Meningioma is one of the most common tumors of the central nervous system, accounting for approximately 24-30% of primary intracranial tumors, of which malignant/mesenchymal meningioma (WHO grade III) accounts for only 1,0%-2,8% [1]. Malignant meningiomas are more common in men and children and have a poor clinical outcome. Due to the evolving diagnostic criteria and the lack of specificity of the pathological and imaging features, there is no standard diagnostic and therapeutic protocol. For this reason, it is necessary to review the biological behavior, pathologic and imaging features, and current status of diagnosis and treatment of malignant meningiomas to guide clinical diagnosis and treatment. The evolution of the definition In 1938, Cushing and Eisenhardt first proposed a malignant subtype of meningioma, and since then, the World Health Organization (WHO) has revised the classification of meningioma several times. 1979, WHO classification indicated that high-grade (WHO grade II/III) meningioma is defined as those with mesenchymal changes, but not yet developed into sarcoma. In 1993, the WHO classification was clearly subjective, making it difficult to clearly distinguish between grade II and III meningiomas. 2000 WHO classification defined malignant (WHO grade III) meningiomas as having 20 nuclear divisions per 10 high-powered views and/or cells with structures resembling carcinoma, sarcoma, or melanoma. 2007 WHO classification reintroduced brain infiltration into the classification. The WHO classification in 2007 reintroduced brain infiltration as one of the diagnostic criteria for atypical (WHO grade II) meningioma, but the diagnostic criteria for mesenchymal/malignant (WHO grade III) meningioma remained unchanged. 2. molecular genetic features and etiological studies 2. 1. gene mutations associated with malignant progression of meningiomas Mutations in the monosomy of chromosome 22 are the most common genetic variants in meningiomas. about half of meningiomas have loss of chromosome 22q12, and NF2 is the major gene in this region. the NF2 gene product is a member of the 4. 1 structural protein family, the latter of which is likely to be a tumor suppressor, and its inactivation may be key to meningiogenesis. Given the similar proportion of NF2 gene mutations in different grades of meningioma, it is believed that this gene is primarily associated with the development of meningioma and is less relevant to the malignant progression of meningioma [2]. Loss of chromosomes 1p, 6q, 10, 14q and 18q and acquisition of 1q, 9q, 12q, 17q and 20q are common in atypical and mesenchymal meningiomas, while acquisition or amplification of 6q, 10, 14q and 17q23 is common in mesenchymal meningiomas, and loss of 9p can also occur [2]. 9p loss can cause loss of CDKN2A, ARF and CDKN2A and CDKN2B, which both encode checkpoint proteins that arrest the cell cycle in the G1-S phase, have altered genetic information. Since most mesenchymal meningiomas have concurrent deletions or mutations of these three genes, it is hypothesized that inactivation of cell cycle checkpoint proteins plays an important role in the malignant progression of meningiomas [3]. 2, 2. Signaling molecules associated with malignant transformation of meningiomas In addition to the specific gene mutations mentioned above, malignant transformation of meningiomas is also influenced by a series of signaling molecules and is often closely related to the above-mentioned gene mutations. Both atypical and mesenchymal meningiomas have alterations in the retinoblastoma protein (pRB)-dependent and p53-dependent signaling pathways, which are associated with poor regulation of CDKN2A, AFR, and CDKN2B genes. mutations in CDKN2A and CDKN2B can cause Cdk/4cyclin D complex impaired regulation of the pRB pathway under control, resulting in abnormal activation of the pRB pathway. Normally, the p53 gene can control pRB pathway activation through ARF. Due to the deletion of ARF gene in mesenchymal meningioma and its loss of contact with p53, it causes the pRB pathway to evade monitoring and appears to be continuously activated [3]. In addition, Notch signaling pathway, WNT and insulin-like growth factor pathway may also be involved in the malignant transformation of meningioma [2]. 3. pathological features 3.1 Histological typing and characteristics The two main histological types of malignant meningiomas are papillary and rhabdoid. The former is commonly seen in children, 75% can invade brain tissue or other structures, 55% develop recurrence after surgery, and 20% develop metastasis; the latter is mainly seen in middle-aged people between 40 and 50 years old, with a higher degree of malignancy and poor prognosis [2]. The main problem of both types of meningioma is local recurrence, and the latter is also the main cause of complications and death of patients. 