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.
5 The Scientific Basis for PREP Assessment A ssessing health risks from conventional tobacco products is simi- lar to that for many environmental and occupational exposures. Tobacco risks, however, are among the more complicated to as- sess for several reasons. The general components of risk assessment (haz- ard identification, dose-response assessment, exposure assessment, and risk characterization) described in Chapter 1 are still useful to consider (see Table 5-1). Hazard identification is challenging because tobacco and the smoke generated upon its combustion are complex mixtures. Some of the hun- dreds (or thousands) of known or suspected toxicants are fairly well un- derstood; however, the relative contribution to overall toxicity of most of the individual compounds is not. In addition, tobacco products contain added constituents or ingredients, but the identity and concentration of these compounds within a specific tobacco product is unknown, due to proprietary concerns. Animal models of tobacco toxicity are limited, pos- ing additional barriers to complete hazard identification. Dose-response assessment is complicated. Because the exposure is a complex mixture, the diseases associated with tobacco exposure are many and the dose-response relationships vary significantly. Assessing the dose in epidemiological studies is complicated in part by the factors described for hazard identification. In addition, the dose a tobacco user is exposed to can change often over a long and variable smoking history. Finally, the responses are most often diseases with long periods of disease progres- sion until diagnosis and from time point of dose estimation. 140
TABLE 5-1 Tobacco and PREP Risk Assessment Hazard Identification Dose Response Exposure Assessment Risk Characterization Risk Management 1983 âRed Epidemiology Epidemiology Dose to which Estimate of the A risk-assessment Bookâ Animal bioassay Low-dose humans are exposed magnitude of the (qualitative or Short-term Studies extrapolation Dose of special public health quantitative) may Comparisons of Animal to human populations problem be one of the bases molecular structure extrapolation Estimation of size of of risk management population poten- tially exposed Challenges in Complex mixture Dose changes for an Changes in smoking For which disease? FTC regarding risk Animal models individual over topography At which point in advertising assessment are limited time Complex mixture smoking history? of tobacco Constituents and Dose of individual additives are toxicants varies proprietary over time information Exposure at time of disease progression Additional Products will change Assessing effect of Changing exposure Need models to FDA authority challenges rapidly with time moving backwards after long-term consider effects on currently exerted of PREP on a dose-exposure higher dose initiation, cessation, only over risk curve, assuming exposure and relapse pharmaceutical assessment long-time previous Some toxicants could PREPs higher exposure increase continues 141
TABLE 5-1 Continued 142 Hazard Identification Dose Response Exposure Assessment Risk Characterization Risk Management Committee 1. Does product 2. Is decreased 1. Does product 4. What are the public 4. What are the public charge decrease exposure exposure associated decrease exposure? health implications? health implications? to the harmful with decreased substances in or harm to health? produced during 3. Are there useful use of tobacco? surrogate indicators of disease that could be used? Disease- 3. Utility of preclinical 1. Dose-response data 2. Validation and 5. Long-term specific research to judge for conventional development of epidemiological summary feasibility tobacco products biomarkers studies and data 2. Validation and 4. Short-term clinical surveillance (Chapter 5; development of and epidemiological Section II) biomarkers studies 4. Short-term clinical and epidemiological studies NOTE: Numbers correspond to research recommendations listed in the Executive Summary.
THE SCIENTIFIC BASIS FOR PREP ASSESSMENT 143 Exposure assessment is difficult for some of the same reasons. There is a multiplicity of tobacco products on the market. The specific exposures associated with any one âbrandedâ product could change throughout time because the product can change. Changes in exposure throughout time are not documented. In addition, smokers of âlow-yieldâ products often compensate (change smoking behavior to increase nicotine exposure), so their exposures to nicotine and tobacco/smoke toxicants are often higher than predicted by a common form of exposure assessment, self-report. The objective of a potential reduced-exposure product (PREP) risk assessment is to determine if the risk of harm from the use of the PREP is less than the risk of harm in the absence of the PREP (see Table 5-1). The risk management objective considered by the committee is not to ban or control the exposure per se, as is the case for environmental and occupa- tional exposures. The risk management objective, as will be made clear in Chapter 7, is primarily to verify whether or not a product is associated with either exposure reduction or harm reduction. A PREP risk assessment involves lowering the dose of a complex mixture in a person (or population) with varying degrees of pre-existing pathology or cellular damage caused by a complex mixture exposure (that of conventional products) and trying to reverse early damage or to stop disease progression. This is problematic at this point, as there are no adequate human or animal studies that replicate this scenario. While some studies report risks in persons who switch from nonfiltered cigarettes to filtered cigarettes, or from high- to low-tar cigarettes, this âswitchingâ did not reduce exposure (due to compensation) significantly in many people. The reduction in risk, if any, would occur only in persons who do not compensate for lower nicotine levels by smoking more or smoking differ- ently. The basic elements of risk assessment can, however, be still consid- ered. The questions become slightly different, and the data required or the study designs might be different from that required for a tobacco risk assessment. For hazard identification, the questions include: â¢ Does the PREP contain (or produce during use) toxicants known to cause adverse health effects? â¢ To what extent are the compounds targeted for reduced exposure causally linked to a tobacco-related disease? â¢ How does its content compare with the toxicants in the conven- tional tobacco product to which it is compared? â¢ Are there unique toxicants in a PREP compared to conventional tobacco products? For hazard identification, it is essential to know the composition of the material to which people will be exposed from the PREP compared to
144 CLEARING THE SMOKE the standard product. Any new material, such as flavors, added to stan- dard products must be included in the analyses. It is important to analyze the product that actually enters the body (for example, the combustion products that are inhaled) rather than the composition of the product as sold. The approach to testing the toxicity of the material to which people are exposed in the tobacco-related PREP compared to standard tobacco products is discussed in Chapter 10. The objectives of the toxicity tests are to determine what toxic effects can be induced by the test materials (the tobacco-related PREP compared to the standard product) and how much of the test materials is required to cause the adverse effect, i.e., the dose- response characteristics in animals of the test materials. Data from animal studies can be used to eliminate new products that are much more toxic than existing ones. A series of comparative potency tests is appropriate. In vitro studies in cultured cells from both animals and humans can be used to determine the ability of the test materials (from the tobacco-related PREP and the standard product) to induce cellular damage, an inflammatory response, or cell death. Assays of the mutagenic or clastogenic activity of the test materials can be done in bacterial or mammalian cell systems. In animal studies, tests for tobacco-related toxicity should include evaluation of the ability to induce adverse health effects or cancer in the respiratory tract, the nervous system, the cardiovascular system, the re- productive and developmental systems, and other organs. Toxicokinetic studies should be used to determine dosimetry to different organs and to suggest biomarkers of internal dose that can be used in humans. Short- term clinical tests in humans should be done to compare the potencies of the test materials to induce acute adverse health effects (such as reduced pulmonary function) and to determine the toxicokinetics of the tobacco- related PREP compared to the standard product. For dose-response assessment, the questions include: â¢ What are the dose-response characteristics of the PREP compared to the conventional tobacco product? â¢ Do smokers use PREPs at a time in their individual smoking his- tory (and therefore of disease progression) that induce different dose-response effects? â¢ Are the patterns of adverse health effects different from PREPs compared to conventional tobacco products? â¢ What is the evidence that reduction in exposure to the targeted compounds in the complex mixture or other hazardous material in the PREP will decrease or reverse the development of disease? â¢ What is the dose-response relationship between the targeted com- pounds and the disease outcome?
