Document O3LDD1MoMbkngn6EeXbnaN57p

environmental research 8, 178-202 (1974) Inhalation Carcinogenesis from Various Forms of Asbestos1 Andrew L. Reeves, Henry E. Puro, and Ralph G. Smith2 Department of Oeeupational and Environmental Health, School of Medicine, Wayne State University, Detroit, Michigan 48201 lieeeiced Novend>er 23, 1973 lints, rabbits, guinea pigs, gcrbils, and mice were exposed to the inhalation of chrysotile, crocidolite, or amosite for 2 years. Mean atmospheric concentrations were 47.9-50.2 mg/m\ but only 0.08-1.82% of the dusts retained fibrous morphology during the dissemination procedure which involved hammer milling. Trace eonlamination especially by chromium and nickel was also increased. Light microscopic fiber counts per ml chamber air were 54 (chrysotile), 1105 (crocidolite 1, and SOI (amosite). A (ibmgenie response to these dusts was observed in all live animal species, the severity corresponding to the extent of exposure (with reaction to chrysotile frequently very slight). Gcrbils developed frequent alveolar proteinosis. Mice developed spontaneous papillary carcinomas in the lungs. Disregarding the latter species, carcinogenic response to asbestos inhalation was restricted to rats and occurred in all three exposure groups. There were 2 lung cancers and 1 pleural mesothelioma after chrysotile inhalation; 4 lung cancers after crocidolite inhalation; and 1 lung cancer and 2 pleural mesotheliomas after amosite inhalation. These eases constituted 7-9% incidence of malignancy among rats with adequate survival record. Hypotheses of asbestos carcinogenesis are reviewed and it is suggested that drib-rent eliologie principles may be involved in the causation of lung cancer and of pleural mesothelioma. Tin; carcinogenic potential of asbestos in tin; environment Inis beet line one of tbe leading public health concerns of ottr time. The hazards of asbestos exposure were well demonstrated recently in miners and mill workers by McDonald et al. (1971), in pipe insulation workers by Sclikoff et al. (1970), in shipyard work ers by Harries et al. (1972), in coke oven operators by Selikoff and Hammond (1971), in cigarette filter manufacturers by Goff and Gaensler (1972), and in automobile brake repairmen by Ilickish and Knight (1970). Asbestos hazard was also considered to originate from battery boxes (Greenberg, 1970), protec tive clothing (Ramber and Butterworth, 1970), and certain types of soils during agricultural work (Burilkov and Michailova, 1970). Nonoccupational and nonrespiratory exposure to potentially significant quantities of asbestos was postu lated for consumers of tale-coated rice in view of the tremolite asbestos in the talc (Merliss, 1971); various bottled beverages including beer (Cunningham and Pontefract, 1971); and parenteral drugs (Nicholson et al., 1972). In the two latter cases, the asbestos content of the fluids was believed to be contributed from the filters during filtration, and it was speculated that migration of the fibers from the gastrointestinal tract to mesothelial tissue might occur in a similar way 1 Supported by Grant No. 5-RO-l-EC 00240 from the U. S. Public Health Service. 2 Present address; School of Public Health, University of Michigan, Ann Arbor, Mich. 48104. 178 Copyright 1974 by Academic Press, Inc. All rights of reproduction in any form reserved. CAHCINOCKN1CS1S FROM AS1SESTOS 179 as from subcutaneous injection sites (Westlake et al., 1965; Kanazawa et al., 1970; Pontefract and Cunningham, 1973). The presence of asbestos fibers in urban air is now well established (Selikoff et al., 1972b), although whether or not the ambient concentration is high enough to constitute a hazard is uncertain (Rickards and Radami, 1971). In the lung of city dwellers, both uncoated asbestos (Langer et al., 1971), and ferruginous bodies (Gross et al., 1969) are common findings, although the central core of the latter is not always asbestos (Gross et al., 1970). Much of the atmospheric asbestos in the cities probably originates from building demolition and weather ing. The significance of vehicle brake emissions is uncertain, since at the tem perature of brake application part or all of asbestos undergoes crystallographic transformation into nonfibrous pyroxenes and forsterite (Speil and Leineweber, 1969). However, the degree of completion of this transformation during actual vehicle operation is not known, nor has the chronic inhalation toxicity of the heat decomposition products of asbestos been thus far adequately investigated. Malignant neoplasia now recognized to result from the inhalation of asbestos fibers includes cancer of the lung, and mesothelioma of the pleura and perito neum (Wagner et al., 1971). Also suspected are links to cancers of tin* gastro intestinal tract (Selikofl et al., 1967), hematopoietic system (Gerber, 1970) and female reproductive system (Graham and Graham, 1967). Some of the former lesions were reproduced experimentally by several workers during the past dec ade. Pleural and peritoneal mesotheliomas were readily produced through local implantation of asbestos fibers into various laboratory animals by Wagner (1962), Smith et al. (1965), Stanton et al. (1969), Wagner and Berry (1969), Donna (1970), and Reeves et al. (1971); a few lung cancers were also produced in rats by inhalation exposure to chrysotile by Gross et al. (1967) or to crocidolite by Reeves et al. (1971). In this paper, we report on the continued production of various neoplasms in the rat after inhalation of chrysotile and crocidolite; and on thi' production of papillary carcinoma of the lung and fibrous or fibrosarcomatous mesothelioma of the pleura in the rat after inhalation of amosite. MATERIALS AND METHODS 1. Dust Characterization Crude samples of amosite, crocidolite, and chrysotile3 were ballmilled in glazed ceramic jars with stainless-steel balls for 10 days, and the resultant dusts were forced through a #8 mesh screen for the distintegration of larger clumps. These specimens were fed into the hopper of hammer mills