Published ahead of print on March 15, 2007, doi:10.1164/rccm.200601-134OC
© 2007 American Thoracic Society doi: 10.1164/rccm.200601-134OC
Two-dimensional Analysis of Elements and Mononuclear Cells in Hard Metal Lung Disease1 Division of Respiratory Medicine, Graduate School of Medical and Dental Sciences, Niigata University; 2 Electron Probe Microanalysis Laboratory, Center of Instrumental Analysis, Niigata University; 3 Department of Internal Medicine, JA Niigata Kouseiren Uonuma Hospital; 4 Division of Dental Biomaterial Science, Graduate School of Medical and Dental Sciences, Niigata University; and 5 Department of General Medicine, Niigata University Medical and Dental Hospital, Niigata, Japan Correspondence and requests for reprints should be addressed to Toshinori Takada, M.D., Ph.D., Division of Respiratory Medicine, Graduate School of Medical and Dental Sciences, Niigata University, 1-757 Asahimachi-dori, Niigata 951-8510, Japan. E-mail: ttakada{at}med.niigata-u.ac.jp
Rationale: Hard metal lung disease is caused by exposure to hard metal, a synthetic compound that combines tungsten carbide with cobalt as well as a number of other metals. Interstitial lung disease caused by hard metal is uniquely characterized by giant cell interstitial pneumonia. The pathogenesis of hard metal lung disease is unclear. Objectives: To elucidate the distribution of inhaled hard metal and reactive inflammatory cells in biopsy lung tissue from patients with hard metal lung disease. Methods: Seventeen patients with interstitial lung disease in which tungsten was detected and five control subjects were studied. Detection and mapping of elements were performed with an electron probe microanalyzer equipped with a wavelength dispersive spectrometer. We immunohistochemically stained mononuclear cells, in tissue samples available from five patients, with anti-human CD4, CD8, CD20, CD68, and CD163 antibodies, and compared the distribution of positive cells with hard metal elements. Measurements and Main Results: Thirteen of 17 patients were pathologically diagnosed as having giant cell interstitial pneumonia. Tungsten and cobalt were accumulated in the centrilobular fibrotic lesions, but were never found in the control lungs. CD8+ lymphocytes and CD163+ monocyte-macrophages were distributed predominantly in centrilobular fibrotic lesions around the hard metal elements. CD163+ colocalized with tungsten. Small numbers of CD8+ and CD163+ cells were also immunohistochemically shown in peribronchiolar areas and alveolar walls. Conclusions: Macrophages may phagocytose inhaled tungsten via CD163 and play an important role in forming the fibrotic lesion of hard metal lung disease with cytotoxic T lymphocytes.
Key Words: electron probe microanalysis • hard metal • interstitial lung disease • CD163 antigen • CD8-positive T lymphocytes
Hard metal is a synthetic compound that combines tungsten carbide with cobalt as well as a number of other metals. It has a hardness nearly that of diamond, which makes it useful in the cutting and grinding of metal tools, stones, and concrete. Occupational hard metal exposure is known to cause hard metal lung disease (HMLD) (1–3). The pathologic findings of HMLD are predominantly those of interstitial pneumonia and fibrosis (4). In many cases, multinucleated giant cells are prominent, resulting in a pattern of giant cell interstitial pneumonia (GIP) (4–6). GIP was originally classified by Liebow as one of the idiopathic interstitial pneumonias (7), but it is now recognized as an uncommon form of interstitial lung disease (ILD) that is caused by exposure to hard metal (8). The characteristic distribution of fibrosis in GIP suggests that the inflammation in the centrilobular area initiated by hard metal accumulation plays an important role in the pathogenesis of HMLD. However, the pathogenesis of the disease is still unclear. Electron probe microanalyzers (EPMAs) irradiate specimens with a finely focused electron beam to obtain information about the elemental composition. Energy-dispersive spectrometers have been applied to the analysis of particles and mapping of chemical elements in lung tissue of HMLD (6, 8, 9). EPMA with a wavelength-dispersive spectrometer (WDS) has a higher sensitivity (10) and this makes it more suitable for detecting elements, which is especially relevant in the case of biosoluble cobalt. We have developed an improved technique for element analysis of tissue sections of 3 µm thickness by means of EPMA with WDS (11). To elucidate the pathogenesis of HMLD, we applied this technique to biopsy lung tissue from patients with HMLD to analyze the distribution of elements in GIP. We also immunophenotyped mononuclear cells infiltrating into the inflammatory lung by immunohistochemistry.
