Tobacco Documents Online

OVERVIEW AND MAJOR CONSIDERATIONS IN THE TOXICITY TESTING OF LOW IGNITION-POTENTIAL CIGARETTES Jeffrey Harris, M.D., Ph.D. Internal Medicine Associates - 2 Wang Ambulatory Care Center 605 Massachusetts General Hospital Parkman Street Boston, MA 02114 21 Aug 1992

page 1 ABSTRACT Both mainstream andsidestream cigarette smoke are complex chemical mixtures. In view of this chemical complexity, it should be no surprise that cigarette smoke has multiple, diverse effects on human health. Nor should it be unexpected that multiple chemicals in cigarette smoke contribute to any single adverse health effect. The diverse human health consequences of cigarettee smoking are briefly reviewed. Many experimental laboratory models have been developed to study the mechanisms of cigarette smoke-induced disease. These laboratory models are not always convertible into practical, standardized test systems that quantitatively compare one cigarette prototype with another. In view of the multiplicity of health effects and mechanisms of smoke-induced health damage, no single test or battery of tests can capture all possible health endpoints. While analyses of smoke constituents and studies in laboratory animals are feasible, human epidemiological studies are not practical for short-term assessment of small differences in the toxic effects of various cigarette prototypes. Cigarette smoke samples for chemical analysis and biological testing need to be collected in a manner that approximates human cigarette puffing as closely as technically feasible. In formulating a testing plan, the CPSC essentially has two options: a design-based testing plan, in which individuaL, pre-selected cigarette design parameters, such as paper porosity or percent expanded tobacco, are systematically varied and tested; and a performance-based testing plan, in which complete cigarette prototypes, and not individual design parameters, are evaluated. Some testing protocols entail a "screening paradigm." Multiple tests are performed in sequence. If a prototype fails any particular test in the sequence, the prototype is rejected and no further tests are performed. Other multi-test protocols allow for tradeoffs among costs and benefits. An unfavorable result at any point along the testing sequence does not necessarily result in rejection. The only governmentally-mandated, health-oriented testing of commercial cigarette brands is the measurement and reporting of "tar," nicotine and carbon monoxide in mainstream smoke by the Federal Trade Commission. With this exception, none of the toxicity tests described by the Expert Panel are routinely performed on existing cigarette brands by any governmental agency. The contents of currently marketed cigarettes are proprietary information. Specific additives, tobacco

page 2 composition, and other design features are not publicly disclosed.

page 3 SCOPE OF THE EXPERT PANEL REPORT This report addresses: scientific aspects of the design of cigarette toxicity testing systems; the selection and sequencing of particular tests; the reliability, feasibility, and costs of particular tests; and the interpretation, limitations, uses and misuses of test results. In addition to the present Overview chapter, the report contains specific chapters on: (i) collection of smoke samples from prototype cigarettes for toxicity testing, by Dr. Harold Pillsbury (Chapter B); (ii) measuring the dosage of smoke constituents actually absorbed by humamsmokers of different cigarette prototypes, by Dr. David Burns (Chapter C); (iii) measuring the amounts of specific chemicals contained in the collected smoke, by Dr. Dietrich Hoffmann (Chapter D); (iv) toxicity testing in single-cell (°'in vitro") systems, by Dr. Gary Gairola (Chapter E); (v) toxicity testing in whole animal ("in vivo") systems, by Dr. Dietrich Hoffmann (Chapter F); and (vi) research needs for developing methods to collect additional data (Chapter G, input needed). The Expert Panel has not made policy recommendations. The Panel members did not perform any testing of prototypes in connection with this Report. SOURCES OF INFORMATION In preparing this Report, the Expert Panel relied upon: the Final Report of the Technical Study Group on Cigarette and Little Cigar Fire Safety under the Cigarette Safety Act of 1984 (28); background papers issued in connection with the Technical Study Group Report (17;27); reports issued by the National Institute for Standards and Technology (and its predecessor, the National Bureau of Standards) in connection with low-ignition potential cigarettes [11;18]; communications from members of the TAG, CPSC staff and DHHS Staff; the published scientific literature; as well as its own expertise and experience. No proprietary or confidential information was requested, offered, or considered. MAINSTREAM VERSUS SIDESTREAM CIGARETTE SMOKE Both smokers and nonsmokers can incur adverse health effects from the smoke of burning cigarettes. Smokers inhale mostly

