Fissural thickening and exposure to asbestos Occurrence determinants and functional impact.код для вставкиСкачать
American Journal of Industrial Medicine 20785-793 (1991) Fissural Thickening and Exposure to Asbestos: Occurrence, Determinants, and Functional Impact Jan Zejda, MD, and Pierre Ernst, M D The occurrence and determinants of thickened interlobar fissures were assessed in 147 insulators, in order to test the hypothesis that radiographic abnormalities in minor (FMN) andlor major (FMJ) interlobar fissures are associated with restrictive respiratory impairment, which may reflect early parenchymal lung fibrosis. According to the diagnostic criteria developed, definite interlobar thickening was found in 33 (FMN) and in 73 (FMJ) men, and was frequently associated with the presence of classical asbestos pleural disease, though not parenchymal abnormality. Multivariate analyses did not demonstrate any association between either FMN or FMJ and lung function. Such an association, however, was found in the case of the usual asbestos-related pleural change. These findings suggest that fissural thickening seen on chest radiography represents a component of asbestos pleural disease and not asbestosis. Key words: asbestos, chest radiography, pleural disease, lung function, ILO classification INTRODUCTION Benign asbestos-related pleural abnormalities may result in lung function impairment [Corris et al., 1988; Cotes and King, 1988; Jarvholm and Sanden, 1986; Oliver et al., 1988; Rosenstock et al., 19881. A restrictive ventilatory impairment is independently associated with roentgenographic evidence of pleural fibrosis [Bourbeau et al., 19901 although the possibility that such impairment may reflect, in part, interstitial lung fibrosis not visible on the chest roentgenographs cannot be unequivocally ruled out [Schwartz et al., 19901. Thickening of interlobar fissures is known to occur in workers with long duration of exposure to asbestos dust. The relationship of this abnormality to either pleural or parenchymal fibrosis is unclear [Solomon and Webster, 1976; Solomon et al., 1979; Solomon et al., 1979a; Rockoff et al., 19871. Kockoff et al.  proposed that fissural thickening might be a marker of early parenchymal lung fibrosis, i.e., asbestosis. If this were true one would expect that, on average, subjects with fissural thickening would have lower lung function Pulmonary Research Laboratory, Department of Epidemiology and Biostatistics, and School of Occupational Health, McGill University, Montreal, Quebec, Canada. Address reprint requests to Dr. Pierre Ernst, Pulmonary Research Laboratory, Department of Epidemiology and Biostatistics, McGill University, 1 1 10 Pine Avenue West, Montreal, Quebec H3A 1A3, Canada. Accepted for publication April 22, 1991. 0 1991 Wiley-Liss, Inc. 786 Zejda and Ernst than those without, after accounting for the effect of smoking and radiologically apparent parenchymal fibrosis and pleural thickening along the chest wall and diaphragm. The principal objective of the present study was to examine the relationship between fissural thickening visible on the chest radiograph and level of lung function. Our secondary objectives were to determine the prevalence and determinants of such fissural thickening in an asbestos-exposed workforce. MATERIAL AND METHODS Subjects The study group consisted of 147 construction insulators, members of Local 58 of the International Union of Frost and Heat Insulation and Asbestos Workers. The subjects were men, all in the trade at the time of examination. They were selected from a total of 644 members of Local 58, if they did not receive compensation for asbestosis, were 50 years old or less, and lived within the Montreal area. The eligible population consisted of 246 workers, 215 agreed to participate. The satisfactory and complete examination data were available in 147 (60%) of the eligible population. Lung Function Testing The assessment of lung function was based on the measurements of spirometric indices of ventilatory capacity and pulmonary diffusing capacity. Spirometry was performed using a Ohio 827 volume displacement spirometer. The best forced vital capacity (FVC) and forced expiratory volume in the first second (FEV,) were obtained according to the recommendations of the American Thoracic Society [ 19791, expressed in both absolute (measured) values and as percentage of the reference values (9% pred.) of Crapo et al. [1981a]. Diffusing capacity (DLCO) was measured by the single breath carbon monoxide method, using Collins testing equipment. The results were accepted for analysis if