Eur Respir J, 1996, 9, 1024–1028 Copyright ERS Journals Ltd 1996

DOI: 10.1183/09031936.96.09051024 European Respiratory Journal

Printed in UK - all rights reserved ISSN 0903 - 1936

POINT OF VIEW

Experimental research into the pathogenesis of

cobalt/hard metal lung disease

D. Lison*, R. Lauwerys*, M. Demedts**, B. Nemery**

Experimental research into the pathogenesis of cobalt/hard metal lung disease. D. Lison, *Unité de Toxicologie Industrielle et Méde-
R. Lauwerys, M. Demedts, B. Nemery. ERS Journals Ltd 1996. cine du Travail, UCL, Brussels, Belgium.
ABSTRACT: In recent years clinical, epidemiological and experimental evidence **Laboratorium voor Pneumologie (Long-
has accumulated indicating that cobalt metal particles, when inhaled in association toxicologie), [Link], Leuven, Belgium.
with other agents such as metallic carbides (hard metals) or diamond dust, may Correspondence: B. Nemery
produce an interstitial lung disease termed "hard metal disease" or "cobalt lung". Laboratorium voor Pneumologie (longtox-
This article summarizes the progress accomplished in our two laboratories to icologie)
understand the pathogenesis of this disease. Gaps and weaknesses in our current Herestraat 49
knowledge have also been highlighted in order to suggest potential avenues for fur- K.U. Leuven
ther research. Whilst animal models have proved useful for the demonstration of 3000 Leuven
the toxic synergy between cobalt and carbides (e.g. tungsten carbide), most animal Belgium
models have remained descriptive and have not provided information on the mech-
Keywords: Acivated oxygen species, cobalt,
anism for this synergy. In particular, the bizarre multinucleated giant cells which diamond, giant cell interstitial pneumoni-
are an important hallmark of the human disease, have not been reproduced con- tis, hard metal, interstitial lung disease,
sistently in experimental animals. Since cobalt is a known sensitizer, there may also occupational lung disease, oxidant injury,
be a need to develop experimental models to test the possible involvement of immuno- tungsten carbide.
logical mechanisms in the pathogenesis of the interstitial disease. In vitro systems
including macrophage cell cultures and physico-chemical tests have been useful to Received: January 18 1996
investigate the mechanism underlying the toxic synergy. Accepted for publication January 24 1996
The recent finding that, in vitro, cobalt and metallic carbides interact with oxy-
gen to produce toxic activated oxygen species opens a new avenue of research and Supported by the Commission of the
European Communities (Directorate General
may offer an alternative interpretation of the fact that only a limited proportion of XII - Research and Technology Department
exposed workers develop interstitial disease. Besides the possible involvement of - Environment), the Belgian Funds for
immunological mechanisms, it may be speculated that individuals with a lower anti- Scientific Research (NFWO/FNRS), and
oxidant defence are more susceptible to the toxic effect of activated oxygen species the Federal Services of the Scientific,
produced by cobalt-containing dusts from hard metal. Technical and Cultural Affairs of the Prime
Eur Respir J., 1996, 9, 1024–1028. Minister's Services (DWTC/SSTC).

