American Journal of Industrial Medicine 22:193-207 (1992)
Asbestos Fiber Analysis in 27 Malignant Mesothelioma Cases Yoshihiro Murai, PhD, and Masanobu Kitagawa, MD
The asbestos body counts per 5 gm wet lung tissue in 27 (23 pleural and 4 peritoneal) malignant mesothelioma cases derived from 19 autopsy and 8 surgical cases were, according to our own criteria, low level exposure in 13 cases (48.2%),moderate level exposure in 2 cases (7.4%),and high level exposure in 12 cases (44.4%).In our previous study on 235 consecutive autopsy cases, the low level exposure was considered to be environmental, the moderate level was secondary or blue collar, and the high level was occupational. In the present study, about half of the cases examined (44.4%,high level exposure) are closely related to some occupational asbestos exposure and the other half (48.2%)to environmental exposure. The type and size of asbestos fibers from the 12 cases of high level exposure were analyzed and the characteristics were compared with those of cases of low level exposure without lung cancer or mesothelioma. Most fibers analyzed (98%)were longer than 5 pm and thicker than 0.10 pm by our counting rules. In the control group, predominant fibers were tremolite or actinolite. In all the 11 pleural mesothelioma cases, the content of amosite fibers was significantly higher than in the controls. In one case of peritoneal mesothelioma, incipient asbestosis was found and the predominant fibers were crocidolite. It is suggested that the presence of amosite and crocidolite is linked to mesothelioma. The mean lengths of amosite and crocidolite, as detected by our resolution capabilities, were 36.0 and 20.9 pm, and the mean diameters were 0.51 and 0.27 pm, respectively. Both amosite and crocidolite fibers had high aspect ratios (94.2 and 115.4). 0 1992 Wiley-Liss, Inc. Key words: malignant mesothelioma,asbestos fibers, asbestos body, electron microprobe analysis, railroad workers
INTRODUCTION
Since Wagner et al. [1960] first pointed out the association of mesothelioma with crocidolite exposure in South Africa, numerous epidemiological and experimental studies have been carried out, and it is now no longer possible to deny the relationship. However, when confronted with a mesothelioma case, it is not always easy to establish the causative relationship of asbestos exposure to this tumor. And as the epidemiological and pathological studies have gradually disclosed that asbestos Department of Pathology, Faculty of Medicine, Toyama Medical and Pharmaceutical University, Toyama, Japan. Address reprint requests to Yoshihiro Murai, PhD, Department of Pathology, Faculty of Medicine, Toyama Medical and Pharmaceutical University, 2630 Sugitani, 930-01 Toyama, Japan. Accepted for publication December 20, 1991. 0 1992 Wiley-Liss, Inc.
194
Murai and Kitagawa
related lung cancer or mesothelioma can develop without asbestosis [Warnock and Isenberg, 19861, it becomes more difficult to evaluate the relationship of the exposure to the development of malignant tumor. On the other hand, because of the widespread use of asbestos in our societies, it is difficult in some cases to determine past asbestos exposure via history alone. Therefore, we have developed an evaluation chart to assess the causative relationship of asbestos exposure to lung cancer or mesothelioma by examining the asbestos body counts in the lung tissue [Kitagawa and Murai, 19881.
