212 Occupational and Environmental Medicine 1994;5l:212-215

Four cases of mesothelioma among Finnish anthophyllite miners Occup Environ Med: first published as 10.1136/oem.51.3.212 on 1 March 1994. Downloaded from

A Karjalainen, L 0 Meurman, E Pukkala

Abstract theless, often constituted a large proportion Four cases of mesothelioma in a cohort of the total fibre burden and could therefore of 999 Finnish anthophyllite miners and not be excluded as a cause of mesothelioma. millers are described. Three deaths were In one of the patients with a reasonably low due to pleural mesothelioma and one to concentration of fibres only anthophyllite was peritoneal mesothelioma among the total detected.5 of 503 male deaths up to 1991. All four In this report we describe four cases of patients with mesothelioma had had long mesothelioma that occurred in the ongoing term (13 to 31 years) exposure in antho- follow up of the cohort of Finnish anthophyl- phyllite mining and milling. The latency lite miners and millers. time from the onset of employment until diagnosis was 39 to 58 years. All four patients were smokers or ex-smokers and Materials and methods had asbestosis. In three of the cases the THE COHORT OF ANTHOPHYLLITE MINERS pulmonary fibre concentration and fibre Two anthophyllite quarries were type were analysed by transmission elec- worked in Finland during the period tron microscopy. High concentrations 1918-75. They were owned by the same (270 to 1100 million fibreslg dry tissue) of company and located within 10 km of each anthophyllite fibres were detected. The other in a sparsely populated area about 50 anthophyllite fibres were thicker and had km from the nearest city. The workers were lower aspect ratios than the values recruited mainly from the farming population reported for crocidolite fibres retained in living in this area. In the 1950s and 1960s the the lungs ofpatients with mesothelioma. annual output was about 10 000 tonnes. The quarries were combined with a dry hammer (Occup Environ Med 1994;51:212-215) mill, which crushed, ground, and classified the broken rock into asbestos products for various industrial applications. Numerous reports from several countries A fairly complete personnel register was have described cases or series of pleural and kept by the company in the form of salary http://oem.bmj.com/ peritoneal mesotheliomas in relation to occu- lists from 1936. The register contained the pational exposure to various types and mix- names, dates of birth, working periods, and tures of asbestos. In 1960 Wagner and roughly the kind of work done. The series of coworkers described 33 cases of pleural miners and millers included all workers mesotheliomas in the neighbourhood of a employed for three months or longer between crocidolite mine in South Africa.' In later 1 January, 1936 and 1 July, 1967 (miners,

studies, mesotheliomas have been found after millers, forestry workers, transportation work- on October 2, 2021 by guest. Protected copyright. occupational exposure to crocidolite, amosite, ers, and office workers). After the exclusion tremolitic material, and chrysotile asbestos.2 of workers who could not be identified, the In Finland, anthophyllite asbestos has been follow up was completed in 999 workers up exceptionally widely used because of its to the end of 1991. The deaths in the cohort domestic production from 1918-75, and spe- were researched through the death files of cial attention has been focused on the role of Statistics Finland. The deaths from cancer anthophyllite asbestos in the aetiology of were based on the files of the Finnish Cancer Institute of Occupational Health, mesothelioma. Anthophyllite fibres have Registry. Helsinki, Finland induced mesotheliomas in animal experi- A Karjalainen ments both after inhalation and intrapleural HISTOLOGICAL VERIFICATION OF THE Department of administration, but epidemiological studies DIAGNOSES Pathology, University ofTurku, Turku, have failed to firmly connect exposure to In all four cases of mesothelioma tissue sam- Finland anthophyllite and mesothelioma in humans. ples taken from the tumour at necropsy or L 0 Meurman In a cohort of about 1000 anthophyllite min- thoracotomy were available for histopatholog- Finnish Cancer ers and millers an increased risk of lung can- ical examination. The diagnoses of mesothe- Registry, Helsinki, cer was found, but no confirmed cases of lioma were confirmed the Finnish Finland by E Pukkala mesothelioma were reported up to 1977.34 In Mesothelioma Panel. Formalin-fixed and a more recent Request for reprints to: study of 29 Finnish patients paraffin embedded tissue blocks were used in Antti Karialainen, Institute with mesothelioma, raised concentrations of light microscopical histopathological, and of Occupational Health, Topeliuksenkatu 41 a A, anthophyllite fibres often occurred in their immunohistochemical examination. In one of SF-00250 Helsinki, Finland. lungs, but usually together with other types of the cases (case 2 in table 1) a morphologically Accepted 28 June 1993 asbestos fibres. Anthophyllite fibres, never- typical pleural mesothelioma with sarcoma- Four cases ofmesothelioma among Finnish anthophyllite miners 213

