Document QJKv7mk0Zj1dYxq5R6mer9Lk6

\99Q S3 2. 168-180 PROCEEDINGS Asbestos in the lungs of persons exposed in the USA A.M. Langer, R.P. Nolan BSTRACT: Asbestos in the lungs of persons exposed in t USA. A.M. Langer. R.P. Nolan. Tissues obtained at autopsy or biopsy from 81 work > and 2 household persons, were chemically digested. The >estos fibres recovered were characterized by analytical emission electron microscopy. Among the 83 causes ieath were 33 mesotheliomas, 35 lung cancers. 12 asbestoand 3 from other cancers. Of the three major commerdai asbestos fibre types, osite was found to be the most prevalent fibre, occur' ; in -76% of the cases, followed by chrysotile in -60% i crocidolite in -24%. Amosite and chrysotile were ouserved as the single commerdai fibre in -22 and -17% of the cases respectively, whereas croddolite and tremolite were found as the single fibre type in only -2.5% of the cases. Among the fifteen cases where chrysotile and tremo- lite occurred together, die amount of chrysotile fibre always exceeded tremolite. However, tremolite was also found in ten additional cases where chrysodle was not detected. Amo site was present in four, amosite plus croddolite in three, and croddolite alone in one. Amosite was present in all of the insulation workers' lungs studied and was found in the highest concentration in this exposure category. The highest chrysotile concen tration was found among workers in general trades. Although most prevalent in shipyard workers lungs, croddolite con centration is not statistically different among the exposure groups studied. Although croddolite was found in twenty cases, amosite accompanied it in eighteen of these. Eleven of the 20 cases were from shipyard workers. Of the 8 mesothelioma cases, 7 also contained amosite. Crocidolite alone only occurred in 1 of the 33 mesothelioma cases analysed. We concluded the following: crocidolite exposure occurred among USA insulators and a large percentage of TOier WOraers as wen: mmiattnn wnrfegrs posed to amosite; mixed fibre exposures are assodated with iriffTmiesutfiellOmas diart single fibre exposures: chrysotile. only exposure is assodated with -12% of the mwnthfiiama cases smdied: and if tremolite exposure Is asiHtP<Lii.-bh chrysotile exposure, the chrysotile amount exceeds that for tbr uriwiatrrt trrmoHtr tXonaldi Arch Chest Dis 1998; 53: 2. 168-180. Keywords: Asbestos, asbestos diseases, exposure, tissues. Environmental Sciences Laboratory. Brooklyn College of the City University of New York. New York. USA. Correspondence: A.M. Langer. Environmental Sciences Laboratory. Brooklyn, New York 11210, USA. Studies to determine the types, concentration and morphology of asbestos fibre retained in human lung tissue by analytical transmission electron microscopy <TEM) are-important in understanding the aetiology of asbestos-related disease. This is especially so in situations where asbestos-related disease occurs, but the nature of the past exposure is uncertain. Such information has been sought for decades ll]. Recently, the limitations of lung content analy sis for asbestos have been viewed negatively not so much for the information gathered but rather for the lack of agreement between epidemiological claims and the results of lung content analysis that appar ently refute them [2], This is especially so for the asbestos fibre type associated with mesothelioma. For example, the 1976 study by Pooley [3] of Can adian chrysotile miners and* millers raised the issue that tremolite tan amphibole mineral sometimes asso ciated with chrysotile asbestos ores) was an addi tional agent in the induction of asbestosis. Laten this was expanded to include mesothelioma [4], Despite all of the criticisms, analysis of the fibres found in lung tissues are cruical to understanding the aetiol ogy of asbestos-related disease. Currentpostulates concerning asbestos and disease There exist in the USA important claims con cerning asbestos exposure and asbestos-related dis ease which lung content analysis can help resolve. Consider the following six postulates: The mortality experience ofinsulation workers can not be related to crocidolite exposure. Outside of the asbestos cement industry, how extensive has croci dolite usage been in the USA? It has been claimed that crocidolite was not imported into the USA much prior to 1930. and it was not used in insulation prod ucts. Therefore. USA insulators were not exposed to it before 1930 [5] (see also Appendix). During the Second World War. US warships used amosite only and military shipyard workers at this time, unlike those in the UK. were either minimally or entirely unexposed to crocidolite. Using these claims an argu ment concludes that chrysotile may be the agent res ponsible for the majority of mesotheliomas observed in pre-1930 exposed American insulation and ship yard workers, especially those whose exposures began before the widespread introduction of amosite around (930. Therefore. the pre- and post-1930 mesothe lioma experience of the insulation workers, which is much the same, is chrysotile-related. These asser tions have influenced the regulatory policy in the USA on fibre type and diseases [6). Is this claim re garding crocidoiite usage in the USA supported or refuted by lung content analysis and crocidoiite con sumption information? Alt the asbestos fibre types induce mesothelioma equally well. The US Department of Labor's Occu pational Safety and Health Administration in its most recent asbestos standard, still does not distinguish among the asbestos fibre types in terms of mesothe lioma risk [7]. This position stands in opposition to that taken by 14 other industrial countries which per mit different workplace exposure levels for the dif ferent asbestos fibre types [8]. Crocidolite's permissible exposure level is frequently one-tenth the value for chrysotile. The relationship between asbestos fibre type and mesothelioma is frequently couched in subtle seman tics. as reflected bv several recent papers in the USA literature and exchanges they provoked (9. 10]. This exchange focused on the importance of fibre type in producing mesothelioma rather than relative fibre potency. A less potent fibre may produce more tumours in a population because it is more widely used. Chrysotile was therefore deemed more dange rous than crocidoiite. Which fibres types are present in the lungs of workers dying with mesothelioma? Chrysotile exposure alone does not cause human mesothelioma. Some investigators claim that chiysotile alone cannot cause pleural and peritoneal mesothe lioma at any dose. .Are there any cases of mesothe lioma in which only chrysotile can be found in the lung? Chrysotile exposure is invariably associated with tremolite exposure. In addition to crocidoiite. tremolite has been the focus of several studies in the fibre carcinogenesis field. It has long been known that inhalation of tremolite asbestos is associated with Ruffian mesothelioma [11]. Tremolite was reported in'the lungs of chrysotile-exposed miners and millers in Thetford, Quebec [3.12]. Tremolite has been found in greater concentration than chrysotile in the lungs of forty Thetford cases f 13]. Results of epidemiolog ical studies suggest that the lung cancer and mesothe lioma experience of chrysotile-exposed workers are so varied that perhaps tremolite played a role [ 14). Tremolite form, asbestos fibre or cleavage fragment, is vet another issue as is the presence of other nonamphibole fibrous minerals, for example, diopside. This claim has been carried into end-product indus tries and industries which receive and process chiysotile fibre from sources outside of the Thetford area. If chrysotile exposure is associated with mesothelioma or lung cancer, and chrysotile alone is weakly or nonmesotheliomagenic. then the logical conclusion must be that the associated tremolite. which may accom pany the chrysotile. is the aetiological agent in these diseases. Are these assertions supported by lung con tent studies? Exposure to mixtures offibres is more dangerous than exposure to a single type. The fibre svnerav hypothesis was proposed almost 20 yrs ago'm an attempt to explain and correlate mesothelioma mor tality with lung content studies (15). Many more me sotheliomas occurred in workers in the UK whose lungs contained more than one fibre type a.s opposed to lungs which contained only one. One American investigator insisted that his fibre consumption data reflected chrysotile exposure in the American insu lation workers before 1930 and to chrysotile-amosite mixtures after 1930. resulting in a very similar meso thelioma proportional mortality experience. The h> - pothesis was abandoned largely because the experience of the American insulation workers was thought not to fit the pattern. w After all. chrysotile fibre is found in pleural tiss ues. and more mesotheliomas are found in cohorts exposed to both amphibole fibres and chrysotile (15. 