By means of animal inhalation experiments, Wagner et al. [ 19741reported that mesothelioma could be induced by all four types of asbestos examined. Many fibrous minerals other than asbestos, such as glass fiber, are capable of causing mesothelioma experimentally after intrapleural inoculation [Stanton et al., 19811; and recently, erionite, a fibrous silicate mineral, was found to be highly active in humans in inducing mesothelioma [Baris et al., 19791, which bore features since replicated in the mouse [Suzuki and Kohyama, 19841. Stanton and his colleagues reported in 1981 that the carcinogenicity of fibers in rats differed when length and diameter (aspect ratio) differed. Attention was therefore directed to the likelihood that the fibrous quality of asbestos, that is, its morphology, plays an important role in carcinogenic action. In the present study, the asbestos burden as to type and fiber sizes was determined in the lungs of mesothelioma cases, and these data are discussed as related to carcinogenicity. MATERIALS AND METHODS
The 27 mesothelioma samples examined in the present study consist of 20 cases from Toyama Medical and Pharmaceutical University Hospital (TMP) and its affiliated hospitals and 7 cases from medical consultation with one of us (M. K.). To verify the diagnosis and classification, the tumors were stained with hematoxylin-eosin and silver impregnation. Alcian-blue, colloid-iron stains, and periodic acid-Schiff reaction were used to evaluate interstitial mucin and cytoplasmic glycogen. Immunohistochemistry with keratin, epithelial membrane antigen, carcino-embryonic antigen, and vimentin were routinely applied. The clinical and pathological outlines of the cases are shown in Table I. The 30 control cases (male 12, female 18; mean age 64.7 years) belonging to the low level exposure group without lung cancer or mesothelioma were selected from consecutive autopsy material in which asbestos body counts had already been determined [Murai and Kitagawa, 19841. Asbestos body counts were carried out by a modified Smith and Naylor [1972] method [Matsuda, 19751, and asbestos fiber counts were by our modified Langer’s [1971] method. Asbestos Body Quantification Five gram individual samples of wet pulmonary parenchyma from the segments 3, 6, 8, and 10, and, whenever possible, from the opposite side of the lesion in the pleural mesothelioma cases, were minced and digested in 200 ml sodium hypochlorite for 48 to 72 hr. Twenty milliliters of chloroform : 50% ethanol were added and well shaken. The solution in the tubes was centrifuged for 10 min at 800 rpm. After the supernatant was discarded by aspiration, 5 ml of 95% ethanol were added to the bottom chloroform layer and the mixture filtered through a millipore filter (SM type,
58lM 62JM 69lM 72JM 58lM 53lM 63lM 72lF 56lM 50lM
60M
Age1 sex 60lM 59lM 45lM 71lM 74lM 44lF 35lM 621F 75lM 56/F 56lM 52JM 68lM 75lF 73lM 67lM
Ironworker Transmission line worker Ironworker SL mechanic' Mariner, carpenter JNR JNR Shipyard worker Housewife Ironworker Insulation worker
JNR
Occupation Fanner Printing type founder Cleaner JNRc (train dispatcher) Tatami maker Electronic part painter Construction worker Housewife Sake dealer Telephone operator HM distributofl Teacher Clerk Housewife stevedore
"Number of asbestos bodies per 5 gm wet lung tissue. %umber of asbestos fibers per 5 gm wet lung tissue. 'JNR, Japan National Railway. dHM, Household Medicine. 'SL, Steam Locomotive.
19 20 21 22 23 24 25 26 27
18
5 6 7 8 9 10 11 12 13 14 15 16 17
4
Case number 1 2 3
Site of tumor R. pleural L. pleural R. pleural R. pleural L. pleural L. pleural R. pleural L. pleural L. pleural Peritoneal R. pleural R. pleural R. pleural Peritoneal Peritoneal L. pleural R. pleural R. pleural R. pleural L. pleural L. pleural L. pleural R. pleural R. pleural R. pleural R. pleural Peritoneal Gross Localized Diffuse Diffuse Localized Localized Localized Localized Localized Localized Diffuse Localized Localized Localized Localized Diffuse Diffuse Diffuse Diffuse Diffuse Diffuse Diffuse Diffuse Diffuse Diffuse Diffuse Diffuse Diffuse
Cell type Sarcomatous Epithelial Desmoplastic Sarcomatous sarcomatous Epithelial Sarcomatous Sarcomatous Sarcomatous Epithelial Epithelial Epithelial Sarcomatous Sarcomatous Epithelial Epithelial Epithelial Epithelial Sarcomatous Biphasic Sarcomatous Epithelial Biphasic Epithelial Epithelial Sarcomatous Euithelial 6 8 8 10 10 13 16 31 41 58 173 264 365 525 549 686 875 1,064 2,256 3,390 17,300 2.720.000
4 4
1 2 3
(a)
2,640 6,380 5,280 880 2,420 6,400 5,900 20,000 30,000 12,000 22,000 7,800 27,000 76,000 8,400 55,000 19.000.000
660
(b) 440 220> 220> 220> 27,500 2,200 220 2,640 1,100
(b)/(a) 440.0 110.0> 73.3> 55.0> 6875.0 366.7 27.5 330.0 110.0 66.0 203.1 398.8 170.3 21.5 41.7 37.0 22.3 54.8 57.1 21.9 32.1 8.9 25.4 33.7 2.5 3.2 7.0
TABLE I. Clinical and Pathological Features of 27 Malignant Mesothelioma Cases (19 Autopsy, 8 Surgical) With Numbers of Asbestos Bodies and Asbestos Fibers Found in Lung Tissue
1%
Murai and Kitagawa
5 pm pore size). The filter was dried and dipped in xylene and mounted without staining. Four filters were prepared for each segment. For those cases in which asbestos body counts were >3,000,the solutions were diluted. Asbestos bodies were counted by searching all fields in the preparations with a light microscope at 100 x magnification. The average number of four segments was used as the number of asbestos bodies per 5 gm wet lung tissue for each case. The core fibers of the asbestos bodies we counted were all asbestos fibers as far as confirmed by energy dispersive electron microscopy [Murai and Kitagawa, 19841.From a study on 235 consecutive autopsy cases, three (3) levels of asbestos exposure were determined using asbestos body counts per 5 gm wet lung tissue: low level exposed group (140);moderate level exposed group (41-150); and high level exposed group (2151).In our previous report [Murai and Kitagawa, 19841,a count of 140 (separated into 1-10 and 11-40) corresponded to low level exposure and was considered to be due to ambient air pollution.
