Document 50J6Jjr1yVdx5e04B4nM7qJv0

PAL XKK ET A L l /yj / ,( / m t '/* * 'b 343 1 the development of emphysema. It is possible that a period long. ? ten weeks reported above may . cythemia. This as well as the de- of cor pulmonale is under investi- his laboratory. .1 Summary mature rats receiving intratraion of papain and with constrict; developed lesions in the lung hose reported by Gross et al in essential features of the expert, compared to the normal controls ows: (1) elevated functional re. icity, (2) elevated pulmonary 3) reduced pulmonary compli. evated blood carbon dioxide ten>) elevated percent of air space :al sections. t was supported by Public Health HE-09532-02 from the National Heart 1 Cowdrey. MD. formerly of the De- 'athology Research, St Luke's Hoa rd. and Paul Gross, MD, of the ;iene Foundation, Pittsburgh, helped ..C.: Measurement of Functional Re in the Rat J Appl Physiol 21:233- and Whittenberger, J.L.: Physical luman Lung Measured During Sponration, J Appl Physiol 5:779-796 '.H.; Munn, J.I.; and Furth, F.W.: a Method for Clinical Hemoglobimed Forces Med J 5:693-703 (May) D.L.: The Standardization of Hemoaaent Amer J Med Sci 277:710-711, G., et al: Blood 0: Content Using ctrode, J Appl Physiol 20:1063-10G9 Vf.S.: Quantitative Methods in the nary Pathology, Thorax 17:329-333 I I ( I tion, the first major cultural s going on about us are m n generations of our ances--Coon, C.S.: The Story of id, ed 2, Ne\ York: Alfred ! i * i l PLAINTIFF'S EXHIBIT DOW-1524 Experimental Asbestosis The Development of Lung Cancer in Rats With Pulmonary Deposits of Chrysotile Asbestos Dust PuRING a study of experimental asbes Materials and Methods tosis involving guinea pigs, rats, and ham Canadian chrysotile asbestos was ball-milled sters that had inhaled chrysotile dust, it was and subsequently hammer-milled, using essen found that pulmonary cancers had devel- tially the method of Holt et al.* The principle ojted in several rats. Spontaneous primary, of this technique is to reflux the comminuted lung cancers, exclusive of lymphoblastoma(a, have been exceedingly rare in our rats we have not seen a single spontaneous pri chrysotile through the high-speed hammer-mill so that .auch of the dust is of ultramicrascopic dimensions (Fig 1). Within less than one year, it became necessary to replace the steel-alloy mary' pulmonary carcinoma or fibrosarcoma hammers with new ones because of excessive in about 10,000 rats examined). Therefore, wear. Also, to maximize the dust concentration a cause-and-effect relationship between the in the chamber, the output of a second ham pulmonary deposition of chrysotile dust and mer-mill was added to the inhalation chamber. tumor development was suspected. Suspi cion increased as the number of malignant lung tumors discovered in rats that died more than 16 months after initiation of the chrysotile exposure increased rapidly. There has been increasing evidence, mostly epidemiologic, linking long-term ex posure to asbestos dust with the develop ment of pulmonary cancer in man.1-2 Since The chamber measured 8 X 8 X 8 ft Initial ly, it had only one four-inch input opening con necting with the outflow duct of a hammer-mill. After 52 weeks, a second four-inch input for the second hammer-mill was installed. There were three exhaust openings in the chamber. Two of these were 1.5 inches in diameter each. Both led to electrostatic precipitators and then con nected with a vacuum pump. The flow of ex haust air through one precipitator was approxi the development due to the pulmonary dep mately 125 liters per minute whereas the flow osition of chrysotile asbestos dust of pri mary lung cancer in an experimental ani mal has not been reported, the occurrence of malignant lung tumors in our asbestotic rats is being reported and a hypothesis ex plaining this phenomenon is offered. through-the other precipitator was adjusted ac curately to 8 liters per minute. The dust in the latter was carefully weighed at the end of each week to determine the average dust concentra tion in the chamber during that week. The third exhaust opening in the chamber, ten inch es in diameter, led to a ten-inch furnace-type Submitted for publication May 4, 1967; accepted lay 24. From the Industrial Hygiene Foundation, Mellon istitute. Pittsburgh. Reprint requests to 4400 Fifth Ave, Pittsburgh "''1 9 m. F*-- filter and then to an exhaust fan. The latter was used to exhaust the chamber rapidly at the end of each day's exposure. The chamber was actually under a slight, but definite, negative pressure during the dust exposure. Arch Environ Health--Vol 15, Sept 1967 a- ii If' % I cc so Kv. 5E1 344 EXPERIMENTAL ASRESTOS1S--GROSS ET AL A total of 131 male white rats were exposed to chrysotile dust in the chamber six hours a day, five days a week for a maximum of 62 weeks. Of these 131 animals, 70 had had an in tratracheal application of 0.05 ml of 57c aqueous sodium hydroxide, the purpose of which was to impair clearance from the lungs of inhaled dust (The 70 rats were the survivors from a total of 61 animals that had received the caustic applications intratracheally.) The in tratracheal caustic applications were accom plished with the animals under ether anesthesia and with the aid of a self-retaining illuminated speculum that allowed visualization of the vo cal chords. A long 16-gauge needle attached to a microliter syringe was used. The end of the needle was plugged, rounded off, and two small openings made on opposite walls about 5 mm from the tip. After expelling the contents of the syringe onto the tracheal wall, the needle was withdrawn with a spiral motion to distribute the caustic as widely as possible. The caustic-treated animals were introduced into the inhalation chamber at different stages of healing of the tracheal lesions. Some were exposed to the dust immediately; others, one, two, and three weeks after the caustic applica tion. The 61 rats not treated with caustic were also exposed in the chamber to chrysotile dust and were assumed to have normal tracheas. Dust exposures were discontinued after 62 Fig 1.