Occupational and Environmental Medicine 1994;51:275-280

Retention and clearance of inhaled ceramic fibresin rat lungs and development of a dissolutionmodel

Hiroshi Yamato, Hajime Hori, Isamu Tanaka, Toshiaki Higashi, Yasuo Morimoto,Masamitsu Kido

AbstractMale Wistar rats were exposed to alu-minium silicate ceramic fibres by inhala-tion to study pulmonary deposition,clearance, and dissolution of the fibres.Rats were killed at one day, one month,three months, and six months after thetermination of exposure. After exposure,

fibres greater than 50 pm in length were

seen with a scanning electron microscopein the alveolar region of the lung. Fibreswere recovered from the lungs with a lowtemperature ashing technique and theirnumber, diameter, and length were mea-sured by scanning electron microscopy.The number of fibres remang in thelungs declined exponentially with timeafter exposure and their silicon contentalso fell. The geometric median diameterof fibres decreased linearly with time. Bysix months after exposure, the surface offibres recovered from the lungs had an

eroded appearance. The results suggestthat ceramic fibres are physically clearedfrom the lung and that they show signs ofdissolution. Finally, the results were usedto develop a theoretical model of fibredissolution that gives a satisfactory fit tothe experimental data.

(Occup Environ Med 1994;51:275-280)

Department ofEnvironmental HealthEngineeringH YamatoH HoriI TanakaDepartment ofWorkSystems and HealthT HigashiDivision ofRespiratory Disease,University ofOccupational andEnvironmentalHealth, JapanY MorimotoM KidoCorrespondence to:Hiroshi Yamato MD,Department ofEnvironmental HealthEngineering, University ofOccupational andEnvironmental Health,Japan, 1-1 Iseigaoka,Yahatanishi Kitakyushu,807 Japan.Accepted 1 October 1993

The physical dimensions, clearance, andbiopersistence of fibres in the lungs are

important factors in predicting their biologi-cal effects. To study the clearance of man

made mineral fibres from the lung, a numberof studies have been performed in which thefibres were given either by intratracheal injec-tion'-3 or by inhalation.4 It seems that manmade mineral fibres given by these routes are

cleared from the lung by transport to the con-

ducting airways and by dissolution. Morganet al I and Holmes et al 2 showed that shortglass fibres were cleared efficiently bymacrophage mediated processes but thatfibres exceeding a critical length (between 10and 30 pm) were not removed. Their resultsalso suggested that the rate of dissolution ofglass fibres depends on their length (longfibres dissolve more rapidly than short fibres)and animal species (glass fibres dissolvedmore rapidly in hamster than in rat lungs).The chemical composition of the fibre is alsothought to be an important factor in deter-mining the durability of fibres in the lung. LeBouffant et al showed that those fibres that

had remained for several months in the lunghad an eroded appearance and the amountsof sodium, calcium, and magnesium weredepleted.5 Leinweber studied the solubility ofsix types of man made mineral fibre in vitroand showed that the solubility of fibres variedover a wide range depending on their chemi-cal composition. He also showed that dissolu-tion rates could be calculated.6 The clearanceof inhaled fibres has also been estimated byanalysing the amount of a major component(for example, silicon) remaining in the lung.4Morgan and Holmes7 reviewed the publishedworks on the solubility of mineral fibres. Thereare few studies, however, that have measuredall these variables after inhalation exposureand discussed them comprehensively.

In this study, we exposed rats to ceramicfibres by inhalation and determined changesin the number, length, diameter, and surfacestructure of fibres at various times. At thesame time, we measured changes in the sili-con content of lungs (silicon is the main com-ponent of the ceramic fibres). An attempt wasmade to model these findings mathematically.

