Document 23MgK5XoZz1QvnrYry2keQJR

SSX^?rrF,SISS^D The Pathogenesis of Mesothelioma Michel* Carbone, Robert A. Kratzke, and Joseph R. Testa About 80% of malignant mesothelioma* (MM) in the Western World develop in individuals with higher than background exposure to asbestos. Only a fraction of those exposed to asbestos develop mesothelioma. In* dtesting that additional factors play a role. Simian virus 4(7 (SV40), a DNA tumor virus that preferentially causes mesothelioma in hamsters, has been detected in several human mesotheliomas. The expression of the SV40 large tumor antigen in mesothelioma cells, and not in nearby stromal ceils, and the capacity of antisense T-antigen treatment to arrest mesothelioma ceil growth in vitro suggest that SY40 contributes to tumor development. The capacity of T-antigen to bind and inhibit cellular pS3 and retinoblastoma (Rb)-family proteins in mesothelioma, together with the very high susceptibility of human mesothelial cells to SV40-me- diated transformation in vitro, supports a causative role of SV40 in the pathogenesis of mesothelioma. As bestos appears to Increase SV40-mediated transforma tion of human mesothelial cells in vitro, suggesting that asbestos and SV40 may be cocarcinbgens. p53 muta tions are rarely found in mesothelioma; pl6, pMARF, and NF2 mutatkms/iosses are frequent. Recent studies revealed the existence of a genetic factor that predis poses affected individuals to mesothelioma in the vil lages of Karain and Tuxkoy, in Anatolia, Turkey. Erio- nite, a type of zeolite, may be a cofactor in these same villages, where 50% of deaths are caused by mesotheli oma. Mesothelioma appears to have a complex etiol ogy in which environmental carcinogens (asbestos and ertonlte), ionizing radiation, viruses, and genetic fac tors act alone or in concert to cause malignancy. Semin Oncol 29:2-17. Copyright 2002 by W.B. Sounders Company. ` tinguished from epithelial celts and fibroblasts be cause of their ultramicroscopic characteristics and their imnumophenotype.1 Mesothelial cells ace the adult equivalent of mesoderm, which covers the celomic cavity. In the adult, the mesoderm is called mesothelium, but to the best of our knowl edge, mesoderm and mesothelium are equivalent. Another characteristic of these cells is their high level of wild-type p53. The levels are four to five times higher than in fibroblasts, and p53 is easily detectable by immunostaining (normally only mu tated p53 is detectable by immunohistochemistry because wild-type p53 does not accumulate to suf ficient amounts to be detectable by this tech nique). Malignant mesotheliomas (MM) are very aggressive tumors that originate from mesothelial cells. Due to the tumor's aggressive nature and resistance to chemotherapy, the median survival from the time of diagnosis is about 1 year. The term "malignant" is often added to that of me sothelioma, because, unfortunately, the latter is also used to describe a number of rare and mostly benign tumors that have little, if anything, to do with MM.1 We will briefly describe these entities and thereafter use the term "mesothelioma" to indicate the classic malignant tumors of the pleura associated with asbestos exposure and recently to Simian virus 40 (SV40) infection and genetic ab normalities. ESOTHELIAL CELLS form the serosal lin BENIGN MESOTHELIOMAS Ming of the pleural, pericardial, and peritoneal Multicystic mesothelioma, also called multiloc- cavities.1 Among the most undifferentiated cells ofular peritoneal inclusion cyst, is a benign mesothe- our body, mesothelial cells are capable of differen lia! lesion, characteristically formed by multiple tiating morphologically into epithelial-like cells or cysts arranged in grape-like clusters. Adenomatoid fibroblast-like cells, Mesothelial cells can be dis- mesotheliomas are benign mesothelial lesions of the genital system. Mesothelioma of the atrioven From the Cancer Zmmunokgj Program, Cardinal Bemardm Cancer Center, Department of Pathology, Loyola University Cfiicod; Department of Medicine. Veterans Affairs Medical Center, Section of Hematology and Oncology, Minneapolis MN; and the Fox Chase Cancer Center, Philadelphia, PA. tricular node is neither a mesothelioma nor a ru mor. This lesion represents congenital heterotopia of the endodermal Sinus in the atrioventricular node. Well-differentiated papillary mesothelioma is found more often in the abdominal cavity of Supported by NJH CA 77220 and ACS 8632 to M.C. Address reprint requests to M. Carbone, MD, PhD, Cardinal Bemardin Cancer Center, Room 205, 2160 S First Ave, May wood,!!, 60153. Copyright 2002 by W.B. Saunders Company 0093-7754/02/290! -0003*35.00/0 young women. Histologically, it is formed by mul tiple papillary structures covered by cytologically benign mesothelial cells. The lesion is benign, but there have been occasional cases in which several years after diagnosis the patient developed a true doi:!O.J053/!onc.2002.30227 mesothelioma. Localized mesothelioma, better re- 2 Semwwrs in Gndaiagy, Vd 29. No [ (February), 2002; pp 2* 17 THE PATHOGENESIS Of MESOTHBJOMA 3 ferred to as localised fibrous rumor of the pleura, is thought to originate from the submesothelial cells, but it is unclear what submesothelial cells are. The tumor cells have a benign fibrous appearance and are characteristically negative for cytokeratin and positive for CD34. which suggests that these cells ate not of mesothelial origin. Occasionally, local ised fibrous tumors of the pleura are histologically and cytologically malignant and clinically these tumors are characterized by multiple recurrences after resection, and a poor prognosis. As stated above, these benign mesotheliomas should not be confused with the classic MM, which are among the most aggressive human tumors. However, to further complicate the matter, histologically ma lignant mesotheliomas have occasionally been as sociated with long survival durations of years or even decades. Whether the latter should be called MM is debatable. ASBESTOS Prior to the 1950s, MM were extremely rare. The first documented case of mesothelioma, ac cording to current diagnostic criteria, was pub lished in 1947.2 In fact, some authors even ques tioned whether mesotheliomas existed at all ' because using current methodologies and review ing 47,000 autopsies performed at the Massachu setts General Hospital from 1896 until 1947 they did not find any evidence of mesothelioma.1 From 1947 until 1990, about 100 mesotheliomas were recorded at autopsy at the Massachusetts General Hospital.3 In the last half century, the incidence of mesothelioma has significantly increased, and, cur rently, 2,000 to 3,000 cases are diagnosed per year in the United States alone/ The exact number of mesotheliomas in the United States is, however, unknown, because these numbers are projections from the Surveillance, Epidemiology and End Re sults (SEER) program. The SEER provides popu lation-based, tumor-specific daca on all histo logically proven cancers occurring in selected geographic sites in the United States, and includes approximately 12% of the US population. Projec tions of these data are used to predict the total number of different tumor types. However, the total number of mesotheliomas appears to be verydifficult to predict because the incidence of me sothelioma is higher in states with asbestos indus tries. The population that provides the SEER data was not chosen according to the presence or ab sence of asbestos sources. Furthermore, the Atlas of Cancer Mortality m the US 1950-1994, compiled by the National Cancer Institute (NCI), combines cancels of the lung, trachea, bronchus and pleura, making it impossible to study the incidence of a relatively rare cancer such as mesothelioma.5 Therefore, the incidence of mesothelioma in the United States cannot be estimated precisely. The increase in mesothelioma has paralleled a much more widespread use of asbestos in this cen tury.6'10 Approximately 80% of mesotheliomas are associated with asbestos exposure, and about 5% of asbestos workers develop mesothelioma.4-10 These observations establish an indisputable link be tween asbestos and mesothelioma4'610 At the same time, these findings indicate that most peo ple exposed to asbestos do not develop mesotheli oma, and that there are about 400 to 600 mesothe lioma cases per year in the United States alone, which are not associated with asbestos. Rare cases of mesotheliomas in children and infants also sug gest that other factors may be involved in the etiology of the tumor.4 Mesotheliomas develop in adults after a very long latency. Unless asbestos acts differently in children than in adults, cases of mesothelioma in children are not the result of asbestos. Thus, it is arguable that other factors, alone or in conjunction with asbestos, play a role in the development of mesothelioma. In addition to mesothelioma, asbestos causes two other lethal diseases: lung cancer and asbestosis,7 and two benign diseases: pleural plaques and pleural effusions.1-7 There is also some evidence that asbestos may increase the risk of laryngeal cancer.4 Asbestos and cigarette smoking have a synergistic carcinogenic effect in the pathogenesis of lung cancer.7 Smoking also increases the risk of asbestosis, because smoke inhibits the mucociliary activity and