Document DM6qXkX6Jvo7gdKVGjNVY2OQ5

Malignant Mesothelioma With Occupational and Environmental Asbestos Exposure in an Illinois Community Hospital Karen M. Wolf, MD; Zdzislaw H. Piotrowski, MS; John D. Engel, PhD; Leonas G. Bekeris, MD; Enrique Palacios, MD; Kenneth A. Fisher, MD Clinical, radiologic, pathologic, and epidemiologic data on 32 patients with diffuse malignant mesothelioma (DMM) diagnosed between 1968 and 1984 at a 427-bed community hospital In Berwyn, III, were reviewed. Independent patholo gists' review of light microscopy, supported by electron mi croscopy, Immunoperoxldase staining, or autopsy, confirmed 29 pleural and three peritoneal DMMs. Clinical and radiologic characteristics were similar to those In published case series. Median age at diagnosis was 67 years, and median survival after diagnosis, seven months. Fourteen patients were women. Exposure histories were obtained through 22 interviews sup plemented by hospital charts and death certificates. Thirty patients (94%) had a history of asbestos exposure through work (15 [47%]) and/or residence near an asbestos facility (27 [34%]). Medical records and death certificates underreported asbestos exposure and DMM. (Arch Intern Med 1987;147:2145-2149) Oince the establishment of the asbestos-malignant mes^ othelioma relationship by Wagner et al in I960,' the majority of reported mesothelioma cases and associated medical data have come from geographic areas of known asbestos exposure. These regions, within Europe, South Africa, Great Britain, Canada, and the United States, include shipbuilding, asbestos mining, and manufacturing sites. The midwestern United States is one locale assumed to have had minor involvement in the asbestos-using industries2 and, therefore, to be at low risk for asbestosrelated diseases. In support of this assumption, only 255 patients with diffuse malignant mesothelioma (DMM) have been reported from the Midwest.*-l> Of these, only 16% were documented to have asbestos exposure (Table 1). Although the majority of the reports were not intended to be epidemiologic studies, these relatively unimpressive data significantly contribute to the perception that the Midwest is a minimally asbestos-affected area. More im portantly, these studies provided little basis to conduct etiologic investigations. From 1968 through 1984, 32 patients with DMM were diagnosed at MacNeal Hospital in Berwyn, 111, a 427-bed community hospital serving approximately 289 000 resi dents of western Cook County. This number of cases of DMM was unexpected not only because of the hospital's location in a low-risk area, but also because MacNeal Hospital is a primary care, not a referral, center. A retrospective study was undertaken, therefore, to (1) confirm the diagnoses with independent pathologists' review and updated pathologic technology, ie, electron microscopy and immunoperoxidase methods, (2) clinically characterize the patients, and (3) investigate asbestos exposure history. Substantively, this study differed from prior midwestern reports by its use of multiple epidemio logic data sources to assess asbestos exposure. Detailed occupational and complete residential histories were com- Accepted for publication Aug 12, 1987. From the Department of Medicine, University of Illinois College of Medicine at Chicago (Dr Wolf); Division of Medical Education, Department of Pathology, and Department of Radiology, MacNeal Hospital, Berwyn, 111 (Mr Piotrowski and Drs Engel, Bekeris, and Palacios); and Department of Medicine, University of the Health Sciences, Chicago Medical School, North Chicago, 111 (Dr Fisher), Reprint requests to Division of Medical Education, MacNeal Hospital, 3249 S Oak Park Ave, Berwyn, IL 60402 (Mr Piotrowski). Arch Intern Med--Vol 147, Dec 1987 piled through next-of-kin or patient interviews for the majority of cases. History of asbestos exposure from hobbies or household contact to an asbestos-exposed worker, or from residential proximity to an asbestos product manufacturing plant, was obtained. In addition, a historical review of the nearby communities