Document 1g2eMRYn7eo71O6oE8zgNLKqE

Smoking Inhibits Asbestos Clearance13 PLAINTIFF'S EXHIBIT DEBORAH MCFADDEN, JOANNE L. WRIGHT, BARRY WIGGS, and ANDREW CHURG Introduction There is a substantial amount of epide miologic and clinical evidence that as bestos workers who smoke cigarettes have more asbestos-related disease than those v.ho do not smoke. Cigarette smoking workers have a risk of lung cancer, which m some studies is as much as 80 to 90 .imes that of the general population (1-3), and the effects of smoking and asbestos exposure in the genesis of lung cancer ap pear to be synergistic (3). There is also epidemiologic and roentgenographic evi dence that workers who smoke have higher attack rates or progression rates for asbestosis (4). Asbestos-related diseases are in general dose-related, and it has been suggested that smoking may exert its effects in part terfering with the clearance of asi .os fibers from the lung, thereby in sing the effective dose (4). Because smoking has been shown to generally re duce particle clearance in humans (5,6), this is a reasonable suggestion, but one which has never been tested. Methods Thirty female Hartley strain guinea pigs (mean weight, 400 g) were given 1 mg U1CC amosite asbestos in 0.5 ml saline by in tratracheal instillation using a procedure simi lar to that previously published (7). The animals were then divided into the following groups: nonsmokers (n = I0, referred to as "nonsmokers"), smoking commencing on the day after asbestos exposure (n = 10, referred to as "post-smoked"), and smoking commenc ing 2 wk before asbestos exposure and con tinuing for the duration of the experiment (n = 10, referred to as "presmoked"). The smoking apparatus is described and illustrated in reference 8. The machine oper ates by drawing 20 ml of air through a single cigarette and then expelling the 20 ml of smoke into a head-only exposure chamber. Each guinea pig is exposed in a separate cham ber, breathing spontaneously in an awake --'e, and each cigarette feeds smoke to only > imber. In this study, smoking animals ,e exposed to the whole smoke of 10 noner standard cigarettes (200 puffs total, 20 ml smoke/puff) 5 days/wk. The cigarettes were obtained from the Tobacco Manufac turers Council of Canada; this type of ciga SUMMARY To determine whether smoking affects the clearance of asbestos fibers, guinea pigs were given amosite asbestos by intratracheal instillation and divided into 3 groups: (1) nonsmokers, (2) animals exposed to smoke only after asbestos instillation, and (3) animals exposed to smoke both before and after asbestos instillation. Asbestos fiber concentrations and sizes were measured at 1 wk and 1 month after exposure. Between 1 wk and 1 month the asbestos burden in nonsmokers decreased on average 6-fold, whereas both smoking groups failed to show a significant decrease. Over this time period, the mean length of retained fibers increased in the nonsmokers but decreased in both smoking groups. This phenomenon was seen in tissue samples and lavage samples, al though lavaged fibers were consistently shorter than tissue fibers. We conclude that in this model, cigarette smoking impedes asbestos clearance, largely by increasing retention of short fibers. This increased pulmonary fiber burden may be important in the increased disease rate seen in asbestos workers who smoke. AM REV RESPIR DIS 1986; 133:372-374 rette has been previously used in studies on permeability of the airway epithelium (8-10) and is equivalent to a commercial nonfilter brand currently on the market. Animals were not exposed to smoke on the day of asbestos instillation. Half of each group was killed by barbitu rate overdose at 1 wk after instillation and half were killed at 1 month. The lungs were excised, and the right lung was fixed in glutaraldehyde by bronchial perfusion at a pressure of 100 cm of water for 30 min; this pressure is required to overcome the intense bronchospasm that occurs in guinea pigs im mediately after death (11). To determine asbestos fiber concentrations and sizes, we used the procedure that we have previously reported (12). Weighed pieces of lung