increases fiber retention. Asbestosis develops in individuals exposed to very high amounts of asbestos (Table 1). There is a threshold level below which asbestosis is not seen, and the lower the exposure, the longer it takes to develop the disease. However, only a fraction of individu als exposed to fibrogenic doses of asbestos develop asbestosis, indicating chat additional and presendy unknown factors play a major role in determining individual susceptibility.7 The relationship between asbestos and me sothelioma is unique. In contrast to lung cancer and asbestosis, smoking does not increase the risk A CARBONE* KRA7ZKE, AND TESTA Table 1. Aibeetot Expwire and Human Ditto*** Baaed an Lunf Content Analyse* Uncoatto' Fibvri7** Croridotoe7^ Controls MtiDthrilonna Asbettoili 0-20 67 690 0.14-1.00 ii(304Ptl-MM) MOO *Fib*f*/g x I03 {unewtod fibers, mosrfy ajncwtt). f Million fberi/dry lur^gt Fibers X IOVj dry kingAbbreviation: PTL-MM, peritonwJ tralignint nwsotheliontt- AnKwfte^f 0.00-0.93 103 (I00PTL-MM) 450. . Amostte't 0.7 0.9 10 of mesothelioma,11 In contrast to asbestosis, in which there is a dose-response relationship {the higher the exposure the higher the risk), this as sociation does not exist for mesothelioma (Table 1), except for chrysotile asbestos, which may be oncogenic only at high doses.7 Asbestos, in fact, is a generic name for a family of naturally occurring silicate minerals with differ ent carcinogenicities.8 The various types of asbes tos are divided into two major groups: serpentine represented by crysotile, the most common and economically important form of asbestos in the Western World; and the amphiboles, which in clude crocidolite, the most oncogenic type of as bestos, amosite, anthophyllite, and tremolite.6'10 The role of crysotile in the pathogenesis of me sothelioma is controversial, whereas the capacity of amphibole asbestos to cause mesothelioma is well established epidemiologically.8'10 There are several reason for the confusion about crysotile carcinogenicity.6,9 First, many patients are exposed to various types of asbestos. For ex ample, Canadian crysotile is contaminated with about 1% of tremolite. Some authors attribute the development of mesothelioma to this contami nant,6 others disagree,9 Second, crysotile miners have more amphibole than crysotile fibers in their lungs.7'9 This is because crysotile can be partially digested, and it is removed from the lung. Amphibole asbestos is more resistant to digestion by cel lular enzymes, and accumulates in the lungs,7 In conclusion, many patients have been exposed to both crysotile and amphibole asbestos, and it is difficult to establish the amount of exposure to the different types of asbestos. When asbestos fibers reach the alveolar re gion, the body's initial response is. an attempt to phagocytose and remove them from the lung, if the fiber is short and easily engulfed by alveolar macrophages, it is efficiently lemoved from the lung via transport to the ciliated epithelium of the smaller airways and eventually to the tra chea and throat where it is swallowed. However, if the fiber is too long to be phagocytosed, it cannot be removed unless it is solubilized, which is not possible with amphibole asbestos. Shorter fibers can also be retained if the dose is so large . that it overwhelms the pulmonary defense mechanism. Asbestos fibers have the capacity to reach the pleura either through the lymphatics, or by direct penetration, and to cause fibrosis, pleural plaques, and eventually mesothelioma.17 In the pleura, amphibole fibers are distributed heterogeneously, and are concentrated in the "black spots" of the parietal pleura.17 These black spots are the anthtacoric deposits observed on the parietal pleura of smokers and of people exposed to coal dust, and correspond to areas of deposit along the lymphatics. It has been sug gested that the concentration of asbestos in the black spots accounts for the similar location of pleural plaques and for the observation that mesothelioma usually originates on the parietal and diaphragmatic pleura rather than on the visceral pleura.12 Whether mesothelioma origi nates from or close to these black spots is un known. What happens when asbestos reaches the pari etal pleura that ultimately leads to mesothelioma is unclear. It is known that asbestos fibers cause mutagenic changes through the production of hy droxyl radicals and superoxide anions. Leading to DNA strand breaks and deletions.4'6'7,13,19 How ever, the same oxygen radicals are made in re sponse to cry5tallin silica7 (which does not cause mesothelioma), suggesting that additional factors THE PATHOGENESIS OF MESOTHELIOMA contribute to asbestos carcinogenicity and to the development of mesothelioma. Alternatively, sil ica, in contrast to asbestos, does not reach the pleura, bur this is unclear. Through mechanically interfering with mitotic segregation, asbestos can also alter chromosomal morphology and ploidy of Syrian hamster embryo cells and cause malignant transformation.15 However, human mesothelial cells are much more susceptible to the toxicity of asbestos than hamster embryo cells, and mesothelial cells that phagocytose crocidolite asbestos usu ally die and do not become transformed.16 In any event, asbestos may induce DMA damage either by direct physical interaction,15 or by the indirect action of reactive oxygen species produced by in flammatory cells in response to asbestos.6'6'7'13-1"1 Asbestos also induces DNA double-strand breaks in cultured cells exposed to chrysotile for 24 hours, and DNA repair mechanisms are critical to permit cell survival following asbestos exposure.17 Asbes tos has also been found to influence the normal activity of genes that regulate the cell cycle and that are altered in certain human cancers. Cro cidolite asbestos was found to stimulate the auto phosphorylation of the epidermal growth factor (EGF) receptor in mesothelial cells, which in cum triggers the extracellular-regulated kinase (ERK) cascade and leads to increases in AP-1 activity and either cell mitosis if the DNA damage can be repaired, or apoptosis if it cannot be repaired6-7 (Fig 1). Recent experiments using cDNA expres sion arrays to screen differentially expressed genes in human bronchial epithelial cells immortalized by human papillomavirus and subsequently trans formed by a single 7-day exposure to chrysotile asbestos demonstrated that activation of the insu lin receptor, and inactivation of DCC (deleted in colon cancer) and Ku70 may cooperate in malig nant transformation.18 It will be important to test if similar results are obtained in mesothelial cells exposed to asbestos. An additional mechanism by which asbestos may favor tumor development is mediated through the immune system, because macrophages produce DNA-damaging oxyradicals and lympholcines that depress immune function upon phagocytosis of asbestos fibers.13'14'19-20 Thus, asbestos is consid ered both a local and systemic immunosuppres sant19. S SIMIAN VIRUS 40 * (SV40) is a DNA tumor virus capable of induc ing mesotheliomas, as well as other tumors, in hamsters and of transforming human cells in vitro.4,21,22 The SV40 genome is a double-stranded circular DNA molecule containing 5243 base pairs that can be divided into two regions, early and late, according to the order in which they are transcribed. The early region encodes three pro teins, large T-antigen (Tag), small t-antigen (tag), and 17kT, and it is responsible for the transform ing ability of the virus. The late region encodes viral coar proteins. Tag binds and inhibits cellular p53, pRb, and several other tumor-suppressor gene (TSG) products.23-25 Tag is also mutagenic and causes numerous chromosomal alterations.23-25 The SV40 tag contributes to malignancy by inhib iting the activity of phosphatase 2A (PP2A) and thus altering the phosphorylation state of several cell cycle-regulatory proteins.26 These effects are highly oncogenic, and Tag is considered one of the most potent carcinogens.27 Although SV40 is endogenous to the rhesus monkey, the virus infected the human population through contaminated polio vaccines, both atten uated and killed, between 1954 and 1963.27,28 Po lio vaccines were prepared in cell Cultures grown on monolayers of infected rhesus monkey kidney cells. Since the virus produces no cytopathic ef fects in these cells, their infection went unrecog nized until 1960, when infectious SV40 was iso lated from some lots of the vaccine. As a result of this contamination, it is estimated that 96 million adults and children in the United States alone were injected with polio vaccines potentially con taminated with SV40. It was estimated that about one third of those vaccines contained various amounts of infectious SV40, which means that some 32 million people in the United States were injected with infectious SV40.4-21-22'27'28 Aside from contaminated polio vaccines, transmission of SV40 to humans may have also occurred between 1957 and 1967, when parenteral adenovirus vac cines contaminated with SV40 were used in the military and, to a limited extent, administered to the general population.21*22 In addition, a high incidence of SV40-neutralizing antibodies has been demonstrated in persons handling rhesus monkeys, as well as laboratory workers handling rhesus monkey