Abstract
Objective
Both primary optic nerve sheath meningiomas (pONSMs) and secondary optic nerve sheath meningiomas (sONSMs) pose clinical challenges because standard treatments such as surgical debulking and radiation therapy can further damage the optic nerve, producing permanent visual loss. The molecular pathology of primary skull base meningiomas is becoming clearer. However, by comparison, pONSMs and sONSMs have not been studied adequately with contemporary high-throughput molecular genetic techniques, which is the primary aim of this study. This is a crucial issue because these tumors may harbor distinct genetic alterations that render them susceptible to targeted therapy, allowing for vision preservation or even visual improvement.
Methods
A total of 18 optic nerve sheath meningiomas, of which 11 were pONSMs and 7 were sONSMs, were obtained from 3 different institutions and underwent next-generation sequencing.
Results
We found that pONSMs and sONSMs harbor gene variants previously identified in other meningiomas but also distinct alterations in genes implicated in cell signaling, transcriptional regulation, and DNA damage repair.
Conclusions
These findings expand our understanding of a relatively understudied specific meningioma with unique therapeutic challenges.
Keywords: optic nerve, optic nerve sheath, meningioma, mutations, NGS, next-generation sequencing
KEY POINTS.
What is known: Optic nerve sheath meningiomas are therapeutically challenging tumors, as standard interventions such as surgery and radiation therapy risk permanent visual loss.
What is known: Compared with other skull base meningiomas, the molecular genetic landscape of primary and secondary optic nerve sheath meningiomas remains poorly characterized.
What this study adds: Using next-generation sequencing, this study demonstrates that primary and secondary optic nerve sheath meningiomas harbor both shared genetic variants seen in other meningiomas and distinct alterations involving cell signaling, transcriptional regulation, and DNA damage repair pathways.
INTRODUCTION
Meningiomas comprise a family of neoplasms showing arachnoid cap cell differentiation and account for approximately 37% of tumors arising in the central nervous system.1,2 Most meningiomas arise in the central nervous system and grow extra-axially in relation to the dura mater, with few exceptions encountered in other sites, such as the choroid plexus.3 A small proportion of meningiomas can be encountered in an ectopic setting, most frequently in the lung.4 An anatomic special site in which meningiomas can be encountered is the orbit.5 In this location, meningiomas can present as (1) primarily arising from the meningothelial optic nerve sheath, (2) extending from the intracranial compartment through the optic canal, and (3) arising in an ectopic manner within the orbital soft tissue.6 The first 2 are known as optic nerve sheath meningiomas (pONSMs) and secondary optic nerve sheath meningiomas (sONSMs) due to their intrinsic anatomic relation to the optic nerve. All types of optic nerve sheath meningiomas (ONSMs) can cause significant visual loss; moreover, those lesions are clinically challenging, given that the optic nerve can be irreversibly damaged by attempts to resect the tumors or radiotherapeutic approaches, particularly when the tumors involve the optic nerve adjacent to the globe.7 Currently, no specific chemotherapeutic or targeted treatments exist for ONSMs.
The mutational landscape underlying intracranial and spinal meningiomas has been investigated in recent years, with many molecular drivers and prognostic markers discovered. Briefly, skull-base meningiomas have been found to be enriched in AKT1, SMO, and TRAF7 alterations, in contrast to spinal/convexity meningiomas, which frequently show 22q deletion and NF2 mutations.8,9 Conversely, ONSM molecular alterations have been less well studied, with a single report of copy-number variations (CNVs) assessed with single-nucleotide polymorphism (SNP) arrays.6
In this study, we aimed to characterize the molecular alterations underlying both pONSMs and sONSMs using next-generation sequencing (NGS) in archival tissue to gain additional insights into the molecular pathogenesis of ONSMs, compared with their more common intracranial and spinal counterparts, and generate hypotheses for novel targetable molecular alterations that could aid in the clinical setting.
MATERIALS AND METHODS
Cases
A total of 18 ONSMs were obtained from the pathology files from Johns Hopkins Hospital, the University of Queensland, and Columbia University Medical Center. Meningiomas that were present on the tuberculum sellae and/or planum sphenoidale at surgery were assumed to have begun there and secondarily extended into the optic canal and, in some cases, into the orbit. Tumors that were confined to the orbit and, in some cases, to the proximal optic canal but not intracranially were considered primary. All cases included in this study were definitively classified as primary or secondary. All cases were centrally reviewed by 1 neuropathologist (FJR) and classified and graded using current criteria (World Health Organization and cIMPACT).