Table 1 Age, tumour type, exposure, and lungfibre concentration in thefour cases ofmesothelioma among Finnish anthophyllite miners Fibre concentration Age at Tumour Year of in lung tissue Fibre types Occup Environ Med: first published as 10.1136/oem.51.3.212 on 1 March 1994. Downloaded from Case No diagnosis Tumour type location diagnosis Years in mining (million fibresig) detected 1 82 Epithelial Peritoneal 1989 31, 1936-67 1100 Anthophyllite 2 75 Sarcomatous Pleural 1987 13, 1929-56 270 Anthophyllite 3 57 Sarcomatous Pleural 1986 28, 1947-75 1100 Anthophyllite 4 62 Epithelial Pleural 1991 23, 1950-75 -= No sample

tous histology remained negative in cytoker- were used. All inorganic particles having atin staining and a sarcoma could not be roughly parallel sides and a length to width completely excluded. ratio greater than three were defined as fibres and counted. The dimensions of the fibres ANALYSES BY ELECTRON MICROSCOPY were measured directly on the screen with In three cases a sample of lung tissue was magnifications up to 100 000. The dimen- available for analysis of the retained sions of 100 fibres per sample were measured. fibres by electron microscopy. In one of these An energy dispersive x ray microanalyser cases (case 2 in table 1) the tissue was stored (Tracor TN 5500) was used to identify the in 4% formalin. In the two other cases the mineral type of each fibre. The intensity specimens were embedded in paraffin, which ratios of Si, Mg, Fe, Ca, and Na were used in was subsequently dissolved with xylene. As the identification by comparing the spectra the tissue weight is known to decrease during peak ratios to standard spectra. Four hundred the process of embedding due to the extrac- fibres per sample were identified. This gave tion of lipids,6 the fibre concentrations of the an analytical sensitivity of <0-25% for fibre paraffin embedded samples were corrected types other than anthophyllite. According to assuming a dry weight loss of 20%. Poisson statistics this equals a detection limit A piece of tissue of about 100 mg wet of <1 0% (four times the analytical sensitiv- weight was taken for the fibre analysis. The ity). samples were prepared according to a previ- ously described method,5 which included low temperature ashing of the lung tissue and fil- Results tration of the remaining ash on polycarbonate In a recent follow up of 999 anthophyllite filters. The filters were further coated with miners and millers, a total of 593 deaths (503 carbon and placed on electron microscopy men and 90 women) were noted up to the grids. The filter material was dissolved with end of 1991; 72 of these were were due to chloroform in a modified Jaffe washer. The lung cancer, four to mesothelioma, and 31 to fibres were counted with a JEOL 100 CX- asbestosis. The diagnosis of asbestosis was ASID4D electron microscope in STEM mentioned in 20% of all death certificates. A

mode on a CRT screen. An acceleration volt- detailed report on the incidence of cancer in http://oem.bmj.com/ age of 100 kV and a magnification of 15 000 this cohort is under preparation. Table 1 presents the pathological charac- teristics of the four mesotheliomas and the Table 2 Dimensions ofanthophyllitefibres in the lungs of three cases of mesothelioma among Finnish anthophyllite exposure data of the patients. Three of the miners workers had pleural and one had peritoneal mesothelioma (one of the pleural mesothe- Case No 1 2 3 liomas had also invaded the abdominal cav- Fibre length (pm): All four were male smokers or on October 2, 2021 by guest. Protected copyright. Range 0-80-27 0 70-44 0-80-31 ity). Mean 6-1 7.9 7-3 ex-smokers and had asbestosis. The three Median 4-1 3.9 5-1 cases with available tissue had a Distribution of lung samples fibre length (%): very high concentration of anthophyllite <1*0 4 2 4 fibres in their about 1% of the fibres 1-04-9 50 55 44 lungs; 5 0-9 9 26 14 29 were coated asbestos bodies. When 400 fibres >10 20 29 23 per sample were identified, only anthophyllite Fibre width (um): fibres were detected. Numerous particles Range 0-05-1 1 0-06-1-5 0-05-1-5 were also detected in the Mean 035 040 0-45 samples. Median 030 035 0-35 All four patients had had long term (13 to Distribution of 31 years) exposure in the mining of asbestos fibre width (%): <0-1 13 8 9 (see table 1). The occupational histories indi- 0-1-0-49 56 62 56 cated no to asbestos other than in 0-5{)-99 30 27 25 exposure )1.0 1 3 10 the Paakkila mine. As well as work in the Paakkila mine, all four had worked in farming Aspect ratio (length/width): Range 3-0-100 3-0-180 3-1-77 and forestry. They all retired when they left Mean 221 25 19 the The time from the Median I14 17 15 company. latency Distribution of onset of employment to the diagnosis was 39 aspect ratio (%): to 58 years. They all had lived in the neigh- <10 332 34 33 10-49 559 55 63 bourhood of the mine before the onset of 50-99 8 8 4 employment. ) 100 3 0 Table 2 summarises the measured sizes of 214 Karjalainen, Meurman, Pukkala