In the past few years, our two laboratories have, more The interest of the UCL team goes back to their long
or less independently, carried out experimental research tradition of research into the toxicity of various metals,
into the pathogenesis of hard metal lung disease. Why which itself is possibly a reflection of Belgium's promi-
should two Belgian teams be particularly interested in nent position in the nonferrous metal industry. In this
what is, after all, a relatively rare condition? The origin "Point of View", we will summarize our present knowl-
of the interest of the K.U. Leuven team stems essentially edge, as well as the questions that are still unanswered,
from their involvement, in the mid-seventies and later, regarding the pathophysiology of hard metal lung dis-
in an "epidemic" of interstitial lung disease among dia- ease, with an obvious emphasis on the possible mecha-
mond polishers [1–3]. The disease had all the features nisms for the pulmonary toxicity of cobalt/hard metals.
of hard metal lung disease. Hard metal is a composite The inhalation of excessive amounts of cobalt-con-
material consisting of metal carbides, mainly tungsten taining dust may produce various respiratory manifesta-
carbide (WC), cemented in a matrix of cobalt (Co). How- tions including upper respiratory tract irritation, asthma
ever, the problem was that these diamond polishers had and a peculiar form of interstitial lung disease, originally
no exposure to hard metal; they were, however, exposed called hard metal pneumoconiosis [4–6]. This interstitial
to cobalt-containing dust, which originated from recently lung disease, the clinical presentation of which may vary
introduced polishing disks, in which microdiamonds were from subacute alveolitis to progressive fibrosis, is char-
also bonded together with cobalt. The finding of "hard acterized, particularly in the early stages, by the pres-
metal lung disease" in subjects who were not exposed to ence of numerous "bizarre" multinucleated giant cells in
hard metal but to cobalt-containing dust provided indi- the lung interstitium and the alveolar lumen, as well as
rect support to what had long been suspected, namely in the bronchoalveolar lavage (BAL) (fig. 1). It has been
that the toxic component of hard metal is cobalt rather suggested that giant cell interstitial pneumonitis (GIP) is
than WC. On this basis, the term "cobalt lung" was pro- almost pathognomonic for hard metal-induced lung dis-
posed [1]. ease [7]. Whilst there is no doubt that exposure to cobalt
 PAT H O G E N E S I S O F C O B A LT / H A R D M E TA L L U N G D I S E A S E 1025

HARDING [11] was first to describe the severe pulmonary

oedema and haemorrhage found in rats administered cobalt
metal powder by intratracheal instillation. Since tungsten
metal or carbide alone did not produce the same effect,
cobalt was deemed responsible for the parenchymal dam-
age observed in the lung of hard metal workers.
Also in the rat, KAPLUN and MEZENCEWA [12] found
that the interstitial lung damage produced by cobalt metal
was exacerbated by the simultaneous addition of tung-
sten or titanium. The morphological changes induced by
the mixtures were identical in nature to that produced by
cobalt alone but they were more marked. The enhanced
toxicity of the mixture was attributed to the higher solu-
bility of cobalt in the presence of tungsten. It is, how-
ever, not clear whether tungsten or tungsten carbide was
used in this study.
KITAMURA et al. [13] examined the pulmonary response
to a single intratracheal administration of a high dose of
hard metal powder in the rat. Six months later, the sur-
viving animals presented pulmonary lesions of patchy
fibrosis in the vicinity of deposited dust, but there was
no evidence of an active inflammatory process in the lung.
At 12 months, the lesions had apparently regressed. The