NGS
All cases underwent NGS with the Illumina panel at Johns Hopkins Hospital using methods previously described in the literature and as described below. In brief, areas enriched with tumor (at least 10% tumor cellularity) were scraped from sequential formalin-fixed, paraffin-embedded (FFPE) sections 5 μm thick (routinely 3-7 unstained slides per sample) and placed into 2-mL tubes for DNA extraction. Extraction and purification were accomplished with the automated Siemens Tissue Preparation System (Siemens Healthcare Diagnostics), while genomic DNA was quantified using the Qubit 2.0 Fluorometer (Life Technologies). DNA libraries were prepared using a custom targeted enrichment with capture probe technology (Integrated DNA Technologies), covering 948 genes relevant to various solid tumors. A final pool of samples was barcoded using the xGEN dual index with unique molecular identifiers and sequenced using 2 × 100 paired-end cycle chemistry on the NovaSeq 6000 platform (Illumina).
Bioinformatics pipeline
Raw binary cluster data generated by the NovaSeq 6000 instrument were demultiplexed using vendor-supplied bcl2fastq2. Raw data qualities were assessed to see if each sample achieved greater than 93% of perfect sample barcode reads and an average base quality score of 35 (Q35). Samples failing this quality control (QC) assessment were resequenced. Samples passing raw QC were carried forward with downstream bioinformatics analysis. Each sample’s sequencing data were mapped to the human genome reference (hg19) using the Burrows-Wheeler Aligner v0.7.10 algorithm with default parameters. Subsequent mapped data in BAM format (.bam) were passed through Picard Tools v1.119 Mark Duplicates with the Mate CIGAR algorithm to remove potential polymerase chain reaction artifacts. Samples with an average unique read depth of 800× were deemed QC passed and carried forward with variant calling and copy number assessment.
Variant calling was performed using an in-house variant caller algorithm (MDLVC v5.0) cross-referenced with HaplotypeCaller (Genome Analysis Tool Kit 3.3) under discovery mode in the coding regions of target genes. Variant calls with at least 50× total allele depth, strand balanced, and greater than or equal to 5% variant allele frequency were used to filter our possible germline status by cross-matching with the dbSNP database. A script was written to use the clinical table search service available from the National Institutes of Health and was used to query Catalogue of Somatic Mutations in Cancer (COSMIC), ClinVar, and dbSNP databases to determine whether the variants in the current dataset have been previously reported. Common SNPs (population allele frequency >1%) were excluded from analysis. Because matched normal (tissue or peripheral blood) was not available for comparison, missense variants with an allele frequency less than 5%, between 45% and 55%, and more than 95% were excluded to reduce the possibility of reporting germline variants. Statistical analysis was performed using GraphPad Prism 10 (GraphPad Software). Fisher exact test was used to compare the percentage of samples in the current study to the percentage of meningeal tumors in COSMIC with variants in each gene. Fisher exact test was also used to compare the percentage of pONSMs and sONSMs with variants in specific genes. Statistical significance was defined at P < .05.
Resulting variant calls were then manually inspected using Integrated Genomics Viewer v2.3.4 (Broad Institute) and annotated appropriately to reference COSMIC to identify likely somatic and ClinVar for pathogenic and likely pathogenic variant status.
Copy number assessment was performed using an in-house pipeline, TMM-CNV, that calculates normalized read coverage over gene targets relative to a stand-alone FFPE-derived pool of normal control samples (n = 21) and using a third-party software, CNVKit. Resulting CNV concordant between both CNV callers were deemed valid CNV calls.
RESULTS
Clinical samples
A total of 18 cases were included, of which 11 were pONSMs and 7 were sONSMs (Figure 1). Tumor samples were collected during an extended period, with variable storage length (between 1998 and 2011), highlighting the relative scarcity of these samples.
Figure 1.
Clinicopathologic features of optic nerve sheath meningiomas. Optic nerve sheath meningiomas can develop primarily in the nerve sheath (A, B) or present as secondary extension from intracranial primaries (C). Most tumors demonstrate features of the meningothelial subtype (D).
Clinicopathologic data of the cases obtained are presented in Table 1. Patients with pONSMs had a mean age at diagnosis of 44.4 years and a female-to-male ratio of 1.75; similarly, patients with sONSMs had a mean age of diagnosis of 48.3 years and a female-to-male ratio of 1.33. All patients had varying severities of monocular visual loss associated with an ipsilateral relative afferent pupillary defect and a swollen, pale, or normal-appearing optic disc in addition to computed tomographic or magnetic resonance imaging evidence of a lesion involving the sheath of the optic nerve. Distribution of different grades and histotypes between the 2 groups was also comparable, with pONSMs showing 10 of 11 (90.9%) grade 1 meningothelial meningiomas and a single (9.1%) grade 2 meningiothelial meningioma, and sONSMs showing 6 of 7 (85.7%) grade 1 (5 meningothelial, 1 transitional) meningiomas and 1 of 7 (14.3%) grade 2 atypical meningiomas.
