Document zQ49XgkK8JpKQzmpyNbY59407

A. K ciences hted to see cases of asly hope the authors, in nal cases in which the day is to establish, as invironment is causing :ss further exposure to me has me perplexed, is among those condi ne. In a morphological ne in any of the cases ue which is formed in is, i.e. these latter reeolar air. To have an ate a barrier between arrier could be a memn noted in the asbesgas transfer would be ization is rarely ob gas transfer would be asbestosis. The fibrohe alveolar capillaries, ir capillary block syne functional disturbmonary ischemia. The logy at later stages in tble, surely there may f breathing as one of e this is an appropriecomes overburdened This stiffens the lung the rhythmic respira i of breathing can be llmer of Bochum, for od in studying early I I S ection VII. A sbestos and N eoplasia: Lxeekicental STUDIES OF CARCINOGENESIS OF ASBESTOS FIBERS AND TH EIR NATURAL OILS* J. S. Harington and F. J. C. Roe Chester Beatty Research Institute, Institute of Cancer Research Royal Cancer Hospital, London, England ! s5 | j The mode of action of asbestos in producing asbestoses and neoplasia j is at present little understood. Pulmonary fibrosis, the main feature of asbestosis, has been ascribed to mechanical or chemical factors, or to a combination of both; the mechanism of carcinogenic action is almost com pletely unknown. Ii Two forms of neoplasia in man are associated with exposure to asbestos: li bronchogenic carcinoma' and diffuse pleural mesothelioma." T he T ypes or A sbestos Involved in Malignancy in Man "Asbestos" is the generic name given to a class of fibrous mineral silicates which may vary considerably in physical and chemical composition. In some studies in the past the term asbestos has been used without proper definition and the type of material used remains unidentified. A priori, the demonstration of differences in the carcinogenic activity between different types of asbestos, either in epidemiological or experimen tal studies, would open the way to a critical investigation of asbestos | carcinogenesis.3 \i Studies of mortality by district of residence^ have suggested that expos \ ure to crocidolite asbestos in the North-Western Cape carries with it a distinct risk of developing malignant disease. In the case of the crocidolite mines in the North-Western Cape, asbestos mining districts showed a three-fold excess of deaths from lung cancer as compared with adjoining districts, and as compared with control districts of similar population and degree of urbanization. This latter comparison is important since in South j Africa, as elsewhere, lung cancer mortality is much lower in rural than urban districts. The excess was seen in both males and females. i Similar figures for cancer mortality in the eastern and north-eastern areas, where chrysotile and amosite asbestos are mined and milled revealed no excess mortality from any form of cancer. In these regions, however, the relative proportion of the population exposed to the asbestos is con siderably reduced by comparison with that in the North-Western Cape. Although the standardized mortality ratios might suggest that exposure to *This investigation has been supported by grants to the Chester Beatty Re search Institute (.Institute of Cancer Research. Royal Cancer Hospital) from the Medical Research Council and the British Empire Cancer Campaign for Research, and by Public Health Service Research Grant No. CA-031S3-08 from the National Cancer Institute, U.S. Public Health Service. 439 ?iIi i! Ii 1 i i i W 440 Annals New \o r k Academy of Sciences crocidolite is a greater cancer hazard for man than exposure to amosite and chrysotile. it would be premature to decide until further investigations of the populations actually exposed have been carried out. For example. Roe and Walters'1recently pointed out that tobacco smoking is so strongly associated with lung cancer that it has to be taken into ac count in all studies of possible associations between the disease and other factors. In Britain there is a 45-fold difference in lung cancer incidence between nonsmokers and heavy cigarette smokers." By comparison the three-fold difference observed by Oettle4 is small. Oettle's work was based on certification of death and involved no match ing with regard to smoking habits nor exposure to many of the other known carcinogens involved in the etiology of lung cancer. The figures therefore, can be regarded as confirmatory only. At least they indicate that even before the risk of mesothelioma became widely recognized, an increased mortality from cancer of the lung was being recorded in the asbestos mining districts of the North-Western Cape. Table l* Production' of Asbestos ( in T housands of Tons) from the North-Western Cafe and Transvaal Areas in South Africa Years 1928-1935 1936-1943 1944-1951 1952-1959 Transvaal amosite crocidolite 35 0.1 118 