6
Thorax 1993;48:390-395 British Association for Lung Research review series New perspectives on basic mechanisms in lung disease 5 S Series editor: Duncan F Rogers Respirable industrial fibres: mechanisms of pathogenicity K Donaldson, R C Brown, G M Brown Background Fibrous materials have much to commend them in industrial applications, where they can offer, in varying degree, reinforcement, thermal and electrical insulation, flexibility, and strength. In the first part of this century the fibrous material that was available, cheaply and in bulk, that could provide these properties to industry was the asbestos group of silicate minerals (table 1). The fibrous, crystalline, asbestiform habit, however, also means that during mining and industrial use fibres are released into the air. Some of these are in the respirable size range and may be deposited in the bronchiolar-alveolar region. Subsequently we have come to realise that asbestos inhalation is associated with several different lung diseases-namely, asbestosis (interstitial fibrosis), carcinoma of the lung, mesothelioma, and small airways disease- and the association is now well documented. This information on the harmful effects of asbestos has led to bans in some countries and considerable litigation and claims for damages throughout the developed world. Regulatory and legal pressures on asbestos have contributed to an increase in the use of alternative materials, both natural and man- made, including some fibrous materials. More importantly, advances in materials sci- ence have identified ar, ever widening range of applications in which the fibrous nature of a material offers positive advantages. Thus in addition to their traditional use in thermal insulation new fibre composites have found new uses, such as metal reinforcement, to which asbestos was never suited (table 2). The finding that in at least some biological assays some non-asbestos fibres could cause the same types of effect as asbestos has raised the question of whether all fibres have the same pathogenic potential as asbestos. A major approach in present research on fibres is therefore to compare any fibre under study with asbestos and to gauge the potential for Table 1 The asbestos minerals Serpentine group Amphibole group Chrysotile Crocidolite Amosite Anthophyllite Tremolite Actinolite Table 2 Industrialfibres other than asbestos Natural Manmade Erionite Slag wool Wollastonite Rock wool Attapulgite Glass wool Sepiolite Continuous filament glass Halloysite Ceramics Alumina Zirconia Silicon carbide Graphite Boron Aromatic amide causing lung disease in the light of experience with asbestos. Thus the spectre of asbestos permeates all discussion of research into res- pirable fibres, of whatever origin, leading to a complex interplay of science, politics, and commercial pressure. This review aims to address only the scien- tific ideas that dominate research into the mechanisms underlying the development of diseases caused by respirable industrial fibres. The constraints of space and the breadth of the published research prevent this review from being comprehensive. Further details can be obtained from recent reviews on mechanisms of fibre pathology' and on non- asbestos fibres and their health effects in general.24 Role of fibre size, shape, and chemistry DEPOSITION A fibre, as defined by the World Health Organisation for measurement of the expo- sure of workers to airborne fibres in factories and mines, is a structure longer than 5 ,um, less than 3 gm in diameter, and with an aspect ratio (length:diameter) greater than 3:1. The deposition of a particle in the lung is a function of the "aerodynamic diameter" (particle size expressed in terms of settling speed). For fibres this is governed predomi- nantly by the actual diameter, density and length being of subsidiary importance. Only fibres with a physical diameter of 3 ,um or less may be deposited in the alveolar region, the efficiency of deposition increasing with decreasing diameter to a physical diameter of about 1 um; beyond this the efficiency of deposition may begin to decrease again. Subsequently thin fibres will be deposited in the alveolar region of the lung even if they are Napier University, Edinburgh EHIO 5DY K Donaldson Institute of Occupational Medicine, Edinburgh EH8 9SU K Donaldson G M Brown Medical Research Council Toxicology Unit, Carshalton, Surrey SM5 4EF R C Brown Reprint requests to: Dr K Donaldson, Institute of Occupational Medicine, Edinburgh EH8 9SU 390 on March 4, 2021 by guest. Protected by copyright. http://thorax.bmj.com/ Thorax: first published as 10.1136/thx.48.4.390 on 1 April 1993. Downloaded from

Respirable industrial fibres: mechanisms of · long fibres and the other composed exclu-sively of short fibres. The short fibres were separated from the long by milling and sedi-mentation

  • Upload
    others

  • View
    4

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Respirable industrial fibres: mechanisms of · long fibres and the other composed exclu-sively of short fibres. The short fibres were separated from the long by milling and sedi-mentation

Thorax 1993;48:390-395

British Association for Lung Research review series

New perspectives on basic mechanisms in lung disease 5SSeries editor: Duncan F Rogers