Asbestos Fiber Quantification For the preparation of asbestos fibers, 0.5 gm wet pulmonary parenchyma from segment 10 of the same lung as the asbestos body preparation was removed and dissolved in 5 ml of 40% potassium hydroxide; this was heated in boiling water for 1 hr. After the tissue had completely dissolved, 5 ml of distilled water was added to the tubes and the suspension centrifuged for 30 min at 2,000 rpm. The sediment was washed with 10 ml distilled water and centrifuged under the same conditions. After two cycles of washing, the suspension was diluted until fiber counting was possible. A total volume of 1 to 5 ml was prepared in almost all cases. A 25 p1 aliquot of the dilute was dropped onto a collodion-coated Maxtaform-grid (HF15) on a dental plate. This grid has printed symbols by means of which it is easy to recognize the location of the objectives. After drying, the grid was examined by transmission electron microscopy (JEOL 200-CX)at X 2,000magnification. All fibers with an aspect ratio 2 3 were photographed for size measurement. The photographed fibers were then analyzed by scanning electron microscopy equipped with an X-ray energy dispersive spectrometer (HITACHI X 650 KEVEX 7000 Q System) for the identification of fiber type. We selected a magnification of X2,000,as we intended to analyze as many fibers longer than 5 pm as possible. Fibers 2 5 pm in length and 2 0.10 pm in diameter were included in the counting procedure. Although with a magnification of ~2,000, fibers longer than 5 pm are detectable, most fibers shorter than 2-3 pm would not be seen. We searched all observable fields in the grids until the number of asbestos fibers exceeded 100. For the high level exposure group, from 100 to 141 (mean 108.5) asbestos fibers for each case were detected and analyzed in 3 to 74 grids (mean 25). The first 100 fibers were used to calculate the percentage of asbestos types and their sizes. The limit of detectability for the high level exposure group was less than 800 in all cases but one. Ten grids were examined in the low and moderate level exposure groups with the limit of detectability being 220 in 12 cases, 440 in one case, and 1,100in 2 cases. Two or four grids were examined in each of the controls and the limit of detectability was 1,100in all. To measure the spread of a 25 p1 drop, ten such drops of distilled water were placed on a dental plate and their diameters measured. The average was 4.24 mm,
+
Asbestos Fibers in Malignant Mesotheliomas
197
corresponding to an area of ~ ( 2 . 1 2 mm2. ) ~ Because the edge of the grid (3 mm diameter) was concealed by the holder of the electron microscope, the observable diameter of the grid was 2 mm. From an enlarged photograph of the grid, we calculated the active area to be 82% of the (2 mm diameter) observable area, or 0 . 8 2 ~ ( 1 )Thus, ~ . the ratio of the total area of a 25 p1 drop on the dental plate to that = 5.5. of the observable grid area is ~ ( 212)2/0.82~(1)2 . The number of asbestos fibers per 5 gm wet lung tissue was calculated by the following formula: Al25 pl x B/C x 5 gdO.5 gm x 5 . 5 , where A is the total suspension volume (pl), B is the total asbestos fiber count, C is the total number of grids, and 5.5 is the ratio of the total area of 25 p1 aliquot placed on the dental plate to the actually observable grid area. RESULTS
From the opening of our hospital (TMP) in October 1979 to December 1988, there were 1,677 autopsy cases over 20 years of age (male 1,040, female 637). Over the same period, there were 430 (male 290, female 140) in Toyama Prefectural Central Hospital (TPC), which is the affiliated teaching hospital of our university. Against this background, six pleural and two peritoneal mesotheliomas were encountered, a frequency of 0.38% of the adult autopsies (8/2107). Clinical and Pathological Outline and Grade of Asbestos Exposure of the 27 Malignant Mesothelioma Cases
