Immunohistochemical characterization of mesothelioma in 6 large felids

Abstract

Mesothelioma has been reported frequently in large felids. These neoplasms present a diagnostic challenge given their highly variable morphology that mimics carcinomas or sarcomas at different locations. Our goal was to characterize mesotheliomas morphologically and immunohistochemically to determine if a panel of antibodies could be used to more accurately support the diagnosis of these neoplasms in large felids. Mesotheliomas from 6 large felids, including 4 clouded leopards, 1 Bengal tiger, and 1 cheetah, were immunohistochemically labeled for vimentin, E-cadherin, pancytokeratin, Wilms tumor 1 (WT1), MUC-1, and calretinin. The mesotheliomas of the 4 clouded leopards and the tiger were of the epithelial subtype; the mesothelioma from the cheetah was biphasic. All 6 mesotheliomas had strong immunohistochemical labeling for vimentin, E-cadherin, and pancytokeratin. All cases had cytoplasmic labeling for WT1, and 2 also had nuclear labeling. The 3 mesotheliomas with distinct papillary fronds were weakly positive for MUC-1. These and one other epithelial mesothelioma were also positive for calretinin. Our study demonstrates that the morphologic and immunohistochemical phenotypes of mesothelioma that have been identified in humans and domestic species can occur in large felids, and a panel of pancytokeratin, vimentin, WT1, and calretinin can be utilized to support the diagnosis of these neoplasms.

Mesotheliomas are malignant neoplasms that arise from mesothelial cells that cover the pleural, peritoneal, and pericardial surfaces. Such neoplasms present a diagnostic challenge, given that they can be difficult to distinguish from reactive mesothelium or from various types of sarcomas or carcinomas. Mesotheliomas occur uncommonly in domestic species,^2,7–9,19 but have been reported more commonly in humans in association with asbestos exposure. ¹³ Mesotheliomas have also been reported in large felids from zoological collections^{3,11,12,18,20}; however, little immunohistochemical characterization of these neoplasms has been performed. We characterized the gross, histologic, and immunohistochemical features of mesotheliomas in 6 large felids from zoological institutions in the United States.

There are 3 main histologic types of mesothelioma: epithelial, sarcomatous, and biphasic. Depending on the morphologic subtype and their location, mesotheliomas must be differentiated from sarcomas or carcinomas arising in similar locations. Immunoreactivity for both epithelial and mesenchymal markers is characteristic of mesothelioma, and panels of antibodies are most useful in the diagnosis of mesotheliomas.^{1,2,5,14,16,17} In human medicine, antibodies that are commonly expressed by mesotheliomas include carcinoembryonic antigen (CEA), Ber-EP4, B72.3, CD15, MOC-31, thyroid transcription factor 1, cytokeratin (CK) 5/6, calretinin, HBME-1, thrombomodulin, Wilms tumor 1 (WT1), mesothelin, D2-40, and podoplanin.^13,15,16 In domestic cats, antibodies that have been demonstrated to be expressed by mesotheliomas are vimentin, CK AE1/AE3, HBME-1, CK 5/6, CEA, and E-cadherin. ²

Six cases of mesothelioma in large felids were identified in archived records from Northwest ZooPath (Monroe, WA, USA; Table 1). The mesotheliomas originated from 4 clouded leopards (Neofelis nebulosa), 1 Bengal tiger (Panthera tigris tigris), and 1 cheetah (Acinonyx jubatus). There were 4 females and 2 males, aged 12–21 y old. All mesotheliomas occurred as primary intrathoracic masses (Figs. 1, 2), and one metastasized to thoracic lymph nodes, the bone of the rib, adrenal glands, kidney, and lung. Pleural effusions were reported for 4 cases.

Table 1.

Case demographics and autopsy findings.

