Document 8O3BBxmJBOONJ07Rwe3X4NZ8d

British Journal of Industrial Medicine 1987;44:96-100 Pulmonary fibrosis in asbestos insulation workers'*with lung cancer: a radiological and histopathological evaluation H M KIPEN,* RUTH LILZ5, Y SUZUKI, J A VALCFUKAS, I J SELIKOFF From the Environmental Sciences Laboratory, Mount Sinai School ofMedicine ofthe City University ofNew York, New York, USA abstract This study was undertaken to determine the relation between radiographic and histological manifestations of pulmonary asbestosis (interstitial fibrosis) in insulation workers who had died of lung cancer. Of 450 confirmed deaths from lung cancer a chest radiograph suitable for determining evidence of pneumoconiosis was obtained in 219. Of these cases, 138 also had a tissue specimen submitted that was suitable for histological study to determine the extent of histological fibrosis. There was a significant albeit limited correlation between the radiographic and histological findings (r -= 0-27, p < 0-0013). All 138 cases had histological evidence of parenchymal fibrosis; in 25(18%), however, there was no radiographic evidence of parenchymal fibrosis. In 10 cases (7%) both parenchymal and pleural disease were undetectable on the radiograph. Thus a negative chest radiograph does not exclude the presence of interstitial fibrosis (asbestosis) in a substantial propor tion of insulation workers previously exposed to asbestos who develop lung cancer. Cancer of the lung is a leading cause of death in workers exposed to asbestos.' For example, it has been found to account for as much as one fifth (21%) of deaths in a large prospective mortality study of asbestos insulation workers.6 In individuals with lung cancer there are many different ways to assess retrospectively whether or not they have had significant exposure to asbestos. In the clinical setting exposure is often inferred from the presence of non-malignant effects of asbestos exposure: pleural fibrosis and interstitial pulmonary fibrosis. Histopathologically, asbestosis includes diffuse interstitial pulmonary fibrosis developing as a con sequence of inhaling asbestos fibres; it is usually asso ciated with the presence of asbestos bodies, although these may vary in number and ease of demonstra tion.7 Clinical studies in manjiowever, often do not use histopathological criteria for the diagnosis of asbestosis but the more easily obtained and less invasive chest x ray (and clinical) data. Such radio- Present address: Department of Environmental and Community Medicine, UMDNJ--Rutgers Medical School, Piseataway, NJ 08854. Accepted 14 April 1986 logical and clinical evidence has been reported to be less sensitive to the presence of asbestosis (and other types of interstitial fibrosis) than histopathological examination.8 9 This study examines further the relation between the radiographic and histo pathological evidence of asbestosis in a population of asbestos insulation workers with lung cancer. As part of a continuing cohort mortality study of 17800 members of the insulation workers' union (International Association of Heat and Frost Insu lators and Asbestos Workers, AFL-CIG, CLC), we examined deaths from lung cancer which occurred between January 1967 and September 1983. Methods IDENTIFICATION OF CASES FOR ANALYSIS On learning from the union of a member's death, we requested death certificates, medical records, chest radiographs, pathology reports, and surgical nec ropsy pathological specimens, or both, from hospitals and attending physicians. These were reviewed and only after pathological confirmation of the presence of primary lung carcinoma was a case eligible for the further analyses described here. 96 Pulmonary fibrosis in asbestos insulation workers with lung cancer 97 RADIOLOGICAL AND PATHOLOGICAL ASSESSMENT Of the 450 pathologically confirmed lung cancers observed by September 1983, we obtained chest radiographs from 229, Of these, 219 were adequate for assessing interstitial pulmonary fibrosis. The chest radiographs had been taken at various times, any where from immediately preceding death to a few years before. When multiple films were available, the film that showed the earliest evidence of carcinoma was