Tobacco Documents Online

) ETS AND NEART DISEASE 465 ence in the workplace. during travel, and at leisure. From the standpoint of' worker ETS exposure. the negative results of Lee et al. (24) are more rele- vant than positive results of spousal studies that do not include ETS expo- sure outside the home environment. Carbon Monoxide Heart disease mortality rates have been reported for workers exposed to high levels of carbon, monoxide from vehicular emissions (tunnell workers. bus drivers, parking attendants) and industrial furnaces (steellfoundry. coke ovenL chemical manufacture) (25.26). However. the results of occupational exposure to high levels of carbon monoxide do noo support~ the argument that this substance contributes to heart disease associated~ with ETS expo- sure. in which reported levels of the gas are a tirny fraction of'the TLV. Carbon Disul,/ide/Carbon-vt Sulfide These two compounds are linke& b~ the fact that the former is ani indus- trial chemical reported to be associated with heart! disease among workers producing viscose rayonifibers. This compound is metabolized to carbonyl sulfide. which happens to be a reported SSS constituent. The concentration of' carbonyl sulfide is so low that it is unlikel'y to attain the TLV (Table 2: 54.945 cigarettes to attain TLV ). However. it is important to discuss mor- ialit~ studies of'ravon viscose workers. because other thanicarbon monoa- ide. carbon disulfide is the onlv industrial chemical for which there are ex- tensive data on, an association with ischemic heart disease. Ini a critical review of the toxicologic literature on carbon d'isulfide. Beauchamp et al. (27)ireviewed data on.the mortality rates of viscose rayon workers. In Fin- land'~ where there is a: high incidence of ischernic hearti disease. a signifi- canth-y higher mortality rate has been reportedlamong exposed workers com- pared to a control group. However, in Japan where there is a notably lower incidence of ischemic heart, disease. no increased mortality rate has been reported among viscose rayon workers. The excess deaths attnibuted to car- boni disulfide became apparenti if predisposing risk factors existed. such ass hypertension, hyperlipidemia. and excessive intake of cholesterol andisatu- rated fats (27,28)! The above observations are essential to consider in attempts to interpret mortality studies on ETS exposure. Dietaryy intakes of cholesterol and fatty food were not considered as a confounding factor in mortality studies relat- ing to workers expose& to the industrial chemicals listed in Table l(with Method A notation). The reported higher susceptibility of Scandinavians too heart disease is reflected by the lower TLV (l5 mg/m') compared to:the TLV in other European countries and the United States (30 mg/m') (18.19).

466 ETS AJx'D HEART DlSE.4SE: Polvcyciic Aromatic kvdrocarbons (PAFI), lt has been suggested by proponents of the ETS-heart disease hypothesis that Scandinavian roofers show excess mortalitvv for ischemic heart disease (3). They extrapolate from PAH-exposed roofers to ETS-exposed workers without recognizing the difference in composition of PAH. Exposures to PAH' among coke oveni workers. creosote wood appliers, and asphalt road builders have not been reported to be associated! with, excess mortality for heart disease but have been reported to be associated with excess mortality for lung cancer. From the standpoint of chemical composition:of PAH ex- posures determined by nature of product. PAH exposures of roofers are ir- relevant to ETS exposure (see also Method F). Heat!Y Metals Mortality studies on work-related exposure have been reviewed by Kris- tensen (161. Lead and cadmium workers have been reported to show a higher mortality rate from heart, disease and' hypertension. Iln the absence of ex- perimental'l animal studies. heart disease is likely to be a complication of hypertension, rather than a direct, effect of lead or cadmium onithe heart and coronary vessels. The suggestion that heart disease may, be associated with workplace exposure to arsenic or cobalt can. be traced to instances of beer drinking contaminated with either of these metals. and subsequent death from cardiomyopathy; Method B: Exercise Testing and Angina Pectoris Exercise testing is essential for the diagnosis of ischemic heart disease (?9). A positive diagnosis is based on the appearance of chest pain or clas- sical angina pectoris after completion of standardized exercise on a treadrnilll or bicyclae ergometer. Exercise testing has also beemused to evaluate sever- ity of arteriosclerotic heart disease based on time of onset of an ischemic pattern in the electrocardiogram as well as the appearance of cardiac arrhythmias. ETS Exposure of Anginal Patients All available reports on exercise testing do not relate to specific occupa- N tional groups comparing two subgroups: with ETS exposure an& no ETS C exposure. There are two studies on anginal patients that suggested to the ~ investigators that ETS exposure during bicycle ergometry may shorten the ~+? time perio& to onset of chest pain. The first study. reporte& in 1978. con- CA sisted of a group of ten American male veterans (30). For variow, reasons. ~ ~ N