kidney cell cultures.21-22 Aside 6 from these known exposures, ocher routes of SV40 transmission to the human population are likely. Individuals bom after 1962 with limited risk of exposure to contaminated polio vaccines had an approximately 5% to 10% positive rate for SV40neutralizing antibodies,22 Since this percentage cannot be accounted for by the aforementioned exposure events, human to human transmission may be the source. While little is known about human to human transmission of the virus, hori zontal and/or vertical transmission cannot be ruled out. In addition, children excreted infectious virus in their stools 3 to 5 weeks after ingestion of the contaminated oral polio vaccine, providing another possible route for person to person trans mission, Finally, a recent report presented docu ments indicating that SV40 may have contami nated polio vaccines even after 1963.27 Thus, whether the use of contaminated polio vaccines introduced a new virus into the human population or simply broadened the infected population is nor known.3',22,27 The actions of SV40 in different cell types are diverse.16 Monkey cells axe permissive, meaning they allow $V40 replication which causes cell lysis. In contrast, rodent cells are nonpermissive, since SV40 does not replicate in these cells and SV40 injection often leads to tumor development because the viral DNA becomes integrated into the host celt genome. Human cells are semipermissive, as they allow SV40 multiplication, and on occasion, following viral integration in the cellular genome also cellular transformation. Recently, it was found that human mesothelia! cells (HM) are unique regarding their susceptibility to SV40,'6 HM are infected much more efficiently than hu man fibroblasts and epithelial cells,'6 However, following infection, the high levels of wild-type p53 normally present in roesothelial cells bind and inhibit Tag-mediated SV40 replication; fewer viral particles are formed, and the cells are not lysed. Therefore, in mesothelia! cells, SV40 can persist in an episomal state.16 The accumulation of Tag in mesothelial cells causes a rate of malignant trans formation in tissue culture which is 1,000 to 100,000 times higher than that observed in other human cell types infected with SV40.'6 Thus, me sothelial cells are very susceptible to SV40 infec tion and transformation.16 CARBONE. KRAT2KE. AND TESTA EXPERIMENTAL MODELS Asbestos Experimental models demonstrating asbestos carcinogenicity have been established in several animals.29 In 1969, Wagner and Berry injected 96 rats intrapleurally with amosite, crocidolite, and chrysotile asbestos and found that mesotheliomas developed in 38, 61, and 55 rats, respectively.30 Therefore, in contrast to the data in humans, in animals, all types of asbestos can induce tumors without notable differences. It was determined that the likelihood of mesothelioma development in rats was directly proportional to the quantity of asbestos injected.30 Hamsters, similar to rats, de veloped mesotheliomas upon receiving intrapleu ral injections of asbestos, and the incidence of the malignancy was also demonstrated to be directly proportional to the dose administered: 20% and 4% of two groups of 50 hamsters each developed mesothelioma when injected with 10 and 1 mg of crocidolite asbestos, respectively.30 In humans, a dose-response relationship between asbestos expo sure and mesothelioma was not found (Table 1). It should also be noted that 10 mg of crocidolite occupies a volume of about 1 cm3, which is a very ' large volume considering the chest volume of a rat. Mesotheliomas in these animals occurred after 18 months, were often small, and were rarely (<10%) the cause of death.31 The adjacent pleura typically showed evidence of fibrosis and local invasion was rarely observed.31 Inhalation studies are considered by some au thors to be a more realistic model of human me sothelioma. When rats were exposed to inhalation of equal weight concentrations of different asbes tos types (10.1 to 14.7 mg/L), the incidence of mesothelioma was one in 146 of those exposed to amosite, two in 145 of those exposed to anthophyllite, four in 141 of those exposed to crocidoSite, four in 137 rats exposed to Canadian crysocile, tero in 144 rats exposed to Rhodesian crysotile, and zero in 126 of the control, nonexposed group.32 After the conclusion of the 2-year exper iment, Wagner et al measured the amount of as bestos retained in the lungs. They found that the weight of amphibole asbestos was about 15 times higher than that of crysotile even though the air the rats breathed contained similar amounts by weight of both asbestos types.32 These results in dicated that crysotile fibers are less persistent in ! THE PATHOGENESIS OF MESOTHELIOMA the lung, and that in animals all types of asbestos are equally oncogenic. SV40 . The oncogenicity of SV40 has been studied mostly in hamsters. Sixty percent of hamsters in jected intracardially with SV40 developed pleural mesotheliomas within 6 months, and when SV40 was injected into the pleural spaces of these ani mals, 100% developed pleutal mesotheliomas in 3 to 6 months-33 Mesotheliomas, however, did not develop in hamsters when SV40 was injected in tracerebrally, subcutaneously, or in the femoral vein. These observations suggest that the route of SV40 inoculation influenced the induction of spe cific tumor types. It is possible that in the case of intracardial injection of SV40, small amounts of virus reached the pleura through the external sur face of the needle. Hamster mesotheliomas were large, mukicentric, and histologically veTy aggres sive, with both epithelioid and sarcomatoid areas, and invasion of adjacent tissues.31'33 No fibrosis of the adjacent pleura was observed. Mesotheliomas were uniformly fatal in hamstets within 3 to 6 months. Thus, in hamsters, there are striking mor phological and clinical differences among me sotheliomas caused by SV40 and those caused by asbestos.31 ASBESTOS EXPOSURE It is generally agreed that background exposure (ie, most people who live in industrial areas or in large cities contain some asbestos fibers in their lungs) does not constitute a health risk, and that mesothelioma and other diseases are associated with higher than background levels of exposure. However, measuring higher than background lev els of exposure is a difficult task prone to errors. Some studies rely on history of exposure, while others look to the amount of asbestos fibers found in the lungs (lung content analyses). Studies based on history of exposure suffer from the inherent imprecision of this methodology. Some history studies are based on patient inter views, others on interviews of relatives or friends, others on death certificates or medical records. Not surprisingly, these studies produce quite dif ferent results, and the proportion of asbestos-asso ciated mesothelioma in the literature varies from 16% to 90%.M Some patients (or friends and rel atives of the patients) may not be aware that they 7 were exposed; others may erroneously think that they were exposed to asbestos, when in fact they were exposed to dust that did not contain asbestos. Histories are often taken decades after exposure and it is often impossible to confirm the patient's recollection of the events. Furthermore, at least for the studies conducted in the United States, there is a strong incentive to report some type of expo sure because the US Courts of Law award substan tial compensatory damages to patients with asbes tos-associated mesotheliomas. Even when these histories are based on cohorts of asbestos workers, it is often very difficult to quantify exposure.10 Different methodologies have been used to mea sure asbestos fiber concentrations in the work place air. Different workers may be exposed to different amounts of asbestos because of different jobs within the same company. These variables make precise estimates of exposure problematic.10 The studies based on lung content analyses may be more reliable because there is an objective amount of asbestos to measure. However, these studies also face difficulties because the amount of asbestos varies in different areas of the same lung, and because the results obtained are so dependent or the instrumentation used, and other unknown factors, that interlaboratory comparisons are con sidered unreliable.7 Therefore, the results of lung content analyses can be compared only with the standards of the laboratory that has conducted the analyses (Table 1). Within these limitations, stud ies of lung content analyses offer an objective measure of asbestos , exposure. However, relatively few of such studies have been conducted because lung tissue is seldom available. A review of some of the published lung content studies seems to indi cate that the incidence of mesothelioma is not directly related to the amount of exposure. Sur prisingly, a higher incidence of mesothelioma was observed in individuals with above background exposure but with exposures considerably lower than those observed in patients with asbestosis (Table 1). The mean latency for asbestosis was 12,6 to 20.2 years following exposure,7 compared to 32 years for mesothelioma35; thus, some individuals may die of asbestosis before they can develop mesothelioma. However, only a fraction