Materials and methodsTEST SUBSTANCESThe aluminum silicate ceramic fibres used inthis study were made by a Japanese manufac-turer. The chemical composition of the fibreis 54% SiO2 and 46% Al,0,. To obtain res-pirable fibre particles having a mass medianaerodynamic diameter of less than 5 pm, bulkceramic fibres were disintegrated three timeswith an ultracentrifugal mill (Retch Co,Germany) at a speed of 15 000 rpm.

EXPOSURE SYSTEMDetails of our system and apparatus havebeen reported in previous papers.A-2 Toobtain a constant concentration, the ceramicfibres were mixed with fluidising particles(small glass beads with a diameter of250 um). The mixture was fed into a hopperand transported smoothly via a continuousscrew feed into a fluidised bed. Dry air flow-ing upwards through the fluidised bed dis-persed the ceramic fibres. Airflow was highenough to transport the ceramic fibres butnot the fluidising particles to the exposurechamber.

CONCENTRATION AND SIZE DISTRIBUTION INTHE EXPOSURE CHAMBEREach day during exposure the mass concen-tration of the ceramic fibres was measured

275

on Novem

ber 3, 2021 by guest. Protected by copyright.

http://oem.bm

j.com/

Occup E

nviron Med: first published as 10.1136/oem

.51.4.275 on 1 April 1994. D

ownloaded from

Yamato, Hon, Tanaka, Higashi, Monimoto, Kido

gravimetrically by isokinetic sampling ofchamber air through a glass filter. The aver-age concentration was 27-2 (SD 9-0) mg/M3.The size distribution of the aerosol in theexposure chamber was determined with anAndersen cascade impactor (Model AN-200,Sibata Sci Tech Ltd, Japan). The massmedian aerodynamic diameter and the geo-metric standard deviation (GSD) of theceramic fibres in the chamber were 3 7,umand 2-2 respectively. The geometric medianlength and the geometric median diameter ofairborne fibres in the chamber were 20-0(GSD 3 3) pm and 1-35 (GSD 1-8) pmrespectively.

ANIMALSTwenty male Wistar rats nine weeks old atthe start of the experiment were used. Theywere exposed to ceramic fibres daily for sixhours a day, five days a week, for two weeks.Groups were then killed at one day, onemonth, three months, and six months afterthe end of the exposure and the lungs weredivided into two parts. The two right lowerlobes were used for measurement of the num-ber, length, and diameter of the fibres. Theleft and remaining right lobes were used toestimate the silicon content of lung tissue bychemical analysis. One of the rats killed afterone day was used to confirm that fibres weredeposited in the alveolar region of the lung.Parts of the left lung of rats killed after sixmonths were prepared for histopathologicalexamination.

INHALED CERAMIC FIBRES AT THE ALVEOLARREGIONThe lung of one rat killed after one day wasfixed in 2-5% glutaraldehyde solution, dehy-drated in graded acetone/water mixture, andfreeze-dried. A section was put on to a scan-ning electron microscope stub, sputter coatedwith platinum, and examined for the presenceof fibres in the alveolar region with a scanningelectron microscope (model S-700, Hitachi,Japan).

DETERMINATION OF CERAMIC FIBRESRETAINED IN RAT LUNGCeramic fibres were extracted on to a 0-2 pmpore sized filter (Nuclepore Corp) from thetwo right lower lobes by low temperatureashing (LAT-25N Yanaco, Japan).'0 Thenumber, length, and diameter of fibres weredetermined with a scanning electron micro-scope according to the reference method formeasuring airborne man made mineral fibresproposed by the World Health Organisation(WHO). 13 The filters were dried, cut, and puton to scanning electron microscope stubs.They were sputter coated with platinum andexamined by scanning electron microscopy.Photomicrographs of the filters were takenrandomly at a fixed magnification ( x 2000).If any of the fibres had only one end withinthe field of view, a second photomicrographwas recorded at lower magnification ( x 1000)and centred on the original field so that thefibre length could be assessed. Only particles

with aspect ratios ) 3:1 were identified asfibres and these were counted and measuredon enlargements (25 x 28 cm). To determinethe number of ceramic fibres, we counted 1 0for fibres with two ends in the photographsand 0 5 for fibres with one end. The numbersretained in each rat were calculated on thebasis of the reference method of WH0.13The fibre sizes were measured on the photo-graphs with a backlight digitiser (KD3030LGraphtec, Japan). The size distributions ofinhaled ceramic fibres were calculated for allrats at each time period.