was conducted to identify asbestos product manufacturing facilities. PATIENTS AND METHODS Case Ascertainment With Pathology Review The presentation of two patients with DMM within a ninemonth period at the MacNeal Hospital weekly medical mortality conference formed the basis to begin a case ascertainment phase. To do so, hospital medical records, tumor registry files, and radiologic and pathologic files were reviewed for all suspected primary pleural or peritoneal neoplasms or pneumoconioses di agnosed at MacNeal Hospital from 1968 through 1984. Fifty-seven potential cases were found in this preliminary search. Nineteen cases were excluded for one of the following reasons: (1) pneumo coniosis was present but there were no signs of malignant disease, (2) the initial clinical suspicion of DMM was not supported by the subsequent medical evaluation, or (3) sufficient pathologic material was lacking. Of the 38 cases identified by the ascertainment phase, one was considered verified by its inclusion in the McDonald and McDonald study.* Histologic materials from the remaining 37 were retrieved and the majority processed for electron microscopy and/or immu noperoxidase staining for carcinoembryonic antigen (CEA) and cytokeratin (KER). These specimens and the original pathology slides were analyzed by one of us (L.G.B.). Two of the patients were accepted without further review because of light microscopic and necropsy characteristics consistent with DMM.11 The light microscopic specimens from the remaining 35 patients were then presented in a blinded fashion to two independent pathologist reviewers. The reviewers were asked to decide if any of the histologic materials demonstrated DMM. They were given the patient's age and sex but no further medical or exposure data. Cases reviewed in this manner were considered pathologically "confirmed" if they met one of the following acceptance criteria: (1) a cumulative score of 4 or less from-the three pathologists' quantification oflight microscopy findings using a scale in which 1 indicates definite; 2, probable; 3, possible; and 4, unlikely (note that the two independent pathologist reviewers had access to only the light microscopic materials), or () a three-reviewer summation score of 8 or less in addition to positive KER and negative CEA staining or characteristic electron microscopic features.'*-'T Six of the 35 cases reviewed were rejected. These 29 reviewed cases, the patient previously described by McDonald and McDonald,* and the two accepted by light microscopy and autopsy criteria com prised the 32 confirmed patients in this case series. Table 2 presents the pathologic review results for each patient. Clinical and Ropntgpnographlc Rvl#w For each confirmed case, presenting symptoms and physical findings were recorded from examinations performed during the first admission in which a pleural or peritoneal neoplasm was considered. Chest roentgenogram data were obtained for all confirmed cases from the first chest roentgenogram report from that same admission (anteroposterior, posteroanterior, with or without left lateral). A blinded reading was conducted by radiol ogists who were not "B-readers" (ie, radiologists not certified in pneumoconiosis radiography as established by the American College of Radiology). < Exposure Review History of occupational asbestos exposure, from age 16 years and up to five years before death, was obtained from personal Malignant Mesothelioma--Wolf et al 2145 20 A -j.-h4; -Table 1.--Summary Data Prom Previously Published Midwestern United States Malignant Mesothelioma Studies .F-i!..,,, , Source, y Godwin,1 1957 Godwin and Jagatic.4 1970 Oels at al,* 1971 Grundy and Miller.* 1972 Hinds,7 1978 McDonald and McDonald,* 1980 Cunningham et al.1 1930 Vogelzang at al," 1984 Adams and Unni,'1 1904 DeKraay," 1985 Lilienlald and Gunderson," 1986 Source of Cate Ascertainment Hospital medical charts Hospital medical charts Hospital autopsy files National death certificate Hies . Tumor registries in Iowa and Detroit National survey of US and Canadian pathologists Tumor registries in Illinois Hospital medical charts No. of Patient! 