Case	Species	Sex	Age (y)	Site	Autopsy findings
1	Clouded leopard	Male	14	Mediastinum	Thoracic mass occupying the mediastinum; pleural and peritoneal serosanguineous effusion
2	Clouded leopard	Male	13	Pleura	Pleura covered in masses with a granular appearance; nodules on pericardium; nodule in lung
3	Clouded leopard	Female	18	Pleura	Pleural effusion, cauliflower-like lesions on the pleura and diaphragm; scattered nodules on pleura, lung, and heart
4	Clouded leopard	Female	12	Pleura	Pleural effusion and mass on left caudal lung lobe; small masses on all surfaces of thoracic cavity
5	Bengal tiger	Female	21	Mediastinum, lung	Nodular mass in the thoracic inlet obstructing the esophagus; similar nodules in the lungs
6	Cheetah	Female	12	Mediastinum, pleura, pericardium	Pleural effusion, mediastinal, pleural, pulmonary, pericardial masses; distant metastases present

Figures 1–10.

Epithelioid mesothelioma in a clouded leopard, case 1. Figure 1. The mediastinum and visceral pleura are covered by large numbers of pale-tan nodules. Figure 2. The parietal pleura is covered by myriad individual-to-coalescing, pale-tan to red nodules. Figure 3. Epithelioid mesothelioma in a clouded leopard, case 2. Neoplastic cells arranged in papillary fronds supported by fibrovascular stroma. H&E. Figures 4, 5. Biphasic mesothelioma in a cheetah, case 6. Figure 4. Neoplastic cells form a solid sheet. H&E. Figure 5. Spindloid neoplastic cells form sheets in some areas. H&E. Figures 6, 7. Epithelioid mesothelioma in a clouded leopard, case 2. Figure 6. Neoplastic cells with strong cytoplasmic labeling for vimentin. Red chromogen, hematoxylin counterstaining. Figure 7. Neoplastic cells with strong cytoplasmic labeling for pancytokeratin. DAB, hematoxylin counterstaining. Figure 8. Biphasic mesothelioma in a cheetah, case 6. Neoplastic spindle cells with strong cytoplasmic labeling for pancytokeratin. DAB, hematoxylin counterstaining. Figure 9. Epithelioid mesothelioma in a tiger, case 5. Neoplastic cells have cytoplasmic and nuclear labeling. WT1. DAB, hematoxylin counterstaining. Figure 10. Epithelioid mesothelioma in a clouded leopard, case 2. Neoplastic cells with cytoplasmic labeling for calretinin. DAB, hematoxylin counterstaining.

Hematoxylin and eosin (H&E)-stained sections were reviewed, to reaffirm the histologic diagnosis. The paraffin-embedded tissue blocks were sent to Michigan State University (Lansing, MI, USA) where representative areas from each case were selected from the original donor block and transferred into a single paraffin block that combined all 6 cases. A section from the block stained with H&E was examined to confirm that neoplastic cells were present from each case. Immunohistochemistry (IHC) for vimentin, pancytokeratin, calretinin, WT1, E-cadherin, and epithelial membrane antigen (EMA, MUC-1) was performed on serial sections from the composite tissue block. IHC was performed for calretinin, E-cadherin, pancytokeratin, and WT1 (BOND-MAX automated staining system, Bond polymer detection system, Vision BioSystems; Leica) with 3,3′-diaminobenzidine (DAB) as the chromogen; for MUC-1 (Dako link 48 automated staining system, peroxidase-conjugated EnVision polymer detection system; Agilent Technologies); and for vimentin (Discovery Ultra automated staining system, UltraMap alkaline phosphatase red detection system; Ventana Medical Systems; Table 2). Non-neoplastic mesothelium from a large felid was used as the positive control for all antibodies. Cases were considered positive for vimentin, pancytokeratin, MUC-1, and calretinin if there was cytoplasmic labeling of neoplastic cells. Cases were considered positive for E-cadherin if there was membranous labeling of neoplastic cells. Cases were considered positive for WT1 if there was cytoplasmic labeling with or without nuclear labeling.

Table 2.

List of antibodies with source, clone, concentration used, retrieval information, and detection method.

