Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
B Case Studies Establishing Active Agents armor Interactions In Complex Mixtures SULFUR DIOXIDE AND SUSPENDED PARTICULATE MATTER The case of sulfur dioxide (SO2) and particulate matter (PM) in ambient air and their effects on mortality and respiratory disease morbidity is a classic case of exposure to complex mixtures in which identification of causal factors has proved extremely difficult. Epidemiologic research has been critical in deter- mining the relative roles of SO2 and PM, whereas laboratory inhalation studies have been critical in identifying active components in the PM. An important caveat is that the critical experimental inhalation studies were done on humans and animals not normally used in standardized bioassays (i.e., dogs, rabbits, and guinea pigs). Also, some of the effects assays particle clearance rates, detailed airway morphometry, and CO-diffusing capacity are not used in conventional rodent bioassay tests. Unconventional animal models and non- routine effects measures were needed because available animal models for the human health effects of interest, such as chronic bronchitis and exacerbations of asthma, have not been widely accepted as relevant to human disease. Air pollution health effects have been associated with coal smoke in London in a qualitative sense at least since the thirteenth century, when Edward I banned the use of coal during sessions of Parliament. However, sustained ef- forts to control coal-smoke pollution did not begin until the London "killer fog" of December 1952 (Great Britain, 19541. Because the effluent from coal combustion contains both SO2 and PM, scientists long debated whether SO2 or PM was the causal agent or whether the effects were due to their combined actions, or synergism. In the absence of clear evidence in the 1950s and 1960s, various authorities responsible for air pollution control made different judg- 133
134 APPENDIX B meets. In the United Kingdom, authorities made efforts to reduce PM, espe- cially the black-smoke component, which was cut by about a factor of 2 be- tween 1952 and 1962. The reduction in SO2 over the same period was substantially smaller. Later reviews of the changes in mortality and morbidity indexes during this period tended to implicate PM as a more important health stressor than SO2. Time-series analyses of daily mortality in New York City (Schimmel, 1978; Schimme] and Murawski, 1975, 1976; Buechley et al., 1973) indicated that mortality increased with increasing SO2 concentrations for both the periods 1963-1966 and 1970-1972. Because the absolute SO2 concentrations dropped by about a factor of 3 over this period, it is highly unlikely that SO2 was the causal factor for the similar association with mortality in the different periods. Rather, the daily fluctuations in SO2 appeared similar to daily fluctuations in other pollutants from the same or similar sources, and other pollutants in the pollutant mixture were probably more directly related to the observed effect. This leads to the conclusion that PM or PM components are better indexes of exposure. The active agents in PM have not yet been clearly identified. Ozkaynak and Spengler (1985) performed regression analyses between daily mortality and four different components of PM: total suspended particulate matter (TSP), inhalable particulate matter (IP), fine particulate matter (FP), and SO42- . TSP has an upper cut-size of ~ 40-60 Am aerodynamic diameter, depending on wind speed and direction. IP has an upper cut-size of 15,um, which is relatively independent of wind speed and direction. FP has an upper cut-size of 2.5 ~m. SO42- is almost all within the FP fraction and varies from less than 10% to more than 50% of FP. All four showed positive regression with daily mortality, with correlation coefficients increasing in the order of TSP < IP < FP < SO42-. Only FP and SO42- had statistically significant coefficients. Although SO42- might be the best predictor of health effects among the commonly measured indexes of PM, it is probably not the active agent. Several physiologic effects of sulfate aerosols have been shown to depend on their H + content. These include specific airway conductance (Utell, 1985) and bron- chial mucocilia~y particle clearance (Schlesinger, 19851. Thus, SO42- might be acting as a surrogate for the strong acid associated with it. Unfortunately, correlation studies between H+ and health effects cannot be performed, be- cause there are too few data available on ambient H+ concentrations (Lipp- mann, 19851. Experimental inhalation studies with SO2 and PM have not played any major role in elucidating the role of these pollutants in population mortality and mor- bidity experience. The only documented human health effect of SO2 exposures at concentrations near those occurring in ambient air are transient changes in airway resistance and compliance. Amdur's guinea pig model (Amdur and Mead, 1958) is much more responsive to SO2 than is any other animal tested
APPENDIX B 135 (U.S. EPA, 19821. In this respect, it is similar in response to asthmatic hu- mans, who, in controlled chamber exposure studies, are about 10 times more responsive (Sheppard et al., 1980) than healthy humans. However, the tran- sient bronchoconstrictive responses seen in these human and guinea pig studies are qualitatively different from those reported for human populations exposed to polluted ambient air. Amdur's guinea pig model also has been used to demonstrate synergism between SO2 and PM. The bronchoconstrictive effects of SO2 are potentiated by coexposure to hydroscopic and catalytic aerosols (McJilton et al., 1976; Amdur and Underhill, 19681. However, both the end point used transient bronchoconstriction and the vein high ~ > 5 mg/m3) concentrations of parti- cles needed to cause the potentiated response make the relevance of these re- sponses to the effects associated with ambient pollutant exposures somewhat questionable. Experimental PM inhalation studies with concentrations and compositions similar to those encountered in ambient air have generally been unproductive (U.S. EPA, 19821. The only constructive interpretation that one can make of the data from such studies is that insoluble materials such as fixed carbon, resuspended soil, and fly ash could not by themselves account for the effects observed. The only experimental inhalation studies producing effects of possible rele- vance to the human experience have been those involving acidic aerosols, either alone or in complex mixtures. Some recent animal inhalation studies by Amdur (1985) demonstrated that effects produced by single exposures at very low acid concentrations can be persistent. She exposed guinea pigs by inhala- tion for 3 hours to the diluted effluent from a furnace that simulates a model coal combuster. The amount of H2SO4 on the surface of the ZnO particles was less than 40 ,ug/m3. These aerosol studies produced significant decrements of total lung capacity (TLC), vital capacity (VC), functional residual capacity (FRC), and CO-diffusing capacity (DLCo). At 12 hours after exposure, there were distension of the perivascular and peribronchial connective tissues and an increase in lung weight. The alveolar interstitium also appeared distended. At 1 hour, there was an increase in lung permeability. At 72 hours, TLC, VC, and FRC had returned to baseline values, but DLCo was still significantly de- pressed. Based on her experience with pure SO2 and pure H2SO4 exposures in the guinea pig model, Amdur concluded that the furnace effluent produced an acid aerosol effect because of its persistence. The persistent changes in func- tion and morphological changes after exposure to acidic aerosol at vein low concentrations suggested that repetitive exposures could lead to chronic lung disease. Another animal inhalation study with implications for human chronic lung disease involved a group of beagle dogs exposed to 16 hours daily for 68 months to a mixture of SO2 at 1,100 ,ug/m3 and H2SO4 at 90 ,ug/m3, either with
136 APPENDIX B
APPENDIX B 137 ther periods of daily exposures needs to be determined. We can expect these issues to be clarified as the results from longer exposure periods become available. The significance of the changes in particle clearance rate in the lungs from repetitive daily exposures to acidic aerosols, in terms of the pathogenesis of chronic respiratory disease, is not yet clear. But the close correspondence be- tween the effects of cigarette smoke and H2SO4 aerosol on mucocilia~y clear- ance after both short-term and long-term exposures, the similarities between the epithelial changes after repetitive H2SO4 inhalation in rabbits and those seen in the lungs of young smokers in postmortem examinations, and the well- established role of smoking in the etiology of chronic bronchitis combine to suggest that chronic bronchitis could result from long-term repetitive expo- sures to H2SO4 (Lippmann et al., 1982~. The fact that the incidence of chronic bronchitis among nonsmokers was higher in the United Kingdom when ambi- ent acidic aerosol concentrations were high lends further plausibility to a causal- relationship. Much more evidence is needed to demonstrate a clear causal rela- tionship, however. Conventional inhalation studies with rodents have not been useful in unrav- eling the effects associated with human exposures to the ambient complex mixtures of sulfur oxides and particles. Research studies with less commonly used experimental animals (e.g., guinea pigs, dogs, and rabbits) have been very useful, but not yet definitive. Perhaps the major lesson is that studies focused on mechanisms of the pathogenesis of chronic lung disease need to be designed with care, with constant attention to the suitability of the animal model and of the relationship to the real-world question. In summary, the excess mortality and morbidity that have been demon- strated in populations exposed to atmospheres containing mixtures of fossil- fuel combustion products might well have been due to a specific component of the mixture for example, the strong acid component of the aerosol or to the combined or synergistic effects of two or more of the many components in the mixture. A more definitive summary is not yet possible, although current labo- rato~y and epidemiologic research might permit one in the near future. LEAD AND NUTRITIONAL FACTORS EFFECTS ON BLOOD PRESSURE Hypertension has long been recognized as a risk factor for cardiovascular disease, and several environmental and nutritional factors have been shown to affect blood pressure in experimental and epidemiologic studies (U.S. EPA, 19841. Among environmental factors that have been associated with blood pressures are lead (Pb) and noise. Among dietary factors associated with blood pressure are calcium (Ca), zinc (Zn), phosphorus (P), alcohol consumption, and vitamins A and C. Although Pb is a single substance, it is always found in
138 APPENDIX B
APPENDIX B 139 An ideal opportunity to separate the role of Pb from a wide range of poten- tially confounding nutritional factors was presented by the large data set from the second National Health and Nutrition Examination Survey of 1976-1980 (NHANES II), based on a random stratified sample of the U.S. population. Pirkle et al. (1985) described the results of their date analyses of 40- to 59-year- old white men from the survey population. After adjustment for age, body- mass index, all measured nutritional factors, and blood biochemistry factors were adjusted for a multiple linear regression model, the relations of both systolic and diastolic blood pressures to blood Pb were statistically significant (p < 0.011. In an examination of the NHANES II population in the age range from 12-74, Harlan et al. (1985) also found a direct relation between blood Pb and systolic and diastolic blood pressures. In the Pirkle et al. (1985) analyses, the authors incorporated additional van- ables, with particular attention directed to the stability and significance of the Pb coefficient in the presence of nutritional factors and blood biochemistry. Their objective was to estimate conservatively the strength and independence of the relationship between blood pressure and blood Pb. Therefore, to provide an unusually rigorous test of the independent significance of blood Pb, 87 nutritional and biochemical variables in NHANES II were included in the stepwise regression. In addition, to account for possible curvilinear relations, squared and natural logarithmic transformations of almost all these variables were also included (see Table B-11. TABLE B- 1 Vanables Included in the Stepwise Regression Analysis, White Males Aged 40-59 years, NHANES II, 1976-1980a Age b Age-squared b Body mass index Dietary sodium C Salt-shaker sodium Dietary sodium x salt-shaker sodium Dietary potassium C Dietary sodium-potassium ratio Dietary calcium C Dietary phosphorus c Dietary protein c Dietary fat C Dietary carbohydrate c Dietary cholesterol c Dietary saturated fatty acids c Dietary oleic acid C Dietary linoleic acid C Dietary iron c a Reprinted with permission from Pirkle et al., 1985. b This variable was forced into each regression to remove any possible age effect on blood pressure. c The natural log and squared transformation of these variables were also included in the stepwise regression. Dietary vitamin A c Dietary vitamin C c Dietary thiamine c Dietary riboflavin C Dietary niacin Serum cholesterol c Serum vitamin C c Serum iron C Serum transferrin saturation Serum zinc Serum copper C Serum albumin C Hemoglobin C Red blood cell count Ethanol consumption per week C Cigarettes smoked per day Total dietary grams c Total dietary calories c
