Document q3BGXb8qnZwd6zQGQX2ZpZzG

January 1995 ^XjjuivaJtency Ppro*pose Wbe and IS. ti 01 was ejection 0? hotogkaliy resolution 69 >, CA: Aca- ai. and ratio3574-607 ary fibrosis and hislo- asbesteis 1088-1099 computed ith normal losls. Clin il workers' latlon with Jiagnostic 32 infiltrativa st radiog- yj losis: CT -658 Jtion CT- raluation arenchy'29-734 s-related 11 aseand 269-275 isolated imputed 71-172 Medical 312-17 vorkers sstosis ' ' eirfor- > iles in scans. -resogenic > , The diagnosis of asbestosis is important. It means that the [6], has been increasing. However, none of these tests pro worker must be removed from asbestos exposure and may vides a specific diagnosis of asbestosis. be entitled to compensation fora work-related injury. Because asbestosis is diagnosed by the presence of peribronchiolar fibrosis with asbestos bodies in lung tissue {1], definitive doc umentation of asbestosis usually requires open-lung biopsy, Correlation Between CT-Determined Abnormalities and Pathologic Asbestosis which is difficult to justify because there is no accepted treat Because of its ability to depict relatively fine morphologic ment for asbestosis. Asbestosis thus differs from most other changes in the lungs and to document pleural plaques, high- interstitial lung diseases because it usually is diagnosed by resolution chest CT has been embraced as an objective clinical criteria. method for confirming the presence of asbestos-related lung The clinical diagnosis of asbestosis [1] involves a reliable fibrosis. The CT criteria for the diagnosis of asbestosis [7] history of exposure and an appropriate time interval between are subpleural curvilinear lines, thickened septal and core exposure and detection (the latency period for the develop lines, subpleural density, parenchymal bands of fibrosis, and ment of asbestosis is usually longer than 15 years). The honeycombing. These abnormalities usually are seen best in presence of small irregular opacities (profusion greater than the posterior subpleural area of the lung, at the lung bases. or equal- to eithefi-l/O.or. l/lyaceording-to-the International in the-prone position.- Gamsu et at.- [8] have discarded sub- Labor Office,[ILO] system) on the chest radiograph is the pleura! density, presumably because of its lack of specificity, next most important criterion. Considerable caution is rec and added subpleural nodules and the useful sign of archi- ommended if this criterion is not met jl]. Other features of tectural distortion. recognized diagnostic value include a restrictive pattern of . The general lack of a pathologic gold standard in the pub physiologic impairment, diminished diffusing capacity, and lished CT studies of asbestosis has led to considerable con- bilateral basal crackles on physical examination. cem about the accuracy of CT diagnosis of this condition. Asbestosis may be present and may cause physiologic Akira et al. (91, with meticulous radiologio-pathologic correla- impairment.when. the. lung parenchyma-appears, normal on - tion-for seven-postmortem lungs, confirmed- that core structhe chest radiograph [2; 3]: Because-the standard-radio ture thickening; intralobularlines, subpleural curvilinear lines, graphic criteria may lead to diagnoses of asbestosis being and honeycombing correlate with pathologic evidence of missed [4], interest in other diagnostic techniques, including fibrosis. However, interlobular septal thickening and ground- examination of exercise physiology [5], scintigraphic evalua glass opacity could be attributable either to fibrosis or to tion of the lungs with gallium-67, and bronchoatveolar lavage edema [9]. The study by Gamsu et al. [8J, which evaluated ' ; ; . ; j 1 ! This article is a commentary on the preceding article by Gamsu el al. 'Departmentol Radiology, Box A030. University of Colorado Health Sciences Center, 4200 E. Ninth Ave., Denver, CO 80262. Address correspondence to D. A, Lynch. AJH 1995;164:69-71 0361-803X/95/1641-69 American Roentgen Ray Society HWBUI0009462 70 LYNCH AJR:164. January 1895.3 the relationship between CT-determined abnormalities and pathologic evidence of asbestosis in 30 asbestos-exposed subjects, is a timely and thought-provoking contribution to our knowledge of the pathologic correlates of CT findings in asbestosis. To interpret the study of Gamsu et al. [8], one must realize that the study population is unique. The authors deserve great credit for accumulating a large series of asbestosekposed subjects from - whom adequate- lung- tissue" was available to make the histologic diagnosis of asbestosis. Because lung biopsy samples were available from only six patients, the remainder of the pathologic material was obtained at autopsy (n = 16) or at lobectomy (presumably for malignancy) (n = 8). In all, 16 patients had malignancy at the time of death, lobectomy, or biopsy. For many of these patients, asbestosis must have been an incidental finding at lobectomy or autopsy. Indeed, in a study of asbestos insula tion workers witfv lung-cancer.