Document rx8mpx9o1KB5g8aZ24xmmdo4G

The Biolosieal Action of Inhaled Bervllium Sulfate A Preliminary Chronic . sx:c::y Study on Rots GIRRIT W. H. SCHIMRS. M.O.. D.St.; THOMAS M. DURKAN, M.I.; ANTHONY I. OILAHANT, and FRANCIS T. CRIIDON. Saranac lake. N. T. One of the first steps in the extraction of beryllium metal from beryl is treatment of the ground ore with suifuric acid in order to form beryllium suifate. In subsequent j... steps the sulfate is purified and converted | to the oxide bv calcination. During these > operations, in certain circumstances, plant 1 personnel can inhale beryllium sulfate in a ! particulate state ``Williams M. For this j reason an experimental investigation of the : biological action of inhaled beryllium sulfate has practical significance. '. From the point of view ot biology, a study | of beryllium sulfate presents the advantage } . that the compound is readily soluble. In investigations of the chronic biological ef fects of beryllium products, the over-all re, suits of inhalation experiments conducted ajj with beryllium metal fume and with the poorly soluble compounds of beryllium have ^ been comparatively unrewarding. Thus, it p has proved particularly difficult to reproduce : - ^ in experimental animals the type of granulo! ' -3 mata characteristic of the human chronic A pulmonary disease (Schepers;). 3 The present paper records the results ob- tained from a series of studies on the long 2 range effects on the rat of inhaled beryllium sulfate. Acute and subacute toxicity studies on inhaled beryllium sulfate have previously been conducted with six species of animals E. - by Sprague. LaBelle. Pettengill. and Stokinger.1 A total of 36 animals were studied s for periods not exceeding 40 days. Of these animals oniv 10 were rats. While a wide range of biochemical, hematological, and Received ior publication Oct. 8. 1956. Director Saranac Laboratory and Trudeau Foundation i Dr Sciieoersi and Associate Di rector Saranac Lanorator-. i Mr D'.irkan < pathological criteria was explored, only the pulmonary reactions are relevant to the present inquiry. The animais showed pul monary edema to a variable degree. It was most marked in hamsters, rats, and rabbits and least evident in guinea pigs. While in flammatory exudate in the terminal bronchi and focal atelectasis characterized the lesions in rabbits, hemorrhages into the alveolar sacs occurred in the hamsters, and the rats showed a moderate degree of infiltration of the pulmonary tissues by neutrophiles and monocytes. These results are significant in that they indicate that beryllium sulfate is capable of provoking considerable pulmo nary damage when inhaled. It should be noted, however, that the above studies were conducted with an aerosol concentration of 90 mg. of BeSO-4HjO per cubic meter-- a very liigh dosage. ^ Systematic investigations into the long- range biological action of beryllium and its various compounds were initiated in the Saranac Laboratory in 1943 bv the iate Dr. Leroy U. Gardner. These studies have con tinued without interruption until the present time. Three of us (T. M. D.. A. B. D., and F. T. C.) have been associated with all these experiments from their onset. One of the projects undertaken was an inhalation study with beryllium sulfate. The lack of a toxic effect of a mist of beryllium suifate inhaled by guinea pigs for three hours daily for periods lasting from two and a naif to four and a quarter months has been re ported by Vorwald.4 The only reaction was the sparse production of large mononuclear alveolar phagocytes. Following the experiments in which beryl lium sulfate mist failed to produce -;?r.;n- I . oniy the it to the >wed pule. It was id rabbits U'hiie in.i bronchi he lesions aiveoiar 1 the rats ration or hiles and incant in juitate is : pulmohould be ere . of iteter-- he longi and its 1 in the late Dr. iave con: present . D.. and with all One ot inalation ack of a : -ulfate jrs daily . naif to .icen retion was onuclcar _jc'iCA- /.' HALED 3ERYLL!l~\t cant puimonary cr.anees in exposeu guinea pig,. ,,rr.:;ar innaiauon stuaies were carried 0it nith rats. Material derived from those itouiti has not been at our disposal, and he reiiiits to be recorded in the present paper are oasea on work subsequently con ducted bv means of gra.ots-in-aid awarded to the -enior investigator i G. \V. H. S.i. c Materials and Methods For the experimental investigation several pounds ot beryllium suifate were obtained in 1949 from the Clifton Products, inc.. through the cooperation ot the Atomic Energy Commission. This original shipment has been the source material tor ail the experimental studies with beryllium suifate sub sequently carneq on at the Saranac Laboratory. The material is a white, coarseiv granular product soiuate in water ana on x-ray diffraction examina tion yields a pattern like that reported for BeS0.4Hs0.` The essential experimental principle comprised exposure of rats to an aerosol of beryllium sulfate in specially constructed octagonal chambers (Ur ban *1 All the chambers employed were idenucai in every respect. These chambers were segregated in a special sealed-off section of an outbuilding both to provide greater safety for personnel and to minimize the risk of accidental exposure of other ammais. The volume of the chamber was about 30 cu. it., and fresh air was drawn in at the rate of about 3 to 10 cu. ft. per minute. Whenever animals were residing within the chamber, a con stant supply of fresh air was maintained for 24 hours a day. This air was passed through a Fibergias filler, the main purpose of which was to guard against any contamination of the atmosphere m the event that, through an accident, the chamber air should happen to be discharged back through the inlet. Ti produce the aerosol, the beryllium sulfate wa; dissolved m distilled water to make a 1% solution, and the solution was atomized into a fine itust, which was introduced into the fresh air stream to the chamber at the point of inlet, located st the top center. The amount of compressed air u-ed inr atomization of the beryllium sulfate soiuU"" was negligible in comparison with the total sniiime o' air passed through the chamber. Smoke studies ot ihe pattern of air flow revealed that 'he aerosol was uniformly diffused throughout each tinmher The outgoing filtered air from alt the dn - ..ers was led int0 a single duct and passed ll" "-i:i a final uitrafiiter before being discharged in'" 'he airrosnhere. 