Malignant pleural mesothelioma is highly associated with asbestos exposure. A study examining populations derived from Los Angeles, New York State, and Veterans Administration Hospitals nationwide estimated that the attributable risk for exposure to asbestos among men with pleural mesothelioma was 88 % [15]. The risk of pleural mesothelioma following exposure to asbestos is dose dependent, as clearly documented in the Wittenoom cohort of crocidolite miners and millers from Western Australia [16]. Risk increases in a linear or supralinear fashion over time, even after exposure cessation [17]. The clinical presentation of mesothelioma lags behind the time of first exposure by about 30 years (latency period) . Thus, the increasing incidence of pleural mesothelioma in the 1970s reflected a surge in asbestos usage in war-related manufacturing during World War II, and the peak incidence of pleural mesothelioma in the USA in the early 1990s reflects the peak use of amphiboles in US manufacturing practice in the 1960s [18]. Worldwide, men are three- to fourfold more likely to die of malignant pleural mesothelioma than women [6]. This observation reflects higher likelihood of more intense asbestos exposure among men than women in post-World War II industrial practice. Studies of women engaged in the manufacture of crocidolite-containing gas masks in England during World War II showed that they too were at significantly increased risk of death from lung and pleural tumors as compared to other malignancies [19].

Environmental exposure to asbestos poses an increased, albeit generally lower, risk as compared to occupational exposure, with family members of asbestos workers and those who live in proximity to asbestos industries showing an increased risk for pleural mesothelioma as compared to other geographic cohorts [20]. Remarkably, high frequencies of pleural mesothelioma have been documented in rural populations where asbestos is present in surface soil and residents have long-standing environmental exposure to the fibers. One cohort study of the rural Dayao community in southwestern China, where crocidolite is prevalent in the soil, estimated an annual mesothelioma mortality rate of up to 365 per million, with mesothelioma accounting for 22 % of cancer deaths [21]. Studies from southeastern Turkey have documented a two- to fivefold increased incidence of malignant pleural mesothelioma among inhabitants of villages where soil-containing tremolite and chrysotile have been used for whitewashing and other household purposes, as compared to villages where asbestos-containing minerals have not been detected [22, 23]. Even in the absence of direct exposure to asbestos-containing soil related to farming or household practices, proximity to sources of naturally occurring asbestos , such as serpentinite and other ultramafic rocks in California, is associated with an increased risk of malignant mesothelioma [24].

As of the mid-2000s, survival in US populations for all comers with mesothelioma was 40.9 % at 1 year, 12.2 % at 3 years, and 3.9 % at 5 years [13]. More recent population data from European cohorts have described similar survival outcomes, with adverse prognostic features including older age, male sex, and sarcomatoid histology [25, 26]. Current use of multimodality therapy, including surgery, radiation , intracavitary chemotherapy, and systemic chemotherapy, has led to some improvement in survival, but these approaches are rarely curative and are controversial [27]. Retrospective analysis of population-based data derived from the SEER dataset of patients diagnosed with malignant mesothelioma demonstrated no significant improvement in overall outcomes over the past four decades, with a median survival of 7.2 months for patients diagnosed in the 1970s versus 7.1 months in the 2000s [13] (Fig. 2.3). There has been a statistically significant improvement in survival among patients with distant disease (5.5 months in 1970s versus 7.0 months in 2000s, p = 0.001); however, this improvement is marginal in clinical terms [28]. In patients undergoing chemoradiotherapy and/or surgery, features associated with adverse survival include male gender, sarcomatoid histology, and advanced pathologic stage according to the American Joint Committee on Cancer staging system [29, 30].

A substantial body of literature is dedicated to the physical and chemical features of asbestos fibers that contribute to carcinogenesis. The results of these studies have resulted in a widely accepted “fiber pathogenicity paradigm” encompassing the characteristics of a pathogenic fiber and the general process by which it leads to tumor formation. In general, a fiber must (a) be inhaled, (b) travel through the upper respiratory tract and be deposited in the lower respiratory tract, (c) persist within the body for a significant amount of time, (d) travel to the parietal pleura, and (e) possess pro-inflammatory and genotoxic physical and chemical properties. In vitro, in vivo, and human epidemiologic studies of asbestos and other fibers have suggested that fibers with high-aspect (length-to-diameter) ratios of > 3:1, small diameters (0.25–0.4 µM), and longer fibers with a minimum length of 5 µM (ideally, lengths of > 10 µM) produce the greatest pathogenic effects [31–36]. Fibers possessing these qualities are present in significant quantities in many widely used types of asbestos and asbestiform materials, and are thought to account for the majority of fiber-related malignant pleural mesotheliomas [35, 37]. While the physical properties of fibers are thought to mediate most of their carcinogenic effects, the chemical composition of a pathogenic fiber is also contributory inasmuch as it affects fiber structure, ability to generate oxidative damage, and biopersistence.