Table 1.
Clinicopathologic Characteristics of the 11 Primary Optic Nerve Sheath Meningiomas (p1-p11) and 7 Secondary Optic Nerve Sheath Meningiomas (s1-s7) Included in the Analysis
| Specimen ID | Histology | Grade | Sex | Age at diagnosis, y |
|---|---|---|---|---|
| p1 | Mengiothelial | 1 | Female | 41 |
| p2 | Mengiothelial | 1 | Female | 43 |
| p3 | Mengiothelial | 1 | Female | 44 |
| p4 | Mengiothelial | 1 | Male | 15 |
| p5 | Mengiothelial | 1 | Female | 32 |
| p6 | Mengiothelial | 1 | Male | 48 |
| p7 | Mengiothelial | 1 | Male | 46 |
| p8 | Mengiothelial | 1 | Female | 60 |
| p9 | Mengiothelial | 1 | Female | 67 |
| p10 | Atypical | 2 | Male | 34 |
| p11 | Meningothelial | 1 | Female | 58 |
| s1 | Meningothelial | 1 | Female | 46 |
| s2 | Meningothelial | 1 | Female | 42 |
| s3 | Transitional | 1 | Male | 28 |
| s4 | Meningothelial | 1 | Female | 45 |
| s5 | Meningothelial | 1 | Female | 45 |
| s6 | Atypical | 2 | Male | 75 |
| s7 | Meningothelial | 1 | Male | 57 |
A total of 1803 coding variations were detected: 1258 in pONSM cases (average number per case: 114) and 545 in sONSM cases (average number per case: 78). Variation types included 808 missense (pONSM: 527; sONSM: 281), 61 truncating (pONSM: 46; sONSM: 15), 337 insertion/deletion (pONSM: 239; sONSM: 98), 571 frame-shift (pONSM: 429; sONSM: 142), and 26 splice (pONSM: 17; sONSM: 9) alterations. All data are available in the Supplementary Table.
CNVs
CNV analysis was performed successfully in 5 cases (27.8%)—3 pONSMs and 2 sONSMs. Figure 2 shows the results of CNV analysis. CCND1 gains were observed in 1 of 3 (33.3%) cases of pONSM and 1 of 2 (50%) cases of sONSM. FGFR4 gain was observed in 1 case of pONSM, and FLT4 gain was detected in 1 case of pONSM. Both cases of sONSM had chromosome 6q loss.
Figure 2.
Optic nerve sheath meningiomas: copy number variation results. Green represents a gain in copy number, and red represents a loss in copy number.
Alterations in genes previously associated with intracranial/intraspinal meningiomas
First, we examined whether ONSMs harbor variants in genes with previously reported variants in meningeal tumors. A visualization of the mutations in genes previously associated with meningiomas identified in the current study is provided in Figure 3. The 20 genes were selected based on the highest percentage of meningeal tumors in COSMIC that harbor mutations in these genes, displayed in descending order of this percentage. The number of samples in the current study that carry mutations is compared with the frequency of mutations in COSMIC in Table 2.
Figure 3.
Genes altered in optic nerve sheath meningiomas and previously reported in extraorbital meningiomas. Genes are listed in descending order of the percentage of meningeal tumors in the Catalogue of Somatic Mutations in Cancer with reported variants. Light blue = missense variations of unknown significance. Red = missense variation that is a putative driver mutation and/or likely pathogenic. Orchid = insertion/deletion variation. Green = frameshift mutation. Orange = truncating mutation. The number in each cell represents the number of the given type of variation found in that sample.
Table 2.