of individuals exposed to high levels of asbestos develop asbestosis,7 and most mesothelioma patients do not appear to have very high levels of asbestos fibers in their lungs 8 Mesothelioma Pathogenesis: Asbestos and SV40 Fig L MmthdiDtnl ptthopneihi postlbto pathogenic mechanisms of mhastot and SY4D. Arrowhudi Indicate a it!m blatory effect; crossed ban Indicate an Inhibitory efface Tag, SY40 large TmHiwi: tag, SV40 small feantigen; ROSt reactive oxygen species; TSG, tumor-wpprator genei; PP2A, phospha* tase 2A. The circle Indicates nonfntegrated SV40; however, In some tumors SV40 has been found Integrated in the host celt DMA; NF2, neurofibromatosis type 2 gene (often deleted in mesothelioma4'7'), (A) Asbestos Induces autophosphorylatiOn of the EGF receptor, which ectivates the mitogen-activated protein (MAP) kinases, and through a set of Intermediaries this process Induces AP-I activity, which leads to cell mitosis and apoptosis.4*4 If the masothelial eeU contains alteration* of pi 6, a CDK Inhibitor, or of the related pl4 AKF, which Inactivates MDM2 and thus interferes with the pW and pRB cellular path ways, cell division ratherthan apoptosis may ensue. Similarly, If the metotheHnJ nil It Infected with 5V40, the Tajpmedlated Inhibition of the cellular damage repair pathway and the antiapoptotic activity oflow amounts of Tag may favor cell dhrfdoiv Furthermore, tag-mediated Inhibition of cellular PP2A, which normally dephosphorylates'and thus Inactivates the MAP Idnaset, may increase the effects of asbestos on AP-I. Mesothe liomas often contain alterations of pit and pi4 ARF, and of several other cell regulatory genes.4 The precise mechanisms that cause these alterations are unknown. (B) The figure pro poses chat the mutagenk ROS produced by mononudear phagocytes following ingestion of asbestos fibers, and the mu tagenic activity of Tag, may both contribute to genetic dam age* Damaged cells will most often execute apoptosis, but rarely they may dhide, accumulate additional genetic damage, such as Ip and NF2 deletions which are often found in mesothe* Home,4 and become malignant. CARBONE, KRATZKE, AND TESTA (Table 1). It is also possible that very high levels of asbestos are so toxic for mesothelial cells that few, if any, cells are left alive near conspicuous asbestos deposits. This is supported by tissue culture exper iments which demonstrate that relatively high lev els of crocidolite asbestos {>2.5 jig/cm!) cause 100% mesothelial cell death within 1 week.16 This hypothesis, summarized in Fig 2, surmises that individuals with higher than background levels of asbestos exposure have an increased risk of devel oping mesothelioma. When the levels of exposure are very high, most mesothelial cells neat amphibole asbestos deposits may be dead and, therefore, there may be fewer potential targets for malignant transformation. This may account for the unex pected findings presented in Table 1. The possible lack of a direct correlation be tween amount of exposure and mesothelioma (Ta ble 1) further suggests that individual susceptibil ity36 and/or additional carcinogens4 play a major role in determining, among exposed individuals, those who will develop mesothelioma. For exam ple, several studies have reported mesothelioma in the wives of asbestos workers (indirect exposure), especially asbestos miners.37 Apparently these women were exposed by washing their husband's FI, 2. Working hypothesis to account ftf the finding! re ported in Tahle I (thick arrow Indict*! & higher frequency compared to the thin arrow). Top: High level! of crocidolite or amoiite asbestos Induce > high percentege of mesotheilal cell death; few mesotheilal cdli remain. Because the number of potential target! for malignancy (ie, the mesothelial cells) is reduced, ubestwis is more frequent than mesothelioma. Mid dle: Moderate exposure above background induces some mesothdlal tell death; however, many mesothelial cells survive and an occasional ceil may become transformed. Bottom: At background levels, a higher risk for mesothelioma or asbestosh Is relatively rare because the amount of exposure Is insufficient to cause disease. THE PATHOGENESIS OF MESOTHELIOMA 9 work clothes, which were contaminated with as bestos. One would surmise that these women were exposed to amounts of asbestos considerably lower than their husbands. The minimum length of time of indirect exposure was 5 years,37 but most me sotheliomas occurred in women whose husbands had worked for 10 or more years in the asbestos industry.37 It has been suggested that for a given amount of asbestos exposure, prolonged exposure to moderate amounts is much more dangerous that intense exposure to the same amount within a brief interval of time.10 A high incidence of mesothelioma in certain families36-33-40 suggests the existence of a putative mesothelioma susceptibility gene. The limited sire of these families, however, hampers the possibility of isolating this putative gene (see below). An alternative explanation is that the entire family was exposed through the contaminated clothes of a family member,36-3S-3? or worked with a contam inated asbestos oven.40 Even if the entire family was exposed to asbestos, the incidence in several family members is too high to be attributed solely to asbestos, because the incidence of mesotheli oma in asbestos miners is 5% or less.10 Thus, exposure to asbestos alone should not account for mesothelioma in several family members, unless additional factors, such as a genetic defect or viral infection, made the members more susceptible to asbestos carcinogenicity. ASBEST05 AND HUMAN MESOTHELIOMA In a mortality study of 1,225 deaths among 7,317 men who worked in crocidolite or amosite mines in South Africa, there were 30 mesothelio mas.10 Twenty of these were in crocidolite miners (in 423 deaths, 4-7%), four were in amosite miners (in 648 deaths, 0-6%), and six in miners with mixed exposure (in 154 deaths, 3.9%).10 Among crocidolite-associated mesotheliomas, six devel oped in individuals exposed from 12 to 95 months, six in miners exposed for 96 to 191 months, and eight in miners exposed for more than 192. months. Among the 10 amosite or mixed exposure mesotheliomas, four developed after exposure of 3 to 11 months, two in miners exposed from 12 to 95 months, and four in miners exposed for more than 192 months. The mean latency of mesothelioma from exposure was 44.9 years for crocidolite min ers, 51 years for amosite miners, and 12-3 years for miners exposed to both crocidolite and amosite. This study indicated that crocidolite is more dangerous chan amosite in humans.10 It also sug gested that there is a threshold level below which the increase in mesothelioma risk cannot be de tected.10 In fact, no mesotheliomas were observed in miners with a period of exposure below 3 months, and the higher risk of mesothelioma was in miners exposed for mbre than 16 years. Lanphear and Buncher35 reviewed 21 reports of a total of 1,690 mesotheliomas and found that 99% of mesotheliomas occurred 15 years or more after asbestos exposure, 96% had a latent period of at least 20 years, and none occurred less than 10 years from exposure. The mean latency was 32 years. What happens during this considerable la tent period remains unclear, but two major theo ries can be suggested. The first postulates that malignant transformation occurs relatively soon after asbestos exposure, but that it takes a long time (years) for the tumor to grow. If true, this scenario would have important clinical implica tions, because it implies that, similar to carcinoma in situ of the cervix and to colon adenomas, we should have many years to detect mesothelioma in its early stages when it can be cured by surgical resection. Unfortunately, mesothelioma, in con trast to cervical carcinomas or colon adenomas, does not have a long premalignant noninvasive phase. If such phase exists, it lasts for a short period of time, because it is nor clinically detect able by current methods.1 It should be noted that other benign lesions of the pleura, such as fibrous tumors, are detected clinically, because these le sions cause pain and accumulation of fluid in the pleural cavity. The second theory argues that genetic alter ations induced by asbestos accumulate over time and eventually after about 30 years lead to a ma lignant cell. This theory requires that either the same cell or cell lineage accumulates damage over more than 30 years. It may appear unlikely that, given the high toxicity of asbestos for mesothelia! cells ,16 a damaged cell or its daughters could sur vive for 30 or more years in the presence of asbes tos while accumulating sufficient genetic damage to become malignant. However, mesothelial cell transformation may be more likely if a key regula tory gene, such as the INK4a/ARF locus on chro mosome 9p21 is deleted or silenced by an epige netic mechanism, ie, promoter methylation- In this event, additional mutations may accumulate 10 relatively rapidly. INK4a/ARF codes for pi6, a cyclin-dependent kinase inhibitor, and for pl4ARF (the murine homolog is pl9ARF), which promotes MDM2 degradation and thus prevents MDM2 neutralization of p53. Both pl6 and pl4ARF are often deleted in mesothelioma.4,25 Most mutations may be incompatible with cell survival, but occa