'i rccuiar intervals ike atmospheric concentra" "i ' er' ::n:m suifate in each cnaniner was de termined Dv chemical analysis, id collecting sam ples cor analysis, air was wnr.orawn irora tne in terior oc tne cnamDer at a point just above the breaming level of the exposed ammais and svas passed througn three fnttea-disk gas washing bottles connected in senes. Distilled water was used as tne collecting riuid in the bodies. Most of the beryllium sulfate trapped by the sampling apparatus was recovered in the first bottle, but a small amount was earned over to the second and even to the third bottle. Routinely, the contents of the fiasks. after the sample had been taken, were acidified slightly ana filtered, and the suifate was precipitated with barium chlonde by the standard gravimemc method. In a few cases, as a check, samples were divided into two parts, one pan being used for the suifate determination and the other part tor a bery llium qetermmauon by the fiuorometrtc method of Klemperer and Martin/ Some difficulty was e.xpenenced during the earlier part of the experiments in maintaining the con centration of the aerosol at a uiuionn level, and on several occasions wide rtuctuations in the beryl lium sulfate content of atmospheric samples were observed. Usually, however, the concentration of beryllium sulfate in each chamber was near the level sought. For the whole penod of each ex posure phase the atmospheric concentration was. on the average, about 12y of BeSOi (ly of Be) per cubic foot of air. The number of animals used, their distribution according to strain, source, and sex. and other pertinent data are recorded in Table 1. As is evident, a total of 126 albino rats were exposed to the beryllium sulfate aerosol for periods up to six months, and a further 139 rats were kept as controls in a normal atmosphere in another build ing. Special precautions were taken to ensure that the control animals would not be exposed acci dentally to any beryllium. The Sherman rats used in the experiments were obtained from two separate commercial supply houses (Manor Farms and Sprague Dawley). The Wistar rats represented animals from the Saranac Laboratory colony which had been bred at the Laboratory during the past 10 years. For practical purposes the female animals predominated in the experiment, but a suificient number of males was included to be sure that the effect of sex would not be overlooked in evaluatinc our results. At the start of the experiments the rats of the Sher man strain ranged in weight between 30 and 110 gm.. and the rats of the Wistar strain between 140 and 210 gm. Two basic experiments were conducted: In one (Study A) the penod of exposure to the beryllium sulfate aerosol ranged from l day to 30 days, while m the other lStudy B) the penod was 180 lavs As part of the experiments special attention was gneu ;o ihe conrrol ammais studies C and 33 A. M. A. ARCHmS n;CL'37R!A 'able I--tabulation or Rai ; m Experiment on the Bioiogtcai .-Icuon of 3eryt,iun: Aerosol Experimental Group -Exposed to Beryllium Sulfate Aerosol ter Different Lengths of Time and Thereafter Allowed to Reside in Normal Air Controls: Never Exposed to Beryllium Sulfate Aerosol Animal Series Group No. of umzialj Sima /l iWhetrjnuurm ( Sprwue Davley Source l Manor Farms i Saranac Laboratory Sex 'l fmeamteale Duration of exposure. days A reran concentration of BeSO*. rcu. ft. Penod in normal air. mo. latcrcumru aeatru Cnder exposure In normal air EeoertmeniaJ Group Airman Exposed to BesO Controls Animals Neeer Exposed to BeiO. D). In the first study (Study A) three groups, each composed oi seven rats, were exposed in the same chamber to the aerosol tor. respectively, one day, one week, and one month. In the case of the one-dav exposure the beryllium sulfate mist was disseminated for 8 hours, but the rats remained in the exposure chamber for a further 16 hours. On removal, four were immediately killed. while three were allowed to live in normal air for a period of three months. A similar procedure was followed for the rats exposed for a week and for a month. It should be noted, however, that a week in this case represents an exposure to the beryllium sulfate aerosol of eight hours a day on five days, four hours on Saturday, and no exposure on Sun day. These nine animals that were allowed to survive in normal air are not reported on in this study but will form the basis of a subsequent communicadon dealing with the delayed effect of transient exposures to beryllium compounds. For the IIS rats which inhaled the beryllium sulfate aerosol for 180 days (Study B), a sched ule similar to that of Study A was followed, but. owing to limitations of space, the animals were exposed in three separate chambers, and the experi ment was conducted in two phases. During the period while the animals were being exposed to the beryllium sulfate aerosol, 46 of the 136 rats died, and a further 29 were killed. Of the latter group 15 were sampled on completion of the full penod of 180 days of exposure, and the other 14 were killed prtor to the end of the third month of exposure. In addition, there were the nine rats mentioned in the preceding paragraph. This left a balance of 52 rats which were transferred to normal air and observed for periods up to 18 months. During this time 12 animals died, and the remainder were killed in the order shown in Table 2. Of the 139 control animals 70 were of the Sher man strain (Study C) and 69 of the Wistar strain (Study D). These animals were killed in sets of three or four at intervals of two months. Only five of the Sherman group and five of the Wistar group died of natural causes. The changes which developed in these two strains of laboratory rats could thus be followed over periods of 20 and 22 months, respectively, and could be contrasted with the lesions provoked in the lungs of the rats which had been exposed to beryllium sulfate aerosol. Intercurrent Deaths 1 Since Sprague. LaBell^, Pettengill. and Stokinger 3 had previously demonstrated that animals which inhaled beryllium sulfate aerosol in relatively high concentrations de veloped an acute form of chemical pneu monitis. the prevalence of spontaneous deaths in our series of animals requires some comment. As is shown in Table 2. the majority of those deaths occurred during the first three months. The fact that no ani mals from the control group died during this period tended to incriminate the beryl-: lium sulfate as a factor in the cause oi ' the deaths. The majority of these animals died from a subacute to chronic type of pleuropericarditis with a tendency to chronic macrophage pneumonitis and multiple ab scess formation. No pathogenic bacteria could be isolated from these lesions. Never theless, in spite of these indications that the beryllium sulfate may have been the cause. 34 :*ois '?r Exposed ie?0. D 59 -9 :G y 0 >t th incr* . istar strain 1 in sets of melts. Only the Wistar nqes which >rntory rats t 20 and 22 trusted with ? r*r* which ngill. and rrated that n suitace ations de.cal pneuontaneous - requires able 2. the ed during :at no am- ; ed during the beryi- cau?e of ,e animals c tvpe of to chronic iltipie ab- , : bacteria' i: Never- ' is that the the cause, ?abl 2 --cjr, of In'nQua 3er\tnum Suirzte Aerosoi on :ne uunqs or Rots Gross 0 bserzations Miscellaneous Lesions z. Gross OoserTsuotu Reactive -xri Pulmonary Tumors period of Experiment. Mo. No. of A nun* is Mod* of Death = Lymphatic 5 = Synem 2 Tumor Exposure to Beryllium Sullu* Aerosol Cs to : and 4 ittd 3 5 31 / IS dted i 13 tilled 24 f 23 died l l killed S 3 died 13 13 killed S4 5 3* : 1 i: 12 Residence in Normal Air After Six months. Etnosure to Servtltum Sulfate Aerosol 6 ' 4 died 34 32 \ 2 killed 3 4 4 killed 4 ; i died ........................... ........................... 41 > 3 / i died { 4 killed .. .. l 3 3-- 4 * 2 2died 0 4 4 killed ....................... * 4 4 1 -> 10 13 13 killed 12 l 1 killed 14 f l died ........................... ........................... ........................... 