Percentage of Optic Nerve Sheath Meningiomas in the Current Study That Carry Variations in the 20 Most Frequently Variant Genes in Meningeal Tumors From the Catalogue of Somatic Mutations in Cancer (COSMIC)
| Gene | Samples with variants in the current study, % (No./total No.) | Meningeal tumors in COSMIC with variants, % (No./total No.) | Fisher exact test P value |
|---|---|---|---|
| LRP1B | 44 (8/18) | 39 (155/401) | .6289 |
| KMT2C | 39 (7/18) | 38 (162/431) | >.9999 |
| NF2 | 44 (8/18) | 37 (946/2581) | .4750 |
| LZTR1 | 0 (0/18) | 37 (153/418) | .0005a |
| TSC2 | 11 (2/18) | 35 (157/450) | .0417a |
| PREX2 | 22 (4/18) | 35 (141/400) | .3182 |
| APC | 33 (6/18) | 32 (149/474) | >.9999 |
| ROS1 | 17 (3/18) | 31 (135/430) | .2960 |
| SMARCA4 | 0 (0/18) | 29 (147/507) | .0053a |
| KDM6A | 11 (2/18) | 29 (130/450) | .1155 |
| MET | 0 (0/18) | 28 (145/512) | .0052a |
| NF1 | 11 (2/18) | 28 (143/517) | .1763 |
| NOTCH1 | 6 (1/18) | 28 (142/511) | .0540 |
| EGFR | 0 (0/18) | 27 (157/580) | .0054a |
| TRAF7 | 17 (3/18) | 26 (371/1431) | .5874 |
| PRKAR1A | 6 (1/18) | 16 (148/899) | .3353 |
| AKT1 | 6 (1/18) | 13 (346/2591) | .4969 |
| KLF4 | 0 (0/18) | 13 (185/1429) | .1527 |
| SMO | 6 (1/18) | 10 (153/1584) | >.9999 |
| TERT | 6 (1/18) | 10 (145/1416) | >.9999 |
Indicates statistical significance defined at P < .05.
With a few exceptions, the ONSMs queried in the current study harbor mutations in genes that have most frequently harbored mutations in previous reports on other meningeal tumors, with frequencies that are congruent. The 3 genes (LRP1B, KMT2C, and NF2) with variants reported in the highest percentage of samples in COSMIC are observed in a similar percentage of samples in the current study (P > .05, Fisher exact test for each comparison). Five of the genes (LZRT1, TSC2, SMARCA4, MET, and EGFR) have statistically significantly fewer samples with variants in the current study compared to what would be expected given the percentage of samples in COSMIC with reported variants (P < .05, Fisher exact test for each comparison).
Genes most frequently mutated in ONSMs
We next identified the genes that most frequently carried variants in ONSMs, regardless of previous association with other meningeal tumors. We used a cutoff of involvement in at least 8 (44%) of the cases, which left us with a reasonable dataset for 14 genes, with proposed functions relating to immune regulation, cell signaling, cytoskeletal organization, cell-cell adhesion, and transcriptional regulation. A visualization of these variants is provided in Figure 4. The frequency of variants in the current study is compared with what is reported in meningiomas in COSMIC in Table 3.
Figure 4.
Genes altered in the highest number of samples of optic nerve sheath meningiomas. Genes are listed in descending order of the number of samples affected. Light blue = missense variation of unknown significance. Red = missense variation that is a putative driver mutation and/or likely pathogenic. Orchid = insertion/deletion variation. Green = frameshift mutation. Orange = truncating mutation. The number in each cell represents the number of the given type of variation found in that sample.
Table 3.
Percentage of Optic Nerve Sheath Meningiomas (ONSMs) in the Current Study and in Meningeal Tumors in the Catalogue of Somatic Mutations in Cancer (COSMIC) That Carry Variations in the 14 Genes With Variants in the Highest Number of ONSMs
| Gene | Samples with variants in the current study, % (No./total No.) | Meningeal tumors in COSMIC with variants, % (No./total No.) | Fisher exact test P value |
|---|---|---|---|
| AR | 61 (11/18) | 1 (4/285) | <.0001a |
| HLA-DRB1 | 56 (10/18) | 2 (5/255) | <.0001a |
| SYNE1 | 56 (10/18) | 4 (10/255) | <.0001a |
| ZNF384 | 56 (10/18) | 0 (0/255) | <.0001a |
| RIMS2 | 50 (9/18) | 5 (12/255) | <.0001a |
| ATN1 | 44 (8/18) | 0 (1/255) | <.0001a |
| COL6A3 | 44 (8/18) | 2 (4/255) | <.0001a |
| FAT3 | 44 (8/18) | 3 (8/255) | <.0001a |
| LRP1B | 44 (8/18) | 39 (155/401) | .6289 |
| MNX1 | 44 (8/18) | 0 (0/255) | <.0001a |
| NF2 | 44 (8/18) | 37 (946/2581) | .4750 |
| TET1 | 44 (8/18) | 0 (1/285) | <.0001a |
| TET2 | 44 (8/18) | 0 (1/286) | <.0001a |
| ZFHX3 | 44 (8/18) | 3 (8/285) | <.0001a |
Indicates statistical significance defined at P < .05.