140 APPENDIX B When they included the nutritional variables, the blood analyses, and their curvilinear transformations, Pb remained significantly associated (p < 0.01) with both systolic and diastolic blood pressures. Furthermore, segmented re- gression analyses indicated that there was not a threshold blood Pb content below which Pb was not significantly related to blood pressure. Between 1976 and 1980, the mean blood Pb levels in the NHANES II popu- lation dropped by 37% owing to reductions in the amount of Pb used in gaso- line. This much reduction in blood Pb in this population would be expected to result in a 17.5% decrease in diastolic blood pressure of at least 90 mm Hg, a value used to define hypertension. Considering the relatively unusual nature of the blood Pb-blood-pressure relation (i.e., characterized by large initial increments in blood pressure at relatively low blood Pb concentrations, followed by leveling off of blood- pressure increments at high blood Pb concentrations), it is not surprising that it was not anticipated by results of animal studies. Many animal studies empha- size results from exposures at higher doses, where results tend to be more definitive. Yet, in retrospect, the human results were consistent with biphasic blood-pressure increases observed in response to blood Pb increases in the rat (Victery et al., 1982a,b). The unusual exposure-response relation might also account for the failure of earlier human studies to find consistent relations between blood pressure and blood Pb in study groups with mild to moderate increases in blood lead. In summary, the use of a very large set of high-quality data covering a wide range of possibly confounding variables allowed a clear-cut determination of the effects of blood Pb on blood pressure for a relatively low range of blood Pb concentrations (5-35 ,ug/dl). The studies also demonstrated the utility of com- prehensive data sets from representative populations for determining the ef- fects of specific components in environmental exposures. RADON DAUGHTERS AND CIGARETTE SMOKE This case study illustrates the interaction of two pulmonary carcinogens, cigarette smoke and radon daughters, to produce an increase in the total inci- dence of bronchogenic carcinoma in humans. Both of these agents are inhaled as gaseous suspensions and are delivered to the epithelium. The result of com- bined exposures is to increase the number of carcinomas, and this increase is interpreted as additive by several investigators, whereas others consider the effect as multiplicative. Exposure to radon daughters and cigarette smoke also appears to decrease the latency induction period for lung cancer. Animal stud- ies in general have supported the human epidemiologic and pathologic findings that the combined exposure to radon daughters and cigarette- smoke yields a greater number of malignancies than that observed for either carcinogen alone. Several of the cigarette-smoke-radon-daughter studies required the transla-
APPENDIX B 14 lion of the natural circumstances and organ systems to demonstrate the in- creased incidence of tumors. For example, McGregor (1976) used cigarette smoke condensate rather than native smoke. In addition, whereas alpha parti- cles are generated by radon-daughter decay, this study used beta particles. Finally, the test organ was not the lung, but rat skin. These changes in the nature of the carcinogenic agents, in the concentration and manner in which they were applied, and in the target organ itself make this example questionable as an analogue of the human events. The studies of Cross et al. (1982), in our opinion, have been incorrectly interpreted and are also of questionable utility. These investigators found more lung cancers in nonsmoking radon-daughter-exposed dogs than in a group ex- posed to cigarette smoke. These findings were interpreted as "protective" ef- fect. Such conclusions are unwarranted, because dogs do not inhale cigarette smoke in a manner or quantity analogous to those of humans. In fact, a major problem in virtually all animal studies in which inhalation of cigarette smoke is a requisite part of the design is the delivery of the smoke to the target organ in concentrations, temperature, age, and quantity identical to that inhaled by people. For one thing, there is a difference in inhalation path- way (i.e., nasal versus oral). In addition, most mammals find the inhalation of cigarette smoke so irritating that they must be forced to inhale. The volume of diluent air, the quantity of the cigarette smoke received, and the depth and frequency of inhalation are nonuniform and unlike the breathing pattern exhib- ited by human smokers. Furthermore, human lung diseases associated with cigarette smoke inhalation require heavy exposures (one or more packs per day) for periods of 20 or more years. Such exposures for such a duration are not feasible in animal studies. Thus, exposures of animals to cigarette smoke by inhalation are not adequate to initiate the disease states associated with human smoking. Radon is an inert gas that results from the radioactive decay of radium-226; it has a half-life of 3.8 days. Other decay products resulting from the sequence of radioactive decay are polonium-218, lead-214, bismuth-214, and polonium- 210. These elements are known collectively as radon daughters and have a collective half-life of 30 minutes. Small amounts of radon are ubiquitous, can be found in all rocks and soil, and are constantly emanating from ground level and diffusing into the atmosphere. Radon concentrations are increased wher- ever closed spaces exist around rock or soil and where phosphates or uranium ores or ore tailings are deposited. When radon decays, its daughter atoms, being heavy metals and ionized, attach to adjacent objects, surfaces, or dust. When inhaled into the respiratory tract, radon decays rapidly. In so doing, it produces irradiation of the adjacent cells before it can be removed by normal clearance mechanisms. Of the radiation dose, 95 % is delivered to the epithe- lium as alpha radiation; the beta and gamma energies are generally insignifi- cant and diffusely distributed. Exposure is measured in working levels (WL).
142 APPENDIX B One WL is the concentration of any combination of short-lived radon daugh- ters in air whose complete decay produces 1.3 x 105 MeV of alpha energy per liter. Exposure at this level for 1 working month, about 170 hours, constitutes an exposure of 1 working level month (WLM). The maximal permissible ex- posure for underground miners is 4 WLM/year, an equivalent of 0.3 WL/ month. Levels in mines have cumulatively reached as high as 10,000 WLM. Average levels in homes appear to be about 0.004 WL in the United States, although some homes have been found to have levels as high as 10 WL. When the first bronchogenic carcinoma was observed in a nonsmoking ura- nium miner in 1965 (described by Archer, 1985), there had already been 80 lung-cancer deaths in cigarette-smoking U.S. uranium miners. Because of the increased expectancy of lung neoplasms in cigarette smokers, it was quite natural to assume that the cause of all lung cancers in this group of miners was cigarette-smoking. In 1969, a second primary bronchogenic carcinoma was observed in a nonsmoking uranium miner (Archer, 19851. This case repre- sented a significant excess of lung cancers from what might be expected in this nonsmoking population. As the mortality study of uranium miners continued, Archer (1985) noted that nonsmoking uranium miners experienced 7 times the incidence of lung cancer observed in nonsmoking persons who were not miners, and smoking uranium miners had 9.5 times the incidence of lung cancers found in non- miners with similar smoking histories. Despite the fact that uranium miners smoked more than the comparably aged nonminer U.S. males, the excess ex- posure to cigarette smoke explained only a small part of the observed increase in lung cancer. This unexplained increase in cancer incidence suggested an interaction between cigarette-smoking and radiation exposure in the miners. Archer (1985) described additional studies which showed that the lung- cancer rate among smoking uranium miners declined after age 65. The expla- nation suggested was that the latency induction period had been shortened in smoking miners. Interpretation of this finding was that cigarette smoke might act as a promoter, whereas radiation acted as the initiator. Axelson and Sundell (1978) suggested that the increased incidence of lung cancer reflected an additive effect between smoking and radon-daughter expo- sure. Damber and Larsson (1985) interpreted their findings as a multiplicative effect. The studies of Hornung et al. (1981), Edling (1982), and Radford and St. Clair Renard (1984) all supported an additive effect, rather than a multipli- cative one. The early studies of the cell type of bronchogenic carcinoma seen in uranium miners suggested that there were greater numbers of undifferentiated small- cell carcinomas in this group than in the lung cancers of nonminers. The in- creased proportion of small-cell or differentiated lung cancers which appeared to be the same in smoking and nonsmoking uranium miners, suggests that mining exposure and radiation were responsible for any difference in the num-
APPENDIX B 143 her of undifferentiated carcinomas. Recent pathologic studies have not con- firmed this increase in undifferentiated small-cell cancers in miners. All cell typesthe squamous carcinoma, the adenocarcinoma, and the small-cell un- differentiated carcinoma have been observed in cigarette-smoking uranium miners in numbers similar to that present in cigarette smokers who are not miners. Factors that might influence the cell type and progression of lung cancers in U.S. uranium miners are age at the start of mining, high radiation exposure rates, and cigarette-smoking. The risk of lung cancer appears to decline in miners who have reached age 65 and have had 25 or more years of latency. Several animal studies have addressed the interaction of cigarette smoke, cigarette-smoke condensate (CSC), and ionizing radiation. McGregor (1976) studied the interaction of CSC and beta particles on rat skin and observed an increase in skin tumors 3 times greater than the increase produced by radiation alone. CSC alone did not produce tumors. Treatment with both agents caused tumors to appear earlier than in the radiation-only group. Nenot (1977) used cigarette-smoke inhalation and americium-241 applied to the lungs in rats. The animals receiving the combination of smoke and radiation had an increased yield of lung cancers, which appeared earlier than with radiation alone. In summary, epidemiologic studies of exposure to radon daughters and ciga- rette smoke have demonstrated two effects. The first is an additive effect of the two agents on the number of cancers induced. The second is the decrease in induction latency period, which produces a different time distribution of tu- mors in smokers and nonsmokers: the tumors in smokers appear earlier. We have seen comparable results in animals exposed to components of cigarette smoke at high concentrations, but not when the animals were exposed to fresh smoke by inhalation. The discordance in the latter case is most likely due to technical limitations in our ability to produce conditions that match human smoking. ASBESTOS EXPOSURE AND CIGA1lETTE-SMOKING This case study is one of the most current and well-recognized examples of how two distinct agents administered together can produce an increased inci- dence of lung cancer that is greater than that predicted from the administration of either agent alone. The increase in lung-cancer incidence in cigarette-smok- ing asbestos workers is considered multiplicative by most investigators who have studied the problem. That is in contrast with the increased incidence of lung cancer in uranium miners, in whom it has been observed that the increase in lung-cancer incidence from the combined exposure is additive. This case study also differs from that of the radon-daughter-cigarette-smoke example, in that asbestos, unlike radon gas, is a fibrous mineral particulate material that enters the lung via the airways and remains in the lung for long
144 APPENDIX B periods. There is ample evidence that asbestos alone is a mild carcinogen. However, the amount of asbestos required to initiate carcinogenesis is not known. Although cigarette-smoking unquestionably increases the incidence of primary epithelial lung neoplasm, it has no apparent effect on the malignant tumors arising in the pleural, peritoneal, orpericardial mesothelium. Mesothe- liomas, as these malignant tumors are called, have a definite relation to inhala- tion of asbestos fibers. However, mesotheliomas have occurred after relatively short and light asbestos exposures, as well as after prolonged and heavy occu- pational exposures. The dose-response relation is therefore variable, and the threshold dose required for mesotheliomas to occur might be exceedingly small or nonexistent. It is generally accepted that the extent and severity of asbestosis, the disease state in which scarring or fibrosis develops, is related to the duration and quan- tity of asbestos to which the person has been exposed. There is no evidence, however, that cigarette-smoking accelerates or modifies the extent of fibrosis produced by asbestos exposure. Animal studies, when properly designed, have demonstrated the production of primary lung cancer, pleural mesotheliomas, and asbestosis. Studies in which combined exposures of animals to cigarette smoke and asbestos fibers were used were difficult to interpret, because the number of lung tumors that occurred as a result of combined exposure was no greater than that produced by asbestos exposure alone. It is likely that the design of that part of the study, which required the delivery of cigarette smoke to the lungs of the animal subjects, was unreliable because of the difficulty, previously noted, in having animals inhale cigarette smoke in the manner and quantities of human inhalation. However, when carcinogenic agents present in cigarette smoke, such as benzoLa~pyrene, are administered with asbestos in a syngeneic tracheal graft or organ culture system, the increase in carcinomas in the target tissue is greater than that initiated by either agent separately. In 1935, Lynch and Smith reported the first case of lung carcinoma associ- ated with asbestosis and silicosis. Sporadic case reports of bronchogenic carci- noma in asbestos-exposed workers appeared in the medical literature between 1935 and 1955. In a landmark 1955 study, Doll reported increased mortality from lung cancer in British asbestos workers. That observation was confirmed by Selikoff et al. in 1968. The latter researchers also evaluated and correlated smoking habits with the incidence of lung cancer in asbestos-insulation work- ers. In the study, no lung cancer was detected in the 48 workers who never smoked, nor in 39 workers who smoked only pipes or cigars. Of the remaining 283 cigarette-smoking workers, 24 died of lung cancer in the followup period, whereas only 2.98 deaths from lung cancer were expected on the basis of statistics in smokers. In 1979, Hammond et al. expanded the evaluation of the role of cigarette- smoking in the development of primary lung cancer among asbestos-insulation workers. Of the 1,332 deaths in that group, 314, or 23.6%, were from lung