THE SCIENTIFIC BASIS FOR PREP ASSESSMENT 145 â¢ What quantitatively happens to disease induction if exposures are reduced? â¢ How much exposure needs to be reduced to result in a measurable benefit? â¢ Are there individual susceptibilities (age, gender, genetic makeup, and prior use of tobacco products) that change this dose-response relationship? Some dose-response information can be obtained from standard pre- clinical animal studies. However, there will be uncertainties in extrapolat- ing from animal data to humans. Additionally, it will be difficult, even in animals, to determine the response to a reduction in dose after a period of higher exposure. For exposure assessment, the questions include: â¢ Do PREP users compensate differently than users of more conven- tional âlow-yieldâ products? â¢ Do PREP users exclusively use PREPs or do they switch back and forth between PREPs and conventional products? â¢ How does the PREP change the exposure pattern for the popula- tion? â¢ Does the introduction of the PREPs increase the number of people initiating use of tobacco products or decrease cessation attempts? â¢ What is the overall balance in exposures (directly and as environ- mental tobacco smoke) with or without the PREP? Past experiences with âlow-yieldâ cigarettes containing less tar and nicotine than other products underscore the need to determine internal dosimetry of toxic material entering the body during use of tobacco- related PREPs. The internal dosimetry of the new products can be com- pared to that of standard products in short term toxicokinetic studies in humans. Biological markers of internal dosimetry of key ingredients can be used when available. The risk assessment process will need to rely on the use of animal preclinical and human clinical studies, in which biomarkers of exposure and potential harm can be measured. Biomarkers of exposure to tobacco products have been validated and are in current use. Unfortunately, few specific early indicators of biomarkers have been validated as predictive of later disease development. It is recognized that today, biomarkers of exposure are better validated compared to biomarkers of potential harm, and that it is more feasible to consider exposure reduction in contrast to risk reduction. However, while an assessment of risk reduction through biomarkers will have more uncertainty, these will need to be included in the PREP risk assessment process in order to enhance confidence that there is no worsening risk, in the least.
146 CLEARING THE SMOKE For risk characterization, the questions include: â¢ Can the exposure levels of the PREP, given its dose-response char- acteristics, be expected to result in reduced risk of one or more tobacco-related diseases than the standard tobacco product to which it is compared? â¢ What is the magnitude of any reduction in risk? â¢ What are the limits in understanding of the risk reduction? â¢ Do fewer tobacco users quit and use PREPs instead? â¢ Do former tobacco users relapse to PREPs? â¢ Do nonusers initiate tobacco use through PREPs? In order to achieve a level of confidence that a PREP will provide meaningful reductions in risk, especially compared to the real possibility of others using this product to initiate or resume smoking, prospective epidemiological studies are required. This could be done in a timely man- ner for some disease endpoints, such as birth outcomes or recurrence of myocardial infarction. For many other diseases associated with tobacco use, however, definitive demonstration of reduced harm will require stud- ies of a long duration. Moreover, it is reasonable to anticipate that the design of PREPs would change rapidly in the coming years so that assess- ing PREP use will be difficult. The claim that a PREP will result in a reduction of the risk for harm requires scientific evidence for the validity of that claim. A discussion of the information relevant to hazard identification and dose-response infor- mation (the first two elements of risk assessment) is given in Chapter 10. Information on the best means of evaluating exposures is given in Chap- ter 11. In the case of tobacco products, animal models of adverse health effects have been problematic in the past, but new animal models show promise for being useful (see Chapter 10). There currently are no popula- tion risk assessment models that mimic the types of predictions hoped for tobacco users who switch to PREPS and how the public health will be effected by the initiation of PREP use by never and former smokers. The committee has evaluated the science base regarding the toxic effects of tobacco on the major diseases known to be caused by tobacco exposure (i.e., cancer and diseases of the cardiovascular, pulmonary, re- productive systems). The committee has done so to arrive at summary conclusions regarding the evidence base that would directly feed into a risk assessment paradigm, such as that described above (see Table 5-1). Specifically, the committee has elaborated in each of the major disease- oriented chapters of Section II and summarized later in this chapter the evidence regarding:
THE SCIENTIFIC BASIS FOR PREP ASSESSMENT 147 1) the dose-response relationship between tobacco smoke and/or con- stituent exposure and health outcomes, 2) identification and development of surrogate markers for disease, 3) the utility of preclinical research in understanding the potential of PREPs to be harm reducing for the disease under review, 4) utility of short-term clinical and epidemiological studies, and 5) the role of long-term epidemiological studies and surveillance. The review of preclinical research and the material in Chapter 10 (Toxicology) provide information on hazard identification and, in part, dose-response assessment. The material on surrogate markers for disease is informative for both dose-response assessment and exposure assess- ment. The review of the short-term clinical studies, epidemiology, and surveillance data provide the proof of reduced harm. In conclusion, the PREP risk assessment process will be challenging because no single definitive study, either human or experimental animal study, would stand up to rigorous scientific scrutiny to be used in such a process. Therefore, today, several types of data will be needed that in- cludes both experimental animal studies and human clinical data, with a definitive plan to conduct epidemiological and surveillance studies. The clinical data is needed because animal studies cannot predict interindividual differences in human behavior that would affect how the PREP is used or cause damage, and because there are too many uncertainties about the use of animal data. During this interim time, and with more research, the regulatory process might be able to identify key types of data needed for the risk assessment process. At this point in time, however, it can only be concluded that both experimental animal and clinical human studies would be needed, and that this would include both the consideration of exposure to individual PREP constituents and as a complex mixture. It is conceivable that the PREP risk assessment process can be simplified, for example by comparing only experimental data for one PREP to another, but this will require substantial experience and characterizing the data for existing PREPs. Sufficient data for streamlining this process in not avail- able today. The remainder of this chapter provides a summary of the major con- clusions and recommendations, arranged by chapter, reached by the com- mittee in Section II of this report. TOBACCO SMOKE AND TOXICOLOGY (SEE SECTION II, CHAPTER 10) Mainstream tobacco smoke and environmental tobacco smoke each is a complex mixture of toxicants composed of carcinogens and other chemi-
148 CLEARING THE SMOKE cals with health effects that alone or in combination are only partially known (see Table 5-2) (Davis and Nielson, 1999). The evaluation of con- ventional tobacco products and tobacco-related PREPs is complicated by a lack of adequate in vivo models of tobacco-related morbidities in man. Toxicology studies, both in vitro and in vivo, provide the opportunity to evaluate the potential harm reduction offered by potential reduced- exposure products. The comparative potency of the PREP can be deter- mined in a series of preclinical studies that include both the PREP and the standard tobacco product that can be replaced by the PREP. The pre- clinical tests should include in vitro tests in both animal and human cells TABLE 5-2 List of Selected Tobacco Mutagens and Carcinogensa Constituent IARC Class Phase Evaluationb Examples N-Nitrosamines Particulate Sufficient in Tobacco-specific nitrosamines (NNK, animals NNN), dimethylnitrosamine, diethylnitrosamine Polycyclic Particulate Probable in Benzo[a]pyrene, benzo[a]anthracene, aromatic humans benzo[b]fluoranthene, hydrocarbons 5-methylchrysene Aryl aromatic amines Particulate Sufficient in 4-Aminobiphenyl, 2-toluidine, humans 2-naphthylamine Heterocyclic Particulate Probable in 2-Amino-3-methylimidazo[4,5-b]- amines humans quinolone (IQ) Organic Vapor Sufficient in Benzene, methanol, toluene, styrene solvents humans Aldehydes Vapor Limited in Acetaldehyde, formaldehyde humans Volatile organic Vapor Probable 1,3-Butadiene, isoprene compounds Inorganic Particulate Sufficient in Arsenic, nickel, chromium, compounds humans polonium-210 NOTE: NNK=nicotine-derived nitroketone; NNN=N-nitrosonornicotine. aThis list is intended to provide a conceptual overview of the complexity of tobacco product exposures. It is not all-inclusive but is presented to allow the reader to understand the number of considerations that must be made in assessing harm reduction strategies. bInternational Agency for Research on Cancer: The classifications here refer to evalua- tions of the compound from any exposure, not just tobacco. Not all chemicals within the class are considered carcinogenic in humans. There is no consideration in this table to delivered dose or route of exposure (IARC, 1986).