SURFACE STRUCTURE OF CERAMIC FIBRES ATHIGH MAGNIFICATIONTo observe the surface structure of ceramicfibres in rat lungs, some fibres recovered aftersix months were investigated at high magnifi-cation ( x 80 000) by scanning electron micro-scopy (model S-900, Hitachi, Japan) withoutsputtering with platinum. Fibres recoveredfrom the mixture of the original material andhomogenised lung tissue by low temperatureashing were compared.

SILICON CONTENT OF LUNGSThe lungs were freeze dried for 24 hours andashed at low temperature for 24 hours. Thesamples of ash were fused with sodium car-bonate in a platinum crucible and the siliconcontent measured by absorptiometry.14

ResultsDEPOSITION OF FIBRES AT THE ALVEOLARREGIONFigure 1 shows ceramic fibres in the alveolarregion one day after the termination of expo-sure. Ceramic fibres more than 50 pm inlength were seen in this region.

NUMBERS OF RESIDUAL CERAMIC FIBRES INRAT LUNGFigure 2 shows scanning electron micro-photographs of the ceramic fibres recovered

5 ALm _FD:: .' ..' ..Y.

Figure I Ceramicfibres in the alveolar region.

276

on Novem

ber 3, 2021 by guest. Protected by copyright.

http://oem.bm

j.com/

Occup E

nviron Med: first published as 10.1136/oem

.51.4.275 on 1 April 1994. D

ownloaded from

Retention and clearance of ceramic fibres

Figure 2 Ceramicfibresrecovered on tofilters atthree months (left) and sixmonths (right) afterexposure.

Figure 3 Changes innumber ofceramic fibresretained in rat lungs afterexposure. 0

r-

x

.0

-0

.Ei

um

0)C._

00z

10I* 0

00

*~~~0

n = 7.55 x 105 x e-01't|(t = month)

on to a nucleopore filter. The left pictureshows fibres recovered after three months andthe right one shows fibres recovered after sixmonths. The numbers of fibres in the lungswere determined from these photographs.Figure 3 shows the changes in the number ofceramic fibres retained in rat lungs afterexposure. The data can be fitted by asingle component exponential expression:

n = 7-55 x 1O x e-Ollt. * . * . * . *~~ . . * ,-

1 2 3 4 5 6 7Clearance time (months)

Table Number, diameter, and length of inhaled ceramicfibres recoveredfrom the ratlung

Clearance Number of Number ofCF GMD (GSD) GML (GSD)time rats (1O'/rat) (Pm) (um)1 day 3 8-7 (1-3) 0-65 (1 4) 9-0 (2-9)1 month 3 6-3 (1-1) 0-63 (1-5) 9-5 (3-2)3 months 7 5-2 (1.4) 0 56*(1 5) 8-6 (3-1)6 months 7 40 (07)* 0-50*(1-7) 7 1 (3 0)

*p < 0 01 v values of one day after the exposure.GMD = Geometric median diameter; GML = geometric median length; GSD = geometricstandard deviation; CF = ceramic fibres. The values of the number of ceramic fibres retained ineach rat were means (SD). Statistical comparisons of the change of fibre number, fibre length,and diameter among each clearance period were analysed by two sample t test. The data forlength and diameter were compared after log transformation.

