Diagnostic challenges in orbital meningioma: A case series of six patients

INTRODUCTION

Meningiomas are the most common primary brain intracranial neoplasms.^[1,2] They arise from meningothelial cells of the arachnoid mater. The most common sites include the cerebral convexities, olfactory grooves, sphenoid ridges, parasellar/suprasellar regions, optic nerve sheath, petrous ridges, and tentorium and posterior fossa.^[3,4] Intraspinal meningiomas are less common than intracranial meningiomas and account for ~12% of all cases. Intraorbital meningiomas are the least frequent and contribute to <2% of all meningiomas. We present a series of six cases of orbital meningioma diagnosed at our hospital over a period of 2 years. In addition, we identified and described various diagnostic challenges in interpreting the small biopsies of these tumors.

CASE SERIES

Case 1

A 65-year-old female presented with a painless mass protruding over the left lower eyelid [Figure 1a]. The patient underwent a magnetic resonance imaging (MRI) scan, which showed that there was a large, lobulated, heterogeneously enhancing, soft-tissue mass measuring 8 cm × 4 cm × 5.2 cm epicentered in the left retro-orbital space, with intracranial extension into the left temporal region. The mass was causing erosion of the sphenoid bone [Figure 1b].

Close

Figure 1: Case 1- Composite clinical, radiological and histopathological images. (a) Clinical photograph showing a painless mass protruding over the left lower eyelid. (b) Coronal post-contrast T1-weighted magnetic resonance images demonstrating a large, lobulated, heterogeneously enhancing retro-orbital mass with intracranial extension (red arrow). (c) Photomicrograph [Hematoxylin and eosin (H&E), 200×] showing dense hyalinisation with tumor cells in a cord-like and slit-like pattern. (d) Immunohistochemistry slide view showing nuclear positivity for progesterone receptor (PR) at 200×. (e) Immunohistochemistry showing focal weak positivity for epithelial membrane antigen (EMA) at 200×. (f) Immunohistochemistry showing CD34 staining highlighting only vessels, with the tumor cells negative (200×).

Export to PPT

A small core biopsy of the orbital tissue mass was performed with a clinical differential diagnosis of sarcoma, poorly differentiated carcinoma or high-grade lymphoma. Microscopy revealed large areas of hyalinisation with tumor cells embedded in the hyalinised stroma, giving a cord-like and slit-like appearance. A few spaces with red blood cells resembling vascular tumors were also observed [Figure 1c].

The tumor cells were seen clustered around these spaces. These cells had abundant cytoplasm with mild to moderate cytological atypia. Focal clusters of cells with syncytial patterns were also observed, with a few scattered mitotic counts. Based on these findings, we shortlisted the morphological differentials of vascular tumor, sarcoma, carcinoma, solitary fibrous tumor/hemangiopericytoma (SFT/HPC) and a remote possibility of meningioma. Immunohistochemistry (IHC) of the tumor tissue sample was conducted. The tumor showed positivity for Vimentin, progesterone receptor (PR) [Figure 1d] and focal weak positivity for epithelial membrane antigen (EMA) [Figure 1e], P63 and D240, while CK7, CK20, smooth muscle actin (SMA), desmin, Myod1, P40, S-100, synaptophysin, and chromogranin were negative. A vascular tumor was one of the most likely differentials; however, CD34 staining [Figure 1f] was negative.

As cells appeared singly, infiltrating, E-Cadherin was added to the IHC panel and was also found to be retained by the tumor cells. The karyotyping index (Ki-67) was 3–4% in the highest proliferative areas. In view of the abovementioned IHC profile, we ruled out carcinomas, sarcomas, neuroendocrine tumors, angiosarcoma and neural tumors. The combination of morphological, immunohistochemical and clinical profiles led to the diagnosis of meningioma. Although reaching the final diagnosis may appear relatively simple after IHC, it is not easy to perform a definitive interpretation on a small biopsy with dense hyalinisation, which obscures morphological features.

Subsequently, the patient underwent wide local excision of the mass along with exenteration of the left eyeball. On histopathological examination of the complete specimen, the tumor displayed classic features of transitional meningioma along with significant hyalinisation, which was the reason for the difficult interpretation of the smaller biopsy. Finally, this case was labelled as a primary intraorbital transitional meningioma of the World Health Organization (WHO) grade 1. As the tumor was completely resected and was WHO grade 1, no adjuvant therapy was indicated. The patient was managed with observation and serial imaging and has remained disease-free at 30 months of follow-up.