13 237 48 446 126 North-Western Cape crocidolite 31 49 85 345 From Keep, 1961.7 Oettle also suggested that asbestos might not have been mined on a sufficiently large scale for a sufficiently long time for an excessive lung mortality to be apparent in the Transvaal amosite (and crocidolite) mining areas. However, production figures for the period 1928-1959 (table 1) show that there is little difference between the production of amosite in the North-Eastern Transvaal and crocidolite in the North-Western Cape. Wagner, i^.eggs and Marchand,- in a study of individual cases, observed an association between pleural mesothelioma and exposure, either occupa tional or nonoccupational (that is, in persons living close to asbestos mines or mills i, to crocidolite asbestos in the North-Western Cape. On the other hand, no cases have been observed in populations exposed to amosite only.*-' ftluis-Cremer' has dealt in this monograph with some of the geo Harington & I graphical and environme difference in carcinogens amphibole asbestos. He explains satisfactorily th vaal. These include mine tensity of production a asbestosis. Recently, in been found in workers exp Previ Sarcomata have been p peritoneal injection of as longer (0.5 to 1.0 cm.) t cates, augite and tremolit tions. Augite belongs to tl group of rock-forming 1 (Ca. Mg, Fe2*, A l)2 (S commonly of fibrous nal (OH, F ) 2Wagner, in 1962,13indu dusts directly into the pi in two animals inoculatei one with finely-divided S' to a chrysotile dust cloud developed in rats which h in progress and reported graph,14 show that croci extent, amosite, all indi intrapleural injection. > crocidolite and amosite,13 carcinogenicity in this sp P ossible M odes i At present the followii der consideration: (1) That carcinogenesis associated with the tion or treatment of (2) that it is due to the complexes in asbesto (Z) that it is an "Opper by the prolonged res incapable of being r mees exposure to amosite irther investigations out. ;hat tobacco smoking to be taken into ac he disease and other ing cancer incidence By comparison the ! involved no matchmany of the other cancer. The figures least they indicate idely recognized, an ing recorded in the he North-Western- RICA North-Western Cape crocidolite 31 49 85 345 : been mined on a an excessive lung :rocidolite> mining 18-1959 (TABLE 1) tion of amosite in rth-Western Cape, ual cases, obsen-ed ure, either occupa i to asbestos mines 'ape. On the other cposed to amosite h some of the geo- H arington & Roe: Asbestos F ibers and T heir Oils 441 j graphical and environmental considerations which apply to the apparent | difference in carcinogenicity in man of these two types of closely-related j amphibole asbestos. He has reported that none of the factors examined ! explains satisfactorily the absence of pleural mesothelioma in the Trans vaal. These include mineralogical differences between the two areas, in tensity of production and the extent of environmental pollution and asbestosis. Recently, in North America, priman.- peritoneal tumors have been found in workers exposed to chrysotile."-1" P reviocs E xperimental Studies Sarcomata have been produced in rats after the subcutaneous and intra peritoneal injection of asbestos of an undisclosed type." The fibers were longer (0.5 to 1.0 cm.. than those used in other studies. Two other sili cates, augite and tremolite, failed to produce tumors under the same condi tions. Augite belongs to the pyroxene group of minerals, the most important group of rock-forming ferromagnesian silicates,15 and has the formula (Ca, Mg, Fe2~, Al)i> (Si, A lls 0<;. Tremolite is an amphibole mineral, commonly of fibrous nature and with the formula Ca2 Mg-, [Siis O22] (OH, F)2- Wagner, in 1962,13induced pleural mesotheliomata by inoculating various dusts directly into the pleural cavity of rats. Pleural tumors were found in two animals inoculated with crocidolite fibers, one with chrysotile, and one with finely-divided silica powder (99.9 per cent SiC>2). A rat exposed to a chrysotile dust cloud also developed a pleural mesothelioma. No tumors developed in rats which had received carbon black. Experiments at present in progress and reported by Skidmore and Wagner elsewhere in this mono graph,11 show that crocidolite, chrysotile, silica powder and to a lesser extent, amosite, all induce pleural mesotheliomata in guinea pigs after intrapleural injection. Malignant tumors have been produced in fowls by crocidolite and amosite.13 though it is too early to determine their relative carcinogenicity in this species. P ossible M odes of A ction- of A sbestos as a C arcinogen- At present the following possible mechanisms of carcinogenesis are un der consideration: (1) That carcinogenesis is due entirely, or in part, to organic materials associated with the fiber, either naturally or as a result of contamina tion or treatment of the asbestos during