16]. What do the fibre data show? " Product type and work environment controls the nature ofexposure, dose and disease. The fibre types and amounts found in American workers' lungs in large measures reflect the products used and the" nat ure of the work environment where the products were manipulated [17]. That is. the insulation trades tended to he exposed hnth to amphibole- and chrvso- tile-containing products, and their work was often eonnnea to cramped and poorly ventilnreri areas J-,- that high concentrations of mixed asbestos dust were generated and subsequendy inhaled. Therefore can ' dose explain, at least in part, the high incidence of asbestos-related disease among insulation workers? In summary, do lung content studies (fibre types, amounts and their character) support or refute these six postulates? Material and methods Almost exclusively, pulmonary tissues were re ferred to our laboratory through consultation. Fur each case asbestos exposure was established on the basis of occupational and/or personal history. The questions asked were often not if exposure to asbestos dust occurred, but rather to which fibre types and concentrations? Further, did these reflect the work ing environment and was there any correlation with a specific disease? The selection was clearly biased toward cases with occupational exposure and asbestos- related diseases. The underlying questions was always: "What is in this lung?" * The tissues were obtained at either autopsy or biopsy, and sent for verification of diagnosis and quantitation of asbestos lung content. The majority were formalin-fixed bulk tissues although some were mounted in paraffin blocks. The 83 tissues examined represent 81 persons occupationally exppsed to asbestos in the work place, and 2 persons thought to have been exposed in a domiciliary setting (table t). A control group for these specimens was not collected during the period of analysis. The specim ens were collected over many years. No case came with a control specimen. Limitations existed for the development of a control group retrospectively. Rather. IflV A.M. LANCER. R.P. NOLAN Table 1. - Information provided with the tissue specimens examines Exposure category Total n=83 Males n=80 Females n=3 Age data u=64 Average age yrs Range Trades identified Category 1 insulator/ pipe coverer 23 23 0 13 60.9*7.2 54-74 Insulator (21) Pipcoverer (2) Category 2 Shipyard wotier 28 28 0 27 60.99.5 45-78 Electrician (3| Welder (2) Machinist (2) Shipfitter (2) Pipefitter (2) Labourer (I) Sheetmetal (1) Sailmaker 11) Seafarer (1,1 Chipper and caulker (1) Shipyard (12) Category 3 Other trades 30 29 1 24 56.59.3* 31-68 Construction (5) Railroad (3) Plasterer (3) Brake repair <2i Pipefitter (2) Plumber (2) Electrician (1) Welder (11 Labourer (1) Sheetmetal (1) Maintenance < 1) Rubber (1) Undercoater autos (1) Other traders (6) Category 4 Family members -> 0 (I 65* . Housewife (2) The average age for ail patients was 59.39.2 yrs. range 31-78 yrs. Data for "Trades identified" are presented as numbers in parentheses. *: the ages of "other trades" are younger than the persons making up the other exposure cat egories (pcO.05 by t-test); *: age available for one housewife only. results were compared with lung content analysis data collected among persons who died in the general population of New York City, during 1966-1968. the years of maximum asbestos consumption in the USA [IX. 19], These persons had no known occupational asbestos exposure and did not die of any asbestosrelated disease. None of the 83 cases reported here had a lung content analysis comparable to anyone in the general population. Information provided for each case The information which accompanied each tis sue specimen varied in detail. The most complete in cluded the' person's name. age. sex. occupational history, years in a specific trade(s). clinical history, primary cause of death (if deceased), secondary or contributing causes of death, e.g.. asbestosis. in lung cancer or mesothelioma, and notation of tobacco and alcohol consumption. Complete data were available for only 63 of the 83 cases studied. Twelve of the 28 shipyard cases were noted as shipyard worker, rather than by specific trade, such as an electrician, in a shipyard. Incomplete data sets are noted (tables I and 2i. Exposure categories After review of the data provided and prelimi nary' analysis of 83 tissues received, it became ap parent that three major exposure categories existed. Category !: insulation workers and pipecoverers. Specimens obtained from twenty three individuals had occupational histories of either insulation work or pipe covering (table 1). Occasionally, the occupa tional history of the insulation worker noted work in other dusty trades. Insulation and pipecovering tra des involved exposure to thermal system products. Only one insulation worker and one pipe coverer had exposure in a shipyard, insufficient to generate a se parate exposure category. They were included as part Table 2. - Distribution of asbestos-related diseases among the exposure categories studied Diseases* Category 1 Insulator/ pipe coverer n=23 Category 2 Shipyard worker n=28 Category 3 Other trades n=30 Category 4 Family members n=2 Total of groups n=83 Mesothelioma Lung cancer Asbestosis* Other cancers 9 (39.1 ins 8 (34.8 ins 6 (26.1)** 0 (0.0) 12 (42.9)ns 12 (42.9 ins 2 (7.1)** 2 (7.1) 12 (40.0)ns 13 (43.3)ns 4 (I3.3)ns 1 (3.4) 0 (0.01 2 (100.0) 0 (0.0) 0 (0.0) 33 (39.76) 35 (42.17) 12 (14.46) 3 (3.61) Data arc presented as number within the category and percentages with that category in parenthesis. *: ascertainment by clinical information, occupational histories, and review of pathology materials. *: asbestosis as a cause of death. Asbestosis was also noted as present in 5/33 mesothelioma cases. 9/35 lung cancers. 1/3 other cancers. Total of ill asbestosis. 27/83 cases or 32.5(5. Note 85.5*5 malignancies among the cases, ns: percentage of mesothelioma and lung cancer among exposure categories 1. 2. 3. not significantly different. **: p<0.01, percentage of asbestosis deaths among categories 1 and 2 arc statistically different. of category l exposure. Category 2: shipyard work ers- Specimens obtained from 28 individuals com prise this group (table 11. Any worker noted as having an exposure component which included a shipyard environment, who was not an insulation worker or pipe eoverer. was included in this category'. Duration of- shipyard work was not a factor in exposure clas sification. For example, in table 1. three electricians are included as shipyard workers because their indi vidual occupational histones indicated employment in this setting. Category 3: other trades and family members. Specimens obtained from 30 individuals comprise this group i table 1 >. If a worker had no ex perience as an insulator or pipe eoverer and never worked in a shipyard environment, the individual was placed in this category. Only 2 of the 83 cases were nonoccupationallv exposed and for exposure comparisons they are grouped with other trades (cate gory 3). Tissue preparation Lung parenchyma specimens were obtained from individuals who developed mesothelioma, lung can cer. asbestosis and other cancers ttable 2). Weighed aliquots of formalin fixed bulk tissues t~i--5 g) were completely digested with 59f potassium hydroxide (20]. A hot water bath, set at ~95C. accelerated the reaction, which was invariably complete after -6 h. Tissues impregnated in paraffin tended to be sma ller in size, some as small as 20 mg. This has not been considered an obstacle in that successful analy sis of small specimens obtained by transbronchial lung biopsies, has been reported [21]. These were subjected to xylene and ethanol baths to dissolve and remove the paraffin. The recovered specimen was washed in distilled water, weighed and subjected to alkali digestion following the above protocol. It has been reported that some paraffin blocks may be contaminated with short chrvsotile (<1 pm) fibres [22]. However, the origin of the fibre was un known. and the issue raised was one which called for the use of control blocks. In this study only fibers >1 pm in length were counted and thereby avoided this confounder. The chemical dtgestates were cen trifuged at lO.OOOxg and resuspended in distilled water 5-8 times. This technique is similar to that used for lung content analysis of exposed persons in the general population (18. 