THE SCIENTIFIC BASIS FOR PREP ASSESSMENT 149 to determine the cytotoxicity and the genotoxicity of the tobacco product to which humans will be exposed. Such tests have recently been reported for a new tobacco-related PREP (Eclipse Expert Panel, 2000). Such a test must include dose-response studies to determine the amount of the expo- sure material required to cause toxicity. Next, studies should be con- ducted in vivo in the best animal models available to determine the com- parative potency of the PREP versus the standard product in producing: (1) pulmonary inflammation, (2) COPD, (3) cardiovascular disease, (4) reproductive toxicology, and (5) pulmonary neoplasms. In vitro studies and in vivo animal studies are useful but limited tools in evaluating the toxicity of products that claim to reduce exposure to tobacco toxicants and potentially reduce tobacco-related harm. In vitro studies may allow rapid, low-cost screening for the toxic properties of conventional tobacco products and tobacco-related PREPs, although the relationship between in vitro toxicity and in vivo human response has not been established for most compounds. These assays include cytotoxicity and genotoxicity assays, which are possible screens for the carcinogenic or inflammatory potential of products. In vivo toxicity testing can be developed to supplement in vitro and clinical studies. Such animal models, if developed, may be useful as a screening assess- ment of the efficacy of PREPs for reduction of various tobacco-attributable diseases (see Chapters 12-16). The committee concludes that animal models should be used to test for the potential adverse health effects of tobacco smoke or any proposed additives. The A/J mouse model, which is sensi- tive to induction of lung adenomas, shows promise as an animal model for screening the potential of tobacco products to induce lung tumors (Witschi et al., 2000; Witschi et al., 1999; Witschi et al., 1997a,b). Future studies should validate the model. These studies (Witschi et al., 2000) indicated that removal of single classes of carcinogens, such as nitrosamine or polycyclic aromatic hydrocarbons (PAHs) may not be protective against induction of lung tumors by smoke. Studies also indicate that some animal models show promise for use in studying the development of symptoms similar to those of chronic obstructive pulmonary disease (COPD), the development of cardiovascular disease, adverse effects on the immune system, intrauterine growth retardation, and poor fetal lung maturation from the inhalation of new or existing tobacco products (see Chapter 10). Testing the general toxicity of smokeless tobacco and evaluating of the potential harm reduction properties of smokeless tobacco (e.g., Swed- ish snus) use in smokers may greatly benefit from assays for genotoxic and cytotoxic potential and the animal models discussed above. Details to be considered in determining the specific set of toxicity tests include species and strain of test animal, duration of test, end points of interest, dose-response considerations, biomarkers of dosimetry and
150 CLEARING THE SMOKE response, and standard comparison products to be tested as positive and negative controls. EXPOSURE AND BIOMARKER ASSESSMENT IN HUMANS (SEE SECTION II, CHAPTER 11) Accurate measures of exposure and the development of biomarkers of adequate specificity and sensitivity are needed to evaluate the toxicity and harm reduction potential of PREPs. Biomarkers can be defined as measurements of tobacco constituents, tobacco smoke constituents, or changes in body fluids (including exhaled air) and organs. The assess- ment of a PREP will have to include markers of external exposure and biomarkers indicative of internal exposure, biologically effective dose (Perera, 1987), and potential harm. The definitions of each are provided in Table 5-3. There have been different definitions of types of exposure as- sessments used previously, but more recent understandings of biomarker uses and limitations, as well as different approaches needed for PREP evaluation lead to a need for clarification and redefinition. The latter three measurements in Table 5-3 improve upon the first by quantifying exposure at the cellular level to characterize low-dose expo- sures or low-risk populations, provide a relative contribution of individual TABLE 5-3 Exposure and Biomarker Assessment Definitions Exposure or Biomarker Assessmenta Definition External exposure A tobacco constituent or product that may reach or is at the marker portal of entry to the body Biomarker of exposure A tobacco constituent or metabolite that is measured in a biological fluid or tissue that has the potential to interact with a biological macromolecule; sometimes considered a measure of internal dose Biologically effective The amount that a tobacco constituent or metabolite binds dose (BED) to or alters a macromolecule; estimates of the BED might be performed in surrogate tissues Biomarker of potential A measurement of an effect due to exposure; these include harm early biological effects, alterations in morphology, structure, or function, and clinical symptoms consistent with harm; also includes âpreclinical changesâ aCategories and definitions reflect concept that the critical exposure is at the level of a biological macromolecule, so that exposure for this discussion is not limited to a measure- ment at the portal of entry to the body.
THE SCIENTIFIC BASIS FOR PREP ASSESSMENT 151 chemical carcinogens from complex mixtures (e.g., tobacco-specific N- nitrosamines in cigarette smoke), and estimate total burden of a particular exposure where there are many sources (e.g., benzo[a]pyrene [BaP] from air, tobacco, diet, and occupation) (Vineis and Porta, 1996). In assessing PREPs through biomarkers, understanding the biological effects of a wide range of exposures will be important. Within the context of this discus- sion, exposure at the level of the cell and critical macromolecules is con- sidered with greater weight, rather than the traditional view of exposure at the portal of entry into a person. Markers of external exposure attempt to measure the dose of tobacco or tobacco smoke constituents that may enter the body and usually in- volve machine testing of products and user questionnaires. Internal expo- sure markers assess the amount of tobacco or tobacco smoke constituents or their metabolites in body fluids or organs. Biomarkers estimating the biologically effective dose measure the internal dose that interacts with cells and macromolecules and may be mechanistically related to disease outcome. Finally, biomarkers of potential harm reflect changes in cells and macromolecules that may lead to disease (see Table 5-4). Measuring the number of cigarettes per day and smoking duration, estimating lifetime exposure, smoking topography, and so forth can pro- vide an effective indicator of exposure that has been associated with harm. However, these measures may be insensitive to changes in risk, are diffi- cult to assess accurately over time, and have not been tested in the context of harm reduction. Also, because there is interindividual variation in the way the body responds to these exposures, such measures might not be sufficiently accurate for new products intended to decrease exposure. Thus, for new products, the relationship of external exposure markers to disease risk might be less predictable. Currently, there is sufficient evi- dence to show that biomarkers can provide better estimates of risk in the context of exposure, and therefore they will likely be able to provide improved assessments for harm reduction products. However, no single biomarker has been sufficiently validated and related to disease risk that it can be recommended as an intermediate biomarker of cancer risk. Thus, different types of biomarkers along the pathway from internal exposure, biologically effective dose, and potential harm are needed, and additional research is necessary to identify the best combination of markers to be used. Experimental toxicity testing (in vitro and animal models) are not sufficient to support a PREP claim because only validated biomarkers can show that the PREP reduces exposure adequately enough to imply risk reduction. The use of intermediate biomarkers as surrogate risk factors for disease may possibly overestimate the number of people who actually develop disease, because not all early changes in morphology or function progress to disease. On the other hand, it may underestimate if, as