processing; (2) that it is due to the presence of certain carcinogenic metals or metalcomplexes in asbestos; (3) that it is an "Oppenheimer Effect," that is to say, that it is induced by the prolonged residence in the tissues of chemically-inert material incapable of being rapidly removed by phagocytosis. AMpRpp m 442 Annals New York Academy of Sciences Oils and Other Organic Material Associated w ith Asbestos Three types of asbestos are used industrially: erocidolite, amosite and chrjsotile. The first two of these, but not the third, contain appreciable amounts of natural oils which may be removed to a large extent by extrac tion with suitable solvents.1"' CTocidolite, amosite and chrysotile may also contain secondary oils as a result of the addition of oil to the asbestos during processing, or as a result of contamination. Asbestos is normally transported and stored in jute bags and the oil from the jute fibers is absorbed to an appreciable extent *70 to 80 per cent) by all three types of fiber. Jute bags are almost always used in the industry, from the point where the fiber leaves the mill to the point where the manufacturer uses it tor his own purposes. In processing, rebagging may be done and there paper-lined bags or plastic may be used, the latter being costlv. An account of the natural (primary) and secondary oils which may be found in asbestos has been given by Harington elsewhere in this mono graph.17 The Saturai Oils As reported elsewhere,17 it is not yet known whether there are any quan titative or qualitative differences between erocidolite and amosite oils, nor whether these oils vary in yield and composition from batch to batch, or within the same sample of fiber. On the other hand, there seems little doubt that finely-divided asbestos yields more oil than the longer fibers in milled material. Large amounts of erocidolite and amosite oils (150 to 200 ml.) have been extracted from 45 kg. amounts of mixtures of both finely-divided and com mercial fiber, using double-distiiied petroleum ether. It is important to record that the solvent used for these extractions was of a high degree of piiritv. Ultra-violet and infra-red spectra indicated that it was apparently free of contaminating materials and no residue was detected when a twoliter sample was evaporated to dryness. The yields for both oils were approximately 0.3 per cent by weight of the crude asbestos fiber before extraction. The two oils differed in color and in infra-red spectra, and more poly cyclic hydrocarbons were detected in amosite than in erocidolite. A sample of tne erocidolite oil was found to contain 1 /xg. benzolajpyrene, 100 g. fiber and the amosite, 5 /xg. benzo a 1pyrene 100 g. fiber. Details of these analyses are given by Harington elsewhere in this Annal.17 The concentrations of benzo(a ;pyrene and related polycyclic hydrocar bons are so low that it is difficult to believe that their presence plays more than a minor carcinogenic role, even in persons exposed most heavily to erocidolite or amosite dusts. The elution by serum of benzofa] pyrene and such hydrocarbons from asbestos is probably equally insignificant. Harington B In order to assess t extractable material, (a) Tests of crude < taneous tissues c These tests are ba; at the sites of subci two species. If this . carcinogenicity of thi before extraction, wo in carcinogenesis by a In controlled exper suspensions of (1) tt (2) the fibers of cro wdth 8 different solve Tests, using BALE those using CBA mic died from intercurre visible at the injectii asbestos fibers, especi cytes. Others were sti area contained asbesti not detected by ordina not been seen in relati BALE, C mice, but is especially in the grouj erocidolite, where fibr< between groups treat have been no injection however, that the exp* (b) Tests of a samph cer had been repc After the associate had been recognized, i of lung cancer in pers tory in the United Sta tion a large con.signmi had been exposed. Sai and biological tests, bo In the biological te.native amosite fiber a> t A sbestos te, amosite and ain appreciable tent by extracchrysotile may to the asbestos :os is normally jute fibers is three types of 'rom the point acturer uses it one and there y. which may be in this mono- are any quanosite oils, nor 1 to batch, or i seems little iger fibers in I.) have been ded and cornimportant to gh degree of s apparently when a twoth oils were fiber before I more polye. A sample rene 100 g. lils of these c hvdrocarplays more heavily to pyrene and ant. H arington & Roe: Asbestos Fibers and T heir Oils 443 Biological Testa for Carcinogenicity In order to assess the carcinogenic role of natural asbestos oils and other extractable material, the following experiments have been undertaken: <a ) Testa of crude and extracted asbestos for carcinogenicity in subcu taneous tissues of rata and mice. These tests are based on the assumption that