Antibody	Host	Source	Clone	Antigen retrieval	Dilution	Chromogen
Vimentin	Mouse	Roche, 790-2917	V9	CC1*	RTU†	UltraMap red
E-cadherin	Mouse	BD Biosciences, 610181	36/E-cadherin	ER1‡	1:300	DAB
Pancytokeratin	Mouse	Agilent, M0821	MNF116	E2§	1:800	DAB
Wilms tumor 1	Rabbit	Abcam, ab15249	Polyclonal	ER2¦	1:500	DAB
Calretinin	Mouse	Agilent, M7245	DAK-Calret 1	ER1	1:100	DAB
MUC-1	Rabbit	LSBio, LS-C30532	Polyclonal	Low PT#	1:500	DAB

^*Heat retrieval for 4 min (CC1 = cell conditioning 1, Discovery Ultra; Ventana).

^†Ready to use.

^‡Heat retrieval for 20 min (ER1 = Bond epitope retrieval solution 1; Leica Biosystems).

^§Enzymatic antigen retrieval for 20 min (E2; Agilent Technologies).

^¦Heat retrieval for 20 min (ER2 = Bond epitope retrieval solution 2; Leica Biosystems).

^#Low pH retrieval for 20 min (Dako PT link; Agilent Technologies).

Histologically, the mesotheliomas in the clouded leopards and the tiger (cases 1–5) were epithelial mesotheliomas. The neoplastic cells in these cases were arranged in tubules and papillary fronds supported by variable amounts of fibrovascular stroma. The cells were polygonal, with moderate amounts of eosinophilic cytoplasm, and distinct cell borders. The nuclei were round-to-ovoid, had finely stippled to vesiculate chromatin, and 1–3 nucleoli. Two mesotheliomas in clouded leopards (cases 2, 3) and the mesothelioma in the tiger (case 5) had large prominent papillary fronds (Fig. 3).

The mesothelioma in the cheetah (case 6) was biphasic and had an area of solid epithelial neoplastic cells (Fig. 4) and sarcomatous neoplastic cells (Fig. 5). In the solid area, the cells were arranged in a dense sheet with scant fibrovascular stroma. These cells were polygonal, had a moderate amount of eosinophilic cytoplasm, and had distinct cell borders. The nuclei were round-to-ovoid, finely stippled to vesiculate, and had 1–3 nucleoli. Anisokaryosis in this area was moderate, and there were multinucleate cells. In the sarcomatous area, the cells were arranged in dense bundles. The cells were spindloid, had a small amount of eosinophilic cytoplasm, and had indistinct cell borders. The nuclei were ovoid-to-elongate, finely stippled, and had 1 or 2 nucleoli. Anisokaryosis was mild in this area.

The diagnosis of mesothelioma was confirmed with positive immunoreactivity for vimentin, pancytokeratin, WT1, and E-cadherin in all cases (Table 3). These cases had diffuse positive cytoplasmic labeling for vimentin (Fig. 6), pancytokeratin (Figs. 7, 8), and WT1 (Fig. 9), and membranous labeling for E-cadherin along the surface of the neoplasm; in case 6, there was membranous labeling of cells invading the stroma and impinging on vessels. Two epithelial mesotheliomas in a clouded leopard and in the tiger (cases 4, 5) also had positive nuclear labeling for WT1 (Fig. 9). The 3 mesotheliomas with distinct papillary fronds (cases 2, 3, 5) were weakly positive for MUC-1 with patchy superficial cytoplasmic labeling, and these mesotheliomas, in addition to another epithelial tumor (case 4), were positive for calretinin with patchy cytoplasmic labeling (Fig. 10).

Table 3.

Histologic and immunohistochemical findings in mesotheliomas in 6 large felids.