sionally this process may lead to a malignant phe notype. The likelihood of this process may be increased if mutations accumulate rapidly. Once the mesothelial cell has reached the malignant transformed phenotype, cell growth will be very rapid and the tumor will soon become clinically detectable. This scenario, would account for the clinical observation that malignant mesotheliomas are very aggressive and fast growing malignancies which develop 20 to 50 years after asbestos expo sure, Figure 3 summarizes these concepts. SV40 AND HUMAN MESOTHELIOMA The discovery that SV40 caused the develop ment of mesotheliomas in hamsters led to poly merase chain reaction (PCR) analysis of human mesothelioma specimens for the presence of the virus. Altogether, 29 of 48 (60%) mesothelioma samples contained SV40 sequences. SV40 Tag expression was also detected by both immunohistochemistry and immunoprecipitation experi ments.41 These results were confirmed by an inde pendent multilaboratory srudy organized by the International Mesothelioma Interest Group,42 and by more than 26 other laboratories worldwide.43'44 Two studies did not detect SV40 in mesothelio mas.43 Furthermore, Hirvonen et al, at the Finnish Institute of Occupational Health, were unable to find SV40 in Finnish mesotheliomas, but in the same study US mesotheliomas tested positive.45 This finding suggests that there may be marked geographical differences regarding the presence of the virus in human tumors. These differences may be related to the fact that SV40-contaminated vaccines were not administered in Finland,45 to the markedly different amounts of contamination among different batches of poliovaccines distrib uted in the United States and Europe,2 l'22-2T'2a or to other unknown factors. Aside from mesotheli omas, SV40 has also been detected in human tumors of the same types that it induces in exper imental animals, including ependymomas, choroid plexus tumors, and osteosarcomas-43 Several tech niques, including PCR, in situ hybridization, mi CARBONE, KRATZKE. AND TESTA crodissection, and immunostaining, have shown that SV40 sequences or antigens are present in tumor celts but not in normal adjacent tissue.46"49 Caution should be used, however, when using anti Tag monoclonal antibodies, because some prepa rations may be contaminated with a protein that causes false-positive reactions.50 These false-posi tive reactions cause a cytoplasmic staining, and thus can be distinguished from true Tag staining, which is nuclear. Furthermore, contaminated Tag monoclonals stain 100% of the cells, regardless of cell type.50 In contrast, pure Tag monoclonals only stain the nuclei of mesothelioma cells and not the nearby stromal cells,41 and stained only two of 38 human mesothelial cell cultures: approximately 96% of the cells of these two cultures and none of the cells of the other 36 SV40-negacive mesothe lial cell cultures.51 Experiments to investigate if Tag was biologi cally active in human mesotheliomas found that Tag and p53 were often coexpressed.49 Immunoprecipitation reactions demonstrated that Tag binds and stabilizes p53 in mesotheliomas, allow ing its detection.49 Furthermore, tumor cells ex pressing Tag failed to induce p21, indicating that p53 was inactivated by its interaction with Tag.49 Treatment of SV40-positive mesothelioma cell lines with Tag antisense restored the p53 pathway and induced p21 expression and growth arrest, which suggests that Tag expression was required for malignant transformation.52 Aside from p53, Tag was also shown to bind, stabilize, and inacti vate members of the retinoblastoma (Rb) tumorsuppressor family--pRb, pl07, and pRb2/pl30 in human mesotheliomas,25 and to cause the inacti vation of RassFl A genes.53 SV40 particles were detected by electron mi croscopy in human mesothelioma cells, and in these cells SV40 caused hepatocyte growth factor production and the activation of the MET onco gene, which in mm promoted mesothelial cell growth.54 These findings were recently presented at a consensus meeting at the University of Chi cago; the consensus was that (1) it has been con vincingly demonstrated that SV40 is present in about 40% to 50% of human mesotheliomas in the United States; and (2) that the role of SV40 in the pathogenesis of mesothelioma has been consider ably strengthened in the past 4 years.44 THE PATHOGENESIS OF MESOTHELIOMA RADIATION AND HUMAN MESOTHELIOMA There have been several case reports in which mesothelioma developed in patients who had re ceived radiation to the thorax or to the abdomen, or who had received Thorotrast inuavascularly.55 Mesothelioma in these patients developed after an interval of 7 to 36 years. In many of these cases, the possibility that asbestos or other factors had contributed to mesothelioma could not be entirely ruled out. However, at least for those mesothelio mas which developed in young adults who had received intensive radiotherapy in their childhood because of Wilm's tumor, radiation appeared as the only possible causative factor. In these cases, a portion of the lung, and thus of the pleura, was included in the radiation fields of the abdomen. Lung content analyses studies were performed in one of these cases and revealed background levels of asbestos only, supporting radiation as the caus ative factor of that particular mesothelioma.5* Studies in rats support the notion that radiation exposure can cause mesothelioma: 27% of rats exposed by imraperitoneal injection to IWPuOz developed epithelial mesothelioma, and 38% de veloped sarcomas, possibly sarcomatous mesothe liomas.57 It seems safe to state that intensive radi ation exposure can cause mesothelioma; however, the overall number of mesotheliomas caused by radiation is probably smalL GENETICS AND HUMAN MESOTHELIOMA In the villages of Karain (population -600) and Turkey (population ~ 1,400) in Cappadocia, a re gion in Central Anatolia, Turkey, characterized by volcanic tuffs and natural caves, 50% or more of deaths are caused by malignant mesothelioma. These two villages, like most others in the region, were built with stones mined from the nearby natural caves. When Dr Y.I. Barish found a very high incidence of mesothelioma in Karain,5a sci entists looked for asbestos, which in the 1970s was the only known causative factor for mesothelioma. Some asbestos was found,55 but subsequent studies demonstrated that in Cappadocia, asbestos is al most everywhere because it is a natural component of that volcanic terrain and because asbestos-based stucco (containing tremolite asbestos) has been widely used in building construction.60 It appeared that asbestos could not account for the unique II high incidence of mesotheliomas in these two vil lages.61 Another type of mineral fit*r, erionite, which had been detected in the lungs of several villagers, was suspected to be the causative agent.58,61 Erionite is a type of fibrous zeolite com monly found in the stones of the houses of Karain and Tuzkoy. When injected intrapleurally into animals, erionite causes mesothelioma, and it was concluded that erionite was the cause of mesothe lioma in these villages.61 Erionite therefore ap peared much more potent than asbestos in causing mesothelioma because more than half of the vil lagers died of this disease. Actually, erionite was claimed as the most potent chemical human car cinogen.61 Studies tried to link erionite to other human tumors, but except for mesotheliomas, there is no significant difference in the incidence of any other tumor types in these two villages compared with the rest of Turkey,67 Why such a potent carcinogen would specifically cause me sothelioma was unknown. Why about 50% of vil lagers appeared to suffer no consequence from ex posure to such a potent carcinogen was also unknown. During repeated visits to these villages one of us (Carbone) suspected that erionite was not the only cause of mesothelioma. In the villages of Karain and Tuzkoy, mesothelioma developed mostly in certain houses, called "the houses of death" because all the residents had died of me sothelioma. Residents of nearby houses had not developed mesothelioma. It was stated that the amount of erionite was higher in the houses of death, bur the only proof for this statement was that residents of those houses had died of mesothe lioma. If this hypothesis was correct there had to be an extremely high amount of erionite in the houses of death, and proportionally a very low amount of erionite in the other houses next to the houses of death. Yet, both types of houses appeared to have been built at about the same time and with the same type of stones. In Karain and Tuzkoy, families live together for multiple generations; therefore, people living in the same house ate often related. It appeared possible that susceptibil ity to mesothelioma was genetically transmitted and that the presumed higher amount of erionite in the houses of death was a misleading hypothesis, The genetic hypothesis was strengthened by the observation that only one mesothelioma had been observed in Karlik, 1.5 miles south of Karain, with a population of about 1,500. This mesothelioma II ' had occurred in a woman from Karain who had moved to Karlik because of marriage. The houses of Karlik' appeared identical to those of Karain and had been built with stones