tumors are likely to arise at the sites of subcutaneous injection of native asbestos fibers in these two species. If this assumption is justified, a comparison of the relative carcinogenicity of the exhaustively-extracted fibers with that of the fibers before extraction, would help to define the role of the extractable material in carcinogenesis by asbestos. In controlled experiments, rats and mice have been injected with saline suspensions of (1; the native fibers of crocidolite, amosite and chrysotile; (2) the fibers of crocidolite and amosite following successive extraction with 8 different solvents (see reference 17 for procedure). Tests, using BALE C mice, have been in progress for 10 months, and those using CBA mice, for only 15 weeks. In BALB C mice which have died from intercurrent disease, sizeable deposits of asbestos have been visible at the injection sites in all groups. Microscopically, some of the asbestos fibers, especially the smaller ones, were seen to lie within phago cytes. Others were still extracellular. Lymph glands draining the injection area contained asbestos-laden phagocytes. Similarly-laden phagocytes were not detected by ordinary microscopic methods in other tissues. Fibrosis has not been seen in relation to the injection sites in any of the asbestos-treated BALB/C mice, but is already apparent on palpation in some of the CBAs, especially in the groups treated with chrysotile. In the cases of amosite and crocidolite, where fibrosis is less frequent, there is no difference in incidence between groups treated with crude and extracted samples. So far, there have been no injection-site tumors in either mouse strain. It is emphasized, however, that the experiments are incomplete. (b) Tests of a sample of amosite from a factory where cases of lung can cer had been reported. After the association between exposure to asbestos and mesothelioma5 had been recognized, attention was drawn to the occurrence of four cases of lung cancer in persons apparently exposed to amosite asbestos in a fac tory in the United States." By good fortune there remained for investiga tion a large consignment of the actual asbestos fiber to which the patients had been exposed. Samples of this fiber have been the subject of chemical and biological tests, both in the laboratory of Smith1"and in our laboratory. In the biological tests undertaken by us, the carcinogenic activity of (1' native amosite fiber and (2 \ amosite extracted with a series of solvents15 I P i f 4 i \ 444 Annals Xevr York Academy of Sciences are being compared. Rats have been injected intrapleurally with 25-mg. amounts of each sample. The experiment is now in the 10th week. (c) Tests for carcinogenicity and cocarcinogenicity on mouse skin. ( I t Asbestos Oils. Experiments have been started in which the oils ob tained by mass extraction of crocidolite and amosite with petroleum ether are applied repeatedly to the skins of mice. These experiments are still only at a very early stage. In preliminary tests the oils (either undiluted or as 50 per cent or 25 per cent dilutions in acetone i were found to be non irritant to mouse skin, causing no apparent epidermal hyperplasia and no dermal changes. Carcinogens of the polycyclic hydrocarbon type produce characteristic changes in mouse skin when applied in high concentra tion.1"-1 The f act that these changes were not seen in our preliminary tests indicates that compounds of this type are probably not present in the oils in high concentration, and this has been confirmed by chemical analysis.17 For this reason a strong carcinogenic response is not to be expected. Though not strongly carcinogenic, it is possible that the asbestos oils possess incomplete carcinogenic activity of either the tumor-initiating or tumor-prorooting type. The lack of strong irritant activity is unlikely to be associated with potent promoting activity. But the possibility that the oils are effective initiators of the carcinogenic process remains entirely open. Our tests for incomplete carcinogenic activity consist of (a) a re peated application of the oils to mouse skin following a single initiating dose of 7,12-dimethylbenz[a]anthracene (DMBA) and (b) a short course of applications of the oils followed by repeated applications of croton oil. (2 t Jute Oils. Heavy contamination of asbestos with jute oils would only be of significance in relation to carcinogenesis if the oils themselves were shown to have relevant biological activity. Accordingly, samples of each of two types of oil used in jute manufacture were tested for carcinogenicity and tumor-promoting activity on mouse skin. Groups of 20 male mice of the Chester Beatty Stock Strain were clipped and treated as follows: Group 1 : 150 g. 7,12-dimethyIbenzfajanthracene (DMBA) in acetone once when the mice were 6 weeks old, followed, after a five-week treatment-free interval, by twice-weekly