Document 2JZBQ8jQzNw2Ve9EyE0rbQD9r

Jitology of [with skin 3\ ran. .. Phila- 5J. (1985) oppression : reserpine a indepenexp. Med. tHLABRB M. studies of ton menta' T cells. J. HAM D.A. orients of tology 89, micK C.H. lination of teous Cans. J. Immu- iofderma- r ^ to ) 'j }rm. Br. }. exp. Path. (1988) 69, 717-737 Cl Bu $(i PB Pk* ' \ ' " >' '} 3': r. : j;-V 1 JomparisoilS t>f the'pithogenicity of long and short fibres of chrysotile asbestos in rats , J.M.G. Davis* and A.D. Jones * Pathology Branch and f Physics Branch, Institute of Occupational Medicine. 8 Roxburgh Place, Edinburgh EH8 9SU, UK Received for publication 12 February 1988 Accepted for publication 10 May 1988 Summary. Long-term inhalation and intraperitonea] injection studies were undertaken with laboratory rats treated with a specially prepared short-fibre sample of Canadian chrysotile asbestos. This was compared, at an equal mass dose, to dust generated from the same chrysotile batch so as to contain the highest possible number oflong fibres. The long-fibre cloud contained roughly five times more fibres > 5/mi in length as seen by phase contrast optical microscopy (PCOM). For increasing lengths, the ratio between the dust clouds increased progressively, reaching over 80:1 for fibres > 30/im in length. Rats treated with long-fibre chrysotile developed six times more advanced interstitial fibrosis (asbestosls) than animals treated with short-fibre chrysotile and three times more pulmonary tumours. At the end ofthe 12-month dusting period, three times more short chrysotile than long had been retained in the rat lung tissues. During the following 6 months, however, the short-fibre chrysotile was removed from the lungs much more rapidly than the long. Following intraperitoneal injection at a mass dose of 25mg of dust, both long and short chrysotile produced mesotheliomas in more than 90% of rats. At a dose level of 2.5mg of dust, the short-fibre chrysotile produced mesotheliomas in only one-third as many rats as the long-fibre dust which still produced mesotheliomas in more than 90% of animals injected. At a dose level of o.25mg of dust, the short-fibre chrysotile produced no mesotheliomas while the long-fibre chrysotile still produced these tumours in 66% of rats. In the two highest doses, where short-fibre chrysotile produced mesotheliomas, the mean tumour induction period was significantly longer than for tumours produced by long chrysotile. Keywords: asbestos, dimensions, carcinogenicity Since the dangers of industrial exposure to asbestos were first realized much experimen tal work has been undertaken to determine ^hich parameters ofany dust cloud are most iportant in disease production. One of the earliest factors to be examined was fibre length and King et al (1946) administered chrysotile samples cut on a special micro tome to rabbits by intratracheal injection. Animals injected with dust of fibre length approximately either 15 or 2.5 pm developed nodular peribronchial fibrosis with some Correspondence: Dr J.M.G. Davis, Pathology Branch, Institute of Occupational Medicine, 8 Roxburgh Place, Edinburgh EH8 9SU, UK. 717 HWBUI0009179 718 J.M.G. Davis and A.D, Jones progressive fibrosis of the alveolar walls, but these changes were much more marked in the animals receiving the long-fibre dust. Similar intratracheal injection studies, using guinea-pigs, were reported by Vorwald et al (1951). This group injected `long' and `short' samples of the main asbestos varieties, chry solite, amosite and crocidoltte as well as anthophyllite, tremolite and brucite. Only the short-fibre samples of chrysotile and amosite would now be considered really short with almost all fibres < 3/rm in length and while these produced little tissue reac tion, long-fibre preparations produced dis tinct pulmonary fibrosis. Klosterkotter (1968) extended these studies by using both the intratracheal and intraperitoneal injec tion of chrysotile and crocidoltte ground to an average fibre length of <5(mi. They found that these samples produced little or no fibrosis in either site. In contrast, longer fibres of the same asbestos types resulted in considerable fibrosis in both areas. Almost identical results were obtained by Hilscher et al. (1970) using very similar techniques. Davis (1972) administered many finely ground mineral samples to mice by intraperi toneal injection. Included in this series of dusts were a normal long-fibre chrysotile sample and two short-fibre chrysotile prep arations. These were synthetic chrysotile with a maximum length of ijon and chiysotile fragmented by ultrasonic treatment in liquid until all fibres were below 1 (im in length. The short-fibre samples produced almost no tissue reaction while the long-fibre chrysotile produced massive peritoneal fibro sis. Wright & Kuschner (1977) injected samples of crocidoltte, synthetic fluoramphibole and glass fibre intratracheally into guinea-pigs. For each mineral there was a long-fibre preparation with up to 80% of fibres >TOjim in length and a short-fibre sample with few fibres > Sfun in length. Ail the long-fibre samples produced widespread pulmonary fibrosis while the short-fibre materials produced little tissue reaction. Stanton & Wrench (1972) demonstrated that, in addition to its effect on fibrosis, fibre length was important in carcinogenesis with the production of mesotheliomas. Carefulb sized samples of a number of mineral fibre! ) types were administered to rats by pleural implantation and it was found that mesothe lioma production was related to the number of long, thin fibres implanted. Also in 1972, Smith et al. reported that hamsters given intrapleural injections of crocidoltte ground to a fibre length < ifim did not develop mesotheliomas while those injected with a long-fibre preparation of the same material did. Stanton's work was further developed (Stanton et al. 1977, 1981) and it was determined that the most dangerous fibres were those > 8(im in length and < l-5/nn in diameter. Similar findings were reported by Pott & Friedrichs (1972) and Pott et al. (1976) and the work of this group has been continued until the present time with many further publications (e.g. Pott (1978,1983), Pott & Ziem (1983)). Fibre length also appears to be important if dusts are inhaled and Vorwald et al. (1951) reported that guinea-pigs, rats and micr treated for up to two years with ball-millei J chrysotile developed no appreciable pulmon- " ary reaction and no suggestion of asbestosis. Complete fibre sizing of the dust clouds was not undertaken but it was reported that while 90% of the dust was non-fibrous, onethird of the fibres present were > io;um in length. These amounted to a concentration of approximately 25 fibres/ml. The importance of long fibres in the pro duction of pulmonary fibrosis following inhalation was further emphasized by Timbrell & Skidmore (1968) and Webster (1970). The former authors exposed rats and guinea-pigs to long and short fibres of amo site and obtained a much greater reaction with the long-fibre sample. Webster treated baboons with a finely ground crocidolite dust with a fibre length below 5fan and obtained only a macrophage reaction in the lungs. Until recently inhalation studies with short-fibre samples of asbestos had not been continued for sufficient time to explore tb production of pulmonary tumours and 1 j HWBUI0009180 is with ''Tefuily '-J "bre -oral nesothenumber n 1972, rs given : ground develop i with a material sveloped . it was us fibres ..5pm in oiled by tt et al. las been th many 1983). wrtantif ' (1951) ad mice *t :bestosis. tads was ted that >us, onero/rm in satration the proollowing by TimWebster .rats and ; of amoreaction r treated olitedust obtained lungs, ies with not been plore the j r Pathogenicity of long and short fibres of asbestos in rats 719 some cases the dust clouds had been insuffi ciently categorized regarding the numbers of fibres in any size range for definite conclu sions to be drawn. Davis et al. (1986b) reported life-time inhalation studies in rats undertaken with a specially prepared shortfibre sample of amosite with almost all fibres < 5pm in length. The effects of this dust were compared to those of a long-fibre dust cloud generated from raw commercial amosite. While the long-fibre dust produced wide spread pulmonary fibrosis and pulmonary tumours