Document B8VJqX7eaVEdOko9zVvmXmKKo

Pergamon PII: S0003-4878(01 )00(M7-3 A*>n *ini/>. f/i}., V4, 45, >*. pp ^M* $*i, Jilitl Rnttvh IKMijsilnml |i\|;ivii' S>Vh'>v PuMt'lW K hl*f hft Sctcikv Ul AU fijdn- nsten od PnnhrJ tit Gumi Hnutin. Case-Referent Survey of Young Adults with Mesothelioma: II. Occupational Analyses J. C. MCDONALDf*, C. W. EDWARDS*, A. R. GIBBS, H. M. LLOYD+, F. D. POOLEYf, D. J. ROSSt and R. M. RUDDj| IfNational Heart and Lung Institute, Imperial College School of Medicine, London, UK: loyal Orthopaedic Hospital, Birmingham, UK; Uandough Hospital, Penarth, Wales, UK; %Universin of Wales, Cardiff, Wales, UK; ||Sr Bartholomew's Hospital, London, UK Objectives: Our study aimed to Identify occupations at increased risk of developing meso thelioma in persons aged SO yr or less, and to relate these occupations to lung tissue concen tration of asbestos fibres by type. In this age group it was thought that most, but not all, work-related exposures would have been since 1970, when the importation of crocidolite, but not amosite, was virtually eliminated. Methods: Eligible cases were sought horn recent reports by chest physicians to the SWORD occupational disease surveillance scheme. Work histories were obtained for 115 men and 13 women, usually with the help of the chest physicians or coroners. Jobs were coded by the Office of National Statistics, so that the observed years spent in each occupation could he compared with expected values from census data, 1960-90. Lung tissue samples were obtained at autopsy from 69 male and four female cases, and mineral fibres identified, sized and counted by electron microscopy. Results; Of 37 industrial occupations analysed, odds ratios were significantly raised in eight; five in the construction industry and the others in shipbuilding, the manufacture of cement products and the manufacture of nan-metallic mineral products (Including asbestos). The concentrations in lung of crocidolite and amosite fibres, which together could account for 80-90% of cases, did not differ between occupational categories; these for amosite were appreciably higher than for crocidolite. Tremolite fibres were rarely found. Conclusion: Mesothelioma in this young age group is dominated by carpenters, plumbers, electricians and Insulators In the construction industry, and is mainly attributable to ampinbole exposure. Work in shipbuilding and manufacture of mineral products was less Important than in earlier studies. Contrary to expectation, however, some 99% of cases were in men who bad started work before 1970. 2001 British Occupational Hygiene Society. Published by Elsevier Science Ltd. AU rights reserved Keywords: mesothelioma in young adults; occupation; lung fibre analysis EXHIBIT 05^ INTRODUCTION The steady rise in mortality from mesothelioma in Britain since the 1950s, predicted to continue for some years to come, was investigated in the present study in persons, so far as possible, aged 50 yr or less at time of diagnosis. It was argued that the occu pations and lung fibre content of these cases would mainly reflect potential exposure to asbestos since Received 26 January 2001; in final form 14 April 2001. Author to whom correspondence should be addressed. Teh: +44-20-7351-8934; fax: +44-20-7351-8091; e-mail: c.mcdonald@ic.ac.uk 1970. By that date the importation of crocidolite, but not amosite. had been virtually eliminated, and stricter controls were in force through the Asbestos Regulations 1969. The use of crocidolite had already tailed off during the 1960s, but some exposure con tinued after 1970 during building maintenance and asbestos treatment and removal. Since 1989, the Surveillance of Work-related and Occupational Respiratory Disease scheme (SWORD) has provided a consistent estimate of the incidence of occupational respiratory disease in the UK (Meredith et a!., 1991). Some 85% of all consultants in respir atory and occupational medicine throughout the coun try have reported new cases of occupational respir- 519 12/U3/Ui ID', U0 PAA OU" OAO OU ,UJ/|>v.l>' *"V" 520 J. C. McDonald <rt I. atory disease on a voluntary and confidential basis, including mesothelioma whether considered occu pational or not All cases of malignant mesothelioma reported by chest physicians to SWORD between 1990 and 1996, born from 1943 onwards, were selec ted for the study. A later date than 1943 would have been preferable, but would have resulted in too few cases for analysis. The Investigation has two separate but related components. The first entailed identification of eli gible cases, followed by detailed recording of work histories and other relevant data. The second required collection of lung tissue at autopsy from as many of these cases as possible, for electronmicroscopic analysis of lung fibre burden, together with samples for comparison from accidental or sudden cardiac deaths. In tbe latter study, substantially higher con centrations of ail asbestos fibres, all amphiboles, crocidolite and, although rarer, tremolite