Document ZBNzbbJqO8gGY3kwx4g2pVLJL

Asbestos Fibers and Pleural Plaques in a General Autopsy Population W. s. F. JUL 11J983 ANDREW CHURG, MD From the Departments ofPathology, University ofBritish Columbia, Vancouver, British Columbia, and the University of California, San Francisco, California It has been claimed that symmetric lower zone pleural or diaphragmatic plaques are markers of asbestos ex posure both in asbestos workers and the general popu lation. In this study, total pulmonary asbestos burden was analyzed for 29 patients selected because pleural plaques were found at autopsy, and the results were compared with values obtained for 25 patients who had no occupational asbestos exposure. The average num ber of asbestos bodies in the plaque groups was 1732/g wet lung, and in the control group, 42/g wet lung. Un coated asbestos fibers were extracted from lung and counted, measured, and identified by morphologic ex amination, electron diffraction, and energy-dispersive x-ray spectroscopy. The total number of fibers/per gram wet lung in the plaque group (114 x 103) was similar to that in the control group (99 x 103), as was the number of chrysotile fibers (51 x 103 versus 68 x 103) and noncommercial amphiboles (13 x 103 versus 29 x 103). However, the plaque patients had a marked increase in the number of the commercially used high aspect ratio amphiboles, amosite and crocidolite (50 x 103 versus 1 x 1(P). A retrospective history of fairly certain asbestos exposure was obtained for 16 of the plaque patients, and such a history correlated strongly with increased numbers of commercial amphiboles in lung. It is concluded that 1) in this general autopsy pop ulation, two subgroups of patients are present. About one half of the patients appear to have developed pleural plaques as a result of asbestos exposure, while the etiology of the plaques in the other half is unclear; 2) the presence of pleural plaques correlates with a modest (50-fold) increase in numbers of long highaspect ratio commercial amphiboles in lung tissue but does not correlate with numbers of chrysotile fibers, noncommercial amphiboles, or the total number of as bestos fibers; 3) asbestos-induced lesions are related to a complex set of mineralogic parameters and not to mere numbers of fibers in lung. (Am J Pathol 1982, 109:8896) PLEURAL PLAQUES are hard, white, slightly ele vated collagenous structures found on the pleurae and diaphragm. They may be completely flat or knobbed, and, at least when associated with exposure to asbes tos, tend to follow the rib outlines over the lower parietal pleurae. Plaques associated with asbestos ex posure are almost always bilateral and most com monly are confined to the lower lung zones; the mor phologically identical structures that may be seen after trauma to the chest or after infection are gen erally unilateral and may be found in any lung field. Plaques are frequently found to be calcified on both pathologic and radiographic examination; the radiographic appearance of calcified diaphragmatic plaques has been considered pathognomonic of asbestos exposure.1-2 The epidemiologic association of exposure to high levels of asbestos and the development of plaques is firmly established.2-3 Whether pleural plaques found in a general population also are invariable markers of asbestos exposure or represent a nonspecific process is uncertain. Selikoff and Lee2 noted that calcified plaques were found in about 2% of chest radiographs of relatives of asbestos workers. Navratil and Trippe4 showed that relatives of asbestos workers and per sons living near an asbestos factory had about 15 times the incidence of calcified plaques, compared Supported by a grant from the Strobel Medical Research Fund of the American Lung Association of San Francisco and by a grant from the National Cancer Institute of Canada. Accepted for publication May 26, 1982. Address reprint requests to Andrew Churg, MD, Depart ment of Pathology, University of British Columbia, 2211 Wesbrook Mall, Vancouver, B.C., Canada V6T 1W5. 0002-9440/82/1011-0088501.25 American Association of Pathologists 88 HWBUI0002105 Vol. 109 No. 1 with controls not living near the factory. Similar re sults for this type of (presumably) modest exposure to commercial exploitation of asbestos were obtained by Hourihane et al,5 by Kiviluoto,6 and by Meurman.7 Burilkov and Mikhailova8 demonstrated that in a dis trict in Bulgaria in which plaques were endemic in both men and cattle, the soil contained amphibole asbestos fibers. Of particular interest are recent re ports from Turkey' showing a high incidence of plaques and mesothelioma in a region in which the soil and local building rocks contain not asbestos but the fibrous mineral erionite. These findings suggest that plaques may appear with rather minimal