Malignant Myopericytoma-like Tumor in a Fischer Rat

Abstract

Myopericytoma is a perivascular tumor that has been recently described in humans, but not in laboratory rodents. The authors encountered an intra-abdominal tumor resembling human malignant myopericytoma in a Fischer rat. Grossly, the tumor was found as two brown-colored masses located in the mesentery of rectum. Microscopically, the tumor was composed of oval to spindle-shaped cells, which were arranged in sheets around numerous thin-walled branching vessels and partly showed a concentric perivascular growth pattern. Mitoses were frequently seen, and the tumor cells showed a local invasion. Immunohistochemically, the tumor cells were strongly positive for alpha-smooth muscle actin and weakly positive for vimentin and desmin. Ultrastructurally, the tumor cells had dendritic processes, actin-like thin filaments with dense bodies, basement membranes, hemidesmosomes, and micropinocytotic vesicles. These findings suggest that the most appropriate term for diagnosis of the present case could be a malignant myopericytoma.

Introduction

Myopericytoma is a recently described perivascular tumor in humans. It is characterized histologically by concentric perivascular proliferation of ovoid to spindle-shaped cells showing apparent differentiation toward perivascular myoid cells or myopericytes (Folpe and McMenamin 2002). It is thought that myopericytes have intermediate features between pericytes and vascular smooth muscle cells (Dictor et al. 1992). The term myopericytoma was originally proposed by Requena et al. in 1996 as an alternative designation for solitary myofibroma with myopericytic differentiation (Requena et al. 1996), and it has now been endorsed by the World Health Organisation (WHO) (Folpe and McMenamin 2002). Myopericytoma generally arises in subcutaneous tissue (Folpe and McMenamin 2002), and very few cases have been reported in other locations, such as intracranial region (Rousseau et al. 2005) and thoracic spine (Cox and Giltman 2003). Both benign and malignant myopericytomas have been recognized in humans (McMenamin and Fletcher 2002; Mentzel et al. 2006). In other animal species, myopericytoma has been reported in dogs (Avallone et al. 2007). However, no report on myopericytomas has been documented in laboratory rodents (Ruben et al. 1997). We encountered an intra-abdominal tumor resembling human malignant myopericytoma in a Fischer rat. This report described the histological, immunohistochemical, and ultrastructural findings of the tumor.

Case Report

The affected animal was a 105-week-old female Fischer (F344/DuCrj) rat (Charles River Japan, Kanagawa, Japan) that was allocated to a high-dose group in a two-year carcinogenicity study of DDT (Harada et al. 2003). The rat was handled during the study in accordance with the Guidelines for Animals Experimentation issued by the Japanese Association for Laboratory Animal Science (1987). The rat was deeply anesthetized with ether and then euthanized by exsanguination on a moribund condition due to marked tremor caused by DDT. At necropsy, two brown-colored neoplastic masses, 35×30×20 mm and 20×10×10 mm in size, were found in the mesentery of rectum (Figure 1). The cut surfaces were firm and white in color and contained areas of necrosis and hemorrhage. Tissues of these neoplastic masses were fixed in 10% neutral-buffered formalin and embedded in paraffin. Paraffin sections (5 μm) were stained with hematoxylin and eosin (H&E). Eight-micrometer frozen sections of neoplastic masses were stained with oil red O. Additional sections were subjected to immunohistochemistry with an avidin-biotin complex method (Vector Laboratories, Burlingame, CA, USA) using the following primary antibodies: anti-swine vimentin monoclonal antibody (V9, ×50), anti-chicken desmin polyclonal antibody (fully diluted), anti-cow S-100 protein polyclonal antibody (x200), anti-cow glial fibrillary acidic protein (GFAP) polyclonal antibody (fully diluted), anti-human chromogranin A polyclonal antibody (×100), anti-human factor VIII-related antigen (f-VIII) polyclonal antibody (fully diluted), and anti-proliferating cell nuclear antigen (PCNA) monoclonal antibody (PC10, ×5,000). A horseradish peroxidase-labeled anti-human alpha-smooth muscle actin (aSMA) monoclonal antibody (1A4) was also used. The primary antibody of f-VIII was obtained from BioGenex (San Ramon, CA, USA), while others were from DAKO (Glostrup, Denmark). The sections were pre-treated with microwave irradiation in 0.1 M citrate buffer at pH 6.0 for ten minutes for desmin, GFAP, and at pH 3.5 for ten minutes for vimentin and chromogranin A. For f-VIII, the section was pretreated with 0.05% pronase solution (DAKO) at room temperature for seven minutes. Immunohistochemical products were visualized with the substrate 3,3′-diaminobenzidine and counterstained with hematoxylin. For electron microscopy, the formalin-fixed masses were postfixed in 1% solution of osmium tetroxide and routinely processed for epon embedding. Ultrathin sections were stained with uranyl acetate and lead citrate and examined with a transmission electron microscopy (JEM-1200EX II, JEOL, Japan).

