Spinal versus intracranial meningioma: Expression of E-cadherin and Fascin with relation to clinicopathological features

Abstract

BACKGROUND:

E-cadherin and Fascin are adhesive proteins that are expressed in many tumors. It was supposed that loss of expression of these proteins is associated with increased aggressiveness of the tumor. Whether spinal and intracranial meningiomas express adhesion proteins in different rates is not yet known.

OBJECTIVE:

We aimed to investigate the expression of E-cadherin and Fascin in a large number of meningioma specimens and determine if clinical and prognostic significance exists

METHODS:

One hundred and thirty-four spinal and intracranial meningioma samples were collected. Manual TMA blocks were constructed and immunohistochemistry for E-cadherin and Fascin was done. Focal or diffuse staining was considered positive.

RESULTS:

Intracranial meningioma occurred in significantly younger age than spinal ones. Most of spinal meningiomas were of transitional histology. E-cadherin was expressed in 38.8% of cases. Spinal meningiomas showed statistically significant negative expression of E-cadherin than intracranial tumors. All atypical meningiomas showed negative E-cadherin expression. Fascin was expressed in 9% of cases with significant expression in atypical cases.

CONCLUSIONS:

Aggressive behavior of meningioma could be explained in part by loss of E-cadherin and overexpression of Fascin especially in spinal meningiomas. Further studies are suggested to explore the biological aspects of spinal and intracranial meningiomas for constructing tailored targeted therapies.

1. Introduction

Meningiomas are a group of tumors arising from the arachnoidal cap cells of the meninges [1, 2]. They are the most common type of CNS tumors and comprise nearly 20% of intracranial tumors [3]. Their incidence increases with age with mostly occurring in females in the sixth and seventh decade of life and are detected incidentally upon imaging or as a result of symptoms such as changes in vision, headaches, seizures, hearing loss and gait dysfunction [1, 2, 3, 4].

Intracranial meningiomas commonly involve the cerebral convexities, falx cerebri, sphenoid ridge, cerebellar convexities and tentorium cerebelli while spinal meningiomas are usually seen in the thoracic region, although they are described in the cervical, lumbar, and rarely the sacral area [1, 5]. In terms of incidence, intracranial meningiomas predominate spinal meningiomas with the former comprising 7.5–12.7% of all meningiomas [5].

Histologically, amongst the benign spinal meningiomas, psammomatous meningioma emerges more common than meningothelial and transitional types; while on the contrary, meningothelial or aggressive anaplastic growth patterns are more common in intracranial meningiomas. Moreover, spinal meningiomas display lesser angioneogenesis compared to intracranial meningiomas thus indicating less vasculature to produce tumors that can cause neurological compromise. Clinically, spinal meningiomas can present early with neurological manifestations even when small in size, while intracranial meningiomas remain silent for many years leading to formation of more intratumoral calcifications than spinal tumors. Spinal meningiomas exhibit more single tumor cell clones compared to their intracranial counterparts that have mostly multiple tumor cell clones, therefore suggesting their predilection to aggressive behavior and relapse potential [6].

Meningiomas are divided into three histopathological grades by the WHO, with 70%–95% of them occupying WHO grade I and hence are benign and slow growing [4, 7]. While they are mostly unifocal, meningiomas can be multifocal in 5–9% of cases and can rarely metastasize outside the central nervous system [1]. Proper histologic grading of meningiomas plays a major role in providing appropriate therapeutic effect and prognosis. However, the disparity between histologic classification and aggressive behavior in meningiomas is around 7–20%. Thus, the investigation of biomarkers in these tumors is crucial to bridge the gap [8]. Of these markers, adhesion proteins have a potential role in meningioma pathogenesis. Epithelial (E-) cadherin, an important 120 kDa glycoprotein receptor involved in cell-cell adhesion that acts as a tumor-suppressor gene and metastasis-suppressor gene, is coded by the gene CDH1 (16q22.1). It interacts with the actin cytoskeleton through beta catenins at the adherens junction, a place of cell-cell contact. Depletion of E-cadherin results in the breakdown of cell-cell adhesions and increased invasive growth properties of tumors while re-expression declines both the proliferative and invasive properties of tumor cells [7, 9, 10].

Fascin, another protein involved in adhesion, is a 55 kDa globular actin binding protein whose gene is located on chromosome 7p22, is involved in assembling F-actin into parallel bundles, producing cell protrusions such as filopodia, lamellipodia, microspikes and promoting cell motility [11, 12, 13].