19], A 10 pL aliquot taken from a 10 mL final dis persed KOH-cleared suspension, was micropipetted onto carbon-coated formvar support, nickel locator grid. Four to six such grid preparations were made and allowed to dm After drying, the grids were ex amined by TEM to insure panicle distribution homo geneity' and integrity of the substrate. The dust residue was examined by analytical electron microscopy. The three diagnostics required for identification of fibres were morphology, structural characteristics determined by selected area electron diffraction, and chemical characterization by energy dispersive spec troscopy [23]. Lung content analysis Mineral residues were examined over a range of magnifications, from h00-50.000x on the TEM view ing screen. Fibre counts were made at Ul.OOO-lUAKXrx magnification on the TEM screen by systematic \y traverses. Fibres were noted on a morphological basis and a preliminary separation of chry sotile' and amphibole was made. It must be noted that some fibres were too thick, thereby preventing transmis sion of diffracted electrons, even when the beam was accelerated at 100 kV. Representative thin fibres had their selected area electron diffraction patterns pho tographed and the plates were analysed to verify the presence of an amphibole or chrysotiie structure. Common Laue zone patterns for the amphibole asbestos minerals were used for identification (24-26). Specific amphibole species were determined by chem ical characterization using energy dispersive \-ray spectrometry'. Each spectrum obtained was compared with standards for identification. Fibres were enumerated and counted on at (east 10 grid openings, and based on the area of the prepa ration scanned, fibre number was calculated and expressed as millions of fibres-g-' of wet tissue. For data presentation, the values are expressed as fibresg-1 of dry lung tissue (TO g wet tissue is equivalent to 1 g dry). Fibre concentration values are expressed as those >1 pm in length present in 1 g of dry lung tissue. The fibre length limitation applied in counting was instituted to avoid incorporating background chrysotiie into the final chrysotiie count. This ubiq uitous object, a single chrysotiie fibril with an aver age length of about <1.0 pm. is commonly found in the lungs of persons who die in New York City [18. 19] with no known exposure to asbestos, especially in the occupational setting. The chrysotiie values given here are therefore biased towards exposures which are point source in origin. This also holds, but to lesser degree, for the amphibole asbestos min erals. The limits of detection ranged from -10.000 100.000 fibres-g-' dry lung tissue. The I pm length limit has been embraced by others for similar rea sons [27], Variance is necessarily associated with indivi dual fibre counts. Under ideal conditions of assay, it is related to the actual concentration of fibres pre sent in the specimen and their distribution on the prepared grids. However, it is also related to the amount of tissue digested, from where in the lung the specimen was obtained and preparation proto col. Fibre count in the present study was carried out on at least 10 grid openings, the values obtained were averaged, and the scatter about the mean calculated. As the number of fibres increased, the spread around Che mean decreased. An interesting discussion of the problems in interpreting the results of lung content analysis (both random and systematic errors) is in the literature [2]. Most of the error sources were well known in the analytical community [28. 29] and some of these have been recently discussed [30]. The concentration of fibres >1 pm in length (fibres-g-' of dry lung tissue), ranged considerably. All the data sets contained high or low value outliers which generated arithmetic averages with large stan dard deviations. Therefore, all statistical comparisons are made on the basis of geometric means (GM) and geometric mean standard deviations (GS) which gre atly reduce the effects of outliers. A.M. IANGER. R.K SOLAN Results Fibre types found in the 83 cases Amosite. chrysotile and crocidolite fibres, the major commercially used asbestos fibres, either sin gularly or in combination, were recovered from all but 2 of the 83 parenchymal tissues studied (table 3i. Amosite was the most prevalent fibre found, oc curring in -7ft# of the tissues examined, followed closely by chrysotile. found in -60#. Crocidolite was quantified in 23# of the cases (table 3). in one tis sue specimen a single crocidolite fibre was found and it was noted as present, raising its prevalence to 24#. Greater amounts of chrysotile were found than either amosite or crocidolite (p<0.01) and greater am ounts of amosite than crocidolite (p<0.01). This quan titative trend reflects the general fibre consumption figures over the past eight decades. However, rather than a chrysotile to amphibole tonnage consumption ratio of 19*: 1. the data indicate only a -2.6:1 lung con tent ratio, favouring chrysotile. The role that chrysotile degradation and/or elimination played in this find ing is unknown. Both amosite and chrysotile asbestos were found as the single asbestos fibre type in 39# of the tis sues. as compared to crocidolite which occurred alone in only two of the tissues studied (table 3 and 4). Tremoiite-actinolite (primarily tremolite with some ferrotremolitei and anthophvllite were also found in the tissues studied, with tremoiite-actinolite occur ring more frequently (in -30# of the tissues stud ied) and twice as the only fibre type detected (tables 3. 5 and 6). When both tremolite and chrysotile oc curred in the same tissue (in 12 cases where tremo lite was quantified) chrysotile concentration always exceeded tremolite (table 5). ' Tremolite was found in 10 cases without chiysotile. but was associated with amosite and/or crocidolite in 8 of these (table 6). Anthophvllite was found in -33# of the cases studied although it never occurred as the single fibre type detected and it was quanti fied in only half the cases. Fibre types and concentration by exposure cate gory ' Among the insulation workers (category 1). all of the cases had amosite recovered from their lung tissues (table 7) and in some, crocidolite was also found. Similar findings have been reported by oth ers [31. 32]. The latter report found crocidolite in 7 of the 13 cases examined, at concentrations slightly higher than those reported here. Amosite concentra tion ranged over almost three orders of magnitude. Amosite occurred with highest concentrations in cat egory 1 as compared to those in the other two expo sure categories (p<0.001). The next highest amosite concentration was found among the category 3 ex posure workers. Chrysotile was found In 12 and crocidolite was found in 3 of 23 cases in category 1 ( table 7). ' Table 3. - Presence and concentrations of fibre types found in lung tissues of the 83 cases Fibre type Cases n Cases # Fibre concentrauon 10t'xfibres-er1 dry lung tissue* CM Range Cases with one fibre type n Cases # Lone fibre cases fibre observed # Amosite 63 75.9- 7.56.5 0.2-350 18 21.7 18/63 (28.6i Crociodolite 2(1- 24.1- 2.75.6 0.1-89 2 2.4 2/19 (10.5) Chrvsotrie Tremolite- 5(1 60.2= 26.3*29.5 0.1-7790 14 16.8 14/50 (28.0) 25* 30.1 6 06.5 0.1-177 n 2.4 2/25 (8.0) acunolite Anthophvllite 27* 32.5 3.44.4 0.1-50 0 0.0 0/27 (0.0) Amphibole onl\. 