3 13 13 11 l1l j1 3 29 2l 13 1 i killed 1* t l died ........................... J1 4 17 4 / 1 died l 3 killed 19 / l died 1 3 killed ........................... ........................... l 2 4- 23 1 ** 4 6 3 figure* in Columns 1 ro 7. under Gross Ohs**mooos. refer to tb* number of mm*i ta which the leswo I to ii. to tn* numoer of tumors found. found: in Columns we are satisfied, for the following reasons, that the deaths represented an intercur rent epizootic. First, deaths ceased rather promptly after sulfathiazole and sodium bi carbonate were added to the drinking water furnished to the animals. This therapy was continued tor a period of several weeks, until there were no more deachs. Second, the deaths were limited to animals in the first two chambers in which beryllium suifate exposures were* maintained. In thethin: chamber, in which exposures were commenced several months after the ex posures in the first two chambers were terminated, no deaths occurred during the full six-month period, nor were there any deaths in the animals exposed for a month "i Study A. In more recent experiments additional sets of 10, 110, and 50 rats have been exposed to a beryllium sulfate aerosol. wfi'vh had about the same concentration as betor.'. ;or periods of nine. six. and three month, respectively, without losing a single ^rnial This suggests rather conclusively t^VJt at a concentration of approximately I2y per cubic foot of air beryllium sulfate by itself does not induce either an acute or a subacute pulmonary inflammatory reaction. The possibility that the inhaled beryllium may have facilitated a latent virus-caused pneumonitis has not, however, been wholly excluded. In this connection it may be mentioned that the rats bred at the Saranac Laboratory have been relatively disease-free for a considerable time. Pulmonary Responses During Exposure Prior to the third month of exposure there were no pulmonary reactions other than those associated with the intercurrent in fections. In those animals that died during the fourth and fifth months of exposure, isolated gross foci of reaction could be iden tified. but they were not clearly linked to the inflammatory process which killed the animals. Histopathological changes were sparse and inconclusive. Mild bronchitis and pulmonary hyperemia sometimes occurred in animals which did not die from pulmonary 35 T A. M A. ABCHC'BS OF ; XDCSTRlAL H0AL7H %& w* TSt ?- 5. H: A Mi 4 f WV . wtV'-v Jj it rw! i*S5?r3 L3r..: ft. ds?,.. >=& > TE, & -. 'AC-^?1 ?_-**. stf 5#^3|i **, y*?w Plate I.--rf, foam-cell duster: Rat 200, exposed to beryllium sulfate aerosol for six months, then killed; x 230. B. mira-aiveolar giant foam cells: Rat 21. exposed to ben ilium sulfate aerosol for 6 months, then in normal air 10 months: X 800. C. focal mural cellular infiltration: Rat 198. exposed to beryllium sulfate aerosol for six months, then killed: x IT5 D. focus of interstitial cellular infiltration and partial fibrosis, surrounded by epitheliaiizci distorted alveoli: Rat 26. exposed to beryllium sulfate aerosol for six months, then m normal air six months; x o5. iluLJ IXh'ALBD SBR) disease Also. :r. the ctaer organs surveyed i liver ioieen. kidney pancreas, tnvroid. suprarenal, puimonarv and hepatic iytnph nodes i no aonormauttes couid be aerr.on-trnted Ituese organs the beryllium suir.itv either proved to oe inert as a direct chemicai pathogen or did not become depo-iteci in concentrations sufficient to pro voke even the least cytological deviations common lv associated with miid metallic toxicttv In the majority of the ammais killed at the end ot six months of exposure to the bervilium sulfate aerosoi. there were macrofcopicaily detectable multipie small grayishwhite -ubpieurai foe: of reaction in variable numbers i Tabie 2't. In most instances the lunv; were moderately well seeded with such discrete lesions, and in about half the cases there was some tendency toward marginal coalescence of individual foci. On histological examination the macroscopically detectable erayish-white foci resolved them selves into zones of reaction of variable composition. F-'am-Ctil Clusters.--Collections of foamy macrophages in several adjacent subpieural lobuies were the chief explanation for the macroscopically detectable lesions 1 Plate IA. 3'< Each of the animals killed at the end of six months possessed such foam-ceil clusters i Table o'), and an aver age of about three cluster units could be counted in single whoie-iung sections passing through the midcoronal plane. There was considerable variation in the prevalence of these areas, some rats showing as many as eieht while others had only one or two. There was variation also in the number of alveolar spaces involved in each unit, but m most instances it was clearly apparent that the foam cells had congregated within discrete pulmonary lobules. Marginal con fluence was achieved through involvement of contiguous lobules rather than through spilla(T- of the reaction beyond the anatomical ""its of individual iebuies. At this eariy 'tatre 'he majority of the arfected areas lay 'mnivdiateiv beneath ire Dieura. In those in stances wnere toam-eeu clusters were .temonstraoie more centrally, they were gen erally confined to lung parenchyma abutting on ramifications of larger biood vesseis or oronchi. The foam-ceil units possessed iurther characteristics worth recording. The com ponent foam cells varied from small mononucieated to large muinnucieated macrophages. The nuclear chromatin stained intensely, and there were large karyosomes. The cytoplasm was relatively excessive and somewhat granular, and the vacuoles which usually impart to such ceils their foamy character were as yet relatively smail. The foam ceils were not yet crowded together excessively, but in most cases each aiveolar space within these unit areas possessed two or more cells per plane of section. The cell boundaries were exception ally sharply defined. The presence of such foam-cell-filled lobules, though arresting in appearance, does not yet constitute a specific type of pulmo nary reaction, as similar cell clusters have been described by other authors (Innes. McAdams, and Yevich '). We have ob served this phenomenon in inhalation ex periments with amorphous silica dust, though the foam cells do not usually occur in quite such abundance as was seen in the present instance. In the normal control ani mals (Table 41 isolated foam-cell clusters occurred only rarely. Focal Mural Infiltration.--At numerous points within the lungs the alveolar walls were chickened either by proliferation of the cells lining the alveoli or due to infiltration by macrophages. The occurrence of such foci of mural infiltration was sometimes re lated to areas where foam-ceil clustering was also present. However, foam-cell clustering occurred in some instances without any as sociated mural thickening. On -the other hand, thickening of the wails unaccompanied by foam-ceil clusters was also observed. In enumerating the foci of mural infiltration for the purpose of constructing Table 3. oniv tlio-e foci were recorded in which no 37 8lOLOCiC.iL ACTiG'* Or IXHALEC ?.?>" i.'L'.U SGLFAiI Table 3--Rate o* Occurrence o' Pulmonary Lesions in Rats Resxamg * So+nt3: A,:' 0iv Pulsocary Lesionj* No. Moie Female Lexioo* it si --o IV) >'*.4 5 3 2 Sherman strain 5. D. 5 0. 6 2-6 - 1 - -- A-l 3-5 12 M2 -1 2 l P-6 i0 :.t. f. u W:r tram 3-1 7 13-18 ?-3 -- 19-24 P-2 1 .. 3 - - -- 3 -- 5. L. A-l 1 12 2-6 P-2 - 2 -- -- -- - .. 