Respirable industrial fibres: mechanisms ofpathogenicity

K Donaldson, R C Brown, G M Brown

BackgroundFibrous materials have much to commendthem in industrial applications, where theycan offer, in varying degree, reinforcement,thermal and electrical insulation, flexibility,and strength. In the first part of this centurythe fibrous material that was available,cheaply and in bulk, that could provide theseproperties to industry was the asbestos groupof silicate minerals (table 1). The fibrous,crystalline, asbestiform habit, however, alsomeans that during mining and industrial usefibres are released into the air. Some of theseare in the respirable size range and may bedeposited in the bronchiolar-alveolar region.Subsequently we have come to realise thatasbestos inhalation is associated with severaldifferent lung diseases-namely, asbestosis(interstitial fibrosis), carcinoma of the lung,mesothelioma, and small airways disease-and the association is now well documented.This information on the harmful effects ofasbestos has led to bans in some countriesand considerable litigation and claims fordamages throughout the developed world.Regulatory and legal pressures on asbestoshave contributed to an increase in the use ofalternative materials, both natural and man-made, including some fibrous materials.More importantly, advances in materials sci-ence have identified ar, ever widening rangeof applications in which the fibrous nature ofa material offers positive advantages. Thus inaddition to their traditional use in thermalinsulation new fibre composites have foundnew uses, such as metal reinforcement, towhich asbestos was never suited (table 2).The finding that in at least some biological

assays some non-asbestos fibres could causethe same types of effect as asbestos has raisedthe question of whether all fibres have thesame pathogenic potential as asbestos. Amajor approach in present research on fibresis therefore to compare any fibre under studywith asbestos and to gauge the potential for

Table 1 The asbestos minerals

Serpentine group Amphibole group

Chrysotile CrocidoliteAmositeAnthophylliteTremoliteActinolite

Table 2 Industrialfibres other than asbestos

Natural Manmade

Erionite Slag woolWollastonite Rock woolAttapulgite Glass woolSepiolite Continuous filament glassHalloysite Ceramics

AluminaZirconiaSilicon carbideGraphiteBoron

Aromatic amide

causing lung disease in the light of experiencewith asbestos. Thus the spectre of asbestospermeates all discussion of research into res-pirable fibres, of whatever origin, leading to acomplex interplay of science, politics, andcommercial pressure.

This review aims to address only the scien-tific ideas that dominate research into themechanisms underlying the development ofdiseases caused by respirable industrial fibres.The constraints of space and the breadth ofthe published research prevent this reviewfrom being comprehensive. Further detailscan be obtained from recent reviews onmechanisms of fibre pathology' and on non-asbestos fibres and their health effects ingeneral.24

Role of fibre size, shape, and chemistryDEPOSITIONA fibre, as defined by the World HealthOrganisation for measurement of the expo-sure of workers to airborne fibres in factoriesand mines, is a structure longer than 5 ,um,less than 3 gm in diameter, and with anaspect ratio (length:diameter) greater than3:1. The deposition of a particle in the lung isa function of the "aerodynamic diameter"(particle size expressed in terms of settlingspeed). For fibres this is governed predomi-nantly by the actual diameter, density andlength being of subsidiary importance. Onlyfibres with a physical diameter of 3 ,um or lessmay be deposited in the alveolar region, theefficiency of deposition increasing withdecreasing diameter to a physical diameter ofabout 1 um; beyond this the efficiency ofdeposition may begin to decrease again.Subsequently thin fibres will be deposited inthe alveolar region of the lung even if they are

Napier University,Edinburgh EHIO 5DYK DonaldsonInstitute ofOccupationalMedicine, EdinburghEH8 9SUK DonaldsonG M BrownMedical ResearchCouncil ToxicologyUnit, Carshalton,Surrey SM5 4EFR C BrownReprint requests to:Dr K Donaldson, Instituteof Occupational Medicine,Edinburgh EH8 9SU

390 on M

arch 4, 2021 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.48.4.390 on 1 April 1993. D

ownloaded from

Page 2: Respirable industrial fibres: mechanisms of · long fibres and the other composed exclu-sively of short fibres. The short fibres were separated from the long by milling and sedi-mentation

Respirable industnialfibres: mechanisms ofpathogenicity

very long (up to 30 or 40 pm). In contrast,for compact particles density is more impor-tant, and particles with diameters greater thanabout 5 pm are unlikely to reach the alveolarregion. The site of the deposition of fibres inthe lungs of rats was elucidated by the elegantstudies of Brody et al5 and Warheit et al 6 whoused microdissection and scanning electronmicroscopy to follow the fate of chrysotileasbestos fibres delivered to the rat lung dur-ing a single exposure. These fibres were

found to be deposited almost exclusivelyaround the terminal bronchioles and the firstalveolar duct bifurcations, where the fibrescould first be seen lying free and later phago-cytosed by macrophages. Our own studies,using amosite asbestos and glass fibres, haveconfirmed this sequence of events and fibresup to 35 pm long can be seen lying on theepithelium of the terminal bronchioles andthe alveolar ducts and inside macrophages(fig 1).

