Document 3JLddQDZMMa8aEzrdpK6YQa1n

EXHIBIT 63 CT Quantification of Interstitial Fibrosis in Patients with Asbestosis: A Comparison of Two Methods Gordon Gamsu1 Christopher J. Salmon1,2 Martha L-Wamock3 Paul D. Blanc4 Received May 20,1994; accepted after revision -v 25.1994. Ttas work was supported in part by National In"`lesot Health grant 2-T32-CA0938S-09. Department ol Radiology, University ot Gafifor505 Parnassus Ave.. San Francisco, CA94143 -'6 Address correspondence to G. Gamsu. Present address: Dep^jtment of Radiology, sgon Health Science University, 3181 S.W. Sam r'-'Son Park Rd,, Portland, OR 97201. Department of Pathology, University of California. Prandsco. CA 341430628. Department of Medicine, University of California, - 'ancisco. CA941430628. -603)095/1641-63 -~e'-can Roentgen Ray Society OBJECTIVE. The likelihood ol interstitial lung disease being detected on highresolution CT scene and having functional significance is often related to the severity of the disease.-The extent and severity of the abnormalities seen on high-resolution CT are usually assessed subjectively. This study was undertaken to Investigate whether a subjective semlquantltative scoring method or a method Using a cumular tlon of the different high-resolution CT features of asbestosis were comparable in suggesting asbestosis In a group of patients with histopathologic confirmation of dis ease. A secondary objective was to compare the results of these two high-resolution CT methods with chest radiographs In the same population. MATERIALS AND METHODS. This study group consisted of 24 patients and six lungs obtained at autopsy. Histopathologic asbestosis was present in 25 of the 30 patients or lungs. The patients or lungs were imaged using selected high-resolution CT scans. The hlgh-resolutlon CT scans were assessed in two ways. One used a subjective semlquartdtattve extent end severity score consisting of four levels of severity, white the other was a cumulative score adding the different types ot hlgh-resolution abnormalities in asbes tosis. The commonest high-resolution CT abnormalities In the cases with confirmed asbestosis were interstitial lines (84%), parenchymal bands (76%), and architectural dis tortion of secondary pulmonary lobules (56%). Subpleural lines and honeycombing were less frequent The histopathologic severity of asbestosis was Independently graded on a four-point scale. Chest radiographs, when available, were classified accord ing to the International Labor Organization (ILO) classification of pneumoconioses. RESULTS. With the subjective semlquantltative high-resolution CT severity score, asbes tosis was suggested in 16 (64%) instances, alt with disease. With the cumulative method,, any one type of abnormality was present in 88% of cases with asbestosis, two types In 76%, arid three in 56%. However, to Include only cases with asbestosis, three different abnormal ities had to be present The high-resolution CT scans were normal or near normal In five Instances of asbestosis. Chest radiographs using the ILO classification predicted aebestosis with a lesser frequency than hlgh-resolution CT in this selected population. CONCLUSION. We conclude that a subjective semlquantltative grading system of the extent and severity of asbestosis and a method using a cumulative addition of the dif ferent findings In asbestosis give similar results In suggesting the presence of disease. Thus, for the hlgh-resolution CT detection of asbestosis, a combination of the cumula tive number of different findings and an assessment of the extent and severity of-the abnormalities could be complimentary. We also conclude that asbestosis can be present htstopathologically with a normal or near normal hlgh-resolution CT scan. AJR 1995;164:63-68 Imaging for detection of asbestos-related pleural and parenchymal lung disease has relied mainly on the chest radiograph, even though it has been demonstrated that the sensitivity for the detection of pathologically documented asbestos-related pleural plaques or asbestosis is only 40-60% [1-5]. CT and high-resolution CT have been used for the detection and characterization of interstitial lung