Document RJVErm0vpLDwwZ6Y41vw0QVEk

/?s-ry nisms of >us glass. >7-821 :) UICC -national ation of isbestos. Research Annals 5-506 n tissue. rent Re in rats sh Jour- MiviesonoOcCygdjjhs dPtoed ferencf iiversity Thomas <f silica. -91. South An- 2, 623- Some etiological considerations of fibre carcinogenesis M. F. STANTON 1 r The exogenous agents which contribute to the cause of cancer generally fall into one of three major groups: ionising radiation, chemicals and viruses. There is a wealth of speculation as to how the mem bers of these groups act to induce cancer, but the mechanisms of their action remain unknown. Asbestos, in all its forms, contains chemicals that are carcinogenic under certain conditions: however, it is of particular interest as a carcinogen because it has attributes of two of the above groups. Firstly, the presence of various metallic ions and polycyclic hydrocarbons, that are either inherent or acquired through processing, would seem the best explanation for the carcinogenicity of asbestos. On the other hand, particles that are within the dimensional range of viruses are abundant in all forms of asbestos; and it is conceivable that these submicroscopic particles could act in a fashion similar to that of viruses, whatever that may be. On the other hand, there is reasonably good evi dence that neither of these attributes is related to the carcinogenicity of asbestos. The evidence for this conclusion can be summarised as follows: (1) There is no indication that any of the asbes toses are sufficiently contaminated with known carcinogenic hydrocarbons to account for their carcinogenicity; and rigorous extraction of those hydrocarbons present in asbestos does not affect its carcinogenicity for the pleura of the rat (Wagner et a/., 1970). (2) Variations in the inherent metal content of various types of asbestos are great, yet these various types of asbestos show only slight differences in car cinogenicity (Harington, 1965: Timbrell, 1970; Wagner, 1970; Stanton & Wrench, 1972). 1 The Laboratory of Pathology, National Cancer Institute, Bethesda, Maryland. USA. (3) Finely particulate metallic nickel, stainless steel or non-crystalline silicon dioxide applied to the pleura of the rat are not sufficiently carcinogenic to account for the carcinogenicity of asbestos by mill contamination (Stanton & Wrench, 1972). (4) Reduction of fibre size by the partial pulverisa tion ofasbestos, a process which increases contamina tion by metallic particles and increases the number of submicroscopic fibrils in asbestos, reduces its car cinogenicity (Stanton & Wrench, 1972). (5) Hand-cobbed crocidolite ore, hand-milled without metallic contamination, is equal in carcino genicity to machine-milled crocidolite (Stanton & Wrench, 1972). (6) Non-asbestiform fibres such as fibrous glass are increasingly carcinogenic as they approach the size range of milled asbestos fibres (Stanton & Wrench. 1972). One must therefore consider that it is the structural features of asbestos that may be the critical factor in its carcinogenicity, and it is toward this hypothesis that we have directed our attention. If the structural features of asbestos are important, then it follows that similar fibres, if sufficiently durable, should also induce tumours. On the basis of this reasoning, we have assessed the distribution of particles by size in a variety of fibrous materials and applied these to the pleura of rats for a period of two years. These ex periments are still in progress: only preliminary re sults of part of them are available and interpretations are limited. However, estimates of final tumour in cidences have been made from the data currently at hand, and the various materials have been segregated into four groups which cause high, moderate, low and negligible tumour incidence (Tables 1-4). The analysis of fibre distribution by size is subject to some error, nevertheless it is sufficient to characterise the general distribution of fibres in the materials. - 289 - HWBUI0007126 290 BIOLOGICAL EFFECTS OF ASBESTOS >0.5>2.5 >5 >10 >20 >40 >80 ||;l - 2. 