The role of expression of E-cadherin and Fascin as adhesive proteins had been previously studied in tumors of the liver [13], biliary duct [14], larynx [15], breast [16], lung [17], stomach [18], colorectum [19], pancreas [20] and urinary bladder [21] depicting mostly the overexpression of Fascin and diminished expression of E-cadherin corresponding to invasiveness and often poor prognosis. Additionally, E- cadherin and Fascin had been investigated in some CNS tumors as in pilocytic astrocytoma (WHO grade I), diffuse astrocytoma (WHO grade II), anaplastic astrocytoma (WHO grade III) and glioblastoma multiforme (WHO grade IV) indicating also that overexpression of Fascin and decreased expression of E-cadherin are associated with increasing grade [12, 22]. E-cadherin expression in meningioma has been investigated in few studies that demonstrated that its downregulation is related to increase in pathological grade and malignant types [9, 23, 24]. However, only one previous study reported increased Fascin expression with increasing pathological grade of meningioma [12].

To the best of our knowledge, the role of both adhesive proteins co-expression hasn’t been examined in meningioma. Moreover, we hypothesized that expression of both markers can be different in spinal versus intracranial meningiomas adding valuable information in better understanding the biological behavior of these tumors. So, in this study we aimed to investigate the immunohistochemical expression of E-cadherin and Fascin in a large number of spinal and intracranial meningioma specimens derived from the same patients and explore the relation of these adhesive proteins expression with different clinicopathological characters trying to explain the biological differences between spinal and intracranial meningiomas.

2. Material and methods

2.1. Samples

Files of all meningioma cases registered at the pathology department, Mansoura, Egypt were reviewed during the period from 2000 to 2007. Cases with incomplete resection or incomplete clinical data were excluded. One hundred thirty-four cases were fulfilling selection criteria, 14 of them were spinal and 120 were intracranial. The patients didn’t receive any pre-operative therapy.

2.2. Clinical parameters and histopathological evaluation

The clinicopathological data of these 134 cases were reviewed along with re-examination of all their slides. The data included: age, gender, location, histological type, atypia and recurrence. Follow up of the cases was done at the Oncology Center and Clinical Oncology and Nuclear Medicine Department, Mansoura University, Egypt.

2.3. Tissue Microarray (TMA) construction

Manual TMA blocks were constructed as previously described [25] using modified mechanical pencil tip method. Multiple cores were punched out from each case, each was measuring 0.8 mm in diameter. Multiple normal tissue cores were inserted to serve as positive and negative internal controls. Positive controls used for both E-cadherin and Fascin were tissue cores of breast, liver, kidney, salivary gland and pancreatic acini. Negative controls were tissue cores of brain, spleen, lymph nade, endometrium, smooth and skeletal muscles. Four micrometer thickness sections were prepared from the TMA blocks for routine H&E and immunohistochemistry (IHC) studies.

2.4. Immunohistochemistry

The slides were immunostained with monoclonal mouse anti-human E-cadherin Ab (Clone 36, Cat. # 790-4497, predilute ready-to-use for IHC, Ventana Medical Systems Inc., Tucson, Arizona, USA) and monoclonal mouse anti-human Fascin Ab (Clone 55k-2, Cat. # 760-2702, predilute ready-to-use for IHC, Ventana Medical Systems Inc., Tucson, Arizona, USA) according to the manufactures’ instructions. Slides were incubated in the oven (57${}^{\circ }$C) for 10 minutes then dewaxed using warm xylene (10 minutes) and then for another 10 minutes at room temperature in xylene. The slides were rehydrated in decreasing concentrations of alcohol then washed by tap water. For antigen retrieval, sections were immersed in Citrate buffer pH6 in a microwave for 20 minutes. Washing of the slides by PBS was done twice for ten minutes each. Drops of alcohol based hydrogen peroxidase were added for twenty minutes to block endogenous peroxidase activity followed by washing with PBS thrice. The antibody was incubated with the sections for one hour at room temperature followed by washing thrice by PBS, each for 3 minutes. Secondary antibody was applied and incubated for thirty minutes followed by washing by PBS thrice, each for 3 minutes. Then DAB-chromogen was applied on the slides for 3–5 minutes and the sections were counterstained by Mayer’s hematoxylin. All the chemicals used for the procedure were from Thermo Fisher Scientific, Waltham, Massachusetts, USA.

Assessment of both E-cadherin and Fascin expressions was done independently by two of the authors (Foda AA and Alam MS) and the cells showing positive expression were counted in each core. Staining was recorded as: 0 for negative, 1 for weak and/or focal and 2 for strong and/or diffuse positive as done in Lusis et al.’s study [26]. A final score was determined for each case by calculating an average score from the scores recorded by each examiner for the individual cores. A final score of 1 or greater was considered positive.