disease, including abnormalities in the pneumoconioses [6-8]. In diffuse lung diseases, high-resolution CT is more precise than chest radiographs for detecting interstitial abnormalities, and more able lo suggest a specific diagnosis [9-11]. Several small studies have correlated specific high-resolution CT features of asbestosis with his- HWBUI0009202 i 64 GAMSU ET AL. AJR:164. January 1995 lopathoiogically established disease J12-14). These studies have validated different high-resolution CT findings in clinical asbestosis, although some of these features can occur in individuals without interstitial disease [15-17]: The frequency of these different CT findings or their value in being able to suggest a diagnosis have not been established. We examined whether a subjective semiquantrtative scoring method of the severity and extent of the interstitial disease is comparable to a scoring method based on fhe cumulative pres ence of the differenthighriesolution CTabnormalities of asbes tosis in suggesting the diagnosis in patients or lungs with histologically confirmed disease. We also compared these two CT methods with chest radiographs in the same population. Materials and Methods Study Population The study group consisted of 30 patients, all with asbestos exposure in shipyards or construction. Patients were being evaluated for pulmo narysymptoms, referred for evaluation for medicolegal purposes, or had died with a documented history of asbestos exposure. There wre si* women and 24 men, with a mean age of 67 years (range, 50-92 years). Their mean length of occupational asbestos exposure was 21 yeans {range, 1-44 years), and mean latency (time between first exposure and time of study) was 53.5 years (range, 19-59 years). Four of the 30 patients (13%) had never smoked. Sixteen of the 30 (53%) had a malig nant neoplasm (bronchogenic carcinoma in 12, mesothelioma in three, gastric carcinoma in one) at the time of death, lobectomy, or biopsy. In six instances, (he lung was obtained from the coroner's office and stud ied postmortem. Limited cOnfcaJ data were available in this group. . Our criteria tor study entry were the availability of (1) a high-resolution CT of the lungs and (2) adequate noncancerous lung tissue for patho logic examination. H interstitial fibrosis was present, it had to be accompa nied by suffident asbestos bodies to fuffill our criteria for asbestosis. We did not indude cases where only tumor was present on the biopsy speci men, or cases with interstitial fibrosis that was Wstopathoiojpcaliy incon sistent with asbestosis. Twenty-one other cases referred over tire same time period did not fuffill ati the Study entry criteria, including five cases with fibrosis that did not meet our strict criteria for being asbestos related. Imaging Studies In six of the 30 patients, high-resolution CT scans were obtained on the fresh excised lung obtained at autopsy. In these cases, the lung was kept continuously inflated with air to 20.cm of water pressure measured with a water manometer. Sx to eight spaced scans were obtained on a 9800 GE CT scanner using high-resolution techniques modified for spec imen studies and an appropriately small field of view (18). In 24 patients imaged during life, high-resolution CT was done with the patients in prone and supine positions, using a 9800 GE CT or a Picker 1200XS CT scanner. Scans were 1.0- or 1.5-mm thick and obtained at six to eight spaced intervals through the thorax. The in vivo CT was performed between3days and 1 year before biopsy or autopsy. In patients who had ' lung tumors, the high-resolution CT scans were obtained at a time when the tumor was focal and did not interfere with the interpretation of the high-resolution CT scans for the presence of interstitial fibrpsis. Each CT examination was viewed simititaneousty by two racMogists who had no clinical or pathologic information, except for the history of occupational asbestos exposure- The findings from the high-resolution CT were agreed on by consensus of the two radiologists'. The subjectivesemiquantitative method of determining the probability of interstitial fibro sis being present and its severity