interpreted for the purposes of this study. To avoid confusion of radiological changes related to tumour with non-malignant changes of fibrosis, the radiological abnormalities in the hemithorax con tralateral to the side of the tumour were interpreted according to the ILO 1980 Internationa! Classification of Radiographs of Pneumoconioses.10 According to the International Classification, cate gory of profusion is based on assessment of the num ber of small opacities per surface unit by comparison with standard radiographs. Use of the ILO Inter nationa! Classification allows stratification of radio logically detectable small opacities (interstitial fibrosis) into the 12 catagories ranging from 0/- to 3/4 (0/0,0/1, 1/0,1/1,1/2, etc). Pleural changes were also graded in the standard fashion regarding width and extent according to the International Classification. Initial chest x ray readings (by RL and HK) were made in the absence of any clinical information except for the date and cause of death--lung cancer. Where no tumour was radiologicatly evident, where previous therapeutic intervention was dearly evident, or where a diagnosis other than lung cancer and asbestosis was considered, clinical and pathology records were reviewed to provide information about the side of the cancer. Films taken before the onset of terminal illness often helped. On six films, the paren chyma was assessed on the side of the tumour because of obscuring contralateral disease such as pleural effusion or cardiac enlargement. On seven additional films, tumours were either not seen or were present bilaterally, so a contralateral side could not be identified. In all these cases we were confident of the accuracy of the (possibly) ipsilateral reading, and these cases are included in the analysis. All findings were reviewed separately with a third reader (IJS), and the consensus of the three readers is reported. All cases have, by definition, a pathological diagno sis of lung cancer by virtue of histological exam ination of tumour tissue. Adequate tissue for micro scopic assessment of the degree of parenchymal fibrosis, however, was not always obtained, as when the diagnosis of neoplasm was established by bron- choscopic biopsy. Of the 219 cases with adequate radiological data, 138 had enough non-malignant pulmonary tissue submitted as part of a surgical specimen, a biopsy, or a necropsy to permit the assessment of the presence of absence of diffuse inter stitial pulmonary fibrosis and its degree, on the level of light microscopy. It is in these 138 cases that we could compare the radiological and microscopical evidence of asbestosis. Consistent with our tissue sources, pathological analysis was often done on the side ipsilateral to the tumour. Grading of the degree of parenchymal fibrosis was performed by YS according to a seven point scale: Mw--No histological evidence of interstitial fibrosis. Minimal--Rare presence of focal, mild thickening of the interstitium.* Mild--Presence of multiple foci of an increased amount of fibrosis in the interstitium. ,, Moderate--Interstitial fibrosis associated with obliteration of alveoli and showing early consoli dation. Severe--Pronounced fibrosis associated with con solidation, honeycomb formation, and obliteration of airways. , `The imersiitiunj includes the peribronchiolar, perivascular, alveolar septal, and sitbpteurai connective tissues. Interestingly, however, asbestos related diffuse pulmonary interstitial fibrosis starts in the peribronchiolar connective tissue. Histopathotogical and radiologicalfindings, ( Complete absence ofx ray changes including pleural abnormalities shown in parentheses) X ray profusion ofsmall opacities Histopathotogicalfibrosis No ofcases Minima! Mild Moderate Severe Totals 0/0 0/1 1/0 1/1 1/2 2/1 2/2 2/3 3/2 3/3 Totals 5(2) 5 2(1) 11(6) 7(1) 20 i1 5 6 13 i 2 20 16 "39 6 6 12 1 5 ? 