5ai - 5^. -10f. -20u -40 m -B0t> -160.u > 20-40p. >160 -320ft iiiiikUM *M;!I i' f'i * i > 2.5- 5 ft >0.5-2.5 m Fig. 1. Graphic illustration of the categories of size used to classify particles in the test samples. Illustrated particles repre sent mean dimensions (u = run). MATERIALS AND METHODS Various specimens of crocidolite, chrysotiie, fibrous glass, and fibrous aluminium oxide were applied by open thoracotomy to the left pleural sur face of 30, 11- to 14-week-old, female OsborneMendel rats at a single standard 40 mg dose level by a method previously described (Stanton et a!., 1969). The unique aspect of these experiments is that all test materials were applied to small 45 mg fibrous glass pledgets prior to application. The glass pledgets are composed of large-diameter fibrous glass which, when intact, has no apparent carcinogenicity in itself. We use it simply as a convenient and accurate means of uniformly applying the test material to a wide sur face area of the pleura. The test materials are listed in Tables 1-4. The UICC standard reference samples of crocidolite and chrysotiie A have been previously described (Timbrell, 1970). Reduction of particle size was accomplished by grinding in a stainless-steel Table 1. High incidence groups (>40% mesotheliomas). Percentage of mass occupied by fibres in each range of size Diameter y.m > 0.5-2.5 > 2.5-5 >20-40 >10-20 >5-10 > 2.5-5 >0.5-2.5 Crocidolite UICC 4- + 4- a 3 1 5 >20-40 >10-20 > 5-10 >2.5-5 >0.5-2.5 Chrysotiie UICC + + + * 8 4 15 >20-40 >10-20 >5-10 >2.5-5 >0.5-2.5 AAA glass crude fibre + + 4- + <1 1 <1 >20-40 >10-20 AAA glass separated *>10x1 urn" > 5-10 4- + + 4- > 2.5-5 1 > 0.5-2.5 83 >20-40 >10-20 >5-10 >2.5-5 > 0.5-2.5 Aluminium oxide whiskers 4- 4- 4- 4- 1 <1 <1 >5-10 1 <1 7 11 14 1 6 3 1 1 Length ym >10-20 >20-40 >40-80 >80-160 >160-320 11 22 1 3 3 17 11 44 2 1 4 4 11 17 22 753 3. 7 5 13 65 3 17 18 10 23 12 44 1 48 26 1 5 2 12 6 25446 + *$ indicate extent of pleural fibrosis most commonly observed. HWBUI0007127 lies repre- igets are ; which, in itself. :e means vide surire listed samples eviously panicle ess-steel f size 11 4 3 7 65 12 26 6 6 SOME ETIOLOGICAL CONSIDERATIONS OF FIBRE CARCINOGENESIS 291 ball mill to produce partially or fully pulverised are shown the five specimens which produced tumour samples. The crude fibres were stripped by hand incidences greater than 40%. These are the U1CC from hand-cobbed ore specimens, and bundles of standard reference samples ofcrocidolite and chryso- fibres were retained as long as possible without con tile A, two samples of very fine fibrous glass with tamination by extraneous mineral. diameters of 3 pm or less, and the aluminium oxide The fibrous glasses were obtained from both the whiskers. All of these samples are composed almost Owens-Corning Fiberglas Corporation, Toledo, entirely of fibres, and further have in common a pre Ohio and the Johns-Manville Research and Engineer dominance of fibres below 5 um in diameter. The ing Center, Manville, New Jersey. We are particu A1203 fibre# are of particular interest because they larly indebted to the latter institution for the size are totally different from those of asbestos and glass, separation of fibrous glasses, which was carried out both in internal structure and in chemical composi through a series of millings and the sedimentation of tion; yet their size distribution is remarkably like exceptionally fine-diameter glass fibres. All of the that of UICC crocidoiite. However, one-third of the glasses were of the usual borosilicate type, whose fibres are slightly longer and thicker than are the mineral oxide contents have been previously re crocidoiite fibres and, since the density of A1203 is corded (Stanton & Wrench, 1972). greater than asbestos, approximately one-sixth as The non-fibrous aluminium oxide and aluminium many fibrous particles are present. The A1203 fibres oxide whiskers were commercial products obtained are most durable and do not fragment into sub- from the Artech Corporation, Falls Church, Virginia. microscopic particles in the manner of asbestos. These are single crystal fibres that are more than Nevertheless, electron microscope study reveals an 99.5% pure Al203. The method of counting fibres abundance ofsubmicroscopic fibrils of less than 0.1 pm has been described previously (Stanton & Wrench, in diameter with lengths comparable to those of 1972). fibres in the lowest optical range. Samples of the materials, suspended in Formvar, Tables 2 and 3 list the six samples of asbestos and were air-dried on glass slides and photographed at glass that fall in the middle ground of carcinogenicity. 