was with a 4-point scale incorporating extent and severity and was as follows: 0 = normal, without interstitial lung disease; 1= a few sites (one to tour) of interstitial abnormality unlikely to represent diffuse interstitial fibrosis; 2 = multifocal abnormalities limited in extent, but in both hemithoraces, or at least two levels in one hemrthorax and consistent with asbestosis; 3 = profuse bilateral interstitial abnor malities visible at multiple sites and on at least several CT scans. In the interpretation of the in vivo high-resolution CT scans, a basal predomi nance of the abnormalities was uniformly present, making the interstitial findings consistent with the distribution found in asbestosis [6], The second method used different high-resolution CT abnormali; ties described in the interstitial fibrosis found with asbestosis. These had to be bilateral or on several scans in one hemithorax to be con sidered present and were as follows: (1) thickening of the interlobu lar septa and centrilobular core structures, grouped together as interstitial lines; (2) parenchymal bands (long scars); (3) subpleurai curvilinear opacities (subpleurai lines); (4) honeycombing; (5) subpleural nodules; and (6) architectural distortion (Figs. 1-4). The high-resolution CT finding of architectural distortion in interstitial . lung disease is not readily defined.[19], and we used a conservative ) i Fig. 1.--Patient with advanced asbestosis. Hlgbessoiuiion CT (can of an isolated lobe shows centrtiobular (curved arrow) snd Interlob- uiar (straightsolidarrows)intsrstiUs! thickening. Subpleurai nodule* (open arrows) and architec tural distortion are seen rst multiple sH&s. Histo pathologic diagnosis wee severe fibrosis. Ftg. 2.--Patten* with minimal asbestosis. High-resolution CT scan of an Isolated lobe shows only focal Interstitial thickening (arrows). The high-resolution CT was judged as mildly ab normal and unlikely to represent asbestosis. Fig. 3.--Patient with severe asbestosis. Highresduiton CT scan through the lower lobes dem onstrates extensive Interstitial disease, Including honeycombing, architectural distortion, thicken ing of interlobular septa and centrilobular cores, end long scars (parenchymal bends). HWBUI0009203 ? a*-3.s> S'S a ? ? S g g 1 9 s' 1 1 I 3 >r33 ") ' iis !( II \ <6*. January 1995 CT QUANTIFICATION OF INTERSTITIAL FIBROSIS 65 . .. ^noo of distorted secondary pulmonary lobules within peripheral ..,as 0f the lung parenchyma. All of the high-resolution CT abnorl';.(ie$ used were consistent with the previous descriptions of the -rersritiai fibrosis of asbestosls [12-171. The 24 in vivo high-resolution CT scans were also observed for -e presence of asbestos-related pleural disease (Table 1). The .-.tteni and severity of pleural plaques were ranked on a 4-point rate from 0 (or no plaques to 3 for the most extensive pleural r agues. Pleural calcification was rioted as present or absent. Chest radiographs were available, for.25 patients:.23 of the 24 A-th in vivo studies and two from the six In vitro autopsy excised ^ngs (Table 1). AH chest radiographs were obtained within 6 months :< me high-resolution CT examinations. The chest radiographs were -:erpreted by consensus by two observers (one a B-reader) using -ie ILO classification for lung disease consistent with asbesiosis 'small irregular opacities) [20]. Pathologic Examination The pathologic material was examined without knowledge of the esulls of the clinical or radiologic studies by a pathologist experi enced In occupational lung diseases. The material examined com prised autopsy specimens (16 patients/lungs), open lung biopsy specimens (six patients), and lobes excised at thoracotomy (eight patients). Histologic specimens obtained from transbronchial biop sies were not used for histopathologic diagnosis in this study. Blocks of pulmonary parenchyma away from sites of tumor were prepared from each specimen. Sections were stained with hematoxylin and eosln, or In those instances where paraffin blocks were available, for connective tissue with van Gieson or trichrome stains. Several slides (mean standard deviation, 4*3; range, .l-r-15) were examined from each specimen. The presence and severity of interstitial fibro sis were ranked on-a-semiquantitative 4-point severity scale: 0 = no interstitial fibrosis; 1 = mild patchy peribronchiolar and interstitial fibrosis (Fig. S); 2 = moderate peribronchiolar and interstitial fibrosis, without architectural distortion (Fig.