3.2 Immunohistochemical features Although mesenchymal meningiomas have a cellular structure similar to that of carcinomas, sarcomas or melanomas, their expression of epithelial cell membrane antigens and wave proteins, lack of expression of keratins, and weak or no expression of S-100 protein often help to differentiate them from these tumors [45]. The cell proliferation marker Ki-67 is one of the commonly used markers for meningioma grading, and because it can be combined with MIB-1 antibodies, the MIB-1 labeling index (LI) can be used as an indicator for histological grading and predicting the risk of recurrence of meningioma [6]. However, some authors believe that there is a large degree of crossover in MIB-1 LI at all levels of meningioma, and because MIB-1 LI is susceptible to technical conditions and counting methods, it should not be used as a reliable indicator for the diagnosis of malignant meningioma [5]. Although 71% of meningiomas show positive expression of progesterone receptors, high-grade meningiomas are mostly negative, and therefore have limited diagnostic value for grade III meningiomas [7]. Hemangiopericytomas (HPCs) have many morphological and immunohistochemical similarities to hemangioblast meningiomas, but immunohistochemical staining for epithelial membrane antigen (EMA), CD99, bcl-2 and claudin-1 However, immunohistochemical staining for EMA, CD99, bcl-2 and claudin-1 and in situ hybridization for 1p, 14q, NF2 and 4 and 1B are useful for differentiation. In general, EMA positivity is more common in mesenchymal meningiomas, whereas CD99 and bcl-2 positivity support the diagnosis of HPCs; chromosomal deletions of 1p32, 14q32, NF2 and 4, 1B (18p11) are common in mesenchymal meningiomas, but quite rare in HPCs. In addition, there are significant differences in their biological behavior: the median survival of mesenchymal meningiomas is less than 2 years, with only occasional metastases; HPCs have a high incidence of metastases outside the CNS of 25% to 60%, but their median survival is 5 to 12 years [8]. 3.3 Nuclear schizogram characteristics The 2007 WHO classification emphasizes that the diagnosis of grade III meningioma should satisfy a number of 20 or more nuclear divisions per 10 high-powered views, but this quantification of mitoses is often influenced by differences in sampling and observation sites and the ability to accurately identify nuclear divisions [4]. The discovery of phosphorylated histone H3, a specific marker of mitosis, has made it possible to rapidly locate the most active areas of mitosis and to clearly distinguish between mitosis and apoptosis [9]. 4. Imaging features Malignant meningiomas have basic imaging features common to all levels of meningiomas, namely well-defined, dura-based, and markedly enhancing extra-axial occupancies. Intratumoral cystic changes, adjacent bone proliferation, bone destruction, tumor progression outward through the skull base, encapsulated arteries and peritumoral edema have been reported to help distinguish benign from high-grade (grade II/III) meningiomas, with intratumoral cystic changes and tumor progression outward through the skull base foramen being the characteristic features of high-grade meningiomas [10]. Some investigators have also concluded that the standard ADC ratio of MRI diffusion-weighted images is highly accurate for the differentiation of high-grade and benign meningiomas [11]. Regarding the application of new techniques in MRI examination, early studies found increased choline compound/creatine and phosphocreatine ratios (Cho / Cr) in high-grade meningiomas on magnetic resonance spectroscopy (MRS) imaging, and the presence of lactate and/or methylene signals also suggested meningioma malignancy [12]; later, it was found that both typical and atypical meningiomas had Cho and N acetylaspartate high expression and Cr low or no expression in both typical and atypical meningiomas, thus MRS has limited value in the diagnosis of grade III meningiomas [13]. Perfusion MRI can show the blood supply to the tumor and can identify benign and malignant meningiomas by measuring the maximum local cerebral blood volume (regional cerebral blood volume, rCBV) within the tumor and in the peritumoral edema area [14]. Since the rCBV of meningioma is significantly and positively correlated with the Ki-67 index, the Ki-67 index of malignant meningioma can also be indirectly understood by measuring the maximum rCBV to predict the risk of recurrence [15]. Because diffusion-tensor imaging (DTI) can show the magnitude and direction of water molecule diffusion, and the movement of intratumoral water molecules in typical meningioma is not as regular as in atypical meningioma, the value of DTI for grading meningioma is superior to diffusion-weighted MRI [16]. There is no standardized treatment plan for malignant meningioma, but it is believed that surgery followed by external beam radiation therapy (EBRT) or stereotactic radiosurgery (SRS) can prolong the survival of patients. The preferred approach for recurrent malignant meningioma is reoperation, but reirradiation or adjuvant chemotherapy is of little relevance. Preoperative embolization can reduce tumor size and intraoperative blood loss. Given that the blood supply of malignant meningioma is comparable to that of benign meningioma, the need for preoperative embolization should be determined by the size and location of the meningioma, the blood supply, and the surgeon’s personal experience. 