14 7 27/37* 3 % 76 36 28 28 Hospital autopsy files Hospital medical charts State death certificates 16 12 8 No. With Definite or Probable Asbestos History 6 (86%)t 10 (27%) t * 15 (54%) 10 (62%) 1/21 Source of Data to Examine Aebesloe Exposure Hospital medical charts Hospital medical charts Hospital medical charts Hcspital medical charts Tumor registries Hospital medical charts and nextof-kin Interviews Tumor registries Hospital medical charts and physician interviews Hospital medical charts Hospital medical charts Death certificates This information was not reported or was not available in the reported study tGodwm and Jagatic reported on six patients with "asbestos bodies ' present. Oels et al reported on ten patients with "asbestos bodies." tor 37 patients presented. 27 were from the Midwest. Asbestos exoosure data were not grouped by geographic region. Frequencies were reported by region. Occupational exoosure data were reported for United Slates and Canada comomed. |{Only two representative case histories were described, one ol which had asbestos exposure. interview when possible. In 21 cases (66%), interviews were conducted with a surviving next of kin and, in one instance, with the patient. Five interviews were refused, and relatives could not be located for five cases (31%). The interviews were conducted using a standardized question naire adapted by the National Cancer Institute from the research by McDonald and McDonald* and Selikoff.1* All job titles, work tasks, and known work-related exposures were ascertained. Ad ditional quprips were made regarding residential histories since birth, hobbies that included asbestos materials, and household members' asbestos-related employment. The location, period of operation, and description of asbestos product manufacturing facilities mentioned during the interviews were confirmed using multiple historical resources. These included Illinois manufactur ers' directories from that period, local newspaper accounts, city directories, and locally published reports.1' For all patients, a premortem histologic diagnosis of DMM had been made. The presence or absence of documentation of meso thelioma as an underlying or contributing cause of death on the death certificate was noted. Any indication of asbestos exposure history in the medical record and the usual occupation entered on the death certificate were recorded. The date of death was confirmed from the death certificate. RESULTS Clinical Findings TWenty-nine patients had primary pleural and three had primary peritoneal involvement The age at diagnosis ranged from 19 to 86 years, with a median of 67 years. Of the 32 patients, 14 (44%) were women. In those with pleural primary tumors, the most common presenting symptoms were dyspnea, chest pain, and cough. The duration of symptoms until the time of presentation varied from one 2146 Arch Intern Med--Vof 147, Dec 1987 week to three years. Of the three peritoneal mesotheliomas, two were found at laparotomy as incidental findings. Presenting chest roentgenogram interpretations docu mented pleural effusions in 27 cases (84%), pleural masses in five, pleural plaques in three, interstitial pulmonary fibrosis in two, bone involvement in two, and pleural thickening in one. Thoracenteses (38 performed) and pleu ral biopsies (17 performed) were nondiagnostic for the majority (807&). Percutaneous lung biopay was positive in ope of the two performed. Excisional biopsies of chest-wall masses (three performed), thoracotomies (25 performed), and laparotomies (three performed) were 100% diagnostic. Data regarding radiation and chemotherapy treatments were insufficient for analysis because of outside referral for the former and infrequent utilization and nonstandardized protocols for the latter. Length of survival from time of diagnosis ranged from one week to 31 months, with a median of seven months. Pathologic Findings Of the 29 patients whose histologic materials underwent light microscopic review (Table 2), 22 (76%) were judged to have definite or probable DMM by all three reviewers using light microscopic criteria alone. The results of electron microscopy and immunoperoxidase staining for CEA and KEIt were as follows. Of the 22 cases rated by the three reviewers as definite or probable by light microscopic analysis, electron microscopy was performed in 16. The electron microscopic findings were supportive of DMM in 12 of the 16 and inconclusive in four. Malignant Mesothelioma--Wolf et al v^ Table 2.