Table I shows the 27 collected cases of mesothelioma, 8 from surgical and 19 from autopsy. Twenty-three of the cases were pleural and four peritoneal. The cases are listed in Table I in order of asbestos body counts: there were 22 males and 5 females with a mean age of 60.9 years. Macroscopically, there were 16 diffuse and 11 localized mesotheliomas. Microscopically, 13 were epithelial, 11 sarcomatous, 2 biphasic, and 1 desmoplastic. These 27 cases were divided into 3 groups according to our criteria for asbestos body counts, 13 belonging to the low level (540) exposure, 2 to the moderate (41-150) level exposure, and 12 to the high (1151) level exposure. Lung tissue from a 50 year-old man with peritoneal mesothelioma gave a count of 2.72 X lo6 asbestos bodies; his was the only case with demonstrable asbestosis. The patient had been an insulation worker for 11 years from age 20 to 31 years. The other patients in the high level category were almost certainly exposed to asbestos in their occupations. Four had worked in the former Japan National Railway and three were ironworkers. All mesotheliomas in the high level group had developed in a diffuse fashion, although this was also found in the low level group. Both biphasic mesotheliomas were in the high level group. Seven epithelial type mesotheliomas were seen in the 12 cases in the high level group. However, no clear-cut relationship between histological type and asbestos exposure was observed. The correlation between asbestos body counts seen at lOOX and asbestos fiber counts seen at 2,000 X is shown in Table I. Here, although the ratio of the body counts to the fiber counts is not always constant, both correlated with each other relatively well (r=0.858, p
Asbestos Fiber Analysis in 27 Malignant Mesothelioma Cases Yoshihiro Murai, PhD, and Masanobu Kitagawa, MD
The asbestos body counts per 5 gm wet lung tissue in 27 (23 pleural and 4 peritoneal) malignant mesothelioma cases derived from 19 autopsy and 8 surgical cases were, according to our own criteria, low level exposure in 13 cases (48.2%),moderate level exposure in 2 cases (7.4%),and high level exposure in 12 cases (44.4%).In our previous study on 235 consecutive autopsy cases, the low level exposure was considered to be environmental, the moderate level was secondary or blue collar, and the high level was occupational. In the present study, about half of the cases examined (44.4%,high level exposure) are closely related to some occupational asbestos exposure and the other half (48.2%)to environmental exposure. The type and size of asbestos fibers from the 12 cases of high level exposure were analyzed and the characteristics were compared with those of cases of low level exposure without lung cancer or mesothelioma. Most fibers analyzed (98%)were longer than 5 pm and thicker than 0.10 pm by our counting rules. In the control group, predominant fibers were tremolite or actinolite. In all the 11 pleural mesothelioma cases, the content of amosite fibers was significantly higher than in the controls. In one case of peritoneal mesothelioma, incipient asbestosis was found and the predominant fibers were crocidolite. It is suggested that the presence of amosite and crocidolite is linked to mesothelioma. The mean lengths of amosite and crocidolite, as detected by our resolution capabilities, were 36.0 and 20.9 pm, and the mean diameters were 0.51 and 0.27 pm, respectively. Both amosite and crocidolite fibers had high aspect ratios (94.2 and 115.4). 0 1992 Wiley-Liss, Inc. Key words: malignant mesothelioma,asbestos fibers, asbestos body, electron microprobe analysis, railroad workers
INTRODUCTION
Since Wagner et al. [1960] first pointed out the association of mesothelioma with crocidolite exposure in South Africa, numerous epidemiological and experimental studies have been carried out, and it is now no longer possible to deny the relationship. However, when confronted with a mesothelioma case, it is not always easy to establish the causative relationship of asbestos exposure to this tumor. And as the epidemiological and pathological studies have gradually disclosed that asbestos Department of Pathology, Faculty of Medicine, Toyama Medical and Pharmaceutical University, Toyama, Japan. Address reprint requests to Yoshihiro Murai, PhD, Department of Pathology, Faculty of Medicine, Toyama Medical and Pharmaceutical University, 2630 Sugitani, 930-01 Toyama, Japan. Accepted for publication December 20, 1991. 0 1992 Wiley-Liss, Inc.