constituting approximately half of the upper (cranial) and half of the lower (caudal) lobe were dissolved in filtered commercial laundry bleach for 24 h. The preparation was then centrifuged, the supernatant was dis carded, and 307o hydrogen peroxide was added. The solution was allowed to stand for 4 h at 60 C. These procedures destroy the organic matter in the lung; our experience has shown that, using laboratory guinea pig lungs, virtually no organic matter remains and no other mineral particles are found after these steps. The solution was then filtered through an 0.45-p pore size Millipore filter. Randomly selected portions of the filter were cut out and placed mineral side down on carbon/formvarcoated electron microscope grids, the grids were placed on pieces of urethane foam im pregnated with acetone, and the filter was dis solved as capillary action drew the acetone onto the filter. This procedure transfers the particulates to the electron microscope grid. At the time of original tissue sampling, an adjacent piece of lung was dried to constant weight to allow expression of results as fibers/ g dry lung. The left lungs were lavaged with 20 ml of saline in volumes of 10, 5, and 5 ml. The lavagate was filtered and prepared as above for fiber analysis. Only fiber sizes were de termined on the lavage material. Using the electron microscope, approxi mately 200 sequentially encountered fibers/ animal were identified and measured from the tissue samples and 100 fibers/animal from the lavage samples. Fiber identity was con firmed by energy dispersive X-ray spectros copy using a Kevex 7000 system attached to the electron microscope. Grid squares to be examined were selected by assigning every square a number and picking the actual squares using random numbers. Squares were selected and counted until the appropriate number of fibers had been found. For the tis sue samples the fiber concentration was de termined by an algorithm, relating weight of sample, and area of filter examined (12). Blank samples, which were run by going through the above steps but omitting tissue, failed to reveal any amosite fibers. Six hundred ninety-seven fibers of the (Received in original form June 25, 1985 and m revised form September 6, 1985) ' From the Department of Pathology and VBC Health Sciences Centre Hospital, University of British Columbia, Vancouver, British Columbia. Canada. 1 Supported by Grants No. MA8051 and N<x MA7820 from the Medical Research Council of Canada. 1 Requests for reprints should be addressed to Andrew Churg, M.D., Department of PathologF University of British Columbia, 2211 Wesbroo Mall, Vancouver, B.C., Canada V6T 1W5. r SMOKING 1 llCC;. | once sta usual I ratios It pies wet I geometi I (SD)val I 5.1 (2.4) I (length Geon eally us; non (13 I groups, . sons wei * her coiu tially aikruskal lowed b\ von test 1 Pheari' j central! ous gro 1 in table I 1 wk to an apprt I fiber co I the 2 sn proxima , centrati, posure, nificantl either o; 1 mont! I cantly k | of the o I presmok I lower co: animals I The ti I groups a ! in table: | asignifk I tween 1 I both sm< I creased 1 month ! animals : groups r. ' cantlv si j smokers. , and the ; I that were | smoked a were not mean f Group f,onsmoker: Pstsrnokec | Pr?smoked ; JKOKING inhibits asbestos clearance 373 'ea pigs mokers, > smoke easured imokere 'crease, creased ties, atmodel, re. This ibeslos 372-374 as fibers/ 20 ml of ml. The as above were de- I 1 approxi'"ers/ i the from 0 con- pectrosached io res to be ig every e actual ires were ropriate rthetiswas de- eight ol I. Blank hrough ailed to 01 the : I . j i | | ' | . j i ( | I | iirui n j l lie -in '' miKi J X1 UiCC amosite sample were sized as a refer ence standard. Because fiber size distribution is usually lognormal, fiber sizes and aspect fatios for both the standard and the test sam ples were expressed as geometric means and geometric standard deviations. The mean (SD) values for the standard were: for length, 5.1 (2-4) p; for width, 0.31 (2.1) p; for aspect (length to width) ratio, 20 (2.3) p. Geometric size data were compared statisti cally using t tests with a Bonferroni correc tion (13). Because of the small sizes of the groups, nonparametric statistical compari sons were used on the concentration data: fi ber concentrations were ranked and were ini tially analyzed in a one-way layout with a Kruskal-Wallis test (H = 15.8, p < 0.01) fol lowed by a nonparametric multiple compari son test based on ranks (14). Results The arithmetic mean and range of con centrations of amosite fibers in the vari ous groups and time periods are shown in table 1. Comparing each group from 1 wk to L month, the nonsmokers had an approximately 6-fold decrease in mean fiber concentration (p < 0.01), whereas the 2 smoking groups each had an ap proximately 20% decrease in fiber con centration (p = NS). At 1 wk after ex posure, the presmoked animals had sig nificantly lower fiber concentrations than either of the other groups (p < 0.01). At 1 month, the nonsmokers had signifi cantly lower concentrations than either of the other groups (p < 0.01), and the presmoked animals continued to have a lower concentration than the postsmoked animals (p < 0.01). The tissue fiber sizes in the different groups at 1 wk and at 1 month are shown in table 2. In the nonsmokers, there was a significant increase in mean length be tween 1 wk and 1 month (p < 0.001). In both smoking groups, fiber lengths de creased significantly between 1 wk and 1 month (each, p < 0.001). Comparing animals at 1 month, both smoking groups retained fibers that were signifi cantly shorter than those in the non smokers, (each comparison, p <0.001), and the presmoked animals had fibers that were shorter than those in the postsmoked animals (p < 0.001). Fiber widths were not significantly different at 1 TABLE 1 MEAN FIBER CONCENTRATIONS (RANGE) x 10s(g DRY LUNG Grcup One Week One Month Nonsmokers Postsmoked animals Presmoked animats 95 (18-167) 100 (53-232) 50 (14-87) 16 (6-27) 78 (52-99) 40 (17-70) TABLE 2 TABLE 4 TISSUE FIBER SIZES IN MICRONS AS GEOMETRIC MEAN [GEOMETRIC SD] (NUMBER OF PARTICLES MEASURED IN PARENTHESES) Group One Week One Month Lengths Nonsmokers Postsmoked animals Presmoked animals Widths Nonsmokers Postsmoked animals Presmoked animals 5.6 [2 4) (1.020) 5.8 [2.5] (1,016) 5.9 [2.5] (1.013) 6 3 [2.3] (1.010) 5.1 [2.3] (1.046) 4.6 [2.7] (1.030) 0.28 (1.9) (1.020) 0.34 [2.0] (1.016) 0.32 [2.0] (1.013) 0 30 [2.0] (1.010) 0.28 [1.8] (1,046) 0.31 [1 9) (1,030) ABSOLUTE FIBER NUMBERS BY SIZE (MEAN CONCENTRATIONS x KWg DRY LUNG) < 10 m > 10 Nonsmokers 1 week 74 21 1 month 13 3 Postsmoked 1 week 75 25 animals 1 month 66 12 in the nonsmokers (p < 0.001), but in the smoking animals there was either no in crease in fiber length with time or a de crease. The absolute numbers of fibers in the nonsmoking and postsmoked groups divided into those shorter and longer than 10 n are shown in table 4. Aspect Ratios Nonsmokers Postsmoked animals Presmoked animals -------------- -- 19 [2.4] (1.020) 17 [2.5] (1.016) 19 [2.4] (1.013) 21 [2.3] (1.010) 18 [2.3] (1.046) 14 [2.4] (1.030) month. The aspect ratios at 1 month showed that the mean values for non smokers were greater than those for postsmoked animals, which in turn were greater than those for presmoked animals (each comparison, p < 0.01). Similar data for the lavage samples are shown in table 3. Overall, the lavaged fibers were shorter, thinner, and of lower aspect ratio than the tissue samples. Again, the mean fiber length increased TABLE 3 LAVAGE FIBER SIZES IN MICRONS AS GEOMETRIC MEAN [GEOMETRIC SD) (NUMBER OF PARTICLES MEASURED IN PARENTHESES) Group One Week One Month Lengths Nonsmokers Postsmoked animals Presmoked animals Widths Nonsmokers Postsmoked animals Presmoked animals Aspect Ratios Nonsmokers Postsmoked animals Presmoked animals 2.3 [2.6] (498) 2.9 [2.5] (499) 2 7 [2.6] (500) 3 0 [2.3] (501) 2.5 [2.1] (500) 2.8 [2.3] (502) 0 19 [2 3] (498) 0 22 [1 8] (499) 0 20 [1 9] (500) 0.22 [1 8] (501) 0.21 [2.0] (500) 0.20 [2 0] (502) 12 [2 1] (498) 13 [2.2] (499) 13 [2.2] (500) 13 [2 1] (501) 13 [2.1] (500) 12 [2.0] (502) Discussion In this series of experiments, we were in terested in examining the question of whether smoking affects asbestos clear ance. The intratracheal instillation