The pleura can be seeded by a wide spectrum of primary and metastatic tumoral conditions. Primary neoplasms affecting the pleura include epithelial, mesenchymal, and lymphoid neoplasms; of these categories, epithelial malignancies are the most common. The focus of this chapter is on primary lesions of the pleura.
Primary epithelial or epithelioid tumors of the pleura comprise various clinical entities, of which mesotheliomas are the most common, although a number of other conditions should be considered in the assessment of pleural biopsy material or resected specimens. These tumors, which may range in degree of malignant potential from benign to low-grade to high-grade malignant neoplasms, include the following:
The most common primary malignant tumor of the pleura is malignant mesothelioma, a condition that nevertheless can be difficult to diagnose. Owing to the legal implications of asbestos exposure and its link with pathogenesis of these lesions and because of their variable histopathologic appearance, they have been the subject of extensive studies. Malignant mesotheliomas are relatively unusual tumors, and the annual incidence in the United States has been estimated to be approximately 3 to 7 cases per 1 million persons, but this rate may be rising. Although mesotheliomas have been associated with exposure to asbestos fibers, approximately 50% of persons affected by mesotheliomas do not report asbestos exposure, indicating that the etiology may be multifactorial.
A careful analysis of radiologic findings along with histopathologic evaluation of appropriate material should provide clinicians and pathologists with the necessary information to make a specific diagnosis. In many instances the clinical and radiologic aspects are clear-cut but the available material for histopathologic examination is not adequate. In such circumstances, any attempt to make a definitive diagnosis should be deferred until additional material is obtained if clinically indicated. Surgical treatment for mesothelioma can be extreme; thus, an unequivocal diagnosis is imperative. Furthermore, other pleural conditions of an inflammatory nature may mimic malignant mesothelioma. Therefore, the clinical and radiologic information should be used not to make a diagnosis per se but rather to guide decisions about a diagnostic approach using immunohistochemical or electron microscopic techniques. Ultimately, the diagnosis of mesothelioma is a histopathologic one.
Wagner may have been the first to describe this tumor in the pleura, in 1870 ; in subsequent years, however, great controversy emerged regarding the characterization of tumors with diffuse pleural involvement. Many of these tumors were assigned different designations, including endothelial carcinoma, sarcoma, lymphangitis proliferans, sarcocarcinoma, and endothelioma. It was not until 1920 that Dubray and Rosson proposed the designation mesothelioma , a term in common use today. Although early reports had questioned the existence of primary pleural tumors, in time, well-documented cases were reported. In 1931, Klemperer and Rabin classified pleural tumors by their macroscopic appearance into localized and diffuse conditions. This distinction gave rise to the current terms diffuse pleural mesothelioma and solitary fibrous tumor of the pleura.
In 1960, Wagner and associates described 33 patients with mesothelioma and suggested the association with asbestos fibers. According to these investigators, all of the patients except one had a history of probable exposure to asbestos. Hirsch and colleagues described 28 cases, in which asbestos exposure was established in 17 cases. Some investigators presented larger series of cases with more emphasis on the association of mesothelioma and asbestos exposure, whereas others explored the histopathologic variability of mesotheliomas. In the past, the diagnosis was established using conventional histologic examination and histochemical stains, such as periodic acid–Schiff reagent (PAS) with and without diastase and mucicarmine, but today the emphasis has shifted to newer modalities such as electron microscopic and immunohistochemical techniques. Nevertheless, the numerous clinicopathologic correlations have contributed greatly to the current understanding of mesotheliomas.
Clinical and radiologic findings are of great importance in establishing the diagnosis. Thus, every effort should be made to correlate histologic features with clinical and radiological information. In general, mesotheliomas are more common in adults older than 50 years of age, but these tumors also may occur in children. History of long-standing exposure to asbestos, whether confirmed or not, should prompt a careful analysis of the biopsy tissue. Mesotheliomas can occur without a history of asbestos exposure, as evidenced by cases described in children and housewives. Other possible etiopathologic factors in the development of mesothelioma include radiation exposure, chronic inflammation, viral infections, and diethylstilbestrol.
If mesothelioma is suspected, a thorough search for evidence of the following diagnostic criteria is warranted:
Diffuse involvement of the pleura
Intraparenchymal tumor nodules or masses (peripheral)
Diffuse thickening of the pleura
Encasement of the lung
Unilateral or bilateral pleural involvement
Pleura-based tumor mass
Patients with mesothelioma may present with nonspecific signs and symptoms such as chest pain, dyspnea, cough, weight loss, and pleural effusions.