toxic effect of these particles on the lung was attributed,
without experimental support, to the cytotoxic action of
cobalt released from the particles. Unfortunately, these
investigators did not test cobalt or tungsten carbide alone.
Using state of the art methods for the evaluation of
pulmonary toxicity (including cellular and biochemical
analysis of BAL fluid, as well as pathology), we [14]
have compared, in the rat, the pulmonary response after
intratracheal administration of tungsten carbide particles,
hard metal powder (a mixture of tungsten carbide with
Fig. 1. – Cytospin of cells obtained by bronchoalveolar lavage in a
24 year old female diamond polisher with severe interstitial lung dis- 6% of cobalt metal particles) or an equivalent dose of
ease. The cell distribution was: 84% macrophages, 3% lymphocytes, cobalt metal particles alone. After a single administra-
9% polymorphonuclear neutrophils and 4% eosinophils. a) It can be tion, the acute and delayed lung responses to hard metal
seen that the cytospin contains numerous multinucleated giant cells powder were much more pronounced than those pro-
(the darker and larger cells). b) A typical example of one of the multi-
nucleated cells to illustrate their "cannibalistic features". Such "bizarre" duced by each individual component administered sep-
cells are characteristically found in giant cell interstitial pneumonitis arately. As found by KITAMURA et al. [13], the alveolar
caused by hard metal/cobalt exposure. (May-Grünwald Giemsa stain ; reaction had completely regressed by 4 months and no
internal scale bar=5 µm). fibrosis developed. In a protocol involving repeated intra-
tracheal administrations of the different particles, no effect
alone may lead to the development of occupational asth- on the parenchymal architecture was found 1 month after
ma, the exact role of cobalt in the pathogenesis of the the last dose in the groups treated with tungsten carbide
parenchymal disease, i.e. the "real" hard metal lung dis- or cobalt alone, whereas clear fibrotic lesions were obs-
ease, has long been debated [8, 9]. Involvement of the erved in the group instilled with hard metal [15]. Again,
lung parenchyma has only been reported, in clinical and as noted by KITAMURA et al. [13], this fibrotic reaction
epidemiological studies, among hard metal and diamond was not accompanied by inflammation, suggesting a scar-
workers, i.e. when cobalt is inhaled in association with ring process rather than active fibrosing alveolitis. We
other components, such as WC or diamond dust. In con- did not detect giant multinucleated cells in BAL or lung
trast, parenchymal toxicity seems to be absent when expo- tissue in any experimental group.
sure is to cobalt alone [10]. As shown below, we have In a guinea-pig model, SCHEPERS [16] and DELAHANT
obtained experimental evidence that supports the con- [17] found that intratracheal instillation of high doses of
cept of a synergy between cobalt and carbides to pro- cobalt metal led to the development of acute pneumo-
duce cellular toxicity, and we have recently clarified the nia. The subchronic response assessed 8–12 months later
underlying mechanism. was characterized by the presence of multinucleated cells,
but a lack of fibrotic reaction within the alveolar walls.
These investigators concluded that cobalt metal was not
Animal studies fibrogenic. In contrast, the instillation of cobalt metal
mixed with tungsten carbide induced a transient imflam-
Most in vivo investigations to assess the lung toxicity matory reaction, with residual fibrosis in the vicinity of
of cobalt and related powders have been performed using retained particles, and multinucleated giant cells were
intratracheal instillation. Different species have been used present in some instances. In inhalation experiments, a sim-
with varying results. ilar mixture caused severe inflammation but no fibrosis.
1026 D . LISON ET AL .

