Tobacco Documents Online

Chronic Inhalation Studies in Mice. 11. Effects of Long-Term Exposure to 2R1 Cigarette Smoke on (C57BL/Cum X C3H/AnfCum)F1 Mice'•2.3 Carol J. Henry °-s and Richard E. Kouri 4.6•7 ABSTRACT-Standardized exposure conditions with Kentucky reference 2R1 cigarettes were used to expose 2,053 (C57BVCum X C3H/AnfCum)Fl female mice (nose only) to fresh, whole cigarette smoke. In addition, 1,014 mice were sham-exposed, and 449 mice were held as shelf controls. The protocol entailed expos- ing mice to smoke (or sham-exposure) on a daily basis, 5 days/ week, for 110 weeks and observing remaining mice until death. A large number of animals was used so that the smoke generation and animal-holding systems could be tested and evaluated and yet provide significant numbers of animals for exposure to cigarette smoke for a major portion of their lifetime. Deposition of smoke particulates was estimated to be about 125-200 µg total particulate matter/lung/day. The only lung cancers observed were diagnosed as alveolar adenocarcinomas (AAC). A total of 19 of 978 smoke- exposed mice and 7 of 651 sham-exposed mice were observed with AAC. The difference between the smoke- and sham-exposed grou0s was not statistically significant at P5.05, but the data sug- gested that the tumors occurred with a shorter latency in the smoke-exposed group (P=.10). The data were analyzed by various methods, including analysis of subsets of the population of ani- mals. A significant increase in the incidence of lung cancer was observed in one subset; however, this difference was not found in the population as a whole or as a result of any other analyses. Under these exposure conditions, 2R1 cigarette smoke would seem to have weak carcinogenic activity in mouse lung tissue. Other changes associated with smoke exposure were increased incidence of pigmented alveolar macrophage accumulation, otitis media, and head and neck fibrosarcomas. However, the incidence of nephritis, hematopoietic cancers (e.g., leukemias, lymphosar- comas, and reticulum cell sarcomas), and pulmonary congestion was significantly higher in the sham-exposed animals.-JNCI 1986; 77:203-212. The experimental design of studies to evaluate the long-term effects of exposure to cigarette smoke in an animal model system must attempt to simulate human exposures. The criteria used to design such studies should be based on those reported to be important in cigarette smoke-associated human diseases. Those fac- tors include high exposure to fresh mainstream cigarette smoke on a daily basis for a major portion of the life- time of the individual (2, 3). Even under these condi- tions, however, resultant smoke-associated diseases have been observed in only a small fraction of the exposed individuals. Our laboratories have been involved in the development of animal model systems and inhalation equipment to assess the biological effects of acute arid chronic inhalation of cigarette smoke. In the develop- ment of the mouse model system, two approaches were taken: 1) quantitation of short-term effects known to be Ctr I-".^'-~r-1'Qdred J D. B. ~.chen associated with smoke exposure and 2) quantitation of long-term effects of model chemicals on carcinogenesis of the respiratory tract (4-6). The experimental condi- tions in this chronic smoke inhalation study were designed to maximize the responses known to be in- volved from the short-term effects and to observe the results of long-term exposure to cigarette smoke. Mice were chosen as the animal model for these stud- ies and, in particular, (C57BL/Cum X C3H/AnfCum)Ft mice. This choice was based on several factors: a) avail- ability of large numbers of animals; b) susceptibility to the induction of lung cancer (squamous cell carcinoma, AAC, poorly differentiated carcinoma, etc.) by model chemical carcinogens (4, 6), which was linked to geneti- cally determined differences in hydrocarbon metabolism capacity (7); c) availability of animals that were free of common adventitious agents, in particular, Sendai virus and pneumonia virus of mice; d) ease and cost effective- ness of handling and manipulating on a daily basis; e) considerable background known on longevity and spon- taneous expression of disease (4); f) smoke particulate deposition characterized with nose-only exposure equip- ment (5); g) certain characterized short-term effects such as induction of aryl hydrocarbon hydroxylase (7), induc- tion of sister chromatid exchange (8, 9), increase in per- ABBREVIATIONS USED: AAC=alveolar adenocardnoma(s); ACN=alve- olar compressing nodules; ANCN =alveolar noncompressing nodules; COHb=carboxyhemoglobin; PAMA=pigmented alveolar macrophage accumulation(s); TPM=total particulate matter. t Received August 19, 1985; revised January 27, 1986; accepted February 24, 1986. 