Twelve of the 14 genes (AR, HLA-DRB1, SYNE1, ZNF384, RIMS2, ATN1, COL6A3, FAT3, MNX1, TET1, TET2, and ZFHX3) had variants observed in a higher percentage of samples in the current study compared to the percentage of meningeal tumors in COSMIC with reported variants (P < .05, Fisher exact test for each comparison). Two of these 12 genes (ZNF384 and MNX1) do not have any variants reported in COSMIC for meningeal tumors. The 2 genes of the 14 that have variants observed in a similar proportion of samples in the current study and in COSMIC (P > .05, Fisher exact test for each comparison) are LRP1B and NF2, as mentioned in the previous section.
Genes differentially altered in either pONSMs or sONSMs
Finally, we identified genes with variants exclusively in either pONSMs or sONSMs to help distinguish the molecular signatures of these tumor subtypes. We used a cutoff of involvement in at least 4 (36%) of the pONSMs or 2 (29%) of the sONSMs, which left us with a reasonable dataset for 11 genes: 6 that had variants exclusively in pONSMs and 5 that had variants exclusively in sONSMs. These genes have a variety of proposed functions, including modulating the immune response, regulating transcription, and contributing to DNA damage response. A visualization of these variants is provided in Figure 5. The frequency of variants in pONSMs and sONSMs is compared with what is reported in meningiomas in COSMIC in Table 4.
Figure 5.
Genes altered exclusively in primary optic nerve sheath meningiomas or exclusively in secondary optic nerve sheath meningiomas. Light blue = missense variant of unknown significance. Red = missense variant that is a putative driver mutation and/or likely pathogenic. Orchid = insertion/deletion mutation. Green = frameshift mutation. Orange = truncating mutation. The number in each cell represents the number of the given type of mutation found in that sample.
Table 4.
Percentage of Primary Optic Nerve Sheath Meningiomas (pONSMs), Secondary Optic Nerve Sheath Meningiomas (sONSMs) in the Current Study, and in Meningeal Tumors in the Catalogue of Somatic Mutations in Cancer (COSMIC) That Carry Alterations in 11 Genes That Have Mutations in Either pONSMs or sONSMs but Not Both
| Gene | pONSMs with variants, % (No./total No.) | sONSMs with variants, % (No./total No.) | Meningeal tumors in COSMIC with variants, % (No./total No.) | Fisher exact test P value (pONSM vs sONSM) | Fisher exact test P value (pONSM or sONSM vs COSMIC) |
|---|---|---|---|---|---|
| MAP3K1 | 55 (6/11) | 0 (0/7) | 16 (68/431) | .0377a | .0042a |
| NCOR1 | 45 (5/11) | 0 (0/7) | 2 (7/286) | .1013 | <.0001a |
| ABCC9 | 36 (4/11) | 0 (0/7) | 1 (2/255) | .1193 | <.0001a |
| CBLB | 36 (4/11) | 0 (0/7) | 1 (3/255) | .1193 | <.0001a |
| RUNX1T1 | 36 (4/11) | 0 (0/7) | 1 (2/255) | .1193 | <.0001a |
| UBR5 | 36 (4/11) | 0 (0/7) | 4 (9/255) | .1193 | .001a |
| CRTC1 | 0 (0/11) | 29 (2/7) | 2 (4/255) | .1373 | .0087a |
| ELOC | 0 (0/11) | 29 (2/7) | 0 (1/285) | .1373 | .0015a |
| FANCL | 0 (0/11) | 29 (2/7) | 1 (3/255) | .1373 | .0059a |
| NCSTN | 0 (0/11) | 29 (2/7) | 0 (0/255) | .1373 | .0006a |
| NLRP1 | 0 (0/11) | 29 (2/7) | 1 (2/255) | .1373 | .0036a |
Indicates statistical significance defined at P < .05.
One gene (MAP3K1) had variants that were statistically significantly enriched in pONSMs compared to sONSMs (P < .05, Fisher exact test). None of the genes that had variants exclusively in sONSMs reached statistical significance when compared to the percentage of samples with observed variants in pONSMs (P > .05, Fisher exact test for each comparison). However, each of the 11 genes (MAP3K1, NCOR1, ABCC9, CBLB, RUNX1T1, UBR5, CRTC1, ELOC, FANCL, NCSTN, and NLRP1) had variants observed more frequently in our study samples compared to meningeal tumors in COSMIC (P < .05, Fisher exact test for each comparison).