Unfortunately, tungsten carbide alone was not tested. It pathogenesis of hard metal disease. Cobalt is a known
should also be emphasized that these investigators observed sensitizer, and cobalt exposure may lead to allergic
multinucleated giant cells in animals treated with a com- contact dermatitis or to bronchial asthma. In both these
bination of tungsten carbide and carbon (without cobalt), conditions, evidence for humoral or cellular immune
raising some doubt about the specificity of this obser- mechanisms has been provided but, so far, "allergy to
vation and its relevance for the human disease. It is also cobalt" has only been demonstrated conclusively in very
unclear whether the multinuclear giant cells found in few patients with the parenchymal form of hard metal
these animal experiments were ordinary foreign body lung disease, in spite of considerable attempts to verify
cells or whether they were similar to the "bizarre" can- this very plausible hypothesis. Although, in general, imm-
nibalistic giant cells found in the human disease. SCHEPERS unologically-mediated interstitial lung disease has proved
[16] in fact seemed to attach more importance to the rather difficult to reproduce experimentally (e.g. beryl-
eosinophil response, which he interpreted as suggestive liosis and hypersensitivity pneumonitis), it appears rele-
of an allergic mechanism, and to the hyperplastic and vant to try and develop an animal model which would
metaplastic changes observed in the alveolar and bronchial combine the toxic properties of hard metal dust and the
epithelium. known allergenic potential of cobalt. This is an avenue
KERFOOT et al. [18] exposed mini-pigs for 3 months to of research in which one of our teams is now engaged.
cobalt concentrations of 0.1 and 1.0 mg· m-3. In view of
a possible involvement of an immunological mechanism
(a type I allergic reaction has been suggested in cobalt In vitro studies
asthma), they first submitted the animals to a "sensi-
tization" period by inhalation. Immediately after exposure, During recent years, the use of in vitro cell culture
lung function studies demonstrated a dose-dependent and systems has played an important role in studying the toxico-
reversible reduction of the lung compliance in cobalt- logy of mineral dusts and in providing insight into their
exposed animals. These compliance changes were inter- mechanism of action. In particular, such studies have
preted as demonstrating functional impairment but they contributed to a better understanding of the basic bio-
were not associated with any radiological or histologi- chemical and molecular mechanisms of how mineral dusts
cal signs of fibrosis, except for some increased collagen interact with cells to elicit inflammation. Numerous stud-
deposition, which was only evident at the electron micro- ies on mineral dust cytotoxicity have focused on the
scopic level. These authors did not present a quantita- macrophage because it is generally accepted that inter-
tive analysis of this collagen excess, nor did they mention actions with this cell type constitute a first critical step
their methodology for selecting the sites for electron in the initiation of the inflammatory process. Many stud-
microscopic examination, which is critical, since there ies have examined cytotoxicity end-points in peritoneal
was no apparent fibrosis at the light microscopic level. and alveolar macrophages, with similar responses. How-
The absence of an "innocuous substance" control group ever, for investigating functional capabilities (e.g. cytokine
is also a major weakness of this study. secretion), peritoneal and alveolar macrophages differ.
Overall, as already suggested by early studies, the high-
er lung toxicity of cobalt when combined with tungsten
carbide has been clearly confirmed by the proper com- Cytotoxicity studies
parison of cobalt alone, tungsten carbide alone, and the
combination thereof under the form of hard metal parti- The different biological reactivities of cobalt and hard
cles. Limited studies in hamsters have also shown that metal particles have been reproduced in a cytotoxicity
the intratracheal administration of cobalt (5 mg· kg-1) and macrophage model [20]. Using LDH release as an index
diamond particles (50 mg· kg-1), caused more acute lung of cytotoxicity, it was found that hard metal powder was
damage, as assessed by lactate dehydrogenase (LDH) almost as reactive as crystalline silica; whereas, when
activity in BAL fluid 2 days after intratracheal adminis- tested separately, tungsten carbide had no effect and pure
tration, than when these particles were given alone. How- cobalt only moderately impaired cell viability. Similar
ever, this was also the case for a similar combination of cytotoxic responses were observed with peritoneal and
cobalt and iron particles, and no long-term (up to 21 alveolar macrophages. Since these responses were con-
days) synergism was found [19]. sistent with the in vivo findings, peritoneal macrophages
So far, all these in vivo studies have remained purely were used to investigate the mechanisms of toxicity of
descriptive and they have failed to offer any definitive hard metal powder. Thus, it was found that, under cer-
information on the mechanism of the human disease. In tain conditions of specific surface area and chemical nat-
particular, an important hallmark of the human disease, ure, cobalt metal particles also interact with carbides other
i.e. the bizarre multinucleated cells has not been prop- than that of tungsten to produce cytotoxicity [21]. These
erly reproduced in an animal model. This represents a studies have also revealed that the unique toxicity of
serious limitation to the further study of the pathogene- cobalt-carbide mixtures could not be ascribed to a non-
sis of hard metal disease and there is a need for inves- specific carrier effect of carbide particles [22]. As men-
tigation of other experimental systems. It is also evident tioned above, it has long been suggested that the enhanced
that intratracheal instillation does not adequately repro- solubility of cobalt and, hence, its increased bioavail-
duce the exposure pattern in the industrial setting, and ability in the presence of a carbide was responsible for
it would still be useful to perform inhalation studies in the high toxicity of hard metal particles [11, 12]. However,
the rat or in other species. we have recently demonstrated, with the in vitro model,
Another important aspect which deserves attention is that cobalt solubilization and bioavailability are not the
the possible involvement of immune mechanisms in the critical factors for explaining this toxicity [22, 23].
 PAT H O G E N E S I S O F C O B A LT / H A R D M E TA L L U N G D I S E A S E 1027

