Pathology of asbestosis--an update of the diagnostic criteria: report of the Asbestosis Committee of the College of American Pathologists and Pulmonary Pathology Society.(Report)

The observations and conclusions reported here represent the work of an international committee of North American, European, and Australasian pathologists, organized under the auspices of the College of American Pathologists and the Pulmonary Pathology Society. This article updates the previous guidelines for the histologic diagnosis of asbestosis and its distinction from other pulmonary fibrotic disorders1 and is intended to be used as a basis for communication between pathologists, pulmonologists, oncologists, radiologists, occupational hygienists, and epidemiologists. The aim of this new edition is to define the morphologic features of asbestosis at its various stages, relate exposure levels to specific tissue reactions, and evaluate the grading scheme published in the first report. Although these guidelines primarily focus on pathologic diagnosis and risk assessment, clinical and radiologic presentations are also included. Epidemiology is considered, and techniques for the assessment of asbestos fiber burden are evaluated.

The initial guidelines also dealt with mesothelioma, lung cancer, and benign, asbestos-related pleural diseases, but attention here is confined to asbestosis and disorders with which it may be confused.

HISTORIC BACKGROUND

The use of asbestos dates back to preclassic times, at least 5000 years. The heat resistance and tensile strength of asbestos were used by ancient Cypriot civilizations to manufacture clothes, whereas, in Finland, archaeologic remains have identified ceramics containing anthophyllite asbestos. The early Greeks used asbestos in lamp wicks, and the Romans incorporated asbestos in cremation clothes.

The hazardous health effects of asbestos were, to our knowledge, first recorded by the Greek geographer Strabo (in Geographia, book 10), reporting on a lung disorder among slaves weaving asbestos cloth. Large-scale industrial exploitation of asbestos began in the 1870s, and in 1898, attention was drawn to "injury to the bronchial tubes and lung" among asbestos-carding, silk-operating, and hemp-spinning [workers,.sup. (2) (p171)] although at the time, the prevailing view was that asbestos was inert in human tissue. The first detailed account of death from asbestosis appeared in 1907, (3) and thereafter, a series of reports described a chronic lung disease in asbestos workers that was neither tuberculosis nor silicosis. In 1917, the radiologic features of this disease were first described, (4) and in 1927, the term pulmonary asbestosis was introduced. (5) In 1930, a landmark epidemiologic study of asbestos textile workers showed a high incidence of lung fibrosis that correlated with exposure intensity and duration. (6) With the recognition of the health hazards came a requirement for governmental regulatory control of asbestos dust in the workplace, which resulted in dust-suppressive measures being introduced in the United Kingdom in 1933.

Until the 1930s, measurements of airborne dust were rare, and hygiene equipment was limited to counting fibrous particulates. In 1922 and 1937, the impinger and the midget impinger were introduced, respectively, in the United States, expressing particle concentrations of millions of particles per cubic foot. Thermal precipitators were the preferred early hygiene tools in the United Kingdom, as was the konimeter, in South Africa. The aim was to control and monitor dust levels in the workplace. Epidemiologists have subsequently used such hygiene measurements to evaluate asbestos exposure in risk analyses.

By the 1930s, there were anecdotal case reports of lung cancer complicating asbestosis, (7-9) but the association between lung cancer and asbestosis was not fully established until 1949, when lung cancer was reported in 13% of 235 men dying with asbestosis, compared with 1.2% of 6884 men dying with silicosis. (10) In 1951, a histologic survey of 1205 industrial postmortems found that 14% of the 121 asbestosis cases were accompanied by primary lung cancer. (11) In 1955, 2 further publications provided irrefutable evidence of the association between lung cancer and asbestosis. The first was a case-control study performed in 11 Californian hospitals, (12) and the second, a retrospective cohort mortality study of asbestos textile workers in the United Kingdom. (13) The latter found an 11-fold increase in lung cancer among long-term workers, all of whom had concomitant asbestosis.

Reports of pleural cancer in persons with asbestosis date from the 1940s and were substantiated by a report of 33 cases of diffuse pleural mesothelioma in a crocidolite mining area of South Africa in 1960, (14) at a time when the very existence of malignant mesothelioma was still controversial. The diagnosis of this tumor would remain problematic until the emergence of immunohistochemistry and electron microscopy in the 1970s.

Methodology for workplace dust monitoring had evolved since its introduction in the first part of the 1920s. In 1965, the modern membrane-filter method and a standardized approach to fiber counting were introduced. Fibers were interpreted as all structures with a length to diameter ratio of 3:1 or more (an arbitrary figure accepted by the Asbestosis Research Council). Pathogenic and respirable fibers were deemed to be those greater than 5 [micro]m long and less than 3 [micro]m in diameter. Fibers with these dimensions became known as regulated fibers or World Health Organization fibers. They were assessed by light microscopy, which meant that only fibers greater than 0.25 [micro]m in diameter were visualized (irrespective of fiber length). Thus, regulated fibers represent only a small proportion of those in a dust cloud. Furthermore, light microscopic methods do not permit an accurate distinction of asbestos from nonasbestos fibers (ie, qualitative data are not possible). To make that distinction would require energy-dispersive x-ray analysis and selected-area electron diffraction, techniques that only later became available.

