Document 6Bb5epQe7YqrLxo4rG5MwDd7E

670 S..P- Hammar et al. effusion restricted to the contralateral side is rare.902 In contrast, with patients with breast cancer and subdiaphragmatic neoplasms (for example, stomach or ovary), there is no such predilection for the ipsilateral side.902 It has been estimated that 50% of patients with dissemi nated breast cancer develop a pleural effusion during the course of their disease, on the same side as the original breast cancer in 60% of the patients, on the contralateral side in 25%, and bilaterally in about 15%.902 In general, the interval between the diagnosis of the primary breast cancer and the subsequent development of an associated pleural effusion is about 2 years, but it can be as long as 20 years or more.902 The size of the pleural effusion in metastatic malig nancy varies greatly. In about 75% of patients the effu sion is moderate to large, within the range of about 500 to 2000mL; in about 10% the effusions are massive (with complete opacification of the affected hemithorax); and in a further 10%, approximately, the effusions are small (less than 500mL).903 About 70% of patients with a massive effusion have an underlying cancer as the basis for the effusion.903 Matthay et al.902 referred to one series of 46 patients with massive pleural effusions from all causes: 31 (67%) had malignant pleural effusions, 27 as a consequence of metastatic carcinoma and one patient had a MM. From an analysis of 500 documented cases of pleural effusion as a consequence of metastatic malignancy, Matthay et al.902 found that the diagnostic yield from cytologic examination of pleural effusion fluid was 66%, versus 46% from pleural biopsy. Matthay et al. com mented that pleural fluid cytologic examination is more sensitive for the diagnosis of metastatic cancer than pleural biopsy, and although cytology and biopsies are complementary to each other, pleural biopsy added little to cytologic examination. Matthay et al. commented further that the lower yield from pleural biopsy may represent operator technique or sampling error, the latter known to be a problem in that metastatic deposits can be widely scattered over the pleural membrane. They suggested that diagnostic yield can be increased by repeat cytology examinations and pleural biopsy. If a diagnosis is not obtained following repeat cytology exam ination and biopsy, thoracoscopy can be considered, and when multiple biopsies are taken at thoracoscopy, the diagnostic yield rises to about 80% to 97%.902 Vargas and Teixeira900 commented that pleural biopsies in cases of malignant pleural effusion establish the diagnosis in about 40% to 75%, but the combination of cytologic evaluation of the effusion fluid and a needle biopsy allows a diagnosis in about 80%. Medford and Maskell906 com mented that "blind" pleural biopsy increased the diagnos tic yield over cytologic examination of effusion fluid by only 7% to 27%, and that at least four samples from one site are required to optimize the diagnostic return. These authors also set forth their perception that "blind" pleural biopsy no longer has a role in the investigation of malig nant pleural disease and that it should be replaced by guided biopsies under imaging control, In general, pleural metastatic deposits are a marker of advanced disease,906 and survival of patients with pleural deposits from cancer of the lung, stomach, or ovary is usually measured in only a few months following diagno sis of the malignant pleural effusion.902 Although it is emphasized that lung and breast cancer and malignant lymphomas account for about 75% of malignancy-associated pleural effusions, almost any cancer with the capacity for metastasis to the lungs in particular also has the capacity for metastasis or spread to the pleura. Such unusual metastases can range from renal cell carcinomas to ependymomas arising in the central nervous system, among many others. Pseudomesotheliomatous Tumors Affecting the Pleura Including Pseudomesotheliomatous Adenocarcinoma of Lung By definition, pseudomesotheliomatous neoplasms affecting the pleura are characterized by diffuse infiltra tion of the pleura in a pattern essentially identical to, and indistinguishable from, pleural MM on gross examination or on radiologic studies, including CT scans.907 In this regard, the