Introduction
Meningiomas are the most common primary neoplasm of the central nervous system [
42]. Traditionally thought to derive from arachnoid cap cells, the more common meningiomas are more likely to derive from the arachnoid barrier cells or dural border cells [
21]. While most of these tumors have a benign course and a relatively good prognosis, a small subset, approximately 1–3%, become malignant and invade local intracranial structures [
13,
18]. Furthermore, in less than 1% of patients these tumors metastasize to distant extracranial locations [
41]. Extracranial manifestation of malignant meningiomas was first reported in 1886 [
36], and since then there have been reports describing varied presentations of this rare phenomenon.
Patients with metastatic meningioma (MM) often have a poor prognosis. Due to their rarity, there is a dearth of effective screening and management protocols for these patients [
16]. Improved and more targeted detection of these cases is needed, as many patients can go several decades before the metastatic disease is found, often on incidental imaging, by which point the care is palliative [
10]. Crucially, no effective treatments exist for these patients.
To add to the current clinical knowledge of metastatic meningioma, we present a case series of six patients from a single center in the UK. We describe the presentation of these patients as well as a reflection on the management strategy adopted for each case. We then systematically review the existing literature to summarize the findings of all published cases of this rare metastatic tumor. We then explore the relationships between the primary tumor and the metastatic lesions in a descriptive analysis. In an attempt to expand the global relevance of this topic, we stratify the cases across countries and economic classifications.
Methods
Case series
Between January 2002 and January 2022, six patients presented to our institution with metastatic meningioma. Clinical information was collected from medical records (Table
1) [
30]. Only adult patients (≥ 18 years) were included. Extracranial metastases were confirmed on histopathology performed by an experienced neuropathologist. We did not include metastases within the cranium, “drop” metastases to the spinal cord, or those due to local or iatrogenic spread (i.e., spread to the tissues overlying the incision site from the surgery). None of our patients had a diagnosis of genetically inherited syndromes such as neurofibromatosis type 2 (NF2). The use of clinical information was approved by the Oxford University Hospitals Foundation Trust Clinical Audit Committee.
Table 1
Case series of metastatic meningioma cases from a single center
59 F | Anaplastic | 3 | 7 | Right occipital | Liver and peritoneum | Increased signal in colon, breast, thyroid, and T7 | Resections and radiotherapy | Radiotherapy and surveillance MRI, 500 mg of dexamethasone |
70 F | Atypical | 2 | 14 | Right frontal | Parotid | Excised with right superficial parotidectomy | Resections and radiotherapy | Died 20 years after first surgery |
63 M | Anaplastic | 3 | 1 | Parasagittal | Lung | Single nodule | Resections and radiotherapy | No further treatment options and referred to palliative care |
47 M | Anaplastic | 3 | 2 | Left convexity | Bone (extracranial) | C7 with C7/8 compression, T2, T5, T9, right sacrum, acetabulum, femur, sternum, 5th rib | Resections and radiotherapy | Spinal radiotherapy |
46 M | Atypical | 2 | 3 | Middle cranial fossa | Liver, lung, kidney, long bone, spine | T5 vertebral extending into canal | Resections and radiotherapy | Died 3 years and 2 months after first surgery |
59 M * | Atypical | 2 | 2 | Falx/SSS junction | Skin, pleura, liver, bone | Right iliac bone | Resections and radiotherapy | Died 3 years and 1 month after first surgery |
Systematic review
VH and RJB performed a search on the MEDLINE database with the following MeSH terms: “metastatic meningioma,” “meningioma metastasis,” and “meningioma metastases.” Forward and backward citation searching was used to capture studies that were not found in our original search. A full list of articles found through this search was pooled together. A first round of selection was then completed independently by VH and RJB by reviewing the title and abstract of the listed studies to ensure they capture the key terms listed above and a full text was accessible. Next, a full-text review was independently reviewed and assessed for inclusion or exclusion. If there was discordance in the decision, the study was reviewed by VH and RJB together with DSJ, and a consensus was reached. Data extraction was completed using a common data template that was constantly reviewed. Studies were excluded if they pertained to the following: tumor-to-tumor metastasis within the primary meningioma, ectopic primaries, “drop” metastases to the spinal cord, or iatrogenic “metastases” (i.e., spread directly to the tissues overlying the incision site from the surgery). Only studies on adult patients were included. Studies were also excluded if they described patients who were found to have older (now-obsolete) classifications on histology. This was of particular importance in earlier descriptions of these patients where the distinction between meningiomas and hemangiopericytomas as separate tumor diagnoses was not yet well established, and descriptors like angioblastic, fibroblastic, and endotheliomatous (in the absence of a CNS WHO grade) cannot be interpreted from the older literature. Only full texts were included; if one could not be obtained through the University of Oxford Library Services or by subsequent request from international libraries and repositories, it was excluded. Papers in languages other than English were translated and, where possible, confirmed with a native speaker. We only included studies published after the release of the publication of the first CNS WHO tumor grading system in 1979. The search was conducted using the PRISMA guidelines [
33]. A summary of the studies included is provided in Supp. Table
1.
Descriptive analysis
To summarize features reported across studies, we performed an analysis specifically focusing on the following features: patient age, sex, location, and CNS WHO grade of the primary intracranial meningioma, time to first metastatic lesion (TTM), and the MM location. We categorized the location of the tumors into those at the skull base (e.g., anterior, middle, and posterior fossa), convexity (e.g., frontal, parietal), falx/tentorium, and others (e.g., intraventricular).