the anthophyllite fibres in the three cases. 29 Finnish patients with mesothelioma, the The dimensions were similar. The median dimensions of fibres found in the lung tissue

lengths of anthophyllite fibres were 3 9 to were measured with the same method as in Occup Environ Med: first published as 10.1136/oem.51.3.212 on 1 March 1994. Downloaded from 5-1 pum with 20% to 29% of fibres exceeding our study.5 In that study, the crocidolite and 10 pm in length. The median aspect ratios amosite fibres were thinner (median width were around 20, and 4% to 1 1% of the fibres 0-06 v 0-35 gm) and had greater aspect ratios had aspect ratios exceeding 50. (median 30 0 v 14-2) than anthophyllite fibres. The size of the anthophyllite fibres (median length 500pm, width 0 35 ,um and Discussion aspect ratio 14-2) was in accordance with our Four histologically confirmed mesotheliomas results (table 2). These median widths are were found among 503 male deaths up to smaller than the 0-61-0-73,um in the earlier 1991 in a cohort of 999 anthophyllite miners report on Finnish anthophyllite miners," and millers. There were no mesotheliomas probably due mainly to the poorer resolution among the 90 female deaths during the same of thin fibres at the lower magnification period. In men, the proportion of deaths from (1500 x) used in that study. mesothelioma (about 0-8% of all male As well as cohorts exposed in mining or deaths) was less than the 1-5% to 18-6% that handling of asbestos, mesotheliomas have have been reported in cohorts exposed to cro- been reported among talc miners, whose cidolite, amosite, , or mixtures con- exposure could have included fibrous tremo- taining these .' The proportional lite.7" Tremolite has also been detected in mortality ratios obtained in different studies geological investigations in the Paakkila area, might, however, not be comparable-for but the amount of fibrous tremolite has not example, because of different age structure of been documented.'3 Exposure to fibrous the person-years in the follow up. For com- tremolite among the Paakkila miners is there- parison, the proportional mortality of fore geologically possible. No tremolite fibres mesothelioma in the Finnish male population were, however, detected among the 400 iden- in 1982-91 according to the statistics of the tified fibres per sample in the lung specimens Finnish Cancer Registry was 0-1%. About in our study. Neither has tremolite been 70% of deaths from mesothelioma in the reported as a contaminant in the Inter- Finnish population are related to occupa- national Union against Cancer (UIGC) tional exposure to asbestos, mainly in ship- anthophyllite standard derived from the building, construction, and maintenance Paakkila mine,'4 15 nor detected in products or occupations.5 Additional deaths from dust samples collected at the mine and mill, mesothelioma in our cohort cannot be com- and analysed in our laboratory. Exposure to pletely ruled out, as only 30% of the death fibrous tremolite among the Paakkila miners certificates were based on necropsy, and pul- is therefore, if not nil, probably very low. monary and abdominal malignancies were The concentrations of fibres found in the included among the cases that were not lungs of the anthophyllite miners were very necropsied. The diagnostics of mesothelioma high. In Finland, with the same microscopy have also greatly improved in Finland during method, similar concentrations have been http://oem.bmj.com/ the past two decades. Therefore the accuracy reported only among asbestos sprayers and of death certificates in the anthophyllite shipyard workers exposed mainly during cohort up to the 1970s may not be compara- asbestos spraying (mostly crocidolite),5 and ble with the data of the general population in an increased risk of mesothelioma (OR 14-4, the 1980s. 90% confidence interval 2-5-178) has been Despite these epidemiological problems, it associated with fibre concentrations exceed-