and blown into the exposure chambers as described previously (Reeves et al., 1971). The milling procedure caused not only substantial reduction of particle size which was intended, but also a loss of fibrous structure in a great proportion of particles which was not intended. By using an axis-diameter ratio of 3:1 or more to define a fiber, we found that only 0.08-1.82% of the dust as collected on Millipore filters in the chambers answered this definition during light microscopic * Amosite, W-3 fibers; crocidolite, 5-blue fibers; and chrysotile, 3-T fibers. These samples were obtained by courtesy of the Johns-Manville Corporation. ISO HK.KVKS, PUUO AND SMITH examination. The upper limit of fiber length observable under the light mieroscope was 55-130 /un (comprising 0.001-0.015% of the dust) in various samples, and the fiber count per ml air was 864 (amosite), 1105 (crocidolite), or 54 (chrysotilc). It is thus apparent that the great majority of light-microscopically visible fibers were destroyed or transformed into submicronic fibrils or apparently nonfibrous crystals. Similar results were observed using X-ray crystallographic technique after a grinding procedure involving cutting action by Occella and Maddalon (1963). It is well known that chrysotile is changed into forsterite at 810C, and the amphiboles into pyroxenes at 900C. Whether or not such high impact temperatures might have occurred during ball milling is not known. If they have, this might be a plausible explanation for the loss of fibrous structure in 98-99% of the particles in the light microscopic range. However, it should be emphasized that even with this degree of destruction of fibrous structure, the animals still inhaled substantial quantities of apparently unchanged asbestos. It can be estimated that atmospheric concentration of the fibrous component alone in the chamber was 40-900 y/m3, and the fiber counts as reported above were 11-220 times higher than the current threshold limit value for asbestos as established by the American Conference of Governmental Industrial Hygienists (Stokinger et al., 1973). Moreover, a similar destruction of fibrous structure was not observed in the electron microscopic range; under the electron microscope it appeared that the majority of particles were fibrous and average length and diameter of dust samples collected in the chambers were as follows; amosite, .'1-5 and 0.2-0.5 /un; crocidolite, 3-6 and 0.4-0.5 / in; and chrysotilc, 6-15 and 0.2 /mi. X-ray diffraction line positions of asbestos samples agreed well with those reported in the literature (Timbrell, 1970). The loss of line intensity (measured as difference in peak heights on samples obtained before and after the dissem ination procedure) was computed as percent of the UICC reference value. The results (Table 1) show about 70% loss of diffractivity for amosite, and about 35% for crocidolite and chrysotile. These values arc uncorrected for the effect of particle size diminution and they indicate the extent of loss of crystallinity only in an approximate manner. Another artefact introduced during dust preparation was heavy metal con tamination. The ballmilling procedure was accomplished with stainless steel balls of /2-2 in. diameter, having the following trace element composition:4 Cr 1.3--1.6%; C 0.9,5-1.1%; Si 0.2-0.35%; Mn 0.25-0.45%; Ni < 0.35%; Cu < 0.25%; Mo < 0.08%; P < 0.025%; S < 0.025%. It is apparent that during the prolonged contact of the balls with the asbestos samples there was potential hazard of trace element adsorption. Analytical-chemical studies were conducted using atomic absorption spec trophotometry in order to compare raw and processed asbestos samples and to relate each to UICC standard reference samples (Rov-Chowdhury et al., 1973). About 1 g specimens of each dust were digested with the minimum amount of 4 According to analysis of flic manufacturer (Atlas Ball Division of SKF Industries, Philadelphia, Pa.) CAltCINOCENESIS KIIOM ASBESTOS 181 TAHITI l X-Ray D IKKttAt TION Diac.uam ok Ashkstos Dusts .1 .\P cmde JM disseminated Sample X-ray line (decrees)" IMC( > or (units) (unit s) UICC'1) (r, or ' , Ins> ol` |M*nk ht (units) l 11 ('C ) during grinding Amnsilt* 10 7 10.11 27.d 27.7 20.1-20.4 22.2-22.S 78 78 12 12 20 41 8 10 100 100 102 122 Crocidolite 10.o-lO.S 67 62 07 10 7-10 0 0 0 100 26.0-26.7 <)<> 14 40 28.7-20.0 10 20 102 22 0 -22.2 14 14 100 24 1 2-1 7 5 5 100 .n .* .{;> 7 s 7 ss ('hrysut ilc 12.0 12.4 18 7 10 8 2-1.2 24,7 60.2-60.7 55 60 10 10 22 10 66 111 100 122 100 20 '\S 62 av 4 21 60 70% 10 26 77 5 28 70 do 20 20 av T) 55 41 :>5r, 18 22 0 48 55 12 86 14 ;i 60 40 <> 75 II ii 71 ; a v r co 10 55' t 16 20 60 5 82 17 " In CtiA'u-radiation, X = 1.24 A. 6 Cnion Internationale Contre le Cancer reference specimens. ` Johns-Manville Corporation eonnnereial specimens. I IF in Pt crucibles and slowly evaporated to dryness. The ash was twice divested with concentrated UNO, and then dissolved in In IICI. The elements Co, Ni, Cr, Mil. and Fe were determined using the 2407, 2302, 3579, 4030, and 3470 A resonance lines, respectively. Iron was also determined spcctrophotomclrically as hydroxylaminc-IiCl and o-phenanthroline complex. The results are sum marized in Table 2. It appears that with UICC samples, our results agreed well with those of Timbrel! et ill. (1968; 1970), and in most cases there was no great discrepancy between these values vs those obtained on crude commercial samples. JM-3T chrysotile resembled UICC-B chrysotile, although its nickel concentration was appreciably lower. Substantial contamination by chromium was introduced during grinding into amosite and crocidolite, and by nickel into crocidolite, the factors being 3X, 9X, and 5x, respectively. The other values appear to be within or close to ex perimental error limits. 