FIBRE LENGTH AND PATHOGENIC POTENTIALThere has been some progress towardsunderstanding the attributes of fibres thatimbue them with the potential for causingpathogenic change. Pioneering studies, suchas those of Wright and Kuschner,7 andStanton and Layard,8 showed that short fibres(below 5 pm) are low in pathogenicity andthat increasing pathogenicity accompaniesincreasing fibre length beyond 8-10 pm. Inparticular, the impressive studies of Stantonand co-workers8 using implantation of fibresinto the pleural space in the rat, showed thata range of quite different fibrous materialscould all cause pleural mesotheliomas pro-

Figure 1 Scanning electron microscope image of the distal airspace ofa rat that had beeninhaling a cloud of 1000 fibreslml of amosite asbestos for seven hours. An alveolar duct isshown with a portion of terminal bronchiolar epithelium present (T) and openings intoalveoli (A). Asbestos fibres can be seen on the alveolar duct surface, both free and partiallyphagocytosed inside alveolar macrophages (M). (Scale bar = 40,um.)

vided that there were fibres of the appropriatesizes. The optimum size range for theproduction of mesotheliomas in this systemwas over 8 pm long and below 0-25 pm indiameter.8

In the early 1980s the Institute ofOccupational Medicine in Edinburghobtained two samples of amosite asbestos,one containing a substantial proportion oflong fibres and the other composed exclu-sively of short fibres. The short fibres wereseparated from the long by milling and sedi-mentation. The two samples were very similarin all other respects, such as diameter, ele-mental composition, and crystallographic fea-tures. Milling reduced a substantialproportion (63%) of the particles in the shortfibre sample to an aspect ratio of less than3:1-that is, they could not be defined asfibres. The remaining 37% were fibres by thiscriterion with an average length of 2 7,pm.These two samples have now been extensivelystudied by inhalation and instillation andsome of the results are summarised in table 3.We believe this to be some of the most com-pelling evidence showing the importance offibre size in determining the biological activityof respirable fibres.One important factor that contributes to

the pathogenic potential of long fibres is theevidence from several studies that long fibresare cleared from the deep regions of the lungless effectively than are short fibres." 12 Thismay be a consequence of failure of macro-phages to ingest long fibres completely or ofdifficulty in moving once they have phago-cytosed a long fibre. Long fibres may eveninteract with several macrophages, which maynot then be able to act in concert. Such areaction might mark the onset of a granulo-matous response. With regard to the latterpossibility, macrophages from the lungs ofrats exposed to asbestos by inhalation havebeen found to have decreased ability tomigrate towards a chemotaxin."3 Thus longfibres could be more harmful to the lungbecause they are retained, whereas the shortfibres are cleared more effectively. Butalthough this presumably contributes to thetoxicity of long fibres the evidence from stud-ies using some in vitro assay systems that donot rely on clearance (see, for example,Brown et al 14 and Donaldson et al 15) confirmsthat longer fibres have intrinsically more bio-logical activity.

Table 3 Biological activity oflong and shortfibresamples of amosite asbestos in animal studies *

Amositefibres

Response measure Short Long

Inhalation % with lung tumours (rats) 0 32-5Mean fibrosis score (rats) 0 15-6Intraperitoneal instillation% with mesotheliomas (rats) 0 88Inflammation score (mice) -/+ + + + +

*Data summarised from experiments described in Daviset aP and Donaldson et al.IO

391 on M

arch 4, 2021 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.48.4.390 on 1 April 1993. D

ownloaded from

Page 3: Respirable industrial fibres: mechanisms of · long fibres and the other composed exclu-sively of short fibres. The short fibres were separated from the long by milling and sedi-mentation

Donaldson, Brown, Brown

ROLE OF CHEMICAL CHARACTERISTICS OF THEFIBRE SURFACEThus the role of fibre size has been wellinvestigated and shown to be important, butless attention has been given to the chemistryof fibres. Of particular interest in this respectis erionite. This naturally occurring fibrousmineral of the zeolite family has been impli-cated in a high prevalence of mesothelioma inthe small Turkish village of Karain, where itis present in the air owing to the weatheringof rock.'6 In animal studies erionite producedan incidence of mesothelioma far higher thanany other fibre, even though the dimensionsof the fibres were not greatly different fromthose of fibres with less pathological poten-tial.'7 Possibly therefore erionite has a compo-sition that contributes unusually to itspathogenicity. Further evidence that the fibresurface can contribute to the biological activ-ity comes from a study by Brown et al,'8 whosubstituted hydrocarbon chains of differentlength on the surface of amosite fibres with-out altering the number or length of thefibres. This treatment appears to alter theability of the fibres to cause mesotheliomaafter instillation into the pleura.