Case 2

The second case was similarly challenging. A 64-year-old male presented with pain and swelling on the right side of his face for a 1-month duration. Computed tomography (CT) of the face revealed an ethmoidal mass extending into the periorbital tissue, destroying the cribriform plate. The clinicoradiological impression was lymphoma versus invasive fungal infection. An endoscopic biopsy showed multiple small tissue bits focally lined by benign respiratory lining epithelium, and the underlying subepithelium showed tumor cells in diffuse sheets and clusters. At some locations, the cells showed a perivascular arrangement. Cells were monomorphic, round to oval, with coarse nuclear chromatin, and were in syncytium [Figure 2a]. Morphologically, we had a differential diagnosis of a vascular tumor, carcinoma, adenoid cystic carcinoma, low-grade sarcoma, etc. On IHC, the tumor cells were diffusely positive for Vimentin and EMA [Figure 2b]. Focal positivity for Pan CK, SMA, P63 and PR [Figure 2c] was observed; however, CD34, synaptophysin, desmin, P40, CD99, TFE3, C-kit and p16 were strongly negative. Ki-67 was 5–8% in the highest proliferative areas. Based on these findings, we ruled out carcinomas, low-grade vascular tumors, carcinoids, squamous cell carcinomas, small round blue-cell tumors, adenoid cystic carcinomas and smooth and skeletal muscle differentiation. Finally, we arrived at differentials of glomus versus a remote possibility of meningioma based on positivity for SMA, vimentin, EMA and PR.

Close

Figure 2: Case 2: Composite histopathological and immunohistochemical images. (a) Photomicrograph at [Hematoxylin and eosin (H&E), 200×] showing monomorphic tumor cells arranged in sheets with a syncytial pattern and perivascular arrangement. (b) Immunohistochemistry staining showing diffuse positivity for EMA at 200×. (c) Immunohistochemistry staining showing focal positivity for PR at 200×. EMA: Epithelial membrane antigen, PR: Progesterone receptor.

Export to PPT

In the resected specimen, classic features of meningothelial meningioma (WHO grade 1) were observed. The tumor cells were arranged in a syncytium and a whorled pattern. The cells had abundant cytoplasm and fine nuclear chromatin. A few cells also exhibited intranuclear inclusions. Morphology and IHC profile clinched the diagnosis of meningioma. The tumor was resected through an endoscopic endonasal approach combined with a transorbital component. The tumor was removed in a piecemeal manner, with anterior skull base reconstruction. As there was no residual tumor on imaging at 3 months and the lesion was WHO grade 1, the patient was managed with observation alone, without adjuvant therapy, and has shown no evidence of recurrence on imaging at >2 years (26 months) of follow-up.

The other four cases were diagnosed based on complete resection specimens with classic features of meningothelial meningioma, and all were grade 1 tumors.

Case 3

A 62-year-old man presented with a history of proptosis of the right eye, which had been gradually progressive for the past 01 year. On evaluation, visual acuity was 6/9 in both eyes. Extra-ocular movements of the affected eye were restricted, with abduction −2 and elevation −4, and the rest of the eye movements were normal. The pupils were round, regular and reacting to light. Intra-ocular pressure was 14 and 16 mmHg in the right and left eyes, respectively. The rest of the anterior and posterior segments were essentially normal. Non-axial proptosis of the right eye was noted, with the globe displaced downward and inward, suggesting growth in the superolateral quadrant of the orbit [Figure 3a]. Measurement with Hertel’s exophthalmometer showed an 8 mm proptosis in the right eye. CT orbit showed an expansile, mixed sclerotic-lytic bony lesion in the superolateral wall and roof of the orbit with extension into the intracranial cavity and posterior ethmoid [Figure 3b]. With this clinicoradiological picture, he was diagnosed as a case of right eye orbital mass with a differential diagnosis of lacrimal gland tumor or sphenoid wing meningioma. Under general anaesthesia, the tumor was approached by a sub-brow incision, and debulking was done. Tissue was sent for histopathological analysis and was confirmed to be a meningioma. On histopathological examination, tumor cells were arranged in a lobular and microcystic pattern. The cells were elongated with cytoplasmic vacuolation and were embedded in fibrocollagenous and myxoid background [Figure 3c]. The cells were monomorphic, round to oval, with coarse nuclear chromatin. The thin, elongated processes formed by cells were seen, which gave a cobweb-like appearance on hematoxylin and eosin staining under ×400 magnification [Figure 3d]. Based on these features, differentials of salivary gland tumors such as pleomorphic adenoma, adenoid cystic carcinoma, dermoid cysts with cystic degeneration and the rare possibility of meningioma were considered. Panel of IHC was carried out, which showed negative staining for P40, CK7 to rule out pleomorphic adenoma and negative staining for c-kit to rule out adenoid cystic carcinoma. Deeper sections did not reveal any endodermal/ectodermal/mesodermal component to rule out a dermoid cyst. IHC by EMA and PR was diffusely positive [Figure 3e and f]. Overall histomorphology and IHC favoured the diagnosis of microcystic meningioma. Following surgical debulking, the patient’s symptoms were relieved. MRI at 3 months post-surgery suggested a small residual mass measuring about 1.5 mm in its greatest dimension. After multidisciplinary discussion, the patient was kept under observation without adjuvant therapy in view of the low-grade (WHO grade 1) meningioma. The patient has remained asymptomatic with no increase in the size of the residual mass at 24 months of follow-up.