ETS A~N1)~~MEAR7 UISEASE ~ 4fu~7~ the 197t+'protocol for exercise testing was evaluated by an ad hoc committee of the Environmental'Protection Agency. In 1983'. the committee concluded that the method used on American male veterans "did not meeta reasonable standard of scientific quality" (13I ). Ini 1987. a: stud'ti of exercise test- ing during ETS exposure was reported by Soviet investigators (32)h The re- sults were essentiallv similar to those reported' from American veterans. lt is this author's opinionithat shortening onset of anginal pain during exercise testing as a result of ETS exposure has not been proved pending evaluation of the Soviet protocol. Anti-anginal drugs sold in the United States arne sup• ported by results of exercise testing in European laboratories that have been approved hy the U.S, Food and Drug Administration and so far. the list d'oes not include arn Soviet laboratories. Influenre of Carbon Monoxide on Exercise Testing Proponents of' the theory that ETS exposure aggravates angina pectoris emphasize the presence of carbon monoxide in ETS. in spite of the fact that the concentration inhal'ed'is 3100 times lower than mainstream smoke. Blood carboxyhemoglobin levels of suhjects exposed to ETS in public places range from I to 3 n among nonsmokers. Slight elevations of blood carboxyhemo- globin level (to 2 and 3.97() have beenireported following administration of carbon monoxide in air (l001and''_30 mgiM') (33). Exercise testing of heart disease patients was reported'to result in an ischemic pattern of electrocar- diogram at these blood carboxyhemoglbbin levels. However, as indicated'in Table 2. this would require more than )00 and '_'G10, cigarettes burning in, a sealed enclbsure of I001m' for carbon monoxide to attain about 2 and'4 times the TLV. respectively. ETS Exposure as Risk Factor jor Angina Pectoris Proponents of the claim that'~ ETS exposure aggravates angina pectoriss have not considered the complexities of the disease separate from other manifestrations or complications of ischemic heart disease (Le., acute myo- cardial infarction and sudden deaths). Althoughi prospective and retrospec- tive studies report that cigarette smoking is one of many risk factors for acute myocardial infarction and sudden deaths. the data on angina pectoris are evenimore complex. The 1983 repornof the Surgeon General onicardio- vascular disease, referring to risk factors, conclud'ed! that "variation in the strength ofassociation between smokingand angina pectoris may be influenced by . . . methodological considerations'" lref. 34. p. 70). More recently. it has beeni argued that the 30-year results of an ongoing prospective studj at Framingham. Massachusetts. indicate that cigarette smoking is a negative risk factor in women li.e., incidence lower in women smokers compare& to

40 ETS AJ1'U H£ART 1)ISEASE uomen nonsmokers) (3i1. The results in men have indicated either a positive or no significant relationship between cigarette smoking and angina pectoris, depending on methodological variation. Some studies relating to cardiac pa- tients admitted to hospitals report that after an initial cardiac episode. the prognosis is not~ influenced b~ smoking (3637). After the initial' infarction. prior smoking was not associated with the severity of subsequent compli'. cations. These observations on cigarette smoking in relation to the prognosis of myocardiallinfarction and the influence of angina pectoris raise additional questions. How can ETS. a dilute mixture of tobacco smoke components in air, aggravate angina pectoris or influence the prognosis of acute myocardial infarction, in light of recent inconsistencies in data derived from smokers° Method C: Coronar% BJood Flos lndicators Coronar~ arteries visualized b\ angiograph\ can shoti obstruction that is organic (arteriosclerosis and'thrombosis) or nonorganic (ivasospasm) in na- ture. Total coronary blood floti is