There is also clear evidence in vitro thatsimply altering the protein that is attached tothe fibre surface will change the biologicalactivity. Treating asbestos fibres with ratimmunoglobulin G in vitro dramaticallyincreased its ability to stimulate macrophagesto release the important inflammatory media-tors superoxide anion'9 and tumour necrosisfactor.'5 Presumably this is a direct effectmediated via the macrophage membranereceptors for the Fc portion of immunoglobu-lin G. Further research on the chemicalnature of the surface of fibres and of thereceptors responsible for the interactions offibres with the cell surface are clearly war-ranted. In this regard Brown et a120 recentlyimplicated the fibronectin tripeptide receptorin the binding of amosite asbestos to cells invitro.The postulated role of free radicals in the

toxicity of fibres2'22 has focused interest in therole of iron at the fibre surface. Because ofthe ability of iron to enhance and catalyse thereactions of free radicals it has been suggestedthat fibres may act as a source of iron in tis-sue, leading to injury by free radicals. Theevidence that iron is important has comefrom studies where iron chelators haveameliorated the toxicity of fibres to cells invitro.2' 2324 Recently iron from crocidoliteasbestos has been shown to be important inthe causation of breaks in DNA strands25 andin the formation of 8-hydroxydeoxyguanosine(RC Brown, unpublished findings). Althoughiron may indeed be important in the toxiceffects of types of fibre with a substantial con-tent of iron, other fibres, such as chrysotileasbestos and some fibrous glasses, have littleor no iron. In other cases it is the iron,mobilised from the fibre, that causes a shortterm effect; the relevance of this to the longterm effect of fibres in vivo is not clear assuch fibres become coated with biological

material, which itself may contain iron.Indeed, many fibres form ferruginous(asbestos) bodies, at least in man, so that ironis deposited on the fibre rather than leachedfrom it. Iron derived from fibres is thereforeunlikely to be an important unifying factor infibre toxicity, but the relative importance ofthe surface and the dimensions of fibres hasyet to be determined fully.One area where fibre chemistry is likely to

be an important factor in determining biolog-ical activity relates to the ability of fibres topersist in the lung once deposited there.Fibres that dissolve or break down in the lungare less likely to be dangerous because theymay become shorter; short fibres are lessactive than longer fibres and are more effi-ciently cleared. In addition, epidemiologicalstudies in workers exposed to the relativelysoluble chrysotile asbestos suggest that thisparticular fibre is much less harmful in manthan are other forms of asbestos.

Experiments with chrysotile fibres thathave been treated to mimic the dissolutioneffects of residence in the lungs have shownthat they do indeed have decreased biologicalactivity26 and are removed faster by the clear-ance systems of the lung." As most glassyfibres are likely to be more soluble in the lungthan even chrysotile asbestos possibly thesewill represent less hazard.A substantial body of research at present is

concerned with the persistence of variousfibre types within the lung. In animal experi-ments even "soluble" fibres may persist in thelungs of rodents for a substantial portion ofthe animal's lifespan and thus are able tocause disease, whereas the same fibres persist-ing for the same time may not have the sameeffect in humans with their longer lifespan.This raises the possibility that "false positive"results might occur in animals because fibresare cleared by purely physicochemical meansunrelated to the animals' lifespan. Negativeresults in such animal experiments shouldtherefore be more reassuring than similarresults with purely chemical toxins.

Mechanisms of disease caused by fibresA substantial body of research has beendirected at the interactions between fibresand cells in the hope of understanding themechanism whereby deposited fibres causethe principal diseases associated with expo-sure.

EPrITELIAL CELLsBrody et a127 showed that inhaled chrysotileasbestos fibres could be found inside epithe-lial cells, and this was interpreted as a mecha-nism whereby fibres were translocated to theinterstitium. Hobson et a128 have also shownthe uptake of fibres by airway epithelial cellsin vitro. The effect of the interaction of fibreswith epithelial cells is both frank toxicity'424and, in some studies, a low level of sublethalgenetic damage.2930 As well as acting as aninitiator in causing direct DNA damage,asbestos is able to increase the frequency of

392

on March 4, 2021 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.48.4.390 on 1 A

pril 1993. Dow

nloaded from

Page 4: Respirable industrial fibres: mechanisms of · long fibres and the other composed exclu-sively of short fibres. The short fibres were separated from the long by milling and sedi-mentation

Respirable industrialfibres: mechanisms ofpathogenicity

transformation effects caused by other car-cinogens at doses at which asbestos alone hasno effect.'1 32 These observations lead to thecontention that asbestos acts as a tumourpromoter, and indeed several different studieshave found that asbestos has activities similarto those of classical promoters."

Increased proliferation of epithelial cells isan early event following inhalation exposureto chrysotile asbestos in rats.'4 Increased cellproliferation is a common effect of pro-moters33 and is likely to represent the restora-tive response to epithelial damage, combinedwith the response to local accumulation ofinflammatory cell derived growth factors. Arole for epithelial hyperplasia in the develop-ment of neoplasia associated with fibrestherefore appears likely.