22/83 (26.5#): chrysotile only. 14/83 (17.0#). cm: geometric mean. *: fibres>lpm: v: 19 of 20 cases report quantities. I case single fibre found (present) and quantities based on 19 cases: *: 21 of 25 cases report quantities. 4 cases sin gle fibre found (present) (see table 6) and quantities based on 2t cases:1>: 14 of 27 cases report quantities. 13 cases single fibre found ipresemi and quantities based on 14 cases: :: significant difference between percentage occurrence of crocidolite and amosite (p<t).02i. crocidolite and chrysotile tp<0.0l >. Crocidolite is less prevalent. Table 4. - Asbestos-related diseases and concentration where a single commercial asbestos fibre type is detected Lone fibre found Mesothelioma Lung cancer Asbestosis Amosite gm+sd Range Cases n Chrysotile gm+sd Ranse Cases n Crocidolite Cases n Total cases n (#i 5.8t>.l 0.2-34.8 6 125.6*7.2 11.0-770.0 4 3.0 1 11 (36.4) 7.1*5.3 0.6-110.0 10 87.5*51.8 0.1-5440 8 7.0 1 19 (54.3) 40.6*1.5 30.0-55.0 338.8*34.9 140-820 * 0 4 (33.3) cm: geometric mean. For amosite: asbestosis > mesothelioma (p=0.02-0.05); asbestosis > lung cancer tp=0.03j: and mesothelioma - lung cancer (ns). For chrysotile: asbestosis > mothelioma (p<0.001): asbestosis > lung cancer ({kO.OD; and mesothelioma > lung cancer (jxO.Ol). Table 5. - Tremolite and chrysotitein pulmonary tissues m quantifiable concentrations Chrysotiie Tremolite KX'xfibres-g-1 dry lung tissue Job description Cause of death 1344.0 1340.0 320.0 610.0 366.0 301.0 280.0 220.0 15.0 7.0 74) 1.0 0.7 0.2 0.1 33.128.6 14.0 14.0 2.0 20.0 50.1) 177.0 60.0 4.0 1.4 2.0 0.6 0.5 Present Present Present 7.06.7 House painter/ plasterer Plasterer Plasterer Shipyard Shipyard Shipyard Insulator Engineer Shipyard Insulator Insulator Shipyard Shipyard Shipyard (machinist) Shipyard (electrician) Lung cancer Lung cancer Asbestosis Asbestosis Lung cancer Lung cancer + asbestosis Mesothelioma Lung cancer + asbestosis Mesothelioma -t- asbestosis Lung cancer Mesothelioma Lung cancer Mesothelioma Mesothelioma Mesothelioma Other fibres BLD BLD BLD Amos. Croc Amos. Croc Amos. Croc Amos BLD Amos Amos Amos. Croc Amos Amos Amos BLD BLD: below level of detection; Amos: amosite: Croc; crocidolite. Chrysotiie: tremolite ratio range 410:1 to 2:1. quantitative val ues (n=l2), arithmetric mean ratio -12:1 and geometric mean ratio -4.7:1. Table 6. - Tremolite presence with chrysotiie below limit of detection Tremolite Amosite Crocidolite Disease Job description lO^fibre-g-1 dry lung 94.0 69.7 12.0* 7.0 3.0 1.9 1.8 1.0 0.3 Presenr cm 4.96.7 . 34.8 - - 22.0 5.8 0.6 15.8 0.4 0.2 gm 3.28.4 . - 7.0 3.0 - - 1.3 3.4 - om 3.12.0 Asbestosis Mesothelioma/asbestosis Mesothelioma Lung cancer Mesothelioma Lung cancer Lung cancer Lung cancer/asbestosis Mesothelioma Mesothelioma Construction worker Insulator Housewife** Housewife Railroad worker Pipecoverer Shipyard worker Maintenance worker Electrician/powerplant Welder gm: geometric mean; *: presence of anthophyilice also (600,000 fibres-g-1 dry lung tissue). **: Occupation of husband unavailable: `t noted as present, (one fibre detected). Table 7, - Fibre type, concentration* by category of exposure (trade) Trade' Amosite Crocidolite Chrysotiie Tremoliteactinoiite Anthophvllite Category 1 GM I5.64.7 1.63.7 13.730.1 22.7 9.8 Insulator/ Range 0.8-320.0 0.6-7.0 0.8-5400.0 0.6-70.0 1.0-44.0 pipe coverer vn=23) % 100 13 52,2 26.1 21.7 Category 2 GM 3.46.9 2.58.2 16.821.9 28.0 13.0 Shipyard worker Range 0.2-91.0 0.1-89.0 0.1-7790.0 0.1-177.0 0.3-50.0 m=28) % 75.0 39.3 67.9 32.1 21.4 Category 3 GM 6.67.1 4.2!5 76.024.0 23.3 4,3 Other trades Range 0.2-350.0 t.3-14.0 0.4-5440.0 1.0-94.0 1.0-12.0 I household (n=32) % 59.4 15.6 62.5 31.3 12.5 cm: geometric mean. *: concentration of fibres are given in lO'xfibres-g-' of dry lung tissue; fibres >1 pm length. A large proportion of shipyard workers < categ ory 2) also had amosite (-75%) and chrysotiie (-68%) recovered from, their lung tissue, but crocidolite (-39%) occurred with greater frequency than in the other two exposure categories (table 7). However, the con* centrations found among the exposure categories were statistically indistinguishable <p>0.05). The other trades (category 3) also had amosite and chrysotiie recovered from lung tissues, and some crocidolite as well. The amount of chrysotiie was greatest in cat egory 3 exposure, with about equal amounts found among cases in the other two categories (table 7). The category 3 group contained tissues from three railroad workers. Amosite fibre was found in two (22X106 and 55xl0*>fibres-g'1 dry lung tissue), and crocidolite in one of these (3xl(> fibres-g*1 dry lung tissue). The mixed amphibole exposure occurred in a person with mesothelioma. The chrysotiie-only exposure hypothesis for all railroad workers [331 is not supported by these findings. 173 V.M. LANCER. R.P. NOLAN Amosite concentration ranged over almost three orders of magnitude. The concentration of chrysotile. greatest in the category 3 exposed workers, ranged over almost five orders of magnitude (tables 3. 4 and 7). The tissues obtained from shipyard workers (category 2) tended to have lower concentrations of amosite and chrysotile as compared to those found m insulation workers and workers in other trades (category 3). Fibre types, concentrations and asbestos-related diseases by exposure category Among the insulation workers studied, asbestosis is cited more frequently as a cause of death as compared to the other exposure categories (table 2). Fibre concentrations in insulation workers' lungs indi cate that only their amosite exposure was more inten se than that experienced by workers engaged in other trades or workers in the shipyard environments (table 7). Workers in all exposure categories had virtua lly the same mesothelioma mortality. However, these cases were selected from a nonrandom sample of persons known to have been occupationally exposed to asbestos (table 2). Four of 33 mesotheliomas (-129) occurred in individuals whose lungs contained chrysotile asbestos as the major commercial fibre type (table 4). It has been reported in Germany that '72 of 843 mesothe liomas (-8.59 l had occurred in workers whose lung content analysis found only chrysotile [27]. The concentration of amosite alone, and chrysotile alone, was greater in those with asbestosis. than in those with mesothelioma (pcO.OI). The chrvsotileonly mesothelioma cases contain very high fibre con centrations (table 4). Twenty-two mesotheliomas are associated with mixed-fibre exposures and eleven w ith exposure to single fibre type exposures. High concentrations of uncoated fibres in the lungs of per sons with asbestosis has been reported previously, and a general correlation was found between degree of scarring and fibre concentration [34. 35]. Discussion Lung content analysis for mineral fibre is of the residual fibre population. The long, durable fibres are preferentially retained while both short and non durable fibres are removed. However, many of the lungs examined in this study indicate the presence of very high concentrations of chrysotile. on a fibre number basis, as compared to the amphibole fibres present in the same tissue specimen. Chrysotile is. in some instances, biopersistent [19. 36]. This would mean, (if the elimination data are in principle cor rect and exposure variables are normalized), that ex traordinarily greater amounts of chrysotile would have been inhaled during the lifetimes of these wor kers. Compare the concentrations of chrysotile and eroeidolite as shown, for example, in table 3. The geometric mean chrysotile content of the lungs stud ied is -26x10*' fibres-g-1 of dry lung tissue as com pared to -2.7x10* tibres-g'1 for eroeidolite. about a 10-fold difference (table 3). If 999 of the chrysotile and 809 of the eroeidolite inhaled during life was eliminated, back-calculation yields an almost 200 fold differential in the original inhaled average dose of these two fibre types: 2.600x10* chrysotile to about 14x10*eroeidolite. deposited in a volume of tissue corresponding to the I g of dry lung tissue. Time since cessation of exposure is an important unknown variable among the cases. However it seems unlikely that the high chrysotile fibre concentrations were determined only in recently retired or current ly employed workers, as opposed to those determi nations made for eroeidolite. Values for different fibre types obtained on single lung specimens support this conclusion. Use of lung content analysis in the res olution of the original postulates raised take these processes into consideration. Quantitation of expos ure based on retained fibre number is a relative index. Examination of the six postulates The mortality experience of. American insulation workers cannot be related to eroeidolite exposure. Crocidolite has been found in the lungs of about 249 of the American workers studied and is more prevalent in the lungs of shipyard workers (tables 7. 