measured by a tracer clearance technique. Patients with ischemic hearti disease shoH a reduction in coronary blood flo\A that is limited to an infarcted'area. When infarction is detected in work- ers previously exposed to carboni monoxide or carbon disulfide. it is not possible to isolate the potential association with chemical exposure from other potentially confounding risk factors. Carbon monoxide alone. by in- creasing carboxyhemoglobin. cani increase coronary blood ilou: but the re- sult would be an oversupply of blood, \vithout reduced oxygen utilizationi beca.use of poisoning oxidative enzymes. Myocardial metabolism requiress the sampling of blood from the coronary sinus and a systemic artery to ob- tain arteriovenous differences of oxygen. carbon dioxide. lipoproteins. and glucose metabolites. There are more directi methods formeasuring coronary blood flow in experimental animals (dog. cat. pig. monkeyl: The relhtive importance of metabolic and neurohumoral control has been evaluated in experimental animals [see reference cited by Bove (3R)]. It has not been possible to reproduce coronary heart disease by exposure to tobacco smoke. which contains nicotine levels higher than ETS, so it is doubtful!that existing animal models can give positive results from ETS exposure. Nitroglycerin and organic nitrates are useful vasodilators for the relief of angina pectoris. The pharmacologic action, of nitrogl'Ncerin is manifeste& in workers who are expose&daily to nitroglti cerin and ethylene gl}~col dinitrate. but after a weekend of nonexposure, developichest pain on Monday morn- ing. Workers suffer from vasospastic angina as a result of nitrate withdrawal during the weekend and are relieved upon resuming nitrate work exposure.. Autopsied workers did not show coronary arterial obstruction, confirming the occurrence of vasospastic angina brought abour: hy weekend withdrawal from nitrate. Workers were acclimatized to the nitrate level in work environ- ment ( 114-16).

ETS AA'l') NEART l)ISEASE 409 CORONARY ATHEROSCLEROSIS The term coronary atlterc,.cclc•rrrai.v used in this chapter refers to histo- patholla:ic changes in arteries leading to i.sclierriic huurt di,~c•n.%c- (see preced- iitc section). Although both terms are included in, cnrorrur-N heurt di.Nru.%c-. there are differences in methodolog~. This section is de~oted to progressive organic lesions of coronary arteries. the methods for detectiom and their evolution, based on human observations and animal' experimentation. The focus is on industrial exposure to carbon disulfide andl carbon monoxide. because of the relativelti greater amounts of data on these substances. The potential relevance of these industrial chemicals to ETS exposure is also discussed. The demonstration of coronary atherosclerosis id'eally should include his- topatihologic evidence derived from autopsy (Method D). This has been ac- complished for worker exposure to carbon disulfide. ti hich has been sup- ported bN the occurrence of hyperlipidemia in exposed workers and coronarn atherosclerosis in experimental animals (Metho& EI. Onithe other hand'. some industrial chemicals are associated'.+ith the development of cor- onars atherosclerosis based on animal experiments onl} or on hematolbcic changes in workers that in animals contribute to aortic atherogenesis (see entries in Table 1). Some of these observations have been used, to supportt the claim that ETS exposure is involved in coronar} atherosclerosis. A dis- tinction is made between concepts derived from human studies (Method D). animal'.experiments (lvlethodlE). and in virro techniques bMethodlFl. Method D: Coronary AngiographN and Histopatholog) The most direct method for diagnosis of coronary atherosclerosis ik b~~ histopathologic examination and coronary angiogram. AlthouFh, there are isolated reports that workers exposed to carbon monoxide suffer from in- creased coronarv atherosclerosis (antemortem or postmorteml. this expo- sure is confounded by competing risk factors such as personal habits. famil- ial historN, and environmentallpollution. Among viscose rayon workers, the occurrence of coronary atherosclerosis reported at autops} of workers dying of heart disease led to mortalitN studies (Method A). Workers are also re- ported tb suffer from hyperlipidemia. which is not entirell due to carbon disulfide exposure. Tt is difficult to replicate earlier studies on workers using modern techniques of diagnosing coronary atherosclerosis, because expo• sure levels have come under, strict regulation. There are no case reports of coronary atherosclerosis iniwor~kers exposed to a single polynuclear aromatic amine because workplace exposure is to mixtures that inclhde tienzolulpyrene. Only research Iaboratory workers in- vestigating benzola)pyrene are candidates for long-term exposure. and so far there has been no report of a higher incidence of heart disease. There are