exposure to asbestos or any other long, thin fibrous material. In contrast, some case-control studies on the gen eral population have failed to show an association of plaques and asbestos exposure. In a recent British report,10 about the same number of plaque and con trol patients gave a history of asbestos exposure, but :he plaque patients had a significantly higher inu'dence of old inflammatory or traumatic events. Rous and Studeny11 also could not find an association jf plaques and asbestos exposure in a Czechoslovadan population, but did find a familial trend, which hey attributed to tuberculosis. "Asbestos" is a generic name for a variety of natually fibrous silicates. More than 90% of the asbestos ised commercially is the serpentine mineral, chrysoile. The remaining forms of asbestos are members of he amphibole group of minerals; of this large group, mly amosite (fibrous grunerite) and crocidolite fibrous riebeckite) have extensive commercial appliation, while actinolite, anthophyllite, and tremolite .re common contamiipnts of other minerals but lave little industrial value. Plaques have been found in persons with high-level 'ccupational exposure to every type of asbestos fiber, 'Ut even in such populations there is little detailed inormation about the types and amounts of dust needed o produce disease and about the actual dust content f the lungs. With the use of the insensitive techiques of counting asbestos bodies or uncoated fibers y light microscopy, it has been shown that persons 'ith plaques and a history of occupational exposure ave more bodies or fibers than control subjects.`-7-11 ,e Bouffant analyzed the plaques themselves in 2 ises and found that only chrysotile was present; the arysotile fibers were considerably more numerous in le plaques than in the lungs.13 Gibbs1-1 suggested lat differences in the mineral species found in local sbestos deposits accounted for the dramatic differtces in the incidence of calcified plaques in Thetford lines, as opposed to Asbestos, both asbestos mining :gions of Quebec. PLEURAL PLAQUES AND ASBESTOS FIBERS 89 For this report, I selected a series of general au topsy subjects who had typical "asbestos-related" plaques and examined both a detailed occupational history and the actual asbestos content of the lungs in an attempt to determine whether history or analysis supported an etiologic role for asbestos in these le sions, and, if so, whether the presence of plaques could be related to total numbers of fibers, sizes of fibers, or some other mineralogic parameter. Materials and Methods Patient Selection The test group for this study consisted of 29 gener al autopsy subjects selected only because they were found to have had lower lobe parietal or diaphrag matic plaques. The control group consisted of 25 pa tients. Of these, 11 men were previously included in a report on asbestos fibers in a general autopsy popula tion15 and were selected on the basis of a lack of his tory of asbestos exposure and an asbestos body count of fewer than 100 bodies/g wet lung, as determined by light-microscopic examination with the use of our previously published methods.16 This number of as bestos bodies appears to indicate a lack of occupa tional asbestos exposure. An additional 14 subjects were selected so as to form a close match for age, sex, and smoking values with the test group, but with out regard to asbestos exposure history or asbestos body count. This latter group included 1 woman and 13 men. The patients had had a variety of occupa tions, largely blue collar; there were a number of nonasbestos miners and quarriers. To obtain better sex matching (since 28 of the 29 subjects in the plaque group were male), I did not include in the present control group the 10 women who were included in the original report on asbestos fibers in the general population.15 A detailed smoking, residential, and occupation al history was obtained for each subject by inTeF-^ view with relatives with a standardized question naire, as previously described.16 Clinical data was obtained by review of the patient's chart. Interviews, chart reviews, and fiber analysis were performed independently. Preparation of Tissue for Light- and Electron-Microscopic Examination For each case, four samples (peripheral lower and peripheral upper lobes, central lower and central up- HWBUI0002106 90 CHURG per lobes) of approximately 2-3 g wet weight of for malin-fixed lung were used. Tissues were prepared by bleach digestion and collection of the sediment on a Millipore filter. This filter was cleared and mounted for light-microscopic quantitation of asbestos bodies or pieces of the filter transferred to electron micro scope grids for evaluation of uncoated fibers. Details of the methods have been previously published.1516 Fibers were identified by a