8). Crocidolite was found in 399 of the tissue spec imens obtained from individuals who had some his tory of shipyard work (table 7i. Five tissue samples with crocidolite were obtained from a railroad work er. an undercoater of automobiles, an electrician, the wife of an insulator, and a maintenance worker (table 8). A trace of crocidolite was found in the tissue of a construction worker. Crocidolite is also present in the lungs of 139 of the insulation workers studi ed (table 7). It is important to note that amosite accompanied crocidolite in -909 of the cases where crocidolite was found (table 8). This suggests that exposure to these fibre types covaried as "the result of either specific products or work sites. Its distrib ution among the exposure categories suggests it was widely used. La.nger and Nolan (17] found that for American insulation workers, amosite was universally present while crocidolite occurred less commonly in their pulmonary tissues^ In all studies, higher prevalence of the amphibole minerals was observed than would be anticipated bv USA commercial consumption fig ures, For example, crocidolite was found in 399 of shipyard workers when only 1-29 of the total asbestos consumed in the USA was this fibre type. Sbjkoff [36] commented during a discussion that crocidolite-contaimng cigarette filters [37J may have been the source of crocidolite fibre found in the lungs of insulation w-orkers reported by Kohvuia and Slzlki [32]. The data now available suggest that this speculation was unwarranted. The historical con sumption data of Selikoff [37J are not accurate (see Appendix) and ample opportunity existed for many American workers to be exposre to crocidolite. Further, crocidolite filters in Kent cigarettes were introduced in1 iqS2 and of -LLQ. metric tonnes consumption-between 1952-1956 [381. This was only a small percentage of the total yearly cro cidolite consumption (-20 metric tonnes-annunr1 as Table 8. - Crocidolite and other fibres, In pulmonary tissues of 20 workers Crocidolite Amosite Chrysotile Job description Cau^e ot death I0hxfibres-g-|dry lung tissue 88.5 90.5 50.0 50.0 20.0 65.0 14.0 1.0 7.0 2!D .7 4 0.4 3.0 22.0 VO ' BLD 3.0 1.0 2.0 5.0 1.7 15.8 1.0 .30.0 1.0 1.0 0.7 14.0 0.7 BLD 0.0 12.0 0.6 16.8 0.4 1.5 0.1 0.2 TR 4.5 301 366 610 BLD BLD BLD BLD BLD BLD 2.0 TR 0.3 BLD BLD BLD 7.0 1.5 BLD BLD BLD Shipyard iweldert Shipyard Shipyard Undercoat autos Insulator Electrician Railroad Shipyard Shipyard Shipyard Maintenance Insulator Shipyard Shipyard (machinist) Wife (insulator) Insulator Shipyard (electrician) Shipyard Shipyard Construction Lung cancer * asbesiosiv Lung cancer Asbestosts Lung cancer AsbestoMs Mesothelioma Mesothelioma Mesothelioma Mesothelioma GI cancer Luna cancer + asbestosis Asbesto.sis Mesothelioma Mesothelioma Lunc cancer Mesothelioma Mesothelioma Lung cancer Lung cancer Lung cancer Geometric mean (sdj for crocidolite 2.45.6. amosite 6.76.6 and chrysolite 17.122.4. BLD: below levei of detection: Gl: eastromtestinai: TR: trace or one fibre detected. Nineteen cases were those in which crociodolite was quantifiable. IX/21) cases as>,ciated with amosite 191K5-1. and 8/20 cases, associated with chrvsotile i-MW. compared to -S.O(X) metric tonneyannunr1 of crocidolite consumed in other applications). It is noteworthy that both insulation and ship yard workers were exposed to amphibole-concaintng and mixed-fibre products. In the former instance, insulator contact was with amosite fibre and. to a lesser degree, depending on time, crocidolite. For ship yard workers, many trades worked in this environ ment. Fibre exposure was. in part, bystander in character with attendant lower quantitative total ex posure (table 5). The first postulate, that crocidolite exposure cat egorically did not occur, especially among insula tors. is rejected on the basis of lung content analysis. Whether or not some or- even all of the early meso thelioma deaths reported among insulators were cro cidolite induced, is unknown. For workers exposed after 1930. the answer is known: yes. some were ex posed to crocidolite: and all were exposed to amosite. All the asbestos fibre types induce mesothelioma equally well. TTie data generated in this study sug gest that the mixed fibre types are^associated with more mesotheliomas (n=22) than, expnsnreln 'finyjjf the three commercial fibers alone. (n=l 1). Among the mesothelioma cases amosite and chrvsotile occur more frequently than crocidolite. Chrysotile exposure alone does not cause human mesothelioma. Compare the data for the different fibre types, in table 4. where concentrations for sin gle asbestos fibre types, by asbestos disease, are given, the amosite and chrysotile exposed populations com prised of almost an equal number of cases (18 and 14 respectively). For mesothelioma. -22 times more chrysotile fibres, on average, are present than amosite. The geometric mean value of -I26xl0h fibres-g'1 of dry lung tissue of chrysotile compared to the 6x10 fibres-g"1 of dry lung tissue for amosite. This diffe rential is similar to that in studies of chrysotile-ex- posed workers who succumbed with mesothelioma [39]. If these numbers are qualitative indices of expo sure and the fibre population represents retained dose, a worker must inhale many orders of magnitude more chrysotile fibre, as compared to amphibole fibre, to achieve the same risk for mesothelioma. The single asbestos fibre type in the mesothelioma pro vides further evidence that chrvs-ntile occurs- without tremolite. and at verv high rlrtMi r-m pm.lm-i-- ral mesothelioma [ 36 ]. Chrysotile exposure is invariably associated with tremolite exposure. Unlike earlier reports of chrysotile miners and millers, in this series of cases when chrysotile and tremolite were found together in the same lung specimen, chrysotile content always exceed ed tremolite. Only in 2 of 83 cases was tremolite the only asbestos fibre type detected. The lung con tent analysis obtained on 94 miners and millers from the Thettord area of Quebec indicated an opposite relationship. Exposure to these ores, some known to contain tremolite. may not reflect a general pattern for all persons who are chrysotile exposed. Based on our data, the tremolite claim does not hold for all chrysotile exposures [14. 35]. Rather, tremolite occurs with highest concen trations in tire lUngs or plasterers (table 5)."Patching, taping and spacklmgcompounils m:irl'-prpri..in the USA are known to have contained tremolitic talc in. addition to chrysotile. This may be a source for iremoliie hut these materials have been shown to con tain cleavage fragments rather than asbestos fl I j. In seven individuals who succumbed with mesothelio ma. amosite and/or crocidolite was found in six. Five of the seven involved shipyard work and the others involved insulation work (table 6). In seven of nine cases where tremolite was found without chrysotile. amosite and/or crocidolite was also present. In four A.M. LANGER, R.P. NOLAN cases where tremolite was found as trace, three con tained amosite. Of the total 25 cases 111 with mesothe lioma) where tremolite was found, nine contained amosite and/or crocidolite. Chrysotile exposure may be associated with tre molite exposure. However, when tremolite is found without chrysotile present, it is generally associated with other types commercial amphiboie asbestos. Calcic-amphibole contamination of amosite needs to be considered as well as other sources of tremolite unrelated to chrysotile. Exposure to mixtures of asbestos fibres is more dangerous than exposure to a single type. These data allow re-examination of risk following expo sure to mixed-fibres as compared with the risk fol lowing exposure to the commercial asbestos minerals alone. Mesothelioma occurred in four cases, where chrysotile was the only asbestos fibre found, as com pared to seven cases where only a single amphiboie asbestos fibre type was found. Mixed-fibre popula tions were associated with 22 of the 33 mesothelio mas. These numbers only roughly approximate the multiplicative model postulated by Acheson and Gardner [15) (22~7 x. 4). Product type and work environment control the nature of exposure, dose and disease. The majo rity of the mesotheliomas (22), occurred in persons whose work environment provided opportunities for intense as well as mixed-fibre exposure. An obser vation which mitigates against dose as a controlling factor (e.g.. insulation workers are exposed to high concentrations of dust) is that the two commercial amphiboie fibres are present in lowest concentrations in individuals with mesothelioma. The data also indi cate that chrysnt'1'* my.