9-1 2s. L. u 12 M2 E-l i P-l -- * - 4 3. L. t. B-l * 13 13-18 P-5 3 3 - - P-l 4 19-22 PI-2 -- - 2 .. Eipioaar.ou of ADbrtTUuona: A - Abccss B -- Breocmtetasu E - Exudation mtn %lrnii P - Pimuboucu uniantitlal) PI P!anrf S. D. - Fpracut Daw 1*7 M. F. - Manor Farms 5. L. - Saranac Laboratory The rtrures >n the column under Dominant Lesion refer to tbe number of animals exbihlttnr each dominant lesion, and the fllures la tbe columns uoder Pulmonary Lesions refer to tbo total aumbor of lesioos seen in all animals killed. foam-cell clusters were present. Such iso lated zones of mural infiltration occurred ottenest in peribronchial or periductal alveoli ("Plate IC). No stromal elements could be identified, the mural thickening be-:-- due to cellular elements. Generally, too. the infiltration produced no gross dis tortion of the peribronchiolar alveolar spaces, but occasionally some atelectasis with resultant focal consolidation supervened < Plate ID). As may be inferred from Table 3. this phenomenon was as yet infrequently observed in the animals killed at the end of the six months' period of exposure. In the control rats of approximately equal age comparable foci of mural cellular infiltra tion were rarely seen. E pith dialication or Peribronchial Alve0,1 --N'ext in extent, but probably no less significant in importance, was the multifocal epithelialization of peribronchial alveolar spaces. In most instances this epithelial re action commenced unilaterally in relation to an alveolar duct (Plate llA), but in other areas it surrounded the bronchioles (Plate IIP). The epithelium covering the surfaces of these alveoli tended to assume a low cuboidal character. The nuclei were closelycrowded and tended to be relatively pyknotic. Sometimes the affected alveoii appeared to be somewhat atelectatic; at other times they seemed to be distended. Foam ceils and macrophages were often seen within such alveolar spaces, but in other instances the alveoli were empty. There was no tendency at this stage for these alveolar epithelial cells to be shed. An indisputable connection between the alveolar ductal epithelium and J9 M A .i.tCHirZS ?C<7RIAL r --st- - .-r >i-s a.s iv B1*4*. i* . SSSSS Plate II.--A. alveolar spaces lined by highly cellular cuboidal epithelium undergoing dis tortion and distention (note foamy macrophages) : Rat 42, exposed to beryllium sulfate aerosol for six months, then in normal air six months, then killed; X 350. B, parabroncmal evagtnaoon of bronchial epithelium: Rat 198, exposed to beryllium sulfate aerosol for six months, then killed: X 175. C, lobular proliferation of alveolar wall septal cells: Rat 144 exposed to beryllium sulfate aerosol for six months, then killed; X 65. D. cytological detail of part C: Rat 143. exposed to beryllium suifate aerosol for six months, then tailed: x dOO 40 he new epithelium could be demonstrated In the control senes no comparable lesions ltl iome areas: in otner instances the two couid be demonstrated ;v>tems did not appear to be continuous Focal Metap'.asuj or A.veoiar Wails.--In Wlth each other, aithougn senai sectioning one of the rats killed at the end of six w);i have to be resorted to before such a months of exposure to the beryllium suifate connection can be ruled out. Where the aerosol, three areas were found in a wnoie- fpitneiiaitzat:on ended, it generally did so iung section taken through the midcoronai abruptly, so that an aiveoius couid some- piane in which aiveoiar walls had become times be seen to be partly lined with cuboidal the seat of focal metaplasia. The ceils were cells and partly devoid of any epithelium. of an unusual type, and nothing comparable One to three such foci could be demonstrated was seen in the control series. per midcoronai whole-lung section in each Gronvlomata.--Three small granuiomata of the rats killed at the end of six months were found in one oi the animals killed at of exposure to the beryllium sulfate aerosol. the end of six months of exposure to the Xo epitheiiaiizations of aiveoiar walls aerosol. They were near the pleura and could be demonstrated in rats of comparable were composed of macrophages and plasma age from the control series. It should be cells with interposed crystal clefts. Such noted that in man and other animals granuiomata were observed in an occasional epithelium sometimes forms in lung areas control animal but never in rats less than which have been isolated by a process of 18 months of age (Table 4V fibrosis. Typical examples of this type of Neoplasia.--At least one indisputable epithelialization have also been observed in adenoma and several other probable ade animals exposed at the Saranac Laboratory nomata were demonstrated in the rats killed to oil smoke. In the present series of ani after six months of exposure to the mals there was no indication that the beryllium sulfate aerosol. In one instance epithelialization of the aiveoiar walls was focal epithelial proliferation had exceeded due to isolation of the affected areas. a benign cytological condition to such an ex Lobular Proliferations of Septal Cells.-- tent that the lesion simulated an early A fourth distinctive type of lesion was dis papillary adenocarcinoma. The cells cover covered at the end of the six months' period ing the distorted alveolar spaces showed of exposure to beryllium sulfate. It man many mitotic figures, and those surface cells ifested itself as discrete lobular prolifera which were observed within the. spaces re tions of alveolar wall surtace ceils. Under sembled neither macrophages nor foam cells, low power these foci were readily discerni and probably represented desquamating neo ble by virtue of the relative prominence of plastic elements. The tumor tissue invaded the chromatin and the close crowding of the the surrounding parenchyma, there being no multiplying ceils lining the aiveoiar walls capsule. No neoplasia was observed in the 'P'.ue IIC). At the edge of each focus control series of rats. of proliferation there was an abrupt zone of transition to the normal pulmonary Postexposure Sequelae parenchyma. It was noted that the alveolar Contrary to expectation, the pulmonary spaces were not significantly occupied by reaction, which had developed toward the cells, though an occasional macrophage oc end of the six-month period of exposure to curred here, and some of the proliferating the beryllium sulfate, did not subside when cells could be seen in the process of being this nocuous stimulus was withdrawn. Not shed into the lumen (Plate IID). In some only was there a progressive increase'Th the cases several of these centers of proliferation frequency of the changes observed in ani could be demonstrated per midcoronal whole- mals killed when the exposure phase was 'dng section, while almost every rat pos terminated, but additional features made sessed at least one such area. their appearance. 