--Characteristics and pathology Evidence of yeHflcation.ofj3asaV;;i:^iV^<i5il5^5S'?>i^T^^i;*. '* Patient/ Sex 1/M 2/M 3/M 4/M 5/M - 6/M 7/M 8/M 9/M Survive), mo Alter Diagnosis 13 12 14 19 21 10 11 8 13 Time of Diagnosis, Age, y/Year 68/1971 62/1972 64/1975 58/1976 69/1977 44/1970 67/1978 19/1980 63/1981 Light Microscopy Reviewer* A 1 1 1 1 1 1 2 1a 1 B 1 1 1 2 1 2 1 1 C 1 1 1 2 1 4 2 1' Electron Microscopy Supportive Supportive Supportive * Supportive Supportive Immunoperoxidase f KER CEA ... Positive Positive Nogatlvs Negative Positive Positive Negative Negative Positive Negative Autopsy Yes 10/M 7 65/1981 1 1 1 Supportive . 11/M 12/M 13/M 14/M 15/M 16/M 17/M$ 18/M$ 1/F 2/F 3/F 4/F S/F 6/F 7/F 8/Ft 9/F 10/F 11/F 12/F 13/F 14/F 12 9 2 3 2 10 4 3 5 6 18 4 30 1 26 26 9 7 5 0.25 5 12 70/1982 69/1983 70/1983 74/1983 70/1983 54/1983 51/1984 67/1984 55/1968 74/1969 62/1971 67/1975 49/1975 70/1977 65/1978 78/1980 6BM9B1 66/1981 79/1982 69/1982 69/1983 86/1984 1 2 1 Supportive Positive Negative 1 1 1 Supportive Positive Negative z 3 1 Inconclusive Positive Negative ... 1 2 1 Inconclusive Negative Negative 11 Supportive Positive Negative 1 2 2 Supportive 111 Positive Negative Yee 11 Positive Negative - Yts 1 12 1 Yes Yea ...I ...1 ...1 1 3 3 Supportive Positive Negative 12 Inconclusive Negative Negative 11 Inconclusive Positive Negative 223 Positive Negative 11 1 3 2 Supportive 11 Inconclusive Positive Negative 1 2 1 Supportive Positive Negative 1 32 Positive Negative 1 3 3 Supportive Positive Negative 11 Supportive Positive Negative Pathologist Quantification of the likelihood ol malignant mesothelioma based on light microscopy: 1 Indicates definite; 2, probable; 3, possible; and 4, unlikely. Only reviewer A (l.G.B.) had access to electron microscopy immunoperoxidase, and autopsy materials . +KER indicates cytokeratm; CEA, carcinoembryonic antigen i ^Peritoneal primary. ~ Autopsy completed after the pathology review. UTissue obtained by pleural biopsy and thoracotomy was reviewed and accepted into the McDonald and McDonald study* Immunoperoxidase staining for KER and CEA was per formed in 15 of these 22 cases. The staining was positive for KER in 13 and negative for CEA in all 15, findings consistent with DMM. Seven of the 29 reviewed cases were given a "less than certain" rating, that is, one or more of the three reviewers had judged the specimen to be possible or unlikely DMM (a three-reviewer summation score s6). For these the diagnosis of DMM was established by the demonstration of characteristic immunoperoxidase stain ing in four, both electron microscopy and immunoperoxi dase criteria in two, and electron microscopy criteria in one. Epidemiologic Findings The yearly rate and trend of DMM at MacNeal Hospital was obtained by dividing the number of patients newly diagnosed with DMM by the total number of admissions for each year (approximately 15000/y) from 1968 to 1984. This yielded an average yearly rate of 12 new patients with DMM per 100 000 MacNeal hospital admissions (an average Arrh Intern Med--Vol 147, Dec 1907 of two cases per year). The differential rate was four per 100000 admissions from 1968 to 1974, ten per 100000 admissions from 1975 to 1979, and 25 per 100 000 admissions from 1980 through 1984. From historical records, two large primary asbestos product manufacturing plants were identified within 4.S km of MacNeal HospitaL The first company, which began operations in 1920 and employed approximately 1200 people during 1949, manufactured gaskets, a specialized appliance of metal covered with asbestos, and asbestos millboard. Chrysotile asbestos was used. This manufac turer was housed in multistory buildings occupying ap proximately l'/2 square blocks. The second site, which began operation in the 1930s and closed in the early 1950s, employed about 200 workers at one time in a one-story plant with an area of 14 400 m* (160 000 sq ft). This company manufactured insulation from chrysotile asbestos. Three smaller asbestos product companies with a combined work force of up to 300 were identified. Whether or not the smaller sites were asbestos factories or asbestos product Malignant Mesothelioma--Wolf et al 2147 3.