at least two measurements were reproducible within 10%. DLCO was calculated as an arithmetic mean of the best two reproducible measurements, and analyzed both as absolute value and % predicted, calculated from the reference equation of Crapo and Morris [ 19811. Radiographic Examination The standard posteroanterior (PA) chest radiographs were available in 163 subjects and lateral films in 158 subjects. They had been read previously by two certified B readers, who blindly assessed parenchymal and pleural abnormalities, according to the ILO International Classification of Radiographs of Pneumoconioses with the standard set of radiographs used for comparison [International Labour Office, 19801. Fissural changes were assessed subsequently by two different readers who did not know the results of the readings of parenchymal and peripheral pleural abnormalities. Parenchymal abnormalities were read according to the 12-point profusion scale described in the ILO classification, and this score (1-12) was used to calculate means. A profusion of 110 or greater was chosen to represent radiographically apparent parenchymal fibrosis. Asbestos-Related Fissural Thickening and Lung Function 787 Pleural abnormalities (excluding interlobar fissures) were expressed as a score obtained by summing up the numeric scores for (1) costophrenic angle, (2) diaphragmatic pleura, and ( 3 ) chest wall pleura as described previously by BCgin and colleagues [ 19841. Obliteration of one costophrenic angle was given 1 point, of two angles 2 points. Similarly, the thickening of diaphragmatic pleura was assigned 1 or 2 points, depending on whether it was visible on one or both sides. Chest wall scores were also computed for each side independently and then added together. Each (left or right) score included the assessmentface on, expressed as 1, 2, or 3 points for the respective extent of changes of 1, 2, or 3. This index was added to the assessment in proJle, expressed as a product of extent category (1,2, or 3) and width category (1, 2, or 3 substituted for the original notation of a, b, or c ) . The final score (1 2 3), with a possible range from 0 to 24 points, was used as a linear variable for multivariate analysis. Subjects were also divided into two groups on the basis of this score as follows: a value of 0 was accepted as category without pleural changes versus the category with any abnormalities (score 1 or greater). No attempt was made to distinguish between discrete pleural plaques and diffuse thickening. The validity of this score is suggested by the previously demonstrated relationship between this score and the degree of lung function impairment [Bourbeau et al., 19901. Thickening of interlobar pleura (fissural thickening) was assessed from the postero-anterior and lateral chest radiographs independently for the minor (horizontal) and major (oblique) fissures, without specifying whether the latter was located in the left or right lung. The following scale, modified from that proposed by Rockoff et al. , was agreed upon by the authors of this paper and used throughout the readings: + + 0 1 2 3 4 5 6 7 fissure not visible; normal (visible) fissure: width ‘‘hairline”; questionable thickening: width <0.5 mm; minimal localized thickening: width 0.5-1 .O mm; length of thickening < 1/2 total length of fissure; minimal diffuse thickening: width 0.5-1.0 mm; length of thickening >1/2 total length of fissure; moderate localized thickening: width >1.O mm; length of thickening <1/2 total length of thickening; moderate diffuse thickening: width >1.O mm; length of thickening > 1/2 total length of thickening; marked thickening. A value of 3 or greater on the above scale was used as a cut-off value for dichotomization of fissural changes into normal and abnormal classes. The linear regression analysis and assessment of between-reader variability were also performed with the above classification rescaled to 5 points instead of 8, in order to increase the numbers of subjects in higher classes. This was achieved by grouping together the categories of minimal localized thickening and minimal diffuse thickening and, separately, moderate localized thickening and moderate diffuse thickening. 