3. Statistical analysis

Data were analyzed using SPSS version 24.0. ${\chi }^2$ (Chi-square) test was used to test significant differences between spinal and intracranial meningiomas in E-cadherin and Fascin expressions, as well as relation of both markers expression with clinicopathological and histological parameters. A $P⩽$ 0.05 was considered significant in all tests.

Figure 1.

a. Strong E-cadherin cytoplasmic/membranous staining in a case of fibroblastic intracranial meningioma. b. Moderate E-cadherin cytoplasmic/membranous staining in a case of transitional spinal meningioma. c. Faint Fascin cytoplasmic staining in a case of psamomatous intracranial meningioma. d. Negative Fascin staining in a case of transitional spinal meningioma (x200).

4. Results

A total of 134 meningioma specimens were analyzed. The clinicopathological and histological features of the tested meningioma cases are listed in Table 1.

Table 1.

Clinicopathological and histological features of intracranial and spinal meningiomas

	Intracranial		Spinal		$P$ value
	No. (%)		No. (%)
Age (y)
60	99	(82.5%)	07	(50%)	0.005${}^{\mathrm{*}}$
$⩾$ 60	21	(17.5%)	07	(50%)
Gender
Male	28	(23.3%)	3	(21.4%)	0.873
Female	92	(76.7%)	11	(78.6%)
Histological type
Meningotheliomatous	48	(40.0%)	0	(0.0%)	0.006${}^{\mathrm{*}}$
Transitional	53	(44.2%)	11	(78.6%)
Fibroblastic	10	(08.3%)	0	(0.0%)
Psammomatous	09	(07.5%)	03	(21.4%)
Atypia
Benign	114	(95.0%)	14	(100%)	0.392
Atypical features	6	(5.0%)	0	(0.0%)
Recurrence
Negative	109	(90.8%)	14	(100%)	0.237
Positive	11	(09.2%)	0	(0.0%)
E-cadherin expression
Negative	59	(54.1%)	11	(84.6%)	0.036${}^{\mathrm{*}}$
Positive	50	(45.9%)	02	(15.4%)
Fascin expression
Negative	102	(89.5%)	14	(100%)	0.202
Positive	12	(10.5%)	0	(0.0%)

${}^{*}P⩽$ 0.05 is significant.

Majority of the samples belonged to females (103 cases; 77%) while 31 cases (23%) were from males. Intracranial meningiomas consisted of 90% of the samples (120 cases) while spinal meningiomas consisted of 10% of the samples (14 cases). Age range at presentation for meningioma was 6 to 80 years (mean, 47 years) and SD $\pm$ 15.18. The most frequent age of presentation was 50 years in females, and 52 years in males. Statistically significant difference ($p=$ 0.005) between meningioma and age of presentation was found, with 82.5% of intracranial cases presented in ages 60 years (Table 1).

Intracranial meningiomas were located mostly at the skull base including sphenoidal wing and olfactory groove (49 cases; 41%) followed by parito-occipital (26 cases; 22%) and frontal (24 cases; 20%) region. The least common presentation was in the temporal region (21 cases; 18%) (Table 1).

There was a significant difference in the histologic type between intracranial and spinal meningiomas ($p=$ 0.006). The most common presenting histologic type among spinal meningiomas was transitional (78.6% of the cases) followed by psammomatous type (21% of the cases). There was no meningothelial or fibroblastic type of spinal meningiomas, while meningothelial type accounted for 40.0% of tested intracranial meningiomas (Table 1).

Ninety-five percent of intracranial and all of the spinal meningiomas were benign, with 6 cases of intracranial meningiomas having atypical features (5%). No statistically significant relationship was found between presence of atypia ($p=$ 0.392). Similarly, there was no statistical significant relationship between the location of the tumor and its recurrence ($p=$ 0.237). There was no recurrence in all spinal tumors, while recurrence was found in 9% of intracranial cases (Table 1).

E-cadherin was expressed in 52/134 cases (38.8%). Spinal meningiomas showed statistically significant negative expression of E-cadherin than intracranial tumors ($p=$ 0.036). Almost 50 cases (46%) of intracranial tumors were positive for E-cadherin, while only 2 cases (4%) of the spinal tumors were positive for E-cadherin expression (Table 1). Moreover, the relations of E-cadherin expression to clinicopathological and histological features of tested cases of meningioma were summarized in Table 2. There was statistically significant relation is between E-cadherin expression and age ($p=$ 0.010) and presence of atypia ($p=$ 0.049). Most E-cadherin positive tumors occur before the age of 60 and all atypical meningiomas showed negative E-cadherin expression (Table 2).

