Low Grade Coal Worker's Pneumoconiosis

Transcription

1 Acta Radiologica ISSN: (Print) (Online) Journal homepage: Low Grade Coal Worker's Pneumoconiosis P. A. Gevenois, E. Pichot, F. Dargent, S. Dedeire, R. Vande Weyer & P. De Vuyst To cite this article: P. A. Gevenois, E. Pichot, F. Dargent, S. Dedeire, R. Vande Weyer & P. De Vuyst (1994) Low Grade Coal Worker's Pneumoconiosis, Acta Radiologica, 3:4, To link to this article: Published online: 04 Jan Submit your article to this journal Article views: 120 Full Terms & Conditions of access and use can be found at

2 Actrr Rarliologicu 3 (1994) Fusc. 4 Prinrrrl in Dcnmork. All rights reserved Copyright 6 Actu Rudiologku 1994 ACTA RADIOLOGICA ISSN FROM THE DEPARTMENTS OF RADIOLOGY AND CHEST MEDICINE, AND THE FONDS DES MALADIES PROFESSIONNELLES, H6PITAL ERASME, UNIVERSITE LIBRE DE BRUXELLES, BRUSSELS, BELGIUM. LOW GRADE COAL WORKER S PNEUMOCONIOSIS Comparison of CT and chest radiography P. A. GEVENOIS, E. PICHOT, F. DARGENT, S. DEDEIRE, R. VANDE WEYER and P. DE VUYST Abstract We compared CT with chest radiography (CR) in the assessment of low grade coal worker s pneumoconiosis (CWP) in a population of 83 subjects. All subjects had a high-voltage p.a. CR, graded according to the ILO classification between 0/0 and 1 /I, a conventional CT (CCT) using contiguous I-cm-thick sections on the entire thorax and a set of 10 high-resolution CT (HRCT) images. CR and CT were separately read by consensus by 2 teams of 2 trained readers. CR were coded 010 in 9 subjects; O/ 1 in 31; 1 / O in 28; I / 1 in 1. Among these groups of patients, micronodules were detected by CT in respectively 2 (22%), 14 (4%1), 17 (61%) and 10 (67%)) patients. In all groups, micronodules were more often detected by CT when the opacities detected on CR were scored as rounded (p, q) than irregular (s, t). Among the patients graded 010 or O/ 1, CT showed micronodules in 40Y0. By contrast, among the patients graded I/O or 1/1, CT did not show micronodules in 37%1, but revealed in numerous patients that opacities detected on CR were related to bronchiectasis and/or emphysema only. Comparative analysis of HRCT and CCT showed that both techniques are complementary and proved the usefulness of CCT in the detection or confirmation of low profusion of micronodules. Key words: CT, high-resolution; lung, CT; -, pneumoconiosis; radiography, chest. In a subject with a documented exposure to coal mine dust, the diagnosis of coal worker s pneumoconiosis (CWP) is usually based on the chest radiograph (CR), histopathologic specimen of the lung being rarely available and not justified for this purpose. Evaluation of the disease severity includes the interpretation of the CR coded according to the 1980 ILO classification (6) and pulmonary function tests. Conventional (CCT) and high-resolution CT (HRCT) have been proved to be more informative than CR in detecting parenchymal changes in diffuse interstitial lung diseases (, 7, 9). The value of this method in evaluation of silicosis has been recently assessed (1, 8). Since the diagnosis of CWP still remains a frequent medicolegal issue, the aim of this study was to compare the value of CR and CT in the assessment of low grade CWP. Material and Methods The 83 men (mean age 9 years; range years) in this study were referred to our department by the Belgian Occupational Diseases Fund for a chest CT examination as part of a medicolegal evaluation, for possible CWP, associated with a minimal exposure of 10 years to coal dust in the coal mining industry. Standard high-voltage p.a. films were exposed at maximal inspiration. The CR were graded for profusion, size and shape of small opacities by 2 trained readers (F. D., P. D. V.) according to the ILO classification (6) and who have a good knowledge of the ILO classification criteria. This scoring system is based on the reader s assessment of small opacities compared with a set of standard radiographs provided by the ILO. Two kinds of shape are recognized: rounded and irregular. Three sizes are differentiated for each of these shapes. The letters p, q and r denote the presence of small rounded opacities, respectively, for a diameter up to about 1. mm, exceeding about 1. mm and up to about 3 mm, exceeding about 3 mm and up to about 10 mm. The letters s, t and u denote the presence of small irregular opacities, respectively, for a width up to about 1. mm, Accepted for publication 20 November

