Malignant mesothelioma is caused in more than 80 % by chronic exposure to asbestos fibers. There is a dose-exposure time relationship: a high dose will result in a mesothelioma within 3–5 years, whereas a medium to medium-low dose will cause mesothelioma only if there is a long duration of exposure [1]. The latency period in these cases can be up to 35 years after cessation of exposure. Different types of asbestos fibers also show differences in the required accumulative dose until mesothelioma development starts [2–6].

There are rare familial cases, for which recently BAP1 gene alteration has been found as a possible cause [7, 8].

Other possible causes are radiation and other inhaled carcinogenic fibers including manmade fibers, SV40 viral antigen, and erionite [6, 9–14]. The contribution of SV40 is controversially discussed, as there are arguments for and against its role in inducing mesothelioma [15–21]. From these investigations and discussions, it is likely that SV40 plays a role as a cocarcinogen, enhancing the development of asbestos fiber-induced mesothelioma [10, 11, 22]. For a more detailed discussion, the reader is referred to the article by Roggli et al. summarizing the current understanding on the epidemiology of mesothelioma and asbestosis [23].

Erionite has gained much interest in the past, because in some villages in Anatolia, Turkey, mesothelioma occurred with high frequency. Even immigrants from these villages presented with mesothelioma after exposure cessation. In the meanwhile, it has been shown that erionite can induce mesothelioma with high penetrance; however, also some genetic abnormalities might be involved too [9, 24–31].

Recurrent pleural effusion is the most common clinical sign of mesothelioma. As a rule recurrence of the effusion within 3 months should raise the suspicion of mesothelioma. Resulting from effusion atelectasis of the lung will cause hypoxia. On X-ray and CT scan, effusion and thickening of the pleura or pleural nodules are the prominent findings. On CT invasion of the chest wall can be seen in advanced disease.

Mesothelioma usually starts from the parietal pleura forming multiple firm white nodules, which later on form white dense plaques. In this stage, mesothelioma might be indistinguishable from pleural metastasis. In late stages, the visceral pleura is involved too and the whole lung can be encased by the tumor (Fig. 21.5). Mesothelioma tends to primarily spread along the lobar fissures and interlobular septa, but invades the lung in late stages. This is usually the gross picture seen at autopsy or sometimes on resection specimen (pleuropneumonectomy). There are few cases of lung carcinomas, which might spread to the pleura in a similar way causing the impression of pseudomesotheliomatous metastasis.

Malignant mesothelioma comes in three forms: epithelioid, mixed biphasic, or sarcomatoid mesothelioma. These varieties also carry different prognoses; sarcomatoid variant has the worst prognosis of an average of 6–8 months survival after diagnosis. Epithelioid mesotheliomas can mimic almost every other tumor: most often epithelioid mesotheliomas will present as tubulo-papillary tumors with abundant acid mucopolysaccharides, positively stained by PAS and Alcian blue. They sometimes form slit-like spaces resembling normal mesothelial structures. On the surface, well-differentiated mesothelioma will form papillae, garland-like rows of cells connecting papillae (Fig. 21.6). By electron microscopy but also in cytological specimen, long microvilli can be seen (Figs. 21.7, 21.8, and 21.9). However, these are not the mesotheliomas, which cause diagnostic problems. It is the variation of structures and patterns, which mesothelioma can form. Epithelioid mesotheliomas can be papillary, acinar, solid, deciduoid, and clear cell. They can mimic pulmonary adenocarcinomas, metastasis from ovarian cystadenocarcinomas, breast carcinomas, and ductal as well as lobular, renal clear cell carcinomas, just to name a few. The sarcomatoid mesotheliomas most often present as a spindle cell proliferation; can be embedded in an inflammatory background (inflammatory sarcomatoid mesothelioma), almost obscuring the tumor cells; and can be pleomorphic like true sarcomas (pleomorphic sarcomatoid mesothelioma). There are even rare forms of mesotheliomas, where the sarcomatoid component simulates an osteosarcoma with osteoid formation. A hard to diagnose sarcomatoid mesothelioma variant is desmoplastic mesothelioma, because tumor cells are scarce and hidden within a fibrous and hyalinized matrix (Fig. 21.10). In mixed mesotheliomas, a combination of epithelioid and sarcomatoid components is present.