applications of undi luted light mineral oil for 20 weeks. Group 2 : Same as Group 1 except that "soluble" oil was given as the second treatment. Group 3: received DMBA in the same way as Groups 1 and 2, but no secondary treatment. Groups i and 5: received no primary treatment with DMBA, but from the age of 11 weeks onwards they were treated with light mineral oil and "soluble" oil, respectively, in the same way as Groups 1 and 2. The 150 /jr.g. dose of DMBA given to Groups 1, 2 and 3 was intended to be sufficient to initiate tumor formation without completing the process. In fact, during an observation period of 45 weeks after application of DMBA, only one skin papilloma arose in the mice of Group 3 which re- Harington & ] ces 3C 3 S- w 5 ' = w XS *rt+r e 3* c*!r 55' fsD 3 S ' " af&O 3o 3p u. 92s r+ 3*3 .. * 3(V *t g t--t " a < <& sf <n & s pcr* w&rt> "sao (T: o h , ^^ o -3 H-i r+ W tr >R a o_ o ' u> ns cPrfr- o<3rrt> ^*f&3at O3 0a9> 3 7j. (sti jS ro 35 C o? a < 3 : *" 3 7> W 75 2 o S 3` wsr 0^9- O o w. r3f- za. Co ' n 2 ; ^ 8 B. 3rt> 5 ckj ' ji 3- *3 ^ O (*"t wI- 5. P3 O- h-, -- 7> 575' n> 3 r+ 0s5f i- p fPtT 3 ,*Ha<--> 3'fZt^, '< <O- 09 O>1 ? P f7ctir ^ $$ *^Oofroot- ft g. 8 05 305 *-f *0 i 3 3- 3 =~ 039- o O cftr O3 2. 075) tT 7) ?>3-t" Jj _ TI-3 Oa* (Cn0 6 0^) oc 0V5I 3Q, r-f 03 0 ,05 a3 , O3M fctr O3 3 2rc*tr<0 c 3CL fCCr~tlf- fV*1t 3O' 37r1+ 5 Tfci *o1 ft *f01t5 Cf-/fls C 3 O<--*' *c(0ir cr oc*r rc+r(0 o 3/1 O 7 2 CSO c c3i* Or- cr f0 X ro O' 3 1X5 j H arington & Roe: Asbestos Fibers and T heir Oils TahIjK 2 j TKSTH KOB CAItCINOCKNtCITY AND TUMC>K-I'K()M0TIN( ACTIVITY OF Jl'TIO OlI.S i Number Primary Secondary treatment Tumor development after 20 weeks of secondary treatment. Group of treatment (0.25 ml X 2 weekly) mico (once only) for 20 weeks .t Surviving Mice with Total Total -i mice skin tumors papillomata carcinomata 1 20 150 /ig- DMBA* Undiluted light material oil 20 used In jute manufacture. 13 118 8 2 20 150 Mg- DMBA* Undiluted "soluble" oil 18 used In jute manufacture. 1 1 0 1 3 20 150 Mg. DMBA* NONE 19 1 1 0 ; 4 20 NONE Undiluted light mineral oil 17 3. 4 0 used In jute manufacture. 5 20 NONE Undiluted "soluble" oil used In jute manufacture. 18 0 0 0 *7 ,12-Dimothylbonz[ a] anthracene. t Secondary treatment was begun 5 weeks after application of DMBA. 8Trtni M 1ISS&&PPP^ mm 445 mm 446 Annals New York Academv of Sciences ceived no other treatment. Mice of Group 1. on the other hand, developed 118 benign and 8 malignant skin tumors during the 20-week period of secondary treatment with light mineral oil after DMBA. This observation indicates that the sample of oil used has considerable tumor-promoting activity. The "soluble" oil showed no such activity. Groups 4 and 5 represented tests for complete carcinogenicity of the two oils. By the 20th week of the experiment three out of 17 survivors of Group 4, treated with light mineral oil only, had developed a total of four benign papillomata. Mice of Group 5 were still without skin tumors at this time. Although no malignant tumors were seen, these results indicate that the light mineral oil probably possesses weak carcinogenic activity. Taken together, these findings, which are summarized in TABLE 2, sug gest that the storage of asbestos in jute bags may introduce a significant variable into subsequent determinations of its carcinogenic activity. Metal Carcinogenesis and A sbestos Many metals have been shown to be capable of inducing cancer.-- Among the more notable ones are arsenic, chromium, nickel, cobalt, cadmium, lead and beryllium. In addition, several other metals, though not apparently carcinogenic in the simple ionic state, became so when combined in certain macromolecular complexes. The best-known example of this is iron, which as ferrous sulfate or gluconate shows no carcinogenic activity but which, in the forms of iron-dextran, iron-dextrin, saccharated iron oxide or fer rous glutamate, is markedly active.'-'' Aluminum is another example in so far as aluminum-dextran is carcinogenic.'1 Iron is in fact one of the more interesting metals in relation to asbestos <tide infra ). Hematite, which contains 70 per cent of iron, all in the ferric state, has been suspected of inducing cancer of the lung in man.-4 How ever, the basis on which this suspicion was founded has recently been questioned.-' Wagner1' reported the induction of pleural mesotheliomata in rats injected with silica powder and this has been confirmed.14 A sample of the powder used in these experiments was examined for iron and a value of approximately 0.06 per cent was obtained. Of this, 0.04 per cent was in the form of ferric oxide tFe^O.-.j and 0.02 per cent as ferrous oxide (FeOj , giving a Fe'1* : Fe2- ratio of 1: 0.6. The values for iron are low and do not support the suggestion that the iron content of the silica powder plays the same kind of role in carcinogenesis as it does in iron-dextran. On the other hand, it is possible that the presence of iron was essentia! to the carcino genic activity of the silica. Small amounts of iron are known to catalyze certain reactions of biolog ical i n t e r e s t .