The principal types of commercial asbestos are chrysotile, amosite, and crocidolite, also known as white, brown, and blue asbestos, respectively, based on their physical coloration (Table 2.1). Amosite and crocidolite are amphiboles, according to their mineralogic properties. Industries in which workers are most likely to be exposed to one of these forms of asbestos include mines, textile industries, and manufacturing of cement, insulation, and brakes (Table 2.2). Of these three fiber types, crocidolite is associated with the highest risk of mesothelioma development; risk of death from mesothelioma following crocidolite exposures is up to one order of magnitude higher than following amosite exposure and two orders of magnitude higher than with chrysotile exposure [38].

Chrysotile is the most commonly employed asbestos fiber, historically accounting for ~ 95 % of total asbestos use [39]. The pathogenicity of chrysotile fibers in development of mesothelioma has been up for debate. Studies of lung tissue from Quebec miners and millers with mesothelioma have demonstrated that very-high-fiber loads of chrysotile can be oncogenic in the absence of significant concentrations of amosite or crocidolite. However, in many cases, chrysotile is accompanied by high levels of tremolite, particularly in mining and textile industries, thus the actual etiologic agent is unclear [40]. Indeed, chrysotile fibers, unlike the amphiboles, are not readily retained in the lung. Chrysotile is relatively fragile and fragments easily, permitting phagocytosis by pulmonary macrophages, and may actually dissolve in lung tissue due to leaching of magnesium out of the fiber. In contrast, amphiboles (which include tremolite) have a straight and broad structure and do not fragment readily, thus they are less susceptible to phagocytosis, and are chemically stable in a biologic environment [41].

Anthophyllite, actinolite, and tremolite are less commonly used in industrial practices in the USA, although these have been mined and used for commercial purposes in other countries and are known contaminants of other industrial minerals including talc and vermiculite. Studies in animal models have suggested that anthophyllite is carcinogenic and contributes to the development of malignant mesothelioma [42]; however, confirmed human cases of anthophyllite-attributable mesothelioma are very rare [43]. Tremolite is a contaminant of other mineral deposits, including chrysotile (see above) and vermiculite, which is used as a form of insulation and a gardening material. In addition to the epidemiologic evidence linking environmental exposure to tremolite in Turkish villages to malignant mesothelioma (see above), occupational exposure to tremolite is linked to development of disease as well. Cohort studies from Libby, Montana, the location of a large tremolite-contaminated vermiculite mine, have shown that miners, millers, and processors of vermiculite were significantly more likely to die of asbestos-related diseases, including mesothelioma, than the general population [44]. Actinolite is chemically similar to tremolite and may be found in combination with tremolite deposits but is less common [45].

Irrespective of these different chemical and biologic properties, all of these fibers are classified together for the purposes of defining workplace regulations under the Occupational Safety and Health Administration (OSHA) [46] and are regulated under the Environmental Protection Agency’s Clean Air Act [47]. The Toxic Substances Control Act banned manufacture and importation of asbestos-containing paper products and flooring felt, as well as any nonhistorical, “new uses” of asbestos . The Clean Air Act and Consumer Product Safety Act have banned the use of materials containing > 1 % asbestos that are sprayed on and asbestos-containing wall-patching compounds [48].

In the USA, asbestos-containing products persist in construction, clothing, and car manufacture and repair [47]. Chrysotile was used in automotive brakes until its use was banned by the Environmental Protection Agency (EPA) in the 1980s. Although OSHA cites an unspecified risk of mesothelioma among automotive mechanics, epidemiologic studies to date have failed to demonstrate an increased incidence among this group relative to background [49]. Similarly, chrysotile was ubiquitous in industrial and residential drywall products until the late 1970s; despite some reports of asbestos-related disease among individuals who used drywall-patching compounds, subsequent epidemiologic studies failed to confirm any health risks associated with using these products. A recent study of Chinese chrysotile-textile plant workers demonstrated an excess risk of lung cancer and respiratory diseases, although the small number of individuals included in the study precluded drawing any conclusions with regard to risk of mesothelioma [50].

Almost all studies concerning non-asbestos fibers as etiologic agents in malignant pleural mesothelioma are based on the assumption that any natural or man-made fiber that fits the “fiber pathogenicity paradigm” (see above) has carcinogenic potential in humans (Table 2.3).