194
Murai and Kitagawa
related lung cancer or mesothelioma can develop without asbestosis [Warnock and Isenberg, 19861, it becomes more difficult to evaluate the relationship of the exposure to the development of malignant tumor. On the other hand, because of the widespread use of asbestos in our societies, it is difficult in some cases to determine past asbestos exposure via history alone. Therefore, we have developed an evaluation chart to assess the causative relationship of asbestos exposure to lung cancer or mesothelioma by examining the asbestos body counts in the lung tissue [Kitagawa and Murai, 19881.
By means of animal inhalation experiments, Wagner et al. [ 19741reported that mesothelioma could be induced by all four types of asbestos examined. Many fibrous minerals other than asbestos, such as glass fiber, are capable of causing mesothelioma experimentally after intrapleural inoculation [Stanton et al., 19811; and recently, erionite, a fibrous silicate mineral, was found to be highly active in humans in inducing mesothelioma [Baris et al., 19791, which bore features since replicated in the mouse [Suzuki and Kohyama, 19841. Stanton and his colleagues reported in 1981 that the carcinogenicity of fibers in rats differed when length and diameter (aspect ratio) differed. Attention was therefore directed to the likelihood that the fibrous quality of asbestos, that is, its morphology, plays an important role in carcinogenic action. In the present study, the asbestos burden as to type and fiber sizes was determined in the lungs of mesothelioma cases, and these data are discussed as related to carcinogenicity. MATERIALS AND METHODS
The 27 mesothelioma samples examined in the present study consist of 20 cases from Toyama Medical and Pharmaceutical University Hospital (TMP) and its affiliated hospitals and 7 cases from medical consultation with one of us (M. K.). To verify the diagnosis and classification, the tumors were stained with hematoxylin-eosin and silver impregnation. Alcian-blue, colloid-iron stains, and periodic acid-Schiff reaction were used to evaluate interstitial mucin and cytoplasmic glycogen. Immunohistochemistry with keratin, epithelial membrane antigen, carcino-embryonic antigen, and vimentin were routinely applied. The clinical and pathological outlines of the cases are shown in Table I. The 30 control cases (male 12, female 18; mean age 64.7 years) belonging to the low level exposure group without lung cancer or mesothelioma were selected from consecutive autopsy material in which asbestos body counts had already been determined [Murai and Kitagawa, 19841. Asbestos body counts were carried out by a modified Smith and Naylor [1972] method [Matsuda, 19751, and asbestos fiber counts were by our modified Langer’s [1971] method. Asbestos Body Quantification Five gram individual samples of wet pulmonary parenchyma from the segments 3, 6, 8, and 10, and, whenever possible, from the opposite side of the lesion in the pleural mesothelioma cases, were minced and digested in 200 ml sodium hypochlorite for 48 to 72 hr. Twenty milliliters of chloroform : 50% ethanol were added and well shaken. The solution in the tubes was centrifuged for 10 min at 800 rpm. After the supernatant was discarded by aspiration, 5 ml of 95% ethanol were added to the bottom chloroform layer and the mixture filtered through a millipore filter (SM type,
58lM 62JM 69lM 72JM 58lM 53lM 63lM 72lF 56lM 50lM
60M
Age1 sex 60lM 59lM 45lM 71lM 74lM 44lF 35lM 621F 75lM 56/F 56lM 52JM 68lM 75lF 73lM 67lM
Ironworker Transmission line worker Ironworker SL mechanic' Mariner, carpenter JNR JNR Shipyard worker Housewife Ironworker Insulation worker
JNR
Occupation Fanner Printing type founder Cleaner JNRc (train dispatcher) Tatami maker Electronic part painter Construction worker Housewife Sake dealer Telephone operator HM distributofl Teacher Clerk Housewife stevedore
"Number of asbestos bodies per 5 gm wet lung tissue. %umber of asbestos fibers per 5 gm wet lung tissue. 'JNR, Japan National Railway. dHM, Household Medicine. 'SL, Steam Locomotive.