DISCUSSION
Based on our cohort of 18 patients, ONSMs tend to present at a younger age than primary intracranial or intraspinal meningiomas1 (median age of 45 vs 66 years) but otherwise share similar clinicopathologic features, including tumor grade. The purpose of this study was to examine whether any molecular alterations may distinguish these tumors. Genetic alterations in ONSMs have previously been investigated using SNP-based array analysis, which demonstrated different patterns of chromosome 22 loss among pONSMs, sONSMs, and ectopic orbital meningiomas.6 Unfortunately, while we lack the CNV data regarding chromosome 22 in our samples due to sample quality issues, we confirmed chromosome 6q loss in our sONSM samples. To our knowledge, dedicated NGS analysis of orbital meningiomas has not been adequately reported in the literature.
In the current study, 14 of the 20 genes with variants most frequently reported in meningeal tumors had variants observed in a similar proportion of our samples. However, we report 5 genes (LZTR1, TSC2, SMARCA4, MET, and EGFR) that have variants in a statistically lower proportion of our samples than what is expected from previous reports and 23 genes (AR, HLA-DRB1, SYNE1, ZNF384, RIMS2, ATN1, COL6A3, FAT3, MNX1, TET1, TET2, ZFHX3, MAP3K1, NCOR1, ABCC9, CBLB, RUNX1T1, UBR5, CRTC1, ELOC, FANCL, NCSTN, and NLRP1) that have variants in a statistically higher number of our samples than what is expected from previous reports. This suggests that ONSMs harbor distinct genetic alterations compared to skull base or intraspinal meningiomas. Despite a small sample size, 1 gene (MAP3K1) showed a variant that was statistically enriched in pONSMs compared to sONMSs, suggesting that these subtypes may have distinct molecular signatures.
The potential prognostic and therapeutic implications from the distinct molecular profiles of these rare neoplasms may not be immediately apparent, but this study is a first step in identifying a candidate list of potential biomarkers and/or drug targets. For example, LZRT1, TSC2, SMARCA4, MET, and NOTCH1 mutations have been reported to be associated with poor prognosis and/or drug resistance in other neoplasms.10‐14 Because significantly fewer of our samples carried variants in these genes than reported in other meningiomas, this may indicate that ONSMs are more amenable to certain medical therapies than their intracranial and intraspinal counterparts. On the other hand, TET1 and TET2 mutations that are relatively unique to ONSMs compared to other meningiomas offer epigenetic therapy as a potential avenue for further research.15
This study has several limitations that need to be considered. First and foremost, suboptimal quality of the specimens analyzed was observed during NGS analysis. This was, to an extent, expected, due to the nature of the materials used (ie, archival FFPE samples). This was necessary given the relative rarity of available specimens at this specific anatomic site compared with tissue from intracranial/intraspinal locations. This mostly affected the CNV analysis, which was successful in 5 of 19 (26%) cases; in this scenario, our data from CNV analysis should be interpreted with caution. We were also unable to perform fluorescence in situ hybridization on these samples and thus cannot determine which of these cases would be classified as NF2-altered meningiomas or fit into other molecularly defined categories. Additionally, in individual cases, multiple sequence variants were identified in some genes that certainly could be secondary to tissue quality, particularly given the high frequency of frameshift variants in some cases. However, we did our best to exclude tumors with severe DNA degradation. We believe that this study is an important first step in outlining the genetic coding variations in ONSMs, which will be progressively clarified through multi-institutional collaborations and increased awareness of variables affecting contemporary tumor banking.
Second, our sample size is somewhat limited (11 pONSMs and 7 sONSMs) because modern neuroimaging has generally eliminated the need to biopsy lesions involving the optic nerve sheath, and as noted above, surgery to remove these lesions largely has been abandoned in favor of radiation therapy.16,17 Finally, we were unable to use matched normal tissue, so it is highly probable that some of the variants we report are germline rather than somatic, despite our attempts to exclude this possibility based on evaluation of public databases and exclusion of certain variant allele frequency (VAF) ranges.
While we successfully identified a candidate list of genes carrying variants in ONSMs that distinguish these rare tumors from intracranial and intraspinal meningiomas, we cannot entirely exclude the presence of technical or methodological artifacts, despite our best attempts to minimize these possibilities. Differences in cohorts, sequencing platforms, and analytical pipelines may also contribute to these discrepancies.
