Brief CommunicationCirculating tumor DNA is present in the most aggressive meningiomas

This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. © The Author(s) 2020. Published by Oxford University Press, the Society for Neuro-Oncology and the European Association of Neuro-Oncology. Circulating tumor DNA is present in the most aggressive meningiomas


Recent discoveries of multiple driver mutations open promising perspectives for targeted therapies in meningioma.
Nevertheless, iterative recurrences of most aggressive meningiomas as extended skull base meningiomas are not systematically operated and histologically documented. This suggests the interest and the relevance of liquid biopsy in meningiomas. In a proof-of-concept study, we detected the NF2 mutation in 2 of 6 cell-free plasma DNAs and 1 of 1 cerebrospinal fluid (CSF) from high-grade recurrent cases, suggesting that identification of the driver mutation in blood and CSF is today feasible. Liquid biopsy could be an interesting tool to adapt the targeted therapy in meningiomas in the near future.
Since the discovery of NF2's involvement, many advances have been performed in the description of meningioma mutational landscape. 1 The therapeutic management of refractory meningioma to iterative surgeries and radiotherapy sessions remains an unmet medical need in neurooncology. Targeted therapies were suggested to be of interest in aggressive meningiomas, particularly for skull base ones given the large pattern of newly discovered driver mutations, which could be targeted by specific inhibitors. Recently, a case of AKT1mutated skull base meningioma treated by AKT1 inhibitor was reported to be successful. 2 PI3K-Akt-mTOR pathway targeting was also demonstrated to be relevant in meningiomas. 3 Although tumor tissue is likely available in recurrent highgrade meningiomas, iterative tumor recurrences are not systematically operated and then often non-histologically documented. Recent studies demonstrated that the mutational pattern of meningioma recurrence could differ from the initial tumor to remote recurrence. 1 Moreover, complex skull base meningiomas are often nonaccessible to surgery, treated without histology, and yet could benefit from targeted therapies. Therefore, in these different situations, liquid biopsy could be of interest in the patient diagnosis, prognosis, and to guide therapeutic management.
The detection of tumor mutation in cell-free plasma DNA (cfDNA) constitutes the most promising biomarker in cancer. In meningioma, in contrast to other brain tumors, the lack of brainblood barrier makes possible the passage of circulating tumor DNA (ctDNA) in the blood. As an alternative solution tumor mutations could be searched in the CSF. In a proof-of-concept study, we analyzed cfDNA in blood from 15 patients and in CSF from 3 other patients, all previously identified with at least one pathogenic variant in their tumor, except one (M17, Table 1).
Eleven patients presented with benign meningioma (WHO grade I and II) and 7 with recurrent high-grade meningiomas were included. Written informed consent was required for each patient. The study was approved by the Aix-Marseille University IRB.
In all cases, the blood sample was taken before surgery. The CSF has been collected during surgery for thoracic benign meningioma (M10) and during ventriculoperitoneal shunt implantation in 2 cases (M11, benign and M17, aggressive). cfDNAs were extracted from 4 mL of serum and from 1.5 to 4 mL of CSF, all previously identified with at least one pathogenic variant in their tumor, except one (M17,  The mutated allelic fraction is the count of mutated alleles out of the total number of alleles (wild type + mutated).
c Coverage of depth: number of times the nucleotide seat of the mutation in the tissue has been read by sequencing. d No mutation detected in the tumor biopsy.

Graillon et al. Circulating tumor DNA in meningiomas
Neuro-Oncology Advances is suitable for rare event detection because random molecular barcodes were incorporated to reduce the error rate and reach a sensitivity of mutation detection less than 0.1%. 4 The bioinformatics analysis was carried out using smCounter2. 5 The previously identified variants in tumor tissue were searched in the VCF files and in BAM files. In BAM files, we eliminated the background noise by seeing if the variant was present in the other samples. If it is present, we considered this variant as true only if its coverage of depth was more than the mean + 2 standard deviations of the other samples.
Among the 11 benign cases (9 cfDNAs and 2 CSF samples), we did not detect any of the previously identified mutations in the tumor (Table 1). Among the 7 aggressive cases (6 cfDNAs and 1 CSF sample), except patient M17, all the patients had a tumor biopsy with an NF2 mutation. In 2 of 6 cases (M16 and M18), the NF2 mutation was detected in cfDNA. In these 2 cases, a subcutaneous extension was observed as shown on MRI ( Figure 1A and B). Moreover, an NF2 mutation at a high allelic frequency (28.7%) was detected in the CSF from a patient bearing a very aggressive tumor (M17). This NF2 mutation was not present in a tumor biopsy (cfDNA was not available). Interestingly, the patient underwent radiation therapy for cavernous sinus meningioma without tumor biopsy in 1999. She was referred for trigeminal neuralgia impairment in 2016. The MRI displayed slight tumor growth ( Figure 1C and D). Re-irradiation was contra-indicated. One year later, she was referred for cognitive troubles and severe visual loss. The MRI displayed a dramatic tumor progression ( Figure 1E-H). An endoscopic transsphenoidal biopsy and internal ventriculoperitoneal shunt implantation with CSF analysis were performed. She deceased many weeks later. Neuropathological conclusions were an atypical meningioma with meningothelial and secretory features.
It was the first study identifying ctDNA mutation driver from meningioma patients. A previous study analyzed cfDNA from 34 meningiomas. 6 In this study, several variants have been identified, but no gene was a driver for meningioma. Moreover, no correlation was made with the tumor; therefore, the tumor origin of the identified variants cannot be asserted. 7 In the patient M17, even with a good quality of biopsy material, no NF2 genetic alteration was detected but the issue could be the area of tumor biopsy in this huge tumor. The lack of NF2 mutation is frequent in skull base meningiomas with meningothelial and secretory features. Two hypotheses may be done to explain the NF2 mutation detected in CSF: (1) a second genetic occurrence with an NF2 mutation leading to extremely aggressive meningioma or (2) a second meningioma occurrence. However, the second hypothesis is not in accordance with the MRI showing a clear tumor progression ( Figure 1).
In benign meningiomas, mutation detection in circulating DNA remains negative suggesting a low level of ctDNA in blood and CSF. In the aggressive tumors, mutation drivers were detectable in cfDNA from 2 of 6 patients and in CSF from 1 of 1 patient. Recurrent high-grade meningiomas represent a low part of meningiomas, but their therapeutic management remains particularly challenging today. The role of the subcutaneous invasion in ctDNA detection remains unknown and requires further studies. It has been shown that patients with tumors limited to the central nervous system have significantly enriched ctDNA in CSF. 8 Interestingly, we were able to detect in CSF an NF2 mutation not found in the tumor, a feature already reported in brain tumor, 8 which may be due to the tumor heterogeneity. 9 Further studies are required to assess the sensitivity and the interest of cfDNA analysis to address some limitations of tissue-based genetics. 10 Circulating tumor DNA could be currently detected in the blood and the CSF of recurrent high-grade meningiomas. The presence of ctDNA in blood or CSF from meningioma patients could be an interesting tool in a near future in a selected population to determine the tumor mutational change in recurrent high-grade meningiomas and to adapt the targeted therapy.

Funding
The project leading to this publication has received funding from the Excellence Initiative of Aix Marseille University-A*Midex-a French "Investissement d'Avenir" program.