Document DGoGXwwoYVKq6bLznd4YQe5ad

y.vv .t ., /, LUNG CANCER MORTALITY -IN THE UNITED ^ 0^ STATES: SHIPYARD CORRELATIONS , t William J. Blot and Joseph F. Fraumeni, Jr. ` Environmental Epidemiolog)' Branch National Cancer Institute Bethesda, Maryland 20014 . Clues to the origins of cancer are often derived from its geographic patterns. This review notes how United States county mortality data have been used in a stepwise fashion to generate hypotheses concerning (he etiology of June cancer and describes how the leads produced by these data are being tested by analytic epidemiologic studies in areas of the country where lung cancer rates are high, focusing on the associations uncovered with the shipbuilding industry. 1 Geographic Patterns Mapping the distribution-of mortality rates, 1950--49. for lung cancer among white males revealed unexpected patterns.1'1 Rates were high in metropolitan centers of the northeast and midwest, but, surprisingly, the highest mortality clustered in coastal areas of the South. Elevated rates were seen along the Gulf Coast from Texas eastward into the Florida panhandle, with nearly every parish in southern Louisiana being affeaed. A series of counties along the southeast Atlantic coast from northern Florida through Charleston, South Carolina also showed consistently elevated mortal* ity rates among white males. Among white females, the geographic pattern was similar, although the southern coastal elTeci was reduced, while in blades, rates were low throuehouwhe South and high in urban areas of the North. f To seek explanations for the geographic variation in lung cancer across the 1 country, we correlated the county mortality rates with demographic and industrial 1 data available at the county level.1 Regardless of the area of the country, the mortality 4 rates showed a positive gradient with urbanization and a weak inverse relation with socioeconomic indices. From census information on industry for the year 1963, when these data were first available to us at the county level on computer tape, the location and approximate number of employees for every manufacturing plant in the United States were obtained.4 With these data, each county of the country was classified as to its relative involvement for each of IS major two-digit Standard Industrial Classifica tion Code manufacturing categories. When the lung cancer rates were examined in relationship to countywide industrial activity, it was found that mortality tended to be higher in counties with certain industries, namely, the chemical, petroleum.'paper. and transportation manufacturing industries. The elevations in rates in these counties were small but were not accounted for by urbanization or other demographic factors and were primarily limited to males. Shipbuilding is listed as a subcategory of the transportation industry under the classification scheme used. By 1963, few counties had shipbuilding industries, but those that did tended to have high lung cancer mortality rates among white males. During World War II, however, ship construction and repair was the major manufac turing industry in the United States. When mortality rates, during 1950-69 were examined for the 49 counties of the country with major shipyards during the war. 313 0077-W23/79/0330-03I3 SJ.7S/0O 1979. NYAS 0 MMMX .* V v * 314 Annals New York Academy of Sciences there was a consistent pattern of excess rates for.respiratory cancer, particularly in the South/ Nearly all of the southern shipyard counties had high rates for cancers of the lung and larynx. Furthermore, in these areas, the rate of increase for respiratory cancer was greater than that seen nationally. Mortality from eancers of the oral cavity, esophagus, and stomach also tended to be high in shipyard counties, whereas rates for most other tumors were roughly comparable to national levels. This constellation of tumors affecting the airways and upper digestive tract has been described previously among workers exposed to asbestos and raises the possibility that wartime exposures to this agent may be related to the current coastal excess of respiratory cancer in the United States. 1 Field Studies in High-Risk areas The mapping and correlation studies raise questions about the causes of cancer rather than answer them. To follow-up on the leads generated by these studies, we have embarked on a series of field studies in several areas of the country where rotes for respiratory cancer are unusual. As an initial step in investigating the excess lung cancer in coastal Georgia, we obtained copies of approximately HOD death certifi cates; half of the deaths were attributed to primary* lung cancer (esses) and half to other diseases (controls).* The statements on occupation and industry on these certificates revealed that more cases than controls worked in the construction industry. There was no differential associated with paper and pulp manufacturing in the "urban** areas of Savannah. Brunswick, and Waycress, but more cases than controls were employed in wood- and paper-related occupations in the surrounding rural counties. Few of the certificates listed shipbuilding as the usual occupation of the deceased. At the same time, a case-control interview study of lung cancer patients was initiated. Information on the lifetime histories of residence, occupation, and tobacco consumption was obtained on approximately 1100 males, including about 500 with lung cancel diagnosed between 1970 and 1976. Most were long-tim^ residents of the coastal area; Local participation was excellent, with interviews completed for more than 97% of those contacted. Comparison of the^responses among the cases and controls revealed an excess lung cancer risk associated with employment in *ea shipyards during World War 11/ Two large shipyards, in Savannah and Brunswick, began operations in 1942 and by 1943 together employed more than 30.000 persons. Both dosed shortly after the war. Over 20% of the lung cancer