(1)

where n = number of ceramic fibres retainedin the rat lung; t = clearance period (inmonths).The solid line in fig 3 shows the calculated

clearance rate of the fibre number from equa-tion (1). The decrease in the number ofceramic fibres retained in the rat lung at sixmonths was statistically significant comparedwith that at one day (p < 0 01; t test). Thetable summarises the result of the change innumber. In our previous study, there was adecrease in the number of fibres with timebut the change was not significant. A possibleexplanation for this difference was the larger

Figure 4 Distribution offibre diameter andfibrelength after one day (x),one month (0), threemonths ([), and sixmonths (0) clearancetime.

90-

-R0)

Ca

E3

70-

50-

30-

10-

- 90 -

U e

0 a m

0 0 oX

0 a K

0 a@x

70 -

50-

30-

10-

.

0od3

0 a0

O n.0 0co

o 03

0 OU

0ONA

0-5 0-7 1-0Fibre diameter (ptm)

. .

3 5 7 10Fibre length (hlm)

03 20 30

I .I . . . . . .

277

on Novem

ber 3, 2021 by guest. Protected by copyright.

http://oem.bm

j.com/

Occup E

nviron Med: first published as 10.1136/oem

.51.4.275 on 1 April 1994. D

ownloaded from

Yamato, Hon,. Tanaka, Higashi, Morimoto, Kido

Figure 5 Surfacestructure ofa controlfibre(left) and afibre recoveredfrom a rat lung at sixmonths (right).

number of rats used in the present study(three to seven rats instead of three to four).

DISTRIBUTION OF FIBRE DIAMETER ANDLENGTHThe size distributions of recovered fibres wereassessed at each time period. Both diameterand length could be fitted by log normal dis-tributions as shown in fig 4. The diameterdistribution shifted from right to left duringthe clearance period corresponding to areduction in diameters. The change in fibrelength was minimal during the first threemonths followed by a slight decrease after sixmonths. The geometric median length anddiameter of the fibres were assessed graphi-cally. The table summarises the results.

SURFACE STRUCTURE OF CERAMIC FIBRES ATHIGH MAGNIFICATIONFigure 5 shows the change in surface struc-ture of ceramic fibres taken at high magnifica-tion ( x 80 000). Fig 5 (left) shows a controlfibre treated by low temperature ashing. Itssurface is smooth and the low temperatureashing treatment had no apparent effect onsurface structure. Fig 5 (right) shows the sur-face of a fibre recovered from rat lung at sixmonths. This has an eroded appearance thatcould not be detected at low magnification.

Figure 6 Mass ofretained ceramicfibresestimatedfrom the

measurement of the siliconcontent of rat lungs.

L- 10

cm

a) 8-L-

a,,.0

(A

4-6-

IU

c -4-

° 2-nwn

E O

Clearance time (months)

This change in surface characteristics is con-sidered to be due to dissolution.

SILICON CONTENT IN RAT LUNGFigure 6 shows the mass of retained ceramicfibres estimated from the measurement of thesilicon content of rat lungs. The values werecorrected for the amount of silicon found inthe lungs of control rats and the percentage ofSiO2 (54%) in the ceramic fibre.

HISTOPATHOLOGICAL FINDINGSAfter six months, parts of the lungs ofexposed and control rats were prepared forhistopathological examination. No granulo-matous response, fibrotic changes, or carci-noma were found either in exposed or controlrats. To make a detailed investigation of thepathogenic response to the inhaled ceramicfibres would have required a longer exposureperiod and an extended study.