- Silica powder not freed of its associated metals (0.6 mg. iron/ gm. silica powderand 0.0016 mg. copper gm. powder and trace amounts of aluminum and other metals/" totally oxidized glutathione in in vitro H arin g to i conditions, but silica ing had no effect." present in the silica, ferrous iron are cap: Against the backgj sider the metal cont The amphibole as nature and are com] sodium in varying sodium ferroso-ferr Mg'- * )a (.Fer,+). Si> 40 per cent iron (a rous forms. The ra Amosite, which exclusively in Sout tains the same amoi form. The Fe34" : F Fe-T) 7 Sk Ose (OH] The serpentine as nature and of formi and 40 per cent ma; iron (2.5 per cent as magnesium in the stt Fe-~ ratio is 1:1.3. With regard to the proved to be hardly 7.4 the following amc 1 tile 6.6 mg., silica po' Chrysotile totally de oxidation. Quite apart from ferric to ferrous iron that carcinogenicity (see TABLES'. In oi large number of min are being tested in m Two other interesl chromium since it is ? cer.-' r' Both may oc< Ni and 0.1 per cent Cr To summarize the \ of crocifiolite and am tain macromolecular iences other hand, developed he 20-week period of BA. This observation ible tumor-promoting inogenicity of the two 17 survivors of Group a total of four benign 1 tumors at this time, ults indicate that the ic activity. ized in table 2, sugitroduce a significant ogenic activity. TOS cing cancer.-- Among cobalt, cadmium, lead sough not apparently n combined in certain of this is iron, which c activity but which, ed iron oxide or fernother example in so n relation to asbestos iron, all in the ferric lung in man.J4 Howd has recently been sural mesotheliomata onfirmed.14 A sample 1 for iron and a value 0.04 per cent was in ferrous oxide (FeO), m are low and do not ica powder plays the extran. On the other ;ntial to the carcino- l reactions of biologated metals (0.6 mg. er and trace amounts itathione in in vitro -l--'f-^- f ` 1*!nJ<1 t ~TiF`-~rfgifi1'ii^ --''itji i^nffliritTSini^~T<liilfiir' "V ' -- , H arington & Roe: Asbestos Fibers and T heir Oils 417 conditions, but silica with a greatly reduced metal content after acid-wash ing had no effect.7' The oxidation process appears to be due to the iron pre.-ent in the silica, since similarly low concentrations of either ferric and ferrous iron are capable of oxidizing glutathione to the same extent. Against the background of these considerations it is interesting to con sider the metal contents of the three types of asbestos. The amphibole asbestos types, crocidolite and amosite. are fibrous in nature and are complex silicates containing magnesium, iron, calcium and sodium in varying proportions. Crocidolite ('`Cape blue asbestos") is a sodium ferroso-ferrisilicate with a composition approximating to Xa2(Fe2-, Mg ).: (Fe k, Sh O-jj (OH and consists of about 50 per cent silica and 40 per cent iron (as oxides >, fairly equally distributed as ferric and fer rous forms. The ratio of Fe; ~ : F e-- is 1: 1.4. Amosite, which is closely related to crocidolite chemically, is found exclusively in South Africa. It is a ferrous magnesium silicate and con tains the same amount of total iron as crocidolite but mostly in the ferrous form. The Fe- : Fe2T ratio is 1:91. The formula approximates to (Mg-'-, Fe-'' )tSis O - (O H )2. The serpentine asbestos, chrysotile, is a magnesium silicate, fibrous in nature and of formula Mg.-. Sic O.-, (OH >4. It consists of 40 per cent silica and 40 per cent magnesium (as oxides) and relatively small amounts of iron (2.5 per cent as oxides*, made up of ferrous or ferric iron replacing magnesium in the structure, and partly of magnetite impurities. The Fe3* : Fe-+ ratio is 1:1.3. With legard to the oxidation of glutathione, the three types of asbestos proved to be hardly less effective than unwashed silica powder." At pH 7.4 the following amounts of the dust oxidized 100 pig. glutathione: chryso tile 6.6 mg., silica powder 50 mg., crocidolite 66.6 mg. and amosite 133 mg. Chrysotile totally destroyed glutathione, the reaction proceeding beyond oxidation. Quite apart from the absolute amounts of iron present, the ratio of ferric to ferrous iron may be important. Such data as is available suggests that carcinogenicity may be associated with a loir ferric: ferrous ratio (see table 3 >. In order to determine whether this association is valid, a large number of minerals with known contents of ferric and ferrous iron are being tested in mice by subcutaneous injection. Two other interesting metallic constituents of asbestos are nickel and " urn since it; is beyond dispute that both these metals cause lung can cer. Both may occur in relatively high concentration (e.g. 0.5 per cent Ni and 0.1 