788 Zejda and Ernst TABLE I. Between-Reader Agreement in Assessment of Radiographic Changes in Interlobar Fissures* No. of fissures read as Crude agreement K- (%) Value 0.499 Fissure Reader 0 1 2 3 4 Minor A 76 64 46 49 39 39 I 10 1 B 1 66.3 A B 13 29 57 40 78 63 10 24 0 2 55,7 Major 0.355 *Fissural thickening measured according to the modified scale: 0, nonvisible fissure; 1, visible but normal fissure or questionable thickening; 2, minimal thickening; 3, moderate thickening; 4, marked thickening. For details concerning modification of the scale, see under Methods. Analysis Statistical analysis was performed using SAS software [SAS Institute, Inc., 19881. Linear regression was used to analyze linear dependent variables and logistic regression for categorical outcomes, both with the level of statistical significance p<0.05. Between-reader variability was assessed by means of the crude percent agreement and the Kappa statistic [Musch et al., 19841. RESULTS Between-Reader Variability in Assessing Fissural Thickening In order to avoid arithmetical distortion of the Kappa statistics resulting from small numbers in some, particularly higher classes, the between-reader variability was assessed in relation to a 5-point instead of the original 8-point scale. Table I presents the number of subjects with different degrees of fissural thickening, as diagnosed by two readers, the crude agreement, and the Kappa values. According to a verbal scale [Musch et al., 19841, the between-reader agreement was “fair” (major fissures) and “good” (minor fissure). Descriptive Statistics The mean values and range of age, smoking status, duration of exposure, the results of lung function testing, and radiographic status are shown in Table 11. Univariate Analysis The differences between subjects with and without each category of chest radiographic abnormality considered alone are shown in Table 111. Small irregular opacities of category 1/0 or greater were found in 12 subjects. They were significantly older than those with normal parenchymal appearance on the chest radiograph (p=O.OOl), had more advanced pleural disease (p = 0.02), but in terms of duration of exposure, smoking habit, fissural changes, and lung function, subjects with small irregular opacities appeared to be no different from those without. Classical pleural abnormalities (thickening of the chest wall pleura and/or diaphragmatic pleura and/or obliteration of the costophrenic angle) were found in 58 subjects. In comparison with those without abnormalities, the men with peripheral pleural changes were significantly older (p = O.OOOl), had longer exposure (p = 0.01), Asbestos-Related Fissural Thickening and Lung Function 789 TABLE 11. Age, Smoking Status, Duration of Exposure to Asbestos and Lung Function in 147 Asbestos-Exposed Insulators Age (years) Amount of smoking (pack-years) Duration of exposure (in thousand hours) Fvc 1 % pred FEV, 1 % pred DLCO ml/min/mmHg % pred Mean SD Range 36.9 15.7 28.2 4.78 100.7 3.85 103.7 28.6 82.1 8.2 12.1 40.36 0.77 11.2 0.67 12.9 5.3 12.8 18-50 0-58 0.37-453.9 2.83-7.35 63.9-135.0 1.97-5.95 61.6-140.7 14.7-41.3 45.3-124.5 TABLE 111. Unadjusted Comparisons of Asbestos-Exposed Subjects According to TvDe of Abnormalitv* Age Category SIO + SIOWALL+ WALLFMN FMN- + FMJ+ FMJ- N 12 135 58 89 33 114 73 74 (yr) 43.4 36.4 41.9 33.6 40.1 36.0 39.6 34.3 Radiography scorea Percent predicted Exposure (h) SIO WALL FMN FMJ FVC FEV, DLCO 38,288 27,347 - 37,139 22,410 47,514 22,486 38,405 18,010 0.4 0.2 2.5 1.o 2.3 0.8 1.6 0.7 1.5 1.5 1.9 1.2 2.7 2.7 3.3 2.3 - - - - 97 101 96 103 96 101 98 102 99 104 100 105 99 105 102 104 80 82 84 80 80 83 80 84 0.2 0.3 0.3 0.2 - - - * S O , small irregular opacities; WALL, thickening of chest wall pleura; FMN, thickening of minor fissure; FMJ, thickening of major fissure. aSIO+ if profusion 2 1/0; otherwise SIO-; WALL+ if WALL>O, otherwise WALL-; FMN+ or FMJ if FMN>2 or FMJ>2. otherwise FMN- or FMJ-. + higher scores for fissural thickening (minor: p = 0.01, major: p = O.OOOl), and lower mean FVC (p=0.0005) and FEV, (p=0.04). In order to simplify comparisons, the score for interlobar fissures was calculated as the arithmetic mean of the scores reported by two readers. Thickening of the minor fissure, defined by the score of 3 or more on the 8 point scale, was observed in 33 men. They were significantly older (p = 0.009), had more advanced thickening of the chest wall pleura (p=0.02), lower FVC (p=0.02) and FEV, (p=O.Ol) than the remaining subjects. Thickening of the major fissure, observed in 73 subjects, was associated with older age (p = O.OOOl), significantly longer duration of exposure (p = 0.003), higher score for pleural disease (p = 0.01) and diminished FVC (p=O.O49). Note that subjects with or without fissural thickening had a similar prevalence of small irregular opacities. Multivariate Analysis This analysis was carried out to determine whether the lower lung function seen in subjects with fissural changes was actually due to the