2Supported by contract CTR•0030 with The Council for Tobacco Research-U.S.A., Inc. 3 Presented in part at the Thirteenth International Cancer Congress, Seattle. WA, September 8-15, 1982 (1). The interpretations and opinions are those of the authors and not necessarily those of The Council for Tobacco Research. 4 Microbiological Associates Inc., 5221 River Rd.. Bethesda, MD 20816. SAddress reprint requests to Dr. Henry at her present address: ICF Incorporated, International Square, 1850 K St., N.W., Washington, DC 20006. 6Present address: International Biotechnologies, Inc., 275 Winchester Ave., New Haven, CT 06511. 7 We thank Dr. Leonard Billups and Dr. William Hall for pathology support; Ms. Denise Avery, Mr. David Dansie, and Mr. H. Doyle Mul- linax for technical support; Mr. Robert Kyanko for computer pro- gramming; and Ms. Cynthia Whitfield for typing the manuscript. JNCI, VOL 77. NO. 1, JULY 1986 203

Chronic Cigarette Smoke inhaiation Studies 2Q: filtered experimental- reference cigarettes and have been characterized under standard Federal Trade Commission orocedures at Oak Ridge National Laboratories, Oak Ridge, TN. Under these conditions, the 2R1 cigarette delivered approximately 44 mg TPM, 2.4 mg nicotine, and 21 ml of carbon monoxide (16). Cigarettes were removed from frozen storage and conditioned at 70-75°F and 60% relative humidity for at least 18 hours before use. Animal weights.-All mice were weighed at monthly intervals at approximately the same time of day, 2 days after cages were changed. Smoke- and sham-exposed animals were weighed at least 2 hours after exposure. Data were recorded for each animal, and the mean and standard deviation were calculated for each group at monthly intervals. TPM deposition and COHb levels.-As determined from parallel radioactive dosimetry experiments, smoke particulate deposition was determined to be between approximately 125 and 200 µg TPM/day/mouse lung for this exposure regimen using 6-8 puffs/cigarette. [See Henry et al. (5) for a description of the methods and results.] COHb levels were determined at monthly intervals. Blood from the retro-orbital sinus was collected from 3-5 mice/group. Natelson blood-collecting tubes (175 µl; Sherwood Medical Industries, Inc., St. Louis, MO) were prepared shortly before use by rinsing the tubes with a drop of sodium heparin (10,000 U/ml; Abbott Laboratories, Baltimore, MD). After blood collection, the tubes were sealed with Critocaps, stored on ice, and analyzed within 30 minutes using an IL-CO-Oximeter (IL-282; Instrumentation Laboratory, Inc., Lexington, MA). Necropsy.-Mice were observed twice daily for evi- dence of illness or respiratory distress. Dates and circum- stances of death were recorded for all mice. Nonauto- lyzed tissues from mice found dead and those killed when moribund were examined microscopically. Lungs were fixed in situ with approximately 1.5 ml of 10% buf- fered Formalin by infusion via the trachea. Lungs were ligated at the trachea, and the thoracic viscera were removed as a single unit and examined grossly. Lung, trachea, esophagus, and thoracic lymph nodes were sec- tioned (6 µm) as a unit at 3 levels, using a frontal plane of section. During the 1st year, respiratory tissues (lungs, larynx, trachea, nasal cavity, and middle ear) and abnormal tissues were examined microscopically from a random sampling of 13% of the animals that died. Dur- ing the 2d and 3d years, respiratory tissues and any abnormal tissues from all animals were fixed and stained. Respiratory tract tissues from 96% of the ani- mals were examined microscopically. In addition, 16 other tissues from major organs were examined micro- scopically from a random sampling of 10% of the ani- mals. These tissues included salivary glands, cervical lymph nodes, spleen, liver, kidneys, adrenal glands, uri- nary bladder, ovaries, uterus, heart, stomach, large and small intestines, pancreas, and thymus. Tissues were examined microscopically from a total of 987, 659, and 369 mice in the smoke-exposed, sham-exposed, and shelf-control animals, respectively. Morphological criteria.