41 A. H. A. ARCHIVES OF ISDVSTRIAL HEALTH' It is of interest to note that significant signs of puimonarv infection disappeared soon after the animals were placed in nor mal air (Table 21. The macroscopicaily detectable foci of re action were observed in increasing numbers and later were seen in all animals. Each focus tended to become somewhat larger, and confluence of distinct units came to be a universal feature until about 14 months after the animals had been transferred to normal air. Thereafter individual foci be came smaller and were inclined to be dis crete. The most outstanding development con cerns the appearance of visible and palpable tumors of increasing size. During the first four months of the postexposure period isolated tumors were detected in the hilar lymph nodes. Thereafter, and until the end of the 18-month period of survival in nor mal air, pulmonary tumors were found in increasing numbers in the animals which were serially autopsied. Not all the rats de veloped macroscopicaily detectable tumors. Some, on the other hand, showed several discrete growths of different size in one or more lobes of the lungs. Though it re mained for histological examination to prove the neoplastic nature of these nodular lesions, the majority had the gross features of new growths. On histological examination of the lungs of these animals (Table 3) the further de velopment of the lesions which were present at the end of the six months of exposure to the beryllium sulfate could be followed. Hyperemia had disappeared. At later stages there was even a tendency to relative ischemia. Emphysema of the atrophic vesicular variety made its appearance in in creasing numbers of animals but did not affect all the rats. Bronchitis was found sporadically throughout the period, and foci of metaplasia of the bronchial epithelium were found in a few of the rats after the ninth month of residence in normal air. At various points, but particularly in alve olar spaces immediately beneath the pleura, spindle-shaped crystals were detected. The 42 nature of these crystals could not be deter mined. but their optical properties suggested that they were composed of choiesteroi. In some instances these crystals were embedded in alveolar wails, and at ocher times they were found within some of the granulomata. Similar crystals have occasionally been ob served in rats never previously exposed to beryllium sulfate. Foam-Cell Clusters.--Increasing numbers of pulmonary lobules became occupied by foam ceils, the incidence per rat lung of these units reaching a peak toward the seventh month in normal air. At this stage a distinct change set in, and in animals killed in subsequent months progressively fewer foam-cell clusters were discovered ('Fig. 1), and each unit became smaller. During the phase of increasing prevalence of the foam-cell areas the individual cells grew in size and increased in number until each alveolar space became virtually choked with these ceils. At the same time the cyto plasmic vacuoles expanded. These units were found in greater number also in re lation to bronchioles and alveolar ducts, al though the tendency to form subpleural foci persisted. It is in the latter locations par ticularly that confluence ofc cells .md in filtration of alveolar walls first became a conspicuous associated phenomenon. With the breakdown of the cell boundaries, nu clear and cytoplasmic debris collected within the alveolar spaces, and sharp crystals could be seen with fair regularity. The greater prevalence of mural infiltra tion at these sites suggests a causal relation ship or a common etiology. It is possible that products of degeneration of these foam I cells became incorporated into the alveolar walls. Soon crystals also appeared :n the i walls, and cellular granulomata formed in relation to them. The latter phenomena were most marked during the phase of decreasing prevalence of the foam-cell units. It is likelv that the clusters ultimately disappeared, both because no new lesions were created and also on account of the progressive break down of the foam cells and the rosor-tion of the resultant detritus. 1 A. M A. ARCHIVES JF lXDCS7RIAL H2.Ai.7i-. slight degree o; ductal stenosis associated with the local murai innitrations may have retarded the egress of the iarge foam ceils so that thev accumulated temporarily in the pulmonary lobules. Focal interstitial ceiiular innitration aiso occurred without undue participation of foam cells. In some instances a small amount of collagen could be demonstrated among the infiltrating ceils, but generally the alveolar walls remained cellular. This fact largely accounts for the subsequent dis appearance of lesions. There is. of course, no proof that in the animals killed after, say, the 10th month in normal air there had previously been greater numbers of foam-cell units and mural infiltrations than were found at later stages. Such a point can be settled com pletely only by means of serial biopsies in the same animals, obviously a technically difficult procedure. The curves of in crementation and regression at least suggest that we are dealing here with evanescent lesions which reflect retarded pulmonary responses to the preceding six months of irritation by the beryllium sulfate aerosol or which may represent a rebound histopathological change caused by the with drawal of the pathogenic agent. Peribronchial Epithelialization of Alve oli.--The fate of the epithelializations of the periductal alveoli and the sequelae of the lobular alveolar wall septal-cell prolifera tions may now be traced. As is shown in Figure 2, the epithelialized type of lesion rapidly became more prevalent with the passage of time and reached a peak pro duction toward the end of eight months after the rats were removed from the beryl lium sulfate environment. Thereafter epitheliaiization diminished as a histological sign with the same rapidity, so that ulti mately it was but rarely encountered. Like wise, the extent of the lesion first increased and thereafter decreased. At the stage of maximum productivity the epithelialized alveolar spaces frequently surrounded the bronchioles almost completely. Two types of reaction appeared in some rats. In the one type of epithelium the r.uciei werpyknotic, and the cvtopiasm scantv in other type the epitheiiai cells possessed a more abundant acidophilic cvtopiasm. It ipossible that infection may have been a fac tor in bringing about these differences, as neutrophile polymorphonuclear leucocytes were commoner in the lesions of the latter type of reaction. What happened to the epithelialized areas after the phase of their peak production de serves some inquiry. At earlier stages the epithelium often appeared to be fuiiv differ entiated. even possessing cilia (Plate IIL-1). These cilia were not observed at later stages. It is quite possible that at least some of the epithelial cells were later desquamated. In many animals such areas of denudation could be found between zones of epithelialization. and it is conceivable that with the passage of time all such newly formed epi thelium was discharged into the respiratory passages. At several sites it would seem as if individual epithelial ceils budded off to become freely roaming spherical alveolar cells. In yet other instances such epithelial cells may have participated in the formation of giant cells after the relevant alveoli had become completely atelectatic. In yet further instances the epithelial cfells became in corporated within areas of interstitial fibrocellular reaction and. through resorption of trapped aif spaces, became reduced to mul tiple circular foci of dark-staining cells within a dominantly stromal area (Plate 1115). These are the regressive changes. In a small proportion of cases epithelial proliferation appeared to have proceeded vir tually autonomously with the formation or papillary processes and secondary acinar recesses (Plate IIIC). The borderline be tween neoplasia and simple hyperplasia could be drawn only with difficulty in many such instances and not at all in others. Epithelialization of the foregoing type did not develop in the rats of the control series during the full period in which they were under observation. Septal-Cell Proliferation.--The lobular foci in which alveolar septal cells under- 44 A. it A. ARCnl-'ES J.- tXDCSTRIAc. -SALT:; went proliferation did not change their the end of the six-month period of exoosure character much as time passed. These lesions "fntteased I5ut~ - ?lbwiy-rm' 'bve?