Simian virus 40 (SV40) is a virus of Asian macaques generally thought not to be infective in humans unless artificially introduced. Large-scale human exposure to SV40 occurred between 1956 and 1966 in areas of Europe, Great Britain, and the USA as a result of widely-distributed contaminated polio vaccines grown in monkey renal-cell cultures. Approximately, 10–15 % of selected populations who were not exposed to the contaminated vaccine are reported to be seropositive for SV40, however, suggesting that other routes of human infection may exist [65, 66]. Interest in the association between SV40 infection and mesothelioma oncogenesis originates from observations in the 1960s that SV40 is oncogenic in rodents and from a study by Cicala et al. in 1993 indicating that intrapleural, intraperitoneal, or intracardiac injection of live SV40 induces pleural or peritoneal mesothelioma in 70 % of exposed hamsters [67]. In vitro studies suggest that malignant transformation of SV40-infected cells is a rare event and likely depends on the integration of SV40 DNA into the host genome. Proposed mechanisms of carcinogenesis include chromosomal damage via SV40 integration into the host genome, suppression of p53 and Rb by the SV40 large T antigen (Tag), and other direct effects of Tag [68, 69].

The role of SV40 in the development of human mesothelioma has been a topic of controversy over the past two decades. The majority of the positive evidence for SV40 oncogenicity in mesothelioma lies in the detection of SV40 DNA, RNA, or protein in patient tumor samples or the finding that SV40 is capable of altering cell proliferation and immortalizing cells in vitro. SV40- or SV40-like sequences have been detected in up to 60 % of frozen and paraffin-embedded mesothelioma samples, and immunohistochemical and western blot evidence of SV40 Tag expression in tumor tissues has been reported [68, 70, 71]. A synergistic effect between SV40 exposure and asbestos exposure on the risk of developing mesothelioma has also been proposed in humans [71]. Other studies, however, have failed to demonstrate significant amounts of SV40 DNA or RNA sequences in tissue samples, including those collected from patients known to be seropositive for SV40 [72–74]. Geographic variation in SV40 exposure has been proposed to account for the variability of results among studies. Other explanations for the variability of results in the literature have been posited. Significant sequence and antigenic overlap exist between SV40 and other papovaviruses that do commonly infect humans, including the John Cunningham (JC) and BK viruses. In addition, SV40 sequences found in commonly used laboratory plasmids may result in polymerase chain reaction (PCR) contamination and false-positive results. Pepper et al. amplified SV40 sequences in six of nine mesotheliomas by PCR; however, all SV40-positive cases were also positive using a broader primer set that amplified a sequence common to the BK, JC, and SV40 viruses [75]. Lopez-Rios et al. subsequently performed a systematic study of 71 mesotheliomas and found that 62 % of cases were positive by PCR using SV40 primers that amplify sequences, also found in commonly used laboratory plasmids, 23 % were positive using plasmid-specific primers, and only 6 % were positive for natural SV40 sequences not known to exist in laboratory plasmids [76]. Some serologic studies in mesothelioma patients using different techniques have suggested a slightly increased, although not always significant, prevalence of SV40 seropositivity in mesothelioma patients compared with control patients, but have also demonstrated inter- and intra-study variability, and the results do not reliably correspond to the presence of SV40 in matching tumor tissue [65, 73, 74].

Epidemiologic evidence of a relationship between SV40 and increased incidence of mesothelioma has not been established, due to the fact that the only definitively proven route of human SV40 infection is administration of contaminated vaccines, and it is often impossible to accurately determine individual vaccination exposure status. Vaccine contamination rates, furthermore, have varied from 10 to 100 % in different countries and in selected populations, making it even more difficult to estimate true SV40 exposure rates [68, 77, 78]. Epidemiologic studies failing to demonstrate association between SV40 exposure and mesothelioma typically examine patient cohorts who are younger than the expected median age of patients with mesothelioma; however, most SV40 exposure is thought to occur in the first few years of life and the follow-up times in many of the largest epidemiologic studies of SV40 and mesothelioma have reflected the expected 30–40-year latency period for mesothelioma development following asbestos exposure. While a few studies have found increased incidence of mesothelioma in populations potentially exposed to contaminated polio vaccine, the results have not reached statistical significance [68]. Strickler et al. found no significant increase in mesotheliomas or other cancers among exposed US populations, but the study was limited by a small patient population [79]. Studies examining cancer incidence in highly exposed populations (86–95 % of all Danish children born between 1955 and 1962 were exposed to contaminated lots of vaccine) have also failed to demonstrate a relationship between SV40 exposure and mesothelioma [78]. Other large studies in the USA and UK have similarly failed to show a consistent relationship between potential SV40 exposure and the development of mesothelioma [66, 77].