--Electron photomicrograph of chrysotile dust obtained by thermal precipitation from the inhalation chamber. weeks. The survivors were then pastured for the remainder of their lives. served as laboratory controls and were not Dust concentrations in the chamber during treated in any manner. They were housed in a the first year, with only one mill in operation, room, remote from the exposure chamber, and ranged from 42 to 106 mg/cu m. During the were pastured for the remainder of their live*. second year, with both mills in operation, the Another group of 15 rats served as controls for range was 67 to 146 mg/cu m. The overall aver the caustic intratracheal applications. They had age was 86 mg/cu m (based on weekly deter survived an intratracheal application of sodium minations) . hydroxide but had not been exposed to asbestos Another group of 64 rats was given intratra dust. Finally, a group of ten rats served as con cheal injections of chrysotile dust obtained trols for the rats injected intratracheally with from the electrostatic precipi tators. The same injection technique was used as that Table 1.--Overall Mortality and Lung Cancer Incidence* used for intratracheal caustic applications except that the 16-Month Survivors Rats With Lung Cancer* 18-gauge needle had its open ing in the tip, the edges of Animal Group No. of Rats No. % of Total % Of No. Survivors which were rounded off. The Chrysotile inhalation with dose injected was 3.5 mg sus intratracheal NaOH 71 31 44 15 48 pended in 1 ml water. Of the 64 rats, 14 animals received but one injection; 14 received two; 1 received three; 9, four; and 26, six. In addition. 30 rats of the same shipment as that of the animals receiving exposure Chrysotile inhalation with out intratracheal NaOH 61 41 67 10 24 Chrysotile, intratracheal 64 19 30 3- 16 Laboratory controls (no treatment) 40 25 63 0 0 NaOH controls (no chrysotile) 15 14 93 0 0 Exclusive of lymphoblastomas (known to occur spontaneously in these rats). Arch Environ Health--Vol 15, Sept 1967 be: pin no; in rr.rr.c or si I b> o 0 T >'5 AT AL EXPEMMEXTAL ASDESTOSIS--GROSS ET AL 345 ST0096760 j.bi^tns dust. They had received intratracheal ejections of water. The survivora of the multiple intratracheal fhn'sot'te injections were killed 21 months aftfX the first injection. fhe lunK f *11 rats were distended with 10% formalin solution under a pressure of 12 ^ water. Sections were cut at 6m from blocks pf lung embedded in paraffin and stained with hematoxylin and eosin. After photomicrogra phy the sections were decolorized and impreg nated with silver according to the method of (Gordon and Sweets. Some sections were also incinerated in order to obtain information on (he amount and distribution of the chrysotile Host. Results iratory controls and were not manner. They were housed in a rom the exposure chamber, and for the remainder of their live*. of 15 rats served as controls for ttracheal applications. They hud atracheal application of sodium ad not been exposed to asbestos group of ten rats served as conts injected intratracheally with I 1 '! * id Lung Cancer Incidence* ' 6-Month Survivors Rats With Lung Cancer* % of % ol N o. Total No. Survivors 3 1 44 1 5 48 - 67 10 . 9 30 3 35 63 0 24 16 0 : 4 93 0 0 n occur spontaneously in these 1 JI ) The earliest lung cancer was found in a rot that had died 16 months after initiation of chrysotile dust exposure in the inhalation chamber. Calculations of the prevalence of lung tumors were based on the number of animals that live 16 months or longer after having been placed in the inhalation cham ber (72 rats) or after they had received the first intratracheal injection (19 rats). Of the 72 surviving rats from the inhalation cham ber, 31 had received an intratracheal ap plication of sodium hydroxide and 41 had not been so treated. (These data are shown in Table 1.) Many lung cancers were microscopic in size but some were visible grossly, the size ranging from 1 to 5 mm in diameter. All tu mors were peripheral in location and none originated in a bronchus or bronchiole. The lumors were light gray-tan in color, bulged slightly above the surface, and were covered by a shiny, thin, transparent pleura. The consistency was firm, verging on hardness. On section, the tumor presented a dull, gray, dry, arid compact appearance. The cut surface did not appear to retract or bulge. The edges of the tumor were smooth and stood out sharply from the surrounding re tracted, moist, generally dark-red lung tis sue. There was no evidence of a capsule. Distinguishing these tumors grossly from the Iymphoblastomata that occurred spon taneously in our aging rats offered no difficulty. The latter tumors were located in the hilar regions, although some of the larg er ones extended to the periphery. They were infiltrating, hence their edges were less Fig 2.--Adenocarcinoma occupies almost the diag onal upper half of the field whereas the lower portion of the field is occupied by a large papillary adenoma. Neither tumor is encapsulated. Compressed alveolar tissue separates the two tumors. This Is from a rat that was exposed to chrysotile dust for 62 weeks be ginning immediately after the Intratracheal caustic ap plication. The animal died 12 months after removal from the exposure chamber (hematoxylin and eosin, X 40). sharply defined. The cut surface was glis tening gray, tended to bulge slightly and, most important, it tended to be lobulated, each lobule having a tiny hole in the center. The hole was the lumen of either a vessel or a bronchus (or-bronchiole), whereas the lobule represented adventitial and periadventitial tumor infiltration. Altogether, there were 28 animals with primary malignant tumors of the lung, ex clusive of spontaneously occurring pulmo nary lymphoblastoma but inclusive of one malignant mesothelioma of the pleura and three questionable, but probable, adenocarcinomata of the lung. The overall prev alence was 31%, based on 91 animals sur viving 16 months or longer. Of the 28 rats with malignant pulmonary tumors, 25 had inhaled chrysotile dust, giving this group a cancer prevalence of 35% (based on a co- Arch Environ Health--Vol 15, Sept 1957 346 F.XrF.niVESTAL ASBESTOS!S--GROSS ET AL 19L960Q1S >1 cc *3 A to * m )U f> m ,t '0 /k Mi Fig 3.