152 CLEARING THE SMOKE TABLE 5-4 Biomarkers of Potential Harmful Effectsa,b Associated Target Dose- with Tissue Specific Variables Used Response Cessation Assay Chemical to Category in Literature Data or Half-life Available Specificity Tobacco Enzymatic Aryl hydrocarbon No >30 d Yes Yes No induction hydroxylase CYP1A2 No NDA Yes Yes No DNA repair NDA Yes Yes NA No enzymes Microarray assays NDA NDA Yes NA No for mRNA expression and proteomics Chromosomal Chromosomal Yes Yes Yes No No alterations aberrations Micronuclei Yes Yes Yes No No Sister chromatid Yes Yes No No No exchanges Loss of Yes Yes Yes No No heterozygosity
THE SCIENTIFIC BASIS FOR PREP ASSESSMENT 153 Related Specific to a Chemical to Disease Specificity Tobacco Riskc Strengths Limitations Yes No Yes Indicates acquired changes in Technically difficult to assess in susceptibility; related to large epidemiological studies DNA-adduct levels Yes No Yes Indicates acquired changes in Technically difficult to assess in susceptibility; related to large epidemiological studies DNA-adduct levels NA No NDA Indicates acquired changes in Technically difficult susceptibility; provides analysis of what is likely to be critical part of carcinogenesis NA No NDA Reflects integrated measure of Difficult to perform; multiple genotypes, provides relationship to disease risk is complex data potentially technically difficult to prove; usable for rapid identification requires extensive laboratory of important risk factors validation; RNA and protein microarray assays are expensive; large-scale studies are needed; refined bioinformatic analysis required No No Yes Can be done in blood as Very nonspecific; relationship to surrogate tissue. Similar target organ is not lesions observed in cancer. established; significant lack of Can be measured in persons specificity and wide overlap without cancer between smokers and nonsmokers No No NDA Facile assay Lack of specificity No No No Easy to do in blood as surrogate Very nonspecific; relationship tissue. Can be measured in to target organ is not persons without cancer established; predictivity for disease risk not established. Association with cancer in case-control studies may have case bias. Significant lack of specificity and wide overlap between smokers and nonsmokers No No NDA Similar lesions observed in Technically complex; cancer relationship to cancer risk unknown continues
154 CLEARING THE SMOKE TABLE 5-4 Continued Associated Target Dose- with Tissue Specific Variables Used Response Cessation Assay Chemical to Category in Literature Data or Half-life Available Specificity Tobacco Mutations in Yes Yes No No No reporter genes (HPRT, GPA) Mutational load NA NDA Yes No No in target genes (p53, K-ras) Mitochondrial Deletions, NDA NDA Yes No No mutations insertions Epigenetic Whole genome NDA NDA Yes No No cancer methylation effects Hypermethylation NDA NDA Yes No No of promoter regions Lipids Blood lipids: Yes NDA Yes Yes No HDL, LDL, oxidized LDL, triglycerides Cardiovascular Heart rate, blood No Yes Yes NA No response pressure Thrombosis Bleeding time No NDA Yes No No Fibrinogen NDA NDA Yes Yes No Prothrombin time, Yes NDA Yes Yes No partial thromboplastin time, plasminogen activator inhibitor, C- reactive protein
THE SCIENTIFIC BASIS FOR PREP ASSESSMENT 155 Related Specific to a Chemical to Disease Specificity Tobacco Riskc Strengths Limitations No No NDA Facile assay in blood Relationship to target tissue or blood unknown No No NDA Target gene specificity Very difficult to do in normal tissues No No NDA Provides corroborative marker Relationship to disease not established No No No Facile assay Relationship to disease unknown No No No Similar lesions observed in Technically difficult; cancers relationship to risk unknown Yes No Yes May be directly related to Levels among heavy smokers disease risk cannot be distinguished. Wide interindividual variation. Many individuals under medication therapy. Significant confounders exist NA No Yes Easy to measure; intraindividual Both interindividual and differences may be important intraindividual differences are for the individual significant. Substantial confounders exist, and many persons are on medications No No No Minimally invasive Very nonspecific Yes No NDA Pathogenically related to Does not distinguish levels of disease smoking. Nicotine might separately affect these parameters so limited use in persons using NRT Yes No NDA Leave a fingerprint at the site of their formation continues
156 CLEARING THE SMOKE TABLE 5-4 Continued Associated Target Dose- with Tissue Specific Variables Used Response Cessation Assay Chemical to Category in Literature Data or Half-life Available Specificity Tobacco Urinary Yes No No Yes No thromboxane and prostacyclins Platelet activation Yes NDA Yes No No and survival Blood cell White blood cell Yes Yes Yes Yes No parameters counts (i.e., lymphocytes, neutrophils, total counts) Hematocrit, Yes Yes Yes No No hemoglobin, red blood cell mass Bronchio- Inflammatory Yes Yes Yes No No alveolar cells, protein, lavage cytokines response Neutrophil Yes Yes Yes No No elastase a1- antiprotease complex Î±1-antitrypsin No No Yes Yes Yes Inflammatory Leukotrienes Yes NDA No Yes No mediators of response Pulmonary FEV1, FVC Yes Yes Yes No No function tests Periodontal Periodontal Yes Yes Yes No No disease height Gum bleeding Yes Yes Yes No No
THE SCIENTIFIC BASIS FOR PREP ASSESSMENT 157 Related Specific to a Chemical to Disease Specificity Tobacco Riskc Strengths Limitations Yes No Yes May be markers of platelet- Technically difficult. Wide vascular interactions; reflect overlap of values due to chronic exposure individual differences in response No No No Platelet activation in vivo might Technically difficult to use for be pathophysiologically large numbers of subjects. related to cardiac artery Significant number of thrombosis confounding variables. Smoking increases platelet counts Yes No Yes Can be a surrogate marker for Relationship to disease several processes including uncertain, although atherosclerosis and thrombosis alterations in levels are linked epidemiologically to disease. Wide interindividual and intraindividual variation and large number of confounders No No No Can reflect both cardiac and Insensitive; wide interindividual respiratory disease risk differences No No NDA Provides different types of data Bronchoscopy is too invasive for with single procedure large epidemiological studies No No NDA Provides different types of data Bronchoscopy is too invasive for with single procedure large epidemiological studies Yes Yes NDA May be specific to tobacco Requires invasive test; short smoke half-life Yes No NDA May be measured in urine, Substantial number of bronchioalveolar lavage, and confounders serum No No Yes Widely available Low sensitivity for mild disease. Decrease in function with aging. Large interindividual variation No No Yes No No Yes continues
158 CLEARING THE SMOKE TABLE 5-4 Continued Associated Target Dose- with Tissue Specific Variables Used Response Cessation Assay Chemical to Category in Literature Data or Half-life Available Specificity Tobacco Osteoporosis Fractures Yes NDA NA No No Bone density NDA NDA Yes No No Skin Premature Yes NDA NA No No wrinkling Fetal and Birth weight Yes Yes Yes No No neonatal effects Weight Weight loss and Yes Yes Yes No No gain NOTE: NA=not applicable; NDA=no data available; FVC=Forced vital capacity; FEV1=Forced expiratory volume in 1 sec; HDL=high-density lipoprotein; LDL=low-density lipoprotein. expected, other mechanisms are involved in the disease process that are not reflected by the biomarkers. Biomarkers may also underestimate the incidence of disease since none are necessarily present in all who develop disease. Therefore, the implication of potential benefit from a harm reduc- tion strategy could be an overestimate or an underestimate, but this limi- tation in the scientific methodology for identifying sufficiently specific biomarkers of risk requires acceptance at the current time. Previously, the most common way to infer exposure reduction (e.g., through use of low-tar cigarettes) has been via methods that simulate human smoking behavior, such as the Federal Trade Commission (FTC) method. Although they provide a standardized way to assess cigarettes, it is clear that these methods have limited usefulness because people smoke cigarettes differently from the machine, with resultant qualitative and quantitative differences in exposure.