-6); and 3 = extensive interstitial fibrosis with architectural distortion or honeycombing (Fig. 7). In this study, the histopathologic diagnosis of asbestosls required the pres ence of peribronchiolar or- interstitial pulmonary fibrosis, (severity scale of 1 or greater) and visible asbestos bodies. [21, 22]. Speci mens with Interstitial fibrosis were thus additionally stained with a Prussian blue stain for Iron, and ferruginous bodies were counted by scanning the sections serially magnified 100 times. In the context of asbestos exposure,' ferruginous bodies with a typical appearance were considered asbestos bodies [22). The surface area of the sec tion was measured by laying the slide on graph paper ruled at 100 lines per square centimeter and counting the squares covered by TABLE t: HRCT and Histology In Asbestosls Histologic Asbestosis Score (0-3) HRCT Asbestosis Score (0-3) HRCT Findings IL PB HC AD SL SN (+/-) (+/-) (+/-> <+/-) (+/-) '(+/-) Plaques Score (0-3) Ca (+/) X-Ray ILO Score 0 0 0 0 2 2 2 2 1 1 1 1. 0 1 3 3 2 3 3 3 2 2 3 3 3 3 3 2 3 3 .0 0 0 0 0 0 0 0 0 1 1 1 .1 -1 2 2 2 3 3. 3 3 3 3 3 3 3 3 3 3 3. -- - .- - + 4 -- - 4 4-' -- - -- * - 4- 4* ++ 4+4 - *- 4 -- 44> - 4- - +- -4 ++ 4- 4* + 4* 4 - 4 4 + 4 + 4* - 4 4 4- 4 4- 4 4 4444 - 4 4 + 4 4 4* 4 4- - - - 4 4 4- 4- + 4* 4 4 + 4- 4 4 ' + 4- + 4- 4- 4- 4 4 - 4 4 4* 4 4 - 4- - 4* 4 + 44 . 4 4- * 4 4- - 1 . . 0/0 2 4- 0/1 * 0/0 * NA 0 - 0/0 1 - 0/0. 0 - 0/1 1 - 0/1 1 . " 1- 0/0 0/0 NA * . 0/0 0 - 0/0 0 - 0/1 2 . - 0/1 3 4- 1/0 2 4 1/1 1 - 2/2 i - 2/3 i 4- 2/2 '3 2/3 NA * NA 1 - 0/0 2 3/3 3 4- 3/3 3 4- NA 2 4 1/2 1 - 1/1 2 212 (0-3) = 4 point scale. (+/-) = present or absent. NA = nor available, = ex vivo scans. IL = interstitial lines. PB = parenchymal bands. HC = honeycombing. AD = architectural distortion. SL = subpleural lines. SN = subpleural nodules. Ca = calcification, ILO = International labor Organization. . HWBUI0009204 66 GAMSUETAL. AJR:164, January 1995 Fig. 4.--Patient with asbestosls with a nor mal sliest radiograph and abnormal CT scan. A, Conecfdown view from the radiograph of the lower lobe of the right tung has a normal ap pearance. B, Hlgh-rasolutloh CT scan through lung bases shows bilateral multifocal areas of inter stitial thickening (arrows). The CT scan was Judged as being moderately abnormal, whereas histopathologic examination showed advanced fibrosis. Fig. 5.--Asbestosls with mild tntenttttjel fibrosis. Photomicrograph of a histologic section shows peribronchiolar (open arrow) and Interstitial (cloned arrow) hbroala. Fig. 6.--Asbestosls with moderate Interstitial fibrosis. Pttotomlcmgraph of a histologic specimen shows mode/ate peribronchiolar (open arrows) and Interstitial (closed snows)fibrosis without architectural distortion. Fig. 7.--Asbestosls with marked IntersttHal fibrosis. Photomicrograph of histologic specimen shows architectural distortion and honeycombing (enlarged Abiotic spaces). ' the section. A minimum of 2 cm2 of lung parenchyma was examined. Results were expressed as number of asbestos bodies per square centimeter. Interstitial fibrosis with a count of greater than 0.5 asbes tos bodies per square centimeter was required for the diagnosis of asbestosis [21J. Histopathologic peribronchiolar and interstitial fibro sis together with asbestos bodies sufficient for a diagnosis of asbes toses were found in 25 of the 30 cases (83%) (Table 1). Analysis We first tested the relationships between the two high-resolution CT methods of assessing asbestosis and the histologic presence ol this disease. We tested the correlation between the first method using the subjective semiquantitative severity score and the histo pathologic score using Spearman's rank correlation or, when appli cable. Fisher's exact test. Statistical relationship between the histopathological fibrosis score and the second method