-A brief description of the lesions observed in these studies is presented below. AAC occurred as discrete grayish-white, firm masses, located in the peripheral portions of the lung, and were often multiple. Some tumors occasionally showed pleu- ral invasion and metastasis. Adenomas are not included in this category. Pulmonary adenomas were scored in two categories: ACN and ANCN. This classification scheme was devised as part of other on-going studies to determine the possi- ble progressive nature of adenomatous lesions to carci- nomas. Details of the chemical inducibility, time course, and transplantability of the ACN and ANCN will be reported elsewhere; however, ACN, as well as AAC, grew into tumors after transplantation into newborn BC3Fl/- Cum mice in about 50% of the cases (Henry CJ, Billups LH, Hall WC, et al.: Manuscript in preparation). Histo- logically, ACN and ANCN were differentiated on the basis of size and extent of compression of the surround- ing lung parenchyma. ACN were rounded masses of hyperplastic alveolo- genic cells, about 2-3 mm in diameter, that did com- press the surrounding parenchyma. Nodules located in the most peripheral portions of the lung tended to invade the pleura. ANCN were similar to ACN, but they did not compress the surrounding parenchyma. These nodules were barely visible grossly. Mitoses were ob- served occasionally. Congestion in the lung was observed as dilated pul- monary veins and capillaries, which were engorged with blood. The congestion could have been the result of an active process resulting in lung irritation, a passive pro- cess resulting in reduced flow of blood from the lung, or an inadvertent delay in necropsy after death. Head and neck fibrosarcomas were malignant neo- plasms of fibrous tissues. They were unertcapsulated and often showed scattered yellow areas of necrosis or red- dish foci of hemorrhage. Many of these observed in this study were highly aggressive, metastatic tumors, com- posed of cells with large spindle-shaped nuclei and scant cytoplasm. Once established, these tumors grew rapidly, invaded veins, and metastasized to other tissues, includ- ing the lung. Hematopoietic cancers were grouped together as can- cers of the hemic and lymphatic systems. They included lymphosarcomas, reticulum cell sarcomas, and lympho- cytic leukemias. Nephritis included all inflammatory lesions of the kidney, glomeruli, and renal pelvis. The inflammatory lesion can be caused by adventitious agents, foreign materials, or other processes. Otitis externa is an inflammation of the external ear. It is usually considered an incidental finding in the mouse. Otitis media is an inflammation of the middle ear and was relatively severe. Usually of bacterial origin, it can be due to ascending infections along the eustachian tube. The eustachian tube, lined by ciliated epithelium, JNCI. VOL. 77. NO. I. JULY 1986

, The surviving fraction of mice as a function of time on test is presented in text-figure 1. These curves were corrected for those animals that were determined to be andomly removed from the study (i.e., they died of exposure-equipment-related factors). The survival curves of the smoke- and sham-exposed groups decreased more rapidly than those of the shelf-control animals, indicat- ing that the daily smoke exposure or sham-exposure caused a detectable level of morbidity and mortality. After 80 weeks on test, however, the shelf-control group was observed to die at a more rapid rate than the smoke- or sham-exposed mice. This was also the time when the shelf-control animals were losing body weight (see next section and text-fig. 2). Body weights.-Body weight analyses of the smoke- exposed, sham-exposed, and untreated shelf-control ani- mals are presented in text-figure 2. The shelf-control mice gained weight until 50-60 weeks and then gradu- ally declined in weight during old age. The rate of weight gain of the shelf-control animals was signifi- cantly greater (P<.05) than that of the smoke- and sham- exposed mice. At about 80 weeks on test, the shelf- control mice had an average weight of 50 g. No differ- ence between the mean body weights of the smoke- and sham-exposed mice was found over the course of the study. Smoke- and sham-exposed mice gained weight slowly until about 60 weeks on test, after which their weight gradually declined. The average maximum weight of BC3F1/Cum mice in the smoke- or sham- exposed groups was 32 g, which was s,een at about 60 weeks on test. Chronic Cigarette Smoke Inhalation Studies 201, Clinical signs.