-il~ prevalence and reached a climax oniy after about 15 months from the time of cessation of the beryllium sulfate exposure. In at least some cases they were the source of pleomorphic types of alveolar cells. Not to the aerosoi of beryiiium sulfate, the granuioma^r made their appearance cniv after, this exposure had been discontinued. In their time of onset they thus lagged behind the phenomena of both foam-ceil production and mural infiltration. However, there is seemingly close biostatistical cor much stroma appeared to have developed in relation among these three phenomena, relation to them. It is suspected that in at which may signify a common mechanism least some instances these focal prolifera of pathogenesis. In fact, it wouid appear tions were the source of certain of the that most of the granulomata had evolved tumors which developed in these animals. within alveolar walls when products of the No comparable lesions were observed in degeneration of foam cells became con the control series. centrated among the infiltrated macrophages. Focal Metaplasia.--At isolated foci in the At least some granulomata appeared aiso to lungs of the animals autopsied six months have arisen in relation to de-aerated alveolar and longer after the exposure to the beryl spaces within foam-cell units. lium sulfate had been discontinued, foci In essence, each granuloma consisted of of metaplasia or the alveolar lining cells a central core of large macrophages and a could be demonstrated. In some instances superficial thin zone of plasma cells. In there was squamous metaplasia (Plate most instances, too, a layer of cuboidal or HID). At other sites ciliated columnar or flat cells covered the surfaces of granulomata pseudostratified epithelium predominated. which faced alveolar spaces. Although crys Though these lesions were not numerous, tals were not universally present, they could their significance as precursors to neoplasia be found within individual granulomata must be recognized. This metaplasia was even at the end of 18 months from the entirely different from that often found in time of cessation of the beryllium sulfate rats with chronic bronchiectasis and sup exposure (Plate IV/4). A marked tendency puration (Passey, Leese, and Knox*). was nofed in this series of ij^ts for granu Comparable metaplasia was not observed in lomata to become confluent as their numbers the control series. increased. Toward the 10th month of sur Granulomata.--There can be no doubt that vival in normal air the majority of the the present study yielded a particularly granulomata were relatively large composite bountiful crop of granulomata in the ex structures (Plate IVS). posed animals. Viewed as an objective of Giant cells of the polymorphonuclear type the study, this result was a very desirable sometimes were present in these small unit one. because for the past 12 years it has lesions but were more usually observed in been an endeavor of the Saranac Laboratory relation to composite granulomata. Crystals to reproduce experimentally in animals the were often interposed between the macro pulmonary granulomata characteristic of a phages within the medulla of the granuioma chronic disease exhibited by some industrial and sometimes crowded the latter to such workers exposed to beryllium compounds. a degree that each cell was reduced tn a flat The evolution of the granulomata produced disk-like structure. At other times the in the rats by the beryllium suifate ex crystals had apparently formed within giant posure is outlined in summary form in cells; alternatively, preformed crystals had Table 3 and Figure 1. Attention should be been phagocytosed by giant cells. It is also drawn to the fact that, with the exception possible that in some cases the giant cells of the three instances found in animals at engulfing these crystals represented macro- -6 the ran: had also ceiis .tro- Plate IV.--A. granuloma with crystals and macrophages: Rat 35. exposed to beryl lium suliate aerosol for 6 months, then in normal air 18 months; X 350. B, composite granuloma 'note fringe of plasma cells): Rat_ 19. exposed to beryllium sulfate aerosol for 6 months, 'hen m normal air 10 months: X 175. C. giant-cell formation through confluence of epithelial eils of atrophic alveoli: Rat 36. exposed to beryllium sulfate aerosol for 6 months, then n normal air 10 months: x 350. D, epitheiialized alveolar space lilted with spheroidal cells: Rat 13. exposed to beryllium suliate aerosol for 6 months, then in normal air 18 months, 'lien Willed: x 800. 47 A. M. A. ARCHIVES OF INDUSTRIAL HEALTH I Table 5 --Prevalence of Various Pulmonary Lesions Rats Exposed to Beryllium Jj Suifate Aerosoi and m Rats Sever Exposed to Any Dust . * Pulmonary Lesion* Foam-cell dusters Murml infiltrations Epitheiiaitiaiioa or oenbrooehtal alreoil Lobular proliferations or sepul calls Focal metaplasia < Alreoui Oraoulomata Neoplaimata L'100 refers to oumoer of lesions per 100 animals. Experiments! Group tlM): fieSO. up to 6 Mo. then Normal \:r up to 18 Mo. No. L.'.OO 3S3 282 ns 242 372 ` 273 188 27 118 70 424 310 78 U Control Orooo M39)Normal Air Ooir up to 22 Mo. No. L. 100 U8 13 H 00 0* 0 00 0 00 phages which had fused together to form the multinucleated giants (Plate IVC). No conchoidal bodies could be demonstrated. In the series of 139 control rats only nine granulomata were found and these were confined to only 3 animals. Virtually all the rats that had been exposed to the beryl lium sulfate aerosol developed granulomata. It may be pertinent at this stage to re view the relative prevalence of the more im portant pulmonary lesions observed in the exposed and in the control series of rats. The total numbers of lesions found have been assembled in Table 5. No refined biostatistical analysis need be invoked to em phasize the fact that the seven criteria listed predominated absolutely in the group of ex perimental animals which had been caused to inhale the beryllium sulfate aerosol as compared with a control series which was not knowingly exposed to any beryllium compound. The seven types of lesion per haps constitute a syndrome specific for beryllium sulfate. It should be further noted that the majority of the lesions manifested themselves only in those 52 animals which were allowed to survive in normal air after their exposure to the beryllium was termi nated. Pulmonary Neoplasia.--The lesions which are probably of the greatest importance be cause of the gravity of their significance are the new growths. Some discrepancy may be noted between the number of tumors detected macroscopicallv and the number identified on microscopy as neoplastic in nature. A total of 89 tumors were seen or palpated at the time of the autopsv on the animals previously exposed to beryllium sulfate (Table 2), while only 76 tumors are listed in Table 3 as histologically neo plastic. "Tbe explanation is simply that in constructing Table 3, only the new growths identified in single midcoronal sections of both lungs were recorded. While it may be apparent that many growths which could not have been identified on naked-eye examina tion are included in the histologically neo plastic group, it should be obvious, too, that many others were not enumerated, because they were not located in the midcoronal plane. The important discovery of this study is that there were such extremeiy large numbers of neoplasmata. The" various types of growth identified have been summarized in Table 6. It is highly probable that all three retesarcomata represented metastases from tumors which probably originated within the abdominal lymphatic nodes. While neither lymphatic system tumors nor primary pulmonary new growths were present in the control series, retesarcomata have occurred from time to time in rats used in other studies. Their presence in the present series of bervllium- Tabu 6.