13 1 10 ? IS I3 4 25 7 ]6 7 2 7 66 63 138 ,98 Kipen, Lilts Suzuki, Vatciukas, Selikoff Fig 1 Distribution patterns ofradiological small irregular opacities (median and mean SD derivedfrom the / to 12 code) and ofhistopathohgwal changes (median andmem SD derivedfrom the 1 to 6 code usedfor numerical transformation of severity offibrosis). (See text.) Intermediate categories of mild-moderate and Results moderate-severe were read in a few cases, yielding seven gradations of fibrosis. (For representational Of the 138 cases, 113 (82%) had radioiogicaliy purposes only the four major categories are shown in detectable small irregular opacities (ILO profusion the table, with intermediate readings included in the equal to or greater than 1/0). The mean degree of next higher category.) profusion was 5-8 (SD 2-3), corresponding to between 1/1 and 1/2 on the ILO scale (fig!). In all 138 cases there were histopathologicai signs statistical analysis of diffuse interstitial fibrosis. The mean degree was 4-7 The major radiographic variable analysed was the (SD 1-3), corresponding to between moderate and category of profusion of small irregular parenchymal severe on the scale (fig I). One or more asbestos bod opacities. Categories of profusion beginning with 0/ ies were seen in 94% of cases (130/138). were assigned sequential integer values front 1 to 12 coded such that 0/- = I, 0/0 = 2, 0/1 =* 3, and so on up to 3/4 12. Pathological gradings of fibrosis from Profusion of radioiogicaliy detectable small irregu lar opacities was significantly correlated with the degree of microscopic fibrosis (r 0-27, p < 0-0GI3) "none" to "severe" were similarly assigned successive (fig 2 and table). integer values from 0 to 6. Of the 138 cases, 94 (68%) had a radioiogicaliy The presence or absence of asbestos related pleural detectable pleura! abnormality (circumscribed or thickening, including diaphragmatic plaques, was the ^diffuse pleural fibrosis, calcification). other major radiographic variable analysed. The The. presence/absence of any pleural abnormality width of tangential pleural thickening was also scored and treated as a separate variable. Descriptive statistics were calculated for each vari on the chest radiograph was not significantly cor related with the profusion of irregular opacities (r = 0-09, p = 0-29) nor with the extent of micro able. Because the variables were ratings, non- scopic fibrosis (r = 0-12, p - 0-16). parametric correlational procedures were used. The width of tangential pleural thickening was Spearman correlation coefficients were calculated for significantly correlated with profusion of small irreg all pair combinations of variables using the SAS ular opacities (r = 0-33, p < 0-0075) but not with program. microscopic fibrosis (r = -- 0 09, p = 0-46). Pulmonary fibrosis in asbestos insulation workers with lung cancer 99 g 3/3|3/2- 5 2/3g 2/2S2fl- r=Q-272 p<0-0013 with the degree of microscopic fibrosis (r = 0-11, p = 0-22) (fig 3). Duration from onset of asbestos work to death in the 25 radiographically negative cases (32'8 y) was not significantly less than that of the group as a whole (34-1 y 8-4 y). g 1*3 i/i- i. i/o- * H 0/1' * * m Discussion The first reports associating lung cancer with asbesr tosis appeared in 1935115 2 and many studies since then have convincingly shown an increased incidence | 0/0 j of lung cancer in asbestos exposed workers.1 ~5 All ' 1-----------fTM Min Mild common cell types occur more frequently.1314 T '''' 1---------- '' 'l Mod Mod Severe In this large study of exposed asbestos insulation Severe workers parenchymal fibrosis was present in all those Degree of histopathotogieol fibrosis Fig 2 Correlation between radiographic and histopathologicai assessments ofinterstitial pulmonary fibrosis in 138 insulation workers with lung cancer fr = 0-27, with lung cancer whose pulmonary tissues were microscopically examined. Jt has been estimated that asbestos exposures in this group were generally in the range of 4~15 fibres per ml.15 Because of the large size p < 0 0013). of this cohort (17 800, with 485 of the first 2275 deaths In 18% of the cases {251138} interstitial pulmonary fibrosis of asbestosis was not radiographically detectable (small Irregular opacities not seen) although diffuse interstitial fibrosis was invariably present on microscopic examination {1381138} (table). Pleural thickening and plaques were also absent in f 0 of the 25 cases without radiographically detectable interstitial fibrosis, making 7% of the 138 films negative for any evidence of benign