APPENDIX B 145 cancer. At 20 years after onset of asbestos exposure, the death rate of nonsmok- ing controls was 11.3 per 100,000, whereas the nonsmoking asbestos workers had a death rate of 58.4 per 100,000 and thus a mortality ratio 5.17 times greater. The death rate of smokers in the control group (no asbestos exposure) was 122.6per 100,000, whereas the smoking asbestos workers had a death rate of 601.6 per 100,000 and thus a mortality ratio 4.9 times greater. The lung- cancer risk for the asbestos-exposed smoking group was 53.24 times that of the nonsmoking, non-asbestos-exposed smoking group, a figure that reflects a multiplicative relation between the two variables (5.17 and 10.85, see Table B-2~. The mortality ratio for lung cancer in pipe and cigar smokers was 7.02. There appeared to be a dose-response relation between the amount of cigarette- smoking and the asbestos exposure. In the same study, ax-smokers with asbes- tos exposure had a mortality ratio of 36.56 over those who smoked less than one pack per day, 50.82, whereas smokers of a pack or more per day had a ratio of 87.36. Similar findings were reported by Berry et al. in 1972. This group studied 1,203 male asbestos workers, 74.5% of whom were cigarette smokers. A dose-response relation of cigarette-smoking with asbestos exposure was ob- served in both men and women: with greater exposure, the ratio of observed to expected lung-cancer deaths increased. McDonald and his colleagues (1980) further confirmed this effect on Canadian miners and millers, as did Meurman and colleagues (1974) in Finnish an~ophyllite miners and millers. The carcinogenic effect of asbestos-fiber inhalation without cigarette-smok- ing has been demonstrated in several independent studies. In addition to the studies of Selikoff et al. (1968), those of Berm et al. (1972), McDonald et al. (1980), Meurman et al. (1974), and Liddell et al. (1982) reported deaths from lung cancer among nonsmoking asbestos-exposed workers. Studies have not established a threshold quantity of asbestos required to increase Me risk of lung TABLE B-2 Age-Standard~zed Lung Cancer Death Ratesa for Cigarette Smoking and/or Occupational Exposure to Asbestos Dust Compared win No Smoking and No Occupational Exposure to Asbestos DusP - Exposure History to Cigarette Death Mortality Mortality Group Asbestos? Smoking? Rate Difference Ratio Control No No 11.3 0.0 1.00 Asbestos workers Yes No 58.4 +47.1 5. 17 Control No Yes 122.6 +111.3 10.85 Asbestos workers Yes Yes 601.6 +590.3 53.24 a Rate per 100,000 man-years standardized for age on the distribution of the man-years of all the asbestos workers. Number of lung cancer deaths based on death certificate information. b From Hammond et al. (1979).
146 APPENDIX B cancer, because the asbestos exposure has not been quantified. Furthermore, as the dose of asbestos decreases, mortality statistics become less accurate. The risk of developing lung cancer after cessation of cigarette-smoking in asbestos-exposed workers appears to decrease in a manner similar to that seen in non-asbestos-exposed cigarette smokers who stop smoking. These findings have been observed by Hammond et al. (1979) and independently by Walker (1984~. Seidman and colleagues (1979) studied workers who received brief but intense exposures to asbestos and were studied 35 years later. The lung-cancer mortality ratio was increased and appeared to remain increased for the entire period after exposure. An occupational exposure to amosite asbestos for as short a time as 1 month was associated with increased lung-cancer mortality. The likelihood of cancer appeared to increase with the duration of exposure after a latent period of at least 25 years. The length of the latent period also appeared to vary with the extent and duration of the exposure. Saracci (1977) reviewed the reports relating cigarette-smoking and asbestos exposure from five major studies and concluded that the relation was consistent with a multi- plicative model. Rats have been exposed to asbestos fibers by inhalation in several carefully performed studies (Davis et al., 1980; Wagner et al. 19801. After exposures of a year or more, approximately 25% of the animals developed primary lung carcinomas, which were most commonly adenocarcinomas. All three major types of asbestos produced these carcinomas, but there was a greater accumu- lation of amphiboles than of ch~ysotiles in the lung. Wehner et al. (1975) reported that hamsters exposed to cigarette smoke and asbestos by inhalation developed adenomas, papillomas, and adenocarcino- mas. The results of the study are difficult to interpret, because the number of tumors that developed in the group receiving the combined cigarette-smoke and asbestos exposure was no greater than that produced by asbestos inhalation alone. Because difficulties in delivery of cigarette smoke to the lung of small mammals undermine our assurance that the animals received the smoke, sev- eral investigators have studied the effect of asbestos in association with ben- zoLa~pyrene (BaP), a polycyclic aromatic hydrocarbon present in cigarette smoke. A striking increase in the number of benign and malignant neoplasms of the lung and airway was seen as a result of this combined exposure, com- pared with the number seen after exposures to asbestos or BaP alone. These findings support the observations in humans of an increased effect from the combination of asbestos and cigarette smoke, compared with either agent alone. Topping and Nettesheim (1980) used an unusual syngeneic tracheal graft system to study the effects of asbestos and polycyclic hydrocarbons. Carcino- mas occurred when the grafts were treated with dimethylbenzanthracene (DMB) first and later with chrysotile asbestos. The DMB alone did not cause tumors an indication that asbestos had acted as a promoting agent. Mossman
APPENDIX B 147 and Craighead (1982) reported similar findings in an organ-culture system of hamster tracheas. Carcinomas developed when 3-methylcholanthrene (3MC) was used to coat the surface of crocidolite fibers that were then placed on the epithelial surface of the tracheal organ culture. Asbestos was considered the carrier of the 3MC, because other minerable dusts, such as kaolin and carbon, also caused tumor formation if coated with 3MC. Although the evidence that epithelial malignant tumors develop in increased numbers as a result of combined exposure to cigarette smoke and asbestos inhalation is compelling, that is not the case for malignant mesotheliomas. Mesotheliomas are rare tumors that are believed to arise from the lining cells of the serous cavities, namely the pleura, peritoneum, and pericardium. Their association with asbestos exposure was first documented by Wagner et al. (1960) and has been confirmed by Selikoff et al. (1968), Newhouse et al. (1972), McDonald et al. (1980), and others. Cigarette-smoking does not ap- pear to influence the incidence, latency, or progression of these tumors. The combined exposure to asbestos and cigarette smoke in humans also has effects on functional responses of the lungs and airways that are modifications of the effects of either agent acting alone. Prolonged exposure to cigarette smoke alone is characterized in humans by airway obstruction, parenchymal destruction with little or no fibrosis, and increased lung volume. There is also often increased production of mucus associated with mucous gland hypertro- phy. However, asbestos inhalation in humans over long periods leads to paren- chymal destruction with fibrosis, reduced lung volumes, and minimal airway narrowing. In asbestos-exposed heavy cigarette-smokers, the predominant functional changes are most commonly airway obstruction and reduced lung volume. This combination of functional changes reflects the combined effects of both agents. Regan and coauthors (1971) evaluated the discriminatory power of various pulmonary Function tests to distinguish the effects of smoking from those of asbestos exposure. Decreases in diffusing capacity for carbon monoxide and decreases in vital capacity were the best measures of severity of obstructive and restrictive lung disease in persons with combined exposures, but these tests do not clearly distinguish between the two agents. A good indicator to distinguish the effect of cigarette-smoking is the 1-second forced expiratory volume (FEV~) as a percentage of forced vital capacity (FVC). The test reflects airflow obstruction in large or proximal airways; however, it is not a good index of small-airway function. Wright and Churg (1984) have reported that asbestos produces a much greater change in membranous bronchioles than does ciga- rette-smoking alone. But the extent of the lesions produced by asbestos and cigarette-smoking, compared with those produced by asbestos alone, was not clearly distinguished. Begin et al. (1983) evaluated small airways in cigarette- smoking asbestos miners and millers in Quebec by physiologic and lung biopsy studies. They found that nonsmokers had relatively normal function, whereas
148 APPENDIX B smoking miners and millers had clear evidence of small-airway obstruction and a 3- to 4-fold increase in upstream resistance at low lung volumes. The small-airway dysfunctions observed in asbestos workers who are not cigarette smokers were not of sufficient severity to cause alterations in the FEV~/FVC ratio. Persons with a long history of cigarette-smoking and asbestos exposure have a higher prevalence of radiologic abnormalities characterized as interstitial fibrosis than do nonsmoking asbestos-exposed persons. The differences are not apparent in populations with a higher prevalence of roentgenographic fibrosis and presumably higher asbestos exposures. The study of Harries et al. (1972) suggests that a shorter asbestos exposure might be required to produce an ab- normal chest roentgenogram in smokers than in nonsmokers. There is no other evidence to suggest that smokers develop more severe fibrosis than non- smokers. A number of immunologic changes have been observed in asbestos-exposed workers, including humoral and cellular immune responses. Lange (1980) re- ported that a proportion of asbestos-exposed workers with roentgenographi- cally demonstrable asbestosis had increased circulating immunoglobins A and G. Those findings were believed to be due to asbestos exposure, rather than to cigarette-smoking. Wagner and colleagues (1960) studied the effect of smok- ing and asbestos exposure on T-lymphocyte number and subset formation. Persons with roentgenographic asbestosis who smoked had a greater number of T helper cells than those who did not smoke, whereas suppressor T cells were not affected. Wagner (1980) reported that asbestos exposure was associated with a decrease in number of total T cells, a decrease in suppressor T cells, and an increase in the helper subset. Smoking history did not influence those results. In other studies, heavy smoking was found to cause an increase in total T lymphocytes, a decrease in the proportion of T helper cells, and an increase in T suppressor cells; this creates a decreased ratio of helper to suppressor cells. DeShazo and colleagues (1983) observed that asbestos workers had sig- nificantly increased proportions and total numbers of B and T lymphocytes in peripheral blood and a proportionate decrease in helper T cells. Those changes had no relation to radiologic categories of pneumoconiosis, cumulative asbes- tos exposure, or pulmonary functional abnormality. There is apparently a lack of uniform results concerning the effect of asbestos and cigarette-smoking on T-lymphocyte number and activity. More work is required to clarify this problem. CIGARETTE-SMOKING AND ALCOHOLIC-BEVERAGE CONSUMPTION Both cigarette-smoking and alcoholic-beverage consumption are risk factors for cancers of the oral cavity. Because these exposures commonly occur to-
APPENDIX B 149 "ether, one can usefully ask whether concurrent exposure to both increases the risk above that expected from simply summing the effects of either exposure alone. That was done in a useful paper by Rothman and Keller (1972), summa- rized here. Before examining the effects of joint exposure, we must define the measure- ments that quantify the population response to individual exposures and define how joint exposures may be classified. Disease risk may be expressed as time- conditioned probability of disease occurrence (e.g., 10 cases per 100,000 per- sons per year). We can assume that the first of two independently acting expo- sures increases risk by an amount, x (i.e., the disease rate is higher in the exposed group than in the unexposed by x). Let us assume that a second expo- sure increases risk by an amount, y. If the result of both exposures is to increase risk by x + y, the exposures are said to act independently. If the risk resulting from both exposures is greater then x + y, the interaction is said to be synergis- tic; if the risk is less than the sum of the two acting alone, the interaction is defined as antagonistic. The relative risk (risk ratio) is commonly used in epidemiologic studies as a measure of risk. It is defined as the probability of disease in an exposed group divided by disease probability among an unexposed population that is other- wise similar (with respect to age, race distribution, socioeconomic status, etc.~. If unexposed persons have a disease rate Ro and exposed persons have a rate Rim, the relative risk is R~/Ro. If another exposure increases the baseline risk Ro by y to a level Ro + y, then independence of action of the two exposure factors implies that the relative risk due to exposure x would be lower in the presence of the other risk factor than in its absence. In general, the proportional or relative increase in risk due to a constant addition to the disease rate among exposed persons decreases with the magnitude of the baseline risk. Thus, a decrease in relative risk from one risk factor in the presence of another is consistent with a simple additive risk model. Rothman and Keller (1972) used data from a study of mouth and pharynx cancers (Keller and Terris, 1965) to examine whether cigarette-smoking and consumption of alcoholic beverages are independent (i.e., simply additive), synergistic, or antagonistic. After excluding people with incomplete smoking or drinking histories, the study included 483 cases and 447 controls. The cases in males were diagnosed with squamous cell carcinoma of the mouth or phar- ynx. Controls were matched to cases by age and sex. Histories of smoking and drinking, routinely taken by the admitting physician, were abstracted from the clinical record. Smoking was expressed as number of cigarettes per day, con- sidering one cigar the equivalent of four cigarettes and one pipeful of tobacco the equivalent of two cigarettes. Alcohol was expressed as ounces of alcohol per day. Although stratified by age, relative risks were calculated for various combinations of smoking and drinking behavior. When smoking and drinking are considered in a simple table dichotomous