Due to the wide variation of patterns, the diagnosis of malignant mesothelioma need to be confirmed by immunohistochemical stains. Many different schemas are currently used; however, a few antibodies have proven to be most reliable: a positive reaction for calretinin, podoplanin, cytokeratin 5/6, vimentin, and thrombomodulin and a negative reaction for monoclonal CEA. Claudin4 and BerEP4 are useful to confirm the diagnosis of malignant epithelioid mesothelioma (Fig. 21.13). As a recommendation, one should use those sets of antibodies, which reliably and reproducibly work in the laboratory. Do not switch to new published antibodies unless these have been validated in several studies against several of the mimickers of mesothelioma.

By cytology the diagnosis is possible, but should not be made straight forward: if there is recurrent effusion and atypical cells are found, I would make a tentative diagnosis of malignant mesothelioma. Immunocytochemistry can be done, preferentially using cell block techniques (Table 21.1).

Besides the commonly used markers for the differentiation of lung and breast adenocarcinomas versus mesotheliomas, the antibodies’ panel might be adapted according to the question addressed. A comprehensive review on marker expression can be found in the articles by N. Ordonez [34–39].

In sarcomatoid mesothelioma, the use of immunohistochemistry is less helpful, as there are no specific markers for mesothelioma. Most sarcomatoid mesotheliomas will express pan-cytokeratin, which is the most useful marker. In cases of coexpression of pan-cytokeratin and vimentin, this constellation is quite specific for malignant sarcomatoid mesothelioma. Calretinin is often negative, thrombomodulin and WT1 are positive in some sarcomatoid mesotheliomas, whereas podoplanin is most often negative. The negative markers as they are focused on epithelial differentiation are useless.

Due to the long latency period of developing mesotheliomas, there are well-documented precursor lesions. Hyperplasia of the mesothelium is the earliest event to be seen on biopsies; however, this is unspecific. Hyperplasia does occur in regeneration after infectious pleuritis. There have been some attempts to better separate regenerative hyperplasia from hyperplasia in developing mesothelioma: in the study by Attanoos et al., desmin and EMA were reported to be most useful in distinguishing benign from malignant mesothelial proliferations, where desmin was preferentially expressed in reactive mesothelium and EMA in neoplastic mesothelium. Immunohistochemical detection of mutated p53 was less useful, but may be used as a second-line marker of neoplastic mesothelium [40]. In the study by Zimmerman, the use of a chromosome 17α probes turned out as another marker for mesothelioma, showing an average CI of 2.23 with less than 44 % diploid nuclei and 50 % of specimens exhibited bizarre signals [41]. In my own experience as well as in two studies, CD146 turned out to be a reliable marker differentiating reactive from malignant mesothelial proliferations [42, 43]. Mesothelioma in situ was used by David Henderson, but usually in the setting adjacent to an already invasive mesothelioma. In situ mesothelioma is defined as an atypical mesothelial proliferation at the pleural surface without invasion into deeper layers (Fig. 21.14). These mesotheliomas in situ may have abundant myxoid stroma and protrude from the surface as small nodular lesions. It is still debated if one should use this term in cases where invasion is absent. Many pathologists use instead the term atypical mesothelial proliferation and ask for additional biopsies. In any case, either mesothelioma in situ or atypical mesothelial proliferation should not cause surgical intervention.

From several studies, it is well established that asbestos fibers first stick to the small bronchi and bronchioles causing chronic inflammation. Fibers transverse the bronchial walls and move through the lung by the respiration movement and finally reach the pleura, where these fibers are deposited, again causing chronic inflammation and damage to the mesothelium but also to the mesenchymal cells of the deeper pleural layer. By recurrent inflammation, aberrations of the DNA occur. Despite this knowledge, the molecular biology of mesothelioma is still not understood. There are many studies focusing on single genes, chromosomal aberrations, mechanisms by miRNA, and also posttranslational protein modifications. Unfortunately none of these studies could be repeated and verified by other studies, leaving many different gene modifications, but no common genetic alterations for a majority of mesothelioma cases.