asbestos related pleuropulmonary changes. Pulmonary fibrosis by histopathologicai assessment was moderate or severe in nine of these ten. Duration from onset of asbestos work to death was significantly correlated with radiological profusion of small irregular opacities (r = 0-24, p < 0-006) but not due to lung cancer6), it seems probable that a large majority (perhaps all) of the lung cancers developing in similarly exposed workers will occur in the pres ence of microscopic fibrosis. This is confirmed by the recent report that 90% of a cohort of British asbestos factory workers who died with lung cancer had histo logical evidence of asbcstosis.5 Radiological data were not"reported for this group, In the insulation workers whom we are following up, an increasing radiographic profusion of small opacities was found with increasing duration from onset of exposure. Longer exposures presumably rep resent greater cumulative exposure to asbestos dust. No such correlation was found between duration from onset of exposure and the pathological grading. This may be explained by the high proportion of cases with moderate or severe fibrosis (93%). Most of these cases had probably passed some relative threshold of exposure beyond which moderate fibrosis usually develops. Analysis of groups with shorter duration from onset of exposure (10 to 20 years) might well yield a higher correlation between the amount of fibrosis on histopathologicai study and the duration from onset as a measure of exposure. It is well known that for interstitial pulmonary fibrosis to become radiologkaily detectable, a certain amount of fibrotie change has to haveJeveloped; the summation effect will then produce the irregular opacities visible on the posteroanterior chest x ray film, allowing the diagnosis of interstitial pulmonary fibrosis to be made. The same summation effect oper ates in the case of small rounded opacities, character Duration tram onset of exposure (years) Fig 3 Correlation between durationfrom onset of insulation work {in years} and radiographic profusion ofsmall irregular istic of silicosis or coal worker's pneumoconiosis. Unless a certain minimum number (and size) of nod ules have developed, rounded opacities do not opacities (r = 0-24, p < 0-006). become recognisable on the radiograph. 100 Kipen, Lilis, Suzuki, Valciukas, Seiikoff Interstitial fibrosis was*_tno radiographically Conclusions detectable in the parenchyma of 25 (18%) of our 138 cases with lung cancer. Pleural abnormalities associ ated with asbestos exposure (pleural thickening, plaques, or calcification) were also absent in 10 of the 25 cases in whom parenchyma! fibrosis was radio logically undetectable, making the presence of asbes tos related disease completely hidden (in the absence of tissue examination) in 7% of the 138 cases of lung cancer in this series. Published reports on the relationship between radiographic and histopathological manifestations of asbestos related disease are sparse. Although no study has previously reported on this relation in individuals with lung cancer, our findings are in accord with two large studies of patients with various intersitial fibroses including asbestosis. From an unselected series of patients with chronic diffuse infiltrative lung disease, six of 58 patients (!0%) with asbestosis We have found that in the presence of lung cancer occurring in asbestos insulation workers with occupational asbestos exposure asbestosis was invari ably present when 'histopathological findings were used as the criterion for diagnosis. Discrepancies in the results of radiological and pathological exam ination for interstitial fibrosis were present in 18% of those heavily exposed insulators and may be even more sion frequent at of pleural lower levels thickening of exposure. The inclu-: as an indicator of s significant asbestos exposure may help to reduce, but will not eliminate, errors made on this basis. These findings indicate the primacy of the history of asbes tos exposure, irrespective of the presence or absence ft 6of non-malignant x ray changes (asbestosis) when considering lung cancer possibly associated with occupational exposure to asbestos. confirmed by biopsy had normal