mode! that we have chosen is, like all in tratracheal instillation models, unphysiologic: the doses given are high, and there is nonuniform fiber distribution within the parenchyma (15). Because all animals were treated in the same fashion, we do , not believe that the method of asbestos administration, perse, should affect our results. However, it should be appreciated that different results might be obtained with an inhalation model, and that changes in the amount of asbestos ad ministered and amount of smoking might also change the results. Thus, our com ments on relevance to humans are mean: to be interpreted with caution. Bearing these caveats in mind, it is evi dent that, in our model, smoking de creased asbestos fiber clearance. Thi.% phenomenon was seen in both smoking groups; neither showed any statistical evi dence of fiber clearance between 1 wk and 1 month. By contrast, in the non smokers only approximately one sixth ol the fibers that were found at 1 wk were still present at 1 month. The 2 different smoking protocols were used because we wished to determine whether exposure to cigarettes commenc ing before exposure to asbestos, i.e., the usual situation in humans, would have the same effect as acute exposure com mencing only after exposure. Although the effects of smoking were parallel in both groups (no clearance), the pre smoked group initially had a lower fiber burden than the group that was smoked only after asbestos instillation. A po< ble explanation for this difference is that n MCFADDEN. WRIGHT, WIGGS, AND CHURG noking sensitizes the guinea pig airways a that bronchospasm is rapidly induced estos administration, and more of ' uministered dose is deposited very *mally, from whence it is rapidly leared, and less of the administered dose caches distal airways and parenchyma, tence the apparent lower retained dose .t 1 wk in the presmoked group. This equence of events has been shown ex>erimentally in humans inhaling bronhoconstrictors and particles (16). These problems do not, however, affect our :onclusion that clearance is impaired in he presmoked animals. A second conclusion from our data is that the increased fiber burden in the smokers compared to that in the non smokers at 1 month is composed in large part, although not entirely, of short fi bers. This phenomenon is best seen in table 4, which we have restricted to the nonsmokers and the postsmoked ani mals, because these groups have almost identical mean concentrations of fibers at l wk. In the smokers, approximately half of the fibers longer than 10 \i were cleared by 1 month, whereas virtually no fibers shorter than 10 n were cleared. However, the clearance failure was not ly confined to short fibers since the King animals still retained 4 times the . iber of long fibers than did the non smoking animals. It has been shown (17, 18) that, ordi narily, short fibers are cleared more ef fectively from the lung than are long fibers, and hence retained fiber length ap pears to increase with time. This phenom enon was seen in our nonsmoking ani mals for both tissue fibers and lavaged fibers. In both groups of smoking ani mals, on the other hand, the length and aspect ratio of retained tissue fibers de creased from 1 wk to 1 month, and a simi lar process was seen in the lavaged fibers. Presumably, the differences in fiber sizes between smoked and nonsmoked ani mals reflect smoke-induced interference with mucociliary transport and/or mac rophage clearance of fibers; phagocyto sis by macrophages is thought to be the first step in clearance of short asbestos fibers from the lung parenchyma (17, 18). It is interesting that the presmoked group had the shortest mean tissue fiber lengths and lowest aspect ratios at 1 month, an observation which may imply that the in terference with clearance mechanisms is related to cigarette dose. There is some data in humans that clearance times are directly proportional to dose as measured by pack-years of smoking (6). The actual mechanisms of carcinogen esis and fibrogenesis by asbestos are not clear, but from observations of disease incidences in laboratory animals, it is of ten claimed that