470 ETS A'ND HEART DISEASE also no case reports of coronary atherosclerosis from prolonged exposure to the heav~ metals and nitrogenous compounds listed in Table l. Method E: Coronary Atherosclerosis in Experimental Animals Repeated attempts to induce coronary atherosclerosis in experimental anr imals by inhalation of cigarette smoke have fail'ed'. Additional feeding with a cholesterol-enriched diet has reportedly led to the development of athero- sclerosis notl involving coronary arteries. ln baboons. after 2-3 years of oral feedine of cholesterol and saturated' fat, and d'aily inhalation of cigarette smoke. arterial lesions were compared between smokers andl controls. Among male baboons. the extent of carotid atherosclerosis was greater in smokers than in controls. but there were no significant differences in athero- sclerosis of the aorta. coronarv arteries. iliac-femoral, and bronchial arter- ies. Among female baboons., there were no significantidifferences in athero- sclerosis between smokers and controls (391'. The same general remarks appl} to experimental testing of carbon mon- oxide in levels far exceeding those reported for ETS exposure. Rabbits. pi- geons. and chickens are reported to need supplementary feeding of choles- terol to show carbon monoxide-induced aortic atherosclerosis (40). Carbon disulfide is the only industrial chemicalireported to cause athero- sclerosis in animals without supplemental cholesterol feeding. Coronary and aortic atherosclerosis and myocardial lesions were detected in rats after 4 months of inhalation exposure (28). There were elevations of serum choles- terol. phospholipid. and triglycerides, indicating similt3rity to the human form of atherosclerosis. Other investigators have tested carbon monoxide and benzo[a]pyrene an&have not observed hyperdipidemia and atheroscle- rosis similar to those reported for carbon disulfide. ln the past. research on carbon monoxide. benzo[a]pyrene. and other polynuclear aromatic hydro• carbons has not been directed to a comparison with carbon disulfide. Polynuclear aromatic hydrocarbons have been reported to induce aortic atherosclerosis in pigeons and chickens (41'--44). It has been speculated that these studies in birds relate to human subjects exposed to ETS ('). There are several reasons for the inapplicability of results of these bird experiments to coronary atherosclerosis: (a) 7,l?-dimethv lbenzo(a,li)anthracene and 3- methycholanthrene are not known, to be present in ETS: (b) although benzo[ki]pyrene is reportedly present in ETS, the d'ose administered. 50 mg- kg injection. is farfetched compare&to concentration levels in SSS. which is 0.00009 mg/cigarette: (c) hepatic metabolism is essentiallfor atherogenesis in one strain, but not in the other strain. a sequence that applies to oral or injected compounds but not to the inhalation route: andl(d) the typical result is aortic atherosclerosis and! rarely coronary atherosclerosis. Aortic athero- sclerosis is different fromi coronary atherosclerosis because ofl myocardial extravascular support in the latter. There are intracardiac mechanisms that influence coronary circulation, which are absent in other arterial beds.

ETS AND HEART DISEASE 471 There are long-term animal experiments designed to study carcinogenicit~ or polvnuclear aromatic hydrocarbons and. so far. coronary atherosclerosis t has not been reported in sacrificed animals. Method F: In Vitro Studies of Hematologic Factors Hematologic factors include alterations in hemoglobin oxygen transporv such as carbon monoxide and' methylene chloride increasing carboxyhemo- globin: andianiline and 2-toluidine leading to methemoglobinemia. The ulti- mate consequence is a reduced supply of oxygen and presumablN athero- sclerosis resulting from carbon monoxide. However. prolonged testing withh methvlene chloride or aniline has not been reported to produce experimental atherosclerosis, suggesting that these two industrial chemicals reduce hemo- globinioxygen transport differently from carbon monoxide. Several techniques have been deveiopedifor the specific purpose of dis- co\ering therapeutic agents for the prevent6on, suppression. and reversal of atherosclerosis. Drugs for influencing blood platelets, blood lipoprotein lev- els. andlend'othelial vulnerability evolved from application of in i•rtro testing of blood derived' from patients with ischemic heart disease. as well as pe- ripheral vascular diseases. The same techniques for identifying therapeutic agents have also been applied'i to investigating ho.