19 20 21 22 23 24 25 26 27
18
5 6 7 8 9 10 11 12 13 14 15 16 17
4
Case number 1 2 3
Site of tumor R. pleural L. pleural R. pleural R. pleural L. pleural L. pleural R. pleural L. pleural L. pleural Peritoneal R. pleural R. pleural R. pleural Peritoneal Peritoneal L. pleural R. pleural R. pleural R. pleural L. pleural L. pleural L. pleural R. pleural R. pleural R. pleural R. pleural Peritoneal Gross Localized Diffuse Diffuse Localized Localized Localized Localized Localized Localized Diffuse Localized Localized Localized Localized Diffuse Diffuse Diffuse Diffuse Diffuse Diffuse Diffuse Diffuse Diffuse Diffuse Diffuse Diffuse Diffuse
Cell type Sarcomatous Epithelial Desmoplastic Sarcomatous sarcomatous Epithelial Sarcomatous Sarcomatous Sarcomatous Epithelial Epithelial Epithelial Sarcomatous Sarcomatous Epithelial Epithelial Epithelial Epithelial Sarcomatous Biphasic Sarcomatous Epithelial Biphasic Epithelial Epithelial Sarcomatous Euithelial 6 8 8 10 10 13 16 31 41 58 173 264 365 525 549 686 875 1,064 2,256 3,390 17,300 2.720.000
4 4
1 2 3
(a)
2,640 6,380 5,280 880 2,420 6,400 5,900 20,000 30,000 12,000 22,000 7,800 27,000 76,000 8,400 55,000 19.000.000
660
(b) 440 220> 220> 220> 27,500 2,200 220 2,640 1,100
(b)/(a) 440.0 110.0> 73.3> 55.0> 6875.0 366.7 27.5 330.0 110.0 66.0 203.1 398.8 170.3 21.5 41.7 37.0 22.3 54.8 57.1 21.9 32.1 8.9 25.4 33.7 2.5 3.2 7.0
TABLE I. Clinical and Pathological Features of 27 Malignant Mesothelioma Cases (19 Autopsy, 8 Surgical) With Numbers of Asbestos Bodies and Asbestos Fibers Found in Lung Tissue
1%
Murai and Kitagawa
5 pm pore size). The filter was dried and dipped in xylene and mounted without staining. Four filters were prepared for each segment. For those cases in which asbestos body counts were >3,000,the solutions were diluted. Asbestos bodies were counted by searching all fields in the preparations with a light microscope at 100 x magnification. The average number of four segments was used as the number of asbestos bodies per 5 gm wet lung tissue for each case. The core fibers of the asbestos bodies we counted were all asbestos fibers as far as confirmed by energy dispersive electron microscopy [Murai and Kitagawa, 19841.From a study on 235 consecutive autopsy cases, three (3) levels of asbestos exposure were determined using asbestos body counts per 5 gm wet lung tissue: low level exposed group (140);moderate level exposed group (41-150); and high level exposed group (2151).In our previous report [Murai and Kitagawa, 19841,a count of 140 (separated into 1-10 and 11-40) corresponded to low level exposure and was considered to be due to ambient air pollution.