3 32 P. A. GEVENOIS ET AL. exceeding about 1. mm and up to about 3 mm, exceeding about 3 mm and up to about 10 mm (6). The readers were not involved in the CT assessment and not aware of the CT results. They had a long experience working with the 1980 ILO classification, reading several hundred images per year. Discrepancies in grading between observers were resolved by consensus. For the purpose of this study, when the CR were graded with 2 different sizes or shapes, only the first letter of the ILO classification was considered. Individuals with a profusion of opacities superior to 1/1 at the consensus reading or with confluences were excluded from the study group. The CT examination consisted in a series of CCT and HRCT images, obtained on a Siemens Somatom Plus unit. CCT was performed by using 10-mm collimation, 120 kvp, 210 ma, and a 1-s acquisition time. Patients were imaged supine at full inspiration from above the lung apices to below the posterior costophrenic angles at contiguous 1 -cm intervals. I.v. contrast medium was never used. Images were reconstructed by using a standard reconstruction algorithm and were photographed at window settings suitable for viewing the lung parenchyma and mediastinum. HRCT was performed in all patients immediately after the CCT and images were obtained from the level of the sternoclavicular joint to the diaphragm at 20 or 30 mm intervals, depending on the height of the thorax, in supine position for a total of 10 images. Images were acquired at full inspiration by using 1-mm collimation, 137 kvp, 2 ma, and a I-s acquisition time. Images were reconstructed with an ultra-high-spatial-frequency algorithm and photographed at window settings suitable for imaging the lung parenchyma. CCT and HRCT were separately interpreted for the presence of micronodules by 2 readers (P. A. G., s. D.). Discrepancies in interpretation between observers were resolved by consensus. According to the definition proposed by REMY- JARDIN et al. (8), we regarded round opacities less than 7 mm in diameter as micronodules. For the purpose of the present study, the CT evaluation was considered consistent with the diagnosis of CWP when micronodules were bilateral and demonstrated on at least 2 CT images at least 3 cm apart. Results The results of CR and CT analyses are summarized in Tables I to 3. The size of the opacities is not included because the number of individuals in each group would be too small. Sixty-six subjects (80%) were scored p or s and 17 subjects (20%) were scored q or t. No patient was classified as r or u. Combined reading of CCT and HRCT revealed micronodules in 43 patients (2%); 3 on CCT and 32 on HRCT, respectively. CT were scored normal in 40 individuals' (48%). Comparative analysis of chest radiographs and CT (Tables Table I Comparison between chest radiography and CT Profusion Shape Patients CT+ CT- o/o o/ 1 Rounded Irregular 1 / O Rounded Irregular 1/1 Rounded Irregular Table 2 Comparison between chest radiography and CT Profusion Patients CT + CT - < 1/ >or>i/o Table 3 Comparison between conventional and high-resolution CT 7 10 I CCT+, CCT+, CCT.-, Profusion Patients HRCT+ HRCT- HRCT+ o/o O/ / and 2). Among the 9 patients with disease classified as profusion category 0/0 disease on CR, combined reading, of CCT and HRCT showed micronodules in 2 patients (22%) (Fig. 1). Among the 31 patients with profusion category 0/1 disease on CR, combined reading of CCT and HRCT showed micronodules in 14 (4%). Among these 14 patients, the shape of the opacities on CR had been classified as rounded in 12 cases and irregular in 2 patients. Among the 17 patients with no micronodules on CT, the shape of the opacities on CR had been classified as rounded in 10 (9%) and irregular in 7 cases (41%). Among the 28 patients with profusion category 1/0 disease on CR, combined reading of CCT and HRCT showed micronodules in 17 patients (61%). Among these 17 patients, the shape of the opacities on CR had been classified as rounded in all. Among the 11 patients with no micronodules on CT, the shape of the opacities on CR had been classified as rounded in 6 patients and irregular in. Among 1 patients with profusion category 1/1 disease on CR, combined reading of CCT and HRCT showed micronodules in 10 patients (67%). Among these 10, the

4 COAL WORKER S PNEUMOCONIOSIS 33 Fig. I. Comparison between CR and CT in the detection of micronodules. Magnified views of the p.a. CR and CT in the upper right lung field. CR (a) classified as profusion category O/ 0. Conventional CT (b) and HRCT (c), performed at the same anatomic level, show numerous micronodules. shape of the opacities on CR had been classified as rounded in 7 patients and irregular in 3. Among the patients with no micronodules on CT, the shape of the opacities on CR had been classified as rounded in 3 and irregular in 2 patients. In Table 2, patients are divided into 2 groups according to the profusion category. Forty patients were graded with a profusion category 0/0 or O/ 1. Combined reading of CCT and HRCT was positive in 16 (40%) of these patients. The individuals of this group were on average 6 years of age (SD 7 years). Forty-three patients were graded with a profusion category superior or equal to 1/0. Combined reading of CCT and HRCT was positive in 27 (63%) of these pa- tients. The individuals of this group were on average 9 years of age (SD 6. years). The difference of the mean ages was not significant. The 16 CT images of the patients with a profusion category superior or equal to 1 /O on CR and no micronodules on CT were reviewed. The opacities were scored as rounded in 9 patients and irregular in 7. CT showed that the parenchymal opacities demonstrated on CR associated with emphysema (Fig. 2) and/or bronchiectasis (Fig. 3) in 11 of these 16 patients (emphysema in 8 cases, mild bronchiectasis in 1 case, bronchiectasis associated with emphysema in 1 case and a Swyer-James syndrome in 1 case). CT was scored normal in the remaining patients. The number of patients