--Adventitial and periadventitial invasion by tumor glands about a pulmonary artery in a rat ex posed to chrysotile dust for 62 weeks beginning three weeks after the intratracheal caustic application. It died 15>/2 months after removal from the exposure chamber (hematoxylin and eosin. X 150). hort of 72); three had been injected intra- Fig 4.--Example of one of the many perplexing le sions found in the asbestotic rat lungs. The lesion is characterized by a collection of fairly large spaces that have occasional irregular evaginations. lined by deeply staining, hyperplastic cuboidal epithelium. The walls are relatively thick but not collagenized. The lesion resembles an adenoma, but more likely represents adenomatoid metaplasia and hyperplasia. From the same lung as in Fig 3 (hematoxylin and eosin, X 150). tracheally with the dust, a prevalence of 16% (based on a cohort of 19 rats). Of the survival time, and nature of the tumor is 25 animals with primary lung cancer in given in Table 2. cluding the mesothelioma in the inhalation No lung cancer was seen in any control group, 15 had received an intratracheal ap animal other than the spontaneously occur plication of lye, giving a cancer prevalence ring lymphoblastoma. of 48% (based on a cohort of 31 rats treated The diagnosis of adenocarcinoma is usu with lye and surviving 16 or more months). ally not difficult since a large nodule com The remaining ten animals with lung can posed of crowded, diminutive and atypical cer that had not had intratracheal lye treat glands makes recognition of this malignant ment, represented a cancer prevalence of tumor relatively easy (Fig 2). Often, how 24% based "on a cohort of 41 rats. ever, other foci of altered morphology in the There were 17 rats with certain adenocar-" same lung or even in the same section lead cinomata and three animals with question to less certain conclusions. The diagnosis able but probable adenocarcinoma. Of seven of some pulmonary lesions is not easy and, in rats with fibrosarcomatous lung tumors, a few instances, was designated as "question three are also listed with the ones having able but probably correct." A fairly common. - adenocarcinoma. Of four rats with a pulmo but fascinating lesion involves regions of nary squamous cell cancer, one also is listed acellular, collagenous scarring affecting re with the 20 having adenocarcinoma. A list spiratory bronchioles, their adjoining alveolar of the animals with tumors, their exposure, ducts, and their evaginating alveoli. In a few Arch Environ Health--Vol !5, Sept 1967 '.'OSS ET AL EXPEHIMEXTAL ASDESTOS1S--CROSS ET AL 347 e nple of one of the many perplexing le-. the asbestotic rat lungs. The lesion a y a collection of fairly large spaces that I irregular evaginations. lined by deeply plastic cuboidal epithelium. The walls hick but not collagenized. The lesion adenoma, but more likely represents letaplasia and hyperplasia. From the Fig 3 (hematoxylin and eosin, x 150). i I ?, and nature of the tumor is hie 2. ancer was seen in any control than the spontaneously occurolastoma. osis of adenocarcinoma is usuTixlt since a large nodule comvded, diminutive and atypical recognition of this malignant ely easy (Fig 2). Often, how:i of altered morphology in the even in the same section lead n conclusions. The diagnosis .narv lesions is not easy and, in s, was designated as "questionblv correct." A fairly common, ig lesion involves regions of agenous scarring affecting rechioles, their adjoining alveolar ir evaginating alveoli. In a few j j j i * | ^ i ' 4 1 ' i | j 57 Table 2.--Rats With Lung Tumors and Their Survival Time Dust Exposure No. of Primary Lung Cancersf Rat 3.5-mg NO. NaOH- 62 Weeks Doses (Mo) Adeno Sq Cell Fibro Meso Adenoma ' '.-304 - + - 16 + 8847 - + - 16 4" 8742 1 + - 16 4- "8752 0 + - 16-/* + ' 8761 1 4- - 16 7. 4- "9068 0 + - 23 4- "8747 3 4- - 25*4 4* 4* "878 1 2 4* - 26 4- + + "8740 0 + - 26/7 + 4- "8834 - 4- - 2 7y. 4- "8835 - + - 2 7 v. 4- 8782 2 + - 2 7v, 4- 8863 8767 - 1 + 4- - 27V< 4* - 28 4- + //> + 8854 -- + - 28U. 4- 4- 1 8743 1 + - 28 V? + C_J 8826 - 4- - 28% 4- <-J 8796 3 + - 28% 4- VO 8768 1 + -- 29% 4- 8811 - + - 30*4 + cr* 8840 8783 10566 2 3 -1+ - 31 + -- 31% + - 32 4- cr* *V1 4- 10567 3 4- - 32% 4- 11175 - + - 34 + 9188 - - 6 21 K:: + 9191 - - 4 21 K 4- 9204 - - 4 21 K + * Applied intratracheally weeks prior to inhalation exposure. Animals exposed to dust immediately after the caustic application are designated by 0. Those that received no caustic application are designated by --. T Exclusive of lymphoblastoma, which is a common spontaneous tumor in these agmg animals. Adeno, adeno* carcinoma; Sq, squamous: Fibro, fibrosarcoma; Meso. mesothelioma. Killed. All other animats were found dead in their cages. scarred foci in many lungs, the air spaces are sive mitotic activity, these foci, if consid lined by darkly staining cuboidal to columnar ered as tumors, should be designated as be epithelium. The combination of distortion of nign, ie, adenomata. But if they are indeed Ihe air spaces with abnormal epithelial lin to be considered as adenomata, they are ing give the impression of neoplasia. Be quite differenffrom the spontaneous pulmo cause the lining epithelium is uniformly nary adenomata seen commonly in mice regular, well oriented, and devoid of exces (and rarely in rats and guinea pigs). Al though the usual pulmonary Table 3.--Trace-Metal Content of Chrysotlle Dust and ot Worn Part From Hammer Mill adenoma is not circum scribed when small, the Chrysotile larger it grows, the more it Element Prior to Hammer Milling. Mg/gm Sample After Hammer Milling Increase. Mg'mg Sample % Worn Hammer From Mill, Mg/gm Sample tends to compress the sur rounding lung tissue and thus develop a pseudocap sule of compressed lung tis N.ekel Cobalt Chromium Lead 134.8 8.8 1 14.6 None 245.5 21.6 153.3 37.9 82 2,338 sue. The adenoma-like foci 145 185 in the lungs of asbestotic 34 rats are irregular in shape 3.5 and although of moderate Arch Environ Health--Vol 15, Sept 1967 348 EXPERIMENTAL ASBEST0S1S--GROSS ET AL ST0096763 2 (ir 3 y* i V* / Fig 5.