2. Chamber Maintenance and Monitoring Three "walk-in chambers" (12x12 ft rooms with perforated walls) were utilized in this experiment, one each for the dissemination of amosite, crocidolite, and chrysotile. These were described in more detail previously (Reeves et al., 1971). The dusts were blown into these chambers through a network of ducts 182 REEVES, PURO AND SMITH a a *f> o CM CM CM eo * - .o <*> I ^ ^ w 2 -H -H +t ~ 2 -H +1 -H Zf o? "H ^g -H +1 +1 V co CMo X*o V cm cm x -* ~ XX N CCO! *f -t CM 01 a t- +1 -H -H cm -H -H +1 rf*o-H-H-H-H ro o CM O V t} *?" -- CO rf CO CO S 0) a -H -H -H *H0 XCM INfl M ^ iO X X -H -H CO CM CO CM *t co *r -H -H -H CM x >o X CM CM CM saa c Ooirs rQof O C1-O*- i--OH +1 -_H -H i <--> i !> ^ t- +1 +1 +! X X CXM C0 CCMO co ,. M CO -H -H -H CO CO f *f fX co ** C-l So. W 4 m-4, ut* ^g o ao m ic 3 *g -H -H -H XCO -f CM C --' t-- CM CM -* +1 -H +1 X X '? o -H 2 -H +1 -H 01 ^ CO o -Sfl3 o S3 sot- SO 3V * s 2 *< S^a C8 cS C I1 32 -o 2 -Oca Cj _ . 3 $3 < y< .2 5 3o <^ 0 a a -rt -Q5^ "-4o3) -g a n. cG5 j? tf e Tj X0>?i: Oa J X S.E-1 2 * , is Eh s -cg EH . oC3 ^ S3 3 < < < < -s a (A O o a 3 <0 " I^ -* o o U> 3 7\ 3 Si w C u u< p o a ^3 e3 l w c -= 0) _ QJ 3 2 *35 o 'c-l x E u JZ <oo CARCINOGENESIS FROM ASBESTOS 183 with the aid of hammer mills and fan systems. The chambers were in operation 4 hours/day, 4 days/week, with each Friday reserved as cleanup day. The ex posure lasted 2 years with 1480 cumulative total exposure hours. Dust concentration was monitored regularly throughout the experiment, through weekly air samples collected in each of the chambers on Millipore filters. The measurements yielded an overall mean (in mg/m3, SD) of 48.6 3.0 for amosite; 50.2 3.6 for crocidolite; and 47.9 3.0 for chrysotile. The aerosols appeared as dense dust clouds readily visible to the naked eye but with little tendency to settle on surfaces. The used asbestos was exhausted from each chamber into plastic collection bags. 3. The Animal Colonies 207 rats, 96 guinea pigs, 60 rabbits, 90 mice, and 204 gerbils6 of both sexes were evenly divided among the amosite, crocidolite, and chrysotile chambers. Exposure TAREK 3 Suuvky of tub Animal Coi.oniks Specimen count (Inven tory of live animals at) (months) Mice Gerbils Rats Guinea Rabbits pigs Amosite Crocidolite Chrysotile Control Total no. of animals 0 (> 12 18 24 No. slides prepared 0 6 12 IS 24 No. slides prepared 0 6 12 18 24 No. slides prepared 0 0 12 IS 24 No. slides prepared 30 68 69 20 23 64 62 14 17 60 56 12 0 f>l 46 10 0 46 19 10 19 58 66 20 30 68 69 20 26 63 63 14 18 58 57 12 0 49 46 9 0 46 22 9 22 49 66 20 30 68 69 20 26 64 62 16 19 56 56 13 0 50 43 11 0 44 14 11 17 59 57 20 10 12 12 12 8 11 97 6 10 76 0 8 54 0 8 33 10 12 12 12 100 216 219 72 32 20 20 13 8 30 32 20 20 14 6 31 32 22 20 14 13 32 12 8 8 5 4 11 108 3 Charles River CD rats; Camm-llartley guinea pigs; Shankin Farms Dutch rabbits; Stout Farms Swiss mice; and Hasenau Mongolian gerbils. 184 REEVES, PURO AND SMITH respectively, with a smaller number (10-12 of each species) being retained as unexposed controls. The animals were 3-6 weeks old when shipped. Sample necropsies were obtained after 3 and 6 months, respectively, with the majority of animals allowed to survive the complete course of exposure. Altogether, out of a total animal count of 715, there were 417 animals necropsied on schedule; 249 animals died in the course of the experiment through attrition; and 49 were lost to cannibalism or otherwise unaccounted for. Attritional losses were heaviest among rats during the fourth semester of exposure when 25-30 animals were lost for unexplained reasons. There were no indications of infectious epizootic and no antibiotic medication was given. It may be that the mortality was the consequence of fully developing pulmonary fibrosis, perhaps in combination with higher summertime temperatures. Mortality among the other animal species was normal, and there was no gross differential effect of exposure to any of the dusts on survival. The complete animal inventories are summarized in Table 3. Necropsy proedures were as reported before (Reeves et al., 1971). Lung weight was obtained after excision, in wet state, but blotted between sheets of filter paper. Histopathologic slides were obtained from the lungs and pleura of alto gether 643 animals, with varying length of exposure. Hematoxylin-eosin, Van Gieson's elastic, and K4Fe(CN) stains were applied to the sections. RESULTS Table 4 summarizes the essential histopathology findings and lung weight/ body weight ratios. It may be seen that exposure to various kinds of asbestos has increased the latter in each species in the following ascending order: cluysotile; amosite; and erocidolitc. Severest effects wen: seen with gerbils (crocidolitccontrol difference t -- 3.14, significant at 99.57 confidence level); and with rats (crocidolitc-control difference t -- 2.36, significant at 95% confidence level). With the other species, the differences were of low or borderline significance, and in each species amosite and cluysotile caused successively less effect than crocidolite. Tumor-bearing animals were excluded from this computation, so that the results are indicative of the degree and extent of asbestosis. The details of histo pathologic impressions with each animal species were as follows. (a) Hats Among control animals there was one bacterial pneumonia but no other sig nificant pathological changes. Among exposed animals, histiocytic and giant cell response was universal and gradually increasing in the course of exposure. Fibrotic response was most severe in the crocidolite group. Fiber deposits were frequently seen but there were few readily apparent ferruginous bodies and al most no necrosis. A few cases of pneumonia and bronchitis were seen. In the crocidolite group, frequent pleural reaction with collagen accumulation occurred, as well as hemosiderin accumulation in the lymph nodes. There was filling of some alveoli with histiocytes, giant cells, and cholesterol. A proliferative response was especially frequent in the crocidolite group, where 8 animals (on the 567, 585, 691, 693, 727, 772, and 785th day, respectively) had squamous metaplasia of the alveolar lining. In 2 erocidolite-exposed animals, epithelial