or mesotheliomas developed in onethird of the animals, neither pulmonary tumours nor pulmonary fibrosis developed in animals exposed to the short-fibre dust. Long-term inhalation studies with a shortfibre chrysotile preparation produced by ballmilling were reported by Platek et al. (1985). These workers examined rats for up to 24 months after the start of dust exposure and monkeys for up to 28 months and reported that the short-fibre chrysotile produced `no compound related lesions including fibrosis'. The dose level used in these studies was, however, only img/m3, which is lower than used by most workers examining fibre patho genicity, and it has also been suggested by Lancer et al. (1978) that short-fibre chryso tile produced by ball-milling is subject to a level of crystal damage. This is sufficient to make results difficult to interpret in relation to hazards resulting from short fibres pro duced during the manufacture of asbestos products or during the subsequent usage of these materials. Jolicoeur et al. (1981) reported on a method of obtaining short-fibre samples of chrysotile by a sedimentation technique which has overcome these difficulties. This preparation contained no fibres >8jum in length and has been used in a number of short-term studies to. examine the early Pathogenicity of the short chrysotile fibres. This short-fibre chrysotile preparation retained the haemolytic activity normally associated with chrysotile asbestos (Pele & Calvert 1983) and was toxic to pulmonary macrophages in short-term in vitro culture (Nadeau et al. 1986). Lemaire (1985) exam ined cell recruitment into rat lungs following intratracheal injection of the short-fibre chrysotile preparation and UICC chrysotile (B). While the UICC material produced a prolonged response with increased numbers of neutrophils in lavage fluid for up to 14 days following injection, the short-fibre chry sotile produce only an insignificant.increase in neutrophils for 1 day following injection. Both chrysotile samples produced increased numbers of pulmonary macrophages which continued for up to 14 days, but the shortfibre dust produced a much lower level of response. In a separate publication Lemaire et al. (1985) reported on changes in pulmon ary morphology produced in rats following the intratracheal injection ofthe two chryso tile samples. While the UICC dust produced fibrosis around distorted and obstructed air ways, the short-fibre dust, although causing an accummulation of inflammatory cells, produced no fibrosis at all. While these studies have suggested that the harmful potential ofvery short chrysotile is much less than that of chrysotile dust containing long fibres, intratracheal injec tion is an abnormal technique which bypasses the normal pulmonary clearance mechanisms. In order to investigate the fibrogenic and carcinogenic potential of any fibre sample in normal situations, inhalation studies are necessary. This paper reports long-term inhalation and injection studies in rats treated either with dust generated from a specially pre pared short-fibre sample of chrysotile or dust from the same type of raw chrysotile gener ated in such a way as to contain a large proportion of long fibres. Materials and methods The short-fibre chrysotile sample used in these studies was supplied by the Institute for Research and Development of Asbestos (IRDA) and was prepared by the differential sedimentation techniques reported by Joli coeur et al. (1981). Chemical, X-ray difirac- HWBUI0009181 720 J.M.G. Davis and A.D. Jones tion and differential thermal analyses have confirmed that the chrysotile structure was substantially unchanged by the sedimen tation process. The long-fibre dust cloud was generated directly from the same batch of chrysotile (grade 4-T-30) that had been used to prepare the short-fibre sample. The dust clouds were generated using a Timbrell fibrous dust dispenser (Timbrel] et al. 1970) modified as described by Beckett (1975), and a fluidized bed aerosol (TSI 3400) for respectively the original long-fibre and short-fibre chrysotile respectively. The fluidized bed aerosol generator was needed to break up particle aggregates in the shortfibre material and to produce a cloud consist ing mainly of separate fibres. The electro static charge carried on the airborne dust was reduced by exposure to ionizing radia tion from /(-sources. The purpose of reducing the charge was to avoid differences in charge level which might have arisen from using different types of dust dispenser (Johnston et al. 1987). The level of electrostatic charge can substantially enhance the deposition of chrysotile fibres (Jones et al. 1983). The respirable dust mass concentrations were measured daily using the Caseila MRE 113A with samples taken for the full 7 h of dust exposure and on each of the 224 days of dusting. The total dust concentrations were measured using open-face 5omm-diameter filter holders facing downwards with yhsamples collected daily on ten days at a flow rate of 21/mihute. The fibre number concen trations were assessed from membrane filter snatch (or very short period) samples col lected using 25mm-diameter Gelman heads (open-faced filters) on 85 separate days. The samples were examined using phase contrast microscopy (at 600 x magnification) and counts made of fibres with length >5/n, aspect ratio greater than 3:1 and diameter <3/tm. The counting procedures are essen tially the same as those used for assessing samples from occupational environments (AIA 1979, HSE 1984). The fibre size distributions were assessed also from snatch samples collected on mem brane filters and nuclepore filters for optical and scanning electron microscopy respect ively. Phase contrast microscopy was used to produce length distributions for the fibres which constituted the number concentra tion, i.e. those fibres longer than 5/im, etc. The scanning electron microscope (SBM) was used to produce size distributions for fibres longer than 0.4/em, with measure ments taken from the video screen at 10000 x magnification. The minimum length corresponds to a detection limit of 0.15/un for the diameters of fibres with an aspect ratio of approximately 3:1. Samples of long and short-fibre chrysotile used in the intraperitoneal injection studies were collected directly from airborne mater ial in the inhalation chambers by an elutriation process chosen to select the respirable fraction ofthe dust cloud (Bolton etal. 1982). For counting and sizing purposes, weighed amounts of dust were resuspended in known volumes of 35% ethanol in distilled water. These suspensions were treated with a brief period of ultrasonication to disperse fibre clumps before filtering on nuclepore filters. The filters were air dried under controlled conditions before being coated with gold for examination by scanning electron micros copy (SBM) at a magnification of 10 000 x. In the inhalation studies two groups of 48 rats of the AF/HAN strain were exposed to clouds of either long or short-fibre chrysotile at a dose level of iomg/m3 of respirable dust for 5 days each week during a period of I year. The animals were aged 3 months at the start of the study. Following the cessation of dusting, four animals from each group were killed to examine early histological changes in the lung tissue and to determine the pulmonary content of retained chrysotile. Four more animals from each group were killed 6 months after the end of dusting but the remainder were allowed to live out their full life span until the experiment was termi nated when the number of survivors in one group dropped to six. This point was reached at 31 months after the start of dusting when the surviving rats were aged 34 months. For the u rats w injection or dose levels fion the ar. ether and tl pulbecco's animals in t to live out a, were killed Forty-eigl groups were unit for theL animals wei to the exper were maints rooms while studies wer Animals we) of ill-health when the reached an The auto] inhalation i ' included ^ major mality, and the visceral Any abnori histological sentative me , the followinj ation whetb dent or not lymph nodes gastro-intest nodes, adrer sueforhistol 10% formal; wax. The Iv prior to exd Sections of a with H&E. V Gordon and i for the stud; sections wej animals being me vals througt HWBUI0009182 I optical ^spect,._y- ' to