associated greater prevalence 790 Zejda and Ernst TABLE IV. Linear Regression Models With Lung Function as Dependent Variable and Parenchymal Changes (SIO), Peripheral (WALL), and Fissural (FISS) Pleural Thickening as Independent Variables* Covariates Independent variables Age (yr) Height (cm) Smoking (pack-years) SIO (score: 0-15) WALL (score: 0-10) FVC Standard error Probability -0.032 0.006 0.060 0.007 c -0.019 0.046 NS -0.059 0.020 c -0.003 0.002 NS FEV, Standard error Probability -0.033 0.005 0.042 0.006 -0.005 0.002 c c a -0.023 0.041 NS DLCO Standard error Probability -0.086 0.129 NS 0.251 0.115 -0.007 0.033 NS -0.357 0.556 NS Dependent variabie a FISS (score: 0-8) Intercept -0.019 0.019 NS -4.103 1.300 -0.044 0.018 -0.015 0.017 NS - 1.940 1.155 NS -0.325 0.262 NS -0.418 0.279 NS -8.702 20.717 NS b b *NS, not significant. "p<0.05. hp<O.Ol. 'p<O.OOl. of pleural plaques and costophrenic angle blunting in these subjects. The contribution of fissural thickening to the observed FEV,, FVC, and DLCO was examined by multiple regression analysis after adjusting for age, height, smoking in pack-years, the grade of small irregular opacities and the degree of pleural thickening. Duration of exposure in the insulating trade was not retained in the models since it did not contribute significantly and its removal did not affect the regression coefficients for the other variables. In the models shown in Table IV, fissural thickening (FISS) represents the sum of the scores for thickening of the major (RMJ) and minor (RMN) fissures. There was no significant relationship between any of the lung function parameters examined and thickening of fissures whether these were considered together or separately (the latter models are not shown). In the models shown, the mean of the scores obtained from the two readers is utilized. The results were similar using either score alone. The only category of radiographic abnormality significantly associated with a decrement in lung function was peripheral pleural thickening which caused a decrease in FEV, and FVC but not DLCO. In order to avoid the masking of an effect by inclusion of nonsignificant fissural changes, the analyses were repeated in subjects with a more obvious thickening of minor fissure and the major fissure (score 3 or more) for each fissure. In neither subgroup did any radiographic variables other than peripheral pleural thickening significantly contribute to the regression model after controlling for age, height and smoking. DISCUSSION In healthy individuals, interlobar fissures are either not visible on standard chest radiographs or are visible but of hairline thickness [Fraser et al., 19881. The scale designed by Rockoff et al. , which was adopted for the purpose of this study used this definition to distinguish normal from thickened interlobar fissures. The Asbestos-Related Fissural Thickening and Lung Function 791 between-reader agreement was better in relation to minor than to major fissures, This can be attributed to the greater difficulty in identifying and quantifying abnormality of the major fissures that may be superimposed as well as have varying degrees of thickening along their considerable length, which is difficult to describe by a summary score. The degree of agreement (crude 55.7-66.3%; K-value 0.355-0.499) is difficult to interpret, since there are no data for comparison, except for the crude agreement of dichotomous classification of changes in the minor fissure (36-46%) reported by Solomon et al. . The obtained level of agreement, however, compares favorably with between-reader variability obtained in assessing the classic signs of asbestos pleural disease [Solomon et al. , 1979; Bourbeau and Ernst, 1988; Robins and Green, 19881. The complete validation of the scale would also require the measurement of within-reader variability and comparison of the results obtained by different teams using the same diagnostic criteria. This is, however, impossible without standardized radiographs demonstrating various degrees of fissural thickening. On the other hand, the assessment of classical pleural changes, characterized by similar variability, has been widely used to investigate the impact of asbsetos pleural disease on lung function. Measurement of thickening of only 0.5 mm on the chest radiograph may be difficult. However, the scale used provided significant correlations between fissural and peripheral changes read by two different readers on different occasions and correlated with lung function. Therefore, the preliminary results concerning the diagnostic feasibility of assessing thickenings