Statistics
Normally distributed data were reported as mean with standard error of the mean (SEM), while non-normally distributed data were reported as median with interquartile range (IQR). Parametric and non-parametric statistical tests were used to assess the difference between variables. Significance was defined as p = 0.05 to test multiple comparisons (with Dunn’s correction). All plots and statistical analyses were generated using custom-made Python code (available upon request).
Discussion
Since Power’s report of the first supposed recorded case of MM in 1886 [
36], the incidence of reports has been relatively low. To date there have been a few attempts to summarize the findings of the literature on this topic, initially in 1963 and most recently in 2013 [
15,
40]. However, these reviews only covered a subset of the available literature. As such, there is a need for a more recent and unified understanding of the relevant descriptive features on these tumors.
The prevalence of distant metastases among all patients with meningioma is low with the most recent reports being 0.18% [
41]. Indeed, these cases represent a relatively rare occurrence in what is otherwise a benign neoplasm that has, in most cases, favorable surgical outcomes [
4]. Nevertheless, the TTM for many of the cases we reviewed is in the order of decades, representing opportunity for diagnosis and a window for intervention (particularly given the changing surgical paradigm of more aggressive and targeted approaches). By way of example, in the case reported by Dincer and colleagues in 2020, a 41-year-old patient was found to have pleural metastases from an intracranial meningioma. A retrospective search from old CT imaging found that it was present 13 years earlier [
12]. Another layer of complexity is that only 51% of patients with systemic metastases were symptomatic at the time of detection of the metastases [
40].
We found that the convexity was the most common primary location. This finding was also the case in other studies where convexity tumors were shown to be of highest prevalence [
39,
45]. Meningiomas that are not located at the skull base have been shown to have double the risk of CNS WHO grade 2 and 3 pathology relative to skull base tumors [
22]. In addition, larger meningiomas are more likely to be CNS WHO grade 2 [
29]. A retrospective study of over 1600 cases found that skull base meningiomas had a relatively low malignant potential [
8].
Our finding that the lung (61%), bone (26%), and liver (19%) were the most common sites of extracranial metastasis, in that order, was similar to the findings in a systematic review by Surov et al. in 2013 [
40] with a frequency of 37.2%, 16.5%, and 9.2%, respectively. While the causes of extracranial metastases from meningioma are largely unknown, one explanation could be through iatrogenic seeding. The disruption of the tumor during resection of the primary lesion could lead to metastatic spread via the lymphatics and venous sinuses to further tissues; much in the same way as CSF is hypothesized to be the cause of “drop” intraspinal metastases [
5,
40]. Knowing the most frequent locations of metastasis will be pertinent in the development of any future targeted screening pathways.
After first biopsy, we found that 47% of the tumors were CNS WHO grade 1, 31% were grade 2, and 22% were grade 3. Surov and colleagues reported that 56.2% of metastatic cases originated from CNS WHO grade 1 or 2 tumors [
40]. We elected to designate the grade of the tumor as the one that was reported at primary biopsy. In several cases, a progression from CNS WHO grade 1 to 2 to 3 was seen in re-operations for recurrence after which a metastasis was discovered. This could explain our higher finding of grade 1 and 2 tumors (78%) relative to the previous systematic review. Alternatively, the broader range of cases in our analysis could equally explain this difference among a relatively small number of cases.
While the CNS WHO grading system is the current gold standard, up to 20% of grade 1 meningiomas can develop aggressive features [
37]. Many centers, including ours, favor examining the methylation status as an additional prognostic indicator. Grade 1 primary tumors made up the greatest proportion in the data. As the TTM in many cases was in the order of decades, it is possible that the primary tumor and its recurrence evolved into more malignant grade 2 or 3 histology before metastasizing. As such, we cannot conclude that the initial grade after primary resection impacts the TTM directly. We did find a statistically significant difference in the TTM in grade 1 and 2 primaries compared to grade 3. Moreover, due to the static nature of imaging and biopsy results, we cannot know when, or if, the malignant change and metastatic spread occurs.
Conclusion
In this study we have highlighted the unique clinical challenges in managing patients with metastatic meningioma. We reported that the convexity was the most likely intracranial primary location, that the initial CNS WHO grade at first available biopsy significantly affects the TTM, and that the lung is the most common site of metastasis. There is a lack of consensus on how to effectively manage these patients and invariably the outcomes are very poor. This dire situation is also seen in the management of recurrent intracranial WHO grade 2 and 3 meningiomas and reflects the paucity of drug therapies for these tumors. While metastatic meningiomas represent an aggressive and extreme end of the spectrum, the failure to manage them also reflects on the failure to control the primary disease. Only with advances in this area will we make progress with metastatic meningiomas. This approach will also be central to the care of these patients in poorly resourced communities around the world. The outcomes for metastatic meningioma are poor even in centers treating large numbers of meningioma patients, such as ours; there is an urgent need for multi-center clinical trials for this underestimated challenge. Only then may we hope to leverage our expanding knowledge on this rare but aggressive subset of meningiomas.
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
The authors have completed an exhaustive review of the literature on metastatic meningioma and pointed out the problem of not being able to effectively treat these metastases when they occur. A more uniform agreement on how to screen patients early on, biopsy suspected metastases when present and use molecular analyses to better select therapy, is needed.
Michael William McDermott
Miami, USA