MACROPHAGES AND INFLAMMATIONMost deposited fibres are likely to be phago-cytosed by alveolar macrophages if theyremain on the lung surface for long enough(fig 2). Early research showed the ability offibres to injure macrophages in vitro, andshortening of the fibres by milling reducedthe toxic effects.'5 At that time fibre mediateddeath of macrophages was seen as an impor-tant pathogenic step. But, although fibres(mostly asbestos) were able to kill macro-phages after phagocytosis a dose-responserelationship could be seen, and low concen-

trations of fibre were not frankly toxic. As theroles of the macrophage became betterunderstood and the subtleties of its inter-actions with other cells via various mediatorsbecame clear, research converged on the sub-lethal effects of low concentrations of fibres.The idea that fibres at low dose could stimu-late macrophages to change their secretoryrepertoire has therefore come to predominate.

Figure 2 Scanning electron microscope image ofan alveolar macrophage from the lungsofa rat exposed to amosite asbestos (see legend to fig 1). A single longfibre, or two shorterfibres, can be seen protrudingfrom the alveolar macrophage, which appears to be acti-vated. (Scale bar = 7.5 ym.)

Particular attention has been given to pro-teases, cytokines, and oxidants because thesehave an obvious potential for mediatinginflammation and pathological change. Table4 shows the various secretory products ofmacrophages that have been found to beincreased by exposure to fibres in vitro or invivo.

Stimulation of macrophages by fibres tosecrete their products, along with the epithe-lial events outlined above, culminates ininflammation, characterised by recruitment ofmore macrophages and neutrophils.41 Ahypothesis can be proposed about the eventsthat occur as a consequence of this localisedaccumulation of inflammatory cells at the ter-minal bronchiolar and proximal alveolarregion. Oxidants cause local injury and socontribute to loss of epithelial integrity,allowing passage of cells and proteins fromthe interstitium to the alveolar space; fibresand macrophages and their products may alsogain access to the interstitium. Macrophagemolecules that stimulate the mesenchymalcells (platelet derived growth factor, tumournecrosis factor, etc) can then signal anincrease in production of fibroblasts and theirmatrix products, leading to fibrosis of thesmall airway walls and the alveolar walls.Oxidants and growth factors may contributeto "promoting" effects, such as proliferation,and so contribute to neoplastic change.

THE PLEURA AND MESOTHELIAL CELLSBecause of the association between asbestosexposure and mesothelioma, the effects offibres on pleural mesothelial cells have beenthe subject of many investigations.Mesothelial cells in culture have beenreported to be particularly susceptible to thetoxic effects of fibres, far more so than epithe-lial cells or fibroblasts.42 Mesothelial cellsphagocytose fibres, leading to chromosomalabnormalities43 and an increase in unsched-uled DNA synthesis as evidence of DNArepair following damage."

With regard to pleural fibrosis, themacrophage events described above in rela-tion to alveolar fibrosis could occur in themost peripheral alveolar units. This would bein close proximity to, and may even be con-tinuous with, the subpleural connective tissueand could lead to localised pleural fibrosis.

Although no clear role for the interactionsbetween mesothelial cells and the leucocytepopulation of the pleural cavity has emerged,changes in the activity of the pleural

Table 4 Some secretory products produced bymacrophages exposed to asbestos

Asbestosfibre Secretory product Reference

Chrysotile Lysosomal hydrolase 36Crocidolite Chemoattractant 37Chrysotile Platelet derived growth

factor 38Chrysotile Superoxide anion 39Chrysotile Arachidonic acid

metabolites 40Amosite Tumour necrosis factor 15

393 on M

arch 4, 2021 by guest. Protected by copyright.

http://thorax.bmj.com

/T

horax: first published as 10.1136/thx.48.4.390 on 1 April 1993. D

ownloaded from

Page 5: Respirable industrial fibres: mechanisms of · long fibres and the other composed exclu-sively of short fibres. The short fibres were separated from the long by milling and sedi-mentation

Donaldson, Brown, Brown

leucocytes after deposition of asbestos fibresin the lung have been described.45 Suchchanges could contribute to the developmentof pleural disease.

FIBRES IN COMBINATION WITH OTHER AGENTSThe clear multiplicative effects of smokingand asbestos exposure in causing lungcancer,46 and the activity of asbestos as a pro-moter as seen experimentally, raise the ques-tion of combined exposures to fibres andother agents. This has not yet been studied indepth but Pinkerton et a147 have shown thatlow dose exposure to ozone results in a dra-matic slowing down of the clearance ofasbestos from the lung. Davis et a148 haveshown that inhaling asbestos fibres in combi-nation with other dust, such as titanium diox-ide or quartz, may produce substantially morelung and pleural tumours than inhaling fibresalone. Thus we need to consider fibre expo-sure in relation to concomitant environmentalexposures to oxidants and dusts.