page 4 "mainstream (MS) smoke" that is drawn through the burning tobacco column and filter tip and exits through the mouthpiece of the cigarette. Nonsmokers inhale mostly "sidestream (SS) smoke" that is emitted into the surrounding air between puffs from the end of the smoldering cigarette. Sidestream smoke is the major source of "environmental tobacco smoke (ETS)." While SS and MS smoke have qualitatively similar chemical compositions, the respective quantities of individual smoke constituents can be quite different [35, Chapt.3; 37, p.88]. For example, in studies of nonfilter cigarettes smoked by machines, the yield'of carbon monoxide (CO) in sidestream~smoke was 2.5 to 4.7-fold that of MS smoke, while the corresponding SS/MS ratio for N-Nitrosodimethylamine (NDMA), an animal carcinogen, was 20 to 100 [35, pp.130-131]. In one compilation of toxic and tumorigenic-agents in cigarette smoke, the SS/MS ratio ranged from 0.03 to 130 [14]. Cigarette modifications that reduce the yields of "tar," nicotine and CO in mainstream smoke do not necessarily reduce the corresponding yields in sidestream smoke. In one study of U1.S. commercial cigarettes, the SS/MS ratios for carbon monoxide were 2.1 and 2.7, respectively, in two nonfilter cigarettes; 3.5 in a conventional filter cigarette; and 26.8 in a perforated filter cigarette. The SS/MS ratios for NDMA were 23.6 and 139 in the nonfilter cigarettes; 50.4 in the filter cigarette; and 167 in the perforated filter cigarette (35, p.131]. The exposure to sidestream smoke constituents, though, may be greatly reduced depending on distance from the cigarette and ventilation characteristics. Modifications of cigarette design intended to reduce ignition potential may likewise have different effects on the compositions of MS and SS smoke. In principle, ignition-reducing chemical agents added to the tobacco column or paper wrapper, such as metals and silicates, may transfer differently into MS and SS smoke. A number of devices have been developed to collect samples of SS smoke for chemical analysis [7]. However, there are no regularly published data on the composition of SS smoke of U~.S. cigarette brands. By contrast, the Federal Trade Commission regularly publishes machine-measured yields of "tar," nicotine and CO of the MS smoke of U.S. commercial cigarettes, as described later in this Report. Still, a testing plan for low-ignition potential cigarette prototypes needs to consider both MS and SS smoke. RANGE OF HUMAN HEALTH CONSEQUENCES

page 5 Cigarette smoke (whether MS or SS) is not a homogeneous ent--ty, but a complex mixture of substances. Some smoke components, such as CO, hydrogen cyanide and nitrogen oxides, are gases. Others, such as nicotine and polyaromatic hydrocarbons (PAH), are contained in the submicron-sized solid particles that are suspended in the smoke. Still others, such as formaldehyde and benzene, are volatile chemicals contained in the liquid-vapor portion of the smoke aerosol (37, p.79; 39, Chapt.14]. In view of this chemical complexity, it should be no surprise that cigarette smoke has multiple, diverse effects on human health. Nor should it be unexpected that multiple chemicals in cigarette smoke contribute to any one adverse health effect. Among the major health effects of cig,arette smoke that need to be considered in the development of a toxicity testing plan are the following: cancer;; non-cancerous lung diseases; atherosclerotic diseases of the heart and blood vessels; and toxicity to the human reproductive system. Cancer Cigarette smoking causes cancers of the lung, esophagus, larynx, oral cavity, bladder, and pancreas in male and female smokers. Smoking has reported to increase the risks of cancers of the kidney, liver, anus, male penis, and female uterine cervix, as well as leukemia [13;31;37;38]. Cigarette smoking is far and away the major cause of lung cancer in the U.S., accounting for 90 percent of cases in men and 79 percent in women (37, p.156]. Numerous epidemiological studies covering the experience of millions of men and women over many years show that smokers' risks of developing cancer increase with the number of cigarettes smoked daily, with the lifetime duration of smoking, and with early age of starting smoking. Smoking cessation gradually reduces cancer risk [37;38]. Filter-tipped and low "tar" cigarettes reduce cancer risk somewhat. Cigarette smoking interacts with other causative agents, including alcohol, asbestos, certain viruses, and certain workplace exposures, in the development of human cancers [31;34;37]i. Mainstream cigarette smoke contains over three dozen distinct chemical species considered to be tumorigenic in humans or animals (14; 31, pp.192-218; 37, p.86],. Some of these chemicals are alone capable of initiating tumors in laboratory animals; others can promote the development of previously initiated cancers. As described later in this Report, condensates collected from cigarette smoke cause mutations and damage to DNA in laboratory assays of mutagenesis (12], as well as malignant transformation in laboratory tests of a chemical's ability to induce malignant changes in mammalian cells [3;8].