Mesotheliomas are tumors with a characteristic gross appearance, rarely posing a diagnostic problem on gross examination. The tumor will display diffuse pleural involvement with thickening of the pleural lining encasing the entire lung ( Fig. 13-1 ). In some cases, tumor growth follows the intrapulmonary septum, and in rare instances the tumor may involve the lung parenchyma, forming small nodules on the surface. The presence of a well-defined tumor mass in the periphery of the lung, even if diffuse pleural involvement also is evident, should alert the pathologist to the possibility of an adenocarcinoma with diffuse involvement of the pleura.
Mesotheliomas may show a variety of histopathologic growth patterns, but the three most common are epithelioid, sarcomatoid, and biphasic (a combination of the epithelioid and sarcomatoid types).
Epithelioid mesothelioma probably is the most common of the three variants, accounting for approximately 70% of all mesotheliomas. Several distinct histopathologic growth patterns of epithelioid malignant mesothelioma have been described, and occasionally, distinguishing among them may pose a diagnostic challenge. These subvariants include the following:
Tubulopapillary. This is the most common growth pattern in epithelioid mesotheliomas. The tumor may show the characteristic papillary growth pattern consisting of medium-sized round to oval cells, with moderate amounts of eosinophilic cytoplasm, round nuclei, and conspicuous nucleoli. In other areas, this cellular proliferation may show elongated tubular structures that appear to anastomose with one another. The tumor is fairly uniform in appearance, with very mild nuclear atypia and minimal mitotic activity. Areas of necrosis and hemorrhage are not commonly seen ( Figs. 13-2 to 13-13 ).
Clear cell. This growth pattern is characterized by a cellular proliferation composed of medium-sized, round to oval cells with round nuclei and conspicuous nucleoli, and clear cytoplasm. A diffuse cellular proliferation dissecting fibroconnective tissue is visible. Mitotic activity, although present, is not prominent, and focal areas of necrosis may be seen. This growth pattern mimics clear cell carcinoma of renal origin ( Figs. 13-14 to 13-17 ).
Glandular. This growth pattern is characterized by the presence of well-formed glands similar to those seen in adenocarcinoma. The glandular proliferation appears to dissect fibroconnective tissue, and in some cases, a desmoplastic reaction with inflammatory infiltrate may be present. This growth pattern closely resembles that in adenocarcinoma ( Figs. 13-18 and 13-19 )
Myxoid/mucoid. This growth pattern consists of a neoplastic cellular proliferation embedded in a myxoid or mucoid matrix, which may show glandular differentiation or the conventional tubulopapillary growth pattern. In some cases, the tumor may display abundant mucoid matrix; this appearance may be confused with that of a mucinous adenocarcinoma. With this pattern, mucicarmine staining does not show intracellular mucin production ( Figs. 13-20 to 13-22 ).
Adenomatoid. This growth pattern closely resembles that of the conventional adenomatoid tumor and is characterized by cords of medium-sized cells with clear cytoplasm and displacement of the nuclei toward the periphery of the cells. Nuclear atypia, mitotic activity, necrosis, and hemorrhage are not common ( Figs. 13-23 to 13-25 ).
Deciduoid. This growth pattern is characterized by a cellular proliferation composed of medium-sized cells with eosinophilic cytoplasm, displaying a “deciduoid” appearance similar to that seen in endometrial cyclic changes ( Figs. 13-26 and 13-27 ).
Cartilaginous and osseous metaplasia. This unusual variant of mesothelioma may pose a diagnostic challenge. The tumor shows areas of formation of “osteoid” or immature cartilage that may be confused with a primary orthopedic tumor. Along with the osseous or cartilaginous changes, a cellular proliferation composed of medium-sized cells with round to oval nuclei and conspicuous nucleoli is characteristic. In some areas the cellular proliferation may exhibit spindle cell features or a mixture of spindle and epithelioid cells ( Figs. 13-28 and 13-29 ).
Among these histopathologic growth patterns, the easiest to recognize, tubulopapillary, is the most common. Nevertheless, at least a theoretical familiarity with the other growth patterns is essential to permit a proper differential diagnosis. Regardless of the histopathologic growth pattern, whether the tumor shows clear cell change, myxoid areas, glandular differentiation, or an adenomatoid pattern, an important radiologic feature that cannot be overlooked is diffuse involvement of the pleura by tumor. This finding should always prompt consideration of the possibility of mesothelioma. In the great majority of cases, only a small biopsy specimen is available for diagnostic analysis.
