INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Pleural plaques have a high specificity for past asbestos exposure, especially when they have the typical radiological characteristics of a sharp, elevated opacity of the parietal pleura with a surface of several square centimeters and a thickness varying from several millimeters to 1 cm. Pleural plaques have been defined and described by the American Thoracic Society (1) and institutions such as the International Labour Office (ILO) (2) or the Pneumoconiosis Committee of the College of American Pathologists and the National Institute for Occupational Safety and Health (NIOSH) (3).

Localized (or circumscribed) pleural plaques are by far the most common asbestos-related disorder. It is believed that in the general population of areas where there are no endemic plaques as a result of particular geologic characteristics, 80- 90% of strictly defined pleural plaques discovered at chest X-rays are due to occupational exposure to asbestos (4). One point that is not clear in the literature relates to the relationship between the size of the plaques and the intensity of past asbestos exposure. It is well established that the prevalence of plaques increases with increasing time since first exposure, but it is not known whether in a group with similar times since first exposure, larger plaques are due to higher cumulative exposures. In other words, is it possible to infer the severity of past exposure from the size of the plaques?

The aim of this study was to examine not so much the prevalence but mainly the size of pleural plaques in a cohort of workers from an asbestos-cement factory, who had been exposed to moderate up to high concentrations of asbestos in the 1960s and 1970s. We attempted to relate the measured size of the pleural plaques to the estimated past asbestos exposure. In addition, we also investigated the possible relation between plaque size and pulmonary function, because this is also a somewhat controversial issue in the literature.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Computed tomography (CT) scans were performed between June 1991 and May 1994 in male workers from an asbestos-cement factory born between 1945 and 1954, hired between 1965 and 1969, and with at least 2 yr of employment. The control group consisted of 21 male workers of similar age from the cleaning or catering departments of the University Hospitals, Leuven, Belgium. The protocol of this study was approved by our institutional Ethics Committee and the factory's Committee of Safety and Health.

Jobs were divided into four categories with increasing dust concentration: offices outside the factory halls, stocking and production of sheets or pipes, product finishing, and handling of raw asbestos. From 1970 until 1985 most fiber measurements were obtained by static sampling during peak installation activities to evaluate dust sources or the necessity or the effectiveness of dedusting equipment. From 1985 on, many fiber counts were obtained by personal monitoring and these are more representative of 8-h time-weighted personal exposures. For each category, the mean asbestos concentration was estimated (5) on the basis of all available measurements (Figure 1), thus allowing a calculation of cumulative exposure, expressed as fiber-years/ml, for each worker (Figure 2A). Exposure during the last 5 yr was ignored, because these recent, much lower exposures had little or no relevance for the presence or extent of pleural plaques. Their inclusion in the analysis does not affect the results.

View larger version (19K): [in this window] [in a new window]   Figure 1.   Estimate of asbestos exposure in three job categories. The mean asbestos fiber concentration in year y (Cy) was calculated by fitting the available measurements to the descriptive formula: Cy = C + (C0  C)/1 + ek(y  z), where C0 and C are asymptotic fiber concentrations at times 0 and , respectively, of two exponential curves with their inflection points when year = z. Three exposure categories were considered: (1) production and storage of asbestos-cement sheets or pipes (molding, pressing, drying, demolding, slate punching), (2) finishing of asbestos-cement products (sawing, drilling, filing, sanding), and (3) handling of raw asbestos material (debagging and preparation of asbestos-cement mix).

View larger version (21K): [in this window] [in a new window]   Figure 2.   (A) Individual cumulative exposure was calculated for each worker by taking the sum ( Cydy) of all calendar period-specific exposures, as estimated for each job category occupied; the shaded area illustrates calculation procedure for a subject who worked in different jobs between times 0 and t. (B) Frequency distribution of the estimated cumulative exposure to asbestos in fiber-years/ ml after exclusion of the last 5 yr of exposure.