--Type of Occupationaland i^^irf^j.'Environmental Asbestos Exposure!^^^^^^: No. (%) Exposure ^.Occupational . .... Asbestos product manufacturing in neighbomood facility (Intervisw data) Men 4/13 (30) Women Combined , -; ;. _ 3/9 (33) 7/22 (32) Probable or definite work- related asbestos exposure (interview and chart dak) 8/18 (44) Total 12/18 (67) 0/14 (0) 8/32 (25) 3/14 (21) 15/3% (47) ^Environmental Household exposure through family member with occupational exposure (interview data) Hobby with exposure to asbestos (interview data) 3/13 (23) 3/13 (23) 3/9 (33) 6/22 (27) 1/9 (11) 4/22 (18) Residence <3.2 km from neighborhood asbestos product manulactunng facility (interview and chart data) Total 15/18 (83) 16/18 (89) 12/14 (86) 13/14 (93) 27/32 (84) 29/32 (91 )t 'Percentages are based on different numDers of patients: interviews. 13 men and nine women; and interviews plus cnart data. 18 men and 14 women. In addition, some patients had more than one route of exposure, so entries in 'Total" rows may not De sums. tOf the three patients who had no history of environmental exposure, one had a history of work exposure to asOestos (patient 17/M). Thus 30 (94%) of 32 patients had occupational or environmental asOestos exposure. user factories could not be determined from available historical records. Women made up approximately 40% of the total work force of the five companies.1' A review of census of manufacturing data in the late 1940s and early 1950a" showed that Illinois was ranked first or second among the contiguous United States in the number of facilities or the number of production employees in asbestos product manufacturing. The asbestos-related industries in Illinois manufactured asbestos products such as textiles, building materials, floor tile, insulation, cement, and gaskets. Raw asbestos was shipped into the Chicago area from both Canada and South Africa, the former primarily producing chrysutile, and the latter, crucidolite asbestos.U,J1 Based on interview and chart data, 30 (94%) of these 32 patients had some history of asbestos exposure through one or more of the following: occupation, household contact, or residential proximity (Table 3). (A detailed breakdown of asbestos exposure by patient is available from Z.H.P.). Twenty-seven (84%) resided within a 3.2-km radius of one of the two major neighborhood asbestos product manufac turing facilities at some time five or more years before diagnosis. Fifteen (47%) had definite or probable work exposure to asbestos (three women, 12 men). Seven of these 15 (four men, three women) had worked in one of the two large neighborhood asbestos facilities. Six had house hold exposure to asbestos through a relative in the home with definite or probable work exposure, and four had exposure through hobbies (eg, insulating their own homes with asbestos-impregnated siding). Only two of the 32 patients had no documented evidence of asbestos exposure of any type. Review of the hospital charts documented probable or definite occupational asbestos exposure for only nine pa tients. No reference was made to residential proximity to an asbestos plant as a risk factor in the recorded medical histories. With personal interviews, six additional patients with definite or probable work asbestos exposure were identified. The death certificates provided information on possible asbestos exposure for ten men and one woman. Nine (28%) of the death certificates failed to cite mesothe lioma as an underlying or contributing cause of death despite premortem tissue diagnosis. COMMENT The midwestern United States is one of many regions assumed to be at low risk for asbestos-related diseases. For the Midwest, this belief may be due in part to the presumption that few or insignificant asbestos-using in dustries existed in the area.