Table 2.

Relation of E-cadherin expression to clinicopathological and histological features of cases of meningioma

	Negative		Positive		$P$ value
	No. (%)		No. (%)
Age (y)
60	48	(68.6%)	46	(88.5%)	0.010${}^{\mathrm{*}}$
$⩾$ 60	22	(31.4%)	6	(11.5%)
Gender
Male	16	(22.9%)	13	(25.0%)	0.783
Female	54	(77.1%)	39	(75.0%)
Location
Frontal	12	(17.1%)	11	(21.2%)	0.054
Temporal	12	(17.1%)	06	(11.5%)
Parito-occipital	16	(22.9%)	08	(15.4%)
Skull base including	19	(27.1%)	25	(48.1%)
sphenoidal wing and
olfactory groove
Spinal	11	(15.7%)	02	(03.8%)
Histological type
Meningotheliomatous	25	(35.7%)	16	(30.8%)	0.682
Transitional	32	(45.7%)	29	(55.8%)
Fibroblastic	07	(10.0%)	03	(05.8%)
Psammomatous	06	(08.6%)	04	(07.7%)
Atypia
Benign	65	(92.9%)	52	(100%)	0.049${}^{\mathrm{*}}$
Atypical features	05	(07.1%)	0	(0.0%)
Recurrence
Negative	63	(90.0%)	48	(92.3%)	0.660
Positive	07	(10.0%)	04	(07.7%)

${}^{*}P⩽$ 0.05 is significant.

Fascin was expressed in 12/134 cases (about 9%). Although none of the spinal meningiomas showed Fascin positivity, there was no statistically significant difference in Fascin expression between spinal and intracranial meningiomas ($p=$ 0.202) as only 10% of intracranial meningiomas were positive (Table 1). The relations of Fascin expression to clinicopathological and histological features of tested cases of meningioma were summarized in Table 3. Similarly, the only statistically significant relation is between Fascin expression and presence of atypia ($p=$ 0.017). Unlike E-cadherin, Fascin was expressed in 2/5 (40%) of cases with atypia, while most of the benign cases (113/123; about 92%) were negative for Fascin (Table 3).

Table 3.

Relation of Fascin expression to clinicopathological and histological features of cases of meningioma

	Negative		Positive		$P$ value
	No. (%)		No. (%)
Age (y)
60	91	(78.4%)	10	(83.3%)	0.693
$⩾$ 60	25	(21.6%)	2	(16.7%)
Gender
Male	24	(20.7%)	05	(41.7%)	0.098
Female	92	(79.3%)	07	(58.3%)
Location
Frontal	18	(15.5%)	05	(41.7%)	0.147
Temporal	18	(15.5%)	01	(08.3%)
Parito-occipital	21	(18.1%)	03	(25.0%)
Skull base including	45	(38.8%)	03	(25.0%)
sphenoidal wing and
olfactory groove
Spinal	14	(12.1%)	0	(0.0%)
Histological type
Meningotheliomatous	38	(32.8%)	08	(66.7%)	0.108
Transitional	60	(51.7%)	04	(33.3%)
Fibroblastic	08	(06.9%)	0	(0.0%)
Psammomatous	10	(08.6%)	0	(0.0%)
Atypia
Benign	113	(97.4%)	10	(83.3%)	0.017${}^{\mathrm{*}}$
Atypical features	03	(02.6%)	02	(16.7%)
Recurrence
Negative	107	(92.2%)	10	(83.3%)	0.295
Positive	09	(07.8%)	02	(16.7%)

${}^{*}P⩽$ 0.05 is significant.

Finally, we tested interrelation between E-cadherin and Fascin expression in meningioma cases. There was no statistically significant correlation between the expression of E-cadherin and Fascin in meningiomas; only 8 cases showed co-expression of both markers ($p=$ 0.610).

5. Discussion

Meningiomas being the most common CNS neoplasms have an annual incidence of 4-5/100.000 individuals and are often associated with serious neurological consequences and decreased quality of life [2, 27]. The overall 5-year survival rate is less than 70% that decreases with age. Although being common, less is known about them in terms of epidemiology, clinicopathological features and therapeutic management compared to gliomas. This can be quite attributed to its mostly benign behaviour and latent course [28]. However, expression of biological markers is necessary for better understanding of the pathogenesis and hence treatment of these tumors.