per cent Cr) in chrysotile.17 To summarize the position with regard to metals in asbestos: in the cases of crocidolite and amosite the iron content is well in excess of that in cer tain macromoleeular complexes of iron, in particular iron-dextran, known m m lum ps wgsppiBmp W tfohni*>li* i ? MMMttUK Sslitt 44S Annals New York Academy of Sciences Table 3 P ossible Relation of the R atio of F erric and F errous Iron in Various Minerals with their Carcinogenicity Mineral ,,Fe3+ : F^e 2-r Carcinogenicity Reference crocidolite chrysotile silica 1:1.4 1:1.3 1:0.6 + + 2.4.13- 14 4.13- 14,19 13-14 amosite 1:91 4,13-14,19 augite tremolite anthophyllite 1:4* 1:8.6* 1:1.7* 1:9.6* | unknown 11 11 * mean of three values for Finnish anthophyllite, determined here. * values from reference 12: mean of eleven values for anthophyllite, of ten for augite and of five for tremolite. All other analyses from reference 17. to be highly carcinogenic in animals. The iron content of chrysotile, though much lower, may still be sufficient to account for carcinogenesis by a mechanism comparable to that of iron-dextran. Silica powder with a much lower content of iron has been shown to be carcinogenic in animals; in this case iron may be playing a different, though nevertheless essential, part in the carcinogenic process. The second mechanism may apply to asbestos also. Finally, the possibility that the ferric: ferrous ratio rather than the absolute content of the iron is important in relation to carcinogenesis re quires further study. Of the other metals in asbestos, the high nickel and chromium contents of chrysotile deserve the most attention. Experiments in Progress Clearly these theoretical considerations provide the basis for many criti cal animal experiments. Because of limitation of facilities a start to these has only recently been made. In the first experiment the carcinogenicity of (1) native crocidolite, amosite and chrysotile asbestos, and (2; acidwashed asbestos17of the same three types, is being compared by the method of subcutaneous injection in mice. Acid washing removes large amounts of metals from asbestos without apparently affecting the structural in tegrity of the fibers, as shown by X-ray diffraction analysis of the fibers. Asbestos Carcinogenesis .as an E xample of T umop.-induction by C hemically I nert Materials In agreement with Schmah!,:: we feel it is unlikely that there is much in common between the mechanism by which asbestos induces cancer and Wr Harington the so-called "Oppenh not be regarded as ch rials studied by the C j essential characteristi i only occurs when larg the same materials ar Recent epidemiolog revitalized the subjec In a parallel paper.1' 1 | discussed. In this papi iI mental methods of e standpoint. The sign, oils purposely or acci constituents, are disci asbestos carcinogenesi We wish to thank 5 us in part of the biolo. help in preparation of vided full support for help and for Professoi 1. Doll, R. 1955. Mor 1 dustr. Med. 12: 8] 2. Wagner, J. C., C. , thelioma and ast Brit. J. Industr. Met I 3. Editorial. 1964. As 4. Oettle, A. G. 1964 Sahara. J. Xat. C; 5. Roe, F. J. C. & M. . cer. Progr. Exp. T 6. Doll, R. & A. B. t observations of Bi 7. Keep, F. E. 1961. A erith Commonwea nesburg, South A 8. Sluis-Cp.emer, G. K and environmenta 9. E n t ic k n a p , J . B. & Brit. J. Industr. M 10. Mancuso, T. F. & studies. Arch. Em 11. Schmahl, Jj . 1953. Ratten. Z. Krebsfc nees ;s Ikon in Various ty Reference 2.4.13- 14 4.13- 14,19 13-14 j 4,13-14,19 11 11 mined here. anthophyllite, of ten -om reference 17. tent of chrysotile, r carcinogenesis by Kin der with a much in animals; in this less essential, part y apply to asbestos tio rather than the carcinogenesis re:he high nickel and ition. tsis for many critiies a start to these he carcinogenicity tos, and (2) acidared by the method ves large amounts the structural in lysis of the fibers. IOR-IXDVCTION bat there is much nduces cancer and H arington & Roe: Asbestos Fibers and T heir Oils 449 the so-called "Oppenheirner Effect." In the first place, native asbestos can not be regarded as chemically inert in the same way as many of the mate rials: studied by the Oppenheimers and their colleagues, and secondiv, the essential characteristic of the Op7,enheimer Effect is that carcinogenesis only occurs when large bodies are implanted -- no effect being seen when the same materials are introduced in powdered form. S ummary Recent epidemiological studies on lung cancer and mesothelioma have revitalized the subject of the mechanism of carcinogenesis by asbestos. In a parallel paper.1' the chemistry of the three main types of asbestos is discussed. In this paper possible mechanisms of carcinogenesis, and experi mental methods of elucidating them, are considered from a biological standpoint. The significance of oils naturally present in asbestos and of oils purposely or accidentally added to it, and the possible role of metal If constituents, are discussed in some detail. It is regarded as unlikely that asbestos carcinogenesis is an example of the Oppenheimer Effect. i A cknowledgments We wish to thank Mrs. M. A. Walters, B.Sc., who is collaborating with us in part of the biological work and Mrs. K. Foster and Miss M. Butt for help in preparation of the manuscript. The Anna Fuller Fund kindly pro vided full support for one of us (J.S.H.). We are most grateful for this help and for Professor A. Haddow's interest and encouragement. R eferences 1. Doll, R. 1955. Mortality from lung cancer in asbestos workers. Brit. J. Industr. Med. 12: 81. 