Vital capacity (VC), forced expiratory volume in 1 s (FEV₁), and maximal flows at 75, 50, and 25% of the forced vital capacity (FVC) were measured with a Morgan TT. Autor-Lind apparatus (P. K. Morgan, Kent, UK) according to current guidelines (6). The transfer factor for carbon monoxide (TL_CO) was measured by the single-breath method. Predicted values were those of Quanjer and coworkers (6).

Chest radiographs were read independently by three readers according to the ILO 1980 classification (7), without knowledge of exposure status. CT scanning was performed on a General Electric 9800 Highlight Advantage (General Electric, Milwaukee, MI). The first 21 subjects were examined, supine, at suspended maximum inspiration with 10-mm-thick CT sections at 10-mm intervals from the lung bases to the apices (140 kV per setting [kVp], 170 mA · s, 2-s acquisition). They then underwent a high-resolution CT with 1.5-mm-thick slices at 30-mm intervals during apnea after deep inspiration in prone and supine positions (102 kVp, 200 mA · s, 2-s acquisition). In the other 73 individuals (52 asbestos-exposed and 21 control subjects) the thick-slice CTs were not done and high-resolution CT scans (1.5-mm-thick slices) were obtained at 10-mm intervals.

CT scan images were visualized on a computer screen using a video camera equipped with a high-quality photo objective. Using the Leica (Wetzlar, Germany) Quantimet system, all visible pleural plaques were overdrawn by the computer mouse and their length was measured on the basis of a 5-cm calibration bar present on each CT picture. The total surface area of the plaques was calculated by multiplying the sum of all the lengths by 1 cm, because the interval of the CT slices was 1 cm.

Data Analysis

Statistical analysis was performed by Student t test, Spearman rank correlation, or multiple regression, as indicated, using StatView for Windows, version 4.53 (Abacus Concepts, Berkeley, CA).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Population

The study cohort consisted of 88 employees, of whom 9 had left the factory before the start of the study (3 of whom had plaques on chest X-ray), 4 were considered mentally incapable of undergoing the CT scan, and 2 refused to undergo these investigations (none of the latter 6 subjects had plaques on chest X-ray). Thus the final study group was composed of 73 exposed employees. The control group consisted of 21 subjects. Table 1 shows the main characteristics of both groups. Sixty-two (85%) of the asbestos-cement workers were smokers or ex-smokers with a mean cumulative smoking history of 10.9 (± 10.4) pack-years (mean ± SD). The control group comprised 16 (76%) smokers or ex-smokers with a mean of 13.4 (± 9.3) pack-years.

Cumulative Asbestos Exposure

Estimated values of exposure until 5 yr before the study ranged from 16.4 to 98.7 fiber-years/ml with a mean of 26.3 ± 12.6 fiber-years/ml (SD). Figure 2B shows the frequency distribution of the cumulative exposure in fiber-years/ml, after exclusion of the last 5 yr of exposure.

Detection of Pleural Plaques

None of the asbestos workers in this cohort demonstrated clear signs of asbestosis and none of the ILO profusion scores was above 1/0. However, by conventional chest X-ray, pleural plaques were seen in 19 (26%) of the asbestos-exposed workers, and no pleural plaques were seen in any of the 21 control subjects. By CT scan, localized pleural plaques were diagnosed in 51 (70%) of the asbestos-exposed employees, while again no pleural plaques were seen in the 21 control subjects.

Seventeen subjects were diagnosed as having pleural plaques on CT scan as well as on conventional chest X-ray, whereas in 34 subjects plaques were seen only on CT scan. The sensitivity of the chest X-ray was, therefore, 33%. On the other hand, pleural plaques were "seen" on chest X-ray but not on the CT scan in two subjects, thus meaning that the specificity of diagnosing pleural plaques by chest X-ray was 91%.

Evaluation of the Extent of the Pleural Plaques

With our method of measuring the length of the pleural plaques directly on CT scan images the mean pleural plaque surface of the exposed group was 43.0 ± 56.8 (SD) cm² for costal plaques only (range, 0-259.3 cm²). After adding diaphragmatic and mediastinal plaques the surface increased to 47.9 ± 61.7 cm² (SD) (median, 22.1 cm²; range, 0-278.4 cm²).