Expression of the adhesion protein E-cadherin had been studied in meningiomas and other CNS tumors as astrocytomas and medulloblastomas. Previous studies revealed contradictory results. Schwechheimer et al. [29] demonstrated the expression of E-cadherin in transitional, psammomatous, fibroblastic and meningothelial variants of meningioma likewise in our study, but also in papillary, angiomatous and microcystic variants. Motta et al. [22] reported that astrocytomas showed more expression of E-cadherin than medulloblastomas with a significant relation between decreased expression and increased invasiveness of these tumors. Grade I and II astrocytomas showed relatively higher E-cadherin expression than grade III and IV. In the current study, we found the same relations in meningiomas. E-cadherin expression was significantly related to higher grading of meningioma. Moreover, most E-cadherin positive meningiomas occurred before the age of 60 and all atypical meningiomas showed negative E-cadherin expression. These findings were also evident in the studies conducted by Zhou et al. [23] and Pecina-Slaus et al. [7], where E-cadherin expression was low in higher grade meningiomas. They also reported significant difference in expression of E-cadherin between invasive and non-invasive meningiomas, strongly suggesting that E-cadherin can be a potentially negative regulator of tumor invasion. However, they found that expression levels of E-cadherin were lower in recurrent than non-recurrent cases (33.33% and 90% respectively). In contrast, we failed to find this relation in the current study. This can be attributed to the low number of cases in their study (49 cases) as compared to ours (134 cases).

Although the role of E-cadherin loss in invasiveness of the tumors is well known, previous studies on CNS and non-CNS tumors revealed contradictory results regarding the relation of recurrence rate/survival of the patients and loss of E-cadherin expression. In the current study, we found no relation of loss of E-cadherin expression with increased recurrence of meningiomas. Motta et al. [22] reported that no relation was also noted between E-cadherin expression and survival of high grade astrocytomas and medulloblastomas patients. Similarly, a meta-analysis done by Xing et al. [30] reported that decreased expression of E-cadherin was considered as a poor prognostic factor in fifteen studies on gastric cancer patients, while ten studies concluded that E-cadherin was not a prognostic indicator for survival and only one study showed better prognosis with reduced E-cadherin expression! The same contradictory results were also reported in colorectal carcinoma [10].

Expression of Fascin had been studied in meningioma and other tumors as well. Previous studies reported that overexpression of Fascin was significantly associated with more advanced tumor grades of meningioma and astrocytoma [12, 31]. Our results were in accordance with these findings. There was a statistically significant relation between Fascin expression and presence of atypia. Fascin was expressed in 40% of cases with atypia, while most of the benign cases were negative for Fascin. Overexpression of Fascin was also related to aggressiveness of many other tumors. Xing et al. [32] reported that Fascin expression was upregulated in highly invasive breast cancer cell lines with no detectable expression in those with low or no invasive potential. They found that forced expression of Fascin significantly decreased adhesion and increased proliferation and invasion while knockdown of Fascin resulted in increased adhesion and decreased invasiveness and hence they concluded that Fascin could be a promising therapeutic target in treatment of breast cancer. In hepatocellular carcinoma, Hayashi et al. [13] found Fascin-1 expression was evidently seen in histologically less differentiated tumors, likewise our study, but correlated it with reduced expression of E cadherin. Mao et al. [14] found a reverse relationship between Fascin and E-cadherin expression in cholangiocarcinoma with significantly poor prognosis with Fascin, but not E-cadherin, expression. Zou et al. [15] demonstrated that laryngeal squamous cell carcinomas with Fascin-1 overexpression and E-cadherin loss were related to poor survival and had significantly higher tumor recurrence rate compared to those with low Fascin-1 expression and E-cadherin overexpression. In contrast to these studies, we didn’t find any significance of co-expression of adhesion proteins in meningioma. However, this is the first study to investigate both markers in spinal versus intracranial meningioma. There was no difference in Fascin expression, while spinal meningiomas showed significant loss of E-cadherin expression than intracranial meningiomas. This addition to our previous knowledge about the biological differences between spinal and intracranial meningiomas needs further molecular and functional studies to further understand the aggressive capacity of meningiomas and tailor targeted therapy for certain subtypes of these tumors.

6. Conclusion

Aggressive behavior of meningioma could be explained by loss of E-cadherin and overexpression of Fascin adhesive proteins. Spinal meningiomas differ than intracranial ones in many histological, clinicopathological and biological aspects. One of these biological differences is that spinal meningiomas showed statistically significant loss of E-cadherin expression than intracranial meningiomas. Further studies are suggested to explore the biological aspects of both types of meningioma in a step for constructing tailored targeted therapies.