* The infrequent reporting of DMM, as summarized in Table 1, may also have contributed to this perception. Hospital medical chart*-4-11-11 and death certificate1-11 data are not reliable for asbestos exposure histories or the detection of DMM. These points are well illustrated by our study's finding that although all but two of 32 patients with DMM were found to have had asbestos exposure, only nine of the medical records mentioned such exposure. Death certificates provided evidence for asbes tos exposure in only 11 instances and underreported the diagnosis of DMM by 28%. Of more importance was the historical review of the community that ensued once the high asbestos exposure rate was determined. A common asbestos exposure source hitherto not suspected was found. Two major asbestos product manufacturing sites in close proximity to MacNeal Hospital were identified. Eighty-four percent of the pa tients had lived within a 3.2-km radius of these plants. We were able to document that 22% had worked in them. An additional eight patients had other definite or probable work-related asbestos exposure, bringing the total with direct occupational exposure to 47% of the entire group. We found that 31% of the patients had household exposure to asbestos through a relative in the household with definite or probable work exposure or through a hobby. Overall, 94% of the patients had documented asbestos exposure. Only two patients had no evidence of asbestos exposure of any type. Exposure to crocidolite asbestos i9 most frequently associated with the development of DMM. "That chrysotile was the predominant asbestos fiber used in the neighbor hood asbestos plants11 does not contradict the theory that these facilities were a common source for the risk of developing DMM. Churg et al" demonstrated only chrysotila ore components within the lungs of five miners and millers with pleural mesotheliomas.0 Furthermore, raw asbestos was shipped into this region from both Canada and South Africa, the latter a major source of crocidolite asbestos. The residential and working populations might have actually been exposed to a mixture of these fibers. Additionally, most published series of DMM have dem onstrated a male-female ratio of 3:1 or 4.I.*1-**-" In contrast, this study found an almost equal number of men and women with DMM (1.3:1). Only three of the 14 women had documented definite work exposure to asbestos. Four had household exposure to family members who worked with and/or had hobbies that involved the use of asbestos. Five women had no known asbestos exposure other than resi dential proximity to the asbestos plants. The observed lower rate of occupational asbestos exposure among women may be due to next-of-kin recall bias. The role of occupa tional and nonoccupational exposure among women war rants further epidemiologic study. ChurgM recently compared the lung asbestos burdens of chrysotile mining town residents with those of non-mining 2148 Arch Intern Med--Vol 147, Dec 1987 Malignant Mesothelioma--Wolf et al wjsvn rosUKsnts. The chryBotile concentration of the mining lung nnbcstoa body counts as high as 9200 asbeatos bodies 4lown residents (median, 1.2 million/R of dry lung) was per grain of wet weight of lung. Chest roentgenography jibout six limds that of the non-mining town residents. A performed in 120 spouses of asbestos workers in a screening licanning electron microscopy study of asbestos fiber con- program in a local pipefitters' union in Chicago showed i-mtrntions by Mowe ct al57 demonstrated that a critical that 17% had classic pleural plaques consistent with asbes ' as low as 1 million fibers per gram of dry lung tissue tos exposure.31 Malignant mesothelioma is a "sentinel v,,,o associated with an increased risk of DMM. Since this health event," the occurrence of which