--To the right of this large adenocarcinomatous mass there are two small foci of peribronchiolar adenomatosis. To the right and below center of the field is a smaller nodule of adenocarcinoma in which glandular epithelium is darker than in larger mass. This is from a rat that received six intratracheal injec tions, each of 3.5 mg chrysotile dust It was killed 21 months from time of the first injection (hematoxylin and eosin, X 150). Fig 6.--Small pleomorphic, darkly-staining tumor cells, many of them more or less spindle-shaped, lin ing the thick walls of slit-like spaces. With silver im pregnation the thick walls are recognizable as septal walls. The pleomorphic cells are associated with fine argyrophilic fibers distinct from the thicker septal fi bers. This is from a rat exposed 62 weeks to chryso tile dust without an intratracheal caustic application. It died about 11 months after removal from the inhalation chamber (hematoxylin and eosin, X 150). size, lack even a suggestion of circumscrip tion. Other significant points of difference between adenoma-like foci in asbestotic rat lungs and the usual spontaneous pulmonary adenoma of small laboratory animals are the presence in the latter of crowded glands with minimal intervening stroma, tall columnar epithelial cells with abundant pale cyto plasm, and papillae with delicate stromal stalks. None of these features are found in the adenoma-like lesions of asbestotic rat lungs. Yet, there are similar foci in which the lining epithelium of the spaces is so crowd ed as to suggest growth activity beyond that consistent with mere hyperplasia; the epi thelium is larger with more abundant cyto plasm; intra-acinar papillae are present, and smaller oval or tubular glandular struc tures are found within expanses of acellular collagenous stroma. These foci are mani festly neoplastic. Malignancy is evidenced by the presence of numerous irregular small glands and cell nests within broad, densely collagenous scars, forming a picture remi niscent of a scirrhous breast cancer. In oth er lungs, foci of malignant neoplastic glands may be characterized by_cellular atypia such as piling up of nuclei with poorly defined cell boundaries and loss of cellular polarity. Occasionally the diagnosis of ma lignancy is facilitated by demonstration of epithelial tissue in the adventitia or the periadventitial stroma of larger vessels (Fig 3). Diagnostic problems arise when lesions intermediate between those manifestly neo plastic and those considered as atypical meta plasia are encountered. Such intermediate lesions are fairly common (Fig 4)! In a number of rats, multiple fod of ade- Arch Environ Health--Vol 15, Sept 1967 4 Fig large and 2 p-epd -nine -:rat r.ont (Kerr r.oa ::or. :he Ir rru 6). *?aand pep Col the del ; mo: 'err : lir ;.-n diiT .t :he r; w.1-' .?**-*.. i.s'.S ET AL ) * I f I EXPERI M EXTAL ASREST0S1S--CROSS ET AL 349 I pleomorphic, darkly-staining tumor nem more or less spindle-shaped, linills of slit-like spaces. With silver imhick walls are recognizable as septal norphic cells are associated with fine -s distinct from the thicker septal fim a rat exposed 62 weeks to chryso: an intratracheal caustic application. II months after removal from the 3er (hematoxylin and eosin, X 150). ,tic. Malignancy is evidenced ce of numerous irregular small 11 nests within broad, densely cars, forming a picture remidrrhous breast cancer. In othaf malignant neoplastic glands acterized by cellular atypia g up of nuclei with poorly oundaries and loss of cellular sionally the diagnosis of macilitated by demonstration of ue in the adventitia or the l stroma of larger vessels (Fig problems arise when lesions etween those manifestly neose considered as atypical metaountered. Such intermediate Iv common (Fig 4)! of rats, multiple foci of ade- 967 Fig 7.--This field represents an undifferentiated, large cell sarcoma involving subpleural parenchyma and also extending over the pleural surface. The silver preparation indicates an intimate association of argyrophilic fibers with these cells. From a rat exposed 62 weeks to chrysotile dust, beginning two weeks after an intratracheal caustic application. The animal died n>/2 months after removal from the inhalation chamber (hematoxylin and eosin, X 300). nocarcinoma are present in the same sec tion, indicating the multicentric nature of the tumor (Fig 5). In several animals, a cellular fibrosarco matous tumor was found in the lungs (Fig 6). On the periphery of such tumors, the air spaces are narrowed, their walls thickened and composed of pleomorphic cells with hyperchromatic nuclei and scanty cytoplasm. Coherent masses of these cells partially fill the air spaces. They are associated with a delicate, argyrophilic supporting stroma. The more substantial stroma of the alveolar septa remains intact. Although superficially resem bling the lymphoblastoma that occurs spon taneously in aging rats, these fibrosarcomas difTer from the latter by the pleomorphism of their cells, their tendency to respect the alveolar architecture at their periphery, and the presence there of a delicate argyTophilic Fig 8.--Combination of adenocarcinoma and fibro sarcoma of the lung. Both tissue elements are well differentiated. Glandular cells secrete mucin whereas stromal cells are spindleshaped and resemble those of endometrium. This field is taken from the lung of the same rat as in Fig 7 (hematoxylin and eosin. X 150). stroma. In contrast, the spontaneous lympho blastoma completely destroys the alveolar architecture whether it is of the lymphocyte or the reticulum cell type. A second form of fibrosarcoma was encountered in a lung, with an adenocarcinoma in which the stroma is composed of crowded, fairly uniform vesicu lar ovoid nuclei and an unusually dense argyrophilic network. In one animal there were foci in which sarcomatous cells resemble decidual cells (Fig 7). Interesting combina tions of adenocarcinoma and fibrosarcoma were found in several animals (Fig 8). The fibrosarcoma may be pure or may contain epithelial components (Fig 9). Perhaps the most interesting of all, be cause of its rarity, was a pleural mesotheli oma in one rat that covered portions of both lungs and the pericardium like a cuirass. The tumor consists of straight and tortuous as well as branching tubular glands seen in longitudinal and cross sections, associated Arch Environ Health--Vol 15, Sept 1967 FXPETIIMFXTAL .1SHFSTOSIS--GROSS FT AL i*. Fig 9.