that used the cumulative CT features was established with chi-square analy sis. We used a standard computerized statistical package (SAS). Statistical comparisons accepted significance at the 0.05 level. Results Of the 25 cases with histopathologic asbestosis, the semi quantitative high-resolution CT score was severely abnormal (grade 3) in 13; moderately abnormal (grade 2) in three; and mildly abnormal but judged unlikely to represent diffuse fibrosis (grade 1) in four (Table 1). The high-resolution CT showed insuf ficient findings to suggest fibrosis (grade 0) in five of- the 25 cases with asbestosis. The semiquantitative grade of asbesto sis on high-resolution CT was significantly associated with the histopathologic fibrosis severity score (r=.78, p = .0001). When we dichotomized these semiquantitative high-resolu tion CT results by combining high-resolution CT grades 0 and 1 (no or unlikely asbestosis) and grades 2 and 3 (consistent with or probable asbestosis), there were 14 grade 0 or 1 cases and 16 grade 2 or 3 cases. Of the 14 patients with high-resolution CT scans graded 0 or 1, five did not have histologic asbestosis and nine did, whereas all 16 of the patients with high-resolution CT scans graded 2 or 3 had histopathologic asbestosis [p = .Q07). The high-resolution CT abnormalities involving the interstitium most commonly associated with the presence of histologic asbestosis were interstitial lines (interlobular and centrilobular thickening), parenchymal bands, and architectural distortion of pulmonary lobules. Interstitial lines were seen in 21 (84%) and parenchymal bands were seen in 19 (76%) of the 25 cases of asbestosis. Among the five cases without histologic asbestosis, ) HWBUI0009205 . ,, (64, January 1995 CT QUANTIFICATION OF INTERSTITIAL FIBROSIS 67 merstitial lines were found in one and parenchymal tends in two. Architectural distortion was present in 14 of the 25 asbestosis cases (56%) and was not seen in any of the five cases without mstopathologic interstitial lung fibrosis. Thirteen of the 14 cases .viih architectural cfistortion fell within the histopathologic fibrosis severity categories 2 or 3. Honeycombing seen on htgh-resoiurion CT in patients with asbestosis has a prevalence of less than 10% 116]. In this selected population, honeycombing was found ,n eight cases (32%), all of which, had severe histopathological asbestosis and the highest scores with the semiquanfitatrve CT scoring method. A subpieural line or band has been described with diffuse interstitial fibrosis {12].- This finding was observed in H cases (44%), all with asbestosis In histopathologic severity categories2or3,andin10ofthe11 with the highest high-resoiution CT severity score using the semiquantitative method. When these five CT abnormal findings were analyzed separately, inter stitial fines and architectural distortion occurred with sufficient fre quency to have an association with histopathologic asbestosis that independently achieved statistical significance. Subpieural noduies have been described as a manifestation of interstitial lung disease [23]. This finding was present mil of the 25 cases with asbestosis and none without asbestosis (Table 1). For high-resolution CT to suggest a diagnosis, we tested toe postulate that a combination of features could be more useful than a single finding. Therefore, we evaluated whether a cumu lative number of high-resolution CT findings would facilitate toe diagnosis (Table 2). Because of an increasing number of highresolution CT features, the frequency with which toe diagnosis of asbestosis was identified decreased. One of toe selected live abnormalities was found in 88% in patients with asbesto sis. two abnormalities in 78%, three In 56%, four in 53%, and all five in 22%. However, with an increasing number of different types of high-resolution CT abnormalities, toe likelihood of asbestosis being present histopathologically increased from 60% with one abnormality to 100% with three or more abnor malities. Because the most common CT features of asbestosis were also present in patients without interstitial fibrosis, a cumulative number of CT abnormalities eliminated