DiscussionCHANGES IN FIBRE NUMBERS AND LENGTHThe numbers of ceramic fibres retained in thelung decreased exponentially (fig 3). Adecrease in numbers of fibres after intra-tracheal injection'-3 and inhalation' has beenreported previously. The studies showed thatshort mineral fibres (less than 10 ,um inlength) are cleared from the lung rapidly dur-ing the first few months and that the remain-der were cleared much more slowly. Thegeometric median length of ceramic fibreused in the present study was 9 0 pm at oneday after exposure, which is in the range ofshort fibres used by Morgan et al (5 and 10pm)I and by Le Bouffant et al.I Our studyshowed that the number of ceramic fibrespresent in the lung initially decreased to 46%during the next six months. This result wasalso in good agreement with that obtained byMorgan et al l and by Le Bouffant et al 5 andis assumed to be the result of macrophage

278

on Novem

ber 3, 2021 by guest. Protected by copyright.

http://oem.bm

j.com/

Occup E

nviron Med: first published as 10.1136/oem

.51.4.275 on 1 April 1994. D

ownloaded from

Retention and clearance of ceramic fibres27

nAA-

1

D = 0*65 0-025 t, r2 = 0978

(t = month)

i

0 1 2 3 4 5 6Clearance time (months)

mediated clearance. If short fibres are cleared

preferentially, then the geometric median

length should increase slightly as reported by

Holmes et al.2 In our study, however, it

seemed that the geometric median length did

not change or even decreased slightly. It is

not clear whether this was due to the frag-

mentation of the fibre or not. Further experi-

ments with longer fibres would be required to

investigate this point.

CHANGE OF DIAMETER AND THEORETICAL

MODEL OF FIBRE DISSOLUTION

As shown in fig 4, the diameter of recovered

ceramic fibres decreased with time due to

erosion of the fibre surface. The following

theoretical model of fibre dissolution is pro-

posed on the assumption of the following

conditions: (1) The fibre is a long and thin

column; (2) the fibre length is constant. This

experiment proved this point; (3) the dissolu-

tion of fibres occurs at their surface and the

rate of fibre dissolution is proportional to the

surface area. The decrease of volume is

described by

4n 2ird =k(rDL+ 27

~k~rDL (2)

where D = fibre diameter (am); L = fibre

length (jim); k = constant of rate of dissolu-

tion (jim/month).

The equation can be rearranged to:

= =(3)

dt

On integration, equation (3) becomes:

D=D-2kt=D-Kt (4)

where Do = the initial diameter of fibre (pm);

K = constant of dissolution rate (pm/month).

Equation (4) shows that the fibre diameter

decreased linearly with time after exposure.

Figure 7 shows the relation between the clear-

ance period and the temporal change in the

geometric median diameters. The solid line in

this figure shows the regression line based on

the equation (4). The decreasing rate of geo-

metric median diameter can be expressed as:

D=O065 -O0025 t (5)

The regression line obtained by the equation

(4) and the geometric median diameter were

in good agreement (r2 = 0-978) which meansthat the assumptions are reasonable. Thislinear decrease of geometric median diameterwas also found in our previous study.'10Morgan et a!'I injected sized glass fibres in

rats and measured the dimensions of fibresextracted from the lung. They showed thatthere was a reduction in the diameter and anincrease of the GSD of fibres recovered fromthe rat lung, and that the effect was morepronounced with long fibres. We plotted thegeometric median diameters shown in theirtable, and found that they decreased almostlinearly with the duration of the clearanceperiod. Holmes et al!2 gave sized glass fibresto hamsters by intratracheal injection andshowed that they dissolved more rapidly inthe hamster lung than in the rat lung. Theyalso showed an almost linear decrease of thegeometric median diameter in their figure andreported an increase in the GSD. Althoughthere might be a difference in physiologicalpH between the animal species (rat and ham-ster) or between the inside and outside ofpulmonary cells in the lung, the lineardecrease in the geometric median diametercould be explained by the equations whichare shown in this paper. As for the increase inthe geometric standard deviation, it can bealso explained by the dissolution of fibres. Insuch cases, the relative decrease in the diame-ter of thinner fibres is greater than that ofthicker ones. Leineweber6 has studied the sol-ubility of man made mineral fibres (four typesof glass fibres, refractory fibre, and mineralwool) in vitro using water and physiologicalsaline. He found that the square root of theweight of fibre decreased linearly with thetime of leaching. The equation developed byLeinweber and our equation (4) are essen-tially identical.

