Immunohistochemical Staining for Claudin-1 Can Help Distinguish Meningiomas From Histologic Mimics

Size: px

Start display at page:

Download "Immunohistochemical Staining for Claudin-1 Can Help Distinguish Meningiomas From Histologic Mimics"

Mitchell Harper
5 years ago
Views:

1 Anatomic Pathology / CLAUDIN-1 IN MENINGIOMAS Immunohistochemical Staining for Claudin-1 Can Help Distinguish Meningiomas From Histologic Mimics Hejin P. Hahn, MD, PhD, Elizabeth A. Bundock, MD, PhD, and Jason L. Hornick, MD, PhD Key Words: Immunohistochemistry; Claudin; Tight junctions; Meningioma; Schwannoma; Solitary fibrous tumor; Hemangiopericytoma; Soft tissue tumors DOI: /G659FVVBMG7U4RPQ Abstract The histologic distinction between meningiomas and other tumors of the central nervous system occasionally can be difficult. Claudin-1 is a tight junction associated protein recently shown to be expressed in anaplastic meningiomas. The purpose of this study was to determine whether immunohistochemical staining for claudin-1 could help distinguish meningiomas from histologic mimics, compared with commonly used markers. Tissue sections from 10 meningothelial meningiomas, 20 fibrous meningiomas, 10 atypical meningiomas, 7 solitary fibrous tumors of the meninges, 5 meningeal hemangiopericytomas, and 7 vestibular schwannomas were stained immunohistochemically for claudin-1, epithelial membrane antigen, S-100 protein, CD34, and glial fibrillary acidic protein. In total, 21 (53%) of 40 meningiomas were immunoreactive for claudin-1, whereas none of the other tumors were positive. In contrast, there was considerable overlap in the distribution of the other antibodies evaluated. Claudin- 1 seems to be a specific marker for meningiomas in this context. Although its sensitivity is relatively low, claudin-1 may be helpful in a panel of immunostains to distinguish meningiomas from histologic mimics. Distinguishing between meningiomas and other tumors arising in the meninges occasionally can be difficult. In particular, fibrous meningiomas are morphologically similar to other spindle cell tumors of the central nervous system, including meningeal hemangiopericytoma (HPC), solitary fibrous tumor (SFT) of the meninges, and vestibular schwannoma. However, because of differences in prognosis and patient management, proper diagnosis is critical. For example, meningeal HPC has a much higher propensity to recur and metastasize than does fibrous meningioma. 1 Although immunohistochemical analysis is helpful for differentiating among meningeal tumors, there is considerable overlap in their immunoprofiles. Epithelial membrane antigen (EMA), a marker commonly used to support the diagnosis of meningioma, is positive in approximately 30% of meningeal HPCs 2 and in a similar fraction of SFTs. 3 CD34, a marker that usually is positive in SFTs, is expressed in up to 60% of fibrous meningiomas. 4 Finally, S-100 protein, although uniformly strongly positive in schwannomas, can be detected in as many as 80% of fibrous meningiomas as well. 4 Thus, new immunohistochemical markers would be beneficial for distinguishing among these entities. Perineuriomas and meningiomas contain numerous cell junctional complexes. 5 Claudin-1 is a tight junction associated protein that recently has been shown to be expressed in perineuriomas and anaplastic meningiomas. 2,6 However, the distribution of claudin-1 in low-grade meningiomas and in other dural-based spindle cell tumors has not been elucidated fully. The purpose of this study was to determine whether immunohistochemical staining for claudin-1 could help distinguish meningiomas from potential histologic mimics, in comparison with markers commonly used in this differential diagnosis. Am J Clin Pathol 2006;125: DOI: /G659FVVBMG7U4RPQ 203