indicates excess concentration may be found in the general population," it disability and untimely death from diseases such as asbes- Idocs strengthen the argument for n "no-threshold" ri.sk for Uhe development of mesothelioma from minor asbestos (exposures. Furthermore, it increases the likelihood that | the "indirect" exposures experienced by these women, ie, I to household asbestos users or residential proximity to tosis, pleural effusions, ami lung, laryngeal, and gastroin testinal tract carcinomas." McDonald and McDonald" re viewed three epidemiologic studies and concluded that about 7% of lung cancer deaths in American men in 1985 resulted from occupational asbestos exposure. j asbestos industries, may be causally related to their devel The national incidence of pleural mesothelioma among opment of DMM. men is estimated to be from seven to 13 per million during It is likely that additional cases of DMM have gone the last ten years. This rate is increasing at a yearly unreported from western Cook County. The stringent average of about 13%. The incidence may be substantially I pathologic criteria used for this study may have led to the exclusion of several patients with DMM. Furthermore, other hospitals within the area serve the same population. A pathologic review should be extended to those hospitals. higher in select geographic areas where asbestos industries existed."34 Manufacturing census data have shown that Illinois along with other midwestern states had many asbestos product manufacturing factories between 1940 and 1960." The MacNeal Hospital patients may represent That an even larger group of patients with asbestos , related disease exists in metropolitan Chicago is probable j for several reasons. Other asbestos product manufacturers a "sentinel community" for malignant and nonmalignant asbestos-related diseases in the Midwest. Other such unrecognized communities may exist worldwide. jhnve been identified within the metropolitan Chicago 5 area.1' Data that support significant asbestos exposure jfrom these manufacturers include the demonstration by SSuta and Levine5' that the Chicago area has an average j atmospheric asbestos concentration of 24 ng/m', a level J comparable with that of American coastal cities such as We are indebted to Robert Spirtas, DrPH, Occupational Studies Section, National Cancer Institute, Bethcsda, Md, and Nicholas J. Vogelzang, MD, Section of Homatology/Oncology. University nf Chicago, for their sugges tions and encouragement. We also thank Jerome Taxi, MO, Lutheran General Hospital, Park Ridge, III; Thomas Victor, MD, Evanston (III) Hospital; and Gregorio Chejfek, MD, Veterans Administration, Hines Hospital, Maywood, III, for their assistance with the pathology studies. In iLos Angeles and New York. Churg and Warnock,'0 in a tcase-control study of Chicago patients with nonmesothe addition, we are especially grateful to Donna Gammers and Raymond Hughes for their patience and support in preparing this manuscript lioma pulmonary and gastrointestinal tract cancers, fouRnedferences 4 Vagner JC, Sleggs CA, Marchand P: Diffuse pleural mesothelioma , s jestoa exposure in the North Western Cape Province. Br J IndMtd ^ 19t>v;17:200-271. J 2. Antman KH: Current concepts: Malignant mesothelioma. N Engl J [Hed 1980;303:200-202. I 3. Godwin MC: Diffuse mesotheliomas with comment on their relation to '(localized fibrous mesotheliomas. Cancer l95f;]U:zyM-:)19. I 4 Godwin MC, Jagatic J: Asbestos and mesotheliomas. Environ Rea i 1970;3:391-11G. i 5. Ocls HC. Harrison EG, Carr DT, etal: Diffuse malignant mesothelioma * of the pleura: A review of 37 cases. Cheat 1971;G0:5G4-570. 6. Grundy GW, Miller RW: Malignant mesothelioma in childhood: Report J of 13 cases. Cancer 1972;30:121G-1218. j 7. Hinds MW; Mesothelioma in the United States: Incidence in the 1970'*. JJOccnp Med 197R;20:4G9-471. 