+ carbon monoxide and ETS might play a role in:atherosclerosis. The interpretation of results de- rived from one test' has been extended to include the entire progressioni of atherosclerosis even though the test was intended to show a therapeutic. rather than toxic. effect of chemical agents.. lii ritco tests have been applied to blood from ETS-exposed subjects. . based on the assumption that any reported effect will contribute to coronary atherosclerosis. It should be pointe& out that chemically induced platelet aggregatiom leads to vascular clot formation. which does not~ necessaril\ in- volve interaction with endothelial cells and the formation of atherosclerotic plaque. Also, in the laboratory. it has not been possible to initiate aortic plaque formation by exceeding the normal level of fibrinogen. Any reported increase in fibrinogen level in the blood of ETS-exposed subjects may not, be relevant to a potential relationship with coronary atherogenesis. Iit is con- ceivable that. for some people. ETS exposure may be perceived as stressful. with release of catecholamines. and that catecholamines are responsible for in vitro testing results. It has not, been possible to conduct a double-blin& testing of ETS exposure since both, investigator and subject can detect ETS presence. Platelet Aggregation Exposure of healthy nonsmokers to ETS is alleged to alter resulus of in rirJ-o testing of platelets in platelet-richi plasma. Aggregation of platelets is

472 ETS AA'D, HEART DlSEASE tested hv the follou inF agents added in riiro: edetic acid and formaldehyde or prostaglandin. The possibility that ETS exposure increases platelet~aggre- gation is alleged to be an important step in the evolution of coronary athero> sclerosis in nonsmokers (''). In ,•iiro studies of platelet aggregation in blood derived'from~smokers have reported inconsistent results, which question the applicability of this method to ETS exposure in nonsmokers. Platelet aggregation testing using whole blood reported no statistically significant differences between nonsmokers and smokers (45). Cigarette smoking is reportedly associatedi k ith altera- tions in platelet factors involved in thrombus formation. but the change has been attributed'to the presence of carbon monoxide levels higher than those reported in subjects exposed to ETS (46). 1,7 vitrr testing does not necessar- ily reflect events in rivo. Although platelets may be activated in viro, the} become attached to ervthrocvtes or form platelet aggregates during the col- lectioni and centrifugation needed to make platelet-rich plasma. There is some evidence that acti%ated platelets are lost from supernatant "platelet- richiplasma." which includes older or less acrive platelets, Plasma Fibrinogen Levels Another in riii-n test for a clotting factor has beeni added to: the list of reports supporting the ETS-heart~ disease hypothesis. Patients with isch- emic hearti disease tiere questioned about their smoking habits. and non- smokers were queried for ETS exposure in the v.orkplace and household. Control suhlects were derived from the same community in Australia (471.. Ih was reported'that the collected blood samples showed higher fibrinogen concentrations among current smokers than nonsmokers. Subjects exposedl to ETS had higher levels thani those not, exposed. The differences Here not statistically significant because of the high variability of measured fibrinogen levels. According to the questionnaire responses. levels of ETS exposure at work were reported to be higher than at home, but the estimated odds ratio for heart disease was less than one. The investigators interpreted their re- sults to indicate inaccurate reporting of ETS exposure or the possihilitythat household exposure to ETS is associated more with heart disease than is workplace exposure. The potential relevance of fibrinogenilevels in relation to ETS exposure is further questioned' by observations that psvchosocial factors may influence the plasma fibrinogen concentration in patients with ischemic heart disease (4Y).. CARDIAC ARRHYTHAIIA AND M19ti'OPATHI The third and last group of methods for establishing cardiac toxicologic profiles for industrial chemicals relate~ to alteration, in cardiac function.

