using octreotide nuclear scintigraphy, or postoperatively in the laboratory via immunohistochemistry.106 The SSTR2A mediates the antiproliferative effects of somatostatin on meningioma tumor cells.107 This was demonstrated in an in vitro study of 80 SSTR2A-positive meningiomas, in which a somatostatin analog significantly decreased cell proliferation in 88% of the tumors but no cell death was observed.108

36.4.4 Growth Hormone Receptors

Growth hormone receptors are present on meningioma tumor cells. Activation of the growth hormone/insulin-like growth factor 1 (GH/IGF-1) axis significantly increases the growth rate of meningiomas and has sparked investigation into inhibitors as treatment options.109

36.5.1 WHO Grade I

Meningothelial, fibrous, and transitional meningiomas are the three most common histological variants of grade I meningiomas. Less common subtypes include psammomatous, angiomatous, microcystic, lymphoplasmacyte-rich, and secretory meningiomas (Fig. 36.1). Chordoid, clear cell, papillary, and rhabdoid variants of meningiomas are often associated with more aggressive tumor behavior and are consequently not classified as grade I tumors. Benign grade I meningiomas may also invade surrounding structures, including the skull, dural sinuses, orbit, and soft tissues.

Meningothelial meningioma is a common variant in which tumor cells form lobules surrounded by thin collagenous septae. Tumor cells are uniform, resembling normal arachnoid cap cells, and may show central clearing. Rounded eosinophilic cytoplasmic protrusions into the nucleus, termed pseudoinclusions, are also seen. Whorls and psammoma bodies, when present, are less well defined than those seen in fibrous, psammomatous, or transitional meningiomas. Fibrous meningiomas are formed by spindle-shaped cells that resemble fibroblasts. These spindle-shaped cells form intersecting fascicles and are embedded in a collagen-rich and reticulin-rich matrix. Transitional or mixed meningioma is a common histological subtype that contains features of both meningothelial and fibrous meningiomas. Both the lobular arrangement of meningothelial meningiomas and the fascicular pattern of fibrous variants may be seen next to one another. They usually demonstrate extensive whorl formation in which the tumor cells form concentric cell layers by wrapping around each other. When the whorl formations hyalinize and calcify, they form structures known as psammoma bodies. Secretory meningiomas are rare grade I meningiomas with secretory globules (pseudopsammomas) that result in extensive peritumoral edema,113 which appears very aggressive on imaging. Although the extent of edema does not correlate with the proliferation index, it is associated with immunohistochemical presence of carcinoembryonic antigen and cytokeratin.114 ^,​ 115 ^,​ 116

36.5.2 WHO Grade II

According to the WHO 2016 criteria, a meningioma is classified as atypical if it demonstrates brain invasion, has ≥ 4 mitoses per 10 HPF, or meets three of the five following criteria: increased cellularity, high nuclear to cytoplasmic ratio (small cells), prominent nucleoli, uninterrupted patternless or sheetlike growth, or foci of spontaneous (not induced by embolism) necrosis.111 Two meningioma variants, clear-cell and chordoid, have been found to have higher recurrence rates even in the absence of the foregoing criteria, leading to their classification as atypical meningiomas (Fig. 36.2).

Clear cell meningioma is a rare meningioma variant composed of sheets of polygonal cells whose clear, glycogen-rich cytoplasm is positive for periodic acid Schiff. They also demonstrate dense perivascular and interstitial collagen deposition. Clear cell tumors are more commonly found in the cauda equina or the cerebellopontine angle and tend to affect younger patients. Chordoid meningiomas are typically supratentorial tumors and have regions that are histologically similar to chordoma. They may have cords of small epithelioid tumor cells that contain eosinophilic cytoplasm residing in a basophilic, mucin-rich matrix.54