of interlobar fissures seems to justify addressing the main objectives of the study. Thickened minor fissures were reported in 25% (reader A) or 30.7% (reader B) of subjects. The respective prevalences for major fissures were 53.9% and 54.6%. Since both fissures were scored using the same numerical scale it cannot be excluded that higher numbers of definite changes in major fissures resulted from the magnification factor inherent in a lateral chest projection. Using a similar scale, Rockoff et al.  found definite thickening in 54.5% of 220 active and retired asbestos workers. In a large epidemiologic study reported by Solomon et al.  in 1,962 asbestos miners, thickening of the minor fissure, as assessed from the PA chest radiograph, was present in 5.1% to 9.4%, depending on the reader. The chest radiographs were read against a “standard” radiograph, showing evident thickening of the minor interlobar fissure. The clearly lower prevalence in asbestos miners in comparison with the results of our study or those of Rockoff et al.  is likely attributed to the different definitions of fissural thickening employed in the studies. On the other hand, it could be due to differences in occupational exposure. The prevalence of classical pleural abnormalities varies from 1.5% in asbestos miners [Enarson et al., 19881, 14.6% in textile workers [Berry et al., 19791 to 17.5-29.4% in plumbers [Rosenstock et al., 19881, and 43.7% in insulators [Lilis et al., 19861. The similarity of the prevalences reported by Rockoff et al. and ourselves must be interpreted with caution, given the different age range of the two populations (our subjects were younger) and their report of a prevalence for both fissures combined only. It does appear, however, that thickening of interlobar fissures is not infrequently observed in asbestos-exposed subjects, particularly thickening of the major fissure which, when assessed from both PA and lateral chest radiographs, may occur in up to 50% of currently employed subjects. 792 Zejda and Ernst Subjects with thickening of interlobar pleura were on average older and had been exposed for approximately twice as long as those without fissural thickening. The two groups did not differ in terms of the presence of parenchymal abnormalities. Subjects with fissural thickening also had more chest wall pleural abnormalities (and vice versa), and these had similar determinants. The close association between interlobar and peripheral thickening of the pleura suggests that they belong to the same category of asbestos pleural disease. That fissural thickening can occur without radiographic changes in the lung parenchyma has been known since the first case reports describing this abnormality in asbestos-exposed subjects [Solomon and Webster, 19761 and was confirmed in the previously cited epidemiologic study in asbestos miners and the case series in asbestos workers. Thus, the present study, as well as the little information available from the literature, suggest that the thickening of interlobar pleura can develop independent of radiographic signs of asbestosis and seems to be as readily detected as the classical appearance of asbestos pleural disease. Computed tomography (CT) scanning could add useful information about the nature of these fissural changes. Fissural thickening has been proposed as an early manifestation of parenchymal fibrosis due to asbestos exposure. If this were the case one would expect subjects with such radiographic changes to have, on average, some deficit in lung function as compared to other exposed subjects without fissural changes or more obvious radiographic evidence of asbestosis. To test this hypothesis, we examined spirometric lung volumes and single-breath diffusing capacity in a workforce with approximately 15-35 years of asbestos exposure but in an age range in which frank asbestosis is unusual today. When examining the relationship between lung function impairment and fissural thickening in isolation by a univariate analysis, subjects with this abnormality had, on average, a lower FVC. However, this association could be completely accounted for by the lung function deficit that resulted from the accompanying classic asbestos pleural disease, as demonstrated in a multivariate analysis. CONCLUSION It appears that fissural thickening is a not infrequent radiographic finding in subjects exposed to asbestos and represents a component of asbestos pleural disease and not asbestosis. Furthermore, any effect on lung function can be explained by the associated pleural disease and fissural thickening does not therefore