were found in cases than controls (McDonald el at, 2001, this issue). Crude odds ratios reflected these differences, with large values for all groups and concentrations of crocidolite, amosite and all amphiboles, increasing with concentration. The odds ratios for tremolite were also elevated, but of limited statistical significance. It was clear that almost all cases could be explained by amphiboles, but as chiysotile fibres do not persist in lung tissue, their contribution is less certain. METHODS Case ascertainment All reported cases of mesothelioma which met the bitth date requirements were identified in the SWORD records. With the agreement of the reporting consultant, a research assistant |HML) visited to obtain a full occupational history with his or her help, using medical records. Where necessary, further information was sought from the coroner and occasionally from the patient's general practitioner. In three cases with no available work history, the patient was visited by a research physician (DJR), after obtaining the consultant's, general practitioner's and patient's consent. Of 180 eligible cases, 18 had been reported twice, 14 were too old, five were overseas and, in three, the diagnosis had been changed to benign pleural disease, inflammatory disease or adenocarcinoma, leaving 140 (126 men and 14 women) suitable for study. OF these, no work history could be found for 11 men and one woman. The latter had as a child lived very close to a large factory which processed crocidolite, where her father worked. Her illness had been officially desig nated os attributable to environmental asbestos exposure. Analysis of work histories There were thus 115 men (mean age 47, interquar tile range 44--19) and 13 women (mean age 42. inter quartile range 37-46) available for the occupational analysis- Work histories were recorded, with industty and occupation for each job (up to 16 per patient). These were tabulated and sent for independent coding to the Office for National Statistics (ONS). Observed and expected person-years by occupation were then calculated for the three-year period 1958-60 and for the quinquennia 1961-5, 1966-70, 1971-75, 197680, 1981-85 and 1986-90 by indirect standardisation as follows: Observed. The recorded history and occupation for all recorded jobs of tbe 115 mgn were coded using the Standard Industrial Classification 1992 (SIC92) (Central Statistical Office, 1992) and the Standard Occupational Classification 1990 (SOC90), (Office for Population Censuses and Surveys, 1992) respect ively. This was done independently by members of the Office for National Statistics (ONS). Codings were then condensed into the major classification of the SIC92 (such as construction and metal processing). A matrix was constructed with each case contribu ting to each quinquennium for each job in that period. Where one person had had more than one job, and the starting and finishing dates were unclear, it was assumed that the jobs were of equal length. Any bias resulting from this assumption would tend to under estimate an association between any particular occu pation and mesothelioma. Each man's career lasted n years, starting in the year of his sixteenth birthday and ending in the year he was reported to SWORD as a case. Each year con tributed an equal fraction (1In) to the analysis. The fractions for all cases in each industrial grouping in each quinquennium were added together for the per iod up to 1980, to give the `observed' total used in Table 1. Table 1. Observed and expected mas-years of employ ment. by occupation Industry Occupation O/E (95% Cl) Construction Carpenters Plumbers Electricians Insulators Unskilled Ship and boat building Mfr. cement products Mfr. non-metallic mineral products (including asbestos) Mfr. glass products Electricity generation Mfi. brick and ceramic products Construction -- other trades Ail other occupations (n=25) 8.9 8.1 5.0 3.4 1.9 6.3 5.7 2.8 2.1 1.5 1.4 !.) <1.0 (7.2-11.0) (6.1-10.6) (3.4-7.1) (25-4.5) (1.4-16) C4.7-8.4) (2.1-12.5) (1.3-5.4) (0.8-425) (0.8-2.6) (0.5--3.1) (0.8-1.6) - HWBUI0010252 rX/UD/UJ. lO'.M rAA <>* '6" 1LU *"W " Occupational analyses of mcKcnheliumu Expected. As industrial codings have remained tion industry. A further four groups Imd raKcd odd-, much more similar during 1980 and 1990 than those ratios, but with confidence intervals which included for occupation, it was easier to group denominators 1.0. A detailed tabulation showing observed ami into industrial than occupational categories, so the lat expected values for all 37 groups in seven periods. ter were used in analyses. 1958-60 and six subsequent quinquennia. 