36.5.3 WHO Grade III

Malignant meningiomas are characterized by a highly elevated mitotic index of > 20 mitoses per 10 HPF and can pathologically resemble sarcoma, carcinoma, or melanoma. In addition, large areas of necrosis can usually be seen. Papillary and rhabdoid meningiomas are consistently associated with malignant behavior and are therefore classified as WHO grade III tumors (Fig. 36.3).117 Papillary meningioma has a propensity to affect younger patients, commonly exhibits brain invasion, and may show distant metastases. The rhabdoid variant is a rare tumor that contains sheets of large rhabdoid cells whose eosinophilic cytoplasm contains whorled intermediate filaments and eccentric nuclei (Perry 1998). The median survival is less than 2 years, and the recurrence rate is 50 to 80% after surgical resection.118

36.5.4 Immunohistochemistry

Epithelial membrane antigen (EMA) is the most commonly used marker in pathological analysis of meningiomas, and most tumors show at least scattered positivity for this antigen.119 All meningiomas also strongly express vimentin.120 Tissue microarray immunohistochemistry (TMA-IHC) has proven to be an efficient and reliable method for analyzing biomarkers in meningioma. Using TMA-IHC, Lusis et al found EMA reactivity in 100% of meningiomas regardless of grade and E-cadherin immunoreactivity in 91% of all meningiomas and 90% of anaplastic meningiomas.93 These markers, however, are not specific for meningiomas.

Immunohistochemistry has also proved useful in quantifying the proliferative index for meningiomas with the antibody MIB-1, which targets the proliferation marker Ki-67. Elevated proliferative indices as measured by MIB-1 labeling have been associated with an increased risk of recurrence.121 Mean MIB-1 labeling indices are 0.7 to 2.2% for grade I meningiomas, 2.1 to 9.3% for grade II meningiomas, and 11 to 16.3% for grade III meningiomas.122 MIB-1 labeling indices of greater than 5% suggest a greater likelihood of recurrence.

Similarly, an antibody that specifically recognizes the phosphorylated histone H3 (PHH3) has been effectively used as an aid in grading meningiomas.123 During mitosis, phosphorylation of the Ser-10 residue of histone H3 reaches a maximum. Consequently, immunostaining with anti-PHH3 allows the observer to rapidly focus on the most mitotically active areas of the tumor. In 2015, Olar et al evaluated the mitotic index (number of mitoses per 1,000 tumor cells) in 363 meningiomas using PHH3 IHC.124 In multivariate analysis, mitotic index was significantly associated with recurrence free survival (P < 0.01) after adjustment for Simpson grade, WHO grade, and MIB-1 index.

36.6.1 Surgery

Complete surgical excision is the treatment of choice for meningiomas. In 1957, Simpson published a landmark paper examining the effect of degree of meningioma resection on recurrence rates following surgery.125 Tumor resections were divided into five categories:

• 
  

  I: macroscopically complete resection of tumor, bone, and dura

  
  
• 
  

  II: macroscopically complete resection of tumor with dural coagulation

  
  
• 
  

  III: complete tumor resection without dural coagulation

  
  
• 
  

  IV: subtotal resection

  
  
• 
  

  V: simple decompression

Kinjo et al further suggested a grade 0 resection in which an additional 2 cm margin of dura is excised as a means for further reducing the rate of recurrence.126 A direct relationship between Simpson grade and 10-year symptomatic recurrence rate was demonstrated as follows: grade I (9%), grade II (19%), grade III (29%), grade IV (44%), grade V (100%). As a result, this scale has been adopted by neurosurgeons as a means of quantifying degree of meningioma resection. However, most attempts to re-create Simpson’s results have failed to demonstrate a significant difference in recurrence rate among grades I, II, and III.127 ^,​ 128 ^,​ 129 ^,​ 130

The extent of resection has since been strongly confirmed as the primary factor influencing meningioma recurrence rate. The 10-year progression-free survival data range from 61 to 80% for a gross total resection and from 37 to 45% for a subtotal resection.131 ^,​ 132 Nanda et al reported 20-year follow-up data for 112 WHO grade I meningiomas, noting a significant difference in tumor recurrence rate between grades 0 and I and grades II and IV (2.9% vs. 31%, p = 0.0001), a finding that was statistically similar when evaluating degree of resection using the Shinsu resection grading scale.133