Close

Figure 3: Case 3: Composite clinical, radiological and histopathological images. (a) Clinical photograph showing non-axial proptosis of the right eye. (b) Computed tomography of the skull showing an expansile, mixed sclerotic-lytic bony lesion of the superolateral orbital wall and roof with intracranial extension seen in both coronal (red arrow) and axial (green arrow) views. (c) Photomicrograph with [Hematoxylin and eosin (H&E), 200×] showing a lobular, microcystic pattern of cells in a myxoid background. (d) Photomicrograph with H&E staining at 400× showing the characteristic cobweb-like appearance. (e) Immunohistochemistry staining at showing diffuse positivity for EMA at 200×. (f) Immunohistochemistry staining diffuse positivity for PR at 200×. EMA: Epithelial membrane antigen, PR: Progesterone receptor.

Export to PPT

DISCUSSION

Meningiomas are the most common primary intracranial neoplasms and are classified according to the WHO grading system. The most commonly encountered meningioma is Grade 1.^[4] These are seen more commonly in females than in males, as discussed in previous studies. It is suggested that this sex disparity is due to differences in sex hormones. Many of these tumors display a combination of oestrogen, progesterone and androgen receptors, which is one of the causes of increased tumor growth during pregnancy and the luteal phase of the menstrual cycle.^[2,5] We, in our study, found four males and two females, and irrespective of gender, we performed PR IHC on all tumor tissues and found positivity in all six (100%) meningothelial tumors, indicating that these tumors express hormone receptors uniformly. Given that our series comprises only six patients, this male-predominant ratio should be interpreted with caution and not be taken as evidence of a true male predominance in orbital meningioma; the well-established female predominance described for meningiomas in general, attributed to hormone receptor expression, may simply not be reflected in a sample of this size.

Meningiomas are generally observed in middle-aged and elderly patients, with a peak in the sixth and seventh decades of life. Childhood tumors are less common but tend to be more aggressive.^[2,3] In our study, there were four males and two females with ages ranging from 10–70 years. The youngest patient in our study was 10 years of age.

Meningiomas involving the orbits are relatively rare. They can be primary or secondary. Primary lesions are derived from cells lining the intraorbital region, the optic nerve sheath, or the orbital face of the sphenoid bone. Secondary intraorbital meningiomas arise intracranially and subsequently extend into the orbit. These tumors most often arise from the inner or outer sphenoid ridge, but can also arise from the clinoid, planum sphenoidale, frontoparietal area, olfactory groove, tuberculum sellae or from the paranasal sinuses.^[3,6-9] In our study, there were four cases of primary orbital origin (Cases 1, 3, 4 and 5) and two cases of secondary meningioma (Cases 2 and 6) with extension into the orbit.

Among the different WHO subtypes, meningothelial, fibrous, and transitional meningiomas are most commonly encountered.^[3] In our study, we found four cases of meningothelial meningioma and one case of transitional meningioma.

It is challenging to diagnose meningiomas accurately on histopathological examination of small biopsies, especially when they arise in uncommon locations, as we experienced in three of our cases. In smaller biopsies, the diagnosis is not straightforward, and morphology and IHC in conjunction could help in pointing toward the diagnosis, whereas in a few cases, excision of the mass may be required to reach a definitive diagnosis. In an old study by Kepes and Kernohan, meningiomas were described as great mimickers on histology, which can often lead to differential diagnosis, particularly those infiltrating the soft tissues of the scalp simulating carcinomas; those with loose oedematous changes in meningiomas with an appearance similar to that of astrocytomas; palisading of cells may serve to set up a close resemblance to neurilemomas.^[10] Boulagnon-Rombi et al.^[11] brought out in their study the similarities between meningioma and other differential diagnoses, which can be challenging. They can mimic any other tumor and exhibit a wide range of histological appearances such as melanoma, haemangioblastoma, gliosarcoma, SFT/HPC, schwannomas, neurofibromas, malignant peripheral nerve sheath tumors, peripheral primitive neuroectodermal tumors (PNET) and synovial sarcomas.^[11]