150 APPENDIX B TABLE B-3 Relative Risks of Oral Cancer According to Presence or Absence of Two Exposures Smoking and Alcohol Consumptiona Alcohol Relative Consumption Smoking Risk No No 1.00 No Yes 1.53 Yes No 1.23 Yes Yes 5.71 a Data from Rothman and Keller (1972). for both exposures, there is a strong suggestion of synergy (Table B-31. In the absence of dnnking, smoking adds 0.53 unit of risk to the baseline risk of 1 .00. In the absence of smoking, drinking adds 0.23 unit of risk to the baseline. If each of these simple dichotomous exposures acted independently, the contri- butions to risk would be additive, and the net contribution of both exposures would be to increase risk by 0.53 + 0.23 for an excess risk of 0.76, instead of the observed excess risk of 4.71. The detailed data were used by Rothman and Keller (1972) to examine risk more carefully. Table B-3 shows the relative risk of oral cancer according to level of exposure to smoking and alcohol. In Table B-3, the result of increasing each exposure is observed at every level ofthe other; that illustrates the individ- ual effects of the exposures. In addition, the observed risks in the lower right portion of Table B-3 are greater than expected on the basis of simple additivity. The expected relative excess risk in the highest smoking and drinking stratum is 1.33 + 1.42 = 2.76, far less than the observed level of 14.5. Caution is warranted in interpreting these results, however. Most of the ex- cess risks that suggest synergy occur in the cells along the bottom row and right-hand column of the table. These strata represent open-ended exposure categories, where an excess of one exposure might account for the excess risk observed, in the absence of synergy. It would not be correct, for example, to TABLE B-4 Relative Risks of Oral Cancer According to Level of Exposure to Smoking and Alcohola Smoking (cigarette equivalents/day) o 1.00 1.40 1.60 2.33 Alcohol, oz/day < 20 20-39 - 40 o <0.4 0.4-1.5 21.6 1.52 1.67 4.36 4.13 1.43 3.18 4.46 9.59 2.43 3.25 8.21 15.5 a Adapted from Rothman and Keller (1972).
APPENDIX B 151 assume that nonsmokers who consumed 1.6 ounces per day consumed the same amount of alcohol as those smoking 20-39 cigarettes per day and dnnk- ing 1.6 ounces of alcohol per day. The exposures are correlated and the categories open-ended with respect to alcohol ingestion. Observed risks in cells adjacent to those in the lower nght-hand corner of Table B-3 also exceed expected values and offer stronger evidence of synergy. But the suggested interaction might have arisen because the categories are too broad. Moreover, a randomly low value in one of the two cells used to generate the expected values can falsely suggest synergy. Rothman and Keller (1972) concluded that the data in Table B-4 suggest a combined effect equal to the sum of two strong individual effects plus a synergistic component. They believed that the evi- dence from this single study was not strong enough to warrant completely discarding the simple model of independent effects. Data from another case-control study of oral cancer by Wynder et al. (1957) were transformed by Rothman and Keller and also presented in tabular form (Table B-51. A different classification scheme was used to produce more stable estimates of relative nsk, but evaluation of synergy should not depend on the classification used. They concluded that the results shown in Table B-5 suggest that action of the two exposures is additive and independent. TRIHALOMETHANES AND OTHER BYPRODUCTS OF CHLORINATION IN DRINKING WATER This case study demonstrates how epidemiologic methods have been used to evaluate human risk from exposure to complex mixtures of chlorination by- products in disinfected drinking water. It is generally accepted that the practice of chlorine disinfection, used since 1909 in the United States (Johnson, 191 1), has been an important factor in reducing morbidity and mortality from venous waterborne pathogens, although numerical estimates are not available. A TABLE B-5 Relative Riska of Oral Cancer According to Level of Exposure to Smoking and Alcoholb Smoking Alcohol, (cigarettes/day) units/days < 15 16-20 21-34 - 35 . < 1 1.00 2.86 1.79 8.40 1-2 1.70 2.05 1.94 3.88 3-6 6.20 7.02 8.91 5.33 > 6 9.69 1 1.6 17.0 19.4 a Risks expressed relative to risk of 1.00 for persons who smoked fewer then 16 cigarettes per day and drank alcohol less than 1 unit per day. b Unadjusted for age. Adapted from Rothman and Keller (1972). c One unit of alcohol equals 1 oz. of whiskey or 8 oz. of beer.
152 APPENDIX B suggestion that the time-tested benefits of chlorination can be partially offset by an increase in the burden of chronic disease from chlorination byproducts was raised in 1974 and 1975. Rook (1974) and Bellar et al. (1974), working independently, showed that chloroform and other trihalomethanes (THMs) are created during chlorine disinfection. Environmental surveys demonstrated that concentrations are much higher in treated surface waters than in water from underground sources. Shortly thereafter, a feeding study of chloroform dem- onstrated its carcinogenicity in rodents (Page and Saff~otti, 1976~. Later work showed that THMs are accompanied by a complex mixture of higher-molecu- lar-weight nonvolatile chlorinated organic substances (Stevens et al., 19851. Specific mutagens, including chlorinated aldehydes and ketones, have been identified (Meter et al., 19851. In most treated waters, more than half the organically bound chlorine is associated with this nonvolatile fraction (Stevens etal., 1985~. Assessment of possible links of chlorination byproducts with cancer in hu- man populations started soon after the discovery of THMs in treated water and continues today. A central issue in epidemiologic studies of chlorination by- products, as well as other complex mixtures, is how best to define exposure. Ideally, the measure used would provide accurate estimates of past exposure to the biologically active factors in the mixture. Most epidemiologic studies of chlorination byproducts have relied on the large differences in contaminant concentrations between water from surface and ground sources and between chlorinated and nonchlorinated water, to de- rive relatively simple, categorical definitions of exposure. A few studies have used recent measurements of THMs to define likely past concentrations. The use of recent THM measurements to impute past exposures, although satisfy- ing a need for quantitative estimates, might tee misleading. THM contents from any supply show large diurnal and seasonal variations. THM estimates from a single measurement, therefore, can provide but a crude estimate of past aver- age exposures. The biologic activity of interest might reside not in the easily measured THMs, but rather in the higher-molecular-weight fraction of the chlorination byproduct mixture. The degree of correlation between THMs and biologically active nonvolatile compounds, although assumed to be high, is not well documented in field measurements and might be variable. In the face of these difficulties, comparisons of risk between groups or individuals exposed to chlorinated surface water, as opposed to persons who used ground- water, could be as valid and useful as comparisons based on actual THM measurements. Just as toxicologic assessment of chlorination byproducts and other chemi- cals in drinking water have evolved from crude evaluations of complex mix- tures to identification of individual chemicals, so epidemiologic studies have progressed from broad assessments, with little ability to distinguish specific exposures or control for potential confounders, to more refined "analytic"
APPENDIX B 153 studies. The first epidemiologic studies were conducted shortly after informa- tion on specific chemical contaminants became available. Most studies were ecologic in design, comparing the geographic distribution of site-specific, age- adjusted cancer mortality rates among U.S. counties with the distribution of water-supply characteristics or measured concentrations of drinking-water contaminants. One of the earliest studies (De Rouen and Diem, 1977) com- pared cancer rates in Louisiana parishes (counties) served by Mississippi River water (with a large number of known organic contaminants) with cancer rates in parishes served by other, mostly groundwater, sources. A multiple-regres- sion model was used in which the independent predictor variables included the percentage of the parish population drinking water from the Mississippi, parish urbanicity, median income, and several parish-level industrial variables. Dependent variables in separate regression models were average annual age- adjusted mortality rates for 1950-1969 for cancers of the gastrointestinal and urinary tracts and total cancer. Statistical tests evaluated the magnitude and statistical significance of correlations of the cancer rate with the drinking- water variable, after other parish characteristics were adjusted for. For the development of exposure indexes, other demographic studies of cancer mortality and water quality used data from a 1963 U.S. Public Health Service Inventory of Municipal Water Supplies (1964) to calculate the percent- age of county populations served by surface or ground sources or by chlori- nated or nonchlorinated supplies. Salg (1977) examined the association be- tween water-quality factors and cancer mortality in the 346 counties of the Ohio River drainage basin, using as exposure variables the percentage of each county's population served by surface water. Another study (Kuzma et al., 1977) looked for associations of water quality and cancer rates in Ohio coun- ties, using a dichotomous exposure variable that indicated whether surface water or groundwater was used by a majority of a county's population. Other studies (e.g., Burke et al., 1983) used as exposures THM contents in the pre- dominant county water supply, according to U.S. Environmental Protection Agency surveys. Reviews of the ecologic studies of cancer and drinking-water contaminants (e.g., Craun, 1985) noted that, although there were many inconsistencies in their results, three cancer sites bladder, colon, and rectum appeared to be associated with one or another water quality variable more often than might have occurred by chance. It was suggested that these three sites deserved fur- ther evaluation in more highly focused case-control studies. The first case-control studies selected deceased cases and controls from computerized listings of state vital-statistics bureaus. Addresses on death cer- tificates served as links to information on drinking-water source and treatment from the community water supply at the time of death. With this information, the most recent drinking-water source and treatment were characterized for each study subject, and the frequency of use of different types of water sources
154 APPENDIX B among cases was compared with that of controls. Although these studies repre- sented a major step forward, in that exposure and disease information was on anindividual, not group, level, they shared some ofthe weaknesses of indirect, ecologic studies. Among the weaknesses are the potential for exposure mis- classification due to changes in water supply, migration or other factors, and the inability to account for other risk factors that can confound relations with water contaminants or interact with them. That can be of special importance when relatively small effects are expected. Five studies (Alavanja et al., 1978; Kuzma et al., 1977; Brenniman et al., 1980; Gottlieb and Carr, 1982; Kanarek and Young, 1982) examined several cancer sites as related to water source at the decedent's last address. In a sixth study (Young and Kanarek, 1983), drir~k- ing water of decedents was defined for the last 20 years of life. Slightly increased relative risks were observed for cancers of the colon, rectum, and bladder in five studies, associated with either surface source (ver- sus ground) or chlorinated source (versus nonchlorinated) . Risk ratios from the five studies were in the range 1.0-2.0. The sixth study, looking only at colon and rectal cancer in relation to exposures up to 20 years before death, found no association for either site. Risk ratios below about 2.0 from epidemiologic studies are often subject to question, even if they are statistically significant. Associations of that magni- tude can be due to confounding from risk factors that are not ascertained. Nevertheless, when exposures are widespread in a population, even small in- creases in relative risk can translate into large numbers of excess malignancies. In addition, if the positive associations from the case-control mortality studies reflected causal relationships, the magnitude of the risk might have been under- estimated, because the net effect of migration would be to decrease the accu- racy of exposure information. Those factors justified continued research into water contaminants and human cancer. Epidemiologic evaluation has continued in the form of case-control studies of incident cancers. Such studies typically gather information on a number of risk factors directly from cancer subjects, controls, or their next of kin. That information permits careful control for potential confounding by other risk factors and increases the precision of risk estimates from water-related expo- sures, because more accurate exposure information is available. Preliminary results from a study of colon cancer in North Carolina (Cragle et al., 1985) show associations with exposure to chlorinated surface water, especially among older groups exposed for at least 15 years. A case-control study of the incidence of colon cancer in Wisconsin (Kanarek and Young, 1982) showed no association of risk with several different mea- sures of past THM intake. Dose differences between exposed and unexposed persons in Wisconsin might not have been as large as in North Carolina, because relatively uncontaminated surface waters were used for drinking- water supplies in Wisconsin and had lower concentrations of chlorination byproducts.