combination of electron optical morphologic study, electron diffraction, and energy-dispersive x-ray spectroscopy. Numbers of fibers and asbestos bodies were calculated as de scribed.15,16 In 4 cases, pieces of pleural plaque were dissolved in bleach and treated with HC1 to remove calcium. Electron-microscopic samples were then prepared as described for lung tissue. A blank was run with each sample. I prepared this by running the entire procedure without any tis sue, preparing an electron microscope grid from a Millipore filter as above, and counting 25-50 squares to detect the presence of asbestos fibers. Only short fibers of chrysotile were seen in these preparations, and then in only occasional samples. When chrysotile fibers were found, the equivalent value of fibers per gram was subtracted from the sample value. The maximum value (minimum limit of detectability) for chrysotile determined in this fashion was 10,000 fibers/g wet lung. All samples were examined at a screen magnification of 18,000 x . To determine the reproducibility of the method, I prepared a series of standards from UICC asbes tos samples, and replicate preparations were made. These showed that the maximum difference between .the highest and the lowest count for the given sample was a factor of 2. Details of these results will be pub lished separately. Histologic Examination Histologic sections of lung and, where possible, gross specimens were reviewed for each case. Sec tions were also prepared from each plaque to ensure that the typical collagenous appearance was present. AJP October 1982 were found. For the purposes of this paper, identifi cation of chrysotile was based on the presence of tu bular forms, a chemical composition of primarily magnesium and silicon, and, when present (which was rare), the typical diffraction pattern of chryso tile. Amphiboles were identified on the basis of a dif fraction pattern consistent with an amphibole (par ticularly the 5.3-A spacing seen in a number of sili cate minerals) and a typical chemical composition. This level of identification corresponds to the "best estimate" category of Chatfield.17 Magnesium, sili con, calcium, and iron containing amphiboles with approximately 10% calcium and equal or greater iron were called actinolite; those with less iron were called tremolite. I have somewhat arbitrarily classified as crocidolite a high-aspect-ratio mineral that produces an amphibole diffraction pattern but has a composi tion closely resembling that of glaucophane, since fibrous high-aspect-ratio glaucophane apparently does not exist (see Results). Statistical Methods Because, in my experience, the distribution of fiber counts is not normal, nonparametric tests for paired data (Wilcoxan signed rank test) and for correlation (Spearman rank correlation) were employed. Results Demographic Data Detailed histories were obtained for 28 of 29 sub jects with pleural plaques. The mean age was 65, with a range of 47 to 82. Twenty eight subjects were men (Table 1). Twenty five of the men and the 1 woman smoked cigarettes; the mean number of pack-years smoked was 54 35 (mean standard deviation). In the control group there were 24 men and 1 woman. The mean age was also 65 years, and the range was 47 to 83. The 1 woman and 22 of the men smoked ciga rettes; the mean number of pack-years smoked was 59 36. .... *, Standards The UICC Standard Reference Asbestos Samples, University of Chicago Tremolite Sample 1611, and a sample of actinolite from San Bernadino County, California, were used as standards. I ran the stan dards through the bleach and peroxide solutions to determine whether compositional changes were pro duced by these procedures. No significant differences Table 1 -- Comparison of Plaque and Control Patients Control group Plaque group (n = 25) (n = 29) Age (mean) Sex (males: females) Smoking (mean pack-years) Asbestos bodies (mean) Total asbestos fibers (mean) 65 24:1 59 42 99 X IQ3 65 28 :1 54 1732 114 X IQ3 HWBUI0002107 Vol. 109 No. 1 PLEURAL PLAQUES AND ASBESTOS FIBERS 91 Table 2-- Numbers and Sizes of Asbestos Fibers in Plaque and Control Patients Fiber type Total fibers (% of all fibers) 1-4.9 (%) Fibers in various size categories 5-9.9 (%) 10+ (%) Plaque patients Chrysotile Noncommercial amphibole Crocidolite Amosite Control patients Chrysotile Noncommercial amphibole Amosite 51 x IQ3 (45%) 13 X 10> (10%) 24 X 103 (22%) 26 x 103 (23%) 68 x 103 (69%) 30 x 1G3 (30%) 1.0 X 103(1%) 44 X 103(86%) 8.6 x 10s (67%) 14 X 103 (56%) 11 X 103 (42%) 60 x 103 (88%) 25 X 103 (84%) 0.3 X 103 (30%) 5.9 x 103(11%) 2.9 X 10* (22%) 7.4 x IQ3 (30%) 8.2 X 103 (31%) 7.3 X IQ3 (11%) 3.9 X IQ3 (13%) 0.6 x 103{60%) 1.5 X 10s (3%) 1.4 X 10*01%) 3.1 x 103 (14%) 7.1 x 103 (27%) 0.7 X 10s (1%) 0.9 X 101 (3%) 0.1 x 103(10%) All values are fibers per gram of