Future research on respirable industrialfibresAlmost all that is known about the mecha-nisms of the diseases caused by inhaled fibresis based on experience with asbestos. The dis-semination of a set of standard samples ofasbestos (Unione Internationale ContreCancer (UICC) samples produced by theBritish and South African Medical ResearchCouncils) was a vital step in allowing inter-laboratory comparisons to be made of thebiological activity of at least these asbestossamples. Whereas elucidation of the cell biol-ogy of asbestos bioeffects has moved aheadrapidly no standard samples of non-asbestosfibres have been generally available to the sci-entific community. Such a standard setshould be well characterised with regard todimensions and surface chemistry. The lackof standard non-asbestos fibres has meantthat progress has stalled with regard to verybasic questions on the similarities and differ-ences between asbestos and other fibres. TheUICC samples were made in 1969 before the

THINIRespirable

More biologically active than thick fibres

? CHEMISTRY

LONGIncompletely phagocytosed

More biologically active than short fibresLess easily cleared than short fibres

DURABLE IPersistent

role of fibre length was fully appreciated; theywere made by milling commercial asbestos toensure that each sample contained mostlyrespirable fibres. These fibres, however, weremostly short and the resulting activity of theUICC series is therefore far from optimal forbiological research. Nearly all the experi-ments carried out so far have used samplescontaining less than 1% of the long activefraction.

Both this dimensional problem and theabsence of standard samples of the manmadefibres have been partially remedied by theUnited States Thermal InsulationManufacturer's Association (TIMA). Thisorganisation and the equivalent Europeanbodies have sponsored a series of animalexperiments on their products. To do thisthey isolated fibres with target mean dimen-sions of 20 4um length and 1 ,um diameterfrom commercial mineral wools. Such fibrescould be expected to be extremely biologi-cally active if dimension is the main determi-nant of activity. As well as carrying out theirown programme of research with these mate-rials the Thermal Insulation ManufacturersAssociation has established a repository,which is making a range of respirable vitreousand ceramic fibres and one long fibre sampleof crocidolite available to bona fideresearchers. This generous step should greatlyimprove fibre research.The asbestos experience has been the dri-

ving force for most research into the newerfibres and this has engendered a cautiousapproach to any respirable fibre. Currently,the idea that any long, thin, durable fibre hasthe potential to be pathogenic dominates thestudy of fibre pathology (fig 3) but there is nogeneral consensus on secure definitions oflong, thin, or durable; the question of the roleof chemical composition, particularly as itaffects durability, continues to be vexing. Thenext 20 years of research into the physico-chemistry of fibres and their interactions withcells should reveal whether "guilt by associa-tion" with asbestos is justice for respirableindustrial fibres. A ban on fibres would beimpractical with the present state of know-ledge, as non-fibrous substitutes do not existfor all known applications. In addition, theenergy efficiency of domestic and other build-ings and industrial processes would be sub-stantially reduced, with consequent increasesin air pollution. A thorough understanding ofthe characteristics of fibres that control theirtoxicity may allow industry to design saferfibres, to the benefit of themselves, the work-force, and society.

KD and GB were supported by the Colt fibre research pro-gramme. We would like to thank Drs Geoff Pigott and BillMacNee for critically reading the manuscript and suggestingimprovements. We would also like to acknowledge the contri-bution of many colleagues in carrying out research on fibrepathogenicity.

Figure 3 The pathogenicity offibres: diagram showing the three properties that arecurrently considered to underlie pathogenicity, indicating how each property is likely tohave its biological effects.

1 Rom, W, Travis, WD, Brody, AR. Cellular and molecularbasis of the asbestos-related diseases. Am Rev Respir Dis1991;143:408-22.

2 Lippmann M. Man-made mineral fibres (MMMF):human exposures and health risk assessment. Toxicol IndHealth 1990;6:225-46.

394

on March 4, 2021 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.48.4.390 on 1 A

pril 1993. Dow

nloaded from

Page 6: Respirable industrial fibres: mechanisms of · long fibres and the other composed exclu-sively of short fibres. The short fibres were separated from the long by milling and sedi-mentation

Respirable industrialfibres: mechanisms ofpathogenicity

3 Wheeler C. Exposure to man-made mineral fibres: a sum-mary of current animal data. Toxicol Ind Health 1990;6:293-307.

4 Liddell D, Miller K, eds. Mineral fibres and health. BocaRaton, USA: CRC Press, 1991.

5 Brody AR, Hill LH, Adkins B, O'Connor RW. Chrysotileasbestos inhalation in rats: deposition pattem and reac-tion of alveolar epithelium and pulmonary macro-phages. Am Rev Respir Dis 1981;123:670-9.