Exposure to external beam radiation has been reported as a risk factor for the development of secondary mesothelioma in the context of treatment for a variety of malignant and nonmalignant conditions, primarily Hodgkin and non-Hodgkin lymphoma, breast cancer, and testicular tumors [80–89] . Most reports consist of single cases and small series, and suggest that radiation-associated mesothelioma occurs within the radiation field after doses ranging from approximately 20 to 90 Gy, affects men and women at equal rates, and has a prognosis similar to asbestos-associated mesothelioma of the same histologic subtype. Reported latency between exposure and development of secondary pleural malignant mesothelioma has ranged from 5 to 41 years after radiation exposure [82].

Many existing studies examining the association between radiation exposure and mesothelioma are limited by small patient cohorts, inadequate information regarding radiation dose, and failure to address occupational history or asbestos exposure. A 20-year review of 1000 recipients of thoracic radiation performed in 1995 at a major cancer center uncovered three instances of presumed secondary malignant mesotheliomas, suggesting a higher incidence compared with the general population, but did not provide further information regarding latency periods, radiation dosages , prognosis , or demographic features [90]. A study published in 1996 examining nearly 1.5 million patients registered in the SEER database reported 33 radiation-associated malignant mesotheliomas [82]. Patients in this cohort were treated for a variety of thoracic, abdominal, and pelvic malignancies, had a median age of 68.5 years, latency of 4.3 years, and sex distribution similar to asbestos-related mesothelioma. Tumors occurred most frequently in patients treated for prostate, colon, and breast cancers. This study, however, included patients who developed mesothelioma within 2 months after primary diagnosis, and, importantly, did not address asbestos exposure as a possible confounder .

Subsequent studies have primarily looked at specific populations of patients (i.e., patients treated for breast cancer, Hodgkin lymphoma, etc.) using updated data from the SEER program and nationwide registries in Norway, Sweden, Finland, Denmark, the Netherlands, and Germany [91–97]. Taking into account all studies that exclude patients who develop mesothelioma within 2 months of primary tumor diagnosis , latency from exposure to diagnosis is 16–28 years, the median reported survival of radiation-associated mesothelioma is approximately 10 months, and sex distribution is similar to that of asbestos-related tumors. Epithelioid histology appears to predominate, and only in rare instances have mixed or sarcomatoid histology been described. The relative risk of mesothelioma in radiated patients with no history of asbestos exposure generally falls in the range of 1.42–3.74 but has been reported to be as high as 19.5 . Multiple factors have been proposed to alter the relative risk of developing mesothelioma and account for the variability between studies, including the type of primary cancer, sex, age at radiation , and synergistic effects of asbestos or chemotherapy exposure. DeBruin et al., for example, reported a markedly increased relative risk of 44.8 in patients who received chemotherapy in addition to radiotherapy; however, this effect has not been noted in other studies [92]. Hodgson et al. studied secondary cancers in nearly 19,000 5-year survivors of Hodgkin lymphoma and reported a significant effect of sex and age at radiation in the development of multiple secondary cancers, including mesothelioma, with female patients radiated at ages younger than 20 possessing the greatest 30-year cumulative risk [94] .

Thorotrast, a solution of thorium dioxide that emits α, β, and γ radiation, was used as an imaging contrast medium throughout Europe and the USA during the 1930s and 1940s. After use during angiography, thorotrast persists in the body, becomes concentrated in the reticuloendothelial system and has been linked to malignancies of the liver, kidney, and bone marrow, among others. The earliest report of an association between thorotrast exposure and mesothelioma described a malignant mesothelioma of the “cervical pleura” in a 43-year-old woman, 25 years after extravasation of thorotrast, during an imaging procedure, but did not define diagnostic criteria or offer a description of the tumor [98]. Subsequent larger studies have confirmed significantly elevated risks of mesothelioma in patients with both systemic and localized exposure to thorotrast in comparison to unexposed patients, despite the smaller relative doses (Gy) of radiation compared to those used during treatment of malignancy [99].

Analyses of individuals with occupational and environmental radiation exposure have suggested some increased risk of mesothelioma as compared to similar populations without known exposure to radiation, but are confounded by concurrent asbestos exposures [99, 100] .

Evidence of a genetic predisposition for the development of mesothelioma is derived primarily from reports of familial clustering, and more recently, observation of syndromic associations, whole-exome sequencing, and genome-wide association studies.