Recently, Arrossi and coworkers evaluated 56 cases of extrapleural pneumonectomy to correlate the original subtype of mesothelioma in the biopsy specimen with that determined using the resected specimen. These workers found that in many cases, mesotheliomas must be reclassified after more complete sectioning is accomplished.
Before the advent of immunohistochemistry, histochemical studies played an important role in the diagnosis of mesothelioma. Currently, they can offer an easy solution in more routine cases. PAS, with and without diastase digestion and mucicarmine, and hyaluronic acid, with and without diastase digestion, have been used in the past with some success. Although both histochemical techniques are very useful, only one is necessary to provide sufficient data for evaluating a particular lesion. As noted, presence of intracellular mucin is strongly indicative of adenocarcinoma, but this finding also has been reported in some mesotheliomas (in up to 5% of the cases). Some mesotheliomas will exhibit abundant extracellular but not intracellular mucin. Thus, in current practice, histochemical techniques often are bypassed in favor of immunohistochemical studies.
A great deal of information regarding immunohistochemical studies in the evaluation of mesotheliomas is available. Recent attempts to define the application and limitation of these studies have been reviewed to discover important practical information in the evaluation of mesotheliomas. Numerous studies attempting to positively identify mesotheliomas have been published, some of which have provided useful information, whereas others have merely attempted to identify adenocarcinoma in order to rule out mesothelioma. Thus, the diagnosis of mesothelioma has in the past been considered one of exclusion. Although many different kinds of immunohistochemical markers are available that may help in the diagnosis of mesothelioma, only a few are used in practice ( Table 13-1 ). These can be classified as positive or negative markers for mesothelioma. The positive markers for mesothelioma include keratin 5/6, calretinin, the Wilms tumor susceptibility gene product (WT-1), HBME-1, thrombomodulin, and mesothelin.
Keratin 5/6 labels epithelioid mesotheliomas in approximately 90% of cases ( Fig. 13-30 ). It is considered a valuable marker for mesothelioma ; however, staining for this antibody also may be positive in carcinomas of extrathoracic origin and in squamous cell carcinoma of the lung. Therefore, this antibody’s value depends largely on the context in which it is used. Calretinin is part of a large family of cytoplasmic calcium-binding proteins and labels approximately 85% of epithelioid mesotheliomas ( Fig. 13-31 ). Of the three antibodies in this family, only calretinin labels mesothelioma and non-neoplastic mesothelium. WT-1 is the product of the Wilms tumor gene, a tumor suppressor gene located at 11p13 in mesangial cells of the kidney. WT-1 shows a strong positive reaction in mesotheliomas; however, it also may react positively in tumor cells of other neoplasms, including ovarian and peritoneal serous carcinomas, malignant melanoma, and renal cell carcinoma. HBME-1 was generated from a human cell line derived from a patient with malignant mesothelioma. This antibody decorates the membrane of mesothelial cells, as opposed to cytoplasmic staining in adenocarcinomas. HBME-1 may not be a highly reliable marker of mesothelioma, however, because a notable percentage of adenocarcinomas and serous tumors of the ovary also may show positive staining. Thrombomodulin (CD141) is a glycoprotein expressed in endothelial cells and in a variety of other cell types, including mesothelial cells. Several studies have been presented in the literature with claims of 60% to 100% staining in malignant mesotheliomas. This marker also may give positive staining in approximately 75% of adenocarcinomas; therefore, its usefulness is limited. Mesothelin is a surface protein that is expressed in the membrane of neoplastic cells in mesotheliomas and in non-neoplastic mesothelial cells. However, mesothelin also may give positive staining in serous carcinomas of the ovary, pancreatic adenocarcinomas, cholangiocarcinoma, colonic adenocarcinoma, and pulmonary adenocarcinoma.
A plethora of markers have been used in the diagnosis of mesothelioma, but mainly to rule it out. When the differential diagnosis is between mesothelioma and adenocarcinoma, the most commonly used negative markers are carcinoembryonic antigen (CEA), MOC31, thyroid transcription factor-1 (TTF-1), Leu-M1 (CD15), and B72.3. Other markers that have been used include Ber-Ep4 and BG-8. The ideal is a positive marker that excludes the possibility of mesothelioma.