5 34 P. A. GEVENOIS ET AL. Fig. 2. Magnified views of the p.a. CR and HRCT in the upper right lung field. CR (a) classified as profusion category 1 /O. HRCT (b) shows that the pulmonary opacities demonstrated on CR are associated to emphysema. with rounded or irregular opacities was similar in each group (abnormal and normal). Comparative analysis of HRCT and CCT (Table 3). Among the 2 patients with profusion category 010 disease on CR and an abnormal CT, HRCT was positive in these 2 patients and CCT in one patient. Among the 14 patients with profusion category O/ 1 disease on CR and an abnormal CT, both techniques showed simultaneous micronodules in 6 patients (42.9%), CCT only in patients (3.7%) and HRCT only in 3 patients (21.4%). Among the 17 patients with profusion category 1 I0 disease on CR and an abnormal CT, both techniques showed simultaneous micronodules in 10 patients (8.8%), CCT only in 4 patients (23.%) and HRCT only in 3 patients (17.6%). Among the 10 patients with profusion category 1 I1 on CR and an abnormal CT, both techniques showed simultaneous micronodules in 7 patients (700/0), CCT only in 2 patients (20%) (Fig. 4) and HRCT only in 1 patient (10Y0). In absence of pathologic proof, the diagnosis of pneumoconiosis was established on the presence of micronodules Fig. 3. Magnified view of the p.a. CR and HRCT in the upper right lung field. CR (a) classified as profusion category 1 / 1. HRCT (b) shows that the pulmonary opacities demonstrated on CR are related to bronchiectasis. on CCT and/or HRCT. Subsequently, the sensitivity and the specificity of these techniques are respectively 81 and 83% for CCT and 74 and 78% for HRCT. Discussion The value of CR in patients with low grade diffuse infiltrative lung disease is limited. EPLER et al. (3) have showed that in patients with such disorders histologically confirmed, 9.6% had normal pre-biopsy CR. During the last few years, CT has made it possible to image the lung with excellent spatial resolution and anatomic detail. Many studies have described the HRCT characteristics of various pulmonary diseases and the utility of HRCT in the clinical management of patients with interstitial lung diseases is well established. Several studies have proved the superiority of CT over CR in the evaluation of infiltrative lung disorders (). REMY-