ETS AA'/2 HE-ART 1)lSEAS£ 47-3 The methods are intended to detect irregularities in heart beat or rhythm. to measure excitabilitN of the intact heart, and to record' electnicaC propertie~ of excised atrium and' papillary, muscle. Cardiac output is measured by° the tracer dilution technique and ventricular imaging in patients: invasive pro- cedures are required for application of the Fick principle in patients and insertion of blood flow recorders in,experimental animals. Perfusion of the excised heart offers an opportunity of measuring myocardial contractility and metabolism. Enzymatic studies and electron microscop~ complete the techniques for~ detecting cardiomyopathy. All these procedures have been applied to determine the occurrence and mechanism for two groups of dis- eases: irregular heart beat or arrhythmia, and cardiomyopathy. Method G: Irregular Heart Beat Industrial chemical poisoning can be manifested' by irregularities of heart beat or cardiac arrhythmia. in the order of increasingseverity: ranging from tach\cardia;or bradycardii3. atrioventricular block. atrial or ventricular ex- tra~_~ stole. atrial fibrillation. to ventricular fibrillation and cardiac arrest. The benign forms lup to atrial fibrillationYare reversible by stopping chem- ical e.posure., but ventricular fibrillation and cardiac arrest: reqNire heroic efforts. Poisonings characterized by cardiac arrhythmies have been reported for the following (see Table 1. Method G): mosti halogenated and nonhalo• genated'solvents. some nitrogenous compounds, one heavy metal (lead), and one oxide (carbon monoxide). The arrhythmia results from a direcn action of the chemical on the heart. specifically by altering excitability. conduction. and refractoriness of one or more of the following: atrial muscle. atrioven- tricular node. conducting system. and ventricular muscle. The effects have been reported in appropriate human studies and animal experimentation. The occurrence of poisoning by industrial chemicals does not support the proposition that since the same chemicals ma~ be reported:at minute le\els in ETS. then ETS also may lead to the development of heart disease in workers. Method H: ExperimentallN Induced Cardiom}opathy The most extreme example of unjustified application of results from ani- mal experiments to ETS exposure of nonsmokers is as follows: in the course of attempting to determine whether long-term cigarette smoking leads to cardiomyopathy; rabbits were exposed in an infant incubator (49). It was reported that all the smoke fromithree burning cigarettes entered the inlet of the incubator throughia mechanical device and': rabbits were kept for 30 min. This description appears to this author as a sealed! chamber with cigarette smoke entering the inlet for 30-min periods. Several groups of rabbits were

474 ETS AA'DHEART UISEASE sacrificed: controls, after one 30-min exposure. twice daily exposure for 2 weeks, and twice daily exposure for 8 weeks. The heart was studied' for mitochondrial', oxidative processes. There was a decrease in respiration as well as in phosphorylation rate that was interpreted by the investigators as card'iotnyopathy: The investigators recognized that carbon monoxide in the incubator was probably responsible for metabolic changes but they did not monitor air or blood levels. Hugod and collaborators (50-52) exposed rabbits to one of the following mixtures: carbon monoxide 220 mg/m' or four times TLV; carbonyl sulfide 130 mg'm' or five times TLV: nitric oxide 6 tng/m' or one-fifth the TLV The rabbits were in air-tight exposure chambers containing freely flowing air or predetermined! mixtures in air for periods ranging from I to 7 weeks. The results of 140 rabbits sacrificed for electron microscopic examination per formed'blindl\ showed no morphological signs of myocardial damage. The four vapor constituents. in lfvels far exceeding ETS levels, were not asso- ciated with ultrastructural changes in rabbit heart. signifying the absence of cardiomyopathy.. The rabbit exposure studies described~ above were extended to include biochemical and histomorphologic investigation of atherosclerosis. Expo- sure to each of the four gas-air mixtures was not related to intimal damage of the aorta and coronary arteries. The negative results noted for carbonyJl sulfide exposed rabbits do not support the claim,thai this known metabolite for carbon disulfide is responsible for coronary atherosclerosis reported' by other investigators. Cardiomyopathy has been reported following exposure to halbgenatedisol- vents, based on case reports of poisoning and experimentallstudies on intact and perfused heart. Cardiomyopathy from heavy metals is described'in case reports of individuals drinking beer from containers that leached' arsenic. cadmium. or lead (16): Cardiomyopathy from hydrocyanic aci&is also based on case reports of poisoning and is readily, supported by biochemical studies of heart muscle. Carbon monoxide is probably the most~ frequently encoun- tered industrial and household chemical associated with death bvy cardio- myopathy. History of exposure to vehicular emissions or household natural gas is verifiable by blood'analysis for carboxyhemoglobin. Among nonhal- ogenated solvents. only phenol has been reportedly related to cardiomyop- athy.(161~ CONCLUDING REMARKS Among more than 32 industrial chemicals potentiallN related to heart dis- ease, only four substances or chemical classes have extensive supportive evidence: carbon monoxide, carbon disulfide. ethylene glycol dinitrate and organic nitrates, and methylene chloride and halogenated solvents. The ef-