res icentraum, etc. ; (SBM) ions for leasureceen at inimum limit of with an lrysotile i studies e materi elutriajspirable 1.1982). weighed a known d water. h a brief rse fibre 'IS. ) Jed , gold for micros- 3600x. jps of 48 :posed to biysotile able dust riod of I ths at the nation of Dup were changes nine the hrysotile. jup were sting but out their fas termi ng in one sreached ing when ,r Pathogenicity of long and short fibres of asbestos in rats 721 For the injection studies six groups of 24 rats were given a single intraperitoneal injection of either long or short chrysotile at dose levels of 25, 2.5 or o.25mg. For injec tion the animals were anaesthetized with ether and the dust was suspended in iml of Dulbecco's phosphate buffered saline. All animals in the injection studies were allowed to live out almost all of their full life span and were killed when moribund. Forty-eight rats of the same age as the test groups were maintained in the experimental unit for their full life span as controls. Control animals were kept under identical conditions to the experimental groups except that they were maintained in normal animal holding rooms while animals used in the inhalation studies were in the inhalation chambers. Animals were killed when they showed signs of ill-health until the study was terminated when the seven surviving controls had reached an age of 3 years. The autopsy schedule for animals in the 4 inhalation studies and the control groups included the macroscopic examination of all major organ systems for signs of abnor mality, and particular attention was paid to the visceral and parietal pleural surfaces. ? Any abnormalities found were taken for histological examination. In addition repre sentative material was routinely taken from ; the following organs for histological examination whether pathological change was evi dent or not: lungs, mediastinal and hilar lymph nodes, liver, spleen, pancreas, kidney, gastro-intestinal tract, mesenteric lymph nodes, adrenal glands, testis and brain. Tis; sue for histological examination was fixed in 10% formal saline and embedded in paraffin Wax. The lungs were fixed by instillation prior to excision from the thoracic cavity. Sections of all tissues were routinely stained with H&E. Van Giesen's collagen stain and Gordon and Sweet's reticulin stain were used for the study of pulmonary fibrosis. Serial sections were cut from the lungs of all ' animals autopsied with groups of sections ; being mounted at approximately imm-inter- v vals throughout the block. This procedure provided material for detailed examination from six to eight levels of each lung. Measurement of pulmonary fibrosis was undertaken by similar methods to those previously described by Davis et al. (1978) except that an electronic image analyser (Graphic Information Systems Limited, GDSi) was available for use in conjunction with the light microscope (Davis et al. 1985). Single lung sections were examined and the sections selected to contain the maximum area of lung parenchyma. As previously described, interstitial fibrosis was estimated using a x 2 microscope objective lens and expressed as a percentage oftotal lung tissue. Peribronchiolar lesions are more numerous and smaller and so the lung tissue was scanned with an eyepiece graticule covering the tissue area of 2.92mm2 and divided into 100 squares. A x 4 objective lens was used. Peribronchiolar lesions were recorded as the percentage of squares containing lesions of this type. The amount of chrysotile retained in the left lung was assayed for rats killed at the end of dust exposure and 6 months later. The right lung from these animals was prepared for histology. The lung burdens of chrysotile were recovered by ashing the lungs in oxy gen plasma in a low-temperature asher (Nanotech P100), washing the residue with distilled water and recovering on a filter. The residue material was formed into a potas sium bromide disc and the chrysotile content determined by infrared spectrophotometry (Dodgson & Whittaker 1973). Comparisons of the dust contents of left and right lungs have indicated that the ratio of the contents is 0.6:1 and this ratio was used to estimate the total lung burdens. For statistical analysis the survival func tions for each experimental group were estimated using the product-limit method (Kaplan & Meier 1958). The survival curves for the different series were tested for homo geneity using the Geralized Wiicoxon (Breslow) statistic in the statistical package BMDP (Dixon et al. 1983). In order to compare the overall mortality of the two groups of ani- HWBUI0009183 J.M.G. Davis and A.D, Jones ^ 722 mals, the sacrifice of an animal as part of a planned kill was treated as a censoring event, all other deaths were treated as responses. Incidence of tumours in particular sites was also compared by estimating survival func tions. All deaths of animals without tumours in the site in question were recorded as censoring events, whilst tumours in the specific site were treated as responses irres pective of whether the death was planned or unplanned. From the intraperitoneal injec tion studies, the number oftumours recorded as well as the survival times of the animals were combined using Cox's method to esti mate the relative hazard of the different asbestos types at all dose levels (Cox 1972, Kalbfleisch & Prentice 1980), Comparisons of levels of pulmonary fibro sis and lung dust burdens were made using the generalized linear models facilities in the statistical package GENSTAT (Alvey et al. 1977). The proportion of lung area with interstitial fibrosis was logarithmically trans formed before analysis using conventional analysis of variance methods. The propor tion of the points on the lung with peribron chial fibrosis was analysed by logistic regres sion techniques. Lung dust burdens were compared by analysis of variance. Results The target mean respirable dust concentra tions of iomg/tn3 were achieved. The total dust mass concentrations and fibre number concentrations corresponding to the respir able dust concentration of iomg/rn3 are shown in Table 1. PCOM examination of samples from the two dust clouds confirmed that the `longfibre' dust cloud did contain many more fibres of all sizes above 5/an than the `shortfibre' cloud. However, the short-fibre mater ial used had obviously contained more long fibres than the original preparations reported on by Jolicoeur et al. (1981) when it was suggested that the sedimentation technique produced chrysotile samples with all fibres <Sfim in length. In the present study the ratio of fibre number concentration between the long and short-fibre dust clouds was 5:1 for all fibres longer than 5/an rising to 20:1 for fibres longer than 20/im and 80:1 for fibres longer than 30pm. The fibre-length distributions obtained by PCOM for the long and short-fibre chrysotile clouds are illus trated in Fig. 1 together with recently obtained figures for UICC chrysotile 'A' which are included for comparison. The fibre-length and diameter distributions obtained by SEM for the three chrysotile varieties are illustrated in Figs 2 and 3, These sizings show that the fibre-length distribu tions of the long-fibre chrysotile cloud gener ated from bulk 4-T-30 material and used in the present study were very close to dust ; 99.95 99.5 95 95 9! 9( .c Ui 8C _QC> CTO 7< SL 0> 6( re CD 5( o> <D 4( 0U3) C OCD 0 CL 1C t Table 1. Airborne dust concentrations Type of measurement Short-fibre Long-fibre UICC chrysotile chrysotile chrysotile Mass of respirable dust mg/m3 (by Casella MRE 113A) Mass of total dust mg/m3 (by downward-facing open somm-filter sampling at 2 1/min) Fibre number obtained by PCOM fibres/ml for fibres longer (pm) than 5 10 20 30 xo.o 13-9 1170 330 33 4 xo-o ir.81 9`9 II.O 55io 1930 670 320 2560 0. 