Research is still required to verify whether the concepts of cellular toxicity, indicating their possible early invol-
derived from these experiments using tungsten carbide vement in the pulmonary toxicity of Co (II) ions [28].
and a number of other metal carbides also apply to the In a later in vitro study on lung slices [29], it was found,
situation encountered with the diamond polishers. The however, that, even at a dose which induced cell dys-
latter were certainly not exposed to tungsten carbide or function, the extent of glutathione oxidation by Co (II)
any of the other well-known metal carbides used for mak- was not of a sufficient magnitude to result in oxidation
ing hard metal, but it has to be admitted that we still do of protein thiol groups, an event which is likely to con-
not know the exact physicochemical characteristics (i.e. stitute the critical consequence of glutathione oxidation
the speciation) of the cobalt, carbon and iron which made in the toxic process. We, therefore, concluded that although
up the bulk of the dusts from these diamond polishing cobalt definitely caused oxidant stress, the oxidation of
workshops. The testing of the toxicity of mixtures of glutathione was, in itself, probably not the direct cause
cobalt and diamond particles should help to elucidate this of the cellular damage induced by cobalt ions.
issue. With the deoxyribose assay to monitor hydroxyl radi-
cal production it was not possible to properly compare
cobalt and hard metal particles due to an interfering reac-
Cytokines and growth factors tion of the carbide with the substrate used. Among the
different antioxidants tested for their ability to reduce the
Cytokines and growth factors have been widely inves- toxicity of WC-Co particles towards macrophages, only
tigated as mediators and markers of pulmonary toxic- butylated hydroxytoluene (BHT) which is a highly lipo-
ity. In particular, attention has been focused on their philic compound was effective, suggesting that peroxi-
role in the lung response to inhaled inorganic dust. In dation of lipidic membranes may constitute a critical step
particular, the potential involvement of cytokines in the in the toxicity of these particles [30, 31]. In hamster lung
pathogenesis of hard metal disease has been suggested slices, we also found that BHT was able to mitigate the
by the demonstration of antigenic tumour necrosis factor- cellular toxicity of cobalt chloride (Lewis et al., unpub-
α (TNF-α) in one case of this disease [24]. In our lab- lished experiments). Recently, using electron spin reso-
oratory, experimental exposure (in vitro or ex-vivo) to nance (ESR) and electrochemical techniques, we have
cobalt alone or to hard metal particles did not stimulate shown that the unique toxicity of WC-Co particles is due
the production of interleukin-1 (IL-1), TNF-α or fibronec- to the production of AOS resulting from the interaction
tin by rat alveolar macrophages [25]. While these results between cobalt metal and tungsten carbide particles (pre-
indicate that the production of these cytokines by alve- sumably hydroxyl radicals) [31]. The precise mechanism
olar phagocytes probably does not contribute to the acute of this interaction is as follows (fig. 2): 1) cobalt metal
lung toxicity of hard metal powder, it cannot be exclud- is thermodynamically able to reduce ambient oxygen but,
ed that other cytokines (e.g. platelet-derived growth fac- due to the surface characteristics of the Co particles, the
tor (PDGF), transforming growth factor-β (TGF-β)), rate of this reaction is very low (cobalt metal is resis-
possibly produced by other cell types (e.g. epithelial tant to oxidation); 2) tungsten carbide is an inert mate-
cells), contribute to the inflammatory process and this rial unable to react with oxygen by itself, but it is a fairly
issue deserves further investigation. good electron conductor and possesses unique surface
properties, which are used in numerous catalysis processes
Activated oxygen species (e.g. to replace platinum in combustion systems and in
petroleum chemistry); and 3) when both particles are
The production of activated oxygen species (AOS) is associated, electrons provided by cobalt metal are easily
another attractive and fashionable hypothesis often in- transferred to the surface of carbide particles, where
voked to explain the toxicity of inhaled inorganic parti- reduction of oxygen can occur at a greatly increased rate.
cles. With regard to cobalt, some recent publications have It is important to emphasize that in this system, solu-
suggested that its toxicity might be mediated by the pro- ble cobalt ions constitute the product and not the source
duction of hydroxyl radicals generated through a Fenton- of the critical reaction and this is very different from the
like reaction, where cobalt replaces ferrous ions [26]. The situation occurring in a Fenton-like reaction. This pro-
possible contribution of AOS in the pathogenesis of the vides a good explanation for the higher solubility and
lung disease observed in diamond polishers was sugges- bioavailability of cobalt when it is associated with a car-
ted in the discussion of a case report describing a rapid bide; the latter is not the cause of the unique toxicity but
deterioration of cobalt lung after oxygen administration rather a consequence of the reaction which generates the
[27]. Different approaches have been used in our respec- toxic species. It now remains to be demonstrated that
tive laboratories to test this hypothesis. AOS are also produced in vivo and effectively cause the
The intratracheal instillation of cobalt chloride (1–1000 toxicity. If this could be confirmed, it may offer a rea-
µg· kg-1) into hamster lungs induced biochemical changes sonable explanation for the fact that only a small pro-
compatible with the development of an oxidative stress: portion of workers exposed to hard metal powders (1–5%)
decreased levels of reduced glutathione, increased levels develop interstitial disease. It may be speculated that indi-
of oxidized glutathione, as well as stimulation of the pen- viduals with lower antioxidant defence are more sus-
tose phosphate pathway. These changes occurred early ceptible to the toxic effect of AOS produced by inhaled
and following the administration of amounts of cobalt hard metal particles. However, other hypotheses, includ-
well below those causing any lung injury. Similar bio- ing immune reactions, are not excluded (see above).
chemical changes were observed in vitro after incubat- In conclusion, starting from various clinical and epi-
ing hamster lung slices with cobalt chloride (0.1–10 mM); demiological observations, our research teams have car-
here again, these manifestations preceded the detection ried out experimental work, both in animals and in vitro,
1028 D. LISON ET AL.