An alternative method for estimating pleural plaque surface is that of Al Jarad and coworkers (8), which uses a four-grade classifying system based on semiquantitative scores (0 to 3) related to the thickness and surface area of the costal pleural plaques. When applying this method to the cohort of this study, the asbestos-cement workers were attributed either grade 0 (negative) or grade 1 (end score 1-6: "minor pleural thickening") with a mean end score of 0.87 ± 0.85 (SD) (range, 0-3.67). None of the asbestos-cement workers reached grade 2 (end score 6-12: "moderate pleural disease"), let alone grade 3 (end score > 12: "advanced pleural disease"). The Spearman rank correlation coefficient between the Al Jarad semiquantitative method and our quantitative measurement results is 0.957, when considering only the costal pleural plaques (the Al Jarad method disregards diaphragmatic and mediastinal plaques).

Correlation between Measured Plaque Surface and Cumulative Asbestos Exposure

Although there is a small tendency for an increase in total plaque surface area with increasing cumulated asbestos exposure (Figure 3), this is not significant (p = 0.24, Spearman rank correlation). When the subjects were dichotomized into those with a cumulative exposure of less or more than 25 fiber-years/ml, the following figures were obtained: 46.2 ± 61.4 cm² (median, 16.8 cm²) for exposures of less than 25 fiber-years/ml and 50.6 ± 63.3 cm² (median, 34.9 cm²) for exposures greater than 25 fiber-years/ml. In the 51 subjects with pleural plaques no correlation was found by multiple regression analysis between pleural plaque surface (after square root transformation to obtain a normal distribution) and cumulative asbestos exposure (r = 0.006, p = 0.94), pack-years (r = 0.177, p = 0.37), or time since first exposure (r = 0.663, p = 0.15). It should be noted that with regard to time since first exposure, the design of the study led to a maximal difference between the subjects of 5 yr.

View larger version (13K): [in this window] [in a new window]   Figure 3.   Relation between calculated pleural plaque surface (cm2) based on direct length measurement on CT scan and cumulative asbestos exposure in 73 asbestos-exposed subjects.

Lung Function Testing

There was no significant difference in lung function test results between the control subjects and asbestos-exposed workers (Table 2). The 51 subjects with pleural plaques did not have a significantly different lung function when compared to the group without pleural plaques (Table 3). There was no significant correlation between lung function test results and the surface of pleural plaques according to multiple regression analysis when taking into account the covariables pack-years and fiber-years (Table 4). Smoking significantly affected FEV₁ in the expected direction.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Summary of Findings

In a group of 73 asbestos-exposed subjects with a cumulative asbestos exposure ranging from 16.4 to 98.7 fiber-years/ml (mean, 26.3 fiber-years/ml), localized pleural plaques were detected by CT in 70% of the subjects. The average calculated plaque surface was 47.9 ± 61.7 cm² (median, 22.1 cm²; range, 0 to 278.4 cm²). To put this in perspective, the total surface area of the parietal pleura, including the costophrenic recesses, amounts to approximately 2,000 cm² in a 70-kg man (9). There was no relation between the measured plaque surface and cumulative asbestos exposure in this group of workers with moderate asbestos exposure. In these workers in whom neither radiological asbestosis nor diffuse pleural thickening was present, neither the presence nor the extent of these moderate plaques was correlated with lung function parameters.

Exposure Characterization

The lack of accurate information with regard to previous asbestos exposure is a common problem in occupational medicine, and epidemiologists have often had to consider mainly the exposure duration as an exposure index because this is the only accurately quantifiable factor. In our study, we were able to consider also the intensity of exposure because a series of measurement results were available and because intensity varied so widely from one job category to another. However, this is also liable to error. Before 1978, asbestos concentrations in this factory were measured less frequently and systematically than in later years. These figures were mostly obtained by static measurements near the exposure source and at moments of peak installation activity. The representivity of these pre-1978 measurements versus the 8-h time-weighted personal exposure of the employees is, therefore, somewhat questionable and lumping together these short-term static measurements with the more precise time-weighted personal monitoring results of later years may entail considerable overestimations. Nevertheless, all measurement results point to large differences between different job categories as well as a considerable decrease in exposure intensity by one to two orders of magnitude over the years. Therefore, historical periods and job categories prove to be much more influential factors than the exposure duration per se. Estimating cumulative exposure on the basis of these factors and the available information on respective exposure intensities, even if inaccurate, seemed preferable to using exposure duration only.