0.0 Fig. 1. Fibre i 'A*. Fibres si parentheses; ) HWBUI0009184 he total ''Number pirare from the he 'longny more le `shortre materlore long i reported ;n it was echnique all fibres study the i between was 5: i I to 20: l 8o: i for re-length : the long are illus- recently jotile `A' The rk*i,,.dons shrysotile 13. These distribu- ud geherid used in e to dust Pathogenicity of long and short fibres of asbestos in rats 723 UICC chrysotile 9`9 11.0 2560 ^ Rg. I. Fibre length distributions of long and short-fibre chrysotile dust clouds as well as UICC chrysotile A'. Fibres sized by phase contrast optical microscopy at a magnification of x 600, Number sized in parentheses; , short fibre (2250); 0, long fibre {2245); a, UICCA (4476). i HWBUI0009185 Fig. 3. Fibre wellasUICC Number sizx HWBUI0009186 D Pathogenicity of long and short fibres of asbestos in rats 725 uds as well er sized in ; Pig. 3. Fibre diameter distributions of long and short-fibre (longer than 0.4/rm) chrysotile dust clouds as well as UICC chrysotile `A'. Fibres sized by scanning electron microscopy at a magnification of x 10 000. Number sized in parentheses: , short fibre (500): 0, long fibre (600): a, UICCA (600). HWBUI0009187 726 J.M.G. Davis and A.D. Jones clouds recently generated in our laboratory from UICC chrysotile (Davis et al. 1988). The fibre diameters of the long-fibre mater ial were, however, somewhat lower than those of the UICC chrysotile. Probably because of this diameter difference the number offibres > 5pm in length counted by PCOM was higher for the long-fibre dust clouds than for clouds of UICC chrysotile of the same respirable dust mass. The survival of animals treated with both long and short-fibre chrysotile and animals in the control group was extremely good with the majority reaching an age of more than 800 days. There were no significant differences between the survival of the two test groups treated with the long and short chrysotile preparations but the control ani mals did survive longer on average (P=0.002). Both groups of rats dusted with the chry sotile preparations developed the same pat tern of pathological change previously reported in similar studies from this Institute (Davis et al 1978, 1985, 19860-0). At the end of the 12-month dusting period the main lesions present were deposits of granulation tissue around the terminal and respiratory bronchioles (Fig. 4). This granulation tissue consisted mainly of macrophages and fibro blasts but foreign-body giant cells were also present. At 12 months after the start of dusting there was marked reticulin staining in the peribronchiolar deposits although relatively little collagen could be demon strated by Van Giesen's stain. With increas ing time after dust exposure, however, colla gen staining became progressively more marked and in old animals the lesions con sisted of mainly acellular fibrous tissue. Both Time after of expos (months Number ol examine PeribroncI fibrosis Interstitial fibrosis The figu Fig. 4. An area of peribronchiolar fibrosis from the lungs of a rat after 12 months exposure to a dust cloud of 'long' chrysotile. x 350. Fig. 5. An exposi ._y] appears To HWBUI0009188 ;)y asuiute At the le main ulation dratory 1 tissue d fibro;re also start of taining though lemonncreas- colla' more as cone. Both 1 i Pathogenicity of long and short fibres of asbestos in rats Table 2. Levels of pulmonary fibrosis produced by long and short-fibre chrysotile dusts 727 Chrysotile dust Long chrysotile Short chrysotile Control group Time after start of exposure (months) Number of rats examined Peribronchiolar fibrosis Interstitial fibrosis 12 4 14.0 (9.4-18.2) 0 18 4 8.9 (6.8-11.2) 0 28-30 11 -- 12.6 (6.4-23.9) 12 4 3-9 (2.9-57) 0.2 {0-0.9) 18 4 2.1 (1.7-2.3) 0 28-30 10 -- 2.4 (0.3-6.7) 33 23 0 0.1 (0-1.9) The figures in brackets are ranges. Fig. 5. An area of early interstitial fibrosis from the lungs of a rat 12 months after the termination of exposure to a dust cloud oflong-fibre chrysotile. The alveolar walls are thickened and the epithelial lining appears to consist almost entirely of rounded type II pneumocytes. x 350. HWBUI0009189 728 J.M.G. Davis and A.D. Jones ) treatment groups showed deposits of this thickening of alveolar walls in response to peribronchiolar fibrosis around many of the treatment with asbestos was described in smallest airways at 12 months (Table 2) but detail by Davis et al. (1986c). In its earliest levels in the animals treated with long-fibre form it was caused almost entirely by hyper chrysotile were significantly higher than for plasia of alveolar lining cells (Fig. 5) but later the group treated with the short-fibre mater there was considerable deposition of reticulin ial (Pco.ooi). By 18 months both treat and eventually collagen in the septal walls. ment groups showed a reduction in peri As shown in Table 2, areas of alveolar bronchiolar fibrosis and taken overall, this interstitial fibrosis became more widespread reduction was also significant (Pco.ooi). in both treatment groups with increasing After 18 months from the start of dusting, time after the end ofthe dusting period. Mean widespread alveolar interstitial fibrosis deve areas of interstitial fibrosis found in those loped and this tended to obscure many ofthe animals that survived to within 2 months of earlier fibrotic deposits. For this reason, the final killing date were 12.6% for the estimations of peribronchiolar fibrosis were group treated with long-fibre chrysotile (11 limited to the first two killing dates. rats) and 2.4% for those treated with short- From about 18 months onwards areas of fibre chrysotile (10 rats). This difference is lung tissue in some animals showed a pro significant (Pco-oooi). In many areas of gressive thickening of alveolar septa. This lung tissue the interstitial fibrotic element of these 1 throve^ Jthe L, became j duce a j Some de. have dev these are the site although discovere certain. 1 neoplasm tial fibros Rats ft either lo developer However, long-fibre nomas, e Fig. 6. An area of advanced interstitial fibrosis from the lungs of a rat 17 months after the termination of exposure to a dust cloud of long' fibre chrysotile. The airspaces which in many cases are enlarged and no longer correspond to the original alveoli are lined with rounded type II pneumocytes. The airspace walls are greatly thickened with reticulin and collagen which can be demonstrated by special stains, x 350. Fig. 7. An fibre c adenoi. HWBUI0009190 Pathogenicity of long and short fibres of asbestos in rats 729 these lesions remained predominant mesotheliomas, animals treated with short- throughout the study (Fig. 6) but in others fibre chrysotile developed only one adenoma ) the hyperplasia of alveolar epithelial cells and six carcinomas (Table 3). These overall ,/per- became progressively more marked to pro tumour numbers are significantly different >ut later duce a pattern of adenomatosis (Fig. 7). (P<0.002). The figures used in Table 3 eticulin i Some definite adenomas could be seen to represent the most serious pulmonary il walls, have developed from the central regions of tumour present in any one rat. Two animals alveolar these areas and it is likely that this was also treated with the long-fibre chrysotile dust espread the site of origin of some carcinomas had benign pulmonary adenomas as well as Teasing although by the time most of these were a pulmonary carcinoma. i.Mean discovered they were too widespread to be While only two pleural mesotheliomas a those certain. All animals developing pulmonary developed in the present study, many of >ntbs of neoplasms had significant levels of intersti those rats examined in advanced age that for the tial fibrosis or adenomatosis in addition. had been treated with either the long or the tile (n Rats from both the groups treated with short-fibre chrysotile preparation showed i short- either long-fibre or short-fibre chrysotile areas of vesicular pleural metaplasia. This rence is developed some pulmonary neoplasms. type of lesion was also found in previous areas of However, while animals treated with the long-term inhalation studies with different ment of long-fibre preparation developed eight ade asbestos preparations (Davis et al. 1986a). nomas, eleven carcinomas and two pleural Areas of metaplasia consisted of loose, nation of jdand no a<" '^lis Fig. 7. An area of advanced interstitial fibrosis from the lungs of a 33-month-old rat treated with longfibre chrysotile. The epithelial lining of the airspaces has become so prominent that the pattern is one of adenomatosis. X350. HWBUI0009191 1 730 J.M.G. Davis and A.D. Jones Table 3. Pulmonary tumours and mesotheliomas found in animals treated with long and short-fibre samples of chrysotile asbestos and in a control group of rats Number of rats Long Short chrysotile chrysotile Controls 40 40 47 Tumour type