APPEND[X B 155 A large case-control interview study (Cantor et al., 1985) of bladder cancer in 10 areas of the United States revealed increasing risk with tap-water intake. The risk gradient with intake was largely restricted to long-term consumers of chlorinated surface water and was not found among consumers of nonchlori- nated groundwater. Relative to long-term consumers of groundwater, bladder- cancer risk increased with duration of exposure to drinking water from treated surface sources, especially among high-quantity consumers. Although most relative risks for bladder cancer were below 2.0, the findings of increased risk were generally consistent among geographic areas and between the sexes, and statistical tests for trends with tap-water intake were highly significant. Results from carcinogenicity testing in animal feeding studies are available for the four major THMs in drinking water (Balster and Borzelleca, 1982), but not the nonvolatile chlorination byproducts. Extensive testing of nonvolatile concentrates has demonstrated mutagenicity in a variety of in vitro testing systems, as well as transformation of mammalian cells in tissue culture. Identi- fication and toxicologic testing (in vitro) of several compounds of the chlorina- tion-byproducts mixture have been conducted and have raised general con- cerns, but more specific implications for adverse human health effects are unclear. COKE-OVEN EMISSIONS The following discussion of the toxic effects of coke-oven emissions is pre- sented to demonstrate that animal toxicity testing of complex mixtures is a valid means of predicting human disease resulting from environmental expo- sure to these mixtures. From a toxicologic point of view, coke-oven emissions and related agents, such as coal tar, are probably the most widely studied complex mixtures. There is an overwhelming body of scientific evidence that these substances are car- cinogenic. Coal-tar derivatives were probably responsible for the first recorded observation of occupational cancer, made in 1775 by Percivall Pott, who noted an excess of scrotal cancer in London chimneysweeps. The first experimental demonstration of chemical carcinogenesis was made by Yamagiwa and Ichi- kawa in 1918, when they applied coal tar to the ears of rabbits. Numerous epidemiologic and animal studies have confirmed the carcinogenic properties of combustion and distillation products of coal. In the present discussion, the relevant question is: Given the animal data, can we predict human health ef- fects of coke-oven emissions? The most extensive demonstrations of human disease resulting from coke- oven emissions have come from mortality studies. These studies show signifi- cant increases in lung and genitourinary cancer mortality associated with expo- sure to coke-oven emissions. The first of these reports was made in 1936 by investigators in Japan and England (Kennaway and Kennaway, 1936) who were studying lung-cancer mortality among persons employed in coal carboni-
156 APPENDIX B zation and gasification processes. Later studies, conducted in the United States by Lloyd (1971) and others, have clearly demonstrated substantial increases in lung-cancer mortality among coke-oven workers. Similar studies have shown that coke-oven workers have a significant increase in mortality due to cancer of the genitourina~y system (IARC, 19851. It is clear that human exposure to coke-oven emissions results in lung and genitourinary cancer. The question is: Could these diseases have been pre- dicted from available animal data? At first glance, the answer appears to be no, because of the designs of the animal experiments and the epidemiologic stud- ies. Coke-oven emissions cannot be generated in an experimental setting. For that reason, coal tar derived from coking operations has been used for animal research. The chemical composition of coal tar is very similar to that of coke- oven emissions, so the effects of coal-tar exposures on animals are thought to approximate closely the effects of coke-oven emissions on humans. For 70 years, investigators have been producing skin cancer in experimental animals by dermal application of coal tar or coal-tar extracts. The results are consistent with observations of increased scrotal cancer in chimneysweeps made two centuries ago, but are not confirmed by modern epidemiologic stud- ies of coke-oven workers. There are two possible reasons for the discrepancy. First, modern coke-oven operations and practices of hygiene have decreased dermal exposures. Second, and more probable, the epidemiologic studies used mortality data, and skin cancer is seldom fatal. The point is that animal data will not predict human disease if means are not available for gathering human data. That applies not only to nonfatal malignancies, but also to other disease end points that do not show up on death certificates. In contrast with skin cancer, lung and genitourinaIy cancers have been de- tected readily in human epidemiologic studies, but not until recently in experi- mental animal studies and even then not genitourina~y cancers. In this case, the discrepancy was probably due to the design of the animal studies, which used dermal exposures. It was not until epidemiologists discovered the relation be- tween coke-oven emissions and lung and genitourinaly cancers that experi- menters began extensive use of nondermal routes of exposure. Experiments confirmed the epidemiologists' findings in the case of lung cancer. A similar situation existed for nonmalignant respiratory diseases that were detected by epidemiologists in coke-oven workers. Those diseases have been overlooked by experimenters, possibly because of the emphasis placed on carcinogenic end points. Animal studies must be designed appropriately, if their results are to predict specific human diseases. It is clear that, for coke-oven emissions, the correlation between animal and human data is far from perfect. When we view the evidence retrospectively, it is evident that the design of animal experiments might not have produced the data necessary to make predictions of specific human diseases before their discovery by epidemiologists. It might never be possible to know a priori pre-
APPENDIX B 157 cisely which experiments to do to determine specific human health effects of complex mixtures. Given the imperfect nature of animal data, can they be used to make decisions concerning health threats posed by complex mixtures? Judg- ing from our experience with coke-oven emissions, the answer is clearly yes. Although there are substantial differences between the human and animal data, we conclude that the adverse effects of coal tar seen in animals are echoed by the effects of coke-oven emissions in humans. The above discussion of coke-oven emissions demonstrates that data gener- ated by animal toxicity testing of complex mixtures can be a reasonable predic- torofhuman disease, if the mixture being tested in the laboratory is representa- tive of mixtures in the human environment, if studies are properly designed to detect diseases, and if we are aware that the specific diseases and target organs in test animals can vale from those in humans. COAL-MINE DUST The studies of lung disease in coal miners have focused principally on chest x-ray alterations, pulmonary function deficits, chronic bronchitis, and patho- logic alterations of lung tissue. Attempts have been made to relate those to each other, as well as to an exposure index (e.g., years spent underground in coal mines) and dose (mass of respirable coal dust per cubic meter of air times total hours exposed). An immense amount of data has been generated. The purpose of this example is to document the correlations, point out the difficulties in- volved, and explain how animal experiments have or have not assisted in ex- plaining some of the biologic response variables. Coal workers' pneumoconiosis (COOP) was originally described in terms of descriptive pathology (Gough, 1947; Heppleston, 19471. The characteristic lesion is a focal collection of coal-dust-laden macrophages at the division of respiratory bronchioles that can exist within alveoli and extend into the peri- bronchiolar interstitium with associated reticulin deposits and focal emphy- sema (Kleinerman et al., 19791. The black lesions have been termed macules; they are 1 - mm in diameter. Other lesions can occur in lungs of coal miners, including micronodules ~ < 7 mm in diameter), macronodules ~ > 7 mm in diameter), silicotic nodules, progressive massive fibrosis (PMF), Caplan's le- sions, and infective granulomas (e.g., in histoplasmosis and tuberculosis). The nodular lesions are palpable and gray to black, and they contain collagen, silica, and silicates, as well as black particles. The silicotic nodules are com- posed of whorled or laminated collagen and silica and can also contain some black particles. The lesions in PMF are black and rubbery to hard, and they can contain inlay fluid. One of the criteria for defining PMF lesions is a minimal size, 1-3 cm (Kleinerman et al., 19791. One of the major subjects of controversy in CWP is whether coal-mine dust causes emphysema and, if it does, whether it is clinically significant. In early
158 APPENDIX B reports, two British pathologists considered focal emphysema around the coal macules to "constitute the characteristic feature ofthe disease" (Gough, 1947) or to represent "the most important feature of the larger lesions," macules 2 - mm in diameter (Heppleston, 19471. The researchers did not try to relate the extent of the emphysema to clinical signs and symptoms. Heppleston (1947) did note, however, that "focal emphysema is occasionally so extensive as to leave but little of the parenchyma unaffected." It was not until recently that controlled studies were performed in an attempt to determine whether coal miners exhibited more emphysema than other work- ers. Cockcroft et al. (1982), in an autopsy series on 39 coal workers and 48 other workers, found that coal workers who smoked exhibited much more centrilobular emphysema than controls who smoked. There were too few non- smoking coal workers in the sample to determine whether work in coal mines by itself could produce significant emphysema. Ruckley et al. (1984) were able to show, in a series of 450 miners, that centriacinar emphysema was signifi- cantly more prevalent in smoking (72%) than in nonsmoking (42%) miners. They also showed that increasing concentrations of lung dust correlated with a higher incidence of centriacinar emphysema (p < 0.051. Although there were no nonsmoking, nonminer controls, the fact that a high percentage (75 %) of the nonsmoking miners with PMF had emphysema strongly suggests that non- smoking miners with PMF have a much higher risk of developing emphysema than do nonsmoking nonminers. Ryder et al. (1970), in a study of 247 coal miners, were able to show a highly significant correlation between extent of emphysema and a decrease in LEVI. It seems clear that coal miners have a higher risk of emphysema than do other workers; however, this trend has been clearly demonstrated only for smokers and possibly for nonsmoking coal miners with PMF. A small but statistically significant decrease in LEVI in active coal miners has also been demonstrated. Rogan et al. (1973) reported a statistically signifi- cant (p < 0.001) progressive reduction in LEVI with increasing cumulative exposure to coal-mine dust in a prospective study on 3,581 coal-face workers. In men without PMF, the reduction in LEVI was calculated to be 120-150 ml over 35 years at mean respirable-dust concentrations of 4 mg/m3. The authors equated that with the loss one might expect from smoking 20 cigarettes/day for the same period. The decrements in LEVI were seen in nonsmokers, as well as in smokers. They found no interaction (synergism) between cigarette-smoking and dust exposure. However, the greatest decrement in LEVI was found in workers with chronic bronchitis, and it could not be explained solely on the basis of dust exposure and cigarette-smoking. In a longitudinal study of 1,677 active coal miners without PMF, Love and Miller (1982) found that the loss of LEVI over approximately 11 years increased with cumulative dust exposure, when the effects of age, height, and smoking were allowed for. They estimated that exposure to dust at the average concentration (117 g x hours/m3) recorded
