INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Coal worker's pneumoconiosis is, in France, diagnosed only at the stage of confirmed disease, as reflected by characteristic changes in pulmonary chest X-ray films according to the classification of the International Labor Organization (ILO) (1). With improvement in knowledge of the pathogenesis and epidemiology of pneumoconiosis and preventive measures against it, a decline in its incidence and prevalence has been observed in coal miners in France (2, 3) as well as in the United States (4), Great Britain (5), and Australasia (6). In France, however, the early stages of pneumoconiosis, defined as grade 0/1 to 1/0 findings on X-ray, are fairly frequent in relatively young active miners, and some of them develop pneumoconiosis with findings classified 1/1 or greater before the age of 50 yr. Several studies have shown an association between pneumoconiosis and respiratory symptoms and lung-function disturbances (7).

For better prevention of pneumoconiosis and also for better management of the coal-mining industry, it may be important to identify those subjects at greatest risk for pneumoconiosis. A question of interest in this regard is whether pneumoconiosis can be predicted from respiratory symptoms, lung-function indices, cumulative coal-mine-dust exposure, and the presence of micronodules on lung computed-tomographic (CT) scans, a technique recently introduced to screening for pneumoconiosis.

In an approach to this problem, we conducted a longitudinal study among the coal miners in the French Houillères du Bassin de Lorraine mine. The study was aimed at assessing respiratory symptoms, lung-function indices, cumulative coal-mine-dust exposure, and the presence of micronodules on lung CT scans as factors prognostic of radiographic worsening and the evolution to pneumoconiosis within 4 yr.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

When this study was done, the Houillères du Bassin de Lorraine, in northeastern France, employed 10,046 underground workers and 5,371 surface workers. Standard high-kilovoltage posteroanterior and lateral chest films at maximum inspiration were made for each worker every 2 yr at the workers' yearly medical examination. These films were routinely read by the collieries' occupational physicians, who were trained to grade small opacities according to the ILO classification (1).

First, we selected 264 subjects, aged 35 to 48 yr, who had worked for 10 yr or more at face work and had no known disease and chest X-ray findings classified as 0/0 to 1/1. We then submitted the radiographs for independent readings by three occupational physicians who had no prior knowledge about the subjects and who read the films in a blinded, random order. We then chose 80 miners with grade 0/1 or 1/0 chest X-ray findings (i.e., who were suspected being in the process of evolution to pneumoconiosis [ES group]). Of these 80 subjects, 72 subjects were classified as falling within these categories by three readers, and eight were classified by two readers (10).

The group exposed to coal mine dust and suspected of pneumoconiosis (ES group) was paired with two control groups matched with it in age (± 2 yr), height (± 5 cm), weight (± 10 kg), and smoking by choosing the nearest subjects. The first control group consisted of 80 miners who had worked for at least 10 yr at coal faces, but with a normal chest X-ray (exposure with normal X-rays [EN group]). The second control group included 80 other underground workers who had worked for less than 2 yr at coal faces, and who had normal X-rays (no exposure, normal X-rays [NN group]).

Design

The protocol comprised: (1) a clinical examination; (2) a questionnaire on respiratory symptoms and smoking habits, prepared by the European Steel and Coal Community (ESCC) and supplemented with questions about job history (11); (3) spirometry; and (4) CT scans.

Spirometric measurements were made with a computerized spirometer based on a Fleisch No. 3 pneumotachograph (Spiromatic; L. Martin, Paris, France). The subject, in the sitting position and with the nose obstructed, generated three valid flow-volume curves (with the control curve on the screen showing the shape of the curve, and a difference in vital capacity [VC] of less than 5%). From the envelope curve (i.e., the composite curve obtained by superimposing the curves at their starting point, or full inspiration, we calculated: FVC, FEV₁, maximum midexpiratory flow (MMEF), and maximal expiratory flows at 50% and 25% of VC (FEF_50% and FEF_25%, respectively) (12).

Lung CT scans were made for all subjects. The lungs were divided into six areas: the upper zones above the carina, the middle zones between the level of the carina and the lower pulmonary veins, and the lower zones below the level of the lower pulmonary veins. The CT slices were taken at FRC, which was controlled through a strain-gauge belt. In each zone, one thick slice (8 mm) and two thin slices (1 mm) were taken about 1.5 cm apart from each other. The results were stored on a floppy disk. A Siemens DRH (Erlangen, Germany) scanner apparatus was used for the thick and thin slices, with a matrix size of 512 × 512 mm. Thin slices were reconstructed with a high-resolution algorithm, the acquisition time being 4 s.