Adenomas Adenocarcinomas Squamous carcinomas Histiocytomas Pleural mesotheliomas Peritoneal mesotheliomas Totals 8 6 5 1 2 1 23 11 61 00 O0 00 10 82 present, ]oper'-:'z\ pleur, J related t< tial fibre tissue an patches < have reai the presi quently f fibre chry short-fibr It is un lar pleure of plena almost a surviving present si type of c mesotheli occurred and neitJ similar to these me; ture shape, ) Noimp< types of tu extra-thoi Table 4. K chrysotile ; Fig. 8. Vesicular hyperplasia of the visceral pleural surface of a rat lung 18 months after the end of exposure to long' fibre chrysotile. Many tissue spaces have been formed in a matrix of loose connective tissue and these are lined with flattened cells of mesothelial type, x 350 fibrous tissue containing large vesicular spaces lined with flattened cells (Fig. 8) and previous TEM studies had shown that these cells were of mesothelial type. Occasionally the walls between vesicular spaces were so thin that they consisted of two closely apposed layers ofextended and flattened cells with no basement membrane between them. Where cells were supported. by areas of fibrous tissue a basement membrane was Number of Organ systi Digestive/pi Urinogenitc Endocrine Musculo, sic Reficulo-en Central ner Totals B, be. HWBUI0009192 hort-fibre Pathogenicity of long and short fibres of asbestos in rats 7 31 ! present. While no method has been deve| loped for the direct quantification of this 1 pleural metaplasia, its occurrence is closely | related to the presence of advanced intersti- i tial fibrosis or adenomatosis in the lung tissue and it is particularly common where i patches of this type of parenchymal lesion have reached the surface. This means that in the present study it was much more fre quently found in animals treated with longfibre chrysotile than in those treated with the short-fibre preparation. It is uncertain whether or not this vesicu lar pleural metaplasia is a potential precursor of pleural mesotheliomas or not. While almost all the chrysotile-treated animals surviving until the last 6 months of the present study showed some evidence of this type of change, only two definite pleural mesotheliomas were found. Both of these occurred in the long-fibre treatment group and neither showed histological patterns similar to the vesicular hyperplasia. One of these mesotheliomas had a papillary strac- ' tare and the other consisted of spindleshaped cells of fibrosarcomatos type. No important differences were noted in the types of non-neoplastic disease arising in the extra-thoracic tissues of animals from the two chrysotile groups or from the controls. Since most of the animals were allowed to survive until of an advanced age, the major ity of lesions present at autopsy were asso ciated with senescence. A large proportion of the rats in both chrysotile treatment groups and the control group developed tumours in non-pulmonary tissues as they reached advanced age. The numbers of tumours recorded are illustrated in Table 4 where individual tumours are recorded although a number of animals had more than one tumour present at autopsy. This applied to three animals in the longfibre treatment group, six animals in the short-fibre group and six animals in the control group. There were no significant differences between the overall numbers of tumours found in the three groups of rats. The lung content ofchrysotile found in the long and short-fibre treatment groups at the end of the dusting period and 6 months later is illustrated in Table 5. Animals treated with short-fibre chrysotile retained nearly three times as much dust at the end ofthe exposure period as animals treated with long fibre. In the subsequent 6 months, however, 89% of the short-fibre chrysotile had been removed from the lung tissue compared to only 54% of 1 Table 4. Non-pulmonary tumours found in animals treated with long and short-fibre samples of | chrysotile asbestos and in a control group of rats be end of jnnective closely aed cells m them, ireas of l' >as Number of rats examined Organ system Digestive/peritoneal iMnogenital Endocrine Musculo, skeletal and integumentary Reticulo-endotheiiai/vascular Central nervous system Totals Long chrysotile 41 BM 11 --. I 66 63 --6 ---- 13 17 B, benign; M, malignant. Short chrysotile 41 BM I1 --3 48 51 --6 ---- 10 19 Controls 47 BM 32 --1 13 4 I5 --3 --I 17 16 HWBUI0009193 732 J.M.G. Davis and A.D. Jones Table 5. Lung dust burdens from the chrysotile inhalation exposures Mean lung burden of asbestos Short-fibre chrysotile Long-fibre chrysotile Time post -------------------------------------- ------------------- ,---------------- exposure Left lung only Both lungs** Left lung only Both lungs 3 days 6 months *392 (86) 42 (to) 1019(224) 109 (26) *35 (75) 62 (15) 35i (i95) 161 ( 39) * Units are micrograms of dust (s.d. in brackets) and are mean figures obtained from groups of four rats. ** Figures obtained by calculation from the left lung dust content Table 7. Meso ) the long-fibre dust. These differences in pul monary deposition and clearance between the long and short-chrysotile preparations are both significant (Pco.oi). Intraperitoneal injection studies Sizing ofdusts by scanning election microscopy. The dust samples oflong and short chrysotile used in these injection studies were collected from the inhalation chambers used in the inhalation studies by an elutriation process in order that they would be as comparable as possible with the dusts inhaled by the rats. It is recognized, however, that dust samples collected in this way will not be identical in fibre size distribution to airborne dusts and a separate SEM sizing exercise was undertaken with the dusts used for injection. The results obtained are illustrated in Table 6 expressed as fibre numbers per milligram of injected dust in each size range. Difficulties in deter mining from these figures an exact number of fibres in any size range needed to produce mesotheliomas are explained in the discus sion section of this paper. The numbers of peritoneal mesotheliomas found following injection of long and shortfibre chrysotile at three dose levels are listed in Table 7 together with the mean tumour Table 6. Numbers oflong and short-fibre chrysotile present in the dustsamples used in the intraperitoneal injection study Fibre length Long-fibre chrysotile Short-fibre chrysotile Ail fibres >1 >3 >15 >8 92.7 57-1 21.2 15.2 10.2 343-0 162.0 24.4 7-4 2.3 These figures represent the number offibres in each size range in each microgram of injected dust They were obtained using scanning electron microscopy at a magni fication of x 10 000. induction pe highest dose was found ir duced by the preparations. 0.2 smg, how material was than the lcuM ofshort-f tumours Within the with either tit preparation increased tu reducing dose previously be and amphiboj Even at the 2 numbers wei induction pe short-fibre pr longer to (P< 0.001). * tion period w also seen at f was equally s. have been see short-fibre cl mesothelioma Discussior The present si HWBUI0009194 Pathogenicity of long and short fibres of asbestos in rats Table 7. Mesothelioma production following injection of long and short chrysotile dust Dose Long-fibre chrysotile Short-fibre chrysotile Mean induction Mean induction Animals with period Animals with period mesothelioma (days) mesothelioma (days) 25 23 361 22 504 2.5 0.25 22 16 511 . 736 8 675 0-- Group size 24 rats in all cases. 