Many asbestos workers in this cohort had been exposed to chrysotile as well as amphibole types of asbestos. Therefore, this study is not informative with regard to any possible differences between asbestos fiber types and their respective potential to cause plaques.

Detection and Surface Measurement of Pleural Plaques

In the investigated cohort, circumscribed pleural plaques were detected by CT in 51 (70%) of asbestos-exposed subjects and in none of the control subjects. CT is increasingly used in the evaluation of asbestos-exposed individuals (10). In the first 21 employees conventional CT scans (10-mm-thick CT sections at 10-mm intervals from the lung bases to the apices) were immediately followed by a high-resolution CT (HRCT) with 1.5-mm thick slices at 30-mm intervals. Conventional CT is recognized as a relatively high-dose diagnostic procedure, whereas HRCT, consisting of 1- to 2-mm-thick sections performed at 10-mm intervals, has an effective radiation dose lower than that of conventional CT, even with high-dose techniques (400 mA · s) (10). Therefore, we decided to switch to high-resolution CT scans only (with 1.5-mm-thick slices at 10-mm intervals) for the other 73 individuals (52 asbestos- exposed and 21 control subjects).

Furthermore, the study by Aberle and coworkers (11) compared HRCT and conventional CT in a significant number of asymptomatic individuals, showing that pleural changes were observed even more frequently on HRCT than on CT. In contrast, Gevenois and coworkers (12) did a similar study in 1994 and their findings were consistent with the higher sensitivity of HRCT than conventional CT for the detection of parenchymal disease but not for the detection of slight pleural abnormalities. This can be explained, however, by the fact that minimal pleural plaques can be located in the skipped area and such lesions may, therefore, not be included in HRCT. We tried to avoid this by taking an image every 10 mm instead of every 30 mm, as was done in the studies previously mentioned. In any case, in our study both imaging systems resulted in a transverse image at every centimeter from lung base to lung top in all the employees who participated in the study: in the first 21 employees, we had continuous slices of 1-cm thickness, and in the second group (n = 73) we took thin slices (1.5 mm) every 1 cm.

Pleural plaques were detected by conventional chest X-ray in only 19 (26%) of the asbestos-exposed workers, confirming that the sensitivity of conventional chest X-ray is low when compared with CT scanning. This sensitivity depends, of course, on the degree of past occupational exposure to asbestos. In the examined cohort, with a mean cumulative exposure of 26 fiber-years/ml, conventional chest X-ray detected pleural plaques in three times fewer subjects than CT scanning. This is comparable with the findings of other investigators. In a review, Järvholm and coworkers (13) concluded that the sensitivity for the detection of pleural plaques by conventional chest X-ray was always less than 50% compared with CT scan. With conventional chest X-ray only the more extensive and thicker plaques are diagnosed and costovertebral plaques are nearly always invisible on standard anteroposterior chest X-rays.

Our study is the first, to our knowledge, to measure the surface of plaques. We measured the total plaque surface areas with and without the diaphragmatic and mediastinal plaques, assuming that diaphragmatic and mediastinal plaques might increase sensitivity when looking for an effect of pleural plaques on lung function. It proved not to make much of a difference. Moreover, detecting and measuring diaphragmatic plaques on the CT scan slices was problematic because of their transverse orientation as they lie on the diaphragm.