wet lung. Asbestos Bodies The mean asbestos body count for the 29 plaque patients was 1732 4l49 bodies/g wet lung (mean standard deviation), with a range of 0 to 19,400 bodies/g. Seventeen of the patients had greater than 100 asbestos bodies/g of lung. For the control pa tients, the mean number of bodies was 42 39/g; 2 of the patients had greater than 100 asbestos bodies/g (120 and 130 bodies/g). for the control group showed that 88% of chrysotile fibers were shorter than 5 p, whereas 70% of the com mercial amphiboles were longer than 5 p and 10% were longer than 10 p. These differences probably re flect both the tendency of chrysotile to fragment into very short fibers, and the fact that most of the chrys otile burden in these patients was derived from at mospheric chrysotile, fibers that are commonly quite short, compared with the fibers used commercially. Uncoaled Asbestos Fibers: Numbers and Sizes For the purposes of this report, the term "commer cial amphibole" is used to refer to amosite and crocidolite fibers and "noncommercial amphibole" to tremolite, actinolite, and anthophyllite fibers, with recognition that the latter types may all contaminate commercial chrysotile. Numbers of fibers for the plaque and control group are shown in Tables 1 and 2. The average number of chrysotile fibers for the 29 plaque patients was 51 X 103 123 X If)3, of noncommercial amphiboles 13 X 103 16 X 103, of crocidolite 24 x 103 87 x 103, and of amosite 26 x 103 + 64 x 103. As shown in Table 2, amosite and crocidolite together accounted for 45 % of the total asbestos fibers, amo site for 23% and crocidolite for 22%. The noncom mercial amphiboles constituted only 10% of the total, and chrysotile 45%. For the 25 control patients, the mean number of chrysotile fibers was 68 x 103 100 x 103, of noncommercial amphiboles 30 x I03 22 x 103, and of amosite 1.0 X 103 + 3.5 x 103. No fibers of crocidolite were identified in the control group. For both groups, distinct differences in fiber size distribution were seen among the different fiber types Table 2). In the plaque group, 86% of the chrysotile >vas shorter than 5 p, whereas almost half of the com mercial amphiboles were longer than 5 p, and ap proximately 20% were longer than 10 p. The results Aspect Ratios and Widths Aspect ratios for the plaque case fibers are shown in Table 3, and widths in Table 4. The mean aspect ratio for all chrysotile fibers was 109, for amosite 89, and for crocidolite 88. There was a consistent increase in aspect ratio with increasing length for chrysotile and the commercial fibers. By contrast, the mean as pect ratio for all the noncommercial amphiboles was 22, and no consistent increasing trend was present. The chrysotile fibers were the thinnest overall, with a mean width for all fibers of 0.054 p. For amosite, the mean width was 0.232 p and for crocidolite 0.182 u. The noncommercial amphiboles were considerably wider (mean width for all fibers 0.533 g), with the longest actinolite fibers reaching a mean width of 1.0 p. All of the amphibole fibers showed a consistent in crease in width with increasing length. The values of aspect ratios and widths for the control group^^gfe Table 3 --Plaque Cases: Fiber Aspect Ratios Aspect ratios for various-sized groups (mean + standard deviation) Fiber type i -4.9 5-9.9 10 + Mean Chrysotile Amosite Crocidolite Tremolite Anthophyllite Actinolite 56.9 63.6 32.4 38.8 43.7 48.0 10.9 12.4 8.9 + 9.5 16.1 29.4 147 153 66.3 + 82.6 93.7 99.1 18.6 29.7 37.0 42.5 41.3 62.6 358 498 123 + 161 156 178 28.1 35.3 45.9 73.8 17.8 15.7 1Q9 89 88 18 27 22 HWBUI0002108 92 CHURG Table 4 -- Plaque Cases: Fiber Width Fiber type Chrysotile Amosite Crocidolite Tremolite AnthophylHte Actinolite 1-4.9 0.043 0.046 0.152 0.201 0.142 0.197 0.368 0.412 0.385 0.360 0.388 0-480 AJP October 1982 Widths for various fiber size groups (values in microns, mean standard deviation) 5-9.9 10 + 0.059 G.Q71 0.200 0.274 0.194 0.301 0.633 0.686 0.418 0.518 0.610 0.812 0.112 Q.133 0.273 0.338 0.234 0.327 0.740 0.838 0.687 0.791 1.000 + 0.812 Mean 0.054 0.232 0.182 0.563 0.463 0.617 quite similar to the values in the plaque group and to the values in our previously published control group and are not reproduced here. Fiber Composition Fiber composition is detailed in Table 5. For most of the fiber types, the plaque cases showed good agreement with the standards. As was true of the control cases,Is and as has been demonstrated in ex perimental animals, chrysotile was extensively leached of magnesium. The crocidolite that we identified was considerably more aluminum-rich and iron-poor than the UICC crocidolite and, in fact, rather re sembled the composition of glaucophane, a mineral closely related to crocidolite.18 However, I have been unable to find reports of an asbestiform glaucophane