In a series of landmark studies in the United States and Germany, researchers drew attention to the importance of fiber length in relation to mesothelioma induction in animals. (15-20) The investigators concluded that the induction of mesothelioma was determined primarily by the physical dimensions of the fibers, the most dangerous being those greater than 8 [micro]mm in length and less than 0.25 [micro]m in diameter. Long fibers were recognized to be more tumorigenic than short fibers. This led to the suggestion that other nonasbestos fibers of similar fiber dimensions might be just as hazardous as asbestos, an hypothesis that was subsequently supported by reports of erionite-induced mesothelioma in Turkey (21) and animal experiments showing lung fibrosis and mesothelioma induction following erionite inhalation. (22) On a dose per dose basis, erionite and amphiboles appeared to be equally potent.

In the 1970s, a number of epidemiologic studies were undertaken to determine the risk of disease after asbestos exposure, culminating in a state of the art document entitled Biological Effects of Mineral Fibres. (23) By the end of the decade, there was firm epidemiologic evidence that asbestos-induced disease related to fiber type as well as to cumulative asbestos dose. Chrysotile appeared to be less fibrogenic and far less tumorigenic than the commercial amphiboles crocidolite and amosite.

Since 1980, there has been considerable investment in the epidemiologic analysis of various asbestos-exposed cohorts to determine, in particular, exposure-response data for asbestosis. This was greatly facilitated by the 1980 International Labour Office system for the radiologic classification of pneumoconiosis (24) and the emergence of sophisticated electron microscopic analytical methods, which allowed more accurate assessment of fiber burden, fiber size, and fiber type.

A number of pathologic definitions of asbestosis exist, and pathologists have not applied them uniformly. During the past 25 years, 2 comprehensive descriptions of the pathologic characteristics of asbestosis have emerged. The Pneumoconiosis Committee of the College of American Pathologists and the National Institute of Occupational Safety and Health provided the first such description in 1982. (1) Their document set forth the proposed minimum histologic criteria for asbestosis and formulated a grading scheme for lung fibrosis. The minimum criteria necessary for a diagnosis of asbestosis were the finding of "discrete foci of fibrosis in the walls of respiratory bronchioles associated with accumulations of asbestos bodies" in histologic sections. Although the essential requirement for asbestosis is the presence of diffuse interstitial fibrosis in association with asbestos bodies, several outstanding issues exist and these problems still bedevil discussions on this topic. These issues concern the minimum number of asbestos bodies necessary and the degree of fibrosis required to enable the pathologist to diagnose asbestosis with confidence. The College of American Pathologists-National Institute of Occupational Safety and Health criteria did not specify the minimum number of asbestos bodies required, despite an inherent understanding that the definition phrasing implies more than one asbestos body. Unfortunately, there was no reference to the required asbestos body number per lung sectional area examined.

The grading scheme for fibrosis (asbestosis) was not universally accepted. Most notably, Churg and coworkers (25) sought to highlight that the disease process asbestosis required "diffuse" interstitial fibrosis and that this condition was distinct and different from small airways mineral-dust disease, where multifocal peribronchiolar fibrosis may be seen after the inhalation of a variety of mineral dusts, including asbestos, silicates, metal oxides, and coal. Similar changes may be seen commonly in the lungs of tobacco smokers. (26) It was recognized that the determination of early/minimal asbestosis in current or ex-smokers would be highly problematic, with potential for misclassification of cases.

A number of key publications emerged in the 1980s. With respect to the pathologic diagnosis of asbestosis, Roggli and Pratt (27) noted that 2 asbestos bodies in a 4-[cm.sup.2], Perl-stained section correlated with a 10-fold increase of asbestos bodies greater than the upper limit for a control reference population. In a series of accepted asbestosis cases studied by Roggli et al, (28) the authors reported more than 5 asbestos bodies per 1 [cm.sup.2] in 95% of cases, and more than 2 asbestos bodies per 1 [cm.sup.2] of lung sectional area in 100% of cases. In iron-stained sections, 2 asbestos bodies in a 2- X 2-[cm.sup.2] section corresponded to approximately 2000 asbestos bodies per gram of dry lung tissue on light microscopy. (29) Because asbestos bodies and fibers were not evenly distributed, more than one section needs to be examined. That evidence was accepted by most authorities and served to underpin the later Helsinki definition of asbestosis (see below). (30)

In 1984, the Ontario Royal Commission heard evidence that the threshold cumulative dose of asbestos necessary for clinical manifestations of asbestosis was between 25 and 200 fibers/ml-yrs (fibers/ml X number of years) of cumulative exposure. In contrast, a no-effect threshold level of asbestos (less than which there was no subsequent risk) was not recognized for malignant mesothelioma. (31)

Also in the 1980s, the mineralogic study of lung tissue in humans and animals was expanded. The focus was now to correlate disease with respirable (ie, inhaled, deposited, and retained) fiber counts at the site of tissue injury and to characterize the retained particulates. Various workers sought to study lung material from published cohorts. Lung fibrosis showed a dose-response relationship with retained amphibole asbestos, but not with chrysotile. Malignant mesothelioma incidence in Canadian chrysotile miners correlated with the contaminant amphibole tremolite, but did not correlate with chrysotile. (32) It emerged that amphiboles are far more potent in the induction of asbestosis, lung cancer, and malignant mesothelioma than chrysotile. Meta-analyses have estimated that, for mesothelioma induction, the risk differential on a fiber per fiber basis is 1:1, 100:1, and 500:1 for chrysotile, amosite, and crocidolite, respectively. For lung cancer, estimates indicate a risk differential of 1:10 to 1:50 for chrysotile and commercial amphiboles, respectively. (33)

In 1997, 19 participants from 8 countries met in Finland to discuss the diversity of asbestos-induced disorders, from which arose the so-called Helsinki Criteria for the definition of asbestosis. (30) The criteria required diffuse interstitial fibrosis and either 2 or more asbestos bodies within …