The presence of coal worker's pneumoconiosis was evaluated mainly in terms of micronodules, their extent, and their site. Micronodules were defined as "opacities" less than 7 mm in diameter, and nodules were defined as opacities from 7 to 20 mm in diameter (13). Micronodules were evaluated in both the parenchymal and subpleural areas. Subpleural micronodules were defined as round and hemispheric or triangular opacities of less than 7 mm in diameter in the subpleural areas, along the chest wall, and along the fissures on both the thin and thick slices. Micronodules, nodules, and other abnormalities, such as "emphysema" and profusion of shadows were recorded according to criteria of the consensus meetings for CT scans established by the Society for Thorax Imagery and the Thorax Club in September 1989 and June 1990 (14), and according to the pathology standards for coal-worker's pneumoconiosis defined by the College of American Pathologists (15). Profusion of abnormalities was graded from 0 to 3 (0 = absence, 1 = rare, 2 = intermediate; 3 = high profusion), and was estimated for the six lung locations examined (upper, middle, and lower parts of the right and left lungs), giving a profusion score of from 0 to 18; a mean profusion may be obtained dividing this score by 6. Two readers had read the scans. The readings of the scans for the first and the second examinations were done at the same time by readers who read independently and were not aware of the examinees' clinical files or other information. If there was a discrepancy, a consensus was reached.

Cumulative coal-mine-dust exposure was calculated with the method proposed by Attfield and Morring (16); for each miner from the miner's job history and from dust measurements at various sites in the coal mine. The estimates were defined by the summations of dust measurements (mg/m³) for the time spent in various workplaces. Cumulative coal-mine-dust exposure was calculated until June 1994 and expressed as milligrams of dust per cubic meter multiplied by years exposed (mg/m³ × yr).

All the examinations and measurements were made in 1990 and 1994 by the same medical team. Of the 240 subjects who had participated in the first survey, 234 (97.5%) were reexamined (one member of the original group was deceased, two refused, and three had left the area). All of these subjects had given their prior written consent to participate in the study. The study protocol was approved by the National Ethics Committee for Medical Research of France.

Statistical Analysis

Comparison of frequencies of symptoms and smoking habits between the different groups were made with the chi-square test or Fisher's exact test of probability. Comparisons between two groups of age, height, weight, tobacco consumption, cumulative coal-mine-dust exposure, scanner micronodule score, and lung-function indices (expressed in percentage of predicted value) were made with the t test. Linear discriminant analyses (17) were made to determine prognostic factors for X-ray worsening and pneumoconiosis at the second examination. Discriminant analyses were also performed to identify the factors distinguishing subjects with worsened X-ray findings from the others. The SAS statistical software package (SAS Institute, Cary, NC) was used for all of these analyses (18).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

At the first examination, the subjects in the ES group had an average age, height, and weight of 42.8 ± 3.4 yr (mean ± SD), 175 ± 6 cm, and 80.3 ± 9.9 kg, respectively. Nonsmokers represented 22%, ex-smokers 22% (4% of them having smoked 20 pack-yr or more), and current smokers 55% (22% having smoked 20 pack-yr or more) of the group. The EN and NN groups were of similar age, height, weight, and smoking history to the ES group. The duration of underground work was closely matched for the ES and EN groups (about 24 ± 6 yr). It was higher than that of NN group (18 ± 9 yr). The cumulative coal-mine-dust exposure of the three groups was, respectively, 85 ± 41, 66 ± 32, and 20 ± 17 mg/m³ × yr.

At the beginning of the study, 57 subjects (71%) in the ES group had X-ray grades of 0/1 and 23 miners (29%) had X-ray grades of 1/0. All subjects in the EN and NN groups had X-ray scores of 0/0, as noted earlier. The micronodule scores on the CT scans for the ES, EN, and NN groups were, respectively, 0: 19% versus 47% and 63%; 1 to 2: 38% versus 40% and 37%; and 3 or more: 43% versus 13% and 0%. At the end of the 4-yr follow-up, 24 subjects in the ES group had worsened X-ray findings, and 10 of them had evolved to having scores of 1/1 or over according to the ILO classification. In the EN group, five subjects had X-ray scores of 0/1 and had an X-ray score of 1/0 at the end of the follow-up. In the NN group, only one subject had an X-ray score of 0/1 at the second survey.

Table 1 shows, that at the first examination, wheezing in the ES group was significantly more frequent and the CT-scan micronodule score significantly higher in miners who had worsened X-ray findings at the end of the follow-up than in the other subjects. The first group also had greater cumulative coal-mine-dust exposure, but the difference was not significant. In the three groups combined, cumulative coal-mine-dust exposure, CT-scan micronodule score, and frequencies of dyspnea (stage 1 or greater) and wheezing were greater in the miners with worsened X-ray findings. No significant difference was observed for cough, expectoration, chronic bronchitis, or any lung-function indices.

The CT-scan micronodule score was significantly higher and the FEV₁/FVC ratio, MMEF, and FEF_25% were significantly lower among those subjects who developed pneumoconiosis. These subjects also had reduced values of FEV₁ and FEF_50% but the differences from the subjects without pneumoconiosis did not reach significance (Table 2).

The results obtained with discriminant analyses, presented in Table 3, show that, in the ES group, wheezing and CT-scan micronodule score could be used to identify the subjects with worsened X-ray findings. However, the percentage of well-classified subjects, though high for the subgroup with worsened X-rays, was low for the subgroup without worsened X-rays. In the three groups (ES, EN, NN) combined, cumulative coal-mine-dust exposure was also found to be a significant determinant of X-ray worsening. This was because the subjects with worsened X-ray findings were mainly in the ES group, which had a higher dust-exposure level. It should be emphasized that the percentages of well-classified subjects were 65% for the subgroup with worsened X-rays and reached 81% for the subgroup without worsened X-rays.