8. McDonald AD, McDonald JC: Malignant mesothelioma in North 1 America. Cancer 1980;4G:1G50-165G. J 9. Cunningham MP, Allen L, Apostol JV, et al: Lung cancer. A report by (j 47 Illinois hospitals on cases diagnosed from 1970-1974, in Reautta fit j TYcating Cancer, report 4. Chicago, American Cancer Society, Illinois j Division Inc, 1980, pp 1-19. | 10. Vogelzang NJ, Schultz SM, lannticci AM, et al: Malignant mesothe- J lioma: The University of Minnesota experience. Cancer 1984;53:377-383. i 11. Adnms VI, Unni KK: Diffuse malignant mesothelioma of pleura: ] Diagnostic criteria based on an autopsy study. Am J Clin Pathol 1984;82:15- 1 23 j 12. DeKrnay WI1: Malignant mesothelioma. WiaMedJ 1985;84:32-34. j 13. Lilionfcld DE, Gunderson PD: The 'missing cases' of pleural maligI nant mesothelioma in Minnesota, 1979-1981: Preliminary report. Public \ Health Rep 198G; 101:395-399. ( 14. Wang N-S: Electron microscopy in the diagnosia of pleural mesothe' liomas. Cancer 1973:31:1040-1054. i 15. Suzuki Y, Churpj .1, KnnncrRtcin M: Ultrastnicture of human mnltftj nant diffuse mesothelioma. Am J I'ntkul 197(i;85:241-262. 16. Warhol M.J, Corson JM: An ultrnstructurnl comparison of mesotheli* : ornaa with adenocarcinomas of the lung and breast. Hum Pathol ID85;16:50- l65- J 1 ** Corson JM, Pinkus GS: Mesothelioma: Profile of keratin proteins and Itf embryonic antigen: An immunoperoxidnse study of 1*0 cases and aCo. /ison with pulmonary ndenoearcinomas. Am J i'nthol 1982;lU8;8U-87. 18. Selikoff [J: Cancer risk of asbestos exposure, in Hiatt Mil, Watson ijJD, Winsten JA (eds): Ononis nf flnwnn Cancer Cold Spring Harbor, NY, |Cold Spring Harbor Laboratory, 1977, pp 17GG-1784. 19. Cramer RE: Mnuufuctnrt in; Structure nf the Cierro District, Met ropolitan Chicago, Department of Geography research paper 27. Chicago, University of Chicago Press, 1953. 20. Untied Stalea Cenaua of Manufacturera: 11)51: Location of Manufac turing Planta by Industry, County, and Employment Size. Washington, DC, US Bureau of Census, 1959. 21. Selikoff IJ, Lee DHK: Historical backgrounds, in Aahcatoa and Diaeaae. Orlando, Fla, Academic Press Inc, 1978, pp.3-32. 22. Walker A: Asbestos-related disease: The epidemiology of mesotheli oma. Cheat 1980;R0(suppl):3G8S. 23. Churg A, Wiggs H, DePaoli L, ct al: Lung asbestos content in chrysotile workers with mesothelioma. Am Rev Reapir Dia 1984; 130:1042 1045. 24. Becklake M R: Asbestos-related diseases of the lung and other organs: Their epidemiology and implications for clinical practice. Am Rev Reapir Dta 1970;114:187-227. 25. McDonald AD. McDonald .1C' Epidemiology of malignant meaotheli oma. In Antman K, Aisncr J (cds): Aabcatoa-Relaled Malignancy. New York, Grune & Stratton, 1987, pp 31-55. 26. Churg A: Lung asbestos content in long-term residents of a chrysotile mining town. Am Rev Reapir Dia 1986;134:125-127. 27. Mowe G, Gylsoth B, Ilartveit F, et al: Fiber concentration in lung tissue of patients with malignant mesothelioma: A case control study. Cancer 1985;6fi:1089-1093. 28. Churg A: Fiber counting and analysis in the diagnosis of asbestosrelated disease. Hum Pathol 1982;13:381-392. 20. Buta HE, l.cvinc BJ; Non-uccupallunal asbestos emissions and exposures, in Michaels L, Chissick SS (cds): Aabeatna: Propertiea, Applicutiona, and Uazarda. New York, John Wiley & Sons Inc, 1979, pp 171-205. 30. Churg AM, Warnock ML: Numbers of asbestos bodies in urban patients with lung cancer and gastrointestinal cancer and in matched controls. Cheat 1979;76:143-149. 31. Holland F.A, Sider L, Davis TM, et al: Radiographic changes in the spouses of occupationally exposed asbestos workers. Read before the Radiologic Society of North America Annual Meeting, Chicago, Dec I, 198G. 32. Kutstein UU, Mullan KJ, Frazier TM, et al: Sentinel health events (occupational): A hnsis for physician recognition and public health surveil lance. AmJ Public Health 1983;73:1054-1002. 33. McDonald JC, McDonald AD: Epidemiology of asbestos-related lung i cancer, in Antman K, Aisncr .I (eds): Aaheatna-Rcliited Mntignuncy. New York, Grune A Stratton, 1987, pp 57-79. 31. Spirtas R, lleebc GW, Connelly R|{, et al: Recent trends in mesothe lioma incidence in the United Slates. Am J Ind Mrd 19HG;9:.'I97-407. 5 Arch Intern Mod--Vol 147, Dec 1987 h Malignant Mosolholioma--Wolf el al 2149 P Ki u