1961-91). Census data from 1960, 1970, 1980 and 1990 were can be obtained from the authors on request. used to calculate the expected distribution of Most of the occupational groups at risk, especially employed workers by region and industry. Data were the eight identified above, were clearly associated available by age for each of these censuses, and only with possible asbestos exposure: the converse is true those age groups required for a cohort bom between of the remainder. Construction, shipbuilding and the 1943 and 1965 were selected. For years between the manufacture of cement and non-metailic mineral pro censuses, distributions were calculated assuming a ducts carried the highest evidence of risk, and linear trend (for example, the proportion in each together accounted for almost all the male cases. Con industrial group in 1961 was calculated as trary to expectation, however, 90% of the cases stud [0.9xproportion in 1960] + [0. lxthat in 1970]). The ied were in men who had started work before 1970. expected distribution was summarised for each quin and mainly before 1965. quennium as a percentage for each industrial group, and multiplied by the total `observed' (for all industries) to give the `expected' total for each indus trial group. Observed/expected (O/E) ratios were then calculated for each industrial group by quinquennium. Further analyses were carried out to allow for time spent in each job, and greater weighting given to the earliest exposures; this produced similar results, confinning the general belief that recent exposures have little effect or mesothelioma risk. Certain major industrial groups -- manufacture of non-metailic mineral products not elsewhere classi fied; manufacture of transport equipment; electrical, and water supply; and construction -- were all Women Of the 13 women with a work history, only one had a clear record of occupational asbestos exposure, but another had been a paint sprayer, another a machine operator, and one of the sales assistants had worked in a textile warehouse. The remaining nine women (seven in teaching, nursing or office work, and one each in glass bottling and retail sales) had worked in jobs not obviously entailing exposure to asbestos. No further systematic analysis was made of the female work histories, but see below in relation to long tissue analyses. Sn to be over-represented among workers with sotbelioma, so die analysis was repeated with these Occupation andfibre content major groups broken down into their subgroups. For The relation between occupation and lung fibre the first three categories, this was straightforward, but content is shown in Table 2, limited to aruphibole the denominator data available for 1960 and 1970 for fibres and to male cases with lung analysis at autopsy. the fourth group (construction workers) could not be As some of the men had worked in more than one subdivided. However, an estimate was obtained from type of occupation, the total number in classified Census economic activity tables kindly supplied by occupational categories (94) is greater than the total ONS which showed the occupational distribution for number of cases (69). construction workers, though without breakdown by Although the number in any given occupation is age. It was therefore assumed that the distribution of small, the distribution of fibre concentrations in them occupations was the same regardless of age. A limi all is remarkably similar, both for crocidolite and for tation of this analysis was that the denominator for amosite, although with higher concentrations overall insulation workers included an unidentifiable number than for crocidolite. AH these values for mesotheli in other skills. oma cases are appreciably higher than in controls, The methods for acquiring lung tissue samples and particularly at a=l f/pg. The data for tremolite, for performing the electronmicroscopic mineral fibre although too scanty for interpretation, appear to fol analyses are described in detail elsewhere (McDonald low a similar pattern. st al2001, this issue). Lung tissue was obtained for analysis from only four of the 13 cases in women. In one of these, 1.8 RESULTS f/pg of crocidolite was found. She had worked mainly in retail'sales (see above) but also, for two periods of >fen five months, as a stewardess on ships calling at South The distribution of observed (O) and expected (E) African ports, and was exposed to clouds of asbestos Blues in years employed are summarised in Table 1 dust during loading. In two other women amosite a descending order of estimated risk. Of the 37 fibres were detected; one with 0.7 f/pg, employed idustrial groups or subgroups analysed, odds ratios most of her life in glass bottling, had been exposed (ere significantly raised (as judged by confidence for two years to asbestos packing material used in the itervals) in eight, five of which were in the construc bottle cartons. The third case, with 0.1 f/pg, had 522 J. C. McDonald el. Table 2. Occupation and amphibole fibre concentrations in lung n* CrocidoUte Amosite Any SelStjlg Any srlf/pg Tremolire Any SrXt/(ig Construction work Plumbers Carpesters Electricians Insulators Unskilled Ship/boat building Cement/non-metaUic mineral products None of die above All cases All controls 15 11 (73%) 3 (20%) 13 (87%) 9 (60%) 4 (27%) OH 10 5 (50%) 1 (10%) 9 (90%) 