6 COAL WORKER S PNEUMOCONIOSIS 3 Fig. 4. Comparison between CCT and HRCT. Magnified views of the p.a. CR and CT in the middle left lung field, at the level of the lingular bronchi. CR (a) classified as profusion category 1 / 1. Conventional CT (b) shows numerous micronodules. HRCT (c) does not show parenchymal micronodules. JARDIN et al. (8) have proved the superiority of combined HRCT and CCT over CR in the evaluation of CWP, with improved sensitivity in the detection of small parenchymal opacities. BBCIN et al. (1) have evaluated the ability of CCT and HRCT to detect early silicosis in 49 exposed workers. They showed that CT significantly identifies more opacities that CR and that CT significantly reduces the interreader variability, despite the absence of ILO type films for CT images. In the present study of workers at risk for CWP, we showed that in each profusion category group, there are discrepancies between CR and CT. Commonly, the diagnosis of silicosis is based on an ILO profusion category equal or superior to 1 /O (10). If CT is considered as a standard of reference, as strongly suggested by the data of REMY- JARDIN et al. (8), and the profusion score of 1/0 as the inferior limit for an abnormal CR, we found 16 false-negatives on 40 negative CR (40%) and 16 false-positives on 43 positive CR (37%) (Table 2). These findings show that in each category, performing CT appears necessary to accurately identify micronodules. Our results show that the CT findings are related to the shape of the opacities on CR. In each profusion category on CR, the CT is more frequently positive when the opacities are classified as rounded; this difference is significant (x ; p < 0.0) in the profusion category 1 /O. DICK et al. (2) have recently reviewed the significance of irregular opacities on CR. The radiographic technique, obesity, age, cigarette smoking and several other occupational exposures are associated with irregular opacities (4). In the present study the mean age was similar in all groups. Particularly, the patients with abnormal CR and no micronodules on CT were not older. Our results demonstrate that emphysema may be the explanation of such opacities. The superimposition of the walls of small bullae, well detected by HRCT, could be responsible of the formation of these irregular opacities recognizable on CR. CT analysis shows that opacities on CR can be related to emphysema and bronchiectasis without associated micronodules on CT. The estimation of disease severity conventionally involves a combination of radiographic findings and pulmonary function tests (4). The exact relationship between very subtle opacities assessed by CT and functional abnormalities remains unknown. The diagnosis of CWP by CT as a basis for medicolegal compensation is debatable: when lung function tests are normal, slight CT abnormalities could be considered evidence more of exposure than of a disease. The medical implication of an occupational disease requiring CT for its detection is probably limited but this is beyond the scope of the present study. It should, however, be insisted that, by using CT, the diagnosis of CWP can be excluded in a significant number of patients with an abnormal CR. The functional impairment can then be safely attributed to another disease (emphysema and bronchiectasis) and should not be economically compensated as an occupational lung disease in the absence of CT signs of CWP. In the past few years, HRCT has become an accurate method for the evaluation of chronic infiltrative lung disorders. On the 43 positive CT, micronodules were only detected by HRCT in 8 cases (19%). Nevertheless, CCT remains useful to detect micronodules. In our study group, these abnormalities were only detected by CCT in 11 patients (26%), both methods being simultaneously positive in 24 patients. Because a minimal profusion of micronodules is necessary to recognize them; thick sections take advantage of their superimposition. In addition, micronodules are easier to distinguish from blood vessels on thick sections, because there is more volume within which the blood vessels can branch, unlike thin-section CT. Furthermore, the failure to identify micronodules on HRCT may have been related to missed nodules due to the spacing of the HRCT slices. On the other hand, when micronodules have too low an attenuation number, they cannot be detected by using thick sections (8). These results show that the use of combined

7 36 P. A. GEVENOIS ET AL. imaging of CCT and HRCT is necessary in the early detection of CWP. In conclusion, these data point out the limited value of CR, graded according to the ILO classification to evaluate low grade CWP in exposed workers, especially when the opacities described on CR are irregular. In this study, we confirmed that CCT and HRCT are more sensitive than CR to detect silicosis (1, 8). We also documented that CT is useful to exclude micronodules in patients with opacities on CR consistent with the diagnosis of CWP. Request for reprints: Dr. Pierre A. Gevenois, Department of Radiology, HBpital Erasme, Route de Lennik 808, B-1070 Brussels, Belgium. REFERENCES BBGIN R., Osncuv G., FILLION R. & COLMAN.: Computed tomography scan in the early detection of silicosis. Am. Rev. Respir. Dis. 144 (1991), 697. DICK J. A,, MORGAN W. K. C., MUIR D. F. G., REGER R. B. & SARGENT N.: The significance of irregular opacities on the chest roentgenogram. Chest 102 (1992), 21. EPLER G. R., McLom T. C., GAENSLER E. A,, MIKUS P. J. & CARRINGTON C. B.: Normal chest roentgenograms in chronic diffuse infiltrative lung disease. N. Engl. J. Med. 298 (1978), FRASER. G., PARB J. A., PARB P. D., FRASER R. S. & GENEREUX G. P.: Diagnosis of diseases of the chest, 3rd edn., p Saunders, Philadelphia GRENIER P., VALEYRE D., CLUZEL P., BRAUNER M. W., LENOIR S. & CHASTANG C.: Chronic diffuse interstitial lung disease. Diagnostic value of chest radiography and high-resolution CT. Radiology 179 (1991), INTERNATIONAL LABOUR OFFICE: Guidelines for the use of ILO international classification of radiographs of pneumoconiosis. Revised edn. International Labour Office Occupational Safety and Health Series 22 (Rev. 80). International Labour Offce, Geneva MATHIESON J. R., MAYO J. R., STAPLES C. A. & MULLER N. L.: Chronic diffuse infiltrative lung disease. Comparison of diagnostic accuracy of CT and chest radiography. Radiology 171 (1989), REMY-JARDIN M,. DEGREEF J. M., BEUSCART R., VOISIN C. & REMY J.: Coal worker s pneumoconiosis. CT assessment in exposed workers and correlation with radiographic findings. Radiology 177 (1990), REMY-JARDIN M., REMY J., DEFFONTAINES C. & DUHAMEL A.: Assessment of diffuse infiltrative lung disease. Comparison of conventional and high-resolution CT. Radiology 18 I (l99l), Task Force on Occupational Respiratory Disease, p. 3. Health and Welfare Canada. Canadian Government Publication, Ottawa 1979.