subjects identified in the survey reported having worked in the shipbuilding industry, almost all for just a few years during the war. The overall relative risk associated with this employment was 1.6, adjusted for age, race, other occupations, and smoking history. However, a strong synergism was seen between shipyard employment and cigarette smoking level, with the greatest excess for shipyard employment among heavy smokers. ' Data on possible asbestos exposures were scanty; few persons reported handling asbestos as part of their shipyard duties. Of the approximately ISO persons in the study who reported working in the shipyards, only one was listed as an insulator. When risk ratios were calculated according to type of job within the shipyards, excesses were seen for a variety of trades, not just those likely to have heavy asbestos exposure. Prompted by the results of the Georgia study, and by reports of mesothelioma among shipyard workers in the United Kingdom and elsewhere in Europe1 and reports of asbestos-related lung abnormalities among current shipyard workers in the Untied States/*10 we have initiated a second case-control interview study of respiratory cancer 0 ^9 /* * *. * * , hi Sciences * iratery cancer, particularly in the ' high rates Tor cancers of the ^ of increase for respiratory r y from cancers of the oral eh in shipyard counties, whereas 'arable to national levels. This upper digestive tract has been stos and raises the possibility that to the current coastal excess of ;k areas lions about the causes of cancer s generated by these studies, we areas of the country where rate a in investigating the excess lung pproximateiy 1700 death cenifi* lung cancer (cases) and half to :upaiion and industry on these rked in the construction industry, and pulp manufacturing in the ass, but more cases than controls ations in the surrounding rural ; as the usual occupation of the :dy of lung cancer patients was sioence, occupation, and tobacco males, including about 500 with 'ere long-time residents of the -rviews completed for more V f controls revealed an excess lung ards during World War II.7 Two operations in 1942 and by 1943 losed shortly after the war. Over ey reported having worked in the :rs during the war. The overall .6, adjusted for age, race, other ng synergism was seen between i, with the greatest excess for ; few persons reported handling jproximatejy 180 persons in the : was listed as an insulator. When thin the shipyards, excesses were ve heavy asbestos exposure, and by reports of mesothelioma elsewhere in Europe* and reports . shipyard workers in the United rview study of respiratory cancer Blot & Fraumeni: Lung Cancer & Shipyards 3! among residents of Tidewater, Virginia. The shipbuilding Industry is the large: employer in the entire state'of Virginia, with large navy and private shipyards in th Norfolk-Newport News area. Data from this survey are not yet available but shoul help to clarify the possible risk associated with shipyard employment. Since th industry continues to be a large employer today, problems in recalling occupation; detail should not be as severe in Virginia as they were in Georgia where th individuals, or their next of kin, were asked to outline events that occurred 3Q or mor years ago. The oldest shipyard company in the United States is located in the county o Sagadahoc on the coast of Maine. Here the lung cuneer mortality rate is elevated am increasing at a faster pace than the national trend. In this area, the hospital charts o cases and controls are being searched as a preliminary step to possible imerviev studies that should clarify the carcinogenic hazards of shipyard employment. If addition, plans are being developed to conduct a case-control interview study of lunj cancer in southern Louisiana. The petrochemical industry is the large manuf&eturint employer in this area, but as of 9 970 over 12.000 residents were employed statewide ir ship construction and repair, with more than three times this number working ir Louisiana Gulf Coast shipyards during the war. The results from this series of epidemiologic studies should help evaluate the contribution of shipyard exposures (to asbestos and possibly other agents) to the geographic patterns of respiratory cancer in the United States. References 1. Mason. T. F. w. McKay. R. Hoover, W. J. Blot A J. F. Fraumeni, Jr. 1975. Atlas of Cancer Mortality for US. Counties: 1950-49. DHEW Publ. (N1H) 75-780, U.S. Government Printing Office- Washington. D.C. 2. Masson, T. J., F. W, McKay. R. Hoover, W. J. Blot St i. F. Fraumeni. Jr. 1974. Atlas of Cancer Mortality Among US. Nonwhiles: 1950-69. DHW Publ. (NIH) 76-I20-. US. Government Printing Office. Washington. D.C. 4 3. BkOT. W. J, Sl J. F. Fraumeni, Jr. 1976. Geographic patterns of lung cancer in the United . States: industrial correlations. Amer. J. Epidemiol. 103: 539-550. 4 US. Department OF Commerce. 1966. Census of Manufactures 1963. US. Government Printing Office. Washington. D.C . 5. -Blot, W. B. J. Stone, J. F. Fraumeni, Jr. ALE. Morris. 1978. Cancer monslky in 6. US. counties with shipyard industries during World War il. Environ. Res. in press. Harrincton, J. M,, W. J. Blot, R. Hoover, w. j. Housworth, C. W. Heath St J. F. Fraumeni, Jr. 1978. Lung cancer in coastal Georgia: a death certificate analysis of occupation. J. Nat. Cancer Inst. 60: 295-298. 7. Blot, W. J,, J. m. Harrincton, a. Toleoo, R. Hoover. C. W. Heath a j. f. Fraumeni, Jr. 1978. Lung cancer and World War II shipyard employment. .V. Enel. J. 8. Med. 299: 620-624. * International acency for Research on Cancer. 1977. Evaluation of the rcin> genic nsk of chemicals to man. Volume 14, Asbestos. lARC. Lyon, France. 9. Murphy, R. L. B. G. Ferris, w. a. Burgess, J. Worcester A E. a. Caensler. i977. Effects of low concentrations of asbestos. N. Engl. J. Med. 2S5: 1271-1278. 10. Seukoff, I. J. A E. C. Hammono. 1978. Asbestos-associated^ disease in United States shipyards. Ca 2& 87-99.