6 Warheit DB, Chang LY, Hill LH, Hook GER, Crapo JD,Brody AR. Pulmonary macrophage accumulation andasbestos-induced lesions at sites of fiber deposition. AmRev Respir Dis 1984;129:301-10.

7 Wright GW, Kuschner M. The influence of varying fibrelengths of glass and asbestos fibres on tissue response inguinea pigs. In: Walton WH, ed. Inhaled particles IV. Vol2. London: Pergamon Press, 1977: 455-74.

8 Stanton MF, Layard M. The carcinogenicity of fibrousminerals. In: Proceedings of the workshop on asbestos: defi-nitions and measurements. Gaithersburg, Maryland, 1977Washington: National Bureau of Standards, 1978:143-51. (Special Publication 506.)

9 Davis JMG, Bolton RE, Donaldson K, Jones AD, SmithT. The pathogenicity of long versus short fibres ofamosite asbestos administered to rats by inhalation andintraperitoneal injection. Br Jf Exp Pathol 1984;67:415-30.

10 Donaldson K, Brown GM, Brown DM, Bolton RE, DavisJMG. The inflammation generating potential of longand short fibre amosite asbestos samples. Br J7 Ind Med1989;46:271-6.

11 Bellman B, Konig H, Muhle H, Pott F. Chemical durabil-ity of asbestos and man-made mineral fibres in vitro. JAerosol Sci 1986;17:341-5.

12 Coin PG, Roggli VL, Brody AR. Deposition, clearanceand translocation of chrysotile asbestos from peripheraland central regions of rat lung. Environ Res 1992;58:97-116.

13 Donaldson K, Brown GM, Brown DM, Slight J,Robertson MD, Davis JMG. Impaired chemotacticresponses of bronchoalveolar leukocytes in experimentalpneumoconiosis. J Pathol 1990;160:63-9.

14 Brown RC, Chamberlain M, Griffiths DM, Timbrell V.The effect of fibre size on the in vitro biological activityof three types of amphibole asbestos. Int J Cancer1978;22:721-7.

15 Donaldson K, Li XY, Dogra S, Miller BG. Asbestos-stimulated tumour necrosis factor release from alveolarmacrophages depends on fibre length and opsonisation.J Pathol 1992;168:243-8.

16 Baris YI, Sahin AA, Ozesmi M, Kerse I, Ozen E, KolacanB, et al. An outbreak of pleural mesothelioma andfibrosing pleurisy in the village of Karain/Urgup inAnatolia. Thorax 1978;33:181-5.

17 Wagner JC, Skidmore JW, Hill RJ, Griffiths DM. Erioniteexposure and mesothelioma in rats. Br J Cancer1985;51:727-32.

18 Brown RC, Carthew P, Hoskins JA, Sara E, Simpson CF.Surface modification can affect the carcinogenicity ofasbestos. Carcinogenesis 1990;11: 1883-5.

19 Scheule RK, Holian A. IgG specifically enhanceschrysotile asbestos-stimulated superoxide anion produc-tion by the alveolar macrophage. Am J Respir Cell MolBiol 1989;1:313-8.

20 Brown RC, Sara EA, Hoskins JA, Evans CE. Factorsaffecting the interaction of asbestos fibres with mam-malian cells: a study using cells in suspension. Ann OccHyg 1991;35:25-34.

21 Mossman BT, Marsh JP. Evidence supporting a role foractive oxygen species in asbestos-induced toxicity andlung disease. Environ Health Perspect 1989;81:91-4.

22 Goodlick L.A, Kane AB. Role of oxygen metabolites incrocidolite asbestos toxicity to mouse macrophagesCancer Res 1986;46:5558-9.

23 Brown RC, Poole A, Turver CJ, Vann C. Role of ironmediated free radical generation in asbestos-inducedcytotoxicity. Toxicology in Vitro 1987;1:67-70.

24 Kamp DW, Dunne M, Anderson JA, Weitzman SA,Dunn MM. Serum promotes asbestos-induced injury tohuman pulmonary epithelial cells. J Lab Clin Med 1990;116:289-97.

25 Lund LG, Aust AE. Iron-catalysed reactions may beresponsible for the biochemical and biological effects ofasbestos. Biofactors 1991;3:83-9.

26 Monchaux G, Bignon J, Jaurand MC, Lafuma J,Sebastien P, Masse R, et al. Mesotheliomas in rats fol-

lowing inoculation with acid-leached asbestos and othermineral fibres. Carcinogenesis 1981;3:229-36.

27 Brody AR, Hill LH, Stirewalt WS, Adler KB. Actin-con-tamiing microfilaments of pulmonary epithelial cells pro-vide a mechanism for translocating asbestos to theinterstitium. Chest 1983;83 (suppl): 11-2S.