APPENDIX B 159 for the 1,677 men studied was associated with about a 40-ml loss in LEVI over an 11-year period. Lloyd (1971) was able to show a significant difference in pulmonary function between active miners and a nonmining control group. Of current miners, 20% had an LEVI less than 80% of predicted, compared with 10% of active telecommunication workers. Hankinson et al. (1977) demonstrated a statistically significant correlation between decrements in flow rates at high lung volumes and years of under- ground exposure. The decrement was more noticeable among the nonsmokers. They suggested that a dust-induced bronchitis was responsible. Those studies, as well as others, have shown that the pulmonary function deficits are rather small (except in workers with PMF). Most of the studies focused on active miners and therefore might have analyzed a self-selected group of people who can tolerate dust exposures without significant effects on pulmonary function. Some support for that idea was recently published by Hurley and Soutar (19861. In a study of 199 men without PMF who had left the coal industry before normal retirement age, a much greater loss of LEVI than expected was found (average loss, 600 ml in those who had experienced moderately high exposures). A higher incidence of chronic bronchitis also has been shown to occur in coal miners than in nonminers (Lloyd, 19711. Because the sample contained so few nonsmokers, however, the comparison was statistically significant only in the smoking group. Rae et al. (1971), in a study of 4,122 face workers from 20 collieries, found a statistically significant association between increasing ex- posure to coal dust and increasing prevalence of bronchitis in the 25-34 and 35 44 age groups. That was true in both nonsmokers and smokers. Coal-mine dust exposure was also found to be significantly related to maximal ratio of mucous-gland thickness to bronchial-wall thickness, an anatomic index of chronic bronchitis (Douglas et al., 19821. The most common method for evaluating CWP in the living is with diagnos- tic chest x rays. A rather elaborate system has been adopted internationally for the grading of x-ray changes in coal workers. CWP is divided into simple CWP and complicated COOP, or PMF. Simple CWP includes all cases in which the x-ray opacities (rounded or irregular) are less than 1 cm in diameter and is sub- divided into several categories that depend on the abundance ("profusion") of these opacities. Jacobsen et al. (1971), in a prospective study of 4,122 coal workers over a 10-year period, found a linear correlation between the progres- sion of chest x-ray category and exposure, measured as the concentration of respirable coal-mine dust per cubic meter. They estimated that exposures to coal-mine dust at 2 mg/m3 for 35 years would lead to fewer than 1 % of miners progressing from category 0/0 (no COOP) to category 2 or higher. On the basis of the studies noted here, it is clear that coal miners exhibit a higher incidence of bronchitis, pulmonary function deficits, and emphysema than nonminers. It is not clear to what extent the changes are caused by coal-
16;0 APPENDIX B mine dust, cigarette-smoking, a combination of coal-mine dust and cigarette- smoking, or other factors. The composition of coal dust and coal-mine dust might vary considerably, depending on the type of coal and the geologic area of the coal seams. Typical constituents in coal in the United States (Schtick end Fannick, 1971) are shown carbon (29.5-81%), volatile matter (5.1-36.8%), moisture (4.3-36.8%), ash (4.3-9.6%), and sulfur (0.5-0.9%~. The ash can contain quartz, aluminum silicates (e.g., mica and kaolin), and iron oxides in substantial concentrations. Coal-mine dusts might contain a variety of other minerals, depending on how much and what kind of rock dust is generated in the process of roof-bolting and cutting the coal seams. In addition, miners are occasionally exposed to fumes from cable fires and explosives. Faced with this wide range of variables and limited resources, most investi- gators studied what they considered to be the most active ingredient in coal- mine dust quartz. The pulmonary effects of coal dusts that have various con- centrations of mineral matter and of quartz have been studied sporadically in animal systems for more than 50 years. The principal end points in the investi- gations were the development of macules and fibrotic nodules in the lungs of the experimental animals. The results of inhalation studies in animals have generally shown that coal dust containing less than 10% quartz has not produced substantial pulmonary fibrosis. Ross et al. (1962) exposed rats to a mixed dust of anthracite coal with a low ash content and quartz added in various concentrations. Exposure was to respirable dust at about 60 mg/m3 for about 3,200 hours over 10-17 months. Rats exposed to mixtures of coal dust with 5% and 10% quartz showed little or no pulmonary fibrosis, whereas those exposed to coal dust with 20% quartz showed grade 2 fibrosis, and those exposed to coal dust with 40% quartz showed graded fibrosis (maximal grade, 51. Weller and Ulmer (1972) exposed animals to coal dust at 45 mg/m3 containing 30-35 % added quartz for 1,800- 2,200 hours over 18 months (rats) and 3 years (monkeys). The rats developed grade 2 fibrosis, and the monkeys developed grade 2-3 fibrosis. The monkeys developed macules thatlooked similar to those seen in coal miners' lungs. Rats did not develop macules with the same appearance as those in coal miners, probably because the anatomy of the rat lung is considerably different from that of the human lung. Gross and Nau (1967) exposed monkeys, rats, guinea pigs, and mice to lignite coal dust at 7. 8 mg/m3 containing 2 % quartz and carbon dust at 8.1 mg/m3 containing 6.5% quartz for 1,820 hours over 1 year. They observed no fibrosis and little or no reticulin in the dust aggregates in either experiment. They observed that the lungs of mice contained less dust than the lungs of rats, which contained less dust than the lungs of monkeys. Martin et al. (1977) exposed rats to coal dust at 200 mg/m3 or with logo quartz for up to 1,300 hours over 2 years. They observed small amounts of reticulin (but no fibrosis) in the dust masses in the lungs of rats exposed to coal dust and massive
APPENDIX B 161 nodular lesions with collagen fibers in the lungs of rats exposed to coal dust Mat contained 10% quartz. The nodules looked similar to fibrotic nodules seen in the lungs of coal miners with COOP, but did not look like the classical silicotic nodules composed of laminated or whorled collagen. The effects of the minerals in coal dust on the f~brogenicity of quartz have also been studied. In relatively short-term studies, Martin et al. (1972) have shown that mineral matter from coal dust can partially inhibit the fibrogenicity of quartz. They also found that quartz treated with a dialysate from coal min- eral matter has a significantly reduced solubility and induced much less colla- gen in the lungs of rats in 3-month expenments. That raises the possibility that quartz in different coal-mine dusts might have different capacities to induce fibrotic changes in the lungs of miners and that the capacity might vale with the composition of the mineral matter that is inhaled with it. Lungs of coal miners might contain quartz at relatively high concentrations (Davis et al., 19831. Lungs with macules but no fibrotic nodules contained quartz at a mean concen- tration of 800 mg per lung the same mean concentration seen in the lungs of workers with slight silicosis (Nagelschmidt, 1965.) The lungs also contained 3,300 mg of kaolin and mica, however. The high concentration of those miner- als might have modified the fibrogenic effect of the quartz. Few pulmonary function studies in animals have been reported. Moorrnan et al. (1977) exposed germ-free and conventional rats to coal dust at 10 mg/m3 for 960 hours. They showed small but statistically significant reductions in flow maximums at 10% of vital capacity in the exposed rats, compared with the controls. Weller and Ulmer (1972) performed pulmonary function tests on rats exposed to coal dust at 45 mg/m3 (with 30-35% quartz) for 18 months. They found small but statistically significant differences only in the values for compliance. No experimental models have been developed to study chronic bronchitis and centnlobular emphysema as they occur in humans. REFERENCES Alavanja, M., I. Goldstein, and M. Susser. 1978. A case control study of gastrointestinal and urinary tract cancer mortality and drinking water chlorination, pp. 395-409. In R. L. Jolley, H. Gorchev, and D. H. Hamilton, Jr. (eds.). Water Chlorination: Environmental Impact and Health Effects. Vol. 2. Ann Arbor Science, Ann Arbor, Mich. Amdur, M. O. 1985. When one plus zero is more than one. Am. Ind. Hyg Assoc. J. 46:467-475. Amdur, M. O., and J. Mead. 1958. Mechanics of respiration in unanesthetized guinea pigs. Am. J. Physiol. 192:364-368. Amdur, M. O., and D. Underhill. 1968. The effect of various aerosols on the response of guinea pigs to sulfur dioxide. Arch. Environ. Health 16:460-468. Archer, V. E. 1985. Enhancement of lung cancer by cigarette smoking in uranium and other miners, pp. 23-37. In M. J. Mass (ed.). Carcinogenesis. Vol. 8. Raven Press, New York. Axelson, O., and L. Sundell. 1978. Mining, lung cancer and smoking. Scand. J. Work Environ. Health 4:46-52.