The ability to achieve a safe, complete resection is influenced by invasion into eloquent cortex, tumor involvement of dural sinuses, growth into cranial nerves such as is commonly seen in cavernous sinus meningiomas, association with vascular structures, tumor location and size, and previous surgery or radiation.110 The vast majority of meningiomas are surrounded by a layer of arachnoid that separates the tumor from the brain, cranial nerves, and blood vessels. By accessing this arachnoidal plane, the surgeon is able to minimize the chance of injury to neurovascular structures. Fortunately, an arachnoid plane can frequently be visualized on MRI T2 fluid-attenuated inversion recovery (FLAIR) images and when present has been correlated with improved surgical outcomes.134 Internal debulking of the tumor facilitates the delineation of the arachnoid plane by allowing the edge of the tumor to collapse inward. Meningiomas may attach to or surround cerebral arteries diminishing the diameter of the vessel, but only very rarely do they invade the arterial walls.

36.6.2 Radiation

36.6.3 Chemotherapy

Use of chemotherapy for treatment-resistant meningioma has been investigated extensively over the years. Unfortunately, the multitude of trials exploring most classes of standard chemotherapeutic agents have been met with largely disappointing results. Of all the chemotherapy drug trials conducted for meningioma, one of the best results came from a prospective study of 14 patients who had progressive malignant meningioma, who were given cyclophosphamide, Adriamycin, and vincristine (3 cycles for GTR and 6 cycles for STR) after surgery and radiation.179 The authors reported a median time to progression of 4.6 years and median overall survival of 5.3 years—a finding significantly better than that associated with surgery alone. Hydroxyurea is an oral ribonucleotide reductase inhibitor that arrests meningioma cell division in the S phase of the cell cycle and induces apoptosis.180 Though this agent has been effective in in vitro and in vivo studies, treatment of patients who have recurrent or unresectable meningiomas has shown little benefit, with the best results being no more than 1 year of progression-free survival.181 ^,​ 182 ^,​ 183

Temozolomide, an alkylating agent that has been used for treatment of malignant gliomas, showed no benefit for treating refractory meningiomas in a phase II trial184; however, in vitro trials of high-mobility group nucleosome-binding protein-5 (HMGN5) inhibition that demonstrated temozolomide sensitization deserve further investigation.185 Irinotecan (CPT- 11), a topoisomerase-1 inhibitor, has been shown to inhibit in vitro cultures of human meningioma cell lines and in vivo studies using a subcutaneous tumor model.186 However, a phase II study evaluating CPT-11 in patients who had recurrent meningioma was stopped prematurely when all patients demonstrated tumor progression within 6 months.187

36.6.4 Hormone Therapy

The presence of hormone receptors on meningioma tumor cells sparked a plethora of trials investigating the clinical efficacy of using inhibitors of these receptors in treatment of recurrent meningioma. Mifepristone (RU486) is a progesterone blocker that has been shown to inhibit the growth of cultured human meningioma tissue and meningioma in animal models.188 In a 2015 double blind, prospective phase III trial, 164 patients who had unresectable meningioma were given either mifepristone or placebo for 2 years.189 At follow-up there were no differences between the two groups in failure-free survival or overall survival. Similarly, a phase II trial of tamoxifen, an ER antagonist, for treatment resistant meningioma was largely ineffective, with tumor progression noted in 10 patients within 6 weeks of treatment.190 Somatostatin receptors are present on meningioma cells, and in vitro studies examining somatostatin analogs have demonstrated an inhibitory effect on growth of meningioma cells. Unfortunately, degree of receptor expression does not correlate with clinical response,191 and thus far, somatostatin analogs have not proven effective for recurrent receptor-positive meningioma in the clinical setting. Specifically, one pilot study and two phase II prospective trials evaluating a total of 39 patients demonstrated 5 partial radiographic responses and a median time to progression of 4 to 5 months.191 ^,​ 192 ^,​ 193 Although somatostatin analog therapy has been mostly unsuccessful, somatostatin receptor-targeted radionuclide therapy with⁹⁰Y-DOTATOC may be another option, a few trials having shown promise.194 ^,​ 195 Yttrium-90 is a radioactive isotope that is combined with DOTATOC, a tetraazocyclo-dodecanetetraacetic acid–modified somatostatin analog. In a trial of 15 patients who had recurrence, receptor positive meningioma stable disease was achieved in 13 patients at 24-month follow-up.195