-tnrliir-<_m<goihi.lirima in humans, hut the fibre exposure required is much higher than that associated with exposure to amosite and crocidolite. If chrysotile consumption was 19 fold greater than that of amphiboles asbestos during the past three-quarters of a centuiy, why are chiysotile- oniy mesotheliomas not more common? The ques tion is not so much whether or not chrysotile is mesotheliomagenic but rather how potent is it? The data indicate that it is not as potent an agent as. for instance, crocidolite on a fibre-for-fibre basis. Conclusion The pulmonary tissues of a group of asbestosexposed persons in the USA. who died of asbestosrelated diseases, were examined. Amosite was observed with the greatest frequency, followed by chrysotile. anthophyllite. tremolite-actinolite and crocidolite. In terms of concentrations, as indexed by fibre number, chrysotile was the most abundant fibre found, follo wed by amosite. tremolite-actinolite. anthophyllite and crocidolite. The frequency and exposure category occurrence of the amphiboie asbestos varieties amosite and cro cidolite suggest that specific products, e.g.. pipe insu lation containing amphiboie fibres, used by specific trades, and work environments, to a large degree ex plain the disease pattern. The frequency of occurr ence and concentration data are so skewed from USA consumption figures, that the usefulness of the lat ter in gauging exposure is questionable. For exam ple. crocidolite consumption in the USA tended to average only about !->% nf thf-mfrtl fibre used in 'commerce over many decades. However, it was found to be present in the lung tissues of 24% of all work ers and -39% of shipyard workers. A similar distri bution holds for amosite. For American insulators, the presence of amosite is -30-fold greater than is indicated by consumption figures. Amphiboie expo sure among shipyard and insulation workers has been noted previously [I3J: risk of peritoneal mesothe lioma is thought to be especially increased follow ing exposure to amosite [40). The Rochdale studies which focus on mesothe lioma are frequently said to reflect gross fibre con sumption figures. i.e,, 98% chrysotile. 2% crocidolite exposure 141]. However, a small portion of the Roch dale works was engaged in manipulating crocidolite only. The lung content found was consistent with the known exposure to chrysotile although crocido lite was also present at a concentration 300-fold higher than that of the general UK population [42]. Again, in such an example, usage of consumption figures are limited, perhaps even misleading, as exposure sur rogates. This argument, on ratios of fibre types, has been used periodically by the investigators to support the contention that the commercial asbestos fibre types have about equal mesotheliomagenic potential. A recent review in the USA accepted the consumption data at face value, which led to the same erroneous conclusion: chrysotile produces as much malignant disease as the amphiboles [43]. Some amosite pipe lagging was made with cal cined diaiomaceous earth and magnesia Manipulation of such a product generates a complex aerosol. The role that crystalline silica and particulates <2.5 pm in diameter in lung fibrosis in insulation workers remains unclear. Such components of the workplace aerosol may be associated with the small, rounded opacities which frequently accompany irregular scar patterns in asbestosis. There may be many explanations for the high concentrations chrysotile observed in tissues: first ly. there is greater opportunity for chrysotile expo sure because more of the fibre was used: secondly, as chrysotile fibre bundles disaggregate, the number of individual fibrils may actually increase with time of tissue residence. leading to a higher exposure in dex; thirdly, chrysotile possesses some biopersis tence: and lastly, the mineral recovery process may produce an increase in fibre number as an artifact of preparation. Fibres >5 pm in length found in this study con stituted from -5-50% of the fibre populations mea sured. Unfortunately, such data are of limited value here. The great range of tissue masses examined, lack of information concerning location of speci men in terms of anatomical site, and the problems associated with number of fibres counted introduces uncertainty [28, 44-46], There appears to be more long fibres in these tissues than have been reported in bulk samples when the latter are examined by transmission electron microscopy [36], Based on the data tn the present study. the ability to unequivo cally distinguish between the effects of dose and fibre length was not possible. Considering the shortfibre elimination bias, tt may never be possible. Acknowledgements- The author, express gratitude and thanks lo colleagues who cute o generously with time, advice, mate rials'. and cntical review M. Ross or the United States Geological Survey. R- Wilson. Dept of Phvsics, Harvard University. We also note the assistance of A. Bentsianoll who provided assis. lance in the stulislicjt Jnalysi. of dafa. The authors thank S'. Svoyskayu tor preparation ot the manuscript. References I Selikoff li. Chura J. - Biological effects of asbestos. ' Ann NY Acad Sci 1965: 132: 720-721. Baker DB. - Limitations in drawing etiologic infer ences based on measurement of asbestos fibers from lung tissues. Ami NY Acad Sci 1991: 643: 61-70. 5. Pooley FD. - An examination of the fibrous mineral content of asbestos lung tissue from the Canadian chrysolite mining industry. >n- Res 1976: 12: 281-298. 4. Churg A. - Chrysolite, tremolite and malignant mesothe lioma in man. Chest 1980: 93: 621--628. 5. Selikoff 11. Hammond EC. Seidman H. - Cancer risk of insulation workers tn the United States. In: Bogovski P. Gilson JC. Timbell V. Wagner JC. eds. Biological Effects of Asbestos. Lvon. France, 1ARC-WHO, 1973: pp. 209-216. ` 6. Selikoff 1J. - Statement on the proposed asbestos stan dard. OSHA-DOL. Em-Sci Lab Report 1984. 7. Federal Reetster. - 29CFR Parts 1910.1001: 1926.1101. 59. No. 153. Wed. Aug. 10. 1994. 8. International Programme on Chemical Safety. Environmen tal Health Criteria 53. - Asbestos and other natural mineral fibers. Geneva. WHO. 1986. 9. Hughes !M. Weill H. - Potency versus importance in fiber pathogenicity. Am J Ind Med 1994: 25: 609-610. 10. Roggli VL. Pratt PC. Brody AR. - Fiber potency c.r. importance. Am J Ind Med 1994: 25: 611-612. 11. Nolan RP. Langer AM. Oechsle GW, Addison J. Colflesh DE. - Association of tremolite habit with biological potential: preliminary report. In: Brown RC. Hoskin JA. Johnson NF. eds. Mechanisms of Fiber Carcinogenesis. New York. Plenum Press. 1991: pp. 231-251. 12. Churg A. Wiggs B. DePaoli L. Kampe B. Stevens B. - Lung asbestos content in chrysolite workers with meso thelioma. Am Rev Respir Dis 1984: 130: 1042-1045. 13. Churg A. - Pathologic reaction to chrysolite and their mineralogtc correlates. Accompl Oncology 1987:1: 54-58. 14. McDonald JC. McDonald AD. - Chrysotile. tremolite and carcinogenicity. Ann Occup Hyg 1997:41: 699-705. 15. Acheson ED. Gardner MJ. - Mesothelioma and expo sure to mixtures of chrysotile and amphibole asbestos. Arch Env Health 1979: 34: 240-242. 