radiographs.8 Simi References larly, 56 of 354 patients (15-8%) with a pathological diagnosis of interstitial lung disease had normal chest 1 M erewether ER A. A nnual report ofthe eltif inspector offactories. London; HMSO, 1947. radiographs-9 Eight of this second group had asbes tosis and the remaining patients had a variety of infiltrative disease on biopsy, The probability that interstitial fibrosis will not be 2 Doll R. Mortality from lung cancer in asbestos workers, Sr J Med 1955;12:81-6. i McDonald JC, Liddell F, Douglas K. Mortality in Canadian miners and millers exposed to chrysolite. Arm NY Acad Set 1979;330:1-9. radiologically detectable in a sizeable proportion of cases of cancer is of considerable significance, especially in the evaluation of a potential aetiologicai 4 Newhouse ML, Berry G. Patterns of mortality in asbestos factory workers in London, Ann NT dead Set 1979;330:53-60. J Newhouse M, Berry G, Wagner JC. Mortality of factory workers In east London 1933-80, fir J fnd Med 1985;42;4~ 11, association between exposure to asbestos and a specific lung cancer. Radiographs negative for evi dence of asbestosis related findings do not exclude the presence of interstitial pulmonary fibrosis as ascer tained by histopathological examination. 6 Seiikoff IJ, Hammond EC, Seidman H. Mortality experience of insulation workers in the United States and Canada. 1943-1976. Ann NY Acad Set 1979;339:91-116. 7 Wamock ML, KuwaharaTJ, Watery O- The relation of asbestos burden to asbestosis and lung cancer. Pathology Annual 1983;18:109-43, ._ In occupational groups with lower asbestos exposure than that in. factory and insulation work lung cancer without radiologically detectable paren chymal opacities may represent a larger proportion of 8 Epler OR, McLoud TC,Gaensler EA, Mtkus JP,CarringtonC8 Normal chest roetgenograms in chronic diffuse infiltrative lung disease, N Engl J Med 1978;298:934-9. 9 Gaensler EA, Carrington C8. Open biopsy for chronic diffuse infiltrative lung disease clinical, roeotgenographie, and physio cases than in such highly exposed groups. Continuing studies indicate an increased incidence of lung cancer (RR 2) in family contacts of asbestos products manufacturing workers.16 The analysis of the pathol ogical data is incomplete at present; however, 17% of these household contacts had parenchymal abnor malities on chest radiography.17 We are currently investigating the histo logical correlations in 302 patients. Am Thorac Sttrg 1980;30:411-26. 10 International Labour Office. Guidelines far the me of 1LO Inter national Classification of Radiographs Of pnenmonconioses. Geneva; ILO, 1980. (Occupational safety and health series No 22 (rev 80).) 11 Lynch KM, Smith WA. Pulmonary asbestosis. III. Carcinoma of lung in asbestos-silicosis. Am J Cancer 1935;24:56-64. 12 Gloyne SR. Two cases of squamous carcinomas of the lung occurring in asbestosis. Tubercle 1935;17:5-10. pathological pulmonary changes in deaths other than those from lung cancer in the insulation worker cohort to see what proportion of workers have evi dence of asbestosis, independent of whether or not j 13 Kannerstein M, Churg J. Pathology of carcinoma of the lung associated with asbestos exposure. Cancer 197230:14-21, 14 Churg A. Lung cancer cell type and asbestos exposure. JAMA 1985;253:2984-5. - 15 Nicholson WJ, Case study I; asbestos-She TLV approach. Ann they developed lung cancer. The question of whether or not asbestosis always precedes lung cancer in sub jects exposed to asbestos is unresolved, and our NY Acad Set 1976;271:152-69. 16 Lilis R, Seiikoff IJ. Asbestos disease in occupationally-exposed workers and household contacts. Zur Beurteilung der Krebsgef.ihr durch Asbest Inslit fur Wasser-, Boden- and Luffhygiene present data do not permit conclusions regarding the aetiologicai relation between lung cancer and pul monary asbestosis per se, as distinct from that of concurrence between lung cancer and exposure to des Bundesgesunheitsamtes. Manfred Fischer und Eckarl Meyer (Herausgeber). Munich: MMV Medizin Verlag, 19*4:117-120. 17 Anderson HA, Lilis R, Daum SM, Seiikoff IJ. Asbestosis among household contacts of asbestos factory workers. Arm NY Acad asbestos. Sc: 1979;330:387-400.