Methylation profiling is an emerging tool for intracranial meningioma subclassification, grading, and prognostication.18‐21 Regarding subclassification, all our samples were successfully diagnosed using histopathologic criteria; the same applies to grading. One of the samples (s7), which was classified as grade 1, carries a variant of unknown significance in the TERT gene. While oncogenic mutations in the TERT promoter would classify this tumor as grade 3 according to the cIMPACT recommendations,22 it is less clear in this case of a variant of unknown significance. Moreover, our NGS assessed a large number of genes for which alterations possess an intrinsic prognostic and/or predictive significance, as discussed. Methylation profiling, in this setting, could represent an additional step to take to further investigate the biological characteristics of our cases, and a comparison between NGS and methylation profiling data would be useful for identifying the most practical and cost-effective method to use in routine practice.
In summary, our study provides an initial overview of the somatic genetic mutational landscape of meningiomas occurring in the orbit, a specifically challenging site from the therapeutic standpoint. Future studies should continue to clarify the biologic and clinical significance of our findings and hopefully provide opportunities for targeted therapies for the optimal benefit of these patients.
Supplementary Material
Contributor Information
Samuel W Hulbert, University of California, Los Angeles, CA, United States; UCSF Medical Center at Parnassus, San Francisco, CA, United States.
Gianluca Lopez, University of California, Los Angeles, CA, United States; Division of Pathology, Fondazione IRCCS Ca’ Granda—Ospedale Maggiore Policlinico, Milan, Italy; Department of Biomedical, Surgical and Dental Sciences, University of Milan, Italy.
Liam Chen, University of Minnesota, Minneapolis, MN, United States.
Suping Chen, School of Medicine, Johns Hopkins University, Baltimore, MD, United States.
Aparna Pallavajjala, School of Medicine, Johns Hopkins University, Baltimore, MD, United States.
Lisa Haley, School of Medicine, Johns Hopkins University, Baltimore, MD, United States.
Aaron J Z Nelson, Duke Health, Durham, NC, United States.
Arpita Maniar, Duke Eye Center, Durham, NC, United States.
Michael Kazim, Department of Ophthalmology and Surgery, Columbia University Irving Medical Center, New York, NY, United States.
Timothy Sullivan, The University of Queensland, Pathology Queensland, Brisbane, Australia.
Charles G Eberhart, School of Medicine, Johns Hopkins University, Baltimore, MD, United States.
Fausto J Rodriguez, University of California, Los Angeles, CA, United States.
Neil R Miller, School of Medicine, Johns Hopkins University, Baltimore, MD, United States.
Author contributions
S.W.H. and G.L. contributed equally to data analysis and interpretation and wrote the first draft of the manuscript; L.C., S.C., L.H., A.P., A.J.Z.N., A.M., M.K., and T.S. contributed to data collection; C.G.E., F.J.R., and N.R.M. contributed to study design, data curation, and supervision; all authors reviewed and approved the final draft of the manuscript.
Supplementary material
Supplementary material is available at American Journal of Clinical Pathology online.
Conflicts of interest
None declared.
Funding
This study was supported in part by National Institutes of Health grant P30 CA006973 to the Sidney Kimmel Comprehensive Cancer Center (PI: W. Nelson) (F.J.R., C.G.E., N.M.) and the UCLA SPORE in Brain Cancer (P50CA211015) (F.J.R).
Data Availability
All data pertinent to the study are presented in the main text, figures, tables, and supplementary material. Additional details are available per written request.
Ethical approval/consent to participate
The study was performed under institutional review board approvals at the participating institutions with patients’ consent or waiver of consent as appropriate.