16. McDonald JC. McDonald AD. - The epidemiology of mesothelioma in histone content. Eur Respir J 1997: 9: 1932-1942. 17. Langer AM. Nolan RP. - Fiber type and burden tn par enchymal tissues of workers occupationally exposed to asbestos tn the United States. In: Bignon J, Peto J. Saracci R. eds. Non-Occupational Exposure to Mineral Fibers. Lyon. [ARC-WHO. 1989: pp. 330-335. Itt. Langer AM. Selikoff [J. Sastre A. - Chrysotile asbestos in the lunes of persons tn New York City. Arch Env Health 1971: 22: pp. 348-361. 19. Langer AM. Nolan RP. - Chrysotile biopersistence tn the lungs of persons tn the general population and expos ed workers. Env Health Perspect 1994: 102: 235-239. 2'J. Lunger AM, McCaughey vVRT - Mesotneitomu m u brake repair worker Jumcei 19X2: ri. 1101-1107 21. KohyamaN. Kvmo H. Yokoyama K.Sera 1 - Evaluation of low-level exposure by transbronchrul lung niupsy with analytical electron Tmcroscopv. J Eu-timn v/u up 1993: 42: 315-327, 22. Lee RJ. Florida RG. Stewart IM. - \sbestos contam ination in paraffin tissue block' In ft Pam Uth Ut,/ 1995: f 19: 528-552. 23. Langer AM. Pooley FD. - Identification of single asbestos fibres in human tissues. In: Bogovski p. Gilson JC. Timbrel! V. Wagner JC. eds Biological Effects As-bestos. I.von. iARC-WHO. I9"3; pp."| |o. -125. ' 24. Champnesv PE. Cliff G. Lonmer GW - The identifi cation of asbestos,./ Uicmwnp 1976; ION 215-225. 25. Nord GL. - State-of-the-art ol the analytical transmis sion electron microscope. Jit: Russell PA. Hutchings AE. eds. Electron Microscopv and X-rav Applications. Ann Arbor Sci 1978: pp. (3.7--147. ' 26. Lully JS. Lee RJ. - Computer indexing of electron dif fraction patterns including the effect of lattice symme try. In: Russell PA. Hutchings AE. eds. Electron Microscopv undX-ray Applications. Aim Arbor Sri 1978; pp. |b9-|X8 27. Friedrichs KH. Brockmann M. Fischer M. Wick G - Electron microscopy analysts ot mineral fibers tn human lung tissue. Am J hid Med 1992: 22: 49-58. 28. Churg A. Wood P. - Observations on the distribution of asbestos fibers in human lunes. Env Re\ 19X3; 31: 374-580. ' 29. Gylseth B. Baunan RH. Overaae L. - Analysis ot fibers in human lung tissue, BrJ hid Med 19X2: 79- 191-195, 30. Becklake M, Case BW. - Fiber burden and asbestos- related lung disease: Determinants of dose-response rela tionships. Am J Respir Crit Care Med 1994; ISO: 1488-1492. 31. Suzuki Y. Kohvama N. - Translocation of inhaled asbestos fibers from the lune to other tissues. Aw J hid Med 1991: 19: 701-704. % 32. Kohyana N. Suzuki Y. - Analysis of asbestos fibers tn lung parenchyma, pleural plaques and mesothelioma (is sues of North American insulation workers. Ann NY Acad Sci 1991: 643: 27-52. 33. Mancuso TF. - Response to CG Ohison. Letter to the Editor. Am J Ind Med 1989: 15: 353-356. 34. Churg A. - Fiber counting and analysis m the diagno sis ot asbestos-related disease. Hum Pathol 1982: 13: 381-392. 35. Churg A. Wright JL. Vedal S. - Fiber burden and pat terns of asbestos-related disease in chrysotile miners and milters. Am Rev Respir Dis 1993: 148: 25-51. 36. Nolan RP. Langer AM. Addison J. - Lung content ana lysis of cases occupationallv exposed to chmoule asbestos. Env Health Perspect 1994: 102: 245-250. 37. Selikoff U. - Discussion. Part 2. Ann NY Acad Set 1991; 643: 71-73. 38. Longo WE. Rigler MW, Slade J. - Crocidolite asbestos fibers in smoke from original Kent cigarettes. Cancer Res 1995: 55: 2232-22351 ` 39. Churg A. Wright JL. - Fiber content of fune in ampht- bole- and chrysoule-tnduced mesothelioma: implications for environmental exposure. In: Bignon J. Peto J. Saracci R. eds. Non-Occupational Exposure to Mineral Fibers. Lyon. IARC-WHO. 1989: pp. 314-318. 40- Health Effects Institute-Asbestos Research. Asbestos in Public and Commercial Buildings. Cambridge. MA. HEI- AR. 1991; pp. 6-19. ' 41. Doll R. Peto J. - Asbestos. Effects on health of expo sure to asbestos. Health & Safetv Comm. London. HMSO. 1985: pp. 58. * 42. Wagner JC. Berry G. Pooley FD. - Mesothelioma and asbestos type in asbestos textile workers: a study of lung contents. Br Med J 1982: 285: 603-605. 43. Huncharek M. - Asbestos and cancer: epidemiological 177 A.M. LANCER. R.P. NuLAS and public health controvetMes. C\mcer Invent 1904: I2f2): 214-222. 44 Dodson RF, Greenberg SD. Williams MG. Com CJ. O'Sullivan MF. Hurst GA. - Asbestos content in lungs of occupattonallv and non-occupationaily exposed indi viduals. JAMA 1984: 254: 68-75 45. Dodson RF. Williams MG. O'Sullivan MF. Com CJ. Greenberg SD. Hurst GA. - A comparison of the feruginous bods and uncoated fiber content tn the lungs of former asbestos workers. Am Rev Respir Dis 1985: 152: 143--147. 4o. Dodson RF. Hurst GA. Williams MG. Com CJ. Greenberg SD. - Comparison of light and electron microscopy for defining occupational asbestos exposure in transbronchial lung biopsies. Chest 1988: 94: 366-570. Appendix on crocidoiite consumption information Crocidoiite usage in the USA in the I960`s. when data were yet to be gath ered concerning mortality experience of working co horts exposed to only one fibre type, the position of many investigators, especially those in the USA. was that all varieties of asbestos possessed virtually the same ability to produce disease. In a presentation at the International Pneumoconi osis Conference.held in Johannesburg in 1969. the mesothelioma mortality experienced among North American insulation workers was reported. Although these workers experienced high mortality from me sothelioma. their exposure history was unknown. Selikoff f 11 speculated that for a century earlier, scant information was available on the composition of USA insulation products. However, he went on to state: "Until 1930 chrvsotiie was almost the entire fibre used, amosite making its appearance later and crocidoiite practically not at .all." "This is not to say that it can be proven that.crocidoiite was never used be lore 1930 for insulation in the USA. Rather, if an\ were so utilized, it could only have been in very small amounts. Therefore, crocidoiite exposure of American insulation workers before 1930 and pro bably before 1940 is unlikely to have occurred to any significant extent, and is still a very minor affair." Crocidoiite was widely us-ed outside the USA for insulation and lagging [2]". and the emerging South African experience dramatically underscored the importance of crocidoiite as a mesotheliomagenic fibre [3.4|. The qualifiers of Selikoff [ 1 ] were noted and data were presented concerning mesothelioma mortality among the American insulation workers: 22 mesotheliomas occurred among 373 deaths t-5.9<r l. 2 prior to 1954 and 20 between 1954 and 1968. Clin ical latencies calculated from onset of exposure sug gested important exposures occurred betore 1930. earlier than the time period the authors had reason tu believe that crocidoiite or amosite. was introduced into commerce. Sixteen of 22 mesotheliomas were peritoneal and therefore amphibole fibre was susp ected by many to be the aetiological agent. The issue regarding inclusion of crocidoiite into insulation was again discussed in a paper published three years later 15). Presumably based on the same > ' evidence as presented in his 1970 paper. Selikoff [5| now stated the following: "Crocidoiite was not used for insulation work in the USA during the peri od covered by our study. Thus, exposure to this fibre cannot explain the cancer risk observed." These arguments were repeated so often that they gradually became accepted by members of the Am erican community. At an international meeting on asbestos diseases held in Johannesburg in 1977. McDonald f6] stated "... fit) is fairly clear that insu lators in North America had not been exposed to crocidoiite." Occasionally studies of fibre exposure in trades other than those involving insulation application and removal are used for discussion. A good example of this involved machinists in steam locomotive repair yards [7. 