THE SCIENTIFIC BASIS FOR PREP ASSESSMENT 159 Related Specific to a Chemical to Disease Specificity Tobacco Riskc Strengths Limitations No No Yes Easily measured Numerous confounders No No Yes No No NA Lack of specificity; involves subjective evaluation No No Yes Data collection is easy Nonspecific; numerous confounders No No Yes Both a biomarker for Some people perceive weight metabolism and an important loss as a benefit of smoking, outcome for some people. despite significant adverse effects associated with smoking aSelected examples; list is not all-inclusive. bReferences are not provided in this table but can be found in the text of this and disease- related chapters. cAny report related to a disease outcome associated with tobacco where the report is plausible but has not necessarily been replicated. Biomarkers may be shown to reveal differences in individual suscep- tibilities and differences in response depending on dose. Thus, biomarkers that measure both complex exposures and single tobacco product con- stituents are needed and should be assessed for the range of possible human exposures, and those that assess complex exposures should carry a greater weight. Also, some biomarkers or sets of biomarkers should be developed that reflect exposure to many tobacco constituents in order to monitor for the introduction of new hazards from PREPs. Today, there remain technical limitations to the use of biomarkers. Depending on the harmful effect, surrogate assays that represent effects in target organs may be easier to perform in humans because the target tissues might not be easily accessible. However, if such is the case, the relevance of the surrogate biomarker to the target organ effect should be demonstrated.
160 CLEARING THE SMOKE The use of a biomarker for harm reduction assessments should in- clude several qualities including reflection of disease pathogenesis, speci- ficity, and sensitivity. Also, consideration must be given to available harm and harm regression dose-response data, target tissue effects, and valida- tion methods. Each biomarker should be validated for its relationship to exposure and harm, and also as a laboratory assay that provides reliable and reproducible data. Separately, the way interindividual variation in response and smoking behavior affects biomarkers should also be consid- ered. The assessment of harm and harm reduction should be made through direct human experience, as these products are used by the general popu- lation. Most of what is known about harmful tobacco products has re- sulted from epidemiology and supported by in vitro studies, laboratory animal studies, and human experiments. However, although epidemio- logical studies can provide the most definitive data about tobacco harm and harm reduction products, the study of diseases with long latency (e.g., cancer, heart disease, COPD) is problematic because such studies often require many years before they provide useful data. Thus, because definitive evidence that a new PREP actually reduces harm will often be unavailable, short-term markers that reflect long-term outcomes are needed. If assessment of the harm reduction potential of a PREP were based only on epidemiological data measuring disease outcome prior to its use by the public, very few if any harm reduction products would be introduced. Importantly, the use of intermediate markers does not re- place long-term follow-up and epidemiological surveillance, but it can be a basis for estimating effects before direct evidence from epidemiological studies is available. Biomarkers of internal exposure, biologically effective dose, or poten- tial harm have been validated to different degrees. It is typically easier to show a relationship between external exposure and biomarkers in the following order: internal exposure, biologically effective dose, and harm. Conversely, it is typically easier to show a relationship between disease outcome and biomarkers in the following order: harm, biologically effec- tive dose, and internal exposure. It might be acceptable to rely on external exposure measurements for considering risk and dose-response, but only with substantial corroborative biomarker data. The best strategy for as- sessing the claims of risk reduction methods is to have several markers that range from exposure to outcome, one being linked to another, and at least one with which a dose-response risk assessment can be made. The recommendation that PREPs be assessed by the use of biomarkers should reflect the available data, which show that individual use and response to tobacco products are affected by cultural and heritable traits.
THE SCIENTIFIC BASIS FOR PREP ASSESSMENT 161 To achieve the greatest confidence that a PREP will reduce risks for per- sons who cannot stop smoking, well-validated methods for predicting risk, including external exposure indicators, and the best available bio- marker assays should be used. NICOTINE PHARMACOLOGY (SEE SECTION II, CHAPTER 9) Nicotine is the addictive component of tobacco products, and the strength of this addiction affects the individualâs ability to stop smoking (U.S. DHHS, 1988). Nicotine is also a component of most PREPs, and therefore, evaluation of the harm reduction potential of PREPs requires evaluation of nicotineâs relative toxicity, especially during long-term use (see Chapter 4). Structurally, nicotine is very similar to acetylcholine (Ach) and inter- acts with specific nicotinic receptors (nAchRs) in the central and periph- eral nervous systems. The interaction between nicotine and its receptor affects the release of numerous neurotransmitters and results in upregula- tion of the nicotinic receptors leading to the physiological, cognitive, and sensory effects associated with tobacco use, addiction, and withdrawal. Nicotine also has well-documented effects on metabolism and on the car- diovascular, gastrointestinal, and hormonal systems (see Chapter 9). Pharmacological nicotine replacement therapy (NRT) has proven to be a remarkably well-tolerated and effective strategy for many, leading to cessation of cigarette smoking at least in the short- to medium-term (Benowitz et al., 1998; Fiore et al., 2000). Although the experience is much more limited, it is also a potential strategy for reducing the number of cigarettes smoked by smokers who cannot or will not quit (FagerstrÃ¶m et al., 1997; Rennard et al., 1990; Shiffman et al., 1998; Transdermal Nicotine Study Group, 1991). There are important considerations in evaluating nicotine products for possible tobacco harm reduction. First, nicotine is addictive, and al- though the daily exposure may be reduced by NRT use, continued usage implies psychological dependence, if not physical addiction. It is arguable whether this should be a concern, given the assumption of an undisputed reduction of risk compared to smoking. However, it would seem reason- able to include surveillance of the dependency potential and abuse liabil- ity of each NRT product. Furthermore, the effects of long-term nicotine intake on such factors as drug and alcohol consumption, the progression of coincidental diseases, the impact of aging on cognitive and other physi- ological functions, and susceptibility to other forms of addictive behavior are largely unknown. For example, observations suggesting that nicotine impairs endothelial function, a property it shares with cigarette smoking,
162 CLEARING THE SMOKE raise questions about its effect on atherogenesis during long-term use (Chalon et al., 2000; Gairola and Daugherty, 1999; Sabha et al., 2000). Such an effect may take many years to emerge and highlights the importance of continued postmarketing surveillance of NRT. Although existing data do not suggest that nicotine is carcinogenic in humans, it would be prudent to have continued surveillance of the incidence of cancer among users of NRT. Studies of long-term nicotine administration on surrogate variables that more closely resemble the mechanism under consideration (e.g., im- aging of plaque progression) and attendant studies in animal models seem timely. Increasingly, the application of genomic and proteomic ap- proaches will help clarify the differential effects of smoking and NRT on the expression and translation of genes related to the development of smoking-related diseases. Finally, the understanding of nicotineâs effect on inflammation and the immune response is confused and limited (Sopori et al., 1998). More research is needed to clarify its effects on cytokine generation, the formation of nitric oxide (NO) and eicosanoids, and oxidative injury. Research should continue to explore other potential therapeutic efficacies of NRT, including for ulcerative colitis, analgesia, weight reduction, Parkinsonâs disease, and cognitive disorders associated with aging and schizophrenia. The continued use of NRT in conjunction with continued, albeit re- duced, smoking prompts additional questions. For example, the constitu- ents of cigarette smoke that mediate tissue injury are not all precisely known, and it is also not known if modulating the coincident nicotine level might influence their absorption, metabolic disposition, mechanism of action, or elimination. Design of such studies will rely upon the devel- opment of more refined and tractable methodology to investigate the in vivo kinetics and dynamics of other constituents of cigarette smoke and their interactions with nicotine. Finally, although ethnicity has been shown to be relevant (Sabha et al., 2000), the factors that determine interindividual differences in nico- tine efficacy, safety, and addictive potential remain largely unexplored. Particular attention might be paid to genetic variation in proteins relevant to nicotine pharmacokinetics and dynamics and their interaction with environmental variables. As with other drugs, one anticipates increasing individualization of nicotine dosage and/or delivery when given as a therapeutic agent. Insight into the interaction of genetic and environmen- tal factors which influence initiation (Gynther et al., 1999; Heath et al., 1999) of cigarette smoking, latency until the practice becomes habitual (Stallings et al., 1999), and the quantity that is then smoked (Koopmans et al., 1999) has been increasing. Clarification of how these factors interact is also likely to afford insights of value in predicting the individual likelihood