5, 2, Surgery and postoperative radiation therapy It has been found that postoperative adjuvant radiotherapy for high-grade meningioma significantly prolongs patient survival, and the dose of radiotherapy is an independent predictor of prognosis, so it is recommended that a 50 Gy dose of radiotherapy should be administered early after surgery for malignant meningioma [17].Dziuk et al [18] concluded that the best option to improve the local control rate of malignant meningioma is sarco-total resection （Mattozo et al [19] concluded that stereotactic radiosurgery (SRS) was the best option for “aggressive” grade I and II meningiomas. ” grade I and II meningiomas for local control, but it is not effective for grade III meningiomas. However, a study by Rosenberg et al [20] demonstrated the effectiveness of SRS for local control of recurrent malignant meningiomas. The largest group of retrospective studies of grade III meningiomas to date (63 cases) showed recurrence-free survival rates of 80%, 57%, and 40% at 2, 5, and 10 years after initial surgery, and overall survival rates of 82%, 61%, and 40%, respectively; reoperation for recurrent grade III meningiomas significantly prolonged survival, whereas secondary surgery followed by radiotherapy was of little significance. Given the significantly longer overall survival with near-total resection + EBRT compared to GTR + EBRT, the authors concluded that it would be detrimental to overly pursue sarco-total resection in some patients instead [21]. 5.3 Proton therapy The advantages of proton therapy are that it allows an increase in the total dose and maximum coverage of the target range independent of tumor size or morphology. Hug et al [22] reported that combined proton and photon radiotherapy increased the 5-year local control rate of atypical and malignant meningiomas to 80%, compared to 17% for photon radiotherapy alone; target area doses >60 Gy/cobalt gray-equivalent (= cobalt – Boskos et al [23] also reported a mean recurrence-free survival of 23 months for those who received combination radiotherapy with a total dose of >60 Gy, and increasing the dose to >65 Gy could further extend the local control of malignant meningioma. If the dose is increased to 65 Gy or more, the survival of patients can be further prolonged. 5.4 Chemotherapy With the exception of oral hydroxyurea, which has been reported to be effective in controlling recurrence of recurrent malignant meningioma after GTR, conventional chemotherapeutic agents that have shown definite efficacy in other sites of tumor have failed to control malignant meningioma growth, and recombinant interferon α-2b has had little effect, while temozolomide and irinotecan have been largely ineffective [24]. Octreotide imaging showed that recurrent meningiomas overexpress growth inhibitor receptors, but only one of the five patients with grade III meningiomas treated with growth inhibitor analogs achieved partial remission and one was stable, while the remaining three cases progressed to varying degrees [25]. In view of the difficulty of controlling the growth of malignant meningiomas with conventional chemotherapeutic drugs, research has been conducted on drugs targeting novel target molecules such as growth factors and their receptors. However, the platelet-derived growth factor inhibitor imatinib (Gleevec) and epidermal growth factor receptor-related cell cycle inhibitors, apoptosis regulators, angiogenesis inhibitors, and hormone receptors have failed to effectively inhibit tumor growth [24]. 6. Summary and outlook Malignant meningioma is a relatively rare, poorly understood, and difficult to treat class of disease in clinical practice. Although distant metastases are uncommon [26], they often have a poor prognosis due to local recurrence. The initial consensus in treatment is a course of fractionated radiotherapy after GTR. Recurrence can be treated surgically again, but the effect of re-radiotherapy is unclear and there is no really effective chemotherapy regimen. It is believed that with the continuous development of neuroimaging technology (especially molecular imaging technology) and the discovery of specific markers for malignant meningioma, its diagnosis will become more accurate, convenient and standardized. Coupled with our deepening understanding of the molecular mechanisms of malignant meningioma development, it is possible to discover new targets that play a key regulatory role in its evolution, thus bringing new hope for specific treatment.