This phenomenon is important in anotherway. According to the Stanton-Pott hypo-thesis,"'16 the fibres longer than 8 ,um andthner than 0-25 gm have high carcinogenic

potential. Even although the original diame-ters of inhaled mineral fibres are thicker thanStanton fibres, they could assume the dimen-sions of Stanton fibres by dissolution in theprocess of time.The differential dissolution rates between

short and long fibres showed by Morgan andHolmes,7 was not confirmed in this studybecause we used only one type of ceramicfibre. It is necessary to give both long andshort ceramic fibres to define whether thereare differential dissolution rates or not.

RESIDUAL MASS OF FIBRE CALCUL-ATED FROMLENGTH AND DIAMETERBecause the numbers and dimensions ofretained fibres were known for each rat, themass of residual ceramic fibres could be cal-culated from fibre density (p = 2 -5 g/cm3).T'he results are shown in fig 8, from which itcan be seen that there was a differencebetween the values of fibre mass obtainedin this way and by measuring the silicon con-tent in the lung (see fig 6). One explanationfor this discrepancy might be that the mass

Figure 7 Decrease ofgeometric mean diameter ofceramic fibres withclearance time. a)

ECOCa

EC.2

E0Co V,14 -

279

on Novem

ber 3, 2021 by guest. Protected by copyright.

http://oem.bm

j.com/

Occup E

nviron Med: first published as 10.1136/oem

.51.4.275 on 1 April 1994. D

ownloaded from

Yamato, Hon, Tanaka, Higashi, Morimoto, Kido

100*Ca

0,

10.E

0

Snco

0

0 1 2 3 4 5 6 7

Clearance time (months)

calculated from the fibre volume was assessedonly for the fibres that had aspect ratiosgreater than 3:1, whereas the fibre massbased on silicon analysis applied to all fibresand ceramic fragments.The decreasing mass of fibres due to their

elimination and dissolution can be expressed:

% of original mass =

nI D2

M -n=1 x n x 100 (6)Mo nO no

Y. D 0n=1

when equations (1) and (5) are substitutedfor equation (6),

% of original mass =

nY (Do -0-025 t)2n=1 xe-Ollt x 100 (7)

nOD2

n=l

The mass of retained fibre can be calculatedfrom equation (7) and was 82-5% after one

month, 55-7% after three months, and 30 0%after six months. Taking the initial mass offibre as 20-2 jug/rat, (the average value at one

day), the calculated rate is shown in fig 8 as a

solid line. The experimental data and calcu-lated line are in good agreement.

CHANGE OF SURFACE STRUCTUREAs shown in fig 5, the ceramic fibres had an

eroded appearance. Le Bouffant et al5 alsoshowed that fibres that had remained for sev-

eral months in the rat lung had an erodedappearance whereas the original fibres had a

smooth surface. LeineweberP showed thatsolvent attack on fibres in vitro was accom-

panied by the build up of a crust or shell anddemonstrated that the degree of surfacechange depended on the fibre's solubility. Inthis study, we used a ceramic fibre that isconsidered to be only slightly soluble in vivo(unpublished data). No surface change wasseen at lower magnifications, but at highermagnifications the eroded appearance of thefibre surface was evident. It was concludedthat the examination of fibres that are resis-

tant to dissolution must be performed at highmagnification. Also, the observation of resis-tant fibres at high magnifications must beperformed without sputter coating becausethe particles of sputtering material interferewith the fine surface structure.

Conclusions(1) The number of initially retained

ceramic fibres decreased exponentially accord-ing to the duration of the clearance period.

(2) The geometric median diameter de-creased linearly with time.

(3) The surface of fibres recovered after sixmonths had an eroded appearance.

(4) The clearance of inhaled ceramic fibre(geometric median length 9-0 pum, geometricmedian diameter 065,um) was mediatedthrough dissolution and macrophage medi-ated clearance.