733 sntical in ists and a idertaken fate results expressed f injected in detert number luce < theliomas ind shortare Usted a tumour aperitoneal induction periods for each group. At the highest dose level of 25mg, little difference was found in the numbers of tumours pro duced by the long and short-fibre chrysotile preparations. At dose levels of 2.5 and o.25mg, however, the short-fibre chrysotile material was significantly less carcinogenic than the long (Pco.ooi). The o.2smg dose ofshort-fibre chrysotile failed to produce any tumours in the group of 24 rats. Within the experimental groups injected with either the long or short-fibre chrysotile preparation there was evidence of an increased tumour induction period with reducing dose of dust. This phenomenon has previously been reported for both chrysotile and amphibole asbestos (Bolton et al. 1984). Even at the 25mg dose level where tumour numbers were similar, the mean tumour induction period was different with the short-fibre preparation taking significantly longer to produce mesotheliomas (Pco.ooi). This increased tumour induc tion period with short-fibre chrysotile was also seen at the 2.5mg dose level where it was equally significant and would probably have been seen with the 0.2 5mg dose if the short-fibre chrysotile had produced any mesotheliomas at this level. Discussion The present study has confirmed the impor tance of fibre length in the pathogenicity of chrysotile asbestos although the results are less clear cut than those recently reported for amosite where the inhalation of short fibres produced neither pulmonary fibrosis nor tumours (Davis et al. 19866). The reason for this lies almost certainly in the quality ofthe short-fibre chrysotile preparation that was available for inhalation studies. In the origi nal report of Jolicoeur et al. (1981) it was stated that the sedimentation techniques which had been developed allowed the pro duction of chrysotile samples with all fibres below 8/un in length. It was found, however, that while this certainly applied to samples of a few milligrams, it was not practicable for the 1.5kg needed for long-term inhalation studies. In the airborne dust clouds at a respirable mass dose of iomg/m3, the shortfibre cloud had only five times fewer fibres > 5/an in length as seen by PCOM than the long-fibre sample used for comparison. How ever, at greater fibre lengths the differences were more substantial with 20 times fewer fibres > 20/un in length and 80 times fewer fibres > 30ym in length. This short-fibre chrysotile sample cer tainly represented the shortest chrysotile that has been available for long-term inhala tion studies and has permitted a meaningful comparison to be made with long-fibre dust using both inhalation and injection tech niques. Because of fibre levels of over 300 HWBUI0009195 734 J.M.G. Davis and A.D. Jones fibres/ml > 10/tm in length, it is not surpris ing that the short-fibre dust cloud produced both pulmonary fibrosis and pulmonary tumours. However, the long-fibre dust with five times more fibres > 5pm in length and six times more fibres > 10pm in length produced approximately six times more advanced interstitial fibrosis in the oldest animals and three times more pulmonary tumours. These figures could indicate that whereas only the relatively long fibres stimu late fibrosis, the shorter fibres (<5gm in length) which were present in much larger numbers in the short-fibre preparation, do have some carcinogenic potential. With an. amosite preparation, however, where ex tremely few fibres were > 5gm in length, neither interstitial fibrosis nor pulmonary tumours developed (Davis et al 1986b). It has been reported previously from this unit (Davis et al. 1986c) that asbestos-related pulmonary tumours in rats tend to develop from areas of interstitial fibrosis via a stage of epithelialization of airspaces which leads to adenomatosis. It is likely, therefore, that these stages of interstitial fibrosis and adeno matosis represent the direct response to the dust, with the progression of 'precancerous' lesions to definite tumours being a more haphazard process. Ifthis is so then a fibrous dust that is unable to produce fibrosis, is unlikely to produce tumours. An understanding ofthe precise number of fibres in any size range necessary to produce tumours has been an important research goal for many years. The problem is, how ever, beset with technical difficulties, par ticularly where inhalation studies are con cerned. It is relatively easy to determine the numbers and sizes offibres in a dust cloud but the proportions reaching the alveoli and retained in the lung for long periods will vary with fibre size and the chemical durability of the fibres. A number of workers including Wagner etal (19 74), Middleton et al. (1977) and Davis et cd. (1978) have shown that following exposure to the same respirable dust mass, much more amphibole than chrysotile was present in rat lungs at the end of both long and short inhalation periods, it was suggested by Timbrell (1973) that this was due to the curly chrysotile fibres being more likely to impact on the upper airways and to be cleared on the mucociliary escala tor within a few hours. Recent work with long and short amosite (Davis et al. 19866) and the present study with long and shortfibre chrysotile have allowed a better under standing of the interrelationships between deposition, mechanical clearance from lung tissue, and fibre dissolution. With both amosite and chrysotile, signifi cantly more short-fibre material than long was present in the lungs at the end of a oneyear dusting period, confirming that long fibres penetrate less easily into the lung parenchyma. The mean figures in question were 3570 and 5640/rg/rat for amosite and 350 and I02ojig/rat for chrysotile. These figures demonstrate, however, that much more long-fibre amosite was present than the very much shorter chrysotile material, indi cating that fibre penetration was not the only factor in determining tire lung dust burden. During the 6-month period after the end of dusting the short-fibre amosite and chryso tile preparations cleared more rapidly than the long from the lung; at this time the long and short-fibre amosite burdens were 3080 and 4470/ig respectively with the compar able figures for chrysotile being 161 and loggg. This more rapid clearance of short fibres would be expected if mechanical remo val ofthe short, easily phagocytosed fibres by pulmonary macrophages was the main pro cess involved. However, the proportion of chrysotile removed was much higher than for the equivalent amosite preparations with clearance of long and short-fibre amosite at 14 and 20% respectively compared to 54 and 89% for the long and short chrysotile. It is possible that macrophages phagocytose and remove chrysotile fibres more readily than amosite but there is no evidence of this and it is more likely that these findings demon strate that dissolution of chrysotile fibres in lung tissue is an important factor in the removal of this dust. These observations fit very we whe. found autopsy, known t< inhaled ( The cc tion and present b offibresi in the iu later. Inf problems should pi number tumours, problem i offibres ii variable 1 are inject fibres froi technical it is diffici fibres on comf With'w* and even to aggref avoiding ultrasonic ration, a examinin There is a will causi larlychry counted. In spite be a wor counts fro tile prepai (Table 6). ably repi number o separator in the tiss> of the nur to stumble regar. ) chrysoumS HWBUI0009196 % ods. It t this Bxt w ays escalark with 1986b) d shortr underjetween >m lung , slgnifian long of a onelat long he lung question >site and e. These it much than the ial, indi; the only burden. f) >f l wfysodly than the long are 3080 comparx6l and : of short calremo1 fibres by nain proortlon of ;her than ions with mosite at to 54 and otile. It is ytose and .dily than this and it s demone fibres in or *n the V ;S fit )J Pathogenicity of long and short fibres of asbestos in rats 735 very well with findings from lung tissue where far more amphibole than chrysotile is found in the lungs of asbestos workers at autopsy, even when chrysotile has been known to constitute by far the bulk of dust inhaled (Pooley 1976, Gylseth et al. 1983). The combination of variable fibre deposi tion and dissolution makes it impossible at present to calculate accurately the numbers offibres in a dust cloud that will be available in the lung to cause disease many months later. Injection studies where there are no problems of deposition and clearance of dust should permit an easier examination of the number of fibres necessary to produce tumours. With chrysotile, however, the problem is still complicated by the separation offibres into individual fibrils which may be a variable process when large masses of dust are injected. Electron microscope counting of fibres from a liquid suspension also presents technical difficulties. With all asbestos types it is difficult to be certain that the filtration of fibres onto nuclepore filters results in a completely even distribution of the material. With chrysotile this is particularly difficult and even in dilute suspensions the fibres tend to aggregate