-Mice were observed during and im- mediately after daily exposure to smoke and sham- treatment. For the smoke-exposed animals, while a cer- tain amount of adaptation occurred, the animals never accepted smoke exposure without some agitation or struggling in the holder. They occasionally demonstrated some shallow breathing or gasping during exposure. Immediately after exposure, mice in the smoke-exposed groups were often lethargic, ataxic, and hypothermic. The sham-exposed animals also demonstrated agitation and struggling while restrained in the holders, but generally they appeared normal after exposure. After 20-30 weeks of exposure, certain animals were observed to have reddened skin and worn-away hair around the neck area that fits into the stocklike holder. In 30-40% of the cases, these irritated areas progressed to open sores as exposures continued. Periodically, during the 2d year of the study the mice were rested and not exposed. Such rests enabled the neck cuts to heal. The length of the rests varied from 1 day to I week. Polycar- bonate inserts for the stocklike restraints were found to cushion the neck slot and could be effective in limiting such neck abrasions. It is recommended that such inserts be used for any long-term studies where animals of dif- ferent ages and/or sizes may be used. Histopathologic Observations The shelf-control mice died carrying a similar spec- trum of diseases to that reported previously for this hybrid strain of mice (4); i.e., about 60% died of neoplas- SMOKE EXPOSURE STOPPED WEEKS ON TEST TEXT-FIGURE L-Surviving fraction of mice presented as a function of time on test. Mice included in this analysis were those that died or were killed when moribund (nonrandom deaths; see text for definition). . JNCI, VOL. 77. NO. 1. JULY 1986

Chronic Cigarette Smoke Inhalation Studies 20s TABLE 2.-Summary of incidence of lesions in BC3Ft/Cum miee during long-term exposure to 2R1 cigarette smoke No. of anim als Lesion Smoke-exposed Sham-exposed Highest level of significance° Present Total Present Total Respiratory tract'' AAC 19 978 7 651 Smoke > sham; P=.10 at 813 days AAC + ACN 49 978 28 651 Smoke > sham; P=.09 at 813 days ANCN + ACN 38 978 21 651 Smoke > sham; P=.39 at 757 days AAC + ACN + ANCN 69 978 38 651 Smoke > sham; P=.18 at 813 days PAMA 158 978 1 651 Smoke >> sham Congestion 183 978 158 651 Sham > smoke; P<.05 at 412 days Rhinitis 14 502 8 319 Smoke > sham; P=.32 at 869 days Otitis media 66 579 35 371 Smoke > sham; P5.05 at 421 days Otitis externa 140 579 78 371 Smoke > sham; P=.14 at 561 days Nonrespiratory tract Nephritis 27 179 26 110 Sham > smoke; P=.16 at 953 days Head-neck fibrosar- 29 987, 8 659 Smoke > sham; P:5.05 at 841 days comas Hematopoietic cancers 125 987 126 659 Sham > smoke; P:5.05 at 729 days "The highest level of significance for the difference between the smoke- and sham-exposed groups and the time at which the difference occurred are given. These data were determined from chi-square analysis of each of the lesions or groups of lesions at 4-wk intervals over the course of the study (up to 1,120 days on test). "Respiratory tract tissues included lung, larynx, trachea, nasal cavity, and middle ear (see "Materials and Methods"). "Materials and Methods"). Thus the total numbers of animals that died could be divided into those that died of factors unrelated to their lesions (i.e., random factors) and those that likely died as a result of their lesions (i.e., nonrandom factors). Assuming that the animals died as a result of these tumors yields the analysis shown in table 3. In this actuarial analysis, the total number of animals at risk was 504 and 379 in the smoke- and sham- exposed groups, respectively. Analysis of this subset of the population showed that there was no difference in the incidence or latency of AAC between the smoke- and sham-exposed mice. Table 4 presents the results of the analysis of another subset of the population, where only those animals that were determined to be randomly taken off the test were studied. In this case the numbers of animals found car- rying lung tumors were compared to the numbers of animals that were observed not carrying the tumor. In this analysis, a total of 7 animals were found with AAC, and all were in smoke-exposed animals. The final inci- dence of 7 AAC of 474 smoke-exposed animals is signifi- cantly higher than 0 AAC of 272 sham-exposed animals (chi-square statistic = 4.18; P=.04). The month-by- month cumulative probability, however, never reached a level of significance of P5.05 (see table 