Archives of the AFIP -- RadioGraphics 1996; 16: 613-644

Because malignant mesothelioma may resemble adenocarcinoma or sarcoma, techniques other than hematoxylin-eosin (H-E) staining are needed to make the correct diagnosis. Current techniques include histochemical staining, immunohistochemical staining, and electron microscopy.

As noted previously, epithelioid malignant mesotheliomas may contain glycogen or hyaluronic acid. Histochemical staining with periodic acid-Schiff (PAS) stain will stain glycogen and mucicarmine; Alcian blue and colloidal iron will stain hyaluronic acid. Unfortunately, these agents also stain the mucin in adenocarcinoma. Predigestion of the tissue with diastase will minimize or eliminate the PAS staining of glycogen in a mesothelioma, leaving PAS-positive mucin unaltered in an adenocarcinoma. Similarly, hyaluronidase predigestion largely removes the hyaluronic acid of mesotheliomas but not the mucin of adenocarcinomas. However, the sensitivity of these techniques is relatively low: 73% for diastase and 41% for hyaluronidase (35,37).

Immunohistochemical techniques are usually used in conjunction with or instead of these histochemical stains. Both adenocarcinoma and mesothelioma can be stained with antibodies to cytokeratins and epithelial membrane antigen, although the staining patterns differ. No antibody is known to specifically identify mesotheliomas. Several antibodies that stain a higher percentage of adenocarcinomas than malignant mesotheliomas will also stain a small percentage of mesotheliomas. As a result, multiple antibodies are used in the evaluation; cytokeratins, carcinoembryonic antigen (38), Leu-M1 (39), B72.3 (40), and BER-EP4 (41) are commonly included. Staining for cytokeratins and more than one of the remaining antibodies in the list effectively excludes a diagnosis of malignant mesothelioma. Staining for cytokeratins and the absence of staining for any of the remaining antibodies strongly suggests mesothelioma. The absence of staining for cytokeratins in the appropriate histologic setting indicates a diagnosis of sarcoma. The search for a specific marker of malignant mesothelioma continues; recently, several antibodies have shown promise, including thrombomodulin (42), ME1 (43), and Wilms' Tumor 1 susceptibility gene (WT1) (44).

Ambiguous results from staining techniques necessitate ultrastructural analysis by means of electron microscopy. The presence of long, slender, curved, branching microvilli with a length-to-diameter ratio of 10-16:1 is regarded as a specific marker of malignant mesothelioma (Figure 9). The ultrastructural features identified will vary with the type of mesothelioma; for example, bushy microvilli are common in the epithelioid variant but are inconspicuous or absent in the sarcomatoid variant. The reverse distribution would be seen for actin fibers. Direct contact of microvilli with collagen fibers through the basement membrane defects is very suggestive of mesothelioma and is not seen in association with adenocarcinoma (45). Electron microscopy can also be used to confirm the absence of features of adenocarcinoma, such as intracellular mucin and membrane-bound secretory granules. Sarcomatoid and biphasic variants will show ultrastructural features transitional between epithelioid and mesenchymal cells that will permit differentiation between mesothelioma and sarcoma or reactive fibroblasts (45). However, in many cases, the diagnosis of mesothelioma can be made without the time and expense of electron microscopy (35).

Abstract

Introduction

Etiology and Epidemiology of MPM

Clinical Features

Microscopic Features

Characterization of MPM with Histochemistry, Immunochemistry, and Electron Microscopy

Gross Pathologic Characteristics

Radiographic Features

CT Features

MR Imaging Findings

Differential Diagnosis

Diagnostic Methods

Staging and Resectability

Treatment

Prognosis

Source Information

References