Subjects Lung biopsy specimens from 85 patients with ILD suspected to be of environmental or occupational origin were analyzed by EPMA between 1993 and 2005. Thirty patients had undergone a surgical biopsy and the other specimens were obtained by transbronchial biopsy. Tungsten (W) was detected in surgical biopsies from 17 patients and in transbronchial biopsy from 3 other patients. These 20 patients were considered to have HMLD, but only the 17 patients with a surgical biopsy were included in this study. EPMA analysis was also performed in lung tissue obtained by lobectomy (for the treatment of non–small-cell lung cancer) from five other patients enrolled as control subjects. The study protocol was approved by the Committee of Ethics of Niigata University (Niigata, Japan).
Sample Preparation and Pathologic Study
Electron Probe Microanalyzer
Light Microscopy, Immunohistochemistry, and Statistical Analysis
Characteristics of Subjects Among the 85 patients with ILD in whom EPMA was performed, no other patient had a record of occupational exposure to hard metal, except for the 20 subjects in whom tungsten was detected in a surgical lung biopsy (17 subjects) or in a transbronchial biopsy (3 subjects, not examined further). Biopsies from two other patients exhibited features of GIP, but no tungsten or cobalt was detected in their lung tissue, and neither had a history of work in the hard metal industry (one subject was a music teacher, and the other had worked for 30 years in a tire factory). Age, sex, smoking history, exposure period, pathologic diagnosis, and elements detected among the 17 patients with HMLD are summarized in Table 1. All subjects with a known duration of exposure had been exposed until the time of lung biopsy, except for patient 7. Thirteen subjects had typical GIP, as illustrated in Figure 1A. Four patients were considered to have an atypical GIP pattern because the features of GIP, that is, centrilobular fibrosis and giant cell accumulation, were mixed to various extents with interstitial inflammation and fibrosis similar to usual interstitial pneumonia (UIP) (Figures 1D, 1E, and 1F).
Two-dimensional Analysis of Biopsy Lung Tissue for Elements, Using an Electron Probe Microanalyzer The centrilobular fibrotic lesion of GIP observed by hematoxylin–eosin staining was two-dimensionally analyzed by EPMA with WDS to detect Al, Si, Ti, Cr, Fe, Co, Ta, W, and Zn (Figure 2A). Using this technique, we were able to determine the distribution of an element of interest in a tissue sample in a semiquantitative manner (Figure 2B) and to superimpose the element map on a pathologic tissue image by coloring the amino nitrogen green (Figure 2C). The results showed that tungsten was distributed in a relatively high concentration almost throughout the peribronchiolar fibrosis in the centrilobular lesion. By qualitative analysis of a selected area (10 x 10 µm area, indicated by the arrowhead in Figure 2D) in a fibrosing lesion of GIP (Figure 2D), we found Al, Si, Ti, Cr, Fe, and Ta in addition to tungsten (W) (Figure 2E). A list of the elements detected in the 17 patients is shown in Table 1. Interestingly, another major component element of hard metal, cobalt, was detected in only 4 of the 17 patients. It seems that the presence of detectable cobalt was not associated with the exposure period or time since the end of exposure. We also mapped elements in control lung specimens from five patients with non-small cell lung cancer (Figure 3). In addition to Si, we found Al, Ti, Fe, and Zn, but never Co or W, in the bronchioles, vessels, and alveoli (data not shown).
Comparison of Hematoxylin–Eosin Staining and EPMA Images We then studied the distribution of elements of interest in GIP fibrosing lesions by comparing hematoxylin–eosin staining and EPMA images. As shown in Figure 4, many elements were accumulated in the peribronchiolar fibrosing tissue. However, the distribution of elements varied widely. Aluminum and silicon were found in that region and were also detected in the surrounding alveolar areas. The results suggest that most of the silicon may be present in the lung tissue as aluminum silicates, with some as silica particles. Areas containing titanium were limited, with most appearing to be distributed at the same position as areas containing tungsten. Cobalt was not detected two-dimensionally at the same magnification as the other elements because it was present in very low amounts (data not shown).
Distribution of Inflammatory Cells in Hard Metal Lung Disease We immunophenotyped lymphocytes and monocyte-macrophages infiltrating lung tissue of HMLD by immunohistochemistry to identify predominant cells in various lesions (Figure 5). The mean numbers of positive cells in 10 fields for each type of structure are summarized in Table 2. The main findings of the immunohistochemical study were as follows: (1) For lymphocytes, CD8+ cells, compared with CD4+ cells, were found predominantly in alveolar walls, peribronchioles, and centrilobular fibrotic lesions as indicated by CD4/8 comparisons; (2) CD20+ cells were found predominantly in lymphoid follicles; and (3) CD163+ monocyte-macrophages were present in alveolar walls, peribronchioles, and centrilobular fibrotic lesions, in addition to their presence in alveolar airspaces as alveolar macrophages. Some of the CD163+ cells were colocalized with particulate black materials and were surrounded by a number of CD8+ lymphocytes.