CEA is considered one of the most reliable markers for distinguishing adenocarcinoma from mesothelioma, because the vast majority of mesotheliomas demonstrate negative reactivity for this antibody. Some studies have suggested that the 5% positivity observed in some mesotheliomas may be due to the use of a heteroantiserum unabsorbed by CEA, which may label unrelated epitopes. In this setting, the use of monoclonal antibodies to specific CEA epitopes is more reliable. MOC31 has been reported in several studies as an important marker for distinguishing mesothelioma from adenocarcinoma, because it purportedly gives positive staining in adenocarcinoma cells. In some cases of mesothelioma, however, MOC31 may give focal and spotty positive staining. TTF-1 is expressed in normal lung and in thyroid epithelial cells. TTF-1 shows high specificity for lung adenocarcinoma, and so far, staining for TTF-1 has been reported to be negative in mesotheliomas. Thus, it is one of the most important markers to separate pulmonary adenocarcinoma from mesothelioma. Leu-M1 (CD15) has a high level of specificity for adenocarcinoma; however, some mesotheliomas, namely peritoneal mesotheliomas, also may demonstrate positive staining in tumor cells. B72.3 is a generic epithelial determinant (tumor-associated glycoprotein-72) that is a high-molecular-weight cell membrane glycoprotein. Although it is a good marker for adenocarcinoma, some mesotheliomas also may demonstrate focal positive staining.
As mentioned earlier, many more antibodies have been presented in the literature as very specific for the distinction between adenocarcinoma and mesothelioma; over time, however, these antibodies have proved unreliable. One such antibody, Ber-Ep4, originally was presented as specific for adenocarcinoma but also has been shown to react with mesotheliomas in more than 20% of the cases. BG-8 is another antibody that may react strongly in cases of adenocarcinoma; however, some mesotheliomas also may show positive staining.
Ultrastructural studies are very important in the diagnosis of mesothelioma; however, in many cases the utility of such studies is hampered by lack of material when it is needed the most. Often, sufficient material does not become available until a more extensive procedure has been performed, but in a majority of the cases, the initial biopsy specimen is the only material available for diagnosis. Electron microscopic features are helpful for evaluation of better-differentiated tumors; but when the tumor is poorly differentiated, the ultrastructural findings are rarely helpful. In most cases when immunohistochemical analysis has failed to provide a clear interpretation, findings on electron microscopy also will be questionable. Nevertheless, the latter study can be very helpful, and every effort should be made to obtain a sample for this examination. The finding of long, slender microvilli is a histopathologic hallmark of mesothelioma.
In the setting of an atypical epithelial cellular proliferation, the most important condition to rule out is either pulmonary adenocarcinoma extending into the pleura, metastatic epithelial tumor of other origin, or most important, mesothelial hyperplasia ( Table 13-2 ). If the cellular proliferation has been deemed to be malignant, then immunohistochemical studies, especially carcinomatous epitopes, will be the next step. A similar approach is appropriate with a metastatic epithelial neoplasm from another source to the pleura; however, the interpretation can be more difficult in cases of mesothelial hyperplasia. In this setting, no available immunohistochemical stain can separate a neoplastic cellular proliferation from a hyperplastic one. Thus, even though the necessary steps have been followed, it is imperative not only to correlate the histopathologic findings with the clinical and radiologic features but also to provide an accurate interpretation of the results of the immunohistochemical studies. Even electron microscopic studies would fail to separate such cellular proliferations. In essence, the diagnosis is a morphologic one that requires careful attention to specific histopathologic features, such as invasion into adipose tissue or skeletal muscle, that are associated with malignant mesothelioma (see Tables 13-1 and 13-2 ).
|Penetration into adipose tissue or muscle||+||0|
|Cellular proliferation in surface||Often 0||+|
|Granulation tissue||Often 0||+|
The sarcomatoid variant of mesothelioma is less common than the epithelial variant and probably accounts for less than 15% of mesotheliomas in its pure form. As its name implies, the characteristic tumor growth pattern is one of spindle cells with elongated nuclei and inconspicuous nucleoli, mimicking sarcoma of soft tissues. In a study of spindle cell tumors of the pleura, Carter and Otis proposed three types, ranging from low grade (possibly benign) to high grade: fibroma (keratin-negative tumor), sarcomatoid mesothelioma (keratin-positive tumor), and sarcoma, or malignant spindle cell tumor (keratin-negative). Because in some cases the histopathologic features may overlap, immunohistochemical studies play an important role in diagnosis.