--Another combination cf adenocarcinoma and fibrosarcoma of the lung showing a small gland in the center surrounded by well differentiated fibrosarcomatous tissue. This is from a rat exposed to chrysotile dust for 62 weeks and died 13 months after removal from the exposure chamber. It had received no intratracheal caustic application (hematoxylin and eosin. x 300). Fig 10,--Portion of thick, pleural mesothelial tumor composed of irregular glands embedded in a dense collagenous tissue. Except for a few small superficial foci, the tumor does not invade lung tissue. This tumor was found in a rat exposed to chrysotile dust for 62 weeks beginning one week after an intratracheal caus tic application. The animal died 13Vi months after re moval from the inhalation chamber (PAS stain, X AO). with a moderately cellular but relatively The smallest consists of stroma only and dense stroma (Fig 10). The glands are lined forms a lentiform pleural thickening. by sharply delimited, cuboidal to low colum Although the squamous cell carcinomata nar epithelium with round, uniform-appear are not difficult to diagnose (Fig 12). occa ing nuclei (Fig 11). The lumen of some sional foci of intra-alveolar atypical squa of the glands contains PAS-positive materi mous cell production have some malignant al, yet the cytoplasm of the lining cells is aspects (Fig 13)._These have been catego PAS-negative. The density of the stroma is rized as probable atypical metaplasia. best seen in the silver-impregnated section Another interesting tumor is a mucinous where crisscrossing thick collagen bundles adenocarcinoma, forming a subpleural mass seem to enclose glandular elements. The about 4 mm in diameter. This tumor, also stromal cells are ovoid in shape and their termed "colloid" carcinoma, consists of ir nuclei of a uniform, moderately dense char regular, mucin-filled cystic spaces that have acter. Over the greater portion of the pleur destroyed and replaced the alveolar archi al surface, the tumor is sharply confined to tecture (Fig 14). Very few viable tumor the pleura. However, in several foci there is cells lining these spaces can be identified superficial invasion of alveolar tissue. Sev ' Fig 15). This type of cancer is more com eral mesothelial tumor nodules appear as monly seen as a primary tumor in the gas isolated foci upon an otherwise normal trointestinal tract. pleura. Whether these are metastases or With the exception of one squamous cell multicentric primaries is difficult to say. carcinoma that metastasized to a satellite Arch Environ Health--Vol 15, Sept 1967 r "vWy ...f.i'?"1' 1 " ; rym - Vi. * 'T*z ^ - issKJ&U is" iP -CROSS ET AL EXPERIMENTAL ASDESTOSIS--CROSS ET AL 351 1 1 .4 ' I Portion of thick, pleural mesothelial tumor f irregular glands embedded in a dense tissue. Except for a few small superficial or does not invade lung tissue. This tumor -i a rat exposed to chrysotile dust for 62 nng one week after an intratracheal causn. The animal died 13'/i months after rehe inhalation chamber (PAS stain, x 40), est consists of stroma only and ntiform pleural thickening. a the squamous cell carcinomata ficult to diagnose (Fig 12), occa- of intra-alveolar atypical squaproduction have some malignant ig 13). These have been categoobable atypical metaplasia, interesting tumor is a mucinous loma, forming a subpleural mass -n in diameter. This tumor, also lloid" carcinoma, consists of iricin-filled cystic spaces that have md replaced the alveolar archig 14). Very few viable tumor these spaces can be identified his type of cancer is more com as a primary tumor in the gas1 tract. exception of one squamous cell :hat metastasized to a satellite \ j ' J< . I j j 1 J | * l. | \ | f 1967 1 Mjr*dK#4.r / Fig 11.--Higher magnification of several tubules of the mesothelioma. Cytoplasm of the lining epithelium is PAS-negative although PAS-positive material is pre sent in the lumen (X 300). lymph node and the possibility that some of the discrete pleural foci of the mesotheli Fig 12.--Irregular cords and nests of wen-differen tiated squamous cells show no keratin formation. Ex tensive keratinization is present in other fields of this pulmonary squamous cell carcinoma. From a rat ex posed to chrysotile asbestos for 62 weeks without caustic treatment. Animal died 14 months after remov al from exposure chamber (hematoxylin and eosin. X 150). oma represent secondary nodules, no evi dence of metastasis has been observed in cause-and-effect relationship between chry satellite nodes or elsewhere. Examination of sotile dust exposure and lung cancer. Until other organs in animals with lung cancer this study, such a cause-and-effect relation failed to demonstrate the presence of a tu ship was only surmised epidemiologically mor that could have produced a pulmonary by a study of workers exposed to chrysotile metastasis. dust in the British asbestos textile industry1 These lung cancers have no apparent re (largely prior to 1933). In the United States lation to the degree of asbestotic involve an epidemiologic study of members of an ment. Not one tumor-bearing rat had asbestos workers' union also indicated in diffuse asbestosis. In all animals the asbes creased prevalence of lung cancer, but the totic involvement was multifocal, with nature of the-dusts to which the workers abundant normal, or near-normal alveolar had been exposed was uncertain.2 tissue between such foci. Lesions were cen The present study, with its 31% prev tered in the proximal portions of the pri alence of pulmonary cancer, is in striking mary lobule. contrast to results reported by Wagner et al* who encountered no lung cancer. It seems Comment unlikely that a difference in susceptibility to lung cancer in the strains of rats used The development of lung cancer in rats could account for the differences in results. caused by the pulmonary deposition of However, there is an excellent probability chrysotile dust confirms experimentally the that a particular quality of chrysotile dust Arch Environ Health -Vol 15, Sept 1967 CO __ . 