Similarly, in a phase II clinical trial of somatostatin-based radiopeptide therapy, DOTATOC was radiolabeled with 90Y and 177Lu and administered to 34 patients who had progressive meningioma.195 In this study, stable disease was achieved in 23 patients (65.6%) with a mean overall survival (OS) of 8.6 years from recruitment. As with other forms of hormone antagonist therapy for meningioma, growth hormone receptor antagonist showed promise in the lab but does not appear to have a significant clinical effect, although the latter has not been studied in prospective controlled fashion. In 2001, McCutcheon et al demonstrated reduction of tumor growth in mice that had been xenografted with human meningioma tumors after administration of pegvisomant, a growth hormone receptor inhibitor.196 Unfortunately, a 2005 case report of a woman who had acromegaly and a skull base meningioma treated with pegvisomant demonstrated no appreciable inhibitory effect on the tumor. Rather, over a period of 5 years of surveillance, the tumor grew in volume to nine times its original size.197

36.6.5 Growth Factor Receptor Inhibitors

Meningiomas express high levels of growth factor receptors.198 ^,​ 199 Unfortunately, trials testing EGFR inhibitors such as erlotinib and gefitinib200 as well as imatinib, a PDGFR inhibitor,201 have failed to demonstrate significant growth inhibition of recurrent meningioma. Antiangiogenic drugs, including sunitinib202 ^,​ 203 and bevacizumab,204 ^,​ 205 ^,​ 206 have been explored as well in case reports, retrospective series, and a few phase II trials, suggesting a possible antitumoral effect in receptor positive tumors, although randomized clinical trials await.

36.6.6 Immunotherapy

Interferon-α is cytokine that has been used as an effective treatment for multiple cancers, including hairy cell leukemia,207 chronic myelogenous leukemia,208 renal cell carcinoma,209 and melanoma.210 Initial lab studies examining the effect of interferon-α on meningiomas demonstrated an inhibitory growth211 and antiangiogenic effect,212 prompting a few clinical studies, including of combination interferon-α and 5-fluorouracil213 and of interferon-α as a solo treatment.214 Although these series demonstrated effective prolongation of time to recurrence in a small group of patients who had aggressive, recurrent meningioma, interferon-α frequently induces a flulike state that is very poorly tolerated.

Immune checkpoint inhibitors that boost the immune response to tumors have received a lot of attention for their successful use in melanoma and other tumors. Early work in this area has demonstrated an elevated number of programmed death ligand 1 (PD-L1) tumor cells in higher-grade meningiomas, suggesting a possible role for immune checkpoint inhibitors in these tumors.215 Accordingly, clinical studies evaluating PD-L1 inhibitors pembrolizumab and nivolumab are under way.

36.6.7 Targeted Molecular Therapy

In addition to immunotherapy, therapeutics that target specific genetic mutations may hold promise as a medical alternative for recurrent, aggressive meningioma. Select mutations, including SMO and AKT1, provide targetable sites for medical treatments that already exist. Accordingly, research is under way to test the efficacy of these medications in mutation-positive meningiomas. At present, a single case report exists demonstrating a positive antitumoral effect from targeted treatment of a patient with a multifocal, skull base, recurrent AKT1-mutant tumor treated using surgery, radiation, and two chemotherapeutic agents: somatostatin and sorafenib.216 After the fourth recurrence, the histology was upgraded to III and pulmonary metastases were identified. Genetic screening identified the AKT1 mutation, and the patient was started on an AKT1 inhibitor. At 2-year follow-up, imaging demonstrated a slight reduction in the volume of intracranial disease and stability of pulmonary disease.

36.6.8 Viral Oncolytic Therapy

Viral oncolytic therapy has been tested in the lab with herpes simplex virus, demonstrating a tumor-killing effect in xenografted mice that had meningioma217 and high-dose adenovirus, producing an oncolytic effect on meningioma cells in culture.218