2 Hahn et al / CLAUDIN-1 IN MENINGIOMAS Materials and Methods Dural-based tumors were retrieved from the neuropathology files of Brigham and Women s Hospital, Boston, MA. Representative H&E-stained slides from each case were reviewed by two of us (H.P.H. and J.L.H.) to reach consensus diagnoses, which were based on standard and widely accepted criteria. 5 Briefly, meningothelial meningiomas are composed of small lobules of uniform bland tumor cells with eosinophilic cytoplasm and ovoid nuclei showing occasional intranuclear pseudoinclusions, surrounded by thin collagenous septa. Fibrous (fibroblastic) meningiomas contain parallel fascicles and lamellae of spindle cells, in a generally prominent collagenous matrix. Atypical meningiomas (World Health Organization grade II) are defined as meningiomas that have 4 or more mitoses per 10 high-power fields or 3 or more of the following histologic features: increased cellularity, small cells with high nuclear/cytoplasmic ratios, prominent nucleoli, sheet-like growth, and foci of geographic necrosis. 5 HPCs are highly cellular tumors composed of plump ovoid cells with scant cytoplasm, accompanied by prominent dilated, branching blood vessels. SFTs are composed of bland ovoid to spindled cells with varying cellularity and a patternless architecture, often with stromal and perivascular hyalinization and occasional HPC-like vessels. Schwannomas contain alternating cellular areas composed of elongated spindle cells with tapering nuclei showing focal nuclear palisading and palely eosinophilic cytoplasm and hypocellular lipidized areas with foci of perivascular hyalinization. All tumors with ambiguous histologic features and those for which discrepant diagnoses were made were excluded. In total, 10 meningothelial meningiomas, 20 fibrous meningiomas, 10 atypical (World Health Organization grade II) meningiomas (including 7 with predominantly spindled morphologic features), 7 SFTs of the meninges, 5 meningeal HPCs, and 7 vestibular schwannomas were selected for study. Immunohistochemical studies were performed on 4-µmthick, formalin-fixed, paraffin-embedded tissue sections. The antibodies, clones, dilutions, pretreatment conditions, and sources are listed in Table 1. The Envision Plus detection system (DAKO, Carpinteria, CA) was used for all antibodies. Appropriate positive and negative control samples were used throughout. Immunoreactivity was graded semiquantitatively as follows: 0, fewer than 5% tumor cells reactive; 1+, 5% to 25% tumor cells reactive; 2+, more than 25% to 50% tumor cells reactive; and 3+, more than 50% tumor cells reactive. For claudin-1, the intensity of staining also was graded as weak, moderate, or strong, with immunoreactivity of perineurial cells in normal peripheral nerve tissue as a reference for strong staining. The Fisher exact test or χ 2 analysis was used, as appropriate, to compare the different groups. A P value of less than.05 was considered statistically significant. The study was approved by the institutional review board of Brigham and Women s Hospital. Results The results of the study are summarized in Table 2. Claudin-1 was positive in 21 (53%) of 40 meningiomas, Table 1 Panel of Antibodies Used in the Study Antigen Clone Dilution Antigen Retrieval Source Claudin-1 Polyclonal 1:50 None Zymed, South San Francisco, CA EMA E29 1:200 None DakoCytomation, Carpinteria, CA S-100 protein Polyclonal 1:3,000 None DakoCytomation CD34 QBEnd 10 1:400 None DakoCytomation GFAP Polyclonal 1:8,500 Microwave DakoCytomation EMA, epithelial membrane antigen; GFAP, glial fibrillary acidic protein. Table 2 Immunohistochemical Staining Results * Meningothelial Fibrous Atypical Solitary Fibrous Meningeal Hemangio- Vestibular Antigen Meningioma (n = 10) Meningioma (n = 20) Meningioma (n = 10) Tumor (n = 7) pericytoma (n = 5) Schwannoma (n = 7) Claudin-1 7 (70) 8 (40) 6 (60) 0 (0) 0 (0) 0 (0) EMA 9 (90) 20 (100) 9 (90) 2 (29) 1 (20) 0 (0) S (0) 18 (90) 0 (0) 0 (0) 0 (0) 7 (100) CD34 0 (0) 8 (40) 6 (60) 7 (100) 3 (60) 0 (0) GFAP 0 (0) 0 (0) 0 (0) 0 (0) 0 (0) 4 (57) EMA, epithelial membrane antigen; GFAP, glial fibrillary acidic protein. * Data are given as number (percentage). 204 Am J Clin Pathol 2006;125: DOI: /G659FVVBMG7U4RPQ