page 6 Undiluted mainstream cigarette smoke is too toxic to be tolerated by laboratory animals such as rodents. In long- term experiments with diluted smoke, these animals still do not inhale the smoke in the same way as humans. In natural human smoking, the smoke is puffed in volumes of about 30 to 70 ml; the puffed smoked is temporarily retained in the smoker's mouth, after which it may be inhaled deeply into the lungs. By contrast, some laboratory animals breath by panting, while others are obligate nose breathers. Even with installation of smoke through irtificial airways, it can be quite difficult to get the animals to inhale deeply, as human smokers do. Accordingly, the distribution and retention of smoke components in the respiratory systems of laboratory animals may not mimic natural human smoking. Nevertheless, as described later in this Report, significant progress has been made in the designofinhalation devices that can expose laboratory animals, especially rodents, to diluted smoke for long periods. Long-term smoke inhalation regularly induces tumors of the larynx in Syrian golden hamsters. Direct installation of cigarette tar into the airways of laboratory animals causes lung cancers [14;31). As discussed later in this Report, the most widely used experimental system is the mouse skin bioassay, in which cancers are induced by the repeated application of condensates of cigarette smoke to the shaved skins of mice. Independent scientific agencies have concluded that environmental tobacco smoke causes lung cancer in nonsmokers [22,•35]. SS smoke, like MS smoke, contains numerous tumorigenic agents. Non-Cancerous Lung Diseases Cigarette smoking is the main cause of chronic obstructive lung disease (,COLD), also called chronic obstructive pulmonary disease (COPD) (33). Smoking accounts for 84 percent of COLD deaths in men and 79 percent in women (37, Chapt.3]i. COLD is a slowly progressive illness that develops after repeated insults to the lung over many years. In the early years after starting to smoke, an individual may report no symptoms. Even at this early stage, however, breathing tests can often detect abnormalities in the small, terminal airways of the lung [2;26;33], and these abnormalities have been directly observed in autopsy studies of young smokers who died suddenly (23]. For smokers in their twenties, there is already a dose-response relation between the extent of abnormal lung tests and the number of cigarettes smoked daily. In random population surveys, from 17 to 60 percent of adult smokers under age 55 have detectable small airways dysfunction (33, pp.27-32].

page 7 Over the course of two decades or more of smoking, a constellation of chronic respiratory changes develops. This picture of chronic lung injury includes: (i) mucus hypersecretion, with chronic cough and phlegm; (ii) airway thickening and narrowing, resulting in obstruction to airflow during expiration; and (iii) emphysema, i.e., abnormal dilation of the air spaces at the end of the respiratory tree, with destruction of the walls lining the air sacs, resulting in further airflow obstruction. These changes can cause significant respiratory impairment, disability, and death. While individual patients vary in the relative contribution of these three changes, those with clinically severe COLD typically have all three. While a minority of cigarette smokers will develop clinically severe COLD, some chronic deterioration in lungstructure or function is demonstrable in the majority of long-term smokers [33, Chapt.2]. Some smokers show more chronic cough and phlegm, others more airway obstruction. In general, breathing function declines as a person's cumulative exposure to smoke, measured in pack-years, increases [6]. Cigarette smoke produces pathological changes in the lungs of smokers by a number of different mechanisms [38, pp.282-285]. Cigarette smoke is toxic to the small hairlike cilia that line the central breathing passages. These cilia, in combination with mucus secretions, defend against deep inhalation of foreign material [33, p.279]. Smoking also induces many abnormalities in the inflammatory and immune systems within the lung [34, p.256]. In particular, cigarette smoke causes inflammatory cells to produce an enzyme called elastase. The enzyme elastase in turn breaks down elastin, an important protein that lines the elastic walls of the air sacs [9; 33, p.431]. Moreover, oxidants present in cigarette smoke can inactivate a separate protective enzyme called alpha-l-antitrypsin, which inhibits the destructive action of elastase (16; 33, p.434]. Researchers have produced various types of acute and chronic lung injury in laboratory animals exposed to cigarette smoke (33, pp.286,428,432,436]. But they have had difficulty inducing genuine emphysema from cigarette smoke alone. As in experimental models of cancer, the laboratory animals do not inhale the smoke deeply. Moreover, very long smoke exposures may be required, as is the case in humans. In one experimental study, hamsters exposed either to low doses of elastase or low doses of smoke alone did not develop emphysema, but the combination of low doses of cigarette smoke and elastase caused emphysema-like changes [15]i. A later chapter in this Report describes a laboratory test for the acute inflammatory effects of cigarette smoke on the lung, in which mice are exposed to cigarette smoke through a nose-only system.