neoplasm characteristically takes the form of multiple nodules, plaques, or a confluent rind of tumor, with an associated pleural effusion in many instances and with frequent obliteration of the pleural cavity in the later stages of the disease, sometimes with invasion into the chest wall, diaphragm, and pericardium, as seen at autopsy. Most pseudomesotheliomatous neoplasms affecting the pleura are thought to originate from the lung,908-923 but pseudomesotheliomatous metastases from carcino mas arising in other sites are well recorded, including the kidney,91S,924_926 thyroid gland,497 larynx,927 stomach,918 and cutaneous malignant melanoma as well as various sarco mas, including malignant phyllodes tumor.928 In addition, with pseudomesotheliomatous carcinomas (PMCs) of the lung, adenocarcinoma is the most frequent histologic type, but other cell types can produce pseudo mesotheliomatous spread, including SCC, small cell car cinoma,773 large cell undifferentiated carcinoma, and carcinosarcoma.5j2 Pseudomesotheliomatous carcinomas of the lung were first described by Babolini and Blasi929 in 1956, to empha size that the symptoms in these patients were related predominantly to involvement of the pleura with recur rent exudative effusion, often accompanied by chest pain and dyspnea. Of five cases reported by Babolini and Blasi, two appear to have represented small cell carci noma and the other three were adenocarcinomas. About 43. Neoplasms of the Pleura 671 Figure 43.145. Pseudomesotheliomatous adenocarcinoma. Pleural biopsy from a 77-year-old man with a right pleural effu sion. At thoracoscopy, the appearances were considered sugges tive of a malignant mesothelioma. The neoplastic acini are embedded in a prominent fibrous stroma ("tubulo-desmoplastic adenocarcinoma"). 20 years later, Harwood et al.908 reported six cases of primary lung cancer with mimicry of mesothelioma in terms of the distribution of the carcinoma within the pleura, and they introduced the term pseudomesothelio matous carcinoma. In two of their six cases there were small intraparenchymal nodules in the underlying lung parenchyma and all tumors were adenocarcinomas, with bronchioloalveolar features in five. The patients were all men of ages 50 to 76 years, and they had symptoms of dyspnea on exertion, chest pain, and weight loss. Koss et al916 also reported an underlying adenocarcinoma in the lung in seven out of 14 autopsy cases. Nonetheless, in some instances, pleural pseudomesotheliomatous adeno carcinomas show no evidence of an underlying intrapa renchymal tumor, probably explicable by overgrowth of a small peripheral primary lung cancer by the predomi nant pleural extension. In their review, Koss et al.916 reviewed 15 previously published pseudomesotheliomatous adenocarcinomas of lung and added a further 15 examples from the files of the Armed Forces Institute of Pathology (AFIP) in Wash ington. Ninety percent of the patients were men with a median age of 61 years, and 17% had possible to definite occupational exposure to asbestos; one patient had proven asbestosis. The prognosis for pseudomesothelio matous adenocarcinoma was similar to that of mesothe lioma: the mean survival time in this series916 was 4.7 months and the longest survival was 25 months. Although PMCs are defined entirely by the gross ana tomic distribution of the neoplasm (or on radiologic examination as a surrogate for gross examination), the acinar structures in pseudomesotheliomatous adenocar cinoma may or may not resemble an epithelial mesothe lioma; that is, these tumors may comprise simplified or isolated glands in a fibrotic stroma, with appearances characteristic of adenocarcinoma; however, in some instances they can show a complex branching and anas tomosing architecture producing a histologic resemblance to epithelial mesothelioma (Figs. 43.145 and 43.146). The acini, tubules, and nests of tumor cells in PMC are char acteristically surrounded by thickened and fibrotic stromal tissue (Fig. 43.145), heightening the resemblance to mesothelioma (an appearance that Hammar and Dodson907 have described as "tubulo-desmoplastic adenocarcinoma''). In the series reported by Koss et al.,916 the main feature used for the diagnosis of pseudomesotheliomatous ade nocarcinoma was the presence of PAS-diastase-positive mucin