352 EXPERIMENTAL ASDESTOSIS--GROSS ET AL Fig 13.--Nests of transitional or partially differ entiated squamous epithelial cells derived by meta plasia from alveolar lining epithelium. Although there is loss of cellular polarity and nuclei are piled up on one another, latter are uniform in appearance and show no mitotic activity. Squamous cells are confined within limits of alveolar spaces. Lesions such as these are not classified as neoplastic. This is from a rat that had an adenocarcinoma in another portion of lung. This animal had been exposed to chrysotile dust for 62 weeks and had died I6V2 months after removal from exposure chamber. There had been no intratracheal caustic application (hematoxylin and eosin, X 150). deposited in the lungs of our rats was sig nificant in producing lung tumors. The pro longed hammer-milling used to prepare the chrysotile dust in our investigation was not employed in other long-term studies. Wag ner et al4 did not further comminute the dust they used for animal exposures but reported obtaining material directly from industrial dust collecting apparatus. We, therefore, presume, although we are not cer tain, that the tracemetal content of the lat ter dust may not have been as high as in the dust to which our animals were exposed (Table 3). The probability exists that fac tors other than the crystalline hydrated mag nesium silicate in chrysotile asbestos dust may account for the severe pathologic re- Fig 14.--Typical mucinous adenocarcinoma, moat commonly primary in gastrointestinal tract but found in lung of rat exposed for 62 weeks to chrysotile dust without a prior intratracheal caustic application. Pe ripheral (subpleural) nodule composed of irregular mucin-filled spaces that have destroyed and completely replaced alveolar tissue. Rat died 19lA months aftar removal from Inhalation chamber (hematoxylin and eosin, X 40). spouse of lung and pleural tissues to this dust. These factors are as follows: 1. The chrysotile dust used in this investigation has a high trace-metal content, con siderably higher than it was prior to the hammer-mill treatment (Table 3). The in crease in content of trace metals after mill ing was nickel, 82%; cobalt, 145%; chrome, 34%. Aanlysis of the worn hammer of the hammer mill revealed a nickel content of 2,338fig/gm of sample. 2. There is a disparity in the published prevalence of lung cancer among those who worked in the British asbestos textile industry prior to 1933* and that among Canadian chrysotile asbestos miners who worked from 1950 to 1956. Although the workers in the British asbestos textile industry had a ten fold greater risk of developing lung cancer than had the general population,1 this in- I * ( ^ _ 1 O CD VO CT> .irv.or ir. du ne c of di ir.; ! er.ee po?u 3CCO' the to a *CT uT.d ctti; ture *:-e f.r.-oi JT'X' r *Lr CCUS Arch Environ Health--Vol 15, Sept 1967 '>'5 ET AL EXPEHIM EX TA L ASBESTOSIS--GROSS ET Ah 353 mary in gastrointestinal tract but found exposed for 62 weeks to chrysotile dust or intratracheal caustic application. Pepleural) nodule composed of irreguar ooces that have destroyed and completely tlar tissue. Rat died 19Vi months after inhalation chamber (hematoxylin ant) ung and pleural tissues to this l factors are as follows: 1 lrysotile dust used in this inves- f a high trace-metal content, con- | icner than it was prior to the ! J treatment (Table 3). The in- I ntent of trace metals after mill- -;el. 82%; cobalt, 145%; chrome, < sis of the worn liammer of the r II revealed a nickel content of j of sample. J is a disparity in the published ( f lung cancer among those who J e- British asbestos textile indus- ^ 19331 and that among Canadian bestos miners who worked from j % .Although the workers in the I tos textile industry had a ten- r risk of developing lung cancer V ? general population,1 this in- / 1967 Fig 15.--Higher magnification of several of the mucin-filled cysts seen in Fig 14, shows the loss of lining epithelium of these cysts, a feature commonly present in this type of cancer (hematoxylin and eosin, X 150). creased risk of lung cancer was not found among workers in the Canadian chrysotile industry.5 Aside from differences in clima tic conditions, ethnic backgrounds, density of dust exposure, and possibly also in smok ing habits, it appears probable that differ ences in the dust itself to which these two population groups had been exposed may account for the disparity. For example, at the mines, chrysotile mineral is subjected lo considerable crushing contact with highstrength steel alloys to "open" the fibers and remove adherent nonfibrous mineral components prior to shipment to manufac turers. However, considerably more exten sive contact with nickel- and chrome-steel alloys, involving more ami mom of the most finely divided fibers, takes place during this process and during carding as well as spin ning, essential processes in the fabrication of asbestos textiles. This, in a manner simi lar to the hammer-milled asbestos, may cause chrysotile asbestos dust to acquire a higher content of potentially fibrogenic and carcinogenic trace-metals* than is present in chrysotile dust at the mine. 3. Preliminary investigations in this lab oratory, using synthetic chrysotile prepared bv W. T. Granquist, PhD, of Mellon Insti tute suggest it is biologically "inert" (Gross, P.; deTreville, R. T. P.; and Granquist, W. T. unpublished data). Something other than pure chrysotile crystals may be responsible for the lesions found in human and animal lungs which contain asbest06. 