mined from the same caves which are located midway between Karain and Karlik on the same side of the mountain on which these two villages were built. Again, the argument was made that there was more erionite in the houses of Karain, but during a local inspec tion of Karlik and Karain, one of the highest amounts of what appeared to be erionite (erionite has. a white color and a soft consistency and looks like an area of decay within the zeolite stones used to build the villages) was found in the stones of the fountain of Karlik, which provides water to the whole village. The marriage of the woman of Karain with a man of Karlik was unique. Residents of Cappadocia consider people from the two me sothelioma villages "weak" and they are afraid that mesothelioma will spread in their families if they marry with villagers of Karain and Tuzkoy. In fact, people from Karain and Tuzkoy have problems even selling their products to the market because residents of nearby villages are afraid to buy any thing that comes from these two villages61 (M. Carbone, personal observations). The conse quence of this incredible situation is that most marriages in Karain and Tuzkoy are between vil lagers and marriages outside the villages are infre quent. Since there was reason to suspect that me sothelioma in Karain and Tuzkoy had a genetic cause, Dr Carbone organized a research team to study this possibility. Iman Roushdy-Hammady, a PhD student in Medical Anthropology at Harvard University, with extensive experience with the Muslim culture and who is fluent in Turkish, agreed to live among the villagers for 2 years 8 months to collect information. She constructed preliminary pedigrees from several families. When these pedigrees were further analyzed by a geneti cist, it was apparent that mesothelioma was genet ically transmitted, possibly as an autosomal-domi nant disease.63 About 50% of the descendenrs of affected parents developed mesothelioma, whereas mesothelioma was absent in other families. When members of unaffected families married into af fected families, mesothelioma appeared in their descendents.63 Mineralogical analyses revealed no differences in the amount or type of erionite among the houses of death and other houses, in cluding those of Karlik (Umran Dogan, University CARBONE, KRATZKE, AND TESTA of Ankara, personal communication). These same analyses confirmed the presence of a very high amount of erionite in the fountain of Karlik. Thus, it appears that in these villages susceptibility to mesothelioma is genetically transmitted. The pos sibility that erionite t$ a cofactor that contributes to mesothelioma in genetically predisposed indi viduals is presently being investigated in our lab oratories. Furthermore, we will try to isolate the putative mesothelioma susceptibility gene that predisposes some families of Karain and Tuzkoy to this malignancy. This should lead to the develop ment of therapeutic approaches for members of these families. It is possible that the same gene that is genetically mutated in Cappadocia may be the target of asbestos and SV40 carcinogenesis, and thus it is hoped that the eventual isolation of (his putative gene might clarify the molecular pathogenesis of mesothelioma, and also benefit all mesothelioma patients. GENETIC ALTERATIONS IN MESOTHELIOMA Whatever the cause, genetics, asbestos, SV40, or radiation, mesothelial cells eventually accumu late a number of genetic alterations and become malignant. Human mesothelioma cells are charac terized by many different genetic abnormalities and separating those that play a causative role during the establishment of the transformed phe notype and tumor progression from those that do not has not been possible to date, particularly since data on early lesions are not available. Karyotypic studies and comparative genomic hybridization (CGH) analyses have revealed mul tiple chromosome alterations in most human mesotheliomasA25-66 Although a specific chromo-. soma! change is not shared by all mesotheliomas, several prominent sites of chromosomal loss have been identified. Deletions of specific regions in the short (p) arms of chromosomes 1, 3, and 9 and long (q) arm of 6, 13, and 15 are repeatedly ob served in these tumors. Loss of a copy of chromo some 22 is the single most consistent numerical change seen in mesotheliomas. Monosomy 4 and monosomy 14 are also common. These recurrent losses frequently occur in combination in a given tumor. Chromosomal gains appear to be less com mon than losses in this disease, although recurrent gains of 5p and 7 have been reported. It is not possible to determine a temporal sequence of these THE PATHOGENESIS OP MESOTHELIOMA 13 pathogenetic events due to the tact of data for early stages of this disease. The high frequency of genomic losses in mesothelioma is consistent with a recessive mechanism of oncogenesis. Loss and/or inactivation of TSGs residing in recurrent sites of chromosomal deletion is thought to contribute to the development and progression of mesotheli' oma. As a prelude to the isolation of these genes, investigators have begun to molecularly map these regions by performing loss of heterozygosity (LOH) analysis with polymorphic DNA markers. Such molecular studies have demonstrated high frequencies of allelic loss from lp22, 3p21, 4q, 6q, 9q21, 13ql3-14, I5qll-15, and 22ql2. To date, TSG loci within two of these regions have been shown to be frequently altered in mesotheliomas. Tumor-Suppressor Genes The TSGs in mesothelioma have been reviewed extensivelyA25'61 A high frequency of mesotheli oma cell lines exhibit homozygous deletion of the 9p21 region. The CDKN2A locus, which encodes the alternative TSG products pl6INK4a and pI4ARF, is located in this region. The pl6 protein is capable of binding to the cyclin-dependent ki nase CDK4, thereby inhibiting the catalytic activ ity of the CDK.4/cyclin D enzymes. Therefore, loss/inactivation of pl6,NK4" would lead to cell cycle deregulation through the loss of a key inhib itor of Gl/S progression. More than 80% of me sothelioma cell lines have homozygous deletions of one or. more pI61NK4a exons, and most of the remainder have greatly downregulated expression of pl6fNK4`\ due to promoter hypermethylation. PCR analyses have documented homozygous dele tions of pJ6,NK4" at a much lower frequency in mesothelioma tissues than in cell lines. This dis crepancy may be an in vitro phenomenon result ing from the selective advantage provided by pjgtNWa deletion during the culturing process. On the other hand, mesothelioma samples contain a significant amount of contaminating normal stroma, which can mask the existence of a ho mozygous deletion in the malignant cell popula tion. Fluorescence in situ hybridization analysis has revealed reduced copy numbers of the pi gene in most mesothelioma specimens, particularly in sarcomatous and biphasic tumors. The majority of these tumors show homozygous loss of the INK4a locus in either all tumor cells or in a significant subset (>10%) of the tumor cell pop ulation, Immunohistochemistry studies suggest that loss of pl61NK4a expression is a universal find ing in mesothelioma tissues.65 Taken collectively, the available data suggest that loss of one copy of pjgiNfOta many 0f these tumors, with the remaining allele silenced by promoter hypermethylation. In a given tumor, a subset of cells may contain homozygous loss of pI6INK4a, with these cells presumably having a proliferative advantage when placed into long-term culture. Importantly, alteration ofpi 6D'*4`' appears to play a critical role in mesothelial cell tumorigenesis, because re-ex pression of pl6,NK4a in mesothelioma cells has been shown to Tesult in cell cycle arrest and apo ptosis, as welt as inhibition of tumor formation and diminished tumor size.66 Homozygous deletions of pI6!NK4<! would in many cases also lead to the inactivation of pJ4ARF, because these two genes share exons 2 and 3, although their reading frames differ. The product of the pI4ARF TSG is required for activation ofp53 in response to the action of oncogenes such as RAS. Since the product of the pl6,NK4a gene in duces a Gl cell cycle arrest by inhibiting the phosphorylation of pRb, homozygous loss of pI4ARF and pI6!NK4a would collectively affect both the p53- and pRb-dependent growth-regulatory pathways, respectively. Recent experiments have demonstrated that adenovirus-mediated transfer of PI4ARF jn mesothlioma cell lines induces Gl phase arrest and apoptosis.67 Together, the avail able data suggest that alteration of either product of the CDKN2A locus, ie, p!4AfiF or pl6,NK4a, contributes to the pathogenesis of mesothelioma. The success of the experiments with p!4AEF and pi 6iN*'4" adenoviral constructs have led investiga tors to propose that such gene therapy-based ap proaches may prove useful in the treatment of mesothelioma patients,68 Loss of pI6 and pI4ARF is also associated with loss of pi5, although the im portance. of this loss to mesothelioma is un known.69 Loss of pl6 is usually associated with retention of wild-type pRB.15,70 p53, instead, is rartly mutated in mesothelioma,71'72 possibly be cause it is inactivated by the interaction with the SV40 Tag.49 Genetic and epigenetic inactivation of pi6 and p/4ARF, together with the inhibitory binding of Tag with