THE SCIENTIFIC BASIS FOR PREP ASSESSMENT 163 of response to the use NRT as a strategy for quitting or reducing tobacco exposure. CANCER (SEE SECTION II, CHAPTER 12) Feasibility of Harm Reduction in Therapy There are sufficient laboratory and human data to suggest that harm reduction for cancer might be an achievable goal for persons who cannot stop smoking. There is evidence of decreased risk of cancer for persons who abstain from smoking, and there is strong evidence of a positive relationship between smoking and risk of developing cancer. The risk varies by the different tobacco products used and how they are used. Clearly, abstinence from smoking is the most effective method for reduc- ing cancer risk, and the cancer risk to former smokers is the lowest-level risk that might occur from the use of any PREP. Importantly, it must be recognized that the use of any harm reduction product will likely increase the risk of cancer at some level as long as there is exposure to tobacco carcinogens, in contrast to abstinence, which stops exposure to all tobacco constituents. Nonetheless, reduction in exposure to tobacco smoke and tobacco products to the lowest possible levels may provide some benefit to individual users and to the general population. However, there are insufficient data from which to conclude how much reduction in expo- sure would yield a measurable benefit and which individuals would ben- efit. Currently, it seems likely that methods that reduce exposure to to- bacco constituents to the greatest extent would likely provide the greatest benefit, but this remains to be proven. A systematic and thorough assessment of PREPs and cancer risk will require analysis of data obtained from well-designed laboratory and human studies. In laboratory animals, the shape of the dose-response curves differs for different tobacco constituents, indicating that the dose-response relationship of tobacco smoke is complex. Dose-Response Relationship In humans, the carcinogenic response increases most around five ciga- rettes per day, and there is relatively little increase in carcinogenicity above 20 cigarettes a day. However, while there are sufficient data to conclude that a dose-response relationship exists for the use of tobacco products and cancer risk, the precise dose-response relationship is really not known in part because exposure is not accurately measured without considering actual smoking behavior. There is some evidence to indicate that when internal exposure is considered through biomarkers, the shape
164 CLEARING THE SMOKE of the curve follows a quadratic equation, indicating a greater benefit in exposure reduction for persons who smoke more. Thus, data are insuffi- cient to predict the harm reducing effect of a change from any intensity of smoking to a PREP. There are sufficient data to suggest that dose-response relationships differ as a function of gender, race, age, and ethnicity, although the actual risk levels have not been sufficiently defined to draw definitive conclusions about risks among groups. Based on these types of data and possible modifiers of cancer risk (e.g., genetic susceptibilities, diet, lifestyle, occupation), it is likely that PREPs would affect risk differ- ently in different people and not at all in some. There is no evidence of a threshold below which tobacco smoking does not increase cancer risk. This conclusion is consistent with the fact that there are many carcinogens in tobacco smoke, and the aggregate might increase risk at any level. Modeling for low-dose exposures indi- cates that there is an increased risk with less than one cigarette per day. Thus, persons who initiate smoking with harm reduction products that contain tobacco would be likely to have an increased risk for cancer, and there is unlikely to be a âsafeâ cigarette. Former smokers who resume smoking with such products would increase their risk further. Regression of risk using PREPs might eventually bring a smoker to a risk equal to some lower level of lifetime exposure to conventional prod- ucts. However, there are insufficient data to validate this assumption or indicate that a decrease in risk would be measurable for some or all smok- ers. There are insufficient data to indicate the shape of the curve for re- gression of risk for any PREP. The data are sufficient to conclude, with some caveats, that filtered cigarettes compared to nonfiltered cigarettes pose a lower risk of lung cancer and possibly other cancers. The caveats are that this occurs only in persons who do not substantially increase the number of cigarettes they smoke per day or otherwise compensate by their smoking behavior for lower levels of nicotine. Also, these studies may be confounded by diet, lifestyle, or other characteristics of people who use filtered cigarettes, which might be different in smokers of filtered compared to nonfiltered cigarettes. The available data are suggestive, but not sufficient, to con- clude that smokers of low-tar cigarettes have a lower cancer risk com- pared to smokers of higher-tar cigarettes, with the same caveats as for the filter smoking studies. However, there are insufficient data to assess the differences in risk for ultralow-, low-, and high-tar cigarettes that are filtered. These cigarettes only became available more recently, so there has not been a long enough latency period in the general population to assess them until recently. There are insufficient data to adequately con- sider how risk changes when switching types of cigarettes.
THE SCIENTIFIC BASIS FOR PREP ASSESSMENT 165 This report has not reviewed potential cancer risks due to fibers re- leased from cigarette filters or tobacco additives, because it is thought that the risk of these exposures are substantially lower than the risk from the constituents of tobacco smoke. However, there are no existing data to prove this assumption. Importantly, as harm reduction products are de- veloped that substantially reduce exposure to tobacco constituents, the relative role of fibers and additives in carcinogenesis might become more important. Thus, fiber and additive exposure should be considered when assessing PREPs. Utility in a Preclinical Setting There are some experimental models (e.g., in vitro cell cultures, labo- ratory animals) that may be useful for the assessment of the carcinogenic- ity of tobacco-related PREPs. Although there are many reasonable models with which to assess individual tobacco smoke products, better models are needed for assessing exposures to complex mixtures. Such studies are not alone sufficient to support claims of potential harm reduction. No claim of potential harm reduction should be allowed without adequate human clinical and epidemiological studies. In vitro and animal studies, however, are very important for (1) determining those products that are not likely to result in measurable harm reduction (e.g., if the product results in exposures that increase genotoxicity, then there would be less enthusiasm for it and so should not be tested in a human clinical study and should not be introduced into the marketplace); (2) identifying unforeseen reactions (e.g., if a product reduces exposure but does not decrease tumors), then there might be some constituent or combination of constituents that is either new or more important than those changed in the product); (3) providing supportive evidence for the use of a particular bioassay in humans (e.g., if a biomarker predicts cancer risk in experi- mental animals); and 4) assessing the dose-response and the shape of the regression of risk for the PREP as exposure is reduced, although the data should be considered qualitative or semiquantitative and cannot be extrapolated directly to human smoking risk. Both in vitro cell culture and experimental animal studies should be used in assessing PREPs, where both can assess genotoxic and nongenotoxic end points, and chronic animal bioassays are needed to assess the end point of cancer risk. It is beyond the scope of the committee to recommend the specific panel of assays, but such a panel will need to be developed. Also, these studies should assess changes due to both specific carcinogens and to complex mixtures, where the latter should be mandatory.