9 (90%) 4(40%) I (10%) 6 3 (50%) 2 (33%) 5 (83%) 3 (50%) 0(-) 0(-) 12 10 (83%) 3 (25%) 11 (92%) 6 (50%) 1(8%) o(-) 12 6(50%) 3 (25%) 10 (83%) 8 (67%) 3 (25%) 0H 12 5 (42%) I (8%) 9(75%) 4 (33%) 3(25%) oh 5 4 (80%) 2 (40%) 5 (100%) 3 (60%) 3 (60%) 0H 22 14 (64%) 4 (18%) 16 (73%) 7 (32%) 2(9%) 0(-) 69 41 (59%) 14 (20%) 56 (81%) 33 (48%) 14 (20%) i (1%) 57 9 (16%) 1 (2%) 23 (40%) 5(9%) 6 (11%) OH "Number of men ever employed in the occupations specified. worked only in clerical or administrative jobs, with no recorded asbestos exposure. DISCUSSION This epidemiological study of mesothelioma is per haps the first to include both occupational and lung burden components in the same investigation of a nationally representative series of recent cases within a narrow and informative age range. It was facili tated -- indeed, made possible -- by the existence of the SWORD surveillance scheme and the full partici pation of the country's chest physicians. The study was greatly simplified by deriving expected values for years employed in ail recorded occupations from agespecific, regional census data, 1960-430. A very large control series would otherwise need to have been selected and interviewed; an expensive, ethically dif ficult and time-consuming task. It should be appreci ated nevertheless that this approach to risk estimation is limited to a comparison of person-years, and allowed no consideration of duration of employment in any given job. Despite the steadily rising mortality from meso thelioma, especially among men, it was important to confirm that there remained little evidence of detect able risk outside a few well-recognised occupations. Confirmation of this does not imply that there are no environmental or household cases, or that there is no risk in other industries, but in magnitude it appears lower than had been feared. Previous populationbased surveys of mesothelioma, predominantly in much older subjects (McDonald, 2000), generally found most cases and highest risks in shipyard work and in asbestos manufacturing and use, particularly for heat insulation. Clearly, this pattern would change with the disap pearance of the main asbestos industries and the great post-war increase in construction work, the first evi dence of which would be seen in the age group we studied. The virtual elimination of crocidolite use before and since 1970, balanced by an enormous increase over the same period in the importation of amosite, mainly for the construction industry, would also be expected to affect the epidemiological picture. It is not surprising therefore to find amosite relatively more prominent than in past studies in the UK (Beny el at., 1989). Regrettably our findings suggest that crocidolite was still important in this young age group, probably because almost all cases were is men who had started work in the 1960s, and also because opportunities' for crocidolite exposure continued in the 1970s, as a result of asbestos removal.' Although construction workers have, since the 1960s, dealt increasingly with amosite-contaimng products, most of their exposure would have been to chiysotile, probably contaminated by some trsmolite. Nevertheless, the lungs of our young cases almost all contained crocidolite or amosite fibres, seldom tremolite, with odds ratios related to retained dose in a man ner suggesting that commercial amphiboles, rather than chrysotile with or without treuiolite, were the cause of most cases. Because of the lack of persist ence in lung tissue of chiysotile, a finding based on lung fibre analysis cannot fully define the role of chiysotile per se. However, the data suggest that despite widespread use of chrysotile, it probably accounted for few cases of mesothelioma. Acknowledgements--This research was conducted with the help of chest physicians, pathologists and coroners throughout the United Kingdom, and was supported by a grant from the Health and Safely Executive. REFERENCES Berry G, Rogers AJ, Fooley FD. Mesothelioma -- asbestos exposure and lung burden. In: Non-occupational exposure to mineral fibres. IAKC Scientific Publications no. 90. Lyon: IARC; 1989. p. 486-96. Central Statistical Office. 1992 Standard Industrial Classi fication of Economic Activities (SIC92). London: HMSQ, 1992. HWBUI0010254 Occupational analyses of mesothelioma McDonald JC. Asbestos, In; McDonald JC, editor. Epidemi ology of work related diseases. 2nd ed. London: BMJ Books: 2000. p. 99-101. McDonald JC, Armstrong BG. Edwards CW. Gibbs AR. Lloyd HM, Pooley FD, Ross DJ, Rudd RM. Case-referent survey of young adults with- mesothelioma: 1 lung fibre analyses. Ann Occ Hyg 2001;45(7):5I3-8. Meredith SK, Taylor VM, McDonald JC. Occupational respir atory disease in the United Kingdom 1`IK`J: ;i report to the British Thoracic Society unci the Society of CX-vitp;mnn:il Medicine by the SWORD project group. Bril J lad Med 1991;48:292-8. Office for Population, Censuses and Surveys tOPCSi. I'Wtt Standard Occupational Classification (SOCWi. Umdnn: HMSO, 1992.