Several studies have focused on chromosomal abnormalities in mesotheliomas. Loss of copies of chromosomes 8, 16, 20, and 18 were found. Translocations were seen in chromosomes 5, 10, and 13 [44]. In another study, partial or total losses of chromosomes 22, 1, 3, 9, and 14 as well as gains of 7 and 11 were reported. Translocations and deletions involving a breakpoint at 1p11-p22 were most common structural aberrations. Copy numbers of chromosome 7p were inversely correlated with survival [45]. Losses of chromosomal regions in 1p, 8p, 14q, and 22q and gains of 5p, 6p, 8q, 15q, 17q, and 20 were also reported in the study by Kivipensas [46].

Subsequent studies focused on aberrations in losses of the short arm of chromosome 9p21-pter. Other losses were frequently detected in the long arms of chromosomes 4q31.1-qter, 6q22-q24, 13, 14q24-qter, and 22q13. A gain was found in the long arm of chromosome 1cen-qter [47]. In the study by Taguchi, chromosomal losses were found in 1p and 9p. The shortest regions of overlap of these losses were at 1p21-p22 and 9p21-p22, respectively. Other common abnormalities included losses of 3p21 and 6q15-q21 and numerical losses of chromosomes 14, 16, 18, and 22 [48]. Losses in 9p21-p22 were more precisely characterized by the markers D9S171 and IFNA/IFNW [49]. It finally turned out that the most frequent lost region was 9p21.3, the locus of CDKN2A and CDKN2B. Other recurrent minimal regions of losses were 1p31.1-p13.2, 3p22.1-p14.2, 6q22.1, 9p21.3, 13cen-q14.12, 14q22.1-qter, and 22qcen-q12.3. Previously unreported gains included 9p13.3, 7p22.3-p22.2, 12q13.3, and 17q21.32-qter [50]. Losses of the p16 protein are a common occurrence in mesotheliomas (Fig. 21.15). Further deletions of 9p21-p22 outside the p16 locus may reflect the involvement of other putative tumor suppressor genes that could also contribute to the pathogenesis [51, 52]. Codeletion of p15 and p16 was found in many mesotheliomas, including all cases with spindle cell components [53].

Allelic loss of 3p21 is another frequent finding in malignant mesothelioma. And one or more putative tumor suppressor genes might contribute to the pathogenesis [54]. Furthermore three independent regions on 6q14, 6q22, and 6q24 are commonly deleted, most likely containing tumor suppressor genes [55]. Underrepresented segments were seen at 22q and 15q1.1-21. The most commonly overrepresented segment was on 5p. A new recurrent site of chromosomal loss within 15q was reported by Balsara et al. The minimal region of chromosomal loss was identified at 15q11.1-15 [56].

In this context of these differences in the reported genetic abnormalities, it does not make sense to focus on single markers but instead focus on pathways activated in mesotheliomas. One of the most common activated pathways is Wnt signaling through Dvl overexpression and downstream signaling through β-catenin [57].

Through the mechanism by asbestos fibers, inflammation plays a major role in the pathogenesis. Yang et al. presented important data that TNFα signaling through NFkB increases the percent of mesothelial cells that survive asbestos fiber exposure, thus increasing the pool of mesothelial cells susceptible to malignant transformation [58]. Moreover, loss of Faf1 and downregulation of Faf1 protein showed that Faf1 regulates TNFα-mediated NFkB signaling in mesothelioma [59]. The importance of asbestos-induced inflammation in the initiation and growth of mesothelioma has been highlighted by the identification of HMGB1 and Nalp3 inflammasomes as key initiators. Asbestos fibers induce necrosis and cause release of HMGB1, which in turn activate Nalp3 inflammasome – a process that is enhanced by asbestos-induced production of reactive oxygen species. HMGB1 and Nalp3 lead to interleukin 1β and TNFα secretion and NFkB activity, thereby promoting cell survival and tumor growth [31].