with such a high aspect ratio. Fiber Distribution The distribution of fibers for the plaque group be tween peripheral and central samples and between upper and lower lobes was compared for all fiber types and for short (less than 5 p long) and long (greater than 5 p long) fibers of all types using the Wilcoxan test. No significant differences were found in the dis tribution of crocidolite, noncommercial amphibole fibers, or chrysotile. Statistically significant accumu lations of amosite fibers were found in the two subpleural samples, compared with the 2 central samples (Z = 1.96, P < 0.05); the difference was even more significant when only long fibers were compared (Z = 2.2, P < 0.03). There were no differences in the distribution of any fibers between upper and lower lobes. Fibers in Pleural Plaques Only chrysotile fibers were found in the four pleural plaques analyzed; no amphiboles were observed, de spite the fact that 1 case, (369), had very large num bers of amphiboles in lung, as shown in Table 6. The numbers of chrysotile fibers in the plaque was less than the lung in 3 of 4 cases, and markedly greater in 1 case. Table 5 --Fiber Composition (Expressed as Elemental Percentages) Mineral Plaque cases Control cases Chrysotile Mg Si Amosite Na Mg Al Si Ca Fe Crocidolite Na Mg Al Si Ca Fe Tremolite Na Mg Al Si Ca Fe AnthophylHte Na Mg Al Si Ca Fe Actinolite Na Mg Al Si Ca Fe 43 + 12.7 57 + 12.2 45 + 9.3 56 9.3 1 3.3 5 2.4 1 * 3.4 41 + 4.6 0 0.8 51 + 4.6 5 + 5.3 4 + 5.2 8 6.2 50 + 6.2 0 1.3 33 10.2 3 4.1 5 + 3.6 0 40 3.6 1 1.0 51 + 3.9 .* 1 1.9 27 3.9 1 1.9 57 + 3.7 11 + 3.2 3 3.1 1 2.0 29 3.5 * 0 1.0 57 3.2 11 + 2.9 2 2.5 0 36 &5 59 3.7 1 2.5 3 4.1 1 2.2 34 + 4.0 0 58 4.0 1 + 1.0 5 3.5 0 18 5.4 1 2.2 50 4.2 11 :t 3.3 20 5.8 0 20 + 5.4 2 2.2 51 3.2 9 1.2 18 6.9 * Not specifically identified in these patients. Reference standard 55 2.4 45 + Z4 t 2.3 7 2.5 0 40 + 3.8 0 52 3.1 8 3.2 2 2.1 1 0.8 40 + 2.1 0 49 + 3.3 1 + 2.0 30 1.6 0 54 0.9 14 + 1.0 00 0+0 35 + 2.5 0 52 2.0 1 1.7 12 3.1 0 25 1.5 1 + 1.5 52 3.0 10 0.8 12 1.1 HWBUI0002109 Voi. 109 No. 1 PLEURAL PLAQUES AND ASBESTOS FIBERS 93 Table 6 --Number of Asbestos Fibers in Pleural Plaques (Fibers per Gram Wet Tissue) Case Chrysotile in plaque Chrysotile in lung Total amphibole in lung 441 17 x 10s 23 X 10a 35 X 1QJ 440 148 X 103 37 X 10* 21 x 10J 369 3 x 103 20 x 101 820 X 10J 359 16 x 10s 37 x 10 48 x 103 Correlations of Numbers of Fibers and Numbers of Asbestos Bodies Correlation of these parameters was performed for the plaque group with the use of the Spearman test. Results are shown in Table 7. A strongly positive cor relation was observed between the number of as bestos bodies and the number of amosite fibers, and between the number of bodies and the number of crocidolite fibers. No correlation at all was observed between numbers of bodies and noncommercial amphiboles or chrysotile fibers. The total number of amosite fibers also correlated strongly with the total number of crocidolite fibers (q = 0.725, P < 0.001). Correlation of Fibers, Bodies, and Occupational Categories Occupations are correlated with asbestos body and fiber counts in Table 8. Eleven of the plaque patients had a history of shipyard work, and 4 more had prob able exposure to asbestos during employment as pipefitters, electricians, construction workers, and in sulators. Nine males %ad a variety of blue collar oc cupations that were not noted for asbestos exposure. Three of the patients (Cases 445, 446, and 369) were considered to have been white-collar workers, but ex amination of the fiber data indicated that patient 369, despite employment as a petroleum chemist, must have had substantial asbestos exposure (see below). Patient 242 made sheetrock for shipyards during World War II for a period of 6 weeks only; she had very large amounts of chrysotile and large amounts of tremolite and actinolite in her lungs (see Churg and Warnock15). Asbestos body and fiber counts for the various oc:upational subgroups are shown in Table 8. The ship yard and the other exposure groups had large numbers of bodies (mean approximately 1400 and 5000/g) and ilso considerably higher numbers of uncoated amosite Ibers (approximately 25,000/g in each group) than he other groups. The distribution of crocidolite tbers was less clearly correlated with occupational group. Noncommercial amphiboles were roughly the same for all categories. Because an asbestos body count of 100 bodies/g of lung appears from previous work'6 to separate ex posed and nonexposed occupational groups, I ex amined numbers of uncoated fibers when the cases were separated by this criterion (Table 9). The major ity of the chrysotile fibers were found in the groups with fewer than 100 bodies/g. Noncommercial am phiboles were the same in both groups, but the com mercial fibers