(5) The clearance rate of inhaled fibre canbe calculated by measuring the number,dimension, and dissolution rate of fibres inthe lung.

We thank Ms S Ishimatsu and Ms T Oyabu for their excellenttechnical help.

1 Morgan A, Holmes A, Davison W. Clearance of sizedglass fibres from the rat lung and their solubility in vivo.Ann OccupHyg 1982;25:317-31.

2 Holmes A, Morgan A, Davison W. Formation of pseudo-asbestos bodies on sized glass fibres in the hamster lung.Ann Occup Hyg 1983;27:301-13.

3 Bellmann B, Muhle H, Pott F, Konig H, Kloppel H,Spurny K. Persistence of man-made mineral fibres(MMMF) and asbestos in rat lungs. Ann Occup Hyg1987;31:693-709.

4 Tanaka I, Akiyama T. Pulmonary deposition fraction of aglass fiber in rats by inhalation. In: Masuda S,Takahashi K, eds. Aerosols-science, industry, health andenvironment. Oxford: Pergamon Press, 1990;1242-5.

5 Le Bouffant L, Daniel H, Henin JP, Martin JC, NormandC, Tichoux G, Trolard F. Experimental study on long-term effects of inhaled MMMvF on the lungs of rats.Ann Occup Hyg 1987;31:765-90.

6 Leineweber JP. Solubility of fibres in vitro and in vivo. In:Proceedings ofWorld Health Organisation/IntemationalAgency for Research on Cancer conference. Biologicaleffects of man-made mineral fibres. Copenhagen: WHO1984;87-101.

7 Morgan A, Holmes A. Solubility of asbestos and man-made mineral fibers in vitro and in vivo: its significancein lung disease. Environ Res 1986;39:475-84.

8 Tanaka I, Akiyama T. A new dust generator for inhalationtoxicity studies. Ann Occup Hyg 1984;28: 157-62.

9 Tanaka I, Akiyama T. Fibrous particles generator forinhalationtoxicity studies. Ann Occup Hyg 1987;31:401-3.

10 Yamato H, Tanaka I, Higashi T, Kido M. Determinantfactor for clearance of ceramic fibres from rat lungs.BrJ Ind Med 1992;49: 182-5.

11 Yamato H, Hori H, Tanaka I, Higashi T, Kido M.Clearance of inhaled ceramic fibers and its dissolutionmodel. In: Gibbs WG, Kido M, Dunnigan J, Higashi T,eds. Health risks from exposure to mineral fibers: an inter-national perspective. Captus University Publications,1993;273-80.

12 Yamato H, Tanaka I, Higashi T, Kido M. Clearance ofinhaled ceramic fibers from rat lungs. Environ HealthPerspect 1994 (in press).

13 World Health Organization. Reference methods for measur-ing airborne man-made mineral fibres (MMMF).Copenhagen: WHO, Regional Office for Europe.1985:36-54. (Environmental health report No 4.)

14 King EJ, Stacy BD. The colorimetrical determination ofsilicon in the micro-analysis of biological material andmineral dusts. Analyst 1955;80:441-51.

15 Stanton MF, Layard M, Tegeris A, Miller E, May M,Kent E. Carcinogenicity of fibrous glass: Pleuralresponse in the rat in relation to fiber dimension. J NatlCancer Inst 1977;58:587-603.

16 Pott F, Huth F, Friedrichs KH. Tumorigenic effect offibrous dusts in experimental animals. Environ HealthPerspect 1974;9:313-5.

Figure 8 Mass ofretained ceramic fibresestimatedfrom fibredensity.

. I I . I I I . I I .

280

on Novem

ber 3, 2021 by guest. Protected by copyright.

http://oem.bm

j.com/

Occup E

nviron Med: first published as 10.1136/oem

.51.4.275 on 1 April 1994. D

ownloaded from