potential false-positive cases and increased toe likelihood of asbestosis being present Wheri honeycombing, architectural distortion, or subpieural lines were seen on high-resolution CT, asbestosis was invariably present histopathologically in this group. Pleural plaques were found on CT In 20 of toe 24 in vivo studies. Calcification was present in plaques on CT scans in seven (35%) of these 20. All 14 living patients who were in highresoiutton CT grades 2 or 3 (likely asbestosis) had visible pleural plaques. In contrast, six of toe 10 patients with high-resolution TABLE 2: Cumulative High-Resolution CT Findings In Asbestosis No. of CT Findings 1 2 3 4 5 Proportion ol Cases of Asbestosis with ,, Findings (%} Proportion of Cases with Asbestosis (%) 88 60 78 80 56 100 53 100 22 100 Noie.--Findings used are interstitial tines, parenchymal bands, honey combing. architectural distortion, and subpieural lines. CT grades 0 or 1 (unlikely asbestosis) demonstrated plaques (p = .02). Among the 24 in vivo studies, there were three cases in which histopathologic asbestosis was present, but no pleural plaques were visible on CT scans. In none of these three was asbestosis suggested from toe high-resolution CT scans. Chest radiographs were available in 25 cases, two from cases in which the in vitro excised lungs were studied, and 23 from the in vi.o high-resolution CT studies (Table 1). Findings on toe chest radiographs were abnormal and consistent with asbestosis (ILO profusion > 1/0) in 11 of toe 25 patients, all with asbestosis in toe highest two histopathologic fibrosis catego ries. However, lOoftoe 14 patients with an ILO profusion score of 0/1 or less had histopatoologicasbestosis, and four did not have asbestosis (p=.10). Overall concordance between histo pathologically confirmed asbestosis and chest radiographs defined by an ILO grade of 1/0 or greater was 11/21 (48%). Discussion In this study, we confirmed a significant correfation.between the abnormal high-resolution CT findings described in asbes tosis and the histopathologic presence of disease. Conventional CT and high-resolution CT have been widely accepted for their diagnostic advantages, both for focal and dif fuse lung diseases. The abnormalities seen in patients -with asbestosis have been described, and their correlation with the presence of clinical asbestosis established [12-17]. In general, patients with clinical asbestosis (exposure, abnormal findings on radiographs, abnormal results of pulmonary function studies) have a high probability of having abnormal findings on highresolution CT scans. The extent of functional impairment in these cases also correlates with the severity and extent of CT abnormalities [16]. In patients occupationally exposed to asbes tos and with normal findings on chest radiograph, high-resolution CT can define a subgroup of patients with reduced lung function compared with those with normal high-resolution CT findings [5]. The distribution and types of high-resolution CT interstitial abnormalities found in asbestosis tend to differ from those seen in idiopathic interstitial fibrosis [24]. Ground-glass opacities are common in idiopathic interstitial fibrosis and uncommon with asbestosis. The reverse is true for parenchymal bands (long scars) and subpieural lines. We found that limited Interstitial lines and parenchymal bands could .be present without indicat ing, the diffuse or multifocal interstitial fibrosis of asbestosis. Thus, either or both of these abnormalities, if not accompanied by additional findings, could be deceptive in suggesting a diag nosis of asbestosis. Only by.having, three abnormal features were all cases without fibrosis excluded. The inflammatory response and fibrosis produced in toe lung by the inhalation of asbestos fibers result in various types of high-resolution CT findings. These are dependent on the stage and severity of the interstitial fibrosis, Our results agree with the study by At-Jarad and colleagues [24] that, in asbestosis, the variety of abnormal ities on high-resolution CT scans can have diagnostic signifi cance: In fact, this cumulative scoring system gave comparable results