papillomata were seen in the bronchi, which appeared histologically CARCINOGENESIS FROM ASHESTOS 185 benign (Fig. 1). There were also the following definite malignant neoplasms in all groups: (1) Amosite. Rat No. 1206-23, male, found dead after 539 days of exposure. The lung and pleura contained a large neoplasm of fibrosarcoma pattern with adhesions to the chest wall (Figs. 2 and 3). Rat No. 1206-26, male, found dead after 564 days of exposure. The lung contained an osteosarcoma, probably metas tasis of a primary bone tumor. Rat No. 1206-42, female, necropsied after 584 days of exposure. The pleura contained a large fibrous mesothelioma (Fig. 4). Rat No. 1206-53, female, found dead on the 18th day alter completion of ex posure. The lung contained a bronchoalveolar carcinoma with papillary pattern, composed of tall columnar cells (Fig. 5). A similar neoplasm was seen on the pleural surface. (2) Crocidolite. Rat No. 1208-48, female, found dead on the 12th day after completion of exposure. The lungs contained a bronchogenic adenocarcinoma. Rat No. 1208-49, male, found dead on the 18th day after completion of exposure. The lungs contained a squamous cell carcinoma. Rat No. 1208-55, female, sacri ficed 45 days after completion of exposure. The lungs contained a squamous cell carcinoma (Figs. 6 and 7). Rat No. 1208-69, female, necropsied 56 days after completion of exposure. The lungs contained a squamous cell carcinoma. (3) Chrysotile. Rat No. 1207-41, male, found dead after 617 days of exposure. The upper mediastinum contained an area of proliferative fibrosis histologically classified as a low-grade fibrosarcoma of probably pleural origin. Rat No. 1207 48, male, found dead after 608 days of exposure. The lungs contained a papillary carcinoma. Rat No. 1207-61, male, necropsied 43 days after completion of ex posure. The lungs contained a well differentiated squamous cell carcinoma (Fig. 8). (b) Rabbits In control animals there were no significant pathological changes. Amosite and crocidolite caused histiocytic and foreign-body responses, with the extent of re action well correlated to the length of exposure. Fight to intermediate fibrosis was observed in animals surviving the complete course of exposure. Inflam matory reaction was infrequent, and chrysotile caused remarkably fewer lesions. There were no malignancies. (c) Guinea Pig,s Control animals had no significant pathological changes. Among exposed ani mals, this species was remarkable for conspicuous abundance of ferruginous Irodies, especially after crocidolite exposure. Foreign-body and giant-cell re sponse, as well as gradually developing fibrosis, were common in all exposure groups. There were a few cases of adenomatosis but no malignant neoplasia. ((I) Garbits Control animals had no significant pathological changes. Fxposed animals showed foreign-body response, light to moderate fibrosis, and a conspicuously high incidence of focal and generalized alveolar proteinosis, especially in the amosite and crocidolite groups. There were no cases of malignant neoplasia. 186 REEVES, PURO AND SMITH 2 -OCj c hfi b2 -- Cs - o. s JQ X 05 ^ r, . o * G. qj Q. Oi 5a5 Cc a05. Cc cs x t^. *2 c p 5 --H -u 3 *%5 g '| | I o a. * --~ a y. og9*- ToXJ S3 >- v ^5 w.mJ SoI <t a S .2 5 *CcX/3 'Korfi S3 S-5 ~ g 95 C- O>> X "S -- ? *a y, u* U- c .2 r j> c .2 X S U- *< gE ^ c 5^ H 0- x + b) b -J I CQ o v: _ CO CO & -- -- w CO O -H t-. o> -H -H +1 +1 -H +1 +1 o o ^ (M t>- ** CO iO --o^ CO CO *o X c -J. * 6 y< as>- as; ^ *rf o t- o'f o *r Ol m* o +1 -H lO X XX CO ~r J13 03 t3*C1 Oc ,pG22 >fc>i Zqj 0> ' o7? O J3 2 A 1 OA A 7'o2 to- tt Vo. a OUU< VO o < C C 03 *3 *k3_*> cc 0X k. -C Oo --au> ss C CS CARCINOGENESIS FROM ASBESTOS *73 Sd bt 9 M beC --es -H a o cS o Ec .S cx3 -- uua 8 to Qtn "W3 o eQj * --^ S .fl O 3ooo* 2gQ.* 6 ato CO -- ^ eo co 55 C*O to OsS _Q o CO o iO -H -H tO o ci --o 'A S /Z- X5) .S .U-5* .3 U* 3C"W0t'")*1 .atkOno. t/2 - o5?- .2 V .a .2 .2 o .2 o o .2.Sf .a js .a jo /-. k K E X OI o O' CO o +1 +1 -H CO X' o to X CO +1 CO uO O o t* to CS IM o CO +1 +1 +1 -H CS 40 CO 40 Tf 187 ce from co n tro l significant a t 9.5 c levelr (fiu / = 2.36). TOS s o oo ojo oc ai 'x o OCC3 i OtoUrn oL. o o < g Sc > 8^ > to s c c too tod 188 KKEVKS, 1'UHO AND SMITH (e) Mice Control animals hud generally no significant pathological changes, except one female (No. 1213-8, necropsied on the 388th day of exposure) which exhibited a papillary carcinoma of the bronchus. Two histologically identical neoplasms caucinocknksis from ashhstos 189 1'ic. 2. (Rat No. 1206-23). Mesothelial fibrosarcoma invading the wall of a bronchus, after amosite exposure. Ueiuatoxylin-eosin stain, X 10. were also seen in 2 females of the eroeidolite group (Nos. 1216-20 and 1216-24, neeropsied on the 360th and 3SSth d;iy of exposure, respectively). Animals ex posed to amosite and eroeidolite, and to a much lesser degree those exposed to chrysotile, had mild to moderate fibrosis associated with pulmonary deposits of asbestos fibers. 190 HKICVES, PURO AM) SMITH Fie. 3. (Hat No. 1206-23). Same specimen as in Fiji. 2, under high magnification, sliow' injt cytologic pleotnorphism and mitotic figures. Hematoxylin-eosin stain, X 100. DISCUSSION Of the 13 malignant neoplasms observed in this study following inhalation of asbestos, one was pulmonary metastatic osteosarcoma in a rat and two were bronchial papillary carcinomas in mice, with a similar carcinoma also occurring CAHCINOGKNESIS FROM ASBESTOS 191 Ibu. 1. (Rut No. 1200-42). Fibrous mesothelioma of the pleura after amosite exposure. 