162 APPENDIX B Balster, R. L., and J. F. Borzelleca. 1982. Behavioral toxicity of trihalomethane contaminants of drinking water in mice. Environ. Health Perspect. 46: 127-136. Batuman, V., E. Landy, J. K. Maesaka, and R. P Wedeen. 1983. Contribution of lead to hypertension with renal impairment. N. Engl. J. Med. 309: 17-21. Begin, R., A. Cantin, Y. Berthiaume, R. Boileau, S. Peloquin, and S. Masse. 1983. Airway function in lifetime-nonsmoking older asbestos workers. Am. J. Med. 75:631-638. Bellar, T. A., J. J. Lichtenberg, and R. C. Kroner. 1974. The occurrence of organohalides in chlori- nated drinking waters. J. Am. Water Works Assoc. 66:703-706. Berry, G., M. L. Newhouse, and M. Turok. 1972. Combined effect of asbestos exposure and smoking on mortality from lung cancer in factory workers. Lancet 2:476-478. Brenniman, G. R., J. Vasilomanolakis-Lagos, J. Amsel, T. Namekata, and A. H. Wolff. 1980. Case- control study of cancer deaths in Illinois communities served by chlorinated or nonchlorinated water, pp. 1043-1057. In R. L. Jolley, W. A. Brungs, R. B. Cumming, and V. A. Jacobs (eds.). Water Chlorination: Environmental Impact and Health Effects. Vol. 3. Ann Arbor Science, Ann Arbor, Mich. Buechley, R. W., W. B. Riggan, V. Hasselblad, and J. B. VanBruggen. 1973. SO2 levels and perturba- tions in mortality: A study in the New York-New Jersey metropolis. Arch. Environ. Health 27: 134-137. Burke, T. A., J. Amsel, and K. P. Cantor. 1983. Trihalomethane variation in public drinking water supplies, pp. 1343-1351. In R. L. Jolley, W. A. Brungs, J. A. Cotruvo, R. B. Cumming, J. S. Mattice, and V. A. Jacobs (eds.). Water Chlorination: Environmental Impact and Health Effects. Vol. 4, Book 2: Environment, Health, and Risk. Ann Arbor Science, Ann Arbor, Mich. Cantor, K. P., R. Hoover, P. Hartge, T. J. Mason, D. T. Silverman, and L. I. Levin. 1985. Drinking water source and risk of bladder cancer: A case-control study, pp. 145-152. In R. L. Jolley, R. J. Bull, W. P. Davis, S. Katz, M. H. Roberts, Jr., and V. A. Jacobs (eds.). Water Chlorination: Chemis- try, Environmental Impact and Health Effects. Vol. 5. Lewis Publishers, Chelsea, Mich. Cockcroft, A., R. M. E. Seal, J. C. Wagner, J. P. Lyons, R. Ryder, and N. Andersson. 1982. Post- mortem study of emphysema in coalworkers and non-coalworkers. Lancet 2(8298):600-603. Cragle, D. L., C. M. Shy, R. J. Struba, and E. J. Siff. 1985. A case-control study of colon cancer and wafer chlorination in North Carolina, pp. 153-159. In R. L. Jolley, R. J. Bull, W. P. Davis, S. Katz, M. H. Roberts, Jr., and V. A. Jacobs (eds.). Water Chlorination: Chemistry, Environmental Impact and Health Effects. Vol. 5. Lewis Publishers, Chelsea, Mich. Craun, G. F. 1985. Epidemiologic considerations for evaluating associations between the disinfection of drinking water and cancer in humans, pp. 133-143. In R. L. Jolley, R. J. Bull, W. P. Davis, S. Katz, M. H. Roberts, Jr., and V. A. Jacobs (eds.). Water Chlorination: Chemistry, Environmental Impact and Health Effects. Vol. 5. Lewis Publishers, Chelsea, Mich. Cross, F. T., R. F. Palmer, R. E. Filipy, G. E. Dagle, and B. O. Stuart. 1982. Carcinogenic effects of radon daughters, uranium ore dust and cigarette smoke in beagle dogs. Health Phys. 42:33-52. Damber, L. and L. G. Larsson. 1985. Underground mining, smoking, and lung cancer: A case-control study in the iron ore municipalities in northern Sweden. J. Natl. Cancer Inst. 74: 1207-1213. Davis, J. M. G., S. T. Beckett, R. E. Bolton, and K. Donaldson. 1980. A comparison of the pathologi- cal effects in rats of the UICC reference samples of amosite and chrysotile with those of amosite and chrysotile collected from the factory environment, pp. 285-292. In J. C. Wagner (ed.). Biological Effects of Mineral Fibers. Vol. 1. IARC Scientific Pub. No. 30. International Agency for Research on Cancer, Lyon, France. Davis, J. M. G., J. Chapman, P. Collings, A. N. Douglas, J. Fernie, D. Lamb, and V. A. Ruckley. 1983. Variations in the histological patterns of the lesions of coal workers' pneumoconiosis in Britain and their relationship to lung dust content. Am. Rev. Respir. Dis. 128: 118-124. DeRouen, T. A., and J. E. Diem. 1977. Relationships between cancer mortality in Louisiana drinking- water source and other possible causative agents, pp. 331-356. In H. H. Hiatt, J. D. Watson, and
APPENDIX B I63 J. A. Winsten (eds.). Origins of Human Cancer. Book A: Incidence of Cancer in Humans. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. DeShazo, R. D., D. J. Hendrick, J. E. Diem, J. A. Nordberg, Y. Baser, D. Brevier, R. N. Jones, H. W. Barkman, J. E. Salvaggio, and H. Weill. 1983. Immunologic aberrations in asbestos cement work- ers: Dissociation from asbestosis. J. Allergy Clin. Immunol. 72:454~61. Doll, R. 1955. Mortality from lung cancer in asbestos workers. Br. J. Ind. Med. 12:81-86. Douglas, A. N., D. Lamb, and V. A. Ruckley. 1982. Bronchial gland dimensions in coalminers: Influence of smoking and dust exposure. Thorax 37:760-764. Edling, C. 1982. Lung cancer and smokingina group ofiron ore miners. Am. J. Ind. Med.3:191-199. Gottlieb, M. S ., and J. K. Carr. 1982. Case-control cancer mortality study and chlorination of drinking water in Louisiana. Environ. Health Perspect. 46: 169-177. Gough, J. 1947. Pneumoconiosis in coal workers in Wales. Occup. Med. 4:86-97. Great Britain, Committee on Air Pollution. 1954. Report. Cmd 9322. Her Majesty's Stationery Of lice, London. Gross, P. and C. A. Nau. 1967. Lignite and the derived steam-activated carbon: The pulmonary re- sponse to their dusts. Arch. Environ. Health 14:450~60. Hammond, E. C., I. J. Selikoff, and H. Seidman. 1979. Asbestos exposure, cigarette smoking and death rates. Ann. N.Y. Acad. Sci. 330:473-490. Hankinson, J. L., R. B. Reger. R. P. Fairman, N. L. Lapp, and W. K. C. Morgan. 1977. Factors influencing expiratory flow rates in coal miners, pp. 737-755. In W. H. Walton (ed.). Inhaled Particles IV. Part 2. Pergamon, New York. Harlan, W. R., J. R. Landis, R. L. Schmouder, N. G. Goldstein, and L. C. Harlan. 1985. Blood lead and blood pressure: Relationship in the adolescent and adult U.S. population. J. Am. Med. Assoc. 253 :530-534. Harries, R. G., F. A. F. MacKenzie, G. Sheers, J. H. Kemp, T. P. Oliver, and D. S. Wright. 1972. Radiological survey of men exposed to asbestos in naval dockyards. Br. J. Ind. Med. 29:274-279. Heppleston, A. G. 1947. The essential lesion of pneumoconiosis in Welsh coal workers. J. Pathol. Bacteriol. 59:453-460. Hornung, R. W., and S. Samuels. 1981. Survivorship models for lung cancer mortality in uranium miners: Is cumulative dose an appropriate measure of exposure?, pp. 363-368. In M. Gomez (ed.). International ConferenceRadiation Hazards in Mining: Control, Measurement, and Medical As- pects, Golden, Colo., October 4-9, 1981. Society of Mining Engineers of the AIME, New York. Hurley, J. F., and C. A. Soutar. 1986. Can exposure to coalmine dust cause a severe impairment of lung function? Br. J. Ind. Med. 43: 150-157. International Agency for Research on Cancer (IARC). 1985. IARC Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans. Volume 35: Polynuclear Aromatic Compounds. Part 4: Bitumens, Coal-Tars and Derived Products, Shale-Oils and Soots. International Agency for Research on Cancer, Lyon, France. Jacobsen, M., S. Rae, W. H. Walton, and J. M. Rogan. 1971. The relation between pneumoconiosis and dust-exposure in British coal mines, pp. 903-919. In W. H. Walton (ed.). Inhaled Particles III, Vol. II. Unwin Brothers, Surrey, England. Johnson, G. A. 1911. Hypochlorite treatment of public wafer supplies: Its adaptability and limitations. J. Am. Public Health Assoc. 1 :562-574. Kanarek, M. S., and T. B. Young. 1982. Drinking water treatment and risk of cancer death in Wiscon- sin. Environ. Health Perspect. 46:179-186. Keller A. Z., and M. Terris. 1965. The association of alcohol and tobacco with cancer of the mouth and pharynx. Am. J. Public Health 55: 1578-1585. Kennaway, N. M., and E. L. Kennaway. 1936. Study of the incidence of cancer of the lung and larynx. J. Hyg. 36:236-267. Khera, A. K., D. G. Wibberley, K. W. Edwards, and H. A. Waldron. 1980. Cadmium and lead levels
164 APPENDIX B in blood and urine in a series of cardiovascular and normotensive patients. Int. J. Environ. Stud. 14:309-312. Kleinerman, J., F. Green, R. A. Harley, N. L. Lapp, W. Laqueur, R. Naeye, P. Pratt, G. Taylor, J. Wiot, and J. Wyatt. 1979. Pathology standards for coal workers' pneumoconiosis. Arch. Pathol. Lab. Med. 103:375-432. Kromhout, D., and C. de Lezenne Coulander. 1984. Trace metals and CHD risk indicators in 152 elderly men (the Zutphen study). Eur. Heart J. 5 (Abstr. Suppl. 1): 101. Kromhout, D., A. A. E. Wibowo, R. F. M. Herber, L. M. Dalderup, H. Heerdink, C. de Lezenne Coulander, and R. L. Zielhuis. 1985. Trace metals and coronary heart disease risk indicators in 152 elderly men (the Zutphen study). Am. J. Epidemiol. 122:378-385. Kuzma, R. J., C. M. Kuzma, and C. R. Buncher. 1977. Ohio drinking water source and cancer rates. Am. J. Public Health 67:725-729. Lange, A. 1980. An epidemiological survey of immunological abnormalities in asbestos workers. II. Serum immunoglobulin levels. Environ. Res. 176-183. Liddell, F. D. K., G. W. Gibbs, and J. C. McDonald. 1982. Radiological changes and fibre exposure in chrysotile workers aged 60-69 years at Thetford Mines. Ann. Occup. Hyg. 26:889-898. Lippmann, M. 1985. Airborne acidity: Estimates of exposure and human health effects. Environ. Health Perspect. 63:63-70. Lippmann, M., R. B. Schlesinger, G. Leikauf, D. Spektor, and R. E. Albert. 1982. Effects of sulphu- ric acid aerosols on respiratory tract airways. Ann. Occup. Hyg. 26:677-690. Lloyd, I. W. 1971. Long-term mortality study of steelworkers. V. Respiratory cancer in coke plant workers. J. Occup. Med. 13:53-68. Love, R. G., and B. G. Miller. 1982. Longitudinal study of lung function in coal-miners. Thorax 37: 193-197. Lynch, K. M., and W. A. Smith. 1935. Pulmonary asbestosis III: Carcinoma of lung in asbesto- silicosis. Am. J. Cancer. 24:56-64. Martin, J. C., H. Daniel-Moussard, L. Le Bouffant, and A. Policard. 1972. The role of quartz in the development of coal workers' pneumoconiosis. Ann. N.Y. Acad. Sci. 200: 127-141. Martin, J. C., H. Daniel, and L. Le Bouffant. 1977. Short- and long-term experimental study of the toxicity of coal-mine dust and some of its constituents, pp. 361-371. In W. H. Walton (ed.). Inhaled Particles IV. Part 1. Pergamon, New York. McDonald, J. C., F. D. K. Liddell, G. W. Gibbs, G. E. Eyssen, and A. D. McDonald. 1980. Dust exposure and mortality in chrysotile mining, 1910-75. Br. J. Ind. Med. 37: 11-24. McGregor, J. F. 1976. Tumor-promoting activity of cigarette tar in rat skin exposed to irradiation. J. Natl. CancerInst. 56:429-430. McJilton, C. E., R. Frank, and R. J. Charlson. 1976. Influence of relative humidity on functional effects of an inhaled SO2-aerosol mixture. Am. Rev. Respir. Dis. 113: 163-169. Meier, J. R., H. P. Ringhand, W. E. Coleman, J. W. Munch, R. P. Streicher, W. H. Kaylor, and K. M. Schenck. 1985. Identification of mutagenic compounds formed during chlorination of humic acid. Mutat. Res. 157:111-122. Meurman, L. O., R. Kiviluoto, and M. Hamaka. 1974. Mortality and morbidity among the working population of anthophyllite asbestos miners in Finland. Br. J. Ind. Med. 31: 105-112. Moorman, W. J., R. W. Hornung, and W. D. Wagner. 1977. Ventilatory functions in germfree and conventional rats exposed to coal dusts. Proc. Soc. Exp. Biol. Med. 155:424-428. Moreau, T., G. Orssaud, B. Juguet, and G. Busquet. 1982. Blood lead levels and arterial pressure: Initial results of a cross sectional study of 431 male subjects. Rev. Epidemol. Sante Publique 30:395-397. (In French.) Mossman, B. T., and J. E. Craighead. 1982. Comparative cocarcinogenic effects of crocidolite asbes- tos, hematite, kaolin and carbon in implanted tracheal organ cultures. Ann. Occup. Hyg. 26: 553-567. Nagelschmidt, G. 1965. The study of lung dust in pneumoconiosis. Am. Ind. Hyg. Assoc. J. 26: 1-7.