166 CLEARING THE SMOKE Clinical Assessment of Tobacco-Related Disease and Biomarkers of Tobacco-Related Disease There is sufficient evidence to conclude that human experimental studies and short-term clinical studies provide evidence of the harmful effects of tobacco products. Thus, such studies can be used in assessing harm reduction. These studies, through the use of biomarkers and surro- gate indicators of cancer risk, can evaluate the manipulation of carcino- gens and nicotine to reduce exposures and how these changes might affect smoking behavior, metabolic activation, enzymatic induction, con- jugation, excretion, biologically effective doses (or their validated surro- gates), and biomarkers of potential harm. Separately, these studies can assess differences in risk and provide evidence for modifying effects due to genetic susceptibilities, diet, lifestyle, occupation, and so forth. How- ever, at the current time, no single biomarker or panel of biomarkers can be considered sufficient indicators of cancer risk by themselves, in part because most have not been sufficiently validated. New technologies are offering new opportunities for biomarkers. Thus, a panel of experts will be needed to devise a set of biomarkers that reflect different exposures, biologically effective doses, and pathways for potential harm. It is clearly possible to assess the effects of PREPs on cancer as the ultimate outcome, and only such studies can provide definitive evidence for the success of a product. However, the long latency for cancer makes these studies infeasible for making such claims today or in the near fu- ture. This relatively long latency period for cancer and the slower decline, probably years or decades, in risk from exposure reduction compared to cardiovascular disease and other tobacco-related diseases will have an impact not only on the time frame of PREP assessment and but also on the health effects experienced by and apparent to the individual. Preneo- plastic lesions or the identification of harmful effects in single cells might be used as indicators for the carcinogenetic pathway, but the technology to identify these in the general population or large epidemiological stud- ies is not yet available. In such studies, the characterization of smoking history and behavior is well validated for recent exposures but less accu- rate for assessing lifetime exposure. Also, self-reported smoking history is insufficient to adequately assess risk in the context of PREP assessments, so biomarkers also are needed to assess exposure, biologically effective doses, and potential harm. Currently, the best approach to assessing PREPs and cancer risk is to focus on lung cancer, because this is the most common cancer and so will provide studies with the greatest statistical power. However, data are sufficient to conclude that there is a risk that the widespread use of PREPs will shift the burden of cancer in the population from one type to another or from cancer to a different disease. Thus, a particular cancer type cannot
THE SCIENTIFIC BASIS FOR PREP ASSESSMENT 167 be the sole indicator for the success of a PREP, and other cancers, diseases, and overall mortality must be evaluated as well. Many studies of nicotine suggest that nicotine is unlikely to be a cancer-causing agent in humans or, at worst, that its carcinogenicity would be trivial compared to that of other components of tobacco. The consideration of nicotine as a carcinogenic agent, if at all, is trivial com- pared to the risk of other tobacco constituents. Some smokeless tobacco products increase the risk of oral cavity can- cers, and a dose-response relationship exists. However, the overall risk is lower than for cigarette smoking, and some products, such as Swedish snus, may have no increased risk. It may be considered that such products could be used as PREPs for persons addicted to nicotine, but these prod- ucts should undergo testing as PREPs using the guidelines and research agenda contained herein. The effects of PREPs on cancer risk from environmental tobacco smoke (ETS) are uncertain because of the difficulties in measuring reduc- tions in exposure. Also, although there is clearly an increased risk of lung cancer from ETS, the determination of changes in risk from the use of PREPs will require studies of large numbers of people, and smoking is currently in this country prohibited in many places where ETS might have occurred. CARDIOVASCULAR DISEASE (SEE SECTION II, CHAPTER 13) Dose-Response Relationship Highly informative information on the existence of a dose-response relationship between cigarette exposure and cardiovascular risk comes from many studies such as the CPS-II (Thun et al., 1997) and Harvard Nursesâ Health Studies (Kawachi et al., 1997). In both instances, there is a relationship between the number of cigarettes smoked and the incidence of cardiovascular events. This is illustrated for the incidence of myocar- dial infarction and stroke. In both cases, the most striking difference is between nonsmokers and individuals who smoke the least number of cigarettes recorded. The relationship becomes somewhat less pronounced as the number of cigarettes smoked per day increases. Interestingly, ex- smokers tend to occupy a space intermediate between nonsmokers and those with the lowest daily smoking frequency. This dose-response curve prompts several considerations. First, there is no persuasive evidence of a threshold below that a cardiovascular risk does not exist. This observation affirms the primary objective of encourag- ing smokers to quit completely. Second, the shallow dose-response rela- tionship, with the impression of a plateau, accords with similar observa- tions relating the number of cigarettes smoked and measurements of
168 CLEARING THE SMOKE systemic bioavailability, such as nicotine and cotinine. The same is true of the relationship with CO (Gori and Lynch, 1985). Although this may rep- resent saturation kinetics of nicotine or CO, the most likely explanation is compensation for lower numbers of cigarettes smoked. Thus, the smoker titrates nicotine delivery toward a range of convergence that is reflected by the measurement of nicotine delivery, which in turn, may reflect dose- dependent convergence of the delivery of additional toxic, but unmea- sured, constituents of cigarette smoke. The relative contribution of distinct constituents of cigarette smoke to smoking-related cardiovascular mor- bidity and mortality is unknown. In this regard, the available information is incomplete regarding any differences between the dose-response rela- tionship of filter or low-tar versus high-tar or unfiltered cigarettes. Feasibility of Harm Reduction in Therapy There are no data that are directly informative on the issue of harm reduction. Thus, although a dose-response relationship exists for cardio- vascular and cerebrovascular events, we do not know if reducing the number of cigarettes smoked results in a quantitative reduction in the risk of these events. However, the intuitively appealing prospect that this is indeed the case is supported by evidence from individuals who quit smok- ing. Thus, quitting results in a time-dependent reduction in the incidence of myocardial infarction and stroke. The latter is most nicely illustrated by data relating to subarachnoid hemorrhage, which was significantly elevated in women smokers in the Nursesâ Health Study. In addition to evidence from such unequivocal clinical events, there is evidence that the increase in biomarkers of oxidant stress, platelet activation, and inflam- mation (Benowitz et al., 1993), all of potential mechanistic relevance to tobacco-related cardiovascular injury, rapidly falls toward the normal range on quitting cigarettes. The offset kinetics of more functional surro- gates, such as endothelial dysfunction, remain to be determined in smok- ers. In summary, the data from quitters encourage the prospect that a graded reduction in cardiovascular risk and in biomarkers of this risk may accompany a reduction in the number of cigarettes smoked in pur- suit of a harm reduction strategy. It is possible, indeed likely, that indi- viduals and perhaps populations, differ in their susceptibility to tobacco- induced cardiovascular risk and, indeed, in their potential benefit from a harm reduction strategy. Data from quitting studies indicate a consider- able variance in the rate of offset of risk, which declines with time. No data are available to address such issues across ethnic groups or gender. Acquisition of such information and research on the environmental and genetic factors that condition interindividual variability in exposure-risk relationships are necessary.