8]. Mancuso [8] stated in his 1988 paper (as underscored by the title of that.paper. "Relative risk of mesothelioma among railroad machinists exposed to chrvsotiie") that during the time period of hire and principal exposure of his mesothelioma cases (1920-1929). the lagging used for steam loco motive: "... was almost exclusively, if not solely chrvsotile". "Exposure occurred to amosite or crocidoiite as packing or other material... would have been com paratively limited in nature and scope." The more circumspect statements concerning exposures are not reflected in this article title nor in its text. The author further buttressed his fibre type argument with the same Department of Commerce importation data and went to state that:"... this lim ited amount of fibre (crocidoiite) was never used for insulation." This latter statement was given further authori ty with a declaration attributed to Hueper [9). In the paper by Mancuso [8] earlier data was repeated and did not reference nor cite his earlier paper in which analysis of six insulation specimens obtained from locomotives showed that four contained amosite only (analyses by F.D. Pooiev. University of Wales). The papers by T.F. Mancuso stood in diametric opposition to advertisements which appeared in trade journals. The Engine, of almost 100 yrs ago (1897) which proclaimed the availability of blue asbestos removable boiler and steam pipe covering, for loco motives. These materials were still advertised 25 yrs later (Asbestos. August. 1921) [10]. Crocidoiite importation into the USA The data used to bolster their exposure theory were later challenged [10], Data for importation of raw crocidoiite fibre into the USA. for the years 1916 through 1930. was from the Department of Commerce [ 1 ]. and those available from the United States Bureau of Mines ias provided in their yearly reports titled "Mineral resources" '! [ 1()|. These data differ consid erably. Prior tn the first import Figures shown [10] in the time period 1916-1929. rather than "no data" for both crocidoiite and amosite (generally inter preted as "none"). 35.040 metric tonnes of crocido iite found its way to USA asbestos manufacturers. It is important to note that these importation figures are for unprocessed fibre, not for finished products. That 35.040 metric tonnes of fibre were used entire ly for valve packings and woven braids is doubtful. i ne aaoesius cement industry wtu, grow mi. juuwusumed much of the fibre. However, crocidoiue was used in insulation products. Crocidolite in USA insulation products importation figures provided by the United States Bureau of Mines showed that crocidolite greatly increased in the USA as the result of lagging required for shipbuilding during World War 1. and expe rienced another period of tonnage importation just prior to World War LI. It was further concluded that based on marketing data found in the trade journal Asbestos that umosite had replaced crocidolite as a component of high temperature insulation rather than the other way around. The American company Asbestos Limited, which held an association with Cape Asbestos. Ltd. of London, was headquartered in New York City*. Its fibre processing plant was located in Boundbrook. New Jersey. It was here that blue fibre was opened and carded, according to customer specifications. Raw crocidolite was derived from deposits in Griqualand (the operation located in the Asbestos Hills of South Africa, north of Koegas and just south of Kuruman. the locality studied by Wagner et al. (3|). Asbestos Limited manufactured magnesia block insulation, advertised as "Best 85# magnesia is made with blue asbestos" (Asbestos. March. 1921). Its application included the product line New Era Insulation, for moderate- and high-temperature systems (Asbestos, July. 1941). Steam locomotives were particularly emphasized. That thermal insulation products contained asbes tos welt past the 1930s was verified in a later study in New York City [12). Of 798 asbestos-containing thermal insulation specimens examined in New York City. II (-1.38%) contained crocidolite. The exposures of American insulation workers were not adequately described in previous reports. Crocidolite was available and did find its way into USA insulation products. It is only its market share which is unknown. - Crocidolite found its way into USA military ship yards Shipyards in the USA. employed the services of thousands of insulators and pipe coverers. Spray insulation containing crocidolite (via the Limpet pro- cos. September. 1936r. *' Naval shipyards during World War II engaged in both construction and repair operations. However, the Navy specified the kind of lagging which was to be used aboard American warships.'and such re quirements were codified by the Navy Department. Insulation products allowed on ships were to be com posed of asbestos fibre with a specific chemistry. When considering the bulk chemistry of available commercial fibres it was obvious that only amosite could be used and if specifically calls for the use of amosite for machinery and piping insulation (I3|. However requirements for ships in the UK dif fered. All three major fibre types were permitted to be used on ships and crocidolite was extensively applied aboard warships [ 14]. Early mesothelioma studies in the UK focused on work in military ship yards [15]. Some international investigators outside the USA interpreted this to mean that crocidolite was also used aboard American ships, and if mesothe lioma occurred among American iasulation workers who worked in military shipyards, this was indirect evidence of crocidolite exposure. However. US Naval specifications rebutted this presumption, and the use argument was no more compelling than the impor tation data referred to earlier. Still other investigators suggested that British ships were re-outfitted in USA ports during the war and they had been the source for crocidolite expo sure to American shipyard workers. This most cer tainly occurred, and citations in the literature support this#. In retrospect, insulation workers, at least those who lived in cities with shipbuilding industries, were probably exposed to at least some crocidolite in this manner. However, those engaged in environments where foreign-ship repairs were undertaken during World War II may have experienced significant expo sures: brief periods of work may have resulted in significant lifetime risks [15. 17. 18]. Appendix references i. Selikoff L). Hammond EC. Churg J. - Mortality expe riences of asbestos insulation workers. I943-I46S. hr. Shapiro HA. ed. Pneumoconiosis. Capetown. Oxford Unix. Press. 1970: pp 97-103. 3. Hendry NW. - The geology, occurrences and major uses of asbestos. Ann NY Atvd Ssi 1965: 132: 12-32. 'Three further points ore noteworthy: the corporate headquarters of J. Manville was at 22 East 40th Street, just across Fifth Avenue Irom Asbestos Limited. The plant in Boundbrook. New Jersey, was located approximately 3 km from the John-Manville (JM) facility in Manville. New Jersey, and the first reports ot mesothelioma in the area, attributed to work at JM only, came from Somerset Hospital in Somerville. NJ. about 4 km from Manville and Boundbrook [111- In retrospect, the mesothelioma experience reported was attributed to exposure at the JM facility, but likely con tained cases exposed at the Asbestos Limited facility as well. The proportion ot cases from each facility when contributed to the total experience is unknown. `Croeiodolite was sprayed, e.g.. in the Sun Shipbuilding and Drydock. Co., in Chester Pennsylvania. Tankers from the Cult Relining Company te.e.. Guff Belle. Gulf Dawni were sprayed under decks and on bulkheads; passenger ferries were sprayed as well. e.c. Princess Anne ot the Virginia Ferry Company. Asbestos il936l. Through President F. Roosevelt's office, and the approval of the Congress. A. Harriman negotiated with and received cooperation Irom East Coast Shipyards to repair British warships in 1440. As an illustrative example, the battleship HMS Rodney was called into action otf Greenland to join tn the pursuit of the Bismark 116). The Rodney was crossing the Atlantic on escort duty after which u was to make for Boston to repair one of its engines which had broken down. It steamed on a single propeller. Alter the sinking ot the Bismark. the Rodney resumed its heading for North America. We assume that British warships carried their crocidolite lagging into USA ports. 174