4. There is a uniqueness not only in the association of asbestos dust exposure with the occurrence of mesothelioma but also in the association with lung cancer insofar as there is no relationship between severity of involvement and occurrence of neoplasia. It nas not been possible to demonstrate con vincingly the presence of the presumed in citing agent at the site of neoplasia. It is very difficult to demonstrate asbestos fibers or asbestos bodies even in non-neoplastic thickened pleurae. If the association of as bestos-dust exposure and neoplasia repre sents a cause-and-effect relationship--and there is little reason to doubt it--then a log ical explanation might be that something accompanying asbestos dust adversely affects tissues located at considerable distances from the site of dust deposition. This could be one or several trace metals that are added to the mineral dust from machinery and become eluted after pulmonary deposition. Of the trace-metals associated with the as bestos dust used in our study, only nickel has been shown to be capable of producing lung cancer experimentally. Sunderman et al* found that rats exposed to nickel carbonyl developed pulmonary squamous cell carci noma. The many different types of primary ma lignant lung tumors in our animals with pulmonary deposits of chrysotile dust and the different types of cells in these tumors bring to mind a concept originated by Wad dell1, and defended by work in this laboratoiy.8 This theory holds that the parent cell of the alveolar membrane-is multipotential, capable of differentiating either into- epi- 5/5 thelial or connective tissue. Except for the mesothelioma, all the lung tumors described in this paper are believed to have originated CD from cells of the alveolar membrane, a uni-VO O'* Arch Enciron Htahh--Vol 15, Sept 1967 cn co i&Zr g-'f *' I' ;.` r Hi *<r * jf si iv* -Sfc . 354 EXPEKIMEXTAL .4SDESTOSIS--GROSS ET AL tarinn concept in support of which are the peripheral location of all tumors in question and failure to demonstrate a bronchial ori gin. The thesis that metaplasia of alveolar epithelium to the columnar type represents a so-called "bronchialization," ie, a downgrowth of bronchiolar epithelium, has been greatly weakened by Totten's demonstra tion that these metaplastic columnar epi thelial cells contain osmiophilic inclusions similar to those seen in normal alveolar sur face cells, inclusions not found in normal bronchiolar epithelium.8 The ability of neoplastic alveolar cells to assume widely different forms was exem plified by the different types of tumor cells observed. The morphology of alveolar tu mor cells found in asbestotic rat lungs ranged from squamous to columnar, from partially differentiatd squamous (Fig 13) to fully differentiated (Fig 14), even with much keratin production; and from darklystaining cuboidal (Fig 4) to the highly differentiated mucin-secreting variety (Fig 8 and 15). The neoplastic connective tissue cell types ranged from undifferentiated small, pleomorphic cells (Fig 6) to large, anaplastic cells (Fig 7); and differentiated types ranged from small spindle cells resem bling those of endometrial stroma (Fig 8) to plump, collagen-producing fibroblast-like cells (Fig 9). The rapidity (within two hours) with which new argyrophilic fibers become de monstrable on an alveolar wall of a sensi tized animal when challenged with tubercu lin,10 suggests that cells of the alveolar membrane rather than the exceedingly few and widely scattered interstitial mesenchy mal cells of the alveolar interstitium are responsible for these fibers. In a recent study of pulmonary adenomas of mice, the apparent ability of alveolar membrane cells to produce argyrophilic stromal fibers was described.8 Finally, the Unitarian concept of the origin of these tu mors is supported by the observations of previous investigators who noted that many metastases and transplants of mouse adeno mas were sarcomatous.11 *- The fact that the prevalence of lung can cer among the rats which had intratracheal caustic applications was more than double that in animals not so treated, underlines the importance of the upper respiratory tract clearance mechanism, particularly jj, eliminating cancerogenic particles. Attention is also directed to the nature of the asbestotic inflammation in rats. Despite a massive exposure to respirable chrysotile dust for 62 weeks, the asbestotic inflamma tion was multifocal rather than diffuse, and 4 characterized by poorly cellularized collagen. The presence of the latter and of normal alveolar tissue between the collagenized foci confirms our opinion about healing and non- tiprogression of chrysotile asbestosis in the rat.13 Summary One hundred and thirty-one rats were ex posed to chrysotile asbestos, finely milled, in an exposure chamber for six hours a day, five days a week for a maximum of 62 weeks. Dust concentrations were very heavy, ranging from 42 to 146 mg/cu m (average 86 mg/cu m). About half the animals exposed alio had intratracheal application of caustic (lye) to reduce lung clearance and maximize dust retention and pathogenicity. Also, 64 additional animals were given intratracheal injections of chrysotile dust, 14 receiving single doses and the remainder multiple in jections of 3.5 mg each. Controls consisted of 30 untreated, unexposed animals and 15 rats treated with caustic but unexposed to asbes tos. Designed to study asbestosis, the most important findings of the investigation re lated to lung tumor production. Other inves tigators have not been successful in produc ing lung cancer experimentally in rats with chrysotile. Tumor_incidence was high. Of those surviving 16 or more months (ie, after appearance of the first tumor) 31% had pri mary malignant tumors of the lung. Occur rence in rats treated with lye was 48%, twice the rate of those exposed but not treated with caustic. Malignant pulmonary tumors in cluded adenocarcinomas, fibrosarcomas, and squamous cell cancers as well as a mesotheli oma and a mucinous adenocarcinoma. A number of pulmonary lymphoblastomas which occur spontaneously in rats were not counted. Also, atypical metaplasia, not mani festly malignant, was commonly noted. Mul ticentricity of malignant lesions was observed in several animals. Results of this study sup- Arch Environ Health--Vol 15, Sept 1967 F915FUD1$ -GROSS EX Kirtanco o( the upper respiratory caranee mechanism, particularly ^ inp canceropenic particles, u lion is also