p53 and pRb-family proteins, may lead to the complete inactivation of the nor mal cellular pathways of these cellular genes and H CARBONE. KRATZKE, AND TESTA 30 YEARS Rg 3. Working hypothesis to accountfor mesothelioma latency* D, DMA damage* According to recent research, at least 3 distinct genetic damages are required to cause malignant transformation of human cells* Tpr Asbestos, represented on the left as a ferruginous body with a dumbbell configuration, induces DNA damage, which Jeads to mesothdial cell proliferation and to a precursor lesion. Additional DNA damage over the years leads to mesothelioma. Middle: Asbestos Induces DNA damage which leads to apoptosis1* (thick arrow). Rarely, the cell survives and eventually continued exposure to asbestos over the years will cause a second DNA damage* Again, most Cells wifi die, but a rare cell may survive- A third DNA damage Will cause apoptosis in most cells, but a rare cell may exceptionally survive all 3 genetic changes and become malignant* Bottom* The first genetic damage Induced by asbestos Involves a cell-cycle regulatory gene such as pi 6 or the cells are infected with SV40, and therefore, the damage/repair and apoptotie pathways are Inhibited. In cither circumstance,a greater number of cells may escape apoptosis and complete mitosis. When celt cycle regulatory genes and/or DNA damage repair genes arc altered, additional mutations may accumulate more easily* the apoptotie pathway Will be less efficient, and the celt may have an Increased possibility of surviving and of becoming malignant. contribute to malignancy. Finally, in contrast to lung cancer, ras mutations are rare in mesothelioma.71'72 Extensive LOH analysis of chromosome 22 losses in mesothelioma has not been performed, because an entire copy of chromosome 22 is lost in most cases. Although the neurofibromatosis type 2 TSGe, NF2, predisposes affected individuals pri marily to tumors of neuroectodermal origin, so- FI, 4. Working hypothesis Illustrating that asbestos by In terfering with the host immune system1* may favor the growth of calls expressing (Vft antigens. Left) SVfWsfsettd mesotheHal eelli express high levels of viral antigens- CDS T lympho cytes tee and kill these cells with the help ofantigen-presenting cells (APC). Right] Asbestos directly and indirectly (it, by in ducing the release Of various cytoldnes by macrophages which phagocytes* asbestos fibers**) interferes with the normal ac tivity Of both CD6 T lymphocyte] and APC. The immunosup pressive effect] of asbestos, together with low levels of ex pression of SV40 antigens often observed In SVdB-poiltivc mesothelioma eelIt," may allow SV40-po*itive mesothelial calls to escape Immune detection and become malignant. I I | THE PATHOGENESIS OF MESOTHELIOMA IS matie mutations of NF2 have occasionally been identified in seemingly unrelated malignancies. NF2 somatic mutations predicting either imeretitial in-frame deletions or truncation of the NF2 gene product (merlin) have been reported in 40% to 55% of mesothelioma cell lines. In many cases, it was possible to confirm the mutation in matched primary tumor DNA.73 In some samples that showed NF2 gene transcript alterations, no genomic DNA mutations were detected, suggest ing that aberrant splicing may constitute an addi tional mechanism for merlin inactivation.73 West . em blot analyses revealed complete absence of merlin expression in cell lines that exhibited alter ations of the NF2 gene, suggesting that truncated forms of the protein are unstable. LOH analyses documented allelic losses at the NF2 locus in more than 70% of mesothelioma cases. All cases exhib iting mutation and/or aberrant expression of NF2 showed allelic losses, implying that inactivation of NF2 in mesothelioma occurs via a "two-hit" mechanism.4'73 Crocidolite asbestos has been shown to induce the proto-oncogenes c-fos and c-jun, which encode transcription factors that activate various genes criticat in the initiation of DNA synthesis. The persistent induction of these transcription activa tors following asbestos exposure may enhance cell division. Mesothelioma may begin with stem cell proliferation via such an epigenetic mechanism, followed by the progressive accumulation, over several decades, of spontaneously occurring muta tions in TSGs and asbestos-induced missegregation of chromosomes. Thus, such activation of pioto-oncogenes and inactivation of TSGs may cooperate in a multistep process to regulate critical events Intrinsic to the pathogenesis of mesotheli oma. The presence of SV40 may accelerate malig nant transformation, because SV40 induces DNA alterations. SV40 may give the cell more time to accumulate sufficient genetic alterations, because SV40 Tag, through its inhibitory binding of p53 and its BCL-2 like domain, inhibits apoptosis (very high levels of Tag, however, can induce apoptosis). Furthermore, the very recent fending that SV40 immortalizes infected human mesothelial cells by inducing telomerase activity, indicates that the presence of SV40 induces a background of cells prone to develop into malignancies when genetic damage is induced.74 At the same time, asbestos by interfering with the immune system may favor the emergence of malignant cells that express SV40 antigens. According to shis hypoth esis, exposure to both asbestos and SV40 may increase the risk of mesothelioma in some individ uals. Figures 1 and 4 summarize these concepts. CONCLUSIONS Most mesotheliomas in the Western World are caused by exposure to asbestos. Certain types of asbestos, especially crocidolite asbestos, appear more carcinogenic than others (crysotile). SV40 in the United States and in parts of Europe, and genetic predisposition in the villages of Karain and Tuzkoy, Have recently been linked to mesotheli oma. Rare cases of mesothelioma have been asso ciated with radiation exposure. Whether these fac tors act independently or interact in causing mesothelioma is currently under investigation. Both scenarios appear possible. There is some ev idence that SV40 and asbestos might be cocar-' cinogens in human mesothelial cells in tissue cul ture. Genetics and erionite may cooperate in causing mesothelioma in Karain and Tuzkoy. NF2, pl6, andpMARF mutation/loss are the most com mon acquired genetic lesions detected in mesothe lioma. Mesothelioma may be a cancer in which genetics, radiation, viruses, and environmental carcinogens such as asbestos and erionite interact to cause malignancy. The eventual future isolation of the putative mesothelioma susceptibility gene that causes mesothelioma in Cappadocia may rep resent the key to understanding the precise patho genesis of mesothelioma and to developing effec tive therapeutic approaches. ACKNOWLEDGMENT We ate in debt to Drs B. Mossman, U. Saffiotti, and N. Vogelsang for critical leading of this manuscript. REFERENCES 1. Batrifora H, MeCaughey WTE: Tumors of the serosal membranes, in: Rosai ], Sobin LH (eds): Atlas of Tumor Pathology- Third Series, Fascicle 15. Washington, DC, Aimed Forces institute of Pathology, 1995, pp 17-99 2. Case Records of the Massachusetts General Hospital (case 33111)- N Engl J Med 236:407-412, 1947 3. Mark EJ, Yokoi T: Absence of evidence for a significant background incidence of malignant mesothelioma apart from asbestos exposure. Ann NY Acad Sri 643:196-204, 1991 4- Testa JR, Pass HI, Carbone M: Molecular Biology of Mesothelioma, in De Vita V, Heilman S, Rosenberg S, Wil liams & Wilkins (eds): Principles and Practice of Oncology (ed6). Philadelphia, PA, Lippincott, 2D01, pp 1937-1943 16 5. Dcvesa SS, Granman DJ, Blot WJ, et al: Atlas of Cancer Mortality in the US. N1H Publication No. 99-4564- Bethesda, MD, National Institutes of Health, 1999, pp 174-183 6. Robledo R, Mossmann B; Cellular and molecular mech anisms of asbestos induced fibrosis. ] Cell Physiol 180:158-166, 1999 7. Mosstnan BT, Churg A: Mechanisms in die pathogenesis of asbestosis and silicosis Am J Respir Ctit Care Med 157:1666 1680, 199B B. Mossman BT, Bignon M, Com M, et al: Asbestos: Sci entific developments and implication for public policy- Science 247:294-301, 1990 9. Nicholson Wj: Comparative dose response relationship of asbestos fiber types: Magnitudes and uncertainties. Ann NY Acad Sci 643:74-84, 1991 10. Sluis-Kremer GK: Asbestos disease at low exposures after long residence times. Ann NY Acad Scl 643:182-193, 1991 11. Muscat JE. Wynder EL; Cigarette smoking, asbestos ex posure, and malignant mesothelioma. Cancer Res 52:2263 2267, 1991 12. Boutin C, Dumoitier P, Rey F, et al: Black spots con centrate asbestos fibers in the parietal pleura. Am J Resp Crit Care Med 153:444-449, 1996 13. Tanaka S, Choe N, Hemenway DR, et al: Asbestos inhalation induces reactive nitrogen species and nyttcsinc for mation in the lungs and pleura of the tat. J Clin Invest 102:445-454. 