3 Anatomic Pathology / ORIGINAL ARTICLE generally with a granular staining pattern Image 1 and Image 2. Of the meningothelial meningiomas (Image 1), 7 (70%) of 10 cases were positive for claudin-1, with 5 cases showing 3+ staining and 1 case each showing 2+ and 1+ staining (6 with moderate and 1 with weak staining). Of 20 fibrous meningiomas, 8 (40%) were positive for claudin-1 (Image 2), with 2 cases showing 3+ staining, 2 showing 2+ staining, and 4 showing 1+ staining (4 each with moderate and weak staining). Of the atypical meningiomas, 6 (60%) of 10 were positive for claudin-1, with 2 cases each showing 3+, 2+, and 1+ staining (4 with moderate and 2 with weak staining). In contrast, no immunoreactivity for claudin-1 was seen in any of the other tumor types. Positive staining for claudin- 1 was significantly more frequent in meningiomas than in the other neoplasms examined (P <.05). As expected, claudin-1 was expressed in the (perineurial) capsule of a schwannoma and was negative in smooth muscle, normal brain tissue, and fibroblasts. The tumors also were stained with other markers commonly used in the differential diagnosis of dural-based neoplasms. There was significant overlap in the immunoprofiles of these tumors (Table 2). Although 9 (90%) of 10 meningothelial meningiomas, all 20 fibrous meningiomas (100%), and 9 (90%) of 10 atypical meningiomas expressed EMA (Image 2), staining for EMA also was detected in 2 (29%) of 7 SFTs (Image 2) and in 1 (20%) of 5 HPCs. Of note, both the meningothelial meningioma and the atypical meningioma that failed to stain for EMA were positive for claudin-1 (3+ and 2+, respectively). Immunoreactivity for S- 100 protein was observed in 7 (100%) of 7 vestibular schwannomas, but also was detected in 18 (90%) of 20 fibrous meningiomas examined (11 cases with 3+, 5 with 2+, and 2 with 1+ staining). All 7 SFTs (100%) were positive for CD34 (5 showing 3+ and 2 showing 1+ staining; Image 2), as were 3 (60%) of 5 HPCs (each 3+). However, 8 (40%) of 20 fibrous meningiomas (3 cases with 2+ and 5 with 1+ staining) and 6 (60%) of 10 atypical meningiomas (3 cases each with 3+ and 1+ staining) also showed immunoreactivity for CD34 (Image 2). Discussion Meningiomas, in particular the fibrous variant, at times can be difficult to distinguish from meningeal HPCs, SFTs, and vestibular schwannomas. Although a panel of immunohistochemical markers can be used to differentiate among these entities, there is significant overlap in their immunoprofiles. Claudin-1 is a tight junction associated protein that recently has been shown to be expressed in anaplastic meningiomas. 2 The purpose of this study was to compare claudin-1 staining with other markers commonly used to differentiate meningiomas from histologic mimics. In accordance with the results of previous studies, we found EMA and CD34 staining in meningiomas and in other spindle cell tumors of the central nervous system. EMA was positive in 95% of meningiomas and in 29% and 20% of meningeal SFTs and HPCs, respectively. Similarly, immunoreactivity for CD34 was observed in all SFTs and 60% of HPCs, as well as in 40% of fibrous meningiomas and 60% of atypical meningiomas. S-100 protein was detected not only in vestibular schwannomas, but also in 90% of fibrous meningiomas, often with diffuse staining. In contrast, claudin-1 expression was specific for meningiomas in this context. Claudin-1 was detected in 53% of meningiomas but not in any of the other tumor types. A B Image 1 Claudin-1 expression in a meningothelial meningioma. A, A meningothelial meningioma showing prominent whorls and nuclear pseudoinclusions (H&E, 400). B, Immunostaining for claudin-1 ( 400). Am J Clin Pathol 2006;125: DOI: /G659FVVBMG7U4RPQ 205

Hahn et al / CLAUDIN-1 IN MENINGIOMAS A B C D Image 2 Comparison between fibrous meningioma (A-D) and solitary fibrous tumor (SFT) of the meninges (E-H).

In contrast, other commonly used markers are positive in both tumor types. Immunoreactivity for epithelial membrane antigen is seen in both fibrous meningioma (C, 400) and a subset of SFT (G, 400).

For example, fibrous meningiomas and meningeal SFTs are composed of bundles of spindle cells with prominent background collagen fibers.

Unfortunately, the markers widely used in clinical practice are not tumor-specific.