WC O2 14. Lasfargues G, Lison D, Maldague P, Lauwerys R.

Co Comparative study of the acute lung toxicity of pure
cobalt powder and cobalt-tungsten carbide mixture in the
rat. Toxicol Appl Pharmacol 1992; 112: 41–50.
×e-
15. Lasfargues G, Lardot C, Delos M, Lauwerys R, Lison
D. The delayed lung responses to single and repeated
intratracheal administration of pure cobalt and hard metal
powders in the rat. Environ Res 1995; 69: 108–121.
16. Schepers GWH. The biological action of tungsten car-
Co×+ -
bide and cobalt. Arch Ind Health 1955; 12: 140–146.
O2 ,..., OH 17. Delahant AB. Experimental study of the effects of rare
Fig. 2. – Proposed mechanism of oxygen reduction by a mixture of metals on animal lungs. Arch Ind Health 1955; 12:
cobalt (Co) and tungsten carbide (WC) particles. When particles are 116–120.
in close contact, electrons (e-) provided by cobalt metal and transferred 18. Kerfoot E, Frederick W, Domeier E. Cobalt metal inhala-
to the surface of WC particles can reduce oxygen and generate acti- tion studies on miniature swine. Am Ind Hyg Assoc J
vated oxygen species, and oxidized cobalt (Cox+) passes into solution. 1975; 36: 17–25.
(From [31] with permission).
19. Nemery B, Erna J, Verbeken EK, Lauweryns JM, Demedts
M. Pulmonary toxicity in the hamster of intratracheally
to understand the mechanisms of hard metal or cobalt administered cobalt particles mixed with tungsten carbide,
lung, which is a fascinating and fairly unique type of diamond or iron (Abstract). Eur Respir J 1990; 3: 341s.
interstitial lung disease. Apart from the possible direct 20. Lison D, Lauwerys R. In vitro cytotoxic effects of cobalt
implications of our research with regard to occupational containing dusts on mouse peritoneal and rat alveolar
health, such as the setting of acceptable air concentra- macrophages. Environ Res 1990; 52: 187–198.
tions for cobalt and cobalt-containing dusts, our research 21. Lison D, Lauwerys R. The interaction of cobalt metal
has also led, or will lead, to interesting spin-offs in the particles with different carbides on mouse peritoneal
wider fields of metal toxicity, oxidant cell injury, inflam- macrophages. Toxicol in Vitro 1995; 9: 341–347.
matory lung disease and pulmonary immunology. 22. Lison D, Lauwerys R. Study of the mechanism respon-
sible for the elective toxicity of tungsten carbide-cobalt
powder toward macrophages. Toxicol Letters 1992; 60:
References 203–210.
23. Lison D, Lauwerys R. Cobalt bioavailability from hard
1. Lahaye D, Demedts M, Van den Oever R, Roosels D. metal particles: further evidence that cobalt alone is not
Lung diseases among diamond polishers due to cobalt? responsible for the toxicity of hard metal particles. Arch
Lancet, 1984; i: 156–157. Toxicol 1994; 68: 528–531.
2. Demedts M, Gheysens B, Nagels J, et al. Cobalt lung 24. Rolfe MW, Paine R, Davenport RB, Strieter RM. Hard
in diamond polishers. Am Rev Respir Dis 1984; 130: metal pneumoconiosis and the association of tumor nec-