Meningiomas can imitate carcinomas, particularly when they invade the skull and adjacent soft tissues of the scalp. In such cases, careful evaluation of other features of meningiomas, such as syncytial pattern, indistinct cell borders, lack of prominent nucleoli and lack of pleomorphism, helps to distinguish them from carcinomas. Although these features are sometimes subjective, pleomorphism and prominent nucleoli can be observed in anaplastic meningiomas. IHC may be of little help because EMA and Pan CK can be positive in both meningiomas and carcinomas. Positivity for vimentin and positive expression of hormone receptors, such as PR, can help support a diagnosis of meningioma over carcinoma. The presence of psammomatous calcification sometimes helps in supporting the diagnosis; however, it becomes difficult when there is a differential diagnosis of metastatic deposits of ovarian and thyroid carcinoma. Careful history taking in these cases may help exclude the diagnosis of metastatic deposits. In case 1 of our series, we received a small biopsy of the orbital tissue mass with a clinical differential diagnosis of sarcoma versus poorly differentiated carcinoma or high-grade lymphoma. A carcinoma was also a differential diagnosis. The above-mentioned features helped to exclude carcinoma in this case.

Meningiomas may undergo partial degeneration or edematous changes.^[11] In such cases, it sometimes becomes difficult to distinguish loose areas of meningiomas from astrocytomas, particularly when there is marked edema. Looking for areas with a glial background and IHC with glial fibrillary acidic protein (GFAP) and isocitrate dehydrogenase helps reach an accurate diagnosis.

The palisading of cells may serve to establish a close resemblance to nerve sheath tumors. Careful search for Verocay bodies, presence of Schwann cells and staining positivity for S-100 help differentiate meningiomas from them. In Boulagnon-Rombi et al.’s^[11] study, S-100 was expressed in fibrous and psammomatous meningiomas also. In this situation, SOX10 was the most useful marker, as it is not expressed by meningiomas, but is expressed by nerve sheath tumors. SOX10 is a neural crest transcription factor crucial for the specification, maturation and maintenance of Schwann cells and melanocytes. It is considered to be more sensitive and specific than S-100.^[12] However, we did not face a differential diagnosis of nerve sheath tumor in any of our cases. SFT/HPC is yet another differential diagnosis, predominantly with the fibrous variant of meningioma. These tumors have histomorphologically overlapping features with meningiomas. EMA and PR for meningiomas, and STAT6 and CD34 for the diagnosis of SFT/HPC, are helpful. The detection of nuclear positivity of STAT6 with IHC indicates the presence of the NAB2-STAT6 fusion or other alterations involving STAT6 and can help differentiate SFTs from other histological mimics such as meningiomas.^[13] In case number 1 of our series, we encountered dense fibrosis resulting in the obscuration of morphological features, which led to one of the differentials of SFT/HPC. SFTs are composed of haphazardly arranged spindle-to-ovoid monomorphic cells admixed with hyalinised, dilated, thin-walled, branching (staghorn-shaped) blood vessels. It has a wide histological spectrum, ranging from a hypocellular type to a highly cellular phenotype, with cells arranged in a patternless manner. The paucicellular end of the spectrum shows abundant stromal keloidal-type collagen. In the fibrous subtype of meningioma, spindle cells are seen in parallel, storiform or interlacing bundles in a collagen-rich matrix mimicking SFT/HPC features; however, subtle features such as nuclei are monotonous and round to oval, and they lack the pseudoinclusions typical of meningioma. In addition, as mentioned previously, IHC for CD34 and STAT6 supports the diagnosis of SFT/HPC, whereas EMA and PR favour meningioma.