seems likely that although anthophyllite fibres ing 1 million fibres/g.'6 All our patients were on October 2, 2021 by guest. Protected copyright. can induce mesotheliomas in humans, they heavily exposed and had asbestosis. This are less potent than crocidolite or amosite might be related to the potency of anthophyl- fibres. This might be related to the differ- lite fibres to induce mesothelioma; a reason- ences in the dimensional characteristics ably high level of exposure is needed for a between anthophyllite, crocidolite, and sufficient amount of fibres of adequate amosite fibres. dimensions to be deposited in the lungs. A model for the carcinogenic potency of a as We express our sincere thanks to Kari Karkola MD, Timo fibre a function of its size has been pre- Romppanen MD, and Helena Malinen MD who helped in sented with the conclusion that long, thin obtaining the tissue material, to the Finnish Mesothelioma Panel (Sisko Anttila MD, Tauno Ekfors MD, and Eero fibres are more carcinogenic and fibrogenic Taskinen MD) for their aid in the verification of the diag- than short and thick fibres, but many short noses, and to Terttu Kaustia MA, who did the linguistic revi- fibres may induce a tumour as easily as a few sion of the manuscript. long fibres.8 It has been suggested that fibres 8 pm in length and <0-2 pum in width (aspect ratio greater than 40) are the most important 1 Wagner JC, Sleggs CA, Marchand P. Diffuse pleural in the causation. of mesotheliomas in mesothelioma and asbestos exposure in the North humans.9 Western Cape Province. BrJ Ind Med 1960;17:260-71. The fibrogenic potential of asbestos 2 International Agency for Research on Cancer. Asbestos. fibres has been considered to correlate with Overall evaluations ofcarcinogenicity: an updating ofIARC monographs. Vols 1-42. Lyon: IARC, 1987:106-16. the total surface area of retained fibres inde- ([ARC Suppl No. 7.) pendently of fibre type.'0 Anthophyllite fibres 3 Meurman LO, Kiviluoto R, Hakama M. Mortality and are on the morbidity among the working population of anthophyl- average much thicker than croci- lite asbestos miners in Finland. Br J Ind Med dolite or amosite fibres." In a recent study of 1974;31:105-12. Four cases ofmesothelioma among Finnish anthophyllite miners 215

4 Meurman LO, Kiviluoto R, Hakama M. Combined effect human lung tissue specimens. Ann Occup Hyg of asbestos exposure and tobacco smoking on Finnish 1988;32:323-40. anthophyllite miners and millers. Ann NY Acad Sci 11 Timbrell V. Deposition and retention of fibres in the 1979;330:491-5. human lung. Ann Occup Hyg 1982;26:347-69. 5 Tuomi T. Fibrous in the lungs of mesothelioma 12 Kleinfeld M, Messite J, Kooyman 0, Zaki MH. Mortality Occup Environ Med: first published as 10.1136/oem.51.3.212 on 1 March 1994. Downloaded from patients. Comparison between data on SEM, TEM and among talc miners and millers in New York State. Arch personal interview information. Am Jf Ind Med Environ Health 1967;14:663-7. 1992;21:155-62. 13 Aurola E, Vesasalo A. On the asbestos occurrences in 6 Roggli LV, Pratt PC, Brody AR. Asbestos content of lung Finland and their technical use. Helsinki: Geological Sur- tissue in asbestos associated disease: a study of 110 vey of Finland, 1954. (In Finnish with English summary.) cases. BrI Ind Med 1986;43:18-28. 14 Rendall REG. The data sheets on the chemical and physi- 7 McDonald AD, McDonald JC. Epidemiology of malig- cal properties of the UICC standard reference samples. nant mesothelioma. In: Antman K, Aisner J, eds. In: Shapiro A, ed. Pneumoconiosis, Proceedings of the Asbestos related malignancy. Boston: Grune and Stratton, International Conference Johannesburg 1969. London: 1987:31-55. Oxford University Press, 1970:23-7. 8 Pott F. Animal experiments on biological effects of min- 15 Timbrell V. Characteristics of the International Union eral fibres. In: Wagner JC, ed. Biological effects of mineral Against Cancer standard reference samples of asbestos. fibres. Lyon: IARC 1980:261-72. In: Shapiro A, ed. Pneumoconiosis, Proceedings of the 9 Stanton MF, Layard M, Tegris A, Miller E, May M, International Conference Johannesburg 1969. London: Morgan E, Smith A. Relation of particle dimension to Oxford University Press, 1970:28-36. carcinogenicity in asbestos and other fibrous 16 Tuomi T, Huuskonen M, Virtamo M, Tossavainen A, minerals. Nad Cancer Inst 1981;67:965-75. Tammilehto L, Mattson K, et al. Relative risk of 10 Timbrell V, Ashcroft T, Goldstein B, Heyworth F, mesothelioma associated with different levels of expo- Meurman LO, Rendall REG, et al. Relationships sure to asbestos. Scand Work Environ Health between retained amphibole fibres and fibrosis in 1991;17:404-8. http://oem.bmj.com/ on October 2, 2021 by guest. Protected copyright.