Comparison of Distribution of Inflammatory Cells and Tungsten Images of immunohistochemical staining and element distribution by EPMA are compared in Figure 6. Most of the CD163+ cells and CD8+ lymphocytes were distributed in fibrotic lesions surrounding tungsten. Some CD163+ monocyte-macrophages colocalized with tungsten. CD8+ and CD163+ cells were also found in the lung interstitium, such as alveolar walls and peribronchioles. However, tungsten and cobalt were not found in those areas (data not shown).
EPMA and Immunohistochemical Analysis of Giant Cells Last, we investigated the relationship between tungsten and giant cells in GIP. EPMA revealed that tungsten was located mainly in giant cells, not in regular alveolar macrophages (Figure 7B). On the other hand, they had less CD163 than alveolar macrophages (Figure 7C).
Exposure to hard metal dust is well recognized as a cause of the ILD known as HMLD and remains an important occupational health concern. We applied an EPMA with WDS technique to lung specimens of GIP from HMLD and successfully showed the distribution of component elements, that is, hard metal tungsten and cobalt, in the centrilobular fibrotic lesion of GIP. We also studied the immunophenotypes of mononuclear cells infiltrating into the lung tissue. EPMA with energy-dispersive spectrometry has been used to analyze elements in HMLD tissue (8, 9, 14, 15). Abraham and coworkers discovered that most patients with a diagnosis of GIP had worked in the hard metal industry, and that 30 of 31 patients with GIP were among those 50 patients with the highest concentrations of tungsten as determined by EPMA with energy-dispersive spectrometry (8). For EPMA with WDS, although the sensitivity to element is almost 10 times better than with an energy-dispersive spectrometer (10), the intense beam required to detect elements probably destroys the tissue sample because of the high temperature. We achieved a high specimen current without sample damage, using a thin section on a pure carbon plate (11). In our case, a specimen current of 0.6 µA was used, which is about 10 times higher than is usual in materials science analysis. Furthermore, the scanning speed was slow, 0.3 to 0.8 s at each pixel. The simultaneous use of both a high specimen current and a slow scanning speed made it possible to detect elements in small volumes, such as respirable particles. We used a pixel size ranging from a 1- to 14-µm step for each axis in this study. Analysis by 1-µm step enabled us to resolve individual particles visible by light microscopy and to estimate the composition of each particle. On the other hand, although analysis using a 14-µm step yields only qualitative data for each pixel, we can obtain a mapping of about 3.6 x 3.6 mm area per scan. Qualitative analysis detected Al, Si, Ti, Cr, Fe, Co, and Ta, besides W, in the patients. Al, Si, Ti, and Fe were also found in control subjects. The main components of hard metal are tungsten carbide (about 90%) and cobalt metal (about 10%), and small amounts of nickel, chromium, and tantalum may be added (16). Therefore, the Al, Si, Ti, and Fe were from environmental dust, whereas other elements, Cr and Ta, should be derived from hard metal. The critical role of cobalt in causing HMLD has been proposed since the 1950s as a result of studies on experimental animals (17–19). Verougstraete and coworkers reported that cobalt exposure contributed to a decline in FEV1 over time (20). However, we could find cobalt in only 4 of the 17 patients in the study. Cobalt was proved to be highly soluble in protein fluids (17) and its location on the surface of the hard metal particles also promotes its rapid dissolution. It was not always found in the lung tissue with a histologic diagnosis of GIP, whereas a pathologic diagnosis of GIP is almost pathognomonic for HMLD and tungsten was found in most patients with GIP (8). Several in vitro and animal experiments suggest that tungsten carbide also plays an important part in the pathogenesis of HMLD as a result of a synergistic effect (21–23). Thus, the presence of tungsten and cobalt in lung tissues from persons with ILD is pathognomonic for HMLD, but the absence of cobalt may not be a negative finding for HMLD diagnosis. GIP was found in Belgian diamond polishers exposed not to hard metal dust, but to cobalt-containing dust (24). Nevertheless, this condition should be called cobalt-associated GIP rather than HMLD. It has been suggested that HMLD may develop in some workers through a hypersensitivity reaction (1, 5). Hypersensitivity pneumonitis (HP) is caused by inhaled allergens that elicit lymphocytic inflammation in the peripheral airways and surrounding interstitium. We found tungsten and cobalt in centrilobular fibrosing lesions, suggesting that inhaled hard metal was trapped at the bronchioles and triggered the inflammation of HMLD. The acute phase of HP is characterized by proliferation of CD8+ cytotoxic T lymphocytes (25). A typical feature of inflammatory cells in bronchoalveolar lavage fluid from a patient with HMLD is that the T lymphocyte counts may be normal or increased and the CD4+ cell:CD8+ cell ratio is inverted (26–28). Our immunohistochemical findings also indicated that CD8+ cells were to be found predominantly in the centrilobular fibrosing lesions, suggesting that HMLD may result from a hypersensitivity reaction. In addition to the typical HP, a clinically subacute or chronic type of HP has also been reported (29–31). Chronic-type HP results from continued low-level exposure to pathogenic antigens. All the patients in this study had an exposure period of at least 2.5 years and more than two-thirds of them were exposed to hard metal for longer than 5 years. Therefore, HMLD may be better classified as chronic HP caused by continued exposure to hard metal. Although tungsten and cobalt were found only in centrilobular fibrosing lesions, CD8+ lymphocytes and CD163+ monocyte-macrophages were found not only in fibrosing lesions around tungsten but also in other areas of the interstitium, such as alveolar walls and peribronchioles. CD8+ lymphocytes play an essential role in the development of pulmonary injury via the Fas–Fas ligand pathway (32). In addition to this pathway, CD8+ cells can also activate alveolar target cells and trigger their expression of inflammatory mediators monocyte chemoattractant protein-1 and macrophage inflammatory protein-2, leading to monocyte-macrophage–mediated interstitial inflammation (33). Pulmonary interstitial injury of HMLD may also develop and progress via either or both of these mechanisms. CD163 belongs to the scavenger receptor cysteine-rich domain family. It is a monocyte-macrophage differentiation antigen and tissue macrophages have a substantially higher expression compared with monocytes (34). CD163 was first characterized as a scavenger receptor for hemoglobin, mediating endocytosis of hemoglobin:haptoglobin complexes (35), but Philippidis and coworkers have reported that it may act as a molecular switch to promote inflammatory resolution and wound healing (36). Thus, CD163+ cells may on the one hand develop monocyte-macrophage–mediated interstitial inflammation, but on the other hand inhibit the progression of interstitial inflammation by CD8+ lymphocytes via their antiinflammatory activity. CD163 was expressed less on giant cells containing tungsten compared with alveolar macrophages. Low expression of CD163 was also reported on multinucleated giant cells from granulomatous diseases (37). That may be because CD163 molecules on the cell surface had just been diluted by fusion with mononuclear cells, and/or it may indicate that the macrophages carrying tungsten lost their scavenger activity when they fused into giant cells. The pathologic findings of HMLD are either desquamative interstitial pneumonia or GIP with or without bronchiolitis obliterans, and varying degrees of interstitial fibrosis (4–6). Most of the patients in this study showed GIP, but we also found a histology of dense fibrosis and fibroblastic foci similar to the UIP pattern. UIP is the histopathologic pattern that identifies patients with idiopathic pulmonary fibrosis (IPF). Epidemiologic studies have indicated that metal dust is one of the risk factors for IPF (38, 39). We are planning to analyze tissue samples of IPF to determine whether elements of hard metal are present, or if not, what kind of elements are predominantly distributed in the UIP tissue from patients with IPF. In conclusion, this study demonstrated the distribution of elements in biopsy lung tissues from patients with HMLD, using a technique of EPMA with WDS. Tungsten was detected mainly in centrilobular fibrosis and peribronchioles and was closely located with a number of CD163+ macrophages surrounded by CD8+ but not CD4+ lymphocytes. These results suggest that CD163+ macrophages may phagocytose inhaled tungsten and play an important role in forming the fibrotic lesion of HMLD with cytotoxic T lymphocytes.
The authors thank the following doctors for the supply of samples: Chiba University Hospital, Dr. Fuminobu Kuroda; Juntendou University Hospital, Dr. Fumiyuki Takahashi and Dr. Shinsaku Jyuuai; Showa University Northern Yokohama Hospital, Dr. Hiroaki Nakajima; Kurobe City Hospital, Dr. Hiroshi Tsuji; Kagoshima University Faculty of Medicine, Dr. Yoshifusa Koreeda; Jikei University Hospital, Dr. Kaoru Aoki; Showa General Hospital, Dr. Masako Kato; University of Fukui Faculty of Medicine, Dr. Masanori Nakanishi; Tohoku University Graduate School of Medicine, Dr. Shinya Okouchi; Nishi-Niigata Chuo National Hospital, Dr. Yasuharu Saito; University of Tsukuba Graduate School of Medicine, Dr. Yukio Ishii; Saitama Cardiovascular and Respiratory Center, Dr. Yoshinori Kawabata; and Kanazawa Graduate University School of Medicine, Dr. Yuko Waseda. The authors offer special thanks to Naofumi Imai for technical assistance and to Dr. Yasuharu Saito for useful discussion.
Supported by a Grant-in-Aid for Scientific Research (16590740). Originally Published in Press as DOI: 10.1164/rccm.200601-134OC on March 15, 2007 Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form January 31, 2006; accepted in final form March 28, 2007
Related articles in AJRCCM:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
AT A GLANCE COMMENTARY