1 [ematoxyliu-eosin stain, X40. in a control mouse. Excluding these tumors as doubtful iu relation to an asbestos etiology, there is a 7-9!f incidence of malignant neoplasia in rats exposed to the inhalation ol finely ground ehrysotile, crocidolite, or amosite. The significance of this finding is threefold. First, the results show that under the conditions of this experiment there was 192 HKKVES. I'UHO AND SMITH Fig. 5. (Hat No. 120fi-5'3). Bronchoahcolor carcinoma after amosite exposure. Hcmatoxylin-oosin stain, X 10. no important dillerence in flic carcinogenic capacities of three asbcstilorm min erals including amosite, which in previous experience (Wagner and Rerry, 1969; Reeves cl cl., 1971, 1972) appeared to be less carcinogenic than crocidolite or chrysotile. Roe (196S) could also obtain pleural and peritoneal mesotheliomas CAHC1NOCENES1S FROM ASBESTOS 193 Fk;. B. (Rat Xo. 1208-55). Squamous cell carcinoma of bronchogenic origin after crocidolite exposure. Hematoxylin-cosin stain. X40. ill mice after subcutaneous injection of crocidolite as well as amosite, although the carcinogenic potential was abolished by solvent extraction; Stanton and Wrench (J972) have produced mesotheliomas after pleural implantation of amosite, croc'dolite, elirysotile, or fibrous glass. Selikolf ct al. (1972a) have called 194 HKKVl.S, WHO AND SMITH attention to the carcinogenicity of amosite in man. In the study reported here, we have obtained the first carcinoma of lung and mesothelioma of pleura after experimental inhalation exposure to amosite. Second, it is remarkable tiiat while the filrogenie capacity of chrysotile in CAKCINOGENESIS FROM ASM'.STOS 195 I'm. 8. (Hat No. 1207-61). Well-differentiated squamous cell carcinoma of the lung after clirysotile exposure. 1 lematoxylin-eosin stain, X40. this experiment was much less than that of crocidolite or amosite, the carcino genic capacities appeared to be about the same. Even though the quantities of the three disseminated dusts were comparable (48.6-50.2 nig/nr'), it appeared that fiber destruction was much more extensive with clirysotile than with crocid olite or amosite (54 vs 864-1105 light microscopically visible fibers per ml cham- 196 HEEVES, I'URO AND SMITH bcr air). Thu sharply reduced fiber concentration of chrysotile is well reflected in the histopathologic data showing marked reduction or near-absence of the fibrotic response in animals exposed to chrysotile. The fact that the carcinogenic response was not comparably reduced suggests either or both of the following postulates. (1) The carcinogenic threshold of chrysotile fibers is orders of mag nitude lower than their fibrogenic threshold. (2) Destruction of the fibrous geometry of chrysotile diminished its fibrogenic potential but did not diminish its carcinogenic potential. Both of these alternatives have highly significant po tential public health implications. Third, this study is remarkable for the finding that among four rodent species exposed to identical aerosols of asbestos and comparable survival record, only rats developed respiratory neoplasia (mice which developed spontaneous bron chi*! cancers are excluded from this consideration). Mesotheliomas were pro duced previously with implanted asbestos fibers in hamsters (Smith et al., 1965), fowls (Peacock and Peacock, 1965), rabbits (Beeves et ah, 1971) as well as rats, but experimental inhalation carcinogenesis from asbestos has thus far been restricted to the rat (Gross et al., 1967; Reeves et al., 1971). In some cases this may have been due to relative efficiencies of pulmonary clearance, and for in stance the unsuitability of hamsters to develop cancers from inhaled asbestos was attributed to a swift and severe fibrotic response which curtailed the life expectancy of the exposed animals (Gross et al., 1967; Reeves et ah, 1971). How ever, guinea pigs showed adequate survival and no carcinogenic response after either asbestos implantation or inhalation. It is also remarkable that after equal asbestos exposure, pulmonary ferruginous bodies were abundant in the guinea pig and rare in the rat. There is now growing evidence to suggest that the formation of ferruginous bodies is the result of a phagocytic process involving several macrophages at a site, which fuse to form giant cells (Suzuki and Clung, 1969; Davis, 1970a). This process results in the coating of fibrous particles with a mucopolysaccharide layer containing hemosiderin granules (Govema and Rosanda, 1972), irrespec tive of whether the phagocytized fiber was asbestos, fibrous glass, or other ce ramic silicate (Davis, 1970b; Botham and Holt, 1971). Ferruginous body for mation is thus perhaps a protective reaction of the organism. In rodents, there appears to be an inverse relation between the frequency of these bodies and the carcinogenic effect attributable to inhaled asbestos. The existence of an immu nologic factor in the causation of ashestosis or asbestos neoplasia has been con sidered (Massey et al., 1971) but serum immunoelectrophoretic studies in as bestos cement workers (Kl-Sewefy and Ilassan, 1971) and pulmonary function studies in asbestos-exposed rabbits with and without administration of the im munosuppressive drug Imuran (Ford, 1971) were thus far inconclusive. Essentially, the carcinogenic entity present in asbestos may be either physical or chemical. The physical hypothesis, according to which the mechanical irrita tion attributable to the embedded fibers or to the ensuing biological response (Oppenheimer eflcct) is the key factor in the etiology of asbestos cancers was not viewed with favor by the early investigators (e.g., Harington, 1965), but re ceived new support from the studies of Stanton and Wrench (1972), who ob CARCINOGENESIS FROM ASBESTOS 197 tained mesotheliomas with fibrous glass if the latter was milled to approach the size range of carcinogenic asbestos fibers. These authors therefore concluded that carcinogenicity was primarily related to the structural shape of these ma terials rather than to physicochemical properties. The chemical hypothesis is espoused by numerous other authors who believe that a chemical or physicochemical factor, associated either with the silicate core of asbestos fibers or with certain adventitious factors, is the ultimate car cinogenic principle. Thus far, the question of adventitious factors has received the most attention. Harington (1962) discovered that certain virgin samples of asbestos contained cyclohexane-extractable oils composed of aromatic hydro