long fibers are consider ably more pathogenic than are short fibers in the genesis of asbestosis, carci noma of the lung, and mesothelioma (19-21). However, the suggestion has also been made that a large number of short fibers might be just as dangerous as a few long fibers (22). Because our guinea pigs retain increased numbers of both long and short fibers, we cannot deter mine whether the increased disease rates seen in smoking asbestos workers should be imputed to long or short fibers; how ever, our data do suggest that short fibers should not be totally disregarded as pos sible agents in the pathogenesis of asbestos-related disease. References 1. Selikoff IJ, Hammond EC, Seidman H. Mor tality experience of insulation workers in the United States and Canada, 1943-1976. Ann NY Acad Sci 1979; 330:91-116. 2. Selikoff IJ, Seidman H, Hammond E, Culyer S. Mortality effects of cigarette smoking among amosite asbestos factory workers. JNCI 1980; 65:507-18. 3. Hammond EC, Selikoff IJ, Seidman H. As bestos exposure, smoking, and death rates. Ann NY Acad Sci 1979; 330:473-92. 4. Weiss W. Cigarette smoke, asbestos, and small irregular opacities. Am Rev Respir Dis 1984; 130:293-301. 5. Cohen D, Arai SF, Brain J D. Smoking impairs long-term clearance from the lung. Science 1979; 204:514-6. 6. Bohning DE, Atkins HL, Cohn SH. Long-term particle clearance in man: normal and impaired. Ann Oceup H\g 1982; 26:259-71. 7. Filipenko D, Wright JL, Churg A. Pathologic changes in the small airways of the guinea pig after amosite asbestos exposure. Am J Pthol 1985; 119:273-8. 8. Simani AS, Inoue S. Hogg JC. Penetration of respiratory epithelium of guinea pigs following ex posure to cigarette smoke. Lab Invest 1974;31:75-87. 9. Hulbert WC, Walker DC, Jackson A, Hogg JC. Airway permeability to horseradish peroxidase in guinea pigs: The repair phase after injury by ciga rette smoke. Am Rev Respir Dis 1981; 123:320-6. 10. Boucher RC, Johnson J, Inoue S, Hulbert W. Hogg JC. The effect of cigarette smoke on the permeability of guinea pig airways. Lab Invest 1980; 43:94-100. 11. Lai Y-L, Lamm WJE, Luchtel DL, Hildebrandt J. Massive postmortem bronchoconstriction in guinea pig lungs. J Appl Physiol 1984; 56:308-14. 12. Churg A. Asbestos fiber content of the lungs in patients with and without asbestos airways dis ease. Am Rev Respir Dis 1983; 127:470-3. 13. Miller RG. Simultaneous statistical inference. 2nd ed. New York: Springer-Verlag, 1981; 67-70. 14. Hollander M, Wolfe D. Nonparametric statisti cal methods. New York: John Wiley & Sons, 1973; 114-36. 15. Brain JD, Knudson DE, Sorokin SP, Davis MA. Pulmonary distribution of particles given by intratracheal instillation or by aerosol inhalation. Environ Res 1976; 11:13-33. 16. Svartengren M, Philipson K, Linnman L, Camner P. Airway resistance and deposition of parti cles in the lung. Exp Lung Res 1984; 7:257-69. 17. Holmes A, Morgan A. Clearance of anthophyllite fibers from the rat lung and formation of asbestos bodies. Environ Res 1980; 22:13-21. 18. Morgan A. Effect of length on the clearance of fibers from the lung and on body formation, in: Wagner JC, ed. Biological effects of mineral fibers. Lyon: 1ARC, 1980; 329-35. 19. Selikoff IJ, Lee DHK. Asbestos and disease. New York: Academic Press, 1978. 20. Davis JMG, Beckett ST, Bolton RE, Collings P, Middleton A. Mass and number of fibers in the pathogenesis of asbestos-related lung disease in rats. Br J Cancer 1978; 37:673-87. 21. Stanton MF, Layard M, Tegeris A. Miller E, May M, Morgan E, Smith A. Relation of particle dimension to carcinogenicity in ampiiibole asbe stoses and other fibrous minerals. JNCI 1981; 67:965-75. 22. Pott F. Animal experiments on biological ef fects of mineral fibers, in: Wagner JC, ed. Biologi cal effects of mineral fibers. Lyon: IARC. 1980; 261-72. i i i i Nitr and Inte JOHN PETEF and G Alvet nary v: muscui cle mu (1-3). .' remodt and ri: velop ( role in in pademphv kyphoventila and in (8). A) hypoxi, mans . extent produv. This qt patient of hypi 12). \\ whethc day wo and pi Hyp pul mo inducr oxia a: strieiio `ascii i entry strict!* chron: (15) or "e a i pretre.: entry nifediascu !ar h\- hypos