into clumps. One method of avoiding this is the use of a brief period of ultrasonication of the solution before filt ration, a process adopted by most workers examining the fibre content of human lungs. There is a risk, however, that this treatment will cause some splitting of fibres, particu larly chrysotile, and so increase the numbers counted. In spite of these problems, it was still felt to be a worthwhile exercise to produce SEM counts from the long and short-fibre chryso lite preparations used in the injection studies (Table 6). For the reasons stated, they prob ably represent an overestimate of the number of fibres injected although with the separation of chrysotile into individual fibrils in the tissues, they will be an underestimate of the number of active fibre units available to stimulate tumour production. However, regardless of the level of accuracy of the chrysotile counts obtained by SEM, it is obvious that very large numbers of fibres must be injected into the peritoneum of rats to produce mesotheliomas. The dose ofshortfibre chrysotile which produced no tumours in 24 rats was calculated to contain 8.6 x io9 fibres of all lengths and 57 x 106 fibres > 8ym in length (Stanton et al. (1977) calculated that fibres > 8^m in length were the most carcinogenic). While the number of animals injected was too small for it to be certain that this represents a definite zero response, the Injected dose was still very substantial in terms of fibre number. Even allowing for the short life span ofthe rat, this seems to be strong evidence that substantial numbers of fibres must reach the pleura in order to make it likely that mesothelioma will develop. This may be of relevance to current fears of the possible carcinogenic effect of extremely low exposures ofhumans to asbes tos, such as occur in the environments of public buildings and schools. Acknowledgements This study was funded by the Institute for Research and Development of Asbestos (IRDA) Montreal, Canada. The authors would like to acknowledge the skilled technical work of Mrs K. Niven, Mrs D. Lyster, Mr D. Hay, Mr S. Clarke and Miss G. Liddle and the secretarial assistance of Mrs M. Brebner. References Altoy N.G., Banmeld C.F., Baxter R.I. et al. (1977) GENSTAT. A general statistical pro gramme. Rothamsted Experimental Station. Asbestos International Association (1979) Reference method for the determination of airborne asbestos fibre concentrations at work places by light microscopy (membranes filter method). (Recommended Technical Method no. 1) London: Asbestos International Associ ation. Beckett S.T. (1975) The generation and evalu ation of UICC asbestos clouds in animal expo sure chambers. Ann. Occup. Hyg. 18,187-198. Bolton R.E., Davis J.M.G., Donaldson K. & ' Wright A. (1982) Variations in the carcinoge- HWBUI0009197 736 J.M.G. Davis and A.D. Jones nicity of mineral fibres. Ann. Occup. Hyg. 26, 569-582. Bolton R.E., Davis J.M.G., Miller B.. Donaldson K. & Wright A. (1984) The effect of dose of asbestos on mesothelioma production in the laboratory rat. In VI International Pneumoconio sis Conference. Bochum: Bergbau-Berufsgenos- senschaft. pp. 1028-1035. Cox D.R. (1972) Regression models and life tables. J. Stat. Soc. B 34, 187-220. Davis J.M.G. (1972) The fibrogenic effects of mineral dusts injected into the pleural cavity of mice. Br. /. exp. Path. 53, 190-204. Davis Beckett, S.T., Bolton R.E., Collings P. & Middleton A.P. (1978) Mass and number of fibres in the pathogenesis of asbestos related lung disease in rats. Br. J. Cancer 37,673-688. Davis J.M.G., Addison J., Bolton R.B., Donaldson K., Jones A.D. & Miller B.G. (1985) Inhalation studies on the effects of tremolite and brucite dust in rats. Carcinogenesis 6, 667-674. Davis J.M.G., Addison J., Bolton R.E., Donaldson K. & Jones A.D. (1986a) Inhalation and injec tion studies in rats using dust samples from chrysotile asbestos prepared by a wet dispersion process. Br.). exp. Path. 67,113-129. Davis J.M.G.. Addison J.. Bolton R.E., Donaldson K., Jones A.D. & Smith T. (19866) The patho genicity of long versus short fibre samples of amosite asbestos administered to rats by inha lation and intraperitonea! injection. Br. J. exp. Path. 67,415-430. Davis J.M.G., Bolton R.E., Brown D. & Tolly H.E. (1986c) Experimental lesions in rats corres ponding to advanced human asbestosis. Exp. Mol Pathol. 4, 207-221. Davis J.M.G., Bolton R.E., Douglas A.N., Jones A.D. & Smith T. (1988) The effects of electro static charge on the pathogenicity of chrysotile asbestos. Br. J. Ind. Med. (in press). Dixon W.J., Brown M.B., Engblman L,, Frare J.W., Hill M.A., Jennrich R.I. & Toporbk J.D. (1983) BMDP Statistical Software. California: University of California Press. Dodgson J. & Whittaker W. (1973) The determi nation of quartz in respirable dust samples by infrared spectrophotometry. Arm. Occup. Hyg. 16, 373-387Gylseth B., Mowe G. & Wannag A. (1983) Fibre type and concentration In the lungs of workers in an asbestos cement factory. Br. J. Ind. Med. 40, 375-379Health and Safety Executive (1984) Asbestos-- control limits--measurment ofairborne dust con centrations and the assessment of control measures. London: Health and Safety Executive (HSE Guidance note EH10). Hilschbr W.. Sethi S. & Friedrichs K.H. (1970) Zusammenhange zwischen Asbestose und Faserlange. Naturwissenschaften 57, 356-357. Johnston A.M., Vincent J.H. & Jones A.D. (1987) Electrical charge characteristics of dry aerosols produced by a number of laboratory mechani cal dispensers. Aerosol Scl. Tech. 6, 1x5-127. JOLIOOEUR C., ROBERGB P. & FORTIER J. (1981) Separation of short fibers from bulk chrysotile asbestos fiber materials: analysis and physico chemical characterization. Can. J. Chem. 59, 1140-1148. Jones A.D., Johnston A.M. & Vincent J.H. (1983) Static electrification of airborne asbestos dust. In Aerosols in the Mining and Industrial Work Environments. Eds V.A Marple & B.Y.H. Liu. Ann Arbor: Ann Arbor Science, pp. 613-632, Kalbfleisch J.D. & Prentice R.L. (1980) The Statistical Analysis of Failure Time Data. New York: Wiley. Kaplan B.L. & Meier P. (1958) Non-parametric estimation from incomplete observations. /. Am. Stats. Assoc. 53, 457-481. King E.J., Clegg J.W. & Rae V.M. (1946) Effect of asbestos and of asbestos and aluminium, on the lungs of rabbits. Thorax 1, 188-197. Klosterkottbr W. (1968) Experimentelle Untersuchungen uber die Bedeutung der Faserlange fur die Asbest-Fibrose sowie Untersuchungen uber die Beeinflussung der Fibrose dutch Polyvinylpyridin-N-oxid. In Biologischt Wirkungen des Asbestos. Internationale Konferenz Dresden. Berlin: Deutsches Zentraiinstitut fur Arbeltsme- dizin. pp. 47-52. Langer A.M., Wolfe M.S., Rohl A.N. & Selikopf I.J. (1978) Variation of properties of chrysotile asbestos subjected to milling. J. Toxicol Environ. Health 4, 173-188. Lbmaikb I. (1985) Characterization of the bronchoaiveolar cellular response in experimental asbestosis. Am. Rev. Respir. Dis. 131,144-149. Lbmairb I., Nadeau D,, Dunnigan J. & Masse S. (1985) An assessment of the fibrogenic poten tial of very short 4-T-30 chrysotile by intratra cheal instillation in rats. Environ. Res. 36,314326. Middleton A.P., Beckett S.T., Davis J.M.G. (1977) A study of the short-term retention and clearance of .inhaled asbestos by rats using UICC reference samples. Inhaled Particles IV. Ed. W.H. Walton. Oxford: Pergamon Press, pp- 247-257. Nadeau D., PouQUErrE-CoimiEB L. & Khorami J- (in press) Mineral fiber toxicity: further evi dence for an active involvement of the surface chemistry. Paper presented at the International Symposium on Respirable Dust in the Mineral T! Industi sit' ' ") PELE ...... J study o fibres o Environ Platek S. L.E.. Fxinhalat Toxicol Pooley F-minera Canadii Res. 12 Pott F. & Ratte n Naturw Parr F., Ergebni nen W: Dentun beim N kenk,, I> 505. PottF. (x ffie car< fibrous PottF fibi J ings fro Halsa IPott F. & Lange * kanzero a confei lag Goal SMrmW.l Biologic short as A162. Stanton L of meso fibrous j HWBUI0009198 aerosols rechani5-127. (1981) aiysotite physicotem. 59, (1983) tos dust. >.ed Work :M. Liu. 13-63280) The 1to. New rametric tions. J. Effect of n, on the le Unterserlange 5? 