carbons, one of which was 3,4-benzpyrene. Furthermore, it was pointed out that besides natural trace constituents which apparently became adsorbed on asbestos fibers during their geological genesis, additional opportunities for con tamination exist during commercial handling. Jute bags (Roe et al., 1966) as well as polythene bags (Commins and Gibbs, 1969) were shown to release aro matic compounds of possible or proven carcinogenic potential. Removal of these contaminants reduced (Harington et al., 1967) or eliminated (Roe et al., 1966) the carcinogenic activity of asbestos, and addition of 3,4-benzpyrene to chrysotile sometimes did and sometimes did not increase its carcinogenic potential (Pott et al., 1972; llylev, 1972). Metals present in asbestos include iron, aluminum, and magnesium as major constituents and chromium, nickel, cobalt, manganese, and others as trace con stituents. Several of the latter are known respiratory carcinogens although it was argued (Stanton and Wrench, 1972) that their levels in asbestos are too small to be biologically significant. However, it was pointed out by Cralley et al. (1967) that during manufacture of asbestos textile products the concentration of carcinogenic metals in asbestos increases substantially through contact with the weaving machinery. Similar increase in heavy metal content of asbestos was also experienced during chamber dissemination as practiced by Gross and DeTreville (1967) and Reeves et al. (1971). Dixon et al. (1970) have attempted to explain the carcinogenic activity of asbestos-associated trace metals on the grounds o| inhibition of benzpyrene hydroxylase, a detoxifying enzyme that presumably protects the organism from the effects of benzpyrene. Gralley (1971) suggested that adsorbed metal on asbestos fibers may become the poles of min iature electrolytic cells, and the ensuing electromotive forces may produce high levels of biologically active cations at localized tissue sites. The in vitro action of asbestos on various model systems including tissue cul ture was studied as a possible adjunct to the understanding of its fibrogenie and carcinogenic properties. Pernis and Castano (1971) have found no cytotoxicity for ehrysotde or erocidolite to mouse peritoneal macrophages under ideal cul ture conditions, and no interruption of cellular lactate biosynthesis was observed (Reck et al., 1971a). However, membrane permeability as measured by eosin uptake was increased after contact with asbestos, which could lead to lysis of the cells in the absence of protective factors (Allison, 1971). Glass fiber, but not glass powder, was shown to have similar effects (Turnock et al., 1971). It was also observed that passage through asbestos filters conferred a growth-inhibitory 198 REEVES, l'UHO AND SMITH property on the medium used in the culture of IIcLa cells (Litterst and Lichtenstein, 1970). Bey and llarington (1971), Miller and llarington (1972), as well as llobock and Klosterkotter (1971) have found higher general cyto toxicity for chrysotile than for the amphiboles; and chrysotile, but not amosite or crocidolite, was found to be a powerful hemolytic agent (MacNab and Har- ington, 1967; Schnitzer and Pundsack, 1970; llarington et al., 1971). That am- phibole dusts (amosite and crocidolite) should nonetheless appear to be more hazardous upon inhalation than chrysotile was explainable on the grounds that the shorter and harsher fibrils of the former have a belter opportunity to pene trate? into the alveoli or pleura while chrysotile, in view of its curved shape, is more likely to get arrested high in the respiratory tract (Robock and Kloster- kotter, 1971). The in vivo solubility of chrysotile was investigated with radioactive tracer techniques by Morgan et al. (1971), detecting migration of certain fractions to the liver. Special vulnerability of the pleural mesothelium to chrysotile was shown by measuring the turnover of tritiated thymidine (Bryks and Bertalanffy, 1971); collagen biosynthesis, measured as rate of proline hydroxylation, was highest in the early stages of tissue contact with asbestos (Davis and Reeves, 1971). ' The toxicity of modified asbestos fibers was investigated only very incom pletely thus far. Beck et ah (1971b) have treated chrysotile with 0.1 x IIC1, leaving a SiO- surface on the fibers, and observed increased acute cytotoxicity as manifested by reduced lactate biosynthesis. Heating an import' specimen of chrysotile (fibrous component about 507, the rest mainly antigorite) to 1000C for 3 hours caused greatly increased acute toxicity after intraperitoncal injection into mice (Jagatic et al., 1967). The heat treatment caused appearance of for- sterite and enstatite detectable by X-ray diffraction, but the acute toxicity may have been due to a volatile component. Selikoff et al. (196S) as well as Berry et al. (1972) have observed that asbestos exposure in man has caused significantly increased mortality from lung cancer only among cigarette smokers. Mortality from mesotheliomas, on the other hand, was apparently independent from cigarette smoking, much less dose-related, and frequently occurred in subjects with little or no asbestosis (Selikoff et al, 1967; Wagner et al, 1971). It is possible that different etiologic entities in the asbestos fiber are responsible for tbe causation of these two diseases. Meso thelioma may be a consequence of mechanical irritation by the fibers or by the fibrous plaques formed in response to the presence of the fibers (Bryson and Biscboff, 1967); pulmonary carcinoma may be related to chemical carcinogens present in or adsorbed on the fibers. More work is needed to throw light on this interesting question. ACKNOWLEDGMENTS Tim authors are indebted to Dr. Anil Hoy-Chowdliury for chemical analysis; to Kenneth H. Mikcl for filler counts; to Duane VanMkcr for operation of the exposure facility; to Leonard Pickett for histological processing; and to Clarence J. Hickey and Walter Dybicki for animal colony maintenance. CARCINOGENESIS FROM ASBESTOS 199 REFERENCES