4 Hahn et al / CLAUDIN-1 IN MENINGIOMAS A B C D Image 2 Comparison between fibrous meningioma (A-D) and solitary fibrous tumor (SFT) of the meninges (E-H). The histologic appearances of fibrous meningioma (A, H&E, 400) and SFT (E, H&E, 400) are similar. Claudin-1 is expressed in fibrous meningioma (B, 400), but not in SFT (F, 400). In contrast, other commonly used markers are positive in both tumor types. Immunoreactivity for epithelial membrane antigen is seen in both fibrous meningioma (C, 400) and a subset of SFT (G, 400). Positive staining for CD34 is common in fibrous meningioma (D, 400) and SFT (H, 400). There is significant overlap in the histologic appearances of spindle cell tumors arising in the meninges. For example, fibrous meningiomas and meningeal SFTs are composed of bundles of spindle cells with prominent background collagen fibers. 7 Particularly in cases in which the diagnosis must be made based on the evaluation of small tumor fragments, immunohistochemical analysis can be very helpful. Unfortunately, the markers widely used in clinical practice are not tumor-specific. For example, EMA is positive in 50% to 100% of meningiomas 8-11 and in normal meninges 10 and often is used to confirm the diagnosis of meningioma. However, approximately 30% of meningeal HPCs 2 and SFTs 3 and occasional schwannomas 10 also are immunoreactive for EMA. Although diffuse S-100 protein staining can be used to support the diagnosis of schwannoma, fibrous meningiomas also frequently strongly express this antigen. 4,7 Finally, there also is significant overlap in the distribution of CD34 expression in SFTs, meningeal HPCs, and fibrous meningiomas. Although SFTs of the meninges and meningeal HPCs often show strong positivity for CD34, 4,7 more than 50% of fibrous meningiomas also express CD34. 4,12 As an aside, because HPCs of soft tissue are morphologically indistinguishable from and show similar immunophenotypic features to the cellular areas of SFTs, these 2 tumor 206 Am J Clin Pathol 2006;125: DOI: /G659FVVBMG7U4RPQ

Anatomic Pathology / ORIGINAL ARTICLE E F G H types now are widely regarded to be closely related (if not synonymous).

Similar to their soft tissue counterparts, meningeal HPCs indeed may also represent uniformly cellular (and malignant) examples of SFT.

2 In striking similarity to our results, the authors detected immunoreactivity for claudin-1 in 7 (54%) of 13 anaplastic (grade III) meningiomas, whereas we found that 50% of grade I

In contrast, Rajaram et al 2 reported that 2 (13%) of 15 meningeal HPCs were focally positive for claudin-1, whereas all cases of meningeal HPCs were negative for this marker in our study.

5 Anatomic Pathology / ORIGINAL ARTICLE E F G H types now are widely regarded to be closely related (if not synonymous). 3 The findings in the present study provide further evidence for the similar antigenic expression patterns in meningeal examples of these tumors. Similar to their soft tissue counterparts, meningeal HPCs indeed may also represent uniformly cellular (and malignant) examples of SFT. Our findings with regard to claudin-1 expression in meningiomas are comparable to those reported recently by Rajaram et al. 2 In striking similarity to our results, the authors detected immunoreactivity for claudin-1 in 7 (54%) of 13 anaplastic (grade III) meningiomas, whereas we found that 50% of grade I meningiomas and 60% of grade II meningiomas were positive for claudin-1. In contrast, Rajaram et al 2 reported that 2 (13%) of 15 meningeal HPCs were focally positive for claudin-1, whereas all cases of meningeal HPCs were negative for this marker in our study. The discrepancy between these results may be due to differences in immunohistochemical staining techniques. Specifically, in contrast with the methods used by Rajaram et al, 2 we did not use antigen retrieval in our study. One additional previous study examined claudin-1 expression in meningiomas; however, this study was published only in abstract form. 13 Bhattacharjee et al 13 reported immunoreactivity for claudin-1 in 17 (85%) of 20 meningiomas, but the Am J Clin Pathol 2006;125: DOI: /G659FVVBMG7U4RPQ 207