in gland lumina or as intracytoplasmic droplets (but "all of the AFIP surgical specimens ... were selected on the basis of mucin-positivity within tumor ceils"). The distinction between mesothelioma and pseudomesothe liomatous adenocarcinoma is usually straightforward on immunohistochemical staining, and the distinction is facilitated by use of a panel of mesothelial cell markers and generic carcinoma-related antibodies (Figs. 43.147 and 43.148), together with immunostaining for TTF-1. A further issue that awaits clarification is whether a causal relationship between these tumors and asbestos exposure differs from other bronchopulmonary carcino mas. In our experience, a high proportion of pseudome sotheliomatous adenocarcinomas appear to have a background of occupational exposure to asbestos, but it is unclear whether this seemingly high proportion is Figure 43.146. Pseudomesotheliomatous adenocarcinoma of pleura. This carcinoma is more cellular than the tumor illus trated in Figure 43.145, with a paucity of stromal tissue. The nuclei of the neoplastic cells are nonhyperchromatic and they show only moderate cytologic atypia. The appearances are similar to those seen in some epithelial mesotheliomas. 672 S.P. Hammar et al. in a pseudomesotheliomatous fashion in one patient. No site of origin within the lung could be identified for either tumor. These authors suggested that the following find ings in such tumors militate against a diagnosis of meso thelioma: (1) neutral mucin production; (2) expression of CEA; (3) squamous differentiation, although squamous differentiation can occur rarely in MMs of epithelial type; or (4) evidence of neuroendocrine differentiation. Serosal-Surface Serous Papillary Tumors Figure 43.147. Pseudomesotbeliomatous adenocarcinoma. (Same case as in Fig. 43.145.) The tumor shows positive staining for carcinoembryonic antigen (CEA). explicable by (1) patterns of referral of cases for which mesothelioma is the differential diagnosis, or (2) whether the clinical and radiologic mimicry of mesotheliomas by these tumors stimulates a more detailed history concern ing asbestos exposure than would be the case for conven tional lung cancers (see also Chapter 27). Spindle Cell Carcinoma and Carcinosarcoma of Lung Although spindle cell (sarcomatoid) carcinomas of lung usually form localized intraparenchymal mass lesions, they can. invade the pleura, with the potential for histo logic mimicry of biphasic mesothelioma. In this regard, Mayall and Gibbs532 reported two carcinosarcomas that presented as pleural tumors, with encasement of the lung Because serous papillary adenocarcinomas arise predom inantly from the ovaries or the peritoneal mesothelium itself, mimicry ofpleural mesothelioma is exceptional, but it can constitute a significant diagnostic problem, espe cially because a high proportion of serous papillary car cinomas show no evidence of CEA expression on immunohistochemistry.U9'9j0""9jJ Even so, three patients with an underlying serous papillary adenocarcinoma of the peritoneum encountered by the authors119 (Figs. 43.149 and 43.150) presented with unilateral pleural effu sion, apparently related to spread from the underlying peritoneal tumor (in at least one of these cases, the primary peritoneal lesion was demonstrable only on CT imaging). The diagnosis in most instances can be made on detailed immunohistochemical studies, for example, including labeling with antibodies such as Ber-EP4 (Fig. 43.150), B72.3,ll9'Mf''932'934 and BG8. In two cases in our files the diagnosis was established primarily by electron microscopy, which demonstrated short blunt microvilli with an antennular glycocalyx characteristic of carcinoma in one case, and by the presence of elongated branched microvilli in another case, where the microvilli lacked the sinuous and serpentine architecture characteristic of mesothelial microvilli.119 The resemblance of such serous Figure 43.148. Pseudomesotheliomatous adenocarcinoma. Positive linear labeling of the neoplastic cells with Ber-EP4. Figure 43.149. Pleural metastasis of a serous papillary adeno carcinoma of the peritoneum (cytology cell, block section). Linear membrane-related staining for epithelial membrane antigen (EMA), essentially indistinguishable from labeling that characterizes epithelial mesotheliomas.