4). Similar analyses were done on all the lesions listed in table 2. The respiratory tract lesions found to be signifi- cantly higher in smoke-exposed mice were PAMA and otitis media. PAMA was found only in the smoke- exposed animals, and approximately 35% of all mice that died during the last 4 months of smoke exposure had this lesion. For some unknown reason, pulmonary congestion was found to be higher in the sham-exposed animals than in the smoke-exposed animals. Certain other nonrespiratory tract lesions were also analyzed by the methods described above. Table 2 shows that the incidence of head-neck fibrosarcomas was sig- nificantly higher in the smoke-exposed "animals than in the sham-exposed animals. Analysis by actuarial methods also showed significant differences (data not shown). However, the incidence of nephritis and hema- topoietic cancers (i.e., leukemias, lymphosarcomas, and reticulum cell sarcomas) was significantly higher in the sham-exposed animals than in the smoke-exposed mice. This latter observation would seem to be an example of competing risks, when the occurrence of a specific lesion was altered by the simultaneous expression of another disease process that removed the animal from the study. DISCUSSION The results of this study suggest that 2R1 cigarette smoke has weak carcinogenic activity in mouse lung tissue. No bronchogenic squamous cell carcinomas were observed in any of these animals. AAC, as well as other alveologenic lesions, were observed in smoke-exposed animals, and the incidence was higher in the smoke- exposed animals than in the sham-exposed animals. The difference in incidence of these AAC and other alveologenic lesions, however, never reached the level of significance of P5.05. The data suggested that the la- tency for the occurrence of these tumors is shorter in the smoke-exposed animals (P=.10; see table 2). Various methods of analyses were performed on these data, including incorporation of other alveologenic lesions with the lung cancers and analysis of subsets of the population of animals that were categorized as to cause of death (i.e., random vs. nonrandom). In general, these analyses yielded no significant differences between the smoke- and sham-exposed groups. However, the analysis of one subset of the population, those animals taken off test randomly, did yield an interesting observa- JNCI. VOL. 77, NO. 1, JULY 1986

cumulative dose of smoke. These data regarding PAMA give biological evidence that the smoke particulates -eached the alveolar regions of the lung. Further evi- dence that the smoke particulates reached other areas of the respiratory tract can be seen from the increased incidence and decreased latency of otitis media. The eu- stachian tube with its ciliated epithelium was_ likely a portal of entry for these inhaled materials and would seem to have been impacted by smoke exposure. Exposure to 2R1 cigarette smoke also resulted in significantly increased incidence and decreased latency of head and neck fibrosarcomas. The mechanism by which these tumors develop is unclear. Whether such factors as localized deposition of smoke particulates in this area and/or the relative severity of neck cuts or abrasions caused by chronic restraint played a role in the formation of these tumors is not known at this time. In rats, chronic exposure to cigarette smoke has been asso- ciated with increased incidence of tumors of the forelegs (21). In these studies the rats were exposed to smoke in whole-body tubes. It was suggested that these tumors resulted from the combination of smoke particulate deposition on the forelegs and the presence of abrasions on the forelegs caused by the method of restraint. The dose of cigarette smoke given in this study was limited by the inherent toxicity of this particular cigarette smoke for this strain of mice. Following expo- sure to 6-8 puffs of 2R1 cigarette smoke, the dose was estimated to be 125-200 ug TPM/lung/day. Previous studies with another cigarette type (high tar and low nicotine) have shown that up to 1 mg TPM/lung/day was given (5). Additionally, the exposure regimen uti- lized about the first one-half of the 2R1 cigarette. It is not known whether the first one-half of the cigarette has different biological activity compared to that of the other one-half. Thus the smoke exposure conditions used in this study (low dose and first one-half of the cigarette) may not have been optimal to detect signifi- cant biological or carcinogenic activity. In drawing