6 Hahn et al / CLAUDIN-1 IN MENINGIOMAS histologic type and grade were not specified. These authors also reported negative staining for claudin-1 in all 10 schwannomas they examined, similar to our results. However, our study is the first to examine claudin-1 expression in SFTs of the meninges, which can be difficult to distinguish from meningiomas with spindled morphologic features; all cases were negative. Meningiomas are thought to originate from the cell clusters capping the arachnoid villi. 14 In addition to other morphologic and ultrastructural similarities, meningiomas and arachnoidal cap cells have numerous cell adhesion structures. However, although tight junctions are seen in arachnoid cells, the cell adhesion structures in meningiomas are desmosome-like. Claudins are understood to be components of tight junctions. So how can claudin-1 expression in meningiomas be explained? There is increasing evidence that tumor cells can show aberrant expression or abnormal cellular localization of claudin-1. For example, despite the fact that tight junctions are not identified by electron microscopy in the spindle cell component of biphasic and monophasic synovial sarcomas, claudin-1 and other tight junction proteins recently have been shown to be expressed in these areas. 15 In addition, Schuetz et al 16 recently demonstrated staining for claudin-1 and other tight junction associated antigens in many cases of Ewing sarcoma/primitive neuroectodermal tumor, although well-formed tight junctions are not detected ultrastructurally in these tumors. Finally, in colorectal adenocarcinomas, claudin-1 expression can be found not only at cell-cell boundaries (as anticipated in epithelial cells) but also in the cytoplasm of tumor cells. 17 Claudin-1 seems to be a specific marker for meningiomas in comparison with other spindle cell tumors arising in the meninges. Although its sensitivity is relatively low, claudin-1 may be helpful when used in a panel of immunostains to distinguish meningiomas from histologic mimics. From the Department of Pathology, Brigham and Women s Hospital and Harvard Medical School, Boston, MA. Presented in part at the 94th annual meeting of the United States and Canadian Academy of Pathology, San Antonio, TX, February 26-March 4, Address reprint requests to Dr Hornick: Dept of Pathology, Brigham and Women s Hospital, 75 Francis St, Boston, MA References 1. Guthrie BL, Ebersold MJ, Scheithauer BW, et al. Meningeal hemangiopericytoma: histopathological features, treatment, and long-term follow-up of 44 cases. Neurosurgery. 1989; 25: Rajaram V, Brat DJ, Perry A. Anaplastic meningioma versus meningeal hemangiopericytoma: immunohistochemical and genetic markers. Hum Pathol. 2004;35: Fletcher CDM, Unni KK, Mertens F, eds. Pathology and Genetics of Tumours of Soft Tissue and Bone. Lyon, France: IARC Press; World Health Organization Classification of Tumours. 4. Perry A, Scheithauer BW, Nascimento AG. The immunophenotypic spectrum of meningeal hemangiopericytoma: a comparison with fibrous meningioma and solitary fibrous tumor of meninges. Am J Surg Pathol. 1997;21: Kleihues P, Cavenee WK, eds. Pathology and Genetics of Tumours of the Nervous System. Lyon, France: IARC Press; World Health Organization Classification of Tumours. 6. Folpe AL, Billings SD, McKenney JK, et al. Expression of claudin-1, a recently described tight junction-associated protein, distinguishes soft tissue perineurioma from potential mimics. Am J Surg Pathol. 2002;26: Carneiro SS, Scheithauer BW, Nascimento AG, et al. Solitary fibrous tumor of the meninges: a lesion distinct from fibrous meningioma: a clinicopathologic and immunohistochemical study. Am J Clin Pathol. 1996;106: Meis JM, Ordonez NG, Bruner JM. Meningiomas: an immunohistochemical study of 50 cases. Arch Pathol Lab Med. 1986;110: Theaker JM, Gatter KC, Esiri MM, et al. Epithelial membrane antigen and cytokeratin expression by meningiomas: an immunohistological study. J Clin Pathol. 1986;39: Winek RR, Scheithauer BW, Wick MR. Meningioma, meningeal hemangiopericytoma (angioblastic meningioma), peripheral hemangiopericytoma, and acoustic schwannoma: a comparative immunohistochemical study. Am J Surg Pathol. 1989;13: Artlich A, Schmidt D. Immunohistochemical profile of meningiomas and their histological subtypes. Hum Pathol. 1990;21: Suzuki SO, Fukui M, Nishio S, et al. Clinicopathological features of solitary fibrous tumor of the meninges: an immunohistochemical reappraisal of cases previously diagnosed to be fibrous meningioma or hemangiopericytoma. Pathol Int. 2000;50: Bhattacharjee M, Adesina AM, Goodman C, et al. Claudin-1 expression in meningiomas and schwannomas: possible role in differential diagnosis [abstract]. J Neuropathol Exp Neurol. 2003;62: Yamashima T. On arachnoid villi and meningiomas: functional implication of ultrastructure, cell adhesion mechanisms, and extracellular matrix composition. Pathol Oncol Res. 1996;2: Billings SD, Walsh SV, Fisher C, et al. Aberrant expression of tight-junction related proteins ZO-1, claudin-1, and occludin in synovial sarcoma: an immunohistochemical study with ultrastructural correlation. Mod Pathol. 2004;17: Schuetz AN, Rubin BP, Goldblum JR, et al. Intercellular junctions in Ewing sarcoma/primitive neuroectodermal tumor: additional evidence of epithelial differentiation. Mod Pathol. 2005;18: Miwa N, Furuse M, Tsukita S. Involvement of claudin-1 in the beta-catenin/tcf signaling pathway and its frequent upregulation in human colorectal cancers. Oncol Res. 2000;12: Am J Clin Pathol 2006;125: DOI: /G659FVVBMG7U4RPQ

A 25 year old female with a palpable mass in the right lower quadrant of her abdomen

May 2016 A 25 year old female with a palpable mass in the right lower quadrant of her abdomen Contributed by: Paul Ndekwe, MD, Resident Physician, Indiana University School of Department of Pathology and