28 Hobson J. Wright JL, Churg A. Active oxygen speciesmediate asbestos fiber uptake by tracheal epithelial cells.FASEB J 1990;4:3135-9.

29 Hesterberg TW, Barrett, JC. Dependence of asbestos- andmineral dust-induced transformation of mammaliancells in culture on fiber dimension. Cancer Res1991;44:2170-80.

30 Jackson JH, Schraufstatter IU, Hyslop PA. Role of oxi-dants in DNA damage. Hydroxyl radical mediates thesynergistic DNA-damaging effects of asbestos and ciga-rette smoke. J Clin Invest 1991;80:1090-5.

31 Topping DC, Nettesheim P. Two-stage carcinogenesis instudies with asbestos in Fischer 344 rats. J7 Natl CancerInst 1980;65:627-30.

32 Brown RC, Poole A, Fleming GA. The influence ofasbestos dust on the oncogenic transformation ofC3HlOT1/2 cells. Cancer Lett 1983;18:221-7.

33 Mossman BT, Craighead JE. Mechanisms of asbestos-associated bronchogenic carcinoma. In: Antman K,Aisner J, eds. Asbestos-related malignancy. Orlando,Florida: Grune and Stratton, 1986:137-50.

34 Brody AR, Overby LH. Incorporation of tritiated thymi-dine by epithelial and interstitial cells in bronchiolar-alveolar regions of asbestos-exposed rats. Am Jf Pathol1989;134: 133-40.

35 Davies R, Chamberlain M, Brown RC, Griffiths DM.Identification of toxic mineral dusts using mammaliancells. Trans R Soc Edinb Earth Sci 1980;71:181-4.

36 Davies P, Allison AC, Ackerman J, Butterfield A,Williams S. Asbestos induces selective release of lysoso-mal enzymes from mononuclear phagocytes. Nature1974;251:423-5.

37 Kagan E, Oghiso Y, Hartmnann DP. Enhanced release ofchemoattractant for alveolar macrophages after asbestosinhalation. Am Rev RespirDis 1983;128:680-7.

38 Baumann MD, Jetten AM, Bonner JC, Kumar RK,Bennett RA, Brody AR. Secretion of a platelet growthfactor homologue by rat alveolar macrophages exposedto particulates in vitro. EurJ7 Cell Biol 1990;51:327-34.

39 Donaldson K, Cullen RT. Chemiluminescence ofasbestos-activated macrophages. Br J Exp Pathol 1984;65:81-90.

40 Kouzan S, Brody AR, Nettesheim P, Eling TE. Produc-tion of arachidonic acid metabolites by macrophagesexposed in vitro to asbestos fibres, carbonyl iron andcalcium ionophore. Am Rev Respir Dis 1985;131:624-7.

41 Donaldson K, Bolton RE, Jones AD, Brown GM,Robertson MD, Slight J, et al. Kinetics of thebronchoalveolar leucocyte response in rats followingexposure to equal airborne mass concentrations ofquartz, chrysotile asbestos, or titanium dioxide. Thorax1988;43:525-33.

42 Lechner JF, Tokiwa T, Yeager H, Harris CC. Asbestos-associated chromosomal changes in human mesothelialcells. In: Beck EG, Bignon J, eds. In vitro effects of min-eral dusts. Third international workshop. Berlin: Springer,1985:197-202 (NATO Advanced Science InstituteSeries G: Vol 3.)

43 Lechner JF, Tokiwa T. LaVeck Asbestos-associated chro-mosomal changes in human mesothelial cells. Proc NadAcad Sci USA 1985;82:3884-8.

44 Renier A, Levy F, Pilliere F, Jaurand M.C. UnscheduledDNA synthesis in rat pleural mesothelial cells treatedwith mineral fibres. Mutation Res 1990;241:261-367.

45 Li XY, Brown GM, Lamb D, Donaldson K. Increasedproduction of plasminogen activator inhibitor in vitro bypleural leukocytes from rats intratracheally instilled withcrocidolite asbestos. Environ Res 1991;55:135-44.

46 Sarracci R. Asbestos and lung cancer: an Analysis of theepidemiological evidence on the asbestos-smoking inter-action. Int3 Cancer 1977;20:323-30.

47 Pinkerton KE, Brody AR, Miller FJ, Crapo JD. Exposureto ozone results in enhanced pulmonary retention ofinhaled asbestos fibers. Am Rev Respir Dis 1989;140:1075-81.

48 Davis JMG, Jones AJ, Miller BG. Experimental studies inrats on the effects of asbestos coupled with the inhala-tion of titanium dioxide or quartz. Int J Exp Pathol199 1;72:501-25.

395

on March 4, 2021 by guest. P

rotected by copyright.http://thorax.bm

j.com/

Thorax: first published as 10.1136/thx.48.4.390 on 1 A

pril 1993. Dow

nloaded from