Allison, A. C. (1971). Effects of silica and asbestos on cells in cultnre. In "Inhaled Particles III" (VV. H. Walton, Ed.), pp. 437--445. Unwin Bros. Ltd., Old Woking, Surrey. Bamber, H. A., and Butterwohth, R. (1970). Asbestos hazard from protective clothing. Ann. Occup. Hyg. 13, 77-79. Beck, E. G., Bruch, J., Friedrichs, K. II., IIilscheh, W., and Pott, F. (1971a). Fibrous silicates in animal experiments and cell cuIture-Morphological cell and tissue reactions ac cording to different physical chemical influences. In "Inhaled Particles III" (W. II. Walton, Ed.), pp. 477-487. Umvin Bros. Ltd., Old Woking, Surrey. Beck, E. G., Holt, P. F., and Nashakah, E. T. (1971b). Effects of chrysolite and acidtreated chrysotile on macrophage cultures. Brit. ). Ind. Alnl. 28, 179-185. Berry, G., Nexvhouse, M. L., and Turok, M. (1972). Combined effects of asbestos ex posure and smoking on mortality from lung cancer in factory workers. Lancet U, 476--478. Bey, E., and Harinuton, J. S. (1971). Cytotoxic effects of some mineral dusts on Syrian hamster peritoneal macrophages. ]. Exp. Alcd. 133, 1149-1109. Botham, S. K., and Holt, P. F. (1971). The development of glass-libre bodies in the lungs of guinea pigs. J. Pathol. 103, 149--150. Bryks, S., and Bertalanffy, F. 13. (1971). Cytodynamic reactivity of the mesotheliumplcural reaction to chrysotile asbestos. Arch. Environ. Health 23, 409--172. Bryson, G., and Bischoef, F. (1907). Silicate-induced neoplasms. Progr. E.xp. Tumor Re search 9, 77-104. Burilkov, T., and Miciiailova, L. (1970), Asbestos content of the soil and endemic pleural aslrestosis. Environ. Res. 3, 443--451. Commins, B. T., and Ginns, G. W. (1909). Contaminating organic material in asbestos. Brit. ]. Cancer 23, 358-302. Crai.ley, L. J. (1971). Electromotive phenomenon in metal and mineral particulate ex posures: relevance to exposure to asbestos and occurrence of cancer. Amur. Inil. Ilyg. Assoc. ]. 32, 653-061. Crai.ley, L. J., Keenan, R. G., and Lynch, N. R. (1907). Kxpoaire to meta's in the manu facture of asbestos textile products. Amcr. Ind. Hyg. Assoc. J. 28, 452--101. Cunningham, H. M., and Pontefract, R. (1971). Asbestos fibers in beverages and drink ing water. Nature ( London) 232, 332-333. Davis, H. V., and Reeves, A. L. (1971). Collagen biosynthesis in rat lungs during ex posure to asbestos. Anwr. Ind. Ilyg. Assoc. ]. 32, 590-032. Davis, J. M. G. ( ]t)7(ta). Asbestos dust as a nueleatiou center in (be ealeilieation ol old fibrous tissue lesions, and the possible association of this process to the formation of as bestos bodies. Exp. Mol. Pathol. 12, 133--147. Da vis, J. Nt. G. (19101)). Further obsrrvations on the ultrastrueture and chemistry of the formation of .asbestos bodies. Exp. Mol. Pathol. 13, 346-358. Dixon, J. R., Lowe, D. B., Richards, D. E., Crai.ley, L. J., and Stokincer, II. E. (1970). The role of trace metals in asbestos carcinogenesis: asbestos cancers. Cancer Research 30, 1068-1071. Donna, A. ( 1970). Tumori Spcrimcntali da Amianto di Crisntilo, Crocidohte, e Amosite in Ratio. Med. Lavoro 61, 1-32. Fl-Seweey, A. Z., and IIassan, F. (1971). Immunoelectrophorctic pattern changes in asbestosis. A/m. Occup. Ilyg. 14, 25-28. Ford, J. E. (1971). The effect of pulmonary macrophage suppression in developing as bestosis. Dissertation, Wayne State University, Detroit. Gerber, M. A. ( 1970). Asbestosis and neoplastic disorders of the hematopoietic system. Amcr. /. Clin. Pathol. 53, 204-208. Goee, A. M., and ( Iaensij-:u, 1',. A. (1972). Asbestosis following brief exposure in cig arette filter manufacture. Respiration 29, 83-93. Governa, M., and Rosanda, C. (1972). A histochemical study of the asbestos body coat ing. Brit. J. Ind. Med. 29, 154-159. 200 REEVES, PURO AND SMITH Greknrerg, M. ( 1970). Asbestos release from battery boxes. Aim. Occup. Ilyp. 13, 79-30. Chaiiam, J., A no Graham, R. (1907). Ovarian cancer and asbestos. Environ, Res. 1, 115 128. Gross, P., and DeTrevici.e., II. T. 1*. (1907). Experimental asbestosis. Art It. Environ. Health 15, 038-049. Gross, P., DeTrkmli.e, H. T. I'., Tolker, E. B., Kaschak, M,, and Baryak, M. A. ( 1907). Experimental asbestosis-The development of lung cancer in rats with pulmonary deposits of chrysotile dust. Arch. Environ. Health 15, 343-355. Gross, P., DeTheville, R. T. P., and Haller, M. X. ( 1909). Pulmonary ferruginous bodies in city dwellers: a study of their central fiber. Arch. Environ. Health 19, 180-188. Gross, 1\, Tuma, J., and DeTkeville, K. T. P. (1970). Eibrous dust particles and fer ruginous bodies. Arch. Environ. Health 21, 38-40. IIarincton, J. S. (1902). Occurrence of oils containing .3,4-beu/pyrene and related sub stances in asbestos. Nature (l.ontlon ) 193, 43-45. IIarincton, J. S. ( 1905). Studies of carcinogenesis of asbestos fibers and their natural oils. Aim. N. Y. Acad. Sei. 132, 439-450. IIarincton, J. S., Miller, K., and MacNar, G. t 1971 ). Hemolysis by asbestos. Environ, lie.s. 4, 9.5-117. IIarincton, J. S., Hoe. !'. J. G., and Walters, M. ( 1907). Studies on the inode of action of asbestos as a carcinogen. S. Aftican Med. J. 41, 800--804. Harries, P. G., Mackenzie, E. A. K., Sheers, G., Kemp, J. II., Oliver, T. P., and Wriciit, 1). S. ( 1972). Radiological surrey of men exposed to asbestos in naval dockyards. Brit. J. lnd. Med. 29, 274-279. lllCKlsil, E. E., and Knight, K. E. (1970). Exposure to asbestos during brake maintenance. Ann. Occup. Hyp. 13, 17-21. Jagatic, J., Rursitz, M. E., Godwin, M. G,, and Weiskopf, R. \V. (1907). Tissue re sponse to intraperitoneal asbestos with preliminary report of acute toxicity of heat-treated asbestos in mice. Environ. Res. 1, 217-230. Kanazawa, K., Bihiieck, M. S. G.,-Garter, R. L., and Roe, F. J. C. (1970). Migration of asbestos fibers from subcutaneous injection sites in mice. Brit. ]. Cancer 24, 90-100. Lancer, A. M., Sei.ikoek, I. J., and Sastre, A. (1971). Chrysotile asbestos in the lungs of persons in New York Gity. Arch. Environ. Health 22, 348-301. Litterst, G. I.., and Lichtenstein, E. P. ( 1970). Toxicity of Hel.a cell growtli medium after passage through asbestos filters. Lab. Pract. 