[3],. asbestosis.was-present in all 138 workers from whom .sufficient nonmalignant-tissue was available for analysis, suggesting that asbestosis is an almost invariable companion to lung cancer in asbestosexposed workers (10]. Thus, the patients in the current study range from those in whom symptomatic asbestosis was suffi cient to cause death or to warrant open-lung biopsy to those who had no asbestosis or for whom asbestosis was an inci dental finding at lobectomy or autopsy. Because of the inevi table selection bias, it is unlikely that this study population is representative of the general population of asbestosexposed workers. One should not use the current study to draw genera! conclusions regarding the accuracy of CT for the diagnosis of asbestosis. The authors' recognition that the CT findings of asbestosis are not perfectly sensitive or specific led them to evaluate two grading systems for determining the likelihood of asbestosis. One system was a cumulative measure of the number of dif ferent types of abnormalities present on CT scans. Not sur prisingly. the likelihood of asbestosis rose as the variety of signs of asbestosis on CT scans Increased. The other grading system was essentially a subjective measure of the profusion and severity of abnormalities on CT scans. Despite the subjec tivity of this scoring system, it correlated significantly (r = .78, p = .0001) with the severity of fibrosis measured by histopathology. It is reassuring that all 16 subjects with grade 2 or grade 3 asbestosis (changes that were bilateral or occurred at multiple levels) had pathologic evidence oh asbestosis. How ever, four of five subjects with grade 1 CT scan abnormalities (unlikely to represent diffuse lung fibrosis) had pathologic evi dence of asbestosis, suggesting that-eyen smalt abnormalities should be regarded as indicating possible asbestosis. In prac tice. it is often difficult to distinguish such small parenchymal abnormalities from the linear densities (crow's feet) commonly seen in subjects with asbestos-related pleural disease [11] or the minor septal thickening or subpleurai lines sometimes seen in normal subjects. Normal CT Scans In Subjects with Asbestosis Five of nine subjects with normal (grade 0) CT findings in the study by Gamsu et al. [8] had histologic evidence of asbestosis. All of these subjects had normal chest] graphs. Some of these patients may have come from tne group in which asbestosis was discovered incidentally at lobectomy or autopsy. It is not known whether any of them had physiologically significant disease. It is dear that there: exists a phase in the development of any interstitial lung dis ease during which the disease is too focal or too mild to' cause a recognizable morphologic change on CT scans [12]. - In the future; image analysis techniques may aid. in the detection, of changes in CT density or-texture before a mor phologic change is apparent [13]. is CT more sensitive than chest radiography for the detection of asbestosis? Friedman et al. [14] concluded that CT rarelyj depicted asbestosis (two of 60 cases) unless this cprttfition wad suspected on the basis of the chest radiograph findings. How ever, Aberie et al. [7] demonstrated that eight of 38 subjects for; whom asbestosis was found highly likely by.CT had normal chest radiographs? In a study by Staples etal:[2} of 1 effasbestos-exposed subjects with normal chest radiographs, 57 Had CT findings suggesting a high probability of asbestosis. The vital capacity and diffusing capacity were significantly lower in these 57 subjects than in 76 subjects with normal or nearly normal CT findings. In the study by Gamsu et al. [8], with its unique popula tion, 11 of 21 subjects with asbestosis (52%) had abnormal chest radiographs (ILO profusion, 1/0), whereas 16of 25 sub jects with asbestosis (64%) had abnormal CT findings. On bal ance, CT seems considerably more sensitive than chest radiography for the detection of asbestosis. Pleural Plaques In Asbestosis Pleural plaques are an important and clinically useful marker of asbestos exposure. However, plaques were absent