Microcystic or clear cell variants of meningiomas can morphologically overlap with hemangioblastomas.^[11] On IHC, hemangioblastomas are positive for inhibin and negative for EMA and PR, whereas meningiomas are negative for inhibin and positive for EMA and PR.^[14] Gliosarcomas also have overlapping morphology with meningiomas. GFAP positivity in gliosarcomas and negativity in meningiomas help in their differentiation. PNETs express CD99 and FLI1; however, they stain negatively for EMA and PR.^[15] Anaplastic meningiomas can be difficult to differentiate from sarcomas, melanomas or carcinomas. Close morphological evaluation and a panel of IHCs can help distinguish between these differential diagnoses. Dense fibrosis can mimic slit-like vascular spaces, mimicking vascular tumors, such as low-grade angiosarcomas, as in one of our cases [Figure 1c]; however, these tumor cells display positivity for CD34 and CD31, whereas meningiomas are negative for the same. Sometimes, even glomus tumors can mimic meningiomas due to their syncytial pattern, as seen in the smaller biopsy in our second case [Figure 2a]; however, the characteristic morphology of punched-out nuclei and arrangement around vessels may help differentiate the two. IHC profiling can also be useful, as CD34 and SMA are characteristically positive in glomus tumors, but negative in meningiomas.^[12-15]

Case 4 in our series, a 10-year-old boy, illustrates the distinct profile of pediatric orbital meningioma. Childhood meningiomas account for <5% of all meningiomas and differ from their adult counterparts in several respects: When an underlying tumor-predisposition syndrome is identified, NF2-related disease remains the most frequent association; pediatric tumors more often display atypical or higher-grade histology and a greater propensity for local recurrence even after a Grade 1 resection; and the extent of resection, rather than grade alone, appears to be the principal determinant of outcome, making prolonged radiological follow-up particularly important in this age group.^[16] Our patient’s tumor showed classical low-grade meningothelial morphology without atypia; nonetheless, given the recognised potential for more aggressive behaviour in pediatric meningioma, we have planned closer-interval imaging follow-up than would routinely be offered to an adult with an identical Grade 1 lesion.

Molecular testing was not available at our institute at the time these six cases were diagnosed, and could not be performed retrospectively because no tissue suitable for sequencing remained archived; this is acknowledged as a limitation of the present study. As an indirect point of reference, our institute has since established an in-house next-generation sequencing facility, with which we have molecularly profiled seven subsequent intracranial meningiomas (unrelated to the present orbital series). NF2 mutations were the most frequent alteration, identified in three of these seven tumors (one with a co-existing NOTCH2 mutation), a telomerase reverse transcriptase (TERT) promoter mutation (c.-146G>A) was found in one atypical (Grade 2) meningioma, an AKT1 p.E17K hotspot mutation was identified in two tumors (one atypical, one transitional), and a CTNNB1 p.Gly34Val mutation with a concurrent FGFR3:TACC3 fusion was found in one atypical meningioma. This distribution broadly mirrors the wider literature, in which NF2 inactivation accounts for roughly half of all meningiomas, non-NF2 alterations such as AKT1, SMO, TRAF7 and KLF4 are enriched in lower-grade, skull-base tumors and TERT promoter mutation or CDKN2A/B loss is associated with higher-grade, more aggressive disease.^[17] Orbital and optic nerve sheath meningiomas specifically have been reported to be largely NF2-intact with few copy number alterations, consistent with their typically indolent clinical course, although rare cases may acquire high-risk copy number alterations and behave aggressively.^[18]

The surgical approach in our series was tailored to the location and extent of each tumor: Wide local excision with exenteration for the predominantly intraorbital Case 1; a combined endoscopic endonasal and transorbital approach with anterior skull base reconstruction for the ethmoid-orbital Case 2; sub-brow debulking for the orbital roof/wall lesion of Case 3; fronto-temporal (pterional) craniotomy with transcranial orbitotomy for the intraorbital tumor with optic nerve compression in Case 4; lateral orbitotomy for the purely intraorbital Case 5; and a combined cranio-orbital (fronto-temporal craniotomy with orbitozygomatic extension) approach for the sphenoid wing-based Case 6. Despite this heterogeneity of approach, complete or near-complete excision was achieved in all six patients, and none required adjuvant radiotherapy. This reflects both the WHO grade 1 histology in all cases and the value of selecting a surgical approach matched to tumor location, and is consistent with the generally favourable outcomes reported for completely resected low-grade orbital meningioma. All six patients remain recurrence-free at last follow-up (range 16–30 months), with only Case 3 showing a small, stable, asymptomatic residual focus on interval imaging.

Table 2 summarises the various differentials of meningioma mentioned in the literature ^[10,13-15] and our differentials in two of our difficult cases involving small biopsies.

CONCLUSIONS

Orbital meningiomas are rare, and their diagnosis on small biopsies can be challenging owing to hyalinisation, fibrosis and morphological overlap with other tumors. A combination of clinicopathological correlation and a focused IHC panel usually allows a confident diagnosis, although excision is occasionally required for definitive confirmation. As this series comprises only six patients, these observations should be viewed as illustrative rather than definitive, and larger multi-institutional studies are needed to validate them.