1998 14. Choe N, Tanaka S, Kagan E: Asbestos fibers and inter leukin-1 upregulate the formation of reactive nitrogen species ifi rat pleural mesochettal cells. Am J Respir Cell Mol Bio! 19:226-236, 1998 . ' 15. Hesterberg TW, Butterick CJ, Oshimura M, et al: Role of Phagocytosis in Syrian hamster cell transformation and cytogenetic effects induced by asbestos and short and long glass fibers Cancer Res 46:5795-5802, 1986 16. Bocchetta M, Di Reset 1, Powers A. et al: Human mesothelial cells are unusually susceptible to SV40 mediated transformation and asbestos eo-carcinogenicity. Proc Natl Acad Sci USA 97:10214-10219, 2000 17- Okayasu R, Takahashi 5, Yamada S, et at: Asbestos and DNA double strand breaks. Cancer Res 59:298-300, 1999 18- Zhao YL. Piao CQ, Wu LJ, et al: Differentially expressed genes in asbestos-induced tumorigenic human bronchial epi thelial cells: implication for mechanism. Carcinogenesis 21: 2005-2010, 2000 19. Rosenthal GJ> Simeonova P, Cnraini E; Asbestos toxic ity: An immunologic perspective. Rev Environ Health 14:11 19, 1999 20. Choe N, Tanaka S, Xla 97, et al; Pleural macrophage recruitment and activation in asbestos induced pleural injury. Environ Health Petspect 105:1257-1260, 1997 21. Carbone M, Fisher S, Poweis A, er ak New molecular and epidemiological issues in mesothelioma; Role of SV40. J Cell Physiol 180:167-172, 1999 22. Burel JE, Lednicky JA: Cell and molecular biology of Simian virus 40: Implications for human infections and disease. J Natl Cancer Jnst 91:119-133, J999 23. Alt SH, De Caprio JA: Cellular transformation by SV40 large T antigen; Interactions with host proteins. Semin Cancer Biol 11:15-22, 200J CARBONE. KRATZKE. AND TESTA 24. Pipes JM, Levine AJ: Role of T antigen interactions with p53 in tumorigenesis. Semin Cancer Biol 11:23-30, 2001 . 25. Testa JR, Giordano A: SV40 and cell cycle perturba tions in malignant mesothelioma. Semin Cancer Biol 11:31-38, 2001 26.. Rundell K, Parakari Rr The rote of the SV40 small t antigen in cell growth promotion and transformation. Semin Cancer Biol 11:5-13, 2001 27. Kops SP: Oral polio vaccine and human cancer: A reassessment of SV40 as a contaminant based upon legal doc uments. Anticancer Res 20:4745-4750, 2000 28. Rino P, Di Resta I, Powers A, et al: Unique strains of SV40 in commercial poliovaccines from 1955 not readily iden tifiable with current testing for SV40 infection. Cancer Res 59:6103-6108,1999 29. Kane AB: Animal models of malignant mesothelioma, in Rom WN (ed): Environmental and Occupational Medicine (ed 3). Philadelphia, PA, Lippincott-Raven, 1998, pp 377-386 30. Smith WE, Hubert DD: The intrapleural route as a means of estimating carcinogenicity, in Karbe E, Parke JF (eds): Experimental Lung Cancer Carcinogenesis and Biostatistics. New York, NY, Springer Verlag, 1974, pp 92-101 31. McConnell EE, Catbone M: A comparison of pleural mesotheliomas induced by asbestos and SV40 virus in Syrian golden hamsters. Inbal Toxicol 12:173-181, 2000 (suppl 3) 32. Wagner JC, Berry G, Skidmore JW, et al: The effects of the inhalation of asbestos in rats- Br J Cancer 29:252-269,1974 33. Cicala C, Pompetti F, Carbone M: SV40 induces me sotheliomas in hamsters. Am J Pathol 142:1524-1533, 1993 34. Walker AM, Loughlin JE, Fricdlander ER, et al: Projec tion of asbestos related disease 1980-2009. J Occup Med 25; 409-425, 1983 35. Lanphear BP, Buncher CR: Latent period for malignant mesothelioma of occupational origin. J Occup Med 34:718 721, 1992 ' 36. Hunchatek M: Genetic factors in the aetiology of ma lignant mesothelioma. EurJ Cancer 31A:1741-l747, 1995 37. Vanna NJ, Polan AK: Non-occupational exposure to asbestos and malignant mesothelioma in females- Lancet 1:1061-1063.1978 38. Hammar SP, Bqckus D, Remington F, et al: Familial mesothelioma: A report of two families. Hum Pathol 20:107 112, 1989 39. Risberg B.NiekelsJ, Wagermark J: Familial clustering of mesothelioma. Cancer 45:2422-2427, 1980 40- Ascoli V, Scaho CC, Bruno C, er al: Familial pleural malignant mesothelioma: Clustering In three sisters and one cousin. Cancer Lett 130:203-207, 1998 41. Carbone M, Pass HI, Risro P, et al; Simian vims 40 like DNA sequences in human pleural mesothelioma. Oncogene 9:1781-1790, 1994 42. Testa JR, Catbone M, Hirvonen A, et al: A multi- institutional study confirms the presence and expression of simian virus 40 in human malignant mesotheliomas. Cancer Res 58:4505-4509, 1998 43. Jasani B, Cristaudo A, Emri SA, et al; Association of SV40 with human tumors. Semin Cancer Biol 11:43-61, 2001 44. Klein G. Powers A, Croce CM: Association of SV40 with human tumors. Oncogene (in press) 45. Hirvonen A, Mattson K, Karjalainen A, et al: SV40-like DNA sequences not detectable in Finnish mesothelioma pa- THE PATHOGENESIS OF MESOTHELIOMA 17 cienrs tioc exposed to SV40 contaminated poliovaccines. Mol Carcindg 26:93-99, 1999 46. Huang H, Reis R, Yonekawa Y, et al: Identification in human brain tumors of DMA sequences specific for SV40 large T antigen. Brain Pathol 9:33-44, 1999 47. Schivapurkar N, Wiethege T, Wistuba IL, et al: Presence of SV40 in. malignant mesothelioma and mesothelial cell ptolifeiations- 3 Cell Biochem 76:181-188, 1999 48. Ramael M, Nagels j, Hcylen H, et al; Detection of SV40 like vital DNA and viral antigens in malignant pleural me sothelioma. Eur RespiT J 14:1381-1386, 1999 49. Carbone M, Ri;io P, Grimly PM, et al: Simian virus 40 large T antigen binds p53 in human mesotheliomas. Nature Med 8:908-912, 1999 50. Pilatte Y, Vivo C, Renter A, et al: Absence of SV40 large T-onrigen expression in human mesothelioma ceil linesAm J Respir Cell Mol Biol 23:788-793, 2000 51- Xu L, Flynn BJ, Ungar S, et al; Asbestos Induction of extended life span in normal human mesothelial cells: Imerindrvjdual susceptibility and SV40 T-antigen. Carcinogenesis 20:773-783, 1999 52. Waheed 1, Guo ZS, Chen GA, et al: Antisense to SV40 early gene region induces growth arrest and apoptosis in Tantigen positive human mesothelioma cells- Cancer Res 59: 6068-6073, 1999 53. Toyooka S, Pass HI, Shivapurkar N, et al: Aberrant methylation and simian virus 40 sequences In malignant me sothelioma. Cancer Res 61:5727-5730, 2001 54. Cacioul F. tibener R, Berta, et al: SV40 replication in human mesothelial cells induces HGF/MET receptor activa tion: A model for viral-related mesothelioma. Proc Nad Acad Sci USA 98:12032-12037, 2001 55. Roggti VL, Sanfilippo F, Shelburne JD: Mesothelioma, in Roggli VL, Greenberg SD, Pratt PC (eds): Pathology of Asbestos Associated Diseases. Boston, MA, little Brown, 1992, pp 109-164 56. Austin MB, Fechnet RE, Roggli VL: Pleural malignant mesothelioma following Wilm's tumor. Am ] Clin Pathol 86: 227-230, 1986 57. Sanders Cl, Jackson TA: Induction of mesotheliomas and sarcomas from hot spots of 2WPuGj activity. Healch Phys 22:755-759, 1972 58. Baris Yl, Sahin A, Osestni M, et ah An outbreak of pleural mesothelioma and chronic fibrosing pleurisy in the village of Kaiain/Uigup in Anatolia- Thorax 33:181-192, 1978 59- Rnhl AN, Langcr AM, Moncure G, et al: Endemic pleutal disease associated with exposure to mixed fibrous dust in Turkey- Science 216:518-520, 1982 60, Coplu L. Dumotiier P, Demit AU, et al: An epidemio. logical study In an Anatolyan village in Turkey environmen tally exposed to tremolite asbestos. J Environ Pathol 15:177- ' 182, 1982 _ 61. Baris Yl: Asbestos and Erionite Related Chest Diseases. Ankara, Turkey, Semih Ofset Matbaacilik, 1987, pp 3-169 62. Baris B, Demir AU, Shehu V, et ah Environmental fibrous zeolite (erionite) exposure and malignant tumors other than mesothelioma. J Environ Pathol Toxicol Oncol 15:183 189, 1996 63. Roushdv-Hammady ], Siegel ], Emri S, et al: Genetic- susceptibility factor and malignant mesothelioma in the Cap padocian region of Turkey, Lancet 357:444-445, 2001 ' 64. Murthy SA, Testa JR: Asbestos, chromosomal deletions, and tumor suppressor gene alterations in human malignant mesothelioma. J Cell Physiol 180:150-157, 1999 65. Krattke RA, Otterson GA, Lincoln CE, et al: Immuno- hismchemical analysis of the pl61NK4 Cyclin dependent kinase inhibitor in malignant mesothelioma. J Natl Cancer Inst 87: 1870-1875, 1995 66. Friiellc SP. Grim J, Zhou J, et al: Re-expression of pl6LNK4s in mesothelioma cells results in cell cycle attest, eel) death, tumot suppression and tumor regression- Oncogene 16: 3087-3095, 1998 67. Yang CT, You I, Yeh CC, et al: Adenovirus-mediated pl4 (ARF) gene transfer in human mesothelioma cells, j Natl Cancer lnsr 92:636-641, 2000 68- Fritelie 5, Rubins JB, Zhou J, er al: Gene therapy of established mesothelioma xenografts with recombinant p!61NK4a adenovirus. Gene Ther 7:1421-1425, 2000 69- Xiao S, Li D, Vijg J, ct al: Codelcrion of pl5 and pl6 in primary malignant mesothelioma. Oncogene 11:511-515,1995 70. Cote RJ, Jhanwar SC, Novick S, et al: genetic alter ations of the p53 gene are a feature of malignant mesothelio mas. Cancer Res 51: 5411-5416,1991 (erratum 51:6399,1991) 71. MetcalfRA, Welsh JA, Benner WP, et at: p53 and K-ias mutations in human mesothelioma cell lines- Cancer Res 52: 2610-2615, 1992 . 72. Pass HI, Mew DJ: In vitro and in vivo studies of me sothelioma. J Cell Biochem 24-.142-151, 1996 (suppl) 73. Bianchi AB, Mitsunaga SI, Cheng JQ, et ai: High fre quency of inactivating mutations in the neurofibromatosis type 2 gene (NF2) In primary malignant mesotheliomas- Proc Natl Acad Sci USA 92.10854-10858, 1995 74. Foddis R, Derienzo A, Broccoli D: SV40 infection in duces telomerase activity in human mesothelial cells. Onco gene (in press) 75. Roggli VL, Pratt PC, Brody AR: Asbestos concent of lung tissue in asbestos associated diseases: A study of 110 cases. Br j Ind Med 43:18-28, 1986 76- Gibbs AR, Pooley FD: Analysis and interpretation of inorganic mineral particles in lung tissues. Thorax 31327-334, 1996