1000 x magnifications. From the photographs, These groups show several outstanding differences in 1000 consecutively counted particles were assigned fibre distribution from those shown in Table 1; both to the 30 ranges of dimension indicated in Figure 1. extremes in the dimensional range of fibres are re Assuming that the panicles in a given range were presented. For example, the two crocidoiite samples normally distributed around the mean size of that show similar decreases in carcinogenicity; however, range, the total mass of all particles could be calcul one is composed almost entirely of long, large- ated; and it is the percentages of the total mass diameter fibre bundles (Table 2); while the other occupied by particles in a given range or size com (Table 3) has nearly half its mass reduced to fibres partment that are presented in Tables 1-4. that are at the very lowest optical range of diameter In the tables, the first entry for each diameter and and length and the rest reduced to particles too small half of the diameters in the second entry in each row to be recognised as fibres by optical standards. This represent particles that are non-fibrous by optical latter fraction is of particular importance, since it standards. The plus figures below the designated consists of particles that are aggregates of submicro specimen indicate the extent of pleural fibrosis most scopic fibrils with diameters below 0.2 pm and with commonly observed in the rats in each experiment. lengths of 5 to 10 pm. Since these very short fibrils of The rats are being observed for two years follow submicroscopic size account for the bulk of this ing application. A necropsy is performed on all material and occur in far greater numbers than in the dead or sick rats, and histologic sections are taken non-puiverised crocidoiite, it follows that their role from the site of treatment and from any other ab in carcinogenesis is probably negligible (Stanton & normal lesion. Wrench, 1972). The distributional array of fibres in the three glass samples of these groups indicates the second dimen CONCLUSIONS sional parameter of carcinogencity. Here it is appar ent that carcinogenicity decreases as more fibres The data are arranged in four tables according to exceed 2.5 pm in diameter and as the fibre length our estimates of mesothelioma incidence. In Table 1 decreases to less than 10 pm. Again, the submicro- Ly I HWBU10007128 292 biological effects of asbestos Table 2. Moderate incidence groups (30-20% mesotheliomas). Percentage of mass occupied by fibres in each range of size scopic fractions of these samples contain an abundance of fibres below 0.2 |xm in diameter, with distributions of length similar to those of fibres present in the lowest optical range. Finally, Table 4 lists those materials that have yielded less than 5% incidence of mesotheliomas, No tumours have developed in rats exposed to either crocidolite or chrysotile pulverised to the extent of reducing all of the optically visible particles to nonfibrous form. Since these excessively milled materials contain all of the elemental components of asbestos and are contaminated by metallic erosion of the mill far more than are the other samples, it follows that neither of these factors is a likely source of asbestos carcinogenicity and that structural integrity of the fibre is essential. The two glass samples in this group again represent both extremes in structure, namely either fibres of great length with diameters greater than IOfimorparticlesthatareshort. thick and essentially non-fibrous by optical standards. In the nonfibrous A1203 experiment, a single tumour has been observed among 30 rats. This tumour is of doubtful significance, but it may reflect the low background incidence of less than 5% induced by partial fragmentation of the vehicle. In summary, the experiments may be compared among single types of material. The results from the seven samples of asbestos indicate that none of the three extremes in fibre distribution yield as high an incidence of mesotheliomas as do the more evenly distributed UICC standard reference samples, Either progressive pulverisation to non-librous form or preservation of the test sample in large bundles of fibres clearly reduces carcinogenicity. In comparing the seven glass samples it is apparent that samples composed over 90% by weight of fibres with diameters of 2.5 jxm or less are the most carcinogenic; and that as length is reduced in these small fibres carcinogenicity is also reduced. Finally, the contrasting results obtained using fibrous and non-fibrous forms of Al203 re-emphasise the importance of structure in carcinogenicity. These exceptionally pure, inert fibres, composed of materials foreign to asbestos and glass, seem to carry the same carcinogenic hazard for the pleura as do those materials, It would seem therefore that carcinogenicity is in HWBUI0007129 l.' 160-320 42 22 10 12 2 20 16 12 14 10 <1 ll flj npared am the of the igh an evenly mples. 