Asbestos Fiber Quantification For the preparation of asbestos fibers, 0.5 gm wet pulmonary parenchyma from segment 10 of the same lung as the asbestos body preparation was removed and dissolved in 5 ml of 40% potassium hydroxide; this was heated in boiling water for 1 hr. After the tissue had completely dissolved, 5 ml of distilled water was added to the tubes and the suspension centrifuged for 30 min at 2,000 rpm. The sediment was washed with 10 ml distilled water and centrifuged under the same conditions. After two cycles of washing, the suspension was diluted until fiber counting was possible. A total volume of 1 to 5 ml was prepared in almost all cases. A 25 p1 aliquot of the dilute was dropped onto a collodion-coated Maxtaform-grid (HF15) on a dental plate. This grid has printed symbols by means of which it is easy to recognize the location of the objectives. After drying, the grid was examined by transmission electron microscopy (JEOL 200-CX)at X 2,000magnification. All fibers with an aspect ratio 2 3 were photographed for size measurement. The photographed fibers were then analyzed by scanning electron microscopy equipped with an X-ray energy dispersive spectrometer (HITACHI X 650 KEVEX 7000 Q System) for the identification of fiber type. We selected a magnification of X2,000,as we intended to analyze as many fibers longer than 5 pm as possible. Fibers 2 5 pm in length and 2 0.10 pm in diameter were included in the counting procedure. Although with a magnification of ~2,000, fibers longer than 5 pm are detectable, most fibers shorter than 2-3 pm would not be seen. We searched all observable fields in the grids until the number of asbestos fibers exceeded 100. For the high level exposure group, from 100 to 141 (mean 108.5) asbestos fibers for each case were detected and analyzed in 3 to 74 grids (mean 25). The first 100 fibers were used to calculate the percentage of asbestos types and their sizes. The limit of detectability for the high level exposure group was less than 800 in all cases but one. Ten grids were examined in the low and moderate level exposure groups with the limit of detectability being 220 in 12 cases, 440 in one case, and 1,100in 2 cases. Two or four grids were examined in each of the controls and the limit of detectability was 1,100in all. To measure the spread of a 25 p1 drop, ten such drops of distilled water were placed on a dental plate and their diameters measured. The average was 4.24 mm,
+
Asbestos Fibers in Malignant Mesotheliomas
197
corresponding to an area of ~ ( 2 . 1 2 mm2. ) ~ Because the edge of the grid (3 mm diameter) was concealed by the holder of the electron microscope, the observable diameter of the grid was 2 mm. From an enlarged photograph of the grid, we calculated the active area to be 82% of the (2 mm diameter) observable area, or 0 . 8 2 ~ ( 1 )Thus, ~ . the ratio of the total area of a 25 p1 drop on the dental plate to that = 5.5. of the observable grid area is ~ ( 212)2/0.82~(1)2 . The number of asbestos fibers per 5 gm wet lung tissue was calculated by the following formula: Al25 pl x B/C x 5 gdO.5 gm x 5 . 5 , where A is the total suspension volume (pl), B is the total asbestos fiber count, C is the total number of grids, and 5.5 is the ratio of the total area of 25 p1 aliquot placed on the dental plate to the actually observable grid area. RESULTS
From the opening of our hospital (TMP) in October 1979 to December 1988, there were 1,677 autopsy cases over 20 years of age (male 1,040, female 637). Over the same period, there were 430 (male 290, female 140) in Toyama Prefectural Central Hospital (TPC), which is the affiliated teaching hospital of our university. Against this background, six pleural and two peritoneal mesotheliomas were encountered, a frequency of 0.38% of the adult autopsies (8/2107). Clinical and Pathological Outline and Grade of Asbestos Exposure of the 27 Malignant Mesothelioma Cases
Table I shows the 27 collected cases of mesothelioma, 8 from surgical and 19 from autopsy. Twenty-three of the cases were pleural and four peritoneal. The cases are listed in Table I in order of asbestos body counts: there were 22 males and 5 females with a mean age of 60.9 years. Macroscopically, there were 16 diffuse and 11 localized mesotheliomas. Microscopically, 13 were epithelial, 11 sarcomatous, 2 biphasic, and 1 desmoplastic. These 27 cases were divided into 3 groups according to our criteria for asbestos body counts, 13 belonging to the low level (540) exposure, 2 to the moderate (41-150) level exposure, and 12 to the high (1151) level exposure. Lung tissue from a 50 year-old man with peritoneal mesothelioma gave a count of 2.72 X lo6 asbestos bodies; his was the only case with demonstrable asbestosis. The patient had been an insulation worker for 11 years from age 20 to 31 years. The other patients in the high level category were almost certainly exposed to asbestos in their occupations. Four had worked in the former Japan National Railway and three were ironworkers. All mesotheliomas in the high level group had developed in a diffuse fashion, although this was also found in the low level group. Both biphasic mesotheliomas were in the high level group. Seven epithelial type mesotheliomas were seen in the 12 cases in the high level group. However, no clear-cut relationship between histological type and asbestos exposure was observed. The correlation between asbestos body counts seen at lOOX and asbestos fiber counts seen at 2,000 X is shown in Table I. Here, although the ratio of the body counts to the fiber counts is not always constant, both correlated with each other relatively well (r=0.858, p