were almost entirely found in the group with greater than 100 bodies/g. The distribution of fiber counts in the groups with fewer than 100 bodies/g is very similar to that in the control population (see Discussion). Histopathologic and Clinical Data Patchy fibrosis was seen in a number of cases and appeared to be represent either nonspecific old scars or the results of current therapy, especially radiation. Diffuse interstitial fibrosis of the type that might be expected in asbestosis was found in only 1 case, but no asbestos bodies were seen in tissue section. Whether this case should be counted as asbestosis is unclear: the fiber count values (chrysotile, 62,000, amosite, 12,000, crocidolite, 4300 fibers/g) are below the aver age for the plaque group as a whole.. Discussion This report has analyzed the occupational histories and pulmonary asbestos burden in a series of 29 pa tients selected from a general autopsy population because they were found to have typical "asbestosassociated" pleural plaques. The results are compared with a series of 25 persons selected from a general au topsy population. These patients did not have plaques. Eleven of 25 were originally chosen because of a doc umented absence of asbestos exposure; the other 14 were selected only to produce a good match for age, sex, and smoking, and general occupational (largely blue collar), but without regard to asbestos Table 7 --Correlations Between Numbers of Asbestos Bodies and Numbers of Asbestos Fibers for Plaque Cases (Spearman's Rank Correlation Test) Sample group Bodies with amosite Bodies with crocidoiite Bodies with noncommercial amphibole Bodies with chrysotile Q 0-71 0.49 .022 .10 Significance P < 0.001 P< 0.006 P = NS P = NS 94 CHURG AjP October 1982 Tables --Asbestos Bodies and Fibers for Plaque Patients by Occupational Category {Fibers per Gram Wet Lung) Category (number of patients) Chrysotile Noncommercial amphibole Crocidolite Amosite Bodies Any history of shipyard work (11) Other probable exposure (construction, electrician, etc.) (4) Remaining blue collar men (9) White collar men (Cases 445 and 446 only) Case 369* Case 242t 28 X 103 60 X IQ3 26 X 103 20 X IQ3 20 x 103 680 x 103 13 x 103 8.8 x 103 8.5 x 103 6.5 X 103 25 x 103 75 X 103 7.8 X 10J 17 X 103 5.2 x 103 11 X 10s 460 X 103 1.6 x 103 25 X 103 23 x 103 5.1 X 10* 0.9 x 103 334 X 103 0 1411 4963 139 62 12,100 36 * See text concerning separation of this case, t Female blue-collar worker. exposure. No patient in the latter set was found to have a history of asbestos exposure. Examination of the demographic data shows that the groups were quite well matched by age, sex, and smoking habits. Most of the patients (22 of 25) were blue-collar work ers. Since we are examining the significance of pleural plaques in a general population, and few persons in such a population will have had occupational asbes tos exposure, the control group appears to be appro priate. This control group could not be used for a de termination of differences in the mineral content of lungs between persons with known exposure with plaques and persons with known exposure without plaques, and the results of this study should not be interpreted in such a manner. The major difference between the two groups is in the number of asbestos bodies and of commercial amphibole fibers. The total number of fibers, and the numbers of noncommercial amphiboles and of chrysotile are roughly comparable, and the distribution of fibers sizes and aspect ratios is also similar between the two groups. The plaque group had an average of 1700 bodies and 50,000 fibers of amosite and crocidolite per gram of wet lung, whereas the control group had an average of 42 asbestos bodies and 1000 fibers of amosite per gram. The values for chrysotile and noncommercial amphiboles in the two groups were quite similar (Table 2). However, the plaque group is not a homogeneous population. When the numbers of fibers in the plaque group are separated according to an asbestos body level of greater or fewer than 100 asbestos bodies per gram (a value previous work has shown to indicate an approximate separation of oc cupational and nonoccupational exposure to asbes tos16), the group with the lower body count has 6900 fibers of amosite and crocidolite per gram of wet lung, whereas the group with the higher body count has 80,000 fibers/g (Table 9). The implication of this finding is that the plaque group, like blue-collar male workers in general,16 is composed of two subpopuiations, one with occupational asbestos exposure and one with exposure similar to that found in the general population. If one attempts to correlate occupation with the number of fibers in the plaque group, the concept that this population is composed of two subgroups is reinforced. The commercial fibers are clearly, in