to the usual subjective, semiquantitative assessment based on the extent and severity of toe interstitial disease. We also found that the subjective semiquantitative scoring system showed a higher detection rate for asbestosis than did chest radiographs classified with the ILO classification (62% vs 48%). The case mix in our series is most likely skewed by con- HWBUI0009206 68 GAMSU ET AL. AJR:164. January 1995 taming a disproportionate number of patients with minimal asbestosis undergoing Sung resection tor cancer and patients with severe asbestosis from which they died. Only six patients had mild asbestosis (ILO categories 0/1 or 1/0) in which highresolution CT may be expected to show diffuse lung disease not apparent from chest radiographs (Table 1). Asbestosis was present in five of these six patients and was shown by high-res olution CT in two. The high-resolution CT sampling technique used in this study could have missed focal areas of interstitial fibrosis, even though we used standard high-resolution CT techniques. Given the slow progression of asbestosis, the time between the CT scans and histopathologic confirmation is . unlikely to account for the missed cases. We also found that asbestosis could be present histopathologically with limited and focal high-resolution CT abnormali ties or with completely normal findings on high-resolution CT scans (nine of 25 cases of asbestosis). Cases of interstitial fibrosis unrelated to asbestosis werenot included in this study, and we cannot comment on the radiographic or high-resolu tion CT distinction between asbestosis and other forms of interstitial lung disease. Our cases preclude collation of the clinical information such as pulmonary function studies with the results of the Imaging studies, or the effects of smoking. For high-resolution GT to be more effective than chest radio graphs in the diagnosis of asbestosis, as with other interstitial fforotic lung diseases, the detection of abnormalities too small or too limited to be visible on radiographs becomes important. How ever, a threshold or lower tart of the extent or severity of abnor mality must be considered. With either the semiquantitative scoring system or the cumulative number of high-resolution CT features, limited abnormalities were not indicative of diffuse or multifocal interstitial fibrosis. About 15% of occupationally exposed individuals who dem onstrate radiographic abnormalities consistent with asbestosis do not have radiographically visible pleural plaques. Autopsy studies, however, have shown that pleural plaques are much more common than asbestosis and in most Instances cl ashesfosis plaques are present but may be too small to be seen radio graphically (5). Because CT is highly sensitive for detecting plaques, lung parenchymal abnormalities consistent with asbes tosis in occupationally exposed individuals are distinctly unusual without demonstrating these plaques [16]: The 14 patients with asbestosis who were in high-resolution CT semiquantitative cat egories 2 or 3 demonstrated plaques on CT. Of the ten patients in semiquantitative categories 0 and 1, six showed pleural plaques and four did not. Although asbestosis was present in seven of these 10, tire diagnosis would not have been sug gested. from high-resolution CT scans. This supports the idea that plaques will usually be present when the high-resolution CT scans show multifocal features consistent with asbestosis. In summary, using asbestosis as a model of a specific type of interstitial pulmonary fibrosis, we have made several observations. In suggesting the diagnosis, usually limited and unilateral interstitial tines or parenchymal bands can be seen in individuals without hfetopathotogical asbestosis and should not be considered by them selves as suggestive of the diagnosis. Abnormalities need to be bilateral or multifocal. High-resolution CT methods of assessment that used either a cumulative number of types of abnormalities or a semiquantitative score of extent and severity showed equivalency with the histologic confirmation of asbestosis. Thus, we propose that the various types of high-resolution CT abnormalities could be used in conjunction with a subjective assessment of extent and severity for suggests a diagnosis in thisjype of interstitial fibrosis. In this study, albeit a stewed population, high-resolution CT was slightly superior to ILO classified chest radiographs for detection of abnormalities consistent with asbestosis. Finally, hisfopathologically proved asbestosis can be present in patients with high-resolution CT scans that show-normal or nearly normal findings. REFERENCES 1. Sefikofl 1J, Lea DHK, ads. Asbestos and disease. San Diego. CA: Aca demic Press, 1978 2. GaendarEA, CarringtonC8. Coutu HE, et aL Pattmtojeal physotogtail, and racft> \bgcsl corafafiors m tfpprreurriocantoses. AmNYAcadSd 1962200:574-60? 3. Kipen HM, tills R, Suzuki , Valdukas JA, Selikoff IJ. Pulmonary fibrosis in asbestos insulation workers with lung cancan a radiological and tiistopathologlcal evaluation. Sr J Ind Med 1987;44:96-100 4. RodasffDS. Schwartz A. Btsan^anoqrepMe undares&nation ol early asbestosis bylLOdasatatioreanalyssordata'andpnibabSties. Gtef1988:93;t088-1099 5. 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 Re4pJrD/s1B89;139:1502-150B . 6. Katz D, Kreel L: Computed tomography in pulmonary asbestosis. Ctai Rsdfof1979;30:207-2t3 7. Remy-Janfin M. Degree! C, Beuscart R, Voisirt C, Remy J. Coal workers' pneumoconiosis: CT assessment in exposed workers and correlation with radiographic findings. (?adfotogy.l90;l77:363-37l - 8. Sperber M, Mohan KK. Computed tomography: a reliable diagnostic modality in pulmonary asbestosis. Comput Radian984:3:125-132 9. Mathieson JR, Mayo JR, Staples CA, MOIIar NL Chronic diffuse infiltrative rung disease: comparison of diagnostic accuracy of CT and chest radiog raphy. Radiology 1989;171 -.111-116 10. MQOerNL Miller RR. Computed tomography of chronic (Effuse infiltrative lung disease (pts 1 and 2). Am Rev HaspirD'1990;143:1200-1215,1440-1448 11. Murata K, Khan A, Herman PG. Pulmonary parenchymal disease: evalua tion with high-resolution CT. Radiology 1989;170:629-635 12. Yoshimirra J, Halakeyama M.. Otsuji H. et al. Pulmonary asbestosis: CT study of subpieural curvilinear shadow. Radiology 19S6; 158:S35-53 13. Aldra M, Yamamoto S, Yokoyama K, et al. Asbestosis: Wgh-resolution CTpathologio correlation. Radiology 1990;176:309-394 14. Akira M, Yokoyama K, Yamamoto S, el al. Early asbestosis: evaluation with high-resolution.GT. Radiology 1B91;178:409-416 15. Aberle DR. Gamsu G. Ray CS.etal, Asbsstos-relaied pleural and parenchy ma) fibrosis: detection with high-resoluBon CT. RaMogy1988:166:729-734 16. Aberle DR, Gamsu G,.Ray CS-High-resoMion CT of benign asbestos-relaled diseases: dinicat and rediogtaphie cotraration. AJR 1988,-151:883-691 17. Friedman AC, Rd SB, Fisher MS,' et aL Asbestos-related pleural disease and asbasosis: a comparison of CT.and chest radiography. AjR1S88;1S0269-275 18. Webb WR. SteTnMG,-FmkbeinerWE. et al. Normal and diseased isolated lungs: high-resolution CT. Radiology 1988;166:81-87 19. Gamsu G. The lungs. In: Moss AA, .Gamsu G, Genant HK. eds. Computed lomogmphyofthe body. Val. 1,2nd ad. PMadelphia: Saundeis, 1992:171-172 20. Classification of radiographs olthe pneumoconioses, trc RusseOAR ed.Msdbtf mdkogmphyarripMography.vd. 57. Rochester, NY: Eastman Kodak, 19812^17 21. Wamock ML Lung asbestos burden In shipyard and construction workers with mesothelioma: comparison with burdens in subjects with asbestosis or lung cancer. Environ Res 1989;50:68-85 22. Churg AM, Wamock ML Asbestos and other ferruginous bodies: their for mation and clinical significance. Am J Pathol 1981 ;102:447-456 23. Remy-Jardin M, Beuscart R, Sauh MC. et al. Subpieural micronodules irr diffuse infiltrative lung diseases: evaluation with thin-section CT scans. Radiology 1990;177:133-139 24. Al-Jarad N, Strickland B, Pearson MC. Rubsns MB, Rudd. RM. High-reso lution computed tomographic assessment of asbestosis and cryptogenic fibrosing alveolitis: a comparative study. Thorax 1992;47:645-650 The reader's attention is directed to the commentary on this article, which appears on the following pages. :________________________________________________________________________ 1 j HWBUI0009207