--Neoplasia m Fifty-Two Rats Exposed to Beryllium Sulfate Aerosol for Six Months and Then Living in Normal Air for Periods up to Eighteen Months Neoplasm Adenoma Squamous carcinoma Acinous adenocarctooma Papillary adenocarcinoma Alveoiar-ctil adenocarcinoma Mucicenous tumor Endothelioma Retnarcoma Total No. Metasuses 18 3i 24 2 11 l 7 l l 3i 78 i 4S jm ? Drs eoin ths of be not naeo* 1 hat } Ait nai if' ied is a:a ich nal itic t\v '.ts, to teir .m* Sfd IS BIOLOGICAL actios OF IXHALED BERYLLIUM SULFATE exposed animals, therefore, is not of too grtMt jigninconce. The criteria employed in deciding that a neoplasm was malignant rather than benign uere primarily cytological and histological. In most instances the malignant tumors were markedly cellular with epithelium-derived tissue predominating or even exclusively present. Mitosis was a prominent feature, and generally at least one mitotic figure could be seen per oil-immersion field. No capsules could be demonstrated, the tumors infiltrating and permeating the surrounding tissue. The individual cells generally were larger than the tissue cells of the normal rat. the nuclei being particularly large with prominent nucieoli. Often the nuciei varied greatiy in size and chromatin content, with pailid and pvknotic nuciei sometimes lying adjacent to one another. In most of these tumors this pleomorphism and anaplasia were associated with functional suppression. However, in one series keratin was pro duced. and in another series there was mucoid secretion. No necrosis, ulceration, or hemorrhages occurred in these earlier stages. The tumors remained relatively avascular. Attention should be drawn also EXPERIMENTALLY INDUCEO PULMONARY NEOPLASIA DATA FOR 84 RATS EXPOSED TO B*SO UP TO 6 MONTHS ANO FOR 52 RATS EXPOSED FOR 6 MONTHS ANO THEN TRANSFERRED TO NORMAL AIR RELATIVE PREVALENCE RATE OF Fig. 2.--Graph showing the evolution of peri bronchial alveolar wall epithelialization, lobular septal-cell proliferations, and pulmonary carcino matosis in rats exposed continually to a beryllium sulfate aerosol for 6 months and thereafter re siding in normal air for periods up to 18 months. Only lesions detected in a single midcoronal plane are recorded. 49 * 4 * 3 A. M. A. ARCHIVES OF ISDVSTRIAL HEALTH | to the rr.uiufocai origin of many of these tumors in several ammais. In some oniv one lesion couid be found ; in others there were numerous separate foci of neoplasia. Other features supporting the malignant character of these tumors were the tendency to metastasize and the successful trans plantation of the growths into the sub cutaneous tissues of new host rats. Some of these secondary and tertiary growths produced metastases. The first unmistakably malignant growth was found in one of the animals killed at the end of six months of exposure to the beryllium sulfate aerosol. A second was detected in one of three animals sampled at the end of a week after the cessation of the exposure. Thereafter tumors became progressively more frequent until the 15th month of residence in normal air had been reached, whereafter the prevalence rate de clined rapidly (Fig. 2). The main varieties of tumor found have been classified in Table 6. Not included in these series are a number of focal lesions in which epithelium-lined lacunae were found filled with spheroidal cells (Plate IVD). These cell masses may represent the product of continued multiplication of the epithelial layer and thus be equivalent to an alveolar granuloma. On the other hand, these cell formations may be examples of carcinoma in situ. Squamous carcinomata occurred both with and without keratin production. The kera tin-forming squamous carcinomata appeared to be slow growing. In several instances typical epithelial pearls couid be found. The stratified epithelium characterizing these tu mors was remarkable chiefly for the exces sive size of its nuclei, the acidophilia of the cytoplasm, and the marked propensity for forming flat comified superficial laminae (Plate V.-l). The variety without keratin presented multiple trabeculae covered with stratified epithelium on both surfaces and exhibiting a variable amount of stroma f Plate VB). These epithelioid tumors tended to be relatively expansile, and cell 50 multiplication took place at the free surface. One of these epitheliomatous growtns was 1 associated with a metastatic deposit tn the j hepatic lymph node. The rumors appear tOj have arisen from alveolar wail epithelium which had undergone metaplasia. Several; such foci of stratification could be seen in relation to foam-cell clusters and at the edges of granulomata. At least one of these cancers originated around a partly degen-' erated granuloma. The adenocarcinomata were the most numerous. They occurred in several varie ties. and it seems highly prooabie that these tumors are the malignant analogues of the epithelial hyperplasias already described. The acinous type of adenocarcinoma was most prevalent (Plate VC). Sometimes the tumor consisted simply, on cross section, of numerous circular groupings of proliferating cuboidal cells, and it had expanded mar- . ginally by the multiplication of these hollow acini. The intervening stroma generally re mained scanty, but rarely some collagen was produced. At the tumor edge there fre quently was an abrupt line of transition 1 from anaplastic to normal tissue. The acini varied' from relatively small rings of cells, with scarcely a lumen, to relatively dilated spaces. Both appeared to be equally ma lignant. and sometimes histological lesions of both types were found in the same tumor. The papillary adenocarcinomata probably # were simply variants of the foregoing type. J in which epithelial hyperplasia had pro- J ceeded at such a disproportionate rate that Ij sessile or villous projections of epithelium, fl either with or without any stroma, protraded into the acini (Plate VD). In some ' instances the acini became virtually occluded by the papillae. At their margins these tumors sometimes sent out solid cellular projections into surrounding noncarctnomatous lung tissue. At other times the acinar processes invaded adjacent areas first. The designation of the alveolar cel! carci nomata as such is based on the observation that these tumors were almost soiid masses 4 of malignant cells at the center, their mar- * i - BIOLOGICAL actios of ishaled beryllium sulfate IV ~Z> SrWBW&rZ m ;4 *" * * r# *. . ' i S' rrl' * *u c V1 A - *A *Vi. .1 ** V. (f5^ *- --*----- 4 *, Vo & 1^1 Plate V.