3 form dies of oaring :mples h diac; and rcinoasting forms ure in inert |>s and azard SOME etiological considerations of fibre carcinogenesis 293 Table 3. Low incidence groups (20-10% mesotheliomas). Percentage of mass occupied by fibres in each range of size Diameter izm > 0.5-2.5 ) >2.5-5 >20-40 >10-20 >5-10 >2.5-5 >0.5-2.5 Crocidoiite partly piilverised + + +a 54 20 >20-40 >10-20 >5-10 >2.5-5 >0.5-2.5 >20-40 >10-20 >5-10 >2.5-5 > 0.5-2.5 AAA glass partly pi iverised +++ 4 12 >20-40 > 10-20 > 5-10 >2.5-5 > 0.5-2.5 >20-40 >10-20 >5-10 >2.5-5 > 0.5-2.5 - >5-10 17 <1 4 3 Length jim > 10-20 j > 20-40 > 40-80 > 80-160 >160-320 r 9 2 16 32 2 e648 2 2 1 '1 3 % +'s indicate extent of pleural fibrosis most commonly observed. Table 4. Negligible incidence groups (< 5% mesotheliomas). Percentage of mass occupied by fibres in each range of size u.m > 0.5-2 5 > 2.5-5 >20-40 >10-20 > 5-10 > 2.5-5 >0.5-2.5 aCrocidoiite fully pulverised + 26 40 <1 >20-40 >10-20 > 5-10 > 2.5-5 > 0.5-2.5 Chrysotiie fully pulverised + 5 3 <1 >20-40 >10-20 >5-10 >2.5-5 >0.5-2.5 Commercial glass whole fibre + >20-40 >10-20 >5-10 > 2.5-5 > 0.5-2.5 AAA glass separated "<5x3 pm' + 5 32 >20-40 >10-20 > 5-10 >2.5-5 > 0.5-2.5 Aluminium oxide non-fibrous + 5 2 <1 >5-10 14 25 2 <1 6 9 3 10 1 Length txm >10-20 >20-40 20 >40-80 >80-160 >160-320 30 28 5 <1 <1 <1 <1 <1 < 1 29 24 20 14 1 <1 28 31 19 4 2 47 24 21 7 1 G -f's indicate extent of pleural fibrosis most commonly observed. HWBUI0007130 294 BIOLOGICAL EFFECTS OF ASBESTOS some way related to the presence of durable particles of fibrous configuration that are particularly long at or perhaps below the smallest diameter which can be recognised optically; and that the carcinogenicity of these fibres has little relation to their chemical composition or to their potential contaminants. SUMMARY Various structural forms of asbestos, fibrous glass and aluminium oxfde have been tested for carcinogenicity on the pleura of rats. Results from all three materials indicate that carcinogenicity is related primarily to fibrous structure rather than to physicochemical properties. A comparison of the dimensional distribution of fibres in those samples of asbestos and glass producing high and low tumour incidence indicate that carcinogenicity may be related to fibres below 2.5 pm in diameter and between 10 to 80 pm in length. ACKNOWLEDGMENTS We wish to thank Dr Vernon Timbrel! for advice in the analysis of the data; Mrs Constance Wrench and Miss Eliza Miller for their diligence in monitoring the experiments; and the staffs of the Owens-Coming Fiberglas Corporation, the Johns-Manviiie Research and Engineering Center,and the M RC Pneumoconiosis Unit, Penarth, UK, for generously providing many of the materials used. REFERENCES Harington, J. S. (1965) Chemical studies of asbestos. Annals of the New York Academy of Sciences, 132, 3147 Stanton, M. F., Blackwell R. & Miller, E. (1969) Experi mental pulmonary carcinogenesis with asbestos. American Industrial Hygiene Association Journal, 30, 236-244 Stanton, M. F. & Wrench, C. (1972) Mechanisms of me sothelioma induction with asbestos and fibrous glass. Journal of the National Cancer Institute, 48, 797-821 Timbrell, V. (1970) Characteristics of the International Union Against Cancer standard reference samples of asbestos. In: Shapiro, H. A., ed., Pneumoconiosis. Proceedings of the international Conference, Johannes burg, 1969, Cape Town, Oxford University Press, pp. 28-36 Wagner, J. C. (1970) The pathogenesis of tumors follow ing the intrapleural injection of asbestos and silica. Morphology of Experimental Respiratory Carcino genesis, AEC Symposium Monograph Series, 21, Oak Ridge, Tennessee, Oak Ridge National Laboratories, pp.347-358 Wagner, J. C., Berry, G. & Timbrell, V. (1970) Mesothe liomas in rats following the intrapleural inoculation of asbestos. In: Shapiro, H. A., ed., Pneumoconiosis. Proceedings of the International Conference, Johannes burg, 1969, Cape Town, Oxford University Press, pp. 216-219 HWBUI0007131 ASBESTOS EMISSIONS FROM BRAKE DYNAMOMETER TESTS APPENDIX D CALCULATIONS OF COLLECTION EFFICIENCY Lining compoadoa tdilute from labomo.'/ 'nalyiit, % SI02 MgO F23 AIjOj Ifi-JI 17.8 2.3 0.3 Chrftatile udeitos 42.4 H20 3.5 > CaCOj 15.4 Zn Oiguuc Total 3.9 m 100.0 Weu debris eadmate, % Decomposed ubotot Decomposed limestone Zinc maul Inorganic 34.9 86 3.9 49.4 Organic Volatile Uncertain Low volatility Organic collectable Total collectable Collected on filter Collectable material not trapped !; idler 10.3 13.7 I 3.1 12.1-27.8 ol .5*77.2 47 14.5-3(1.2* "This material ; .*.ama!ly on shoe cither, talij'vt. tutor. ^hc*.t. an.! tire. AS-56 HWBUI0007132