conclusions from this study, several alter- natives are possible. The first is that the mouse may not be a good model for long-term human cigarette smoke inhalation studies. Mice are obligate nose breathers, have a turbinate structure different from that of humans, have a shorter life-span, do not actively inhale smoke as do humans, and are more sensitive to certain smoke- associated chemicals (e.g., nicotine) that limit the dose of smoke that can be given. All of these factors tend to mitigate against rodents as an implicit model for human inhalation studies. On the other hand, the mouse appears to be a good model for the study of the effects of smoke inhalation as noted by the following; Lung and kidney aryl hydrocarbon hydroxylases are increased in smoke-exposed mice (6, 7, 10); the number of sister chromatid exchanges in bone marrow cells is increased in smoke-exposed mice (8, 9); PAMA is increased in smoke-exposed mice; DNA repair capacity of lung tissue from smoke-exposed mice is approximately 50% inhib- ited (12); DNA synthesis in lung tissue from smoke- exposed mice is increased up to twentyfold (10); and Chronic Cigarette Smoke Inhalation Studies 211 squamous cell carcinoma can be produced in these mice at high incidence with known chemical carcinogens (4, 6). A further alternative to this latter observation is that polycyclic aromatic hydrocarbons are poor model chemi- cals to use for biologic activity of cigarette smoke. In general, rodents have not been observed to develop high incidences of lung cancer following exposure to cigarette smoke [see reviews (22, 23)]. Notable among these many studies is that bronchogenic squamous cell carcinoma has been observed only in 1 rat (F344) after exposure to cigarette smoke (21). In this case, 10 respiratory tract tumors were observed in 7 of 80 animals (9%), with the difference between smoke-exposed and sham-exposed rats reported to be significant at P<.05. The tumors were reported to be I adenocarcinoma and I squamous cell carcinoma of the nasal cavity and 5 adenomas, 2 alveologenic carcinomas, and I squamous cell carci- noma of the lung (21). Extensive studies by Dontenwill et al. (24, 25) using randombred Syrian golden hamsters have indicated sig- nificant changes in laryngeal tissues after smoke expo- sure, including laryngeal carcinoma, with little or no changes reported in the lung. Using inbred Syrian golden hamsters, Bernfeld et al. (26, 27) have reported neoplastic changes in the larynx of hamsters exposed to smoke, with few changes reported in the lung. Broncho- genic squamous cell carcinomas have not been reported in either mice or hamsters after exposure to cigarette smoke, although the incidence of AAC has been reported increased in mice (28, 29). A second alternative conclusion to this study is that the smoke exposure conditions did not reflect the "true" biological potential of this 2R1 cigarette smoke. Ap- proximately the first one-half of the cigarette was used, and relatively low levels of TPM were deposited,,~tlainly because the high nicotine content of the cigarette limited the dose. Additional studies with other cigarette types will have to be done to test this •alternative. A third alternative is that cigarette smoke alone has fairly weak biologic activity and that the data in rodents reflect this activity. The high biological activity and response observed in human smokers (23, 30) may be aa result of a combination of factors, which so far have failed to be duplicated in the laboratory setting. REFERENCES (1) HENRY CJ, Btt.r.UPS LH, HALL WC, et al. The effect of lifetime exposure to whole cigarette smoke in BC3Fl/Cum mice. In: Proceedings of the Thirteenth International Cancer Congress, Seattle, WA. Geneva: International Union Against Cancer, 1982. (2) PIKE MC, HENDERSON BE. Epidemiology of polycyclic hydro- carbons: Quantifying the cancer risk from cigarette smoking and air pollution effects. In: Gelboin H, ed. Polycyclic hydro- carbons and cancer. New York: Academic Press, 1981:317-334. (3) DoLL R, PETO R. The causes of cancer: Quantitative estimates of avoidable risks of cancer in the United States today. JNCI 1981; 66:1191-1308. (4) HENRY CJ, BILLUPS LH, AVERY MD, et al. Lung cancer model system using 3-methylcholanthrene in inbred strains of mice. Cancer Res 1981; 41:5027-5032. JNQ. VOL. 77, NO. 1, JULY 1986

ited (12); DNA synthesis in lung tissue from smo e- system using 5-m avnncarl rnirP is inrrPased un to twentvfold (10); and Cancer Res 1981; 41:5027-5032.