page 8 A large number of organic and inorganic chemicals in the gaseous, volatile and particulate phases of cigarette smoke appear to contribute to its toxicity to the respiratory system [33, pp.289,415], including hydrocarbons, aldehydes, ketones, organic acids, phenols, cyanides, acrolein, and nitrogen oxides. Some components contribute to the development of chronic mucus hypersecretion in the central airways, while others play a greater role in the production of small airway abnormalities and emphysematous injury to the peripheral air sacs ['33, p.425]. As noted above, oxidizing agents in smoke inhibit the enzymes that defend against the destruction of lung elastin. Passive exposure to environmental tobacco smoke produces respiratory irritation in nonsmokers, particularly in the children of smoking parents [,33, Chapt.7; 35, p.37]. Infants and children ot smoking parents are at increased risk of acute respiratory infections, chronic cough and wheezing, and measurable declines in lung function [35, pp.38-59]. These early-life infections can have long-term adverse effects. In adults passively exposed to ETS,, some studies have reported measurable changes in lung function. Overall, the effect appears to be too small to implicate passive smoking alone as a cause of full-blown COLD ['35, p.62]. Atherosclerotic Cardiovascular Diseases Cigarette smoking is a major contributing cause to coronary heart disease, stroke, and other atherosclerotic diseases of the circulatory system [32;37). Atherosclerosis is a chronic disease that can affect the arterial blood vessels in virtually every part of the human body, including the coronary arteries that supply blood to the heart muscle; the aorta that carries the blood directly from the heart; the carotid arteries that carry blood to the brain; and the iliac and femoral arteries that carry blood to the legs. The common underlying lesion of atherosclerosis is the plaque, which occurs within the wall of the affected artery. As the plaque enlarges and matures, the artery becomes narrowed, and blood flow is reduced. If the narrowed artery carries blood to the heart, then chest pain on exertion (angina) is produced. If the affected artery carries blood to the leg, then calf pain on walking (claudication) is produced. If the affected artery carries blood to the brain, then transient neurological symptoms, such as fainting, loss of vision, movement, or speech (transient ischemic attacks) are produced. If the affected artery carries blood to a man's penis, impotence can result. A sufficiently narrowed artery is susceptible to complete blockage by a superimposed blood clot. If the blocked artery carries blood to the heart, then a heart attack (myocardial

page 9 infarction) is produced. A blockage of an artery supplying a limb canproduce gangrene. A blockage to the arteries supplying the brain can cause a stroke. The most important form of atherosclerosis in the U.S. i's coronary atherosclerosis. Its manifestations, which include angina, heart attack, heart failure, and sudden death, are described by the inclusive term coronary heart disease (CHD). Atherosclerosis involving the arteries supplying the brain is a form of cerebrovascular disease (CVD). Atherosclerosis involving the arteries to the limbs is called peripheral vascular disease (PVD). Atherosclerotic plaques take years to develop. The earliest lesion is called a fatty streak, which consists of deposits of cholesterol-within the arterial wall. These fatty streaks can be observed in young people with no symptoms, and even in children. There is a progressive inflammatory reaction to the fatty deposits, and a collection of fibrous debris, muscle cells, and more fatty deposits is incorporated into the developing plaque. Cholesterol is a fatty substance that does not dissolve readily in water. It circulates in the blood mostly by attaching to specialized proteins. These cholesterol-protein complexes, which also contain other fatty substances, form particles of various sizes, which are called lipoproteins. The lipoprotein particles are classified by their density. There are very-low density lipoprotein (VLDL), intermediate density lipoprotein (IDL), low-density lipoprotein (LDL) and high-density lipoprotein (HDL) particles. The fundamental event in the initiation of a fatty cholesterol deposit appears to be the transfer of LDL particles from the blood across the inner lining (endothelium) of the arterial wall. This transfer may require prior injury to the inner lining of the artery, in order to expose the raw surface to LDL transfer. When a person's blood cholesterol is measured, the amount that is specifically attached to LDL is called the LDL-cholesterol, or popularly the "bad cholesterol." On the other hand, HDL particles work in the opposite direction, removing cholesterol from LDL and transporting it back to the liver. Because of this reverse-transport function of HDL, the amount of cholesterol attached to HDL is popularly termed the "good cholesterol." In epidemiological studies of humans, certain measurable personal characteristics have been consistently found to be predictors of the risks of atherosclerotic disease. These predictors are sometimes called risk factors. For example, male gender is a risk factor for coronary heart disease. This does not mean that maleness per se causes CHD. Still, the fact that