THE SCIENTIFIC BASIS FOR PREP ASSESSMENT 169 Utility in a Preclinical Setting Studies in cell culture and model systems can afford much needed information on tobacco-related cardiovascular risk. These might include a profiling of gene expression and translation in cardiovascular tissues in response to cigarette smoke, constituents of smoke, and potential harm reduction substituents. These might identify proteins of potential func- tional relevance to the transduction of cardiovascular risk. Such studies might be coupled with gene inactivation and overexpression studies to address the role of these proteins in vivo. Similarly, studies of exposure to cigarette smoke or to discrete constituents of smoke might be deployed to investigate effects on atherosclerosis progression, susceptibility to vascu- lar injury, thrombotic stimuli, graft rejection, cardiovascular development, or endothelial dysfunction in model systems such as mice. Studies of cardiovascular genomics and ultimately proteomics can also be extended to model systems to investigate gene expression and translation in re- sponse to exposure to tobacco-related products in vivo. These observa- tions may, in turn, be related to the pattern of gene expression and trans- lation in cardiovascular tissues obtained from cigarette smokers. Biomarkers of Tobacco-Related Disease The predominant mechanisms by which cigarette smoking induces cardiovascular injury is unknown. However, small studies in smokers of potentially relevant biomarkers of platelet and vascular activation, lipid peroxidation, and inflammation afford evidence of a dose-response rela- tionship and a decline on quitting. There is even evidence of a signal in individuals exposed to ETS in the case of some of these markers. More mechanism-based clinical studies are required to confirm and expand these findings. Where possible, these should be related to surrogate mea- surements of cardiovascular function, such as hemodynamics, flow- mediated endothelial function and estimates of plaque progression by ultrasound or electron-beam computerized tomography (EBCT). Further- more, biomarker studies can usefully be integrated into many studies in model systems as well as studies of clinical outcome to afford their ulti- mate validation. Clinical Assessment of Tobacco-Related Disease The time course of offset of myocardial infarction and stroke in people who stop smoking suggests that cardiovascular disease represents a trac- table scenario in which one might evaluate harm reduction strategies. Clearly, the health effects experienced by the individual and the assessment
170 CLEARING THE SMOKE of the impact of such events can occur in a more reasonable time frame than from cancer in which declines in risk from tobacco exposure reduc- tion may only be apparent after years or decades. NONNEOPLASTIC RESPIRATORY DISEASE (SEE SECTION II, CHAPTER 14) In evaluating harm reduction strategies for tobacco-related lung dis- ease, three major nonneoplastic respiratory diseases linked to cigarette smoking are considered: COPD, asthma, and respiratory infections. Res- piratory diseases are major tobacco-related illnesses, and there is a clear need to mitigate the harmful effects of exposure to both mainstream and secondary tobacco smoke. It is generally accepted that cessation of smok- ing slows or stops the progression of the lung diseases related to smoking and it is plausible that decreasing smoking will reduce the severity of chronic lung diseases and the incidence of respiratory infections. How- ever, there is no adequate scientific evidence to support this because the effects of reduced smoking on harm reduction have not been extensively studied in man. Dose-Response Relationship There is a need to determine dose-response relationships more pre- cisely and to develop biomarkers of respiratory disease. Rational design of studies to assess harm reduction requires knowledge of the dose- response relationship. At present, such data for respiratory diseases are limited and of uncertain quality. Study design would also incorporate biomarkers of disease, and the testing of current and new biomarkers might be done concurrently in the models and populations studied for dose-effects. The Cancer Prevention Studies I and II, large-scale prospec- tive studies, however, do suggest a direct dose-response relationship be- tween cigarettes smoked per day and mortality rates from COPD (NIH, 1996, 1997), indicating that decreasing the number of cigarettes smoked may lead to fewer deaths from COPD. Biomarkers of Tobacco-Related Disease There are currently no specific molecular biomarkers of the nonneo- plastic respiratory diseases due to smoking tobacco products. No unique molecular or genetic defect specific for tobacco-related respiratory disease has been identified. The processes involved, such as inflammation and increased levels of oxidants, are not unique to tobacco-related respiratory
THE SCIENTIFIC BASIS FOR PREP ASSESSMENT 171 diseases. Identifying unique biomarkers is further confounded by the heterogeneous nature of these diseases, the complex mixture of tobacco smoke, and the range of individual susceptibilities to the harmful effects of tobacco smoke. The most widely used markers of tobacco-related respi- ratory diseases in population studies are symptom questionnaires and pulmonary function testing. These have well-known limitations of speci- ficity and sensitivity, particularly for detecting the early effects of tobacco smoke on lungs (U.S. DHHS, 1989). Subtle effects of tobacco smoke expo- sure on the lung can be detected by sampling fluid in the lower respira- tory tract via a bronchoscope inserted into the airways, but the signifi- cance of these changes for clinically important pulmonary disease has not been established. Newer approaches such as sampling the subjectsâ urine (Pratico et al., 1998) or exhaled gas (Ichinose et al., 2000) for metabolic products due to tissue injury have the advantage of noninvasive sam- pling but must be validated. Clearly, the greatest obstacle for rational development of a specific biomarker is the lack of fundamental informa- tion on mechanisms of how tobacco smoke exposure causes specific respi- ratory diseases. The availability of dose-response data and validated biomarkers may improve the design of contemplated intervention studies and allow greater confidence in the results. However, the time frame for generating dose-response data and testing biomarkers is uncertain. The inclusion of dose-response considerations and biomarkers in the design of clinical trials on reduction of harm from respiratory diseases must also be vali- dated. Clinical Assessment of Tobacco-Related Disease An alternative is to proceed with interventional trials based on cur- rent knowledge if there are uncertainties about the added value of dose- response data or untested biomarkers to study design. As an example, an intervention study of the effect of smoking reduction on COPD could be considered, similar in design to the Lung Health Study (Anthonisen et al., 1994), a large prospective trial of the effects of smoking cessation on rate of decline of FEV1 (forced expiratory volume at 1 second) in middle-aged smokers with mild COPD. Another approach is to conduct a trial using a low-tar and moderate-nicotine product made available from a noncom- mercial source to avoid product endorsement issues. Design of population studies for harm reduction of major respiratory diseases is challenging because of uncertainties about effectiveness and long-term compliance with harm reduction interventions. Reduction in the burden of tobacco-related respiratory diseases through harm reduc- tion strategies should be a major priority for the nationâs public health.
172 CLEARING THE SMOKE REPRODUCTIVE AND DEVELOPMENTAL EFFECTS (SEE SECTION II, CHAPTER 15) Feasibility of Harm Reduction in Therapy Cigarette smoking is a major cause of fetal and infant morbidity and mortality (U.S. DHHS, 1988, 1990; Kleinman et al., 1988). This is particu- larly true for the associations with low-birthweight and its consequences, as well as preterm delivery and SIDS (CDC, 2000; Leach et al., 1999; Shah and Bracken, 2000; U.S. DHHS, 1983). For several important adverse reproductive effects of maternal smoking, a decrease in smoking has been found to be associated with a decrease in risks to the fetus and infant (Li et al., 1993; Hebel et al., 1988). The greatest benefit, of course, comes from smoking cessation. However, the smoking cessation rate for women smokers who become pregnant is very low and remain comparable to those in the general population, despite knowledge of the harmful effects of smoking and personal experience with adverse fetal and infant conditions. More- over, as current rates of smoking increase slowly among adolescent women, these adverse effects associated with tobacco smoke exposure while pregnant are likely to worsen. Dose-Response Relationship On average, infants exposed to maternal smoking in utero are 200 grams lighter and 1.4 cm shorter than those unexposed (Wang et al., 1997). A strong dose-response relationship has been supported in numer- ous studies (Li et al., 1993),and a decrease in dose (number of cigarettes) in controlled studies has led to increased birthweights in a predictable pattern (Wang et al., 1997). What is known about the mechanism of effect of cigarette smoke on the fetus suggests that several agents in tobacco smoke contribute to the adverse effects. There is evidence that CO plays a major role in growth retardation through increased tissue hypoxia (Benowitz et al., 2000). Nicotine has also been thought to play a role through increasing vasoconstriction and decreasing perfusion through the placenta. Although nicotine replacement products and buproprion are cur- rently not approved by the Food and Drug Administration for use by pregnant women, the Agency for Healthcare Research and Qualityâs (AHCRQ) Clinical Practice Guidelines for Treating Tobacco Use and De- pendence (Fiore et al., 2000) recommend that âPharmacotherapy should be considered when a pregnant woman is otherwise unable to quit, and when the likelihood of quitting, with its potential benefits, outweighs the risks of the pharmacotherapy and potential continued smokingâ. It is generally thought that NRT can reasonably be used with pregnant patients
THE SCIENTIFIC BASIS FOR PREP ASSESSMENT 173 if prior behavioral modifications have failed and the patient continues to smoke at least 10-15 cigarettes per day (ACOG, 1997). There are no data regarding the efficacy of potential reduced-exposure products (PREPs) during pregnancy, but there is the presumption that the tobacco-related PREPs are likely to have adverse effects at some level and that until fur- ther evidence is produced, existing guidelines concerning pharmacologic PREPs still pertain. Clinical Assessment of Tobacco-Related Disease and Utility in a Preclinical Setting To practically assess the health effects of PREPs, reliable measures of health outcomes that can be utilized in a relatively short time are desired. Among the reproductive outcomes of maternal smoking, intrauterine growth retardation resulting in low-birthweight babies has been studied extensively, and a large body of evidence has