2. Wagner, J. C., C. A. Sleggs & P. Marchand. 1960. Diffuse pleural meso i thelioma and asbestos exposure in the North Western Cape Province. Brit. J. Industr. Med. 17: 260. 3. Editorlal. 1964. Asbestos and malignancy. Brit. Med. J. 2: 202. 4. Oettle, A. G. 1964. Cancer in Africa, especiallv in regions south of the Sahara. J. Nat. Cancer Inst. 33:383. 5. Roe, F. J. C. & M. A. B alters. 1965. Some unsolved problems in lung can cer. Progr. Exp. Tumor Res. 6 (in press.) 6. Doll, R. & A. B. Hill. 1964. Mortality in relation to smoking: ten years' observations of British doctors. Brit. Med. J. 1 :1399. "" I 7. Keep, F. E. 1961. Amphibole asbestos in the Lnion of South Africa. In Sev enth Commonwealth Mining and Metallurgical Congress. : 1-28. Johan nesburg, South Africa. " 8. Sluis-Cremer, G. K. 1965. Asbestosis in South Africa: certain geographical and environmental considerations. This Annal. 9. Extickxap, J. B. & W. J. S mither. 1964. Peritoneal tumours in asbestosis. Brit. J. Industr. Med. 21: 20. 10. Mancuso. T. F. & E. J. Coulter. 1963. Methodology in industrial health studies. Arch. Environm. Hith. 6: 210. 11. Schmahl, D.195S. Cancerogene Wirkung von Asbest bei Implantation an Ratten. Z. Krebsforsch. 62: 561. i im'tVtwii 450 Annals Xew York Academy of Sciences 12. Deer, W. A., R. A. Howie & J. Zussman. 1963. Rock-forming Minerals. 2. Chain Silicates. Longmans. London. England. 13. Wagner. J. C. 1962. Experimental production of mesothelial tumours of the pleura by implantation of dusts in laboratory animals. Nature 196: 180. 14. Skidmore. J. W. & J. C. Wagner. 1965. Asbestos dust deposition and re tention in animals. This Annal. 15.. P eacock, A. & P. R. Peacock. 1964. Asbestos-induced tumours in fowls. Abstracts oth Ann. Meeting British Assoc, for Cancer Res. Edinburgh, Scotland. 16. Harington, J. S. 1962. Natural occurrence of oils containing 3:4-benzpy- rene and related substances in asbestos. Nature 193: 43. 17. Harington, J. S. 1965. Chemical studies of asbestos. This Annal. 18. Selikoff. I. J. 1964. Personal communication. 19. Smith, W. E. & L. Miller. 1965. Tests for carcinogenicity of asbestos. This Annal. 20. P vllinger, B. D. 1940. The first effects on mouse skin of some polycyclic hydrocarbons. J. Path. Bact. 50: 463. 21. Svntzeff, V., E. V. Cgwdp.y & A. Croninger. 1955. Microscopic visualiza- tion of the degeneration of sebaceous glands caused by carcinogens. Can- cer Res. 15: 637. 22. Roe, F. J. C. & M. C. Lancaster. 1964. Natural, metallic and other sub- stances as carcinogens. Brit. Med. Bull. 20: 127. 23. Haddow, A., C. E. Dukes & B. C. V. Mitchley. 1961. Carcinogenicity of iron preparations and metal-carbohydrate complexes. 39th Ann. Rep. British Empire Cancer Campaign (2) : 74; Richmond, H. G. 1959. In- duction of sarcoma in the rat by iron-dextran complex. Brit. Med. J. 1: 947: Haddow, A. & E. S. Horning. 1960. On the carcinogenicity of an iron-dextran complex. J. Nat. Cancer Inst. 24: 109. 24. Faulds, J. S. & M. J. Stewart. 1956. Carcinoma of the lung in haematite workers. J. Path. Bact. 72: 353. 25. Muller, E. & W. E rhardt. 1956. Experimenteller Beitrag zur Frage der Cancerogenitat von Eisenoxydstaub. Z. Krebsforsch. 61: 65. 26. Schofield, P. J., B. J. Ralph & J. H. Green. 1964. Mechanisms of hydrox- yiation of aromatics on silica surfaces. J. Phys. Chem. 68: 472. 27. Nutt, A. & J. S. Harington. 1964. The reaction of reduced glutathione with quartz powder and with associated iron and copper. La Med. del Lavoro 55: 176. 28. Doll, R. 1958. Cancer of the lung and nose in nickel workers. Brit. J. Ind. Med. 15: 217. 29. Grushko, M. Y. 1961. Chromium as a carcinogenic substance. Probl. Oncol. 7: 116. - IHUtt I PE Healt , ' O, The word " mineral silica! j | properties am i mechanical me , Asbestos is i } and amphiboli 3 anthophyllite, i j tinolite. The s mercially, sine : asbestos. The < ; i 1 purposes in ir been discussed ; In studying ' ; tors to considei ! in the molecula * The texture ' harshu When theoretical qua As the quantity- increases to a ] needle.?.3-* Impi produce semiha able in asbestos growth is poor! between the fibe Amphibole fii molecular watei also affect the te; To evaluate b must be known, minerals are pr This work wa CA 07681 from tl