appear to have an independent effect on spirometric lung volumes or diffusing capacity. ACKNOWLEDGMENTS We would like to thank the members of Local 58 of the International Association of Heat and Frost Insulators and Asbestos Workers. This work was supported by a grant from the Environmental Research Fund of the School of Occupational Health. Dr. Jan Zejda is in receipt of The Romer Foundation Scholarship. Dr. Pierre Ernst is a Chercheur-Boursier, Fonds de Recherche en Santk du Quebec (FRSQ). Asbestos-Related Fissural Thickening and Lung Function 793 REFERENCES American Thoracic Society (1979): Snowbird workshop on standardization of spirometry. Am Rev Respir Dis 119:881-888. Btgin R, Boctor M, Bergeron D, Cantin A, Berthiaume Y, PCloquin S, Bisson G, Lamoureux G. (1984): Radiographic assessment of pleuropulmonary disease in asbestos workers: Postero-anterior, four view films and computed tomograms of the thorax. Br J Ind Med 41:373-383. Berry G, Gilson JC, Holmes S, Lewinsohn HC, Roach SA (1979): Asbestosis: A study of dose-response relationships in an asbestos textile factory. Br J Ind Med 36:98-112. Bisson G, Drapeau G, Lamoureux G, Cantin A, Rola-Pleszczynski M, BCgin R (1983): Computer based quantitative analysis of Ga-67 uptake in normal and diseased lungs. Chest 84513-518. Bourbeau J, Ernst P, Armstrong B, Chrome J, Becklake MR (1990): The relationship between respiratory impairment and asbestos-related pleural abnormality in an active workforce. Am Rev Respir Dis 1421837-842. Bourbeau J, Ernst P (1988): Between- and within-reader variability in the assessment of pleural abnormality using the ILO 1980 International Classification of Pneumoconioses. Am J Ind Med 14: 537-543. Coms PA, Best JJK, Gibson GJ (1988): Effects of diffuse pleural thickening on respiratory mechanics. Eur Respir J 1:248-252. Cotes JE, King B (1988): Relationship of lung function to radiographic reading (ILO) in patients with asbestos related lung disease. Thorax 43:777-783. Crapo RO, Morris AH (1981): Standardized single breath normal values for carbon monoxide diffusing capacity. Am Rev Respir Dis 123:185-189. Crapo RO, Morris AH, Gardner RM (1981a): Reference spirometric values using techniques and equipment that meet ATS recommendations. Am Rev Respir Dis 123:185-189. Enarson DA, Embree V, MacLean L, Grzybowski S (1988): Respiratory health in chrysotile asbestos miners in British Columbia: A longitudinal study. Br J Ind Med 45:459-463. Fraser RG, Pare JAP, Pare PD, Fraser RS, Genereux GP (1988): “Diagnosis of Diseases of the Chest.” 3rd Ed. Philadelphia: W.B. Saunders Company, pp 663-685. International Labour Office (1980): “ILO/UC International Classification of Radiographs of Pneumoconioses 1980.” Occupational Safety and Health Series No. 22, revised. Geneva: ILO. Jarvholm B, Sanden A (1986): Pleural plaques and respiratory function. Am J Ind Med 10:419-426. Lilis R, Selikoff IJ, Lerman Y, Seidman H, Gelb SK (1986): Asbestosis: interstitial pulmonary fibrosis and pleural fibrosis in a cohort of asbestos insulation workers: Influence of cigarette smoking. Am J Ind Med 10:459-470. Musch DC, Landis R, Higgins I n , Gilson JC, Jones RN (1984): An application of kappa-type analyses to interobserver variation in classifying chest radiographs for pneumoconiosis. Stat Med 3:73 -83. Oliver LC, Eisen EA, Greene R, Sprince NL (1988): Asbestos-related plaques and function. Am J Ind Med 14:639-656. Robins TG, Green MA (1988): Respiratory morbidity in workers exposed to asbestos in the primary manufacture of building materials. Am J Ind Med 14:443-448. Rockoff SD, Kagan E, Schwartz A, Kriebel D, Hix W, Rohatgi P (1987): Visceral pleural thickening in asbestos exposure: The occurrence and implication of thickened interlobar fissures. J Thorac Imag 2:58-66. Rosenstock L, Barnhart S, Heyer NJ, Pierson DJ, Hudson LD (1988): The relation among pulmonary function, chest roentgenographic abnormalities, and smoking status in an asbestos exposed cohort. Am Rev Respir Dis 138:272-277. Schwartz DA, Fuortes LJ, Galvin JR, Burmeister LF, Schmidt LE, Lestikow BN, Lamarte FP (1990): Asbestos-induced pleural fibrosis and impaired lung function. Am Rev Respir Dis 141:321-326. Solomon A, Webster I (1976): The visceral pleura in asbestosis. Environ Res 11:128-134. Solomon A, Irwig LM, Sluis-Cremer GK, Glyn Thomas R, Du Toit RSJ (1979): Thickening of pulmonary interlobar fissures: Exposure-response relationship in crocidolite and amosite miners. Br J Ind Med 36:195-198. Solomon A, Sluis-Cremer GK, Goldstein B (1979): Visceral pleural plaque formation in asbestosis. Environ Res 19:258-264.