REFERENCES
- 1. Ostrom QT, Cioffi G, Gittleman H, et al. CBTRUS statistical report: primary brain and other central nervous system tumors diagnosed in the United States in 2012-2016. Neuro Oncol. 2019;21:v1-v100. 10.1093/neuonc/noz150 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Buerki RA, Horbinski CM, Kruser T, Horowitz PM, James CD, Lukas RV. An overview of meningiomas. Future Oncol. 2018;14:2161-2177. 10.2217/fon-2018-0006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. McIver JI, Link MJ, Giannini C, Cohen-Gadol AA, Driscoll C. Choroid plexus papilloma and meningioma: coincidental posterior fossa tumors: case report and review of the literature. Surg Neurol. 2003;60:360-365. 10.1016/s0090-3019(03)00157-5 [DOI] [PubMed] [Google Scholar]
- 4. Oh JH, Cho HS, Hwang HS, Ji W. Primary pulmonary meningioma presenting as multiple lung nodules: a case report. Thorac Cancer. 2022;13:141-143. 10.1111/1759-7714.14270 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Saeed P, Rootman J, Nugent RA, White VA, Mackenzie IR, Koornneef L. Optic nerve sheath meningiomas. Ophthalmology. 2003;110:2019-2030. 10.1016/S0161-6420(03)00787-5 [DOI] [PubMed] [Google Scholar]
- 6. Ho CY, Mosier S, Safneck J, et al. Genetic profiling by single-nucleotide polymorphism-based array analysis defines three distinct subtypes of orbital meningioma. Brain Pathol. 2015;25:193-201. 10.1111/bpa.12150 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Subramanian PS, Bressler NM, Miller NR. Radiation retinopathy after fractionated stereotactic radiotherapy for optic nerve sheath meningioma. Ophthalmology. 2004;111:565-567. 10.1016/j.ophtha.2003.06.025 [DOI] [PubMed] [Google Scholar]
- 8. Clark VE, Erson-Omay EZ, Serin A, et al. Genomic analysis of non-NF2 meningiomas reveals mutations in TRAF7, KLF4, AKT1, and SMO. Science. 2013;339:1077-1080. 10.1126/science.1233009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Brastianos PK, Horowitz PM, Santagata S, et al. Genomic sequencing of meningiomas identifies oncogenic SMO and AKT1 mutations. Nat Genet. 2013;45:285-289. 10.1038/ng.2526 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Abe T, Kanno S, Niihori T, et al. LZTR1 deficiency exerts high metastatic potential by enhancing sensitivity to EMT induction and controlling KLHL12-mediated collagen secretion. Cell Death Dis. 2023;14:556. 10.1038/s41419-023-06072-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Zhu QY, He ZM, Cao WM, Li B. The role of TSC2 in breast cancer: a literature review. Front Oncol. 2023;13:1188371. 10.3389/fonc.2023.1188371 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Peng L, Li J, Wu J, et al. A pan-cancer analysis of SMARCA4 alterations in human cancers. Front Immunol. 2021;12:762598. 10.3389/fimmu.2021.762598 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Ko J, Jung J, Kim ST, et al. MET gene alterations predict poor survival following chemotherapy in patients with advanced cancer. Pathol Oncol Res. 2022;28:1610697. 10.3389/pore.2022.1610697 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Lu Z, Du W, Jiao X, et al. NOTCH1 mutation and survival analysis of tislelizumab in advanced or metastatic esophageal squamous cell carcinoma: a biomarker analysis from the randomized, phase III, RATIONALE-302 trial. J Clin Oncol. 2025;43:1898-1909. 10.1200/JCO-24-01818 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Li D, Guo B, Wu H, Tan L, Lu Q. TET family of dioxygenases: crucial roles and underlying mechanisms. Cytogenet Genome Res. 2015;146:171-180. 10.1159/000438853 [DOI] [PubMed] [Google Scholar]
- 16. Miller NR. Stereotactic fractionated radiation therapy for optic nerve sheath meningiomas. In: Lim M, Hsu W, Rigamonti D, Kleinberg LR, eds. Handbook of Radiosurgery in CNS Disease. Demos Medical; 2013:93-102. [Google Scholar]
- 17. Pandit R, Paris L, Rudich DS, Lesser RL, Kupersmith MJ, Miller NR. Long-term efficacy of fractionated conformal radiotherapy for the management of primary optic nerve sheath meningioma. Br J Ophthalmol. 2019;103:1436-1440. 10.1136/bjophthalmol-2018-313135 [DOI] [PubMed] [Google Scholar]
- 18. Sahm F, Schrimpf D, Stichel D, et al. DNA methylation-based classification and grading system for meningioma: a multicentre, retrospective analysis. Lancet Oncol. 2017;18:682-694. 10.1016/S1470-2045(17)30155-9 [DOI] [PubMed] [Google Scholar]
- 19. Paramasivam N, Hübschmann D, Toprak UH, et al. Mutational patterns and regulatory networks in epigenetic subgroups of meningioma. Acta Neuropathol. 2019;138:295-308. 10.1007/s00401-019-02008-w [DOI] [PubMed] [Google Scholar]
- 20. Choudhury A, Magill ST, Eaton CD, et al. Meningioma DNA methylation groups identify biological drivers and therapeutic vulnerabilities. Nat Genet. 2022;54:649-659. 10.1038/s41588-022-01061-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Nasrallah MP, Aldape KD. Molecular classification and grading of meningioma. J Neurooncol. 2023;161:373-381. 10.1007/s11060-022-04228-9 [DOI] [PubMed] [Google Scholar]
- 22. Sahm F, Aldape KD, Brastianos PK, et al. cIMPACT-NOW update 8: clarifications on molecular risk parameters and recommendations for WHO grading of meningiomas. Neuro Oncol. 2025;27:319-330. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All data pertinent to the study are presented in the main text, figures, tables, and supplementary material. Additional details are available per written request.