directed to the nature of stotic inflammation in rats. Despit* ve exposure to respirable chrysotil* 62 weeks, the asbestotic inflamma5 multifocal rather than diffuse, and irized by poorly cellularized collagen, fence of the latter and of norma) tissue between the collagenized focj our opinion about healing and non ion of chrysotile asbestosis in the i * a Summary undred and thirty-one rats were ex3 chrysotile asbestos, finely milled, posure chamber for six hours a day s a week for a maximum of 62 lust concentrations were very heavy, from 42 to 146 mg/cu m (average j m). About half the animals exposed intratracheal application of ca'istic reduce lung clearance and maximize ention and pathogenicity. Also, 64 al animals were given intratracheal s of chrysotile dust, 14 receiving )ses and the remainder multiple inof 3.5 mg each. Controls consisted of ated, unexposed animals and 15 rats vith caustic but unexposed to asbesgned to study asbestosis, the most it findings of the investigation relung tumor production. Other inveshave not been successful in produceancer experimentally in rats with e. Tumor incidence was high. Of -viving 16 or more months tie, after ice of the first tumor) 31% had priilignant tumors of the lung. Occurrats treated with lye was 48%, twice )i those exposed but not treated with Malignant pulmonary tumors indenocarcinomas, fibrosarcomas, and 3 cell cancers as well as a mesotheli! a mucinous adenocarcinoma. A of pulmonary lymphoblastomas cur spontaneously in rats were not Also, atypical metaplasia, not manilignant, was commonly noted. Mul: y of malignant lesions was observed animals. Results of this study sup- I J 4 J 1 I I * ' * J rept 1967 EXPERIMENTAL ASBESTOSIS--GROSS ET AL 355 port the Unitarian concept of tumor origin from alveolar cells. Despite extreme expo sure conditions, diffuse asbestosis was not sfen. Fibrosis observed was multifocal, cen tered in the proximal portion of primary lobules with normal appearing tissue inter vening. Cancer production with chrysotile in this study appears to involve trace metals intro duced during hammer milling which in creased nickel, chrome, and cobalt. Previous investigators have not reported subjecting their dusts to hammer milling. The increase in content of trace metals after milling was follows: nickel, 82%; cobalt, 145%; O A A noKrcir iko - 1--------------- of the hammer mill revealed a nickel con tent of 2,338/ig/gm of sample. This finding, if validated by further studies, should pro vide an important aid to guide industrial hygiene controls of potential health haz ards. This investigation was supported by Public Health Service grant 5-RO1-OH00132-05. The chrysotile asbestos was obtained from JohnsManville Corporation, New York, through K. W. Smith, MD. Analyses of the trace materials were done at the National Center for Urban and Industrial Health. Department of Health, Education and Welfare. US Public Health Service, Cincinnati, through Lewis J. Cralley, PhD, scientist director, who is director of Epidemiology and Field Studios of the center's Occupational Health Program. James M. McNemey participated in the early stages of this investigation. References 1. Doll. Re Mortality From Lung Cancer Among Asbestos Workers, Brit J Industr Med 12:81-86, 1955. 2. Selikoff, I.J.; Churg, J.; and Hammond, E.C.: The Occurrence of Asbestosis Among Insulation Workers in the United States, Ann NY Acad Sci 132:139-155, 1965. 3. Holt, P.F.; Mills, J.; and Young, D.K.: The Karlv Effects of Chryaotile Asbestos Dust on the Rat Lung, J Path Bad 87:15-23, 1964. 4 Wagner, J.C.: Asbestosis in Experimental Ani mals. Brit J Induetr Med 20:1-12, 1963. 5. Braun, D C., and Truan, TX>.: An Epidemio logic Study of Lung Cancer in Asbestos Miners, Arch Induetr Health 17*34-653, 1958. 6. Sunderman, F.W., et al: Nickel Poisoning: IX. Carcinogenesis in Rats Exposed to Nickel Carbonyl, Arch Industry Hyg 20:36-41, 1959. 7. Waddell, W.R.: Organoid Differentiation of the Petal Lung: A Histologic Study of the Differentia tion of Mammalian Fetal Lung in Utero and in Ti ansplants. Arch Path 47:227-247, 1949. 8. Gross. P., and Poel, W.: Pulmonary Adenoma: The Fibrogenic Potential of Its Epithelium, Arch Path 77:424-428, 1964. 9. Totten. R-S-, and Pierce, G.B.: Cortisone and Atypical Pulmonary "Epithelial" Hyperplasia: Fur ther Studies Including Electron Microscopy, Tissue Culture, Animal Transplantation and Long-Term Observations. Amer J Path 45*77-990, 1964. 10. Gross, P,; Westrick, ML: and McNeroev. J.M.: The Pulmonary Response to Tuberculin, Arch Path 70:233-243, 1960. 11. Andervant, H.B.: Pulmonary Tumors of Mice: VH. Further Studies on the Serial Transmission of Lung Turnon Occurring in Inbred Mice, Public Health Rep 64:1519-1524, 1939. 12. Stewart, H.L.; Grady, H.G.; and Andervoot H.B.: Development of Sarcoma at Site of Serial Transplantation of Pulmonary Tumors in Inbred Mice, J Nat Cancer Inst 7:207-225, 1947. 13. Gross, P,, and deTreville, R.T.P.: Experimen tal Asbestosis: Studies on the Progressiveness of the Pulmonary Fibrosis Caused by Chrysotile Dust, Arch Environ Health, to be published. SCIENCE VERSUS LIFE When scientists and engineers undertake to apply science to practical problems, how ever, they find it impossible to stick to quantitative data and rules of logic. The problems include too many other aspects, and in dealing with them they naturally behave like human beings. And so weary cynics like lawyers and political scientists take a certain malicious pleasure in showing that the natural scientists are led by confidence in their special scientific techniques to try to make judgments on political values or policy deci sions, without realizing all the other components of a nonscientific nature that go into those decisions. The lesson is a valid and important one, and needs to be driven home to the general public much more than to the scientists, most of whom know it already even if they do not like for nonscientists to remind them of it.--Price, D.K.: The Scientific Estate, Cambridge, Mass: Harvard University Press, 1965, quoted in Amer Sci 52:528 (Dec) 1965. Arch Environ Health--Vol 15, Sept 1967 CO