Microscopically, both neoplastic masses had similar morphological features and were composed of oval to spindle-shaped cells, which were arranged in sheets around numerous thin-walled branching vessels (Figure 2A) and partly showed concentric perivascular proliferation (Figure 2B). The tumor cells had round to oval nuclei and pale eosinophilic or vacuolated (oil red O-negative) scant cytoplasms without distinct cell borders (Figure 2C). There were some foci of whorled arrangements that were embedded in myxoid matrix. Mitoses were seen frequently (Figure 2C), and the tumor cells showed a local invasion into the surrounding connective tissues, with evidence of hemorrhage and hemosiderin deposition, indicating destruction of blood vessels (Figure 2D). Focal necrosis and hemorrhage were also seen in the central part of tumors. These findings suggest that the tumors are malignant state, although evidence of any metastases or invasions within blood vessels or lymph nodes was not observed. Immunohistochemically, tumor cells were strongly positive for aSMA (Figure 3A), weakly positive for vimentin and desmin (Figure 3B), and negative for S-100 protein, GFAP, chromogranin A, and f-VIII. The nuclei of many tumor cells were strongly positive for PCNA. These histopathological features are similar to myopericytoma, which is typically composed of ovoid to spindle-shaped aSMA-positive myoid-appearing cells with a striking tendency for concentric perivascular growth (Granter, Badizadegan, and Fletcher 1998; McMenamin and Fletcher 2002). Ultrastructurally, tumor cells had dendritic interlocking cytoplasmic extensions that formed interdigitating processes with the adjacent cells (Figure 4A). The tumor cells possessed actin-like thin filaments with dense bodies, and many micropinocytotic vesicles were situated inside the plasma membrane (Figure 4B). There were occasional hemidesmosomes and partial basement membranes between the tumor cells. Intercellular spaces and collagen fibrils were not found. These ultrastructural features were consistent with those of myopericytes, which are intermediate cells between pericytes and vascular smooth muscle cells. Pericytes that may differentiate into smooth muscle cells have an elongated cell body with long cell processes and encircle the capillary wall (Hirschi and D’Amore 1996; Requena et al. 1996). They share the morphologic properties of cytoplasmic actin microfilaments, dense plaques, micropinocytotic vesicles, and complete investment by a basal lamina with smooth muscle cells (Dictor et al. 1992). Smooth muscle cells differ morphologically from pericytes by lack of dendritic processes and the presence of dense bodies among more numerous actin filaments. It is thought that myopericytes as transitional cell forms have short dendritic processes, as noted in pericytes, as well as dense bodies, which are characteristic of smooth muscle cells (Dictor et al. 1992). Overall, the morphological, immunohistochemical, and ultrastructural features observed in the present tumors were similar to those seen in human malignant myopericytomas.

Most perivascular tumors have been diagnosed as hemangiopericytomas. In this case, tumor cells arranged in sheets around branching vessels with hemangiopericytoma-like appearance were present. However, hemangiopericytomas are now recognized to include several different tumors, which share the presence of thin-walled branching blood vessels histologically, and they are no longer considered a specific entity. According to the current classification organized by WHO Classification of Tumors, myopericytoma showing clear evidence of myoid/myopericytic differentiation is newly classified into categories of pericytic (perivascular) tumors instead of conventional hemangiopericytoma. Hemangiopericytoma is now reclassified into fibroblastic/myofibroblastic tumors as cellular variant of solitary fibrous tumor, which is mostly negative for actin isoforms (Folpe and McMenamin 2002; Gengler and Guillou 2006).

Myopericytoma is recognized as a tumor derived from myopericyte and shares features with other perivascular myoid tumors, such as myofibroma, glomus tumors, and smooth muscle tumors. Differential diagnosis of perivascular tumors is presented in Table 1. Myofibroma forms a morphological continuum with myopericytoma, because myopericytes share many features with myofibroblasts (Dictor et al. 1992). Both myopericytes and myofibroblasts are uniformly positive for aSMA and vimentin and are sometimes weakly positive for desmin (Dray et al. 2006; Folpe and McMenamin 2002). A typical growth pattern of tumor cells is of help in distinguishing myopericytoma from myofibroma. Myopericytoma is histologically characterized by concentric perivascular proliferation of myopericytes, while myofibroma shows a biphasic growth pattern composed with myofibroblastic fascicles and hemangiopericytoma-like appearance. The fascicles of myofibroblasts with elongated nucleus and abundant eosinophilic cytoplasm tend to be arranged around a central area composed of more primitive round to polygonal cells with a hemangiopericytoma-like vascular pattern (Dray et al. 2006). The tumor cells observed in our case showed perivascular concentric growth, not biphasic pattern. Glomus tumors are well-defined perivascular tumors composed of cuboidal cells with abundant cytoplasms and distinct cell borders (Folpe and McMenamin 2002). They do not usually show striking concentric perivascular growth pattern. The current case is also distinguished from angioleiomyomatous tumor, which is composed predominantly of elongated spindle-shaped cells surrounding thick-walled vessels. Malignant myopericytoma and leiomyosarcoma are often confused, as both are composed of cells that showed myoid differentiation with evidence of the immunopositivity for aSMA. Leiomyosarcomas often show positive staining with desmin, whereas tumor cells in myopericytoma are negative or sometimes weakly positive for desmin (McMenamin and Fletcher 2002). Therefore, immunohistochemical staining is not of help in the distinction of these two tumors. We ruled out leiomyosarcoma as a diagnosis in our case, because it generally has a more fascicular pattern than myopericytoma and is composed of fascicles of smooth muscle cells with cigar-shaped nuclei. Additionally, the tumor cells in our case ultrastructurally had many dendritic processes, which are not seen in smooth muscle cells.

Based on the findings described above, the most appropriate term for diagnosis of the present case was considered to be a malignant myopericytoma. To our knowledge, this report describes the first case of malignant myopericytoma in laboratory rodents. In rats, most perivascular tumors have been diagnosed as hemangiopericytomas. Further investigations of hemangiopericytoma-like perivascular tumors are required to establish the term myopericytoma in laboratory animals.