19, 1221-1223. MacNar, C., and IIarincton, J. S. (1907). Haemolytic activity of asbestos and other mineral dusts. Nature (London) 214, 522-523. McDonald. J. C., McDonald, A. D., Gums, G. \V,, Siemiatycki, J., and Rossn er, G. E. ( 1971 ). Mortality in the chrysotile asbestos mines and mills of Quebec. Arclt. Environ. Health 22, 077-087. Massey, D., Lemonde, J., and Kournier-Massey, G. (1971). Les facteurs immunologiques dans 1' amiantose. Union Med. Canada 100, 1588-1591. Merlins, R. R. (1971). Talc-treated rice and Japanese stomach cancer. Science 173, 1141 1142. Miller, K., and IIarincton, J. 8. (1972). Some biochemical effects of asbestos on macro phages. Brit. ]. Exp. Pathol. 53, 397-405. Morgan, A., Holmes, A,, and Gold, G. (1971). Studies of the solubility of constituents of chrysotile asbestos in vivo using radioactive tracer techniques. Environ. Res. 4, 558-570. Niciioi.son, \V. J., Magciore, G, J., and Selikoi e, I. J. (1972). Asbestos contamination of parenteral drugs. Science 177, 171-173. Occella, E., and Maddai.on, G. ( 1963). X-ray diffraction characteristics of some types of asbestos in relation to different techniques of comminution. Med. Lavoro 54, 628-636. Peacock, P. R,, and Peacock, A. ( 1965). Asbestos-induced tumors in white Leghorn fow ls. Aim. N. Y. Acad. Sci. 132, 501-503. CARCINOGEN liSIS from asbestos 201 Pernis, B., and Castano, P. (1971). Effect of asbestos on cells in vitro. Med. Lavoro 62, 120-129. Pontefract, R. D., and Cunningham, II. M. (1973). Penetration of asbestos through the digestive tract of rats. Nature (London) 243, 352-353. Pott, F., Hutu, F,, and Friedrichs, K. II. (1972). Tumors of rats after i.p. injection of powdered chrysotile and benzo(a)-pyrene. Zbl. Bakt. Hyg. B 155, 463-469. Reeves, A. L., Puro, H. E., Smith, R. G., and Vorwald, A. J. (1971). Experimental as bestos carcinogenesis. Environ. Res. 4, 496-511. Reeves, A. L., Puno, H. E., Smith, R. G., and Vorwald, A. J. (1972). Fibrogenicity and carcinogenicity of amosite, crocidolite, and chrysotile in animal experiment. Amcr. Ind. Uyg. Assoc. } 33, A 79. Rickards, A. L., and Badami, D. V. (1971). Chrysotile asbestos in urban air. Nature ( London) 234, 93-94. Rorock, K., and Klosterkotter, W. (1971). Biological action of different asbestos dusts with special respect to fibre length and semiconductor properties. In "Inhaled Particles III" (W. H. Walton, Ed.), pp. 465--475. Unwin Bros. Ltd., Old Woking, Surrey. Roe, F. ]. C. (1968). Experimental asbestos carcinogenesis. Food and Cosmet. Toxicol. 6, 566-568. Roe, F. J. C., Walters, M. A., and IIaringtox, J. S. (1966). Tumor initiation by natural and contaminating asbestos oils. Intern. J. Cancer 1, 491--495. Rov-Chowdhury, A. K., Mooney, T. F., and Reeves, A. L. ( 1973). Trace metals in as bestos carcinogenesis. Arch. Environ. Health 20, 253-255. Ryi.ev, L. N. (1972). Morphological lesions in rat lungs induced by intratracheal injection of pure chrysotile asbestos and a chrysotile asbestos-ben/.o()pyrcne mixture. Vop. Oncol. 18, 40-45. Schnitzer, R. J., and Puxdsack, F. L. (1970). Asbestos hemolysis. Environ. Res. 3; 1-13. Selikoff, 1. J., and Hammond, E. C. (1971). Asbestos exposure to coke men operators. ]. Occup. Med. 13, 496-497. Selikoff, I. J., Bader, R. A., Bader, M. K., Ciiurg, J., and Hammond, E. C. ( 1967). As bestos and neoplasia. Amer. ]. Med. 42, 487--496. Selikoff, 1. J., Hammond, E. J., and CnuRt:, J. (1968). Asbestos exposure, smoking and neoplasia. J. Amer. Med. Assoc. 204, 104-110. Selikoff, I. J., Hammond, E. C., and Chukg, J. (1970). Mortality experiences of asbestos insulation workers. In "Proceedings of the International Conference of Pneumoconiosis" ( II. A. Shapiro, Ed.), pp. 180-186. Oxford Univ. Press, Cape Town. Selikoff. I, J., Hammond, E. C., and Ciiurg. J. ( 1972a). Carcinogenicity of amosite as bestos. Arch. Environ. Health 35, 183-186. Selikoff, I. J., Nicholson, W. f., and Lancer, A. M. (1972b). Asbestos air pollution. Arch. Environ. Health 25, 1-13. Smith, W. E., Miller, I,., Ciium;, J., and Selikoff, I. J. ( 1965). Mesotheliomas in ham sters following intraperitoneal injection of asbestos. J. Mt. Sinai Hasp. N. V. 32, 1-8. Speii., S., and Leineweher, J. P. (1969). Asbestos minerals in modern technology. En viron. Res. 2, 166--208. Stanton, M. F., Blackwell, R., and Miller, E. (1969). Experimental pulmonary car cinogenesis with asbestos. Amer. Ind. Ilyg. Assoe. J. 30, 236--244. S i an ion, M. F., and Wrench, C. (1972). Mechanisms of mesothelioma induction with asbestos and fibrous glass. /. Nat. Cancer. Inst. 48, 797-821. Stokinceh, H. E. (Chairman) (1973). Threshold limit values for chemical substances in workroom air. Am, Conf. Govt. Indust. Hygienists, Cincinnati. Suzuki, Y., and Ciiurg, J. ( 1969). Formation of the asbestos body-A comparative study with three types of asbestos. Environ. Res. 3, 107-118. Ti.murell, V. (1970). Characteristics of the International Union Against Cancer Standard Reference Samples of Asbestos. In "Proceedings of the International Conference on Pneu moconiosis" (H. A. Shapiro, Ed.), pp. 28-38. Oxford Univ. Press, Cape Town. 202 REEVES, PURO AND SMITH Timbrell, V., Gilson, J. C., and Webster, I. (1968). UICC standard reference samples of asbestos. Inter. }. Cancer 3, 406-408. Turnock, A. C., Bhyks, S., and Bertalanffy, F. D. (1971). The synthesis of tritiumlabeled asbestos for use in biological research. Environ. Res. 4, 86-94. Wagner, J. C. (1962) Experimental production of mesothelial tumors of the pleura by implantation of dusts in laboratory animals. Nature (London) 196, 180-181. Wagner, J. C., and Berry, G. (1969). Mesotheliomas in rats following inoculation with asbestos. Brit. J. Cancer 23, 567-581. Wagner, J. G., Gilson, J. C., Berry, G., and Timriiell, V. (1971). Epidemiology of asbestos cancers. Brit. Med. Bull. 27, 71-76. Westlake, G. E., Spjut, II. J., and Smith, M. N. (1965). Penetration of colonic mucosa by asbestos particles. Lab. Invest. 14, 2029-2033.