--A. pulmonary squamous carcinoma; incipient keratogenic growth with multiple, pearls and proliferating basal layers: Rat 1. exposed to beryllium sulfate aerosol for six months, then in normal air six months, then killed; X 175. B, pulmonary squamous carcinoma; epithe lioid growth showing stratified bilamtnated trabeculae which do not produce keratin: Rat 19, exposed to beryllium sulfate aerosol for 10 months, then in normal air 10 months, then killed; x 250. C. pulmonary acinous adenocarcinoma; acinous adenocarcinoma with small lacunae and delicate collagen bundles: Rat 15. exposed to beryllium sulfate aerosol for 6 months, then in normal air for 18 months, then killed; x 175. D, pulmonary papillary adenocarcinoma: stromacontaining pedunculated papilla in a papillary adenocarcinoma: Rat 56, exposed to beryllium sulfate aerosol for 6 months, then in normal air 10 months, then killed; X 750. THE library of the 51 AMERICAN MEDICAL ASSOCIATION S35 NORTH DEARBORN STREET CHICAOO IQ, ILLINOIS A. M A. ARCHIVES OF IXDCSTRIAL HEALTH Plate VI.--A, pulmonary alveolar cell carcinoma; neoplastic cells derived from septal-cell dements beginning to fill an alveolar space: Rat 6, exposed to beryllium suliate aerosol tor 6 months, then in normal air 10 months, then killed; x 800. B, pulmonary mudgenous tumor: conliuent mucus-containing epithelium-lined loculi (note advanctng columns and peripheral round-cell reaction) : Rat JO. exposed to beryllium sulfate aerosol for 6 months, then in normal air 10 months, then killed: X 10. C. endothelioma of the pleura: Rat 18. exposed to ben-ilium sulfate aerosol for 6 months, then in normal air 12 months: x J50. D. argentophiitc rer.cuium supporting ceils of retesarcoma: Rat 55. exposed to bery llium suiiate aerosol for o mcnti.s. then in normal air 10 months: x J50. 1 BIOLOGICAL ACT'OS' OF IXHALED BERYLLIUM SULFATE gini showing that the component cytologicai eieinents were derived by universal anaplastic proliferation of alveolar lining cells into the puimonarv aiveoli < Plate VIA). Slightly more centrally from the tumor edge, the cells from opposite sides of the alveoli couid sometimes be seen to meet and thus to obliterate the air spaces. At yet other points neoplastic cells were shed into the alveolar lumen. Fundamentally, there* tumor with sequential conduence of mucoid masses. Thus, a multiloculated chairacter was imparted to some of these tumors. Along the expanding edges of the tumor the cvtological elements were better preserved, and it could be seen that the growth invaded surrounding lung parenchyma by means of solid cords of cells which subsequently acquired a lumen in which secretions collected. Such advancing Xo attempt was made by the alveolar-cell tumors to evolve either new acini or to become truly papiilomacous. In some of the les? vigorously developing tumors, the originai alveolar architectonic pattern was preserved, but the air sacs were tilled with neoplastic cells which contrasted sharply There was also a fairly distinct tendency to lymphocytic and plasma-cell reaction to the presence of these invading columns, a feature not observed in connection with other tumors of this series, The source of these mucoid tumors was not readily apparent. There is a strong metrically proliferating alveolar cells tended to form central bodies of polyhedral cells attached to the alveolar wall along one edge but unattached elsewhere. The provisional indications appear to be that these tumors arose out of the lobular alveolar septal-cell proliferations already described. A number of mucigenous carcinomata were discovered. Their malignant nature was heid in doubt, and they were at first classified as mucigenous adenomata until at bronchi (Plate IIS), The single example of an endothelioma or mesothelioma encountered in these animals was located on the pleural surface of the apex of a basal lobe. It showed no tendency to invade the subjacent lung but had spread laterally and into the pleural fissure. In structure it consisted of solid cords of pleomorphic proliferating cells with some superficial palisading. Delicate vascular or lymphatic channels were interposed between could be traced to one of these tumors. The of independent pleural mesotheliomata has of their distended acini belied their danger- this skepticism it is well to record that in cells among those .forming the lining of epi- incipient squamous carcinoma, one papillary thelial membranes. Some of these cells re- adenocarcinoma, and one alveolar-cell carci- the nrigmal acinus denuded ("Plate VIB1. instance could represent a pleural metastasis toward the centers of the tumors, thus creatlnR an illusion of aceilularity here. There was .liso a tendency for interruption of c'nymal tumors, its cvtological and histological features did not closely resemble any of these three lung tumors. While it must be a. m. a. archives of industrial health pleteiv exciude the metastatic origin or the lesion. it does at least lend support to the possibility that the endotheiioma was a fourth independent tumor. For the sake ot completeness, and perhaps because their existence in this series of ani mals may be more than coincidental, refer ence has been made to the pulmonary meta static retesarcomata. These tumors did not differ materially from those which have been seen from time; to time in our own rat colony and in animals obtained from other sources. They were characterized by the exceedingly cellular nature of the tumor and the large size of the polyhedral pleomorphic ceils and of the cell nuclei. A delicate argentophilic reticulum supported most of these ceils i Plate VID). Of more particular interest in the present series of rats is the fact that the pulmonary metastases showed a predilec tion for localization in relation to areas of foam-cell clustering and granulomatosis. This observation helps to illustrate the point that the lymphatic pathways to and from the foam-cell clusters and granulomata were not obstructed by the pulmonary reaction to the inhaled beryllium sulfate. Comment The success with which the features of human chronic beryllium disease have been duplicated in this series of rats deserves comment. For a considerable time the main diagnostic criterion of human chronic beryl liosis has been the gTanuloma. Skepticism has gradually come to surround the validity of the human lesion as a pathognomonic diagnostic criterion because of failure to reproduce this lesion in animals and because of the occurrence of somewhat similar gran ulomata in Boeck's sarcoidosis. Since gran ulomata can be reproduced abundantly in the lungs of at least one animal species by exposures to beryllium sulfate, a feat ac complished in the present experiment, the discovery of similar lesions in personnel of beryllium-utilizing industries becomes the more convincingly significant. The tissue changes which accompanied and preceded its emergence seem, however. The exceedingly minute quantities of beryllium sulfate which effectively induced pulmonary changes in these rats should be noted. It is estimated that at | most each rat could have inspired about 150y of beryllium sulfate during the exposure phase. On bio chemical analysis beryllium was recovered from the lung tissue, as shown in Figure 3 Although beryllium sulfate is a soluble com pound. it is calculated that about 02<~c of the inhaled beryllium evidently was at first tenaciously retained by the pulmonary pro toplasm. and appreciable excretion of the element occurred only after three to five months' residence in normal air. However, *t the initial pulmonary tissue response does not appear to have been a function of the higher pulmonary concentration of beryl lium. Emergence of the biological reaction coincided better with the phase of reduced pulmonary retention of the beryllium, h is possible, therefore, that we are dealing here with either a delayed response to the previ ous phase of maximal concentration or a rebound phenomenon brought about by \\ith-