the case of amosite, and less strongly in the case of cro cidolite, associated with known occupational exposure (Table 8). In this particular group, the 11 men with histories of shipyard work ahd the 4 men with other likely exposure had 10 times the amount of amosite and 5 times the amount of crocidolite as the 9 other blue-collar male workers. If one compares the num bers of fibers found in two white-collar male workers (Cases 445 and 446),jhen the group with exposure by Table 9 --Numbers of Fibers in Plaque Patients With Greater and Fewer Than 100 Asbestos Bodies per Gram of Lung (Fibers x ICF/g Wet Lung) Group Asbestos bodies Chrysotile Total amphibole Noncommercial amphibole Commercial amphibole Amosite Crocidolite Less than 100 bodies/g lung Greater than 100 bodies/g lung 46 34 2886 :fc 5165 82 190 32 29 17 21 93 + 191 10 20 14 12 3.3 6.4 42 8! 3.6 6.1 38 110 HWBUI0002111 Vo!. 109 No. 1 occupational history had more than 10 times the amount of amosite, but only 2.5 times the amount of erocidolite. However, we have emphasized in the past15 the dangers of relying on occupational history to rule in or rule out asbestos exposure, and Case 369 in this report illustrates this problem. This patient was a petroleum engineer who had no history of oc:upational exposure. He did, however, do home con struction work, which may have been the source of he high numbers of fibers in his lungs. The result shown in Table 8 is actually not surprisng in light of the statistical data in Table 7, which show that there is a strongly significant correlation tetween the number of bodies observed by light mi:roscopy and the number of amosite and erocidolite ibers. This observation reinforces the concept extressed in our study of the control population'5 that lumbers of bodies are an indicator of long, and trobably of high-aspect-ratio, amphiboles and pro'ide no information about chrysotile content of the ung. Moreover, this correlation is only statistical, md numbers of asbestos bodies are probably a valid creen for total asbestos burden only when large lumbers are found. The 1 woman with plaques (Case 142) had extremely large numbers of chrysotile fibers nd of noncommercial amphiboles present despite an sbestos body count of 36/g. It is noteworthy that, with the possible exception if I case, none of the patients in our plaque populaion had pathologic evidence of asbestosis. Comparion with fiber number data from another laboratory,19 /here patients with asbestosis were found to have on re order of IQ7-10s fibers/g dry lung (106~107 bers/g wet lung), reinforces the conclusions drawn rom epidemiologic studies that the pleura is much tore sensitive than the parenchyma to the fibrosing ffects of asbestos.2 In our previous report on the control patients,,s we bserved that short chrysotile fibers tended to accumlate under the pleura, an observation also made y others in man and experimental animals. In the laque group we were unable to observe any differices between pleura and deep parenchyma for chrystile, but did find significantly greater numbers of nosite fibers in both subpleural samples; no differlces were seen between upper and lower lobes. Whether this implies a direct physical concentration feet in the genesis of pleural plaques is unclear. It is jteworthy that I found no amphibole fibers in the aques themselves, but only chrysotile fibers. The imbers of chrysotile fibers in the few plaques exnined appeared to have no relation to the numbers the underlying lung. PLEURAL PLAQUES AND ASBESTOS FIBERS 95 The results presented in this paper thus suggest that bilateral lower-zone or diaphragmatic plaques in the general population are frequently but not consis tently associated with asbestos exposure. It may be argued that those persons in the plaque group who did not have large numbers of commercial amphi boles had all been exposed to chrysotile that had subsequently dissolved. However, my experience in analyzing chrysotile miners and the experience of others20 has been that chrysotile is usually accom panied by contaminating tremolite and aetinolite fibers, which over the long term persist in large num bers, often larger than the remaining chrysotile bur den. The 1 patient in this group who appeared to have plaques because of chrysotile exposure in fact showed a large number of noncommercial amphibole fibers (Table 8, Case 242). Additionally, if the argu ment that plaques are induced by chrysotile that has disappeared is correct, it would be difficult to account for the good correlation between occupational his tory of exposure and increased numbers of commer cial amphiboles, and it would also be difficult to ac count for the excellent correspondence for numbers of each type of mineral fiber between our control population, who appear to have had only environ mental exposure, and the values for the subgroup of plaque patients who had fewer than 100 asbestos bodies per gram, a number that appears to mark en vironmental exposure. It is more likely that two subgroups are present in this general population. One group had pleural plaques as a result of exposure, most likely occupational ex posure, to commercial high-aspect-ratio amphiboles; the number of commercial fibers averaged about 50 times that of the control population. This argument should not be taken to imply that chrysotile exposure does not produce pleural plaques but points out that in this particular population the exposure, with one exception, was to amphiboles. The other subgroup had pleural plaques of unknown etiology. The cause might have been infection or trauma, as has been suggested, but might also conceivably have been exposure to other types of mineral fibers, since long, thin mineral fibers of many sorts appear to_S|lSi., pleural disease.21 We are currently evaluating the nonasbestos mineral fibers in this population. From the point of view of understanding the patho genesis of asbestos-related disease, the useful point emerges from this study that asbestos-induced lesions may not be merely reflections of numbers of fibers; -.they probably depend on complex mineralogic pa rameters. Thus, in this instance the plaque and con trol groups had about the same number of total HWBUI0002112 96 CHURG fibers, but the plaque group had much larger num bers of long high-aspect-ratio amphiboles. Detailed studies of this type will be required for evaluation of the role of asbestos in cases of malignancy and inter stitial fibrosis, and may have a role in setting expo sure standards. References 1. Roberts GH: The pathology of parietal pleural plaques. J Clin Pathol 1971, 24:348-353 2. Selikoff IJ, Lee HK: Asbestos and Disease. New York, Academic Press, 1978, pp 189-233 3. Becklake MR: Asbestos-related diseases of the lung and other organs: Their epidemiology and implications for clinical practice. Am Rev Respir Dis 1976, 114: 187-227 4. Navratil M, Trippe F: Prevalence of pleural calcifica tion in persons exposed to asbestos dust, and in the general population in the same district. Environ Res 1972, 5:210-216 5. Hourihane DO, Lessof L, Richardson PC: Hyaline and calcified plaques as an index of exposure to asbestos: A study of radiological and pathological features of 100 cases with a consideration of epidemiology. Br Med J 1966, 1:1069-1074 6. Kiviluoto R: Pleural calcification as a roentgenologic sign of nonoecupational endemic anthophyllite-asbestosis. Acta Radio! 1960, Suppl 194 7. Meurman L: Asbestos bodies and pleural plaques in a Finnish series of autopsy cases. Acta Pathol Microbiol Scand 1966, ISl(Suppl): 1-107 8. Buriikov T, Michaiiova L: Asbestos content of the soil and endemic pleural asbestosis. Environ Res 1970, 3: 443-451 9. Baris YI, Artvinli M, Sahin AA: Environmental meso- AJP October 193. thelioma in Turkey. Ann NY Acad Sci 1979, 330:423432 10. British Thoracic and Tuberculosis Association (BTTA) A survey of pleural thickening: Its relation to asbesto: exposure and previous pleural disease. Environ Re 1972, 5:142-151 11. Rous V, Studeny J: Aetiology of pleural plaques. Tho rax 1970, 25:270-284 12. Whitwell F, Scott J, Grimshaw M: Relationship be tween occupations and asbestos-fibre content of tht lungs in patients with pleural mesothelioma, lung can cer and other diseases. Thorax 1977, 32:376-377 13. Le Bouffant L: Investigation and analysis of asbesto: fibers and accompanying minerals in biological materi als. Environ Health Perspect 1974, 9:149-153 14. Gibbs GW: Etiology of pleural cacification: A study oi Quebec chrysotile asbestos miners and millers. Arcl Environ Health 1979, 34:76-83 15. Churg A, Warnock ML: Asbestos fibers in the genera population. Am Rev Respir Dis 1980, 122:669-678 16. Churg A, Warnock ML: Correlation of quantitativt asbestos body counts and occupation in urban patients Arch Pathol Lab Med 1977, 101:629-634 17. Chatfield E: Analytical protocol for the identification ol asbestos fibers, SEM 1982. Edited by O Johari O'Hare, 111, SEM Inc. (In press) 18. Deer WA, Howie RA, Zussman J: Rock forming min erals. Vol 2. Chain Silicates. London, Longmans 1963, 333-351 19. Ashcroft T, Heppleston AG: The optical and electror microscopic determination of pulmonary asbestos fib concentration and its relation to human pathologica reaction. J Clin Pathol 1973, 26:224-234 20. Rowlands N, Gibbs GW, McDonald AD: Asbestos fi bres in the lungs of chrysotile miners and millers. In haled Particles V, 1981 (In press) 21. Stanton MF, Layard M, Tegeris A, Miller E, May M; Kent E; Carcinogenicity of fibrous glass: Pleural re sponse in the rat in relation to fiber dimension. J Nal Cancer Inst 1977, 58:587-603 ""f HWBUI0002113