Malignant spindle cell tumors (keratin-positive) of the pleura can be further subdivided into three distinct categories based on their growth pattern:
Spindle cell type (fibrosarcoma-like or malignant fibrous histiocytoma–like). The histopathologic diagnosis of either one of these variants is rather straightforward. In the fibrosarcoma-like pattern, the tumor is composed of a spindle cellular proliferation that may show the so-called herringbone pattern with interdissecting fascicles of spindle cells with indistinguishable cell membranes, moderate amounts of light eosinophilic cytoplasm, elongated nuclei, and inconspicuous nucleoli. Nuclear atypia is present, and mitotic activity is readily visible. In the malignant fibrous histiocytoma–like pattern, the tumor displays features of a high-grade sarcoma with a fascicular growth pattern characterized by the presence of spindle or oval cells, or both, with elongated or round nucleus and conspicuous nucleoli. In addition, the tumor also may exhibit multinucleated malignant giant cells intermixed with the spindle cell proliferation. Nuclear atypia is prominent, and mitotic activity is readily visible ( Figs. 13-32 to 13-40 ).
Desmoplastic mesothelioma. This variant is the one that poses a challenge in diagnosis, mainly when only a small biopsy specimen is available for interpretation. The initial description by Kannerstein and Churg in 1980 has been followed by a few more series. Cantin and associates reported 27 cases in which the clinical course was often rapid, and the mean survival for cases of pure sarcomatoid tumor was approximately 6.18 months. In their experience, desmoplastic mesothelioma also showed a greater tendency toward metastatic disease at 60%, compared with 40% of the nondesmoplastic variant. Mangano and coworkers reported a series of 31 cases in which the emphasis was on separating desmoplastic mesotheliomas from fibrous pleurisy. These workers noted the presence of p53 in these two conditions and concluded that reactivity for this marker can be positive in both, and that although p53 is more commonly seen in desmoplastic mesothelioma, the difference was not statistically significant. Histologically, these tumors may show extensive areas of collagenization with a very discrete spindle cell proliferation that may be missed in a cursory review of the histologic sections. The following are the most important histopathologic features that have been associated with the diagnosis of desmoplastic mesothelioma:
Invasion of chest wall or lung
Foci of bland necrosis
Frank sarcomatoid foci
These criteria apply mainly to resected specimens ( Figs. 13-41 to 13-44 ), pleural peeling, or a very generous pleural biopsy specimen. In a small sample, establishing this diagnosis may prove to be very difficult, if not impossible. Colby has warned about the care that must be exercised in making such a diagnosis and has stated that because no proven therapy for desmoplastic mesothelioma is recognized, underdiagnosis is preferable to overdiagnosis. This is especially true nowadays, because the current trend is to perform extrapleural pneumonectomies for the treatment of mesothelioma.
Lymphohistiocytoid. This type of mesothelioma is included in the subcategory of sarcomatoid mesotheliomas, although in some cases the epithelioid component may be formed by oval cells instead of by spindle cells. This subtype is unusual and is characterized by a prominent lymphoid component admixed with a cellular proliferation composed of epithelial cells with a “histiocytoid appearance” ( Figs. 13-45 to 13-47 ).
Histochemical studies, such as those using PAS with and without diastase, mucicarmine, or hyaluronic acid with and without diastase digestion, have no role in the diagnosis of these tumors.
In the setting of a spindle cell mesothelioma, the role of immunohistochemical studies is relatively limited whether the tumor is desmoplastic or not, because most of the antibodies used in diagnosis of conventional epithelioid mesotheliomas have no practical use in identification of sarcomatoid mesotheliomas. Broad-spectrum keratin is by far the most important ( Fig. 13-48 ). All other carcinomatous epitopes are known not to react with sarcomatoid tumors. The use of calretinin and keratin 5/6 is rather limited because positivity may vary, and negative results do not mean that the tumor in question is not a mesothelioma. Fibrous pleurisy cannot be distinguished from desmoplastic mesothelioma by means of immunohistochemical techniques, because both lesions may react with keratin antibodies. Use of immunohistochemical studies is relevant to rule out other spindle cell tumors of different lineage, including leiomyosarcomas, malignant fibrous histiocytoma, and other mesenchymal tumors.
When the neoplastic nature of the tumor is not in question, the most important distinction is that with another spindle cell neoplasm of mesenchymal origin. In this setting, immunohistochemical studies or electron microscopy will lead to the appropriate interpretation. In cases of sarcomatoid carcinoma involving the pleura in a diffuse manner, the radiographic finding of an intrapulmonary tumor mass will lead to the appropriate interpretation. This distinction may prove to be very difficult on histopathologic grounds, however. By far the most challenging entities to rule out in the differential diagnosis are fibrinous pleuritis and fibrous pleurisy of a reactive or inflammatory nature. With these processes, the diagnosis is based on morphologic grounds, because immunohistochemistry cannot solve the problem ( Table 13-3 ).
|Granulation tissue||Often 0||+|
|Inflammatory reaction||Often 0||+|
|Smooth muscle actin||+||+|