'iraangen Dresden. beitsme- Skukoff hrysotile ' Environ. he bronsrimental 44-149. Masse S. Ic potenintratra- 36,314- s J.M.G. ition and its using !es IV. Ed. ress. pp- iorami J. ther evi&r--<ace Pathogenicity of long and short fibres of asbestos in rats 737 Industry. Held at Pennsylvania State Univer sity. Symposium report. Pelb J.P. & Calvert R. (1983) A comparative study on the haemolytic action ofshort asbestos fibres on human, rat and sheep erythrocytes. Environ. Res. 31, 164-175. Plater S.F., Groth D.H., Ulrich G.E., Stettler L.E., Finnell M.S. & Stoll M. (1985) Chronic inhalation ofshort asbestos fibres. Pundam Appl Toxicol. 5, 327-340. Pooley F.D. (1976) An examination of the fibrous mineral content of asbestos lung tissue from the Canadian chrysotile mining industry. Environ. Res. 12, 281-298. Pott F. & Friedrichs K.H. (1972) Tumoren der Ratte nach I.p. Injektion Faserformiger Staube. Naturwissenschaften 59, 318. Pott F., Friedrichs K.H. & Huth F. (1976) Ergebnisse aus Tierversuchen zur kanzerogenen Wirkung faserformiger Staube und ihre Dentung in Hinblick auf die Tumorenstehung beim Menschen.. Zentrcdbl. Bakteriol. Parasitkenk.. Infectionskr. Eyg. Abt. I Orig. 162, 467- 505Pott F. (1978) Some aspects on the dosimetry of the carcinogenic potency of asbestos and other fibrous dusts. Staub 38,486-490. Pott F. (1983) Animal experiments with mineral fibres. In Short and Thin Mineral Fibres. Proceed ings from a symposium, Stockholm. Arbete och Halsa 19, 133-1651. Pott F. & Zibm U. (1983) Zur Bedeutung von Lange und Durchmesser einer Faser fur ihre kanzerogene Potenz. In Fibrous Dusts. Report of a conference, Strasbourg. Dusseldorf: VDI Ver- lag GmBH. pp. 235-241. Smith W.E., Hubert D.D. & Badollet M.S. (1972) Biological differences in response to long and short asbestos fibres. Am. Ind. Eyg. Assoc. J. 3, A162. Stanton M.F. & Wrench C. (1972) Mechanisms of mesothelioma Induction with asbestos and fibrous glass. J. Nat Cancer Inst. 48, 797-821. Stanton M.F., Layard M., Tegeris A., Miller E., May M. & Kent E. (1977) Carcinogenicity of fibrous glass. /. Nat. Cancer Inst. 58, 587-603. Stanton M.F., Layard M., Tegeris A., Miller E., May M., Morgan E. & Smith A. (1981) Relation of particle dimension to carcinogenicity in amphibole asbestos and other fibrous minerals. J. Nat. Cancer Inst. 67, 965-975. Timbrbll (1973) Physical factors as etiological mechanisms. In Biological Effects of Mineral Fibres. Scientific Publication No. 8. Ed. P. Bogovski. Lyon: International Agency for Research on Cancer, pp. 295-303. Timbrbll V. & Skidmore J.W. (1968) Significance of fibre length in experimental asbestosis. In Biologische Wirkungen des Asbestos. Internatio nale Konferenz, Dresden. Deutsches Zentralinstitut fur Arbeitsmedizin. pp. 52-56. Timbrell V., Skidmore J.W., Hybtt A.W. & Wagner J.C. (970) Exposure chambers for inhalation experiments with standard refer ence samples of asbestos of the International Union Against Cancer (UICC). J. Aerosol. Sci. 1, 215-223. Vorwald A.J., Durkan T.M. & Pratt P.C. (1951) Experimental studies of asbestosis. AMA Arch. Ind. Eyg. Occup. Med. 3, 1-43. Wagner J.C., Berry G., Skidmore J.W. & Timbrell V. (1974) The effects of the inhalation of asbestos in rats. Br. J. Cancer 29, 252-269. Wbbster I. (1970) The pathogenesis of asbestosis. In Pneumoconiosis. Proceedings of the Interna tional Conference, Johannesburg. Ed. H. Sha piro. Oxford: Oxford University Press, pp. 117- 119Wright G.W. & Kuschner M. (1977) The influence of varying lengths of glass and asbes tos fibres on tissue response in guinea pigs. In Inhaled Particles IV. Ed. W.H. Walton. Oxford: Pergamon Press, pp. 455-477. i HWBUI0009199 is a latent and J.F. 'O .educed m with function > Kuroiw mental py of the dy of the coccal milk uenza In the rat ?ujiwara, lepatic sseter HWBUI0009200 The British Journal of Experimental Pathology editor R. Marian Hicks associate editors: N. Woolf and D. R. Katz P.P. Anthony Exeter C.L. Berry London C.C. Bird Edinburgh A.W. Boylston Leeds D. Crawford London A.D. Dayan London C.S.F. Easmon London editorial board P. Grasso Guildford R.B. Heath London P. Isaacson London P.L. Lantos London P.N. Magee London C.A.C. Mims London F.W. O'Grady Nottingham J.R. Pattison London I.F.H. Purchase ICI Macclesfield T. Slater Hillingdon G.R. Smith London D.R. Turner Nottingham M. Waterfield London N.A. Wright London aims and editorial policy The British Journal of Experimental Pathology was founded In 1920 to publish the results ofresearch into the cause, diagnosis and cure ofdisease in man. Itis an international Journal reporting original experimental investigations into the biogenesis and progression of pathologic processes which afford new insights into the basic mechanisms underlying human disease. Both the results of clinical research and ofinvestigations based on appropriate in vivo and in vitro experimental models are published. Experimental path ology encompasses the use ofmultidisciplinary scientific techniques and includes research into such diverse fields of Interest as the causes of Impaired resistance to infection, stress-related diseases, cardiovascular disease, environment-related toxicity, the effect ofdiet on disease processes including carcinogenesis, biochemical mal functions, Iatrogenic disease, the pathology of ageing, the development and use of In vitro models of human pathology and many other comparable problems in which an experimental approach is able to yield new and clinically useful data. It Is the editorial policy to give priority to the publication of manuscripts on topics of particular current interest or ofImmediate clinicopathologtca! concern. subscription information The British Journal of Experimental Pathology is published bi-monthly (1 volume per annum) and the subscription prices for 1988 are 82.50 {UK), 98.50 (Overseas except North America), 8167.50 (US and Canada), in all cases post free. Orders for current subscriptions and back issues should be sent to Blackwell Scientific Publications Ltd., P.0. Box 88, Oxford OX2 ONE; all other business correspondence, including orders for offprints and advertising space, should be addressed to Blackwell Scientific Publications Ltd., Osney Mead, Oxford OX2 oEL (Telephone no. 0865 240201, telex 83355 Medbok G, fax 0865 721205). dispatch The Journal is dispatched within Europe by surface mail, to other continents by various forms of air speeded delivery: to the U.S.* by air freight for forward ing by second class post, to India by air freight for guaranteed local delivery, and to all other countries by Accelerated Surface Post. copyright and photocopying 1988 Blackwell Scientific Publications Ltd, on behalf of The British Journal of Experimental Pathology. Published by Blackwell Scientific Publications Ltd on behalf of the British Journal of Experimental Pathology, a company limited by guarantee. Registered Office: 1 Old Burlington StreetLondon Wi. Registered in England 355858. Directors: F. W. O'Grady cbe to (President), A. C. Thackray (VicePresident), J. V. T. Gostling, R. Marian Hicks, P. J. Lachmann prs, J. R. Pattison, C. E. Gordon Smith cb, D. A. J. Tyrrell cbe prs, N. Woolf (Secretary). Authorization to photocopy for internal or personal use or the Internal or personal use of specific clients is granted by Blackwell Scientific Publications Ltd. for libraries and other users registered with the Copyright Clearance Center (CCC) Transactional Reporting Service, provided that the base fee of 803.00 per copy is paid directly to the CCC, 27 Congress Street, Salem, MA 01970, USA. This consent does not extend to other kinds of copying, such as copying for general distribution for advertising or promotional purposes, for creating new collective works or for resale. Special requests should be addressed to the Editor. 007-1021/88 $03.00. business correspondence Orders and remittances relating to subscriptions .and offprints should be sent to Blackwell Scientific Publications, PO Box 88, Oxford. advertisements Advertisement bookings and enquir ies should be sent to The Advertisement Manager, Blackwell Scientific Publications, Osney Mead, Oxford 0x2 oel (Telephone 0865-240201, Fax 0865721205). * Second class postage paid at New York, NY. Post Master, send address changes to British Journal of Experimental Pathology, c/o Expediters of the Printed Word Ltd., 515 Madison Avenue, New York, NY 10022. 1 Do yoi D. Mac This boc tables in use to cl publishec encomps values, ar will enab the planr 1987. 22 HWBUI0009201