APPENDIX B 165 Nenot, J. C. 1977. In IAEA. Proceedings of the IAEA Symposium. International Atomic Energy Agency, Chicago., Newhouse, M. L., G. Berry, J. C. Wagner, and M. E. Turok. 1972. A study of the mortality of female asbestos workers. Br. J. Ind. Med. 29: 134-141. Ozkaynak, H., and J. D. Spengler. 1985. Analysis of health effects resulting from population expo- sures to acid precipitation precursors. Environ. Health Perspect. 63:45-55. Page, N. P., and U. Saff~otti. 1976. Report on Carcinogenesis Bioassay of Chloroform. National Cancer Institute, Division of Cancer Cause and Prevention, Bethesda, Md. (60 pp.) Pirkle, J. L., J. Schwartz, J. R. Landis, and W. R. Harlan. 1985. The relationship between blood lead levels and blood pressure and its cardiovascular risk implications. Am. J. Epidemiol. 121 :246-258. Pocock, S. J., A. G. Shaper, D. Ashby, T. Delves, and T. P. Whitehead. 1984. Blood lead concentra- tion, blood pressure, and renal function. Br. Med. J. 289:872-874. Pocock, S. J., A. G. Shaper, D. Ashby, and T. Delves. 1985. Blood lead and blood pressure in middle- aged men, pp. 303-305. In T. D. Lekkas (ed.). International Conference: Heavy Metals in the Environment. Vol. 1. September; Athens, Greece. CEP Consultants, Edinburgh, United Kingdom. Pott, P. 1775. Cancer scroti, pp. 63-68. In Chirurgical Observations Relative to the Cataract, the Polypus of the Nose, the Cancer of the Scrotum, the Different Kinds of Ruptures, and the Mortifica- tion of the Toes and Feet. Printed by T. J. Carnegy for L. Hawes, W. Clarke, and R. Collins, London. Radford, E. P., and R. G. St. Clair Renard. 1984. Lung cancer in Swedish iron miners exposed to low doses of radon daughters. New Engl. J. Med. 310: 1485-1494. Rae, S., D. D. Walker, and M. D. Attfield. 1971. Chronic bronchitis and dust exposure in British coalminers, pp. 883-896. In W. H. Walton (ed.). Inhaled Particles III. Volume II. Unwin Brothers, Surrey, England. Regan, G. M., B. Tagg. J. Walford, and M. L. Thomson. 1971. The relative importance of clinical, radiological and pulmonary function variables in evaluating asbestosis and chronic obstructive air- way disease in asbestos workers. Clin. Sci. 41:569-582. Rogan, J. M., M. D. Attfield, M. Jacobsen, S. Rae, D. D. Walker, and W. H. Walton. 1973. Role of dust in the working environment in development of chronic bronchitis in British coal miners. Br. J. Ind. Med. 30:217-226. Rook, J. J. 1974. Formation of haloforms during chlorination of natural waters. Water Treat. Exam. 23:234-243. Ross, H. F., E. J. King, M. Yoganathan, and G. Nagelschmidt. 1962. Inhalation experiments with coal dust containing 5 per cent, 10 per cent, 20 per cent and 40 per cent quartz: Tissue reactions in the lungs of rats. Ann. Occup. Hyg. 5: 149-161. Rothman K., and A. Keller. 1972. The effect of joint exposure to alcohol and tobacco on risk of cancer of the mouth and pharynx. J. Chron. Dis. 25:711-716. Ruckley, V. A., S. J. Gauld, J. S. Chapman, J. M. G. Davis, A. N. Douglas, J. M. Fernie, M. Jacobsen, and D. Lamb. 1984. Emphysema and dust exposure in a group of coal workers. Am. Rev. Respir. Dis. 129:528-532. Ryder, R., Lynons, J. P., H. Campbell, and J. Gough, . 1970. Emphysema in coal workers' pneumoco- niosis. Br. Med. J. 3:481-487. Salg, J. 1977. Cancer Mortality Rates and Drinking Water Quality in the Ohio River Valley Basin. Ph.D. dissertation, University of North Carolina at Chapel Hill. (144 pp.) Saracci, R. 1977. Asbestos and lung cancer: An analysis of the epidemiological evidence on the asbestos-smoking interaction. Int. J. Cancer20:323-331. Schimmel, H. 1978. Evidence for possible acute health effects of ambient air pollution from time series analysis: Methodological questions and some new results based on New York City daily mortality, 1963-1976. Bull. N.Y. Acad. Med. 54:1052-1108. Schimmel, H., and T. J. Murawski. 1975. SO2Harmful pollutant or air quality indicator? J. Air Pollut. Control Assoc. 25:739-740.
166 APPENDIX B Schimmel, H., and T. J. Murawski. 1976. The relation of air pollution to mortality. J. Occup. Med. 18:316-333. Schlesinger, R. B. 1985. The effects of inhaled acids on respiratory tract defense mechanisms. Envi- ron. Health Perspect. 63:25-38. Schtick, D. P., and N. L. Fannick. 1971. Coal in the United States, pp. 13-26. In M. M. Key, L. E. Kerr, and M. Bundy (eds.). Pulmonary Reactions to Coal Dust: A Review of U.S. Experience. Academic Press, New York. Seidman, H., I. J. Selikoff, and E. C. Hammond. 1979. Short-term asbestos work exposure and long- term observation. Ann. N.Y. Acad. Sci. 330:61-89. Selikoff, I. J., E. C. Hammond, and J. Churg. 1968. Asbestos exposure, smoking, and neoplasia. J. Am. Med. Assoc. 204:106-112. Sheppard, D., W. S. Wong, C. F. Uehara, J. A. Nadel, and H. A. Boushey. 1980. Lower threshold and greater bronchomotor responsiveness of asthmatic subjects to sulfur dioxide. Am. Rev. Respir. Dis. 122:873-878. Stevens, A. A., R. C. Dressman, R. K. Sorrell, and H. J. Brass. 1985. Organic halogen measurements: Current uses and future prospects. J. Am. Water Works Assoc. 77(4):146-154. Topping, D. C., and P. Nettesheim. 1980. Two-stage carcinogenesis studies with asbestos in Fischer 344 rats. J. Natl. Cancer Inst. 65:627-630. U.S. EPA (Environmental Protection Agency), Environmental Criteria and Assessment Office. 1980. Long-Term Effects of Air Pollutants in Canine Species. EPA-600/8-80-014. U.S. EPA, Cincinnati, Ohio. U.S. EPA, Environmental Criteria and Assessment Office. 1982. Air Quality Criteria for Particulate Matter and Sulfur Oxides. Vol. III. EPA-600/8-82-029a. U.S. EPA, Research Triangle Park, N.C. U.S. EPA, Environmental Criteria and Assessment Office. 1984. Air Quality Criteria for Lead, Re- view Draft. EPA-600/8-83-028. U.S. EPA, Research Triangle Park, N.C. (4 vole.) U. S. Public Health Service, Division of Water Supply and Pollution Control. 1964. Inventory of Municipal Water Facilities (1963). Region V. Illinois, Indiana, Michigan, Ohio, and Wisconsin. PHS Publication No. 775. Vol. 5. U. S. Government Printing Off~ce, Washington, D.C. (Available from NTIS as PB-218194.) (152 pp.) Utell, M. J. 1985. Effects of inhaled acid aerosols on lung mechanics: An analysis of human exposure studies. Environ. Health Perspect. 63:39~4. Victery, W., A. J. Vander, H. Markel, L. Katzman, J. M. Shulak, C. Germain. 1982a. Lead exposure, begun in utero, decreases renin and angiotensin II in adult rats (41398). Proc. Soc. Exp. Biol. Med. 170:63-67. Victery, W., A. J. Vander, J. M. Shulak, P. Schoeps, and S. Julius. 1982b. Lead, hypertension, and the renin-angiotensin system in rats. J. Lab. Clin. Med. 99:354-362. Wagner, M. M. F. 1980. Immunology and asbestos, pp. 247-251. In J. C. Wagner (ed.). Biological Effects of Mineral Fibers. Vol. 2. IARC Scientific Pub. No. 30. International Agency for Research on Cancer, Lyon, France. Wagner, J. C., C. A. Sleggs, and P. Marchand. 1960. Diffuse pleural mesothelioma and asbestos exposure in the North Western Cape Province. Br. J. Ind. Med. 17:260-271. Wagner, J. C., G. Ber~y, J. W. Skidmore, and F. D. Pooley. 1980. The comparative effects of three chrysotiles by injection and inhalation in rats, pp. 363-372. In J. C. Wagner (ed.). Biological Effects of Mineral Fibers. Vol. 1. IARC Scientific Pub. No. 30. International Agency for Research on Cancer, Lyon, France. Walker, A. M. 1984. Declining relative risks for lung cancer after cessation of asbestos exposure. J. Occup. Med. 26:422~26. Wehner, A. P., R. H. Busch, R. J. Olson, and D. K. Craig. 1975. Chronic inhalation of asbestos and cigarette smoke by hamsters. Environ. Res. 10:368-383. Weiss, S. T., A. Munoz, A. Stein, D. Sparrow, and F. E. Speizer. 1986. The relationship of blood lead to blood pressure in a longitudinal study of working men. Am. J. Epidemiol. 123:800-808.
APPENDIX B 167 Weller, W., and W. T. Ulmer. 1972. Inhalation studies of coal-quartz dust mixture. Ann. N.Y. Acad. Sci. 200:142-154. Wright, J. L., and A. Churg. 1984. Morphology of small-airway lesions in patients with asbestos exposure. Hum. Pathol. 15:68-74. Wynder E. L., I. J. Bross, and R. M. Feldman. 1957. A study of the etiological factors in cancer of the mouth. Cancer 10: 1300-1323. Yamagiwa, K., and K. Ichikawa. 1918. Experimental study of the pathogenesis of carcinoma. J. Cancer Res. 3:1-29. Young, T. B., and M. S. Kanarek. 1983. Matched pair case control study of drinking wafer chlorination and cancer mortality, pp. 1365-1380. In R. L. Jolley, W. A. Brungs, J. A. Cotruvo, R. B. Cumming, J. S. Mattice, and V. A. Jacobs (eds.). Water Chlorination: Environmental Impact and Health Ef- fects. Vol. 4, Book 2: Environment, Health, and Risk. Ann Arbor Science, Ann Arbor, Mich.