130–135. rosis factor-alpha. Am Rev Respir Dis 1992; 146:1600–
3. Nemery B, Casier P, Roosels D, Lahaye D, Demedts M. 1602.
Survey of cobalt exposure and respiratory health in dia- 25. Huaux F, Lasfargues G, Lauwerys R, Lison D. Lung
mond polishers. Am Rev Respir Dis 1992; 145: 610–616. toxicity of hard metal particles and production of inter-
4. Cugell DW. The hard metal diseases. Clin Chest Med leukin-1, tumor necrosis factor-α, fibronectin, and cys-
1992; 13: 269–279. tatin-c by lung phagocytes. Toxicol Appl Pharmacol
5. Balmes JR. Respiratory effects of hard metal dust expo- 1995; 132: 56–62.
sure. Occup Med State of the Art Reviews 1987; 2: 327– 26. Moorhouse CP, Halliwell B, Grootveld M, Gutteridge
344. JMC. Cobalt(II) ion as a promoter of hydroxyl radical
6. Nemery B. Metal toxicity and the respiratory tract. Eur and possible "crypto-hydroxyl" radical formation under
Respir J 1990; 3: 202–219. physiological conditions: differential effects of hydroxyl
7. Abraham JL. Exposure to hard metal (Letter to the edi- radical scavengers. Biochim Biophys Acta 1985; 843:
tor). Chest 1985; 87: 554. 261–268.
8. Cullen MR. Exposure to hard metal (Letter to the edi- 27. Nemery B, Nagels J, Verbeken E, Dinsdale D, Demedts
tor). Chest 1985; 87: 554. M. Rapidly fatal progression of cobalt lung in a diamond
9. Demedts M, Ceuppens JL. Respiratory diseases from polisher. Am Rev Respir Dis 1990; 141: 1373–1378.
hard metal or cobalt exposure: solving the enigma. Chest 28. Lewis CPL, Demedts M, Nemery B. Indices of oxida-
1989; 95: 2–3. tive stress in hamster lung following exposure to cobalt(II)
10. Swennen B, Buchet JP, Stanescu D, Lison D, Lauwerys ions: in vivo and in vitro studies. Am J Respir Cell Mol
R. Epidemiologic survey on workers exposed to cobalt Biol 1991; 5: 163–169.
oxides, cobalt salts and cobalt metal. Br J Ind Med 1993; 29. Lewis CPL, Demedts M, Nemery B. The role of thiol
50: 835–842.
oxidation in cobalt(II)-induced toxicity in hamster lung.
11. Harding HE. Notes on the toxicology of cobalt metal.
Biochem Pharmacol 1992; 43: 519–525.
Br J Ind Med 1950; 7: 76–78.
12. Kaplun ZS, Mezencewa NW. Experimentalstudie über 30. Lison D, Lauwerys R. Evaluation of the role of reac-
die toxische Wirkung von Staub bei der Erzeugung von tive oxygen species in the interactive toxicity of carbide-
Sintermetallen. J Hyg Epidemiol Microbiol Immunol cobalt mixture in macrophages in culture. Arch Toxicol
1960; 4: 390–399. 1993; 67: 347–351.
13. Kitamura H, Yoshimura Y, Tozawa T, Koshi K. Effects 31. Lison D, Carbonnelle P, Mollo L, Lauwerys R, Fubini
of cemented tungsten carbide dust on rat lungs following B. Physicochemical mechanism of the interaction between
intratracheal injection of saline suspension. Acta Pathol cobalt metal and carbide particles to generate toxic acti-
Jpn 1980; 30: 241–253. vated oxygen species. Chem Res Toxicol 1995; 8: 600–606.