on the chest radiographs of 32% of subjects with lung cancer and pathologic evidence of asbestosis [3]. Con versely, plaques may be present on chest radiographs in the absence of histologically determined-asbestosis.. I n the study; by Friedman et al. [1.4], two of 21 subjects with CT findings of; asbestosis had no plaques on CT scans. Gamsu et al.' [8].i confirmed that the absence of plaques on CT scans does not] exclude asbestosis. A more frequent CT sampling interval i may increase the likelihood of the detection of plaques. Summary CT is more sensitive than clinical-evaluation for-the detec tion of asbestosis but is inevitably less sensitive and less specific than -pathologic evaluation. For the asbestosexposed individual, CT is useful for the evaluation of sus pected lung masses, particularly rounded atelectasis [15], for identifying pleural plaques, and for confirming unequivocal asbestosis (grade 2 or grade 3 [8]). CT also will identify and quantify emphysema as a cause of physiologic impairment. Because clinicians commonly use CT to resolve clinical uncertainties, radiologists often feel pressured to categorize disease as unequivocally present or absent. Gamsu et a show that the borderline between normal and abnorma, .</ not always sharply defined. In the absence of pathologic proof, the diagnosis of asbestosis must be based on a HWBUI0009463 O xjR:l6t, January 1995 COMMENTARY 71 the intally at ' of them Tat there lung dis. > mild to arts [12], d- in the e a mor- detection )T rarely ition was js. HowJjects for 1 normal 9asbesr ''hadCT rhe vital in these irmaiCT populabnormal 25subOn bali chest ) thoughtful evaluation of the likelihood of asbestosis by use of afl available clinical, physiologic, and radiologic information. The scoring systems used by Gamsu et al. [8] offer a practi cal approach to defining the likelihood of asbestosis based on CT appearances. acknowledg ment The author wishes to express appreciation for the very helpful comments.madaby Dr. Newman and the expert assistance of Mary. Ann Hamm. REFERENCES 1. Murphy RL, Beddake MR, Gaensler EA, atal. The diagnosis of nonmallgnanl diseases related to asbestos. Am RavResplrDls 1986; 134:363-368 2. Staples CA, Gamsu G, Ray CS. Webb WR. High resolution computed tomography and lung function In asbestos-exposed workers with normal chest radiographs. Am Rev RespirOis 1989;139:1502-1508 3. (Open HM. tills R, Suzuld Y, Vatdukas JA, Selikoft IJ. Pulmonary fibrosis inasbestas- Insulation-workers wtthlungeancaR aradiologiral -and Wstopalhotogicai evaluation. BrJtnd Med 1987;44:96-100:' :* 4. Rockoff SD. Schwartz A. Roentgenographie underestimation of earfy asbestosis by International labor organization classification. Chest 1986; 93:!988-1091 5. Smith D, Agostoni P. The discriminatory value of the P(A-a)02 during exercise In the detection of asbestosis in asbestos exposed workers. Chest 1989;95:52-55 6. Begin R. Ostiguy G. Rlion R, Groleau S. Recent advances in the earty diagnosis of asbestosis. Semin /?oenfgerrof1992;27;f2f-t39 7. Abode DR, Gamsu G, Ray CS. High-resolutlon CT of benign asbestos-rotated cfeaases: cSnteal and radiographic oorrefaSoa AIR 1338;1 SI:883-891 8. Gamsu G. Salmon CJ. Wamock ML, Blanc PD. CT quantification of inter stitial fibrosis in patients with asbestosis: a comparison of two methods. AJR 1995;164:63-68 9. Akira M, Yamamoto S, Yokoyama K, et at. Asbestosis: high-resolution CTpathotogiccdrrelatlon. Radiology'tB9ty,i7H:3SS:-^94 10. ChurgA. Asbestos-related disease in the workplace and the environment controversial Issues. MonogrPathol1933;36;54~-77 11. HMerdal G. Non-maligriant asbestos-pleura! disease. -Thorax-1981;36: 669-675 12. Lynch DA, Rose CS, Way D, King TJ. Hypersensitivity pneumonitis: sensitivity of Mgh-resdutbn CTin a population-based study. AJR1992159:469-472 13. Wollmer P, Jakobsson K, Aibin M. et al. Measurement of lung density by X-ray computed tomography: relation to lung mechanics In workers exposed to asbestos cement. Ores! 1387:91-.865-869 14. Friedman AC. Fist SB. Fisher MS, Fladecki PD, Lev-ToadAS. Caroline DF. Asbestos-related-pteurad-dtsease -and asbestosis-. a comparison -of- CT an<rchestradiography.A/R 1988;150:269-275 . 15. Lynch DA, Gamsu G. Ray CS. Aberte DR Asbestos-related focal lung masses: manifestations on conventional and high-resolution CT scans. Radiology 1988;169:603-607 useful 3 were ith lung I- Cons in the e study lings of : al. [8] oes not nterval s. : j detecrd less restosif sus15], for livocal fy and "ment. finical jotize a' r<] }ogic on a > > ' > > 22E5SS9HH HWBUI0009464