With regard to grades of X-ray abnormality, CT-scan micronodule score, wheezing, and cumulative coal-mine-dust exposure were also significant discriminant factors in distinguishing X-ray grades of 1/0 or over from lower X-ray grades, with the percentages of well-classified findings in each category being 77% and 84%, respectively. For pneumoconiosis, the CT-scan micronodule score and the FEF_25% value appeared to be significant factors, but provided modest percentages of well-classified subjects (60% and 76%). In the three groups combined, very good results were obtained with the CT-scan micronodule score and the MMEF for pneumoconiosis, with percentages of well-classified subjects reaching 80% and 88%, respectively (Table 4).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The present study assessed the prognostic role in coal miners of cumulative coal-mine-dust exposure, tobacco consumption, lung CT scans, respiratory symptoms, and lung-function indices for chest X-ray worsening and for evolution to pneumoconiosis during a 4-yr follow-up. Because miners involved in this study were relatively young (aged 35 to 48 yr at the beginning of the study), practically none of them retired during the follow-up (97.5% of the subjects participated in the second survey). The study protocol was used by the same medical team and under the same conditions in the two surveys.

For the chest X-ray readings, we used three trained readers because the readings of chest films were well codified. For the readings of CT scans we used only two readers because the findings with this relatively new technique were still not codified. However, the latter two readers were trained readers and members of the French consensus team (14).

At the end of the 4-yr follow-up period, the radiologic classification was worsening for 24 of the 78 subjects in the ES group, and 10 reached the ILO level of 1/1 or more that allowed them to be classified as having pneumoconiosis. In the EN group, only six of 78 subjects had worsened X-rays, and in the NN group only 1 of 78 subjects had worsened X-rays.

An association between cumulative coal-mine-dust exposure and X-ray worsening was observed in the three groups combined. This finding was consistent with that by Hurley and colleagues (19), who found a relationship between radiographic grade and exposure to dust in British coal miners, and the finding of Attfield and Seixas (20), who showed a positive relationship between the percentage of pneumoconiosis and cumulative dust exposure in coal miners in the United States.

The lung-function indices measured in our study were not differently significant for subjects with X-ray worsening versus others. This was in agreement with the results obtained by Collins and coworkers (21), which showed that lung function is not altered in the early stages of pneumoconiosis. However, the FEV₁/FVC ratio, MMEF, and FEF_25% were lower at the first examination among subjects with evolution to pneumoconiosis during the 4-yr follow-up period. This was consistent with the results reported by Dechoux and associates (22) and by Worth and Smidt (23), which demonstrated a correlation between decreased lung function and increasing pneumoconiosis grade for more advanced coal workers' pneumoconiosis.

Tobacco consumption was not found to be different for subjects with X-ray worsening, in contrast to other variables associated with the evolution to pneumoconiosis. This may have been due to low consumption as a result of campaigns against smoking. However, the miners who developed pneumoconiosis clearly had greater tobacco consumption, the differences being nonsignificant because of the small number of subjects in the study.

Among clinical signs, wheezing was found to be significantly more frequent among subjects who had X-ray worsening and pneumoconiosis. Lung scanning improves not only the precision of the diagnosis, but also the predictability of the evolution to pneumoconiosis. The CT-scan micronodule score was significantly higher in the ES group than in the EN and NN groups at first examination, and this score was significantly higher among the miners with X-ray worsening, and even more so among those who developed pneumoconiosis. Our results agreed with those of Remy-Jardin and colleagues (13, 24), Grenier and associates (25), and Collins and colleagues (26), pointing to the utility of CT and its superiority to chest radiography in evaluating "simple silicosis" (13). However, as stated Wagner and colleagues (27) "even with the relative insensitivity of radiographs to pneumoconiosis the chest X-rays remained a useful basis for investigating temporal and geographic trends to pneumoconiosis."

The results we obtained indicated that the 4-yr change in chest X-ray findings seen in our study was due mainly to coal- mine-dust exposure. This could not rule out the role of "individual susceptibility" emphasized by Begin and associates (28) and Jakobsson and coworkers (29)a term describing the difference in distribution of inhaled air (30) and lung structure (31), but also alveolar dust clearance capacity (28), and immunologic and histocompatibility factors (32).

The present study provided interesting information about the early stages of pneumoconiosis, which have rarely been studied, as attested by the paucity of their discussion in the literature.

We concluded that worsening X-ray findings and pneumoconiosis were more often observed in coal miners with micronodules on lung CT scans, wheezing, low values of MMEF and FEF_25%, and high dust exposure at the first examination. This offers the possibility of identifying those miners who need to be relocated to less exposed workplaces and to be monitored. Our study also pointed to the utility of lung CT scanning, mainly among workers with suspect shadows on classic X-ray films. However, it is not our intention to recommend CT-scans for all underground coal workers. The X-ray remains useful, and very few subjects with an ILO score of 0/0 on chest X-ray showed micronodules on chest X-rays 4 yr later.