Case	Histologic type	Vimentin	Pancytokeratin	WT1	E-cadherin	Calretinin	MUC-1
1	Epithelioid, tubulopapillary	+	+	+	+	−	−
2	Epithelioid, tubulopapillary	+	+	+	+	+	+
3	Epithelioid, tubulopapillary	+	+	+	+	+	−
4	Epithelioid, tubulopapillary	+	+	+	+	+	+
5	Epithelioid, tubulopapillary	+	+	+	+	+	+
6	Biphasic	+	+	+	+	−	−

Given that the histologic features of mesothelioma are variable and often do not allow differentiation from either sarcomas or carcinomas, IHC is commonly needed to confirm the diagnosis. The infrequency of reported cases of mesothelioma in veterinary medicine poses an additional challenge to the diagnostician. Positive labeling for epithelial and mesenchymal markers is characteristic of mesotheliomas.^{1,2,5,14,16,17} Positive labeling for pancytokeratin in mesenchymal mesotheliomas helps to distinguish these cases from sarcomas or a marked scirrhous response. Although labeling for pancytokeratin and vimentin in epithelial mesotheliomas aids in distinguishing them from carcinomas, it is not a feature unique to mesotheliomas and can be observed commonly in some carcinomas, such as pulmonary carcinoma, a primary differential of thoracic mesotheliomas. ⁴ Therefore, additional antibodies are commonly needed to further support a diagnosis of mesothelioma or to exclude the specific primary differentials that may vary based on location and morphology of the mesothelioma.

All mesotheliomas in our series were also positive for E-cadherin, and this antibody may be used as an alternative or additional epithelial marker. MUC-1 (EMA) is another marker of epithelial lineage and had been proposed in human medicine to differentiate malignant from benign mesothelial proliferations. Although increased expression of MUC-1 has been reported in malignant mesothelial proliferations, the immense variation in expression between benign and malignant mesothelial lesions made MUC-1 a less useful marker to identify malignant mesotheliomas.^14,16 In our study, MUC-1 labeled the 3 mesotheliomas with the most differentiated tubulopapillary pattern. Although MUC-1 may be useful as part of a panel to help distinguish mesotheliomas from sarcomas, it is of little benefit to differentiate mesotheliomas from carcinomas given that the reaction patterns of different carcinomas in large felids are mainly unknown.

Two of the more discriminating antibodies in our study included WT1 and calretinin. The WT1 gene is a tumor suppressor gene originally discovered as a mutation in Wilms tumors in humans. In humans, in addition to labeling Wilms tumor, this antibody has been proven useful to distinguish neoplastic mesothelial cells.^{10,13,14,16,17} All 6 mesotheliomas in our study had cytoplasmic labeling for WT1, and 2 cases had nuclear labeling, confirming that this antibody may be useful in the diagnosis of mesotheliomas in large felids. In the human literature, it is reported that nuclear labeling for WT1 is required to differentiate mesothelioma from carcinoma. Cytoplasmic labeling can also be observed in mesothelioma; however, this is less specific, and some adenocarcinomas may have cytoplasmic labeling. ¹⁰ Calretinin is a calcium-binding protein that plays a role in cell-signaling, buffering intracellular calcium levels, and neuronal excitability. ⁶ In addition to expression in neuronal tissue, it is also expressed in mesothelium. This antibody is considered one of the more sensitive and specific markers for mesothelioma in human medicine and is routinely included in diagnostic mesothelioma panels.^6,10,16,17 Although in our study, only 4 of the 6 cases were positive, especially in neoplasms that were also positive for pancytokeratin and vimentin, calretinin essentially confirmed the diagnosis of mesothelioma. Interestingly, the only biphasic mesothelioma in our series was negative for calretinin. The other negative mesothelioma was a less well-differentiated tubulopapillary mesothelioma.

Mesotheliomas represent a diagnostic challenge, even for experienced pathologists. Although morphologic features are important to establish mesothelioma as a differential diagnosis, IHC is essential to confirm the diagnosis. Depending on the morphologic subtype of mesothelioma and its location, differentials may vary, and different panels of antibodies will be necessary to exclude such differentials. A panel composed of pancytokeratin, vimentin, WT1, and calretinin may be most helpful in aiding the diagnosis of mesothelioma in large felids.