According to the semiquantitative method of Al Jarad and coworkers (8), the employees of our cohort would be categorized in the lowest categories: 0 (no pleural thickening) or 1 (minimal pleural thickening) (see RESULTS). When comparing this semiquantitative method with our actual surface measurements, there is good correlation between the two methods. This means that our time-consuming and sophisticated quantitative method can, in practice, safely be replaced by the semiquantitative method. Nevertheless, our quantitative method represents a more detailed method for detecting possible influences on lung function or differences in the extent of pleural lesions due to cumulative asbestos exposure and it can be used for further research purposes.

Relation of Plaque Size to Past Asbestos Exposure

A possible reason why we did not find a significant correlation between the size of the pleural plaques and the cumulative asbestos exposure might be that the variations in both the size of the plaques and the intensity of exposure were rather limited. The lack of correlation between the surface of pleural plaques and latency since first exposure to the moment of the investigation is explained by the fact that, by design, the range of latencies was kept narrow with all subjects having started work within a defined period of 5 yr. It is likely that wider variations in the time since first exposure would have confirmed the influence of this well-known factor, as found by several authors (14), but this was not the purpose of our study. We had intentionally decided to reduce the influence of this factor in order to look at the effect of cumulative asbestos exposure.

Lung function tests of asbestos-cement workers were comparable to those of the control subjects. No cases of asbestosis were detected in this study. This is in agreement with the existence of a threshold for clinical or radiological asbestosis that has been estimated at 25 fiber-years/ml (20, 21), although it must also be recognized that our population was relatively young (all subjects being less than 50 yr old) and, hence, the possibility cannot be excluded that pulmonary fibrosis might develop in some subjects. However, when using this threshold as an exposure index to divide the study group in two subgroups, we found that those with an exposure of more than 25 fiber-years/ml had a mean plaque surface of 50.6 ± 63.6 cm² (median, 34.9 cm²), whereas exposure of less than 25 fiber-years/ml resulted in a mean plaque surface of 46.2 ± 61.4 cm² (median, 16.8 cm²), the difference being nonsignificant.

Relation of Plaque Size to Pulmonary Function

There was no significant correlation between either the presence or the extent of pleural plaques and the lung function parameters. These findings are in agreement with the opinion that moderate pleural plaques, in the absence of other asbestos-related disorders, have little or no effect on lung function tests (1, 22). It is unlikely that any significant healthy worker effect was the cause for not finding a correlation between pleural plaques and lung function. For most of the dropouts, recent conventional chest-X-rays and lung function tests were available and there was no evidence of the existence of more respiratory disorders in the dropouts than in the studied subjects.

In some studies (27, 28) it has been claimed that pleural plaques are associated with some lung function decrease. However, most studies failed to consider the magnitude of exposure (and, hence, the probability of subradiological asbestosis) and to exclude the presence of other asbestos-related diseases such as diffuse pleural thickening or asbestosis. It is well documented that (subradiological) asbestosis or diffuse pleural thickening can lead to significant lung function impairment (22, 23, 29), particularly when the costophrenic angle is obliterated (29, 32). As pointed out by Bégin (33), inclusion of some diffuse pleural thickening cases in a group of subjects with localized pleural plaques might lead to finding a relationship between pleural disease and physiological effect. Moreover, there is also a possibility of confounding by exposure, because higher cumulative exposures increase both the probability of developing plaques and the chance of decreased lung function. In the present study, we were fortunate to be able to study a group of subjects in whom neither diffuse pleural thickening nor radiological asbestosis was present and for whom exposures were neither extremely high nor varied by several orders of magnitude between subjects. This allowed us to isolate any effect of the plaques themselves better than in most other studies, although the disadvantage is that our subjects did not have extensive plaques, which could affect diaphragm shortening (32). In our study, neither the presence nor the extent of localized pleural plaques was found to be associated with lung function. In this cohort no correlation could be found between smoking and pleural plaques, either. This is consistent with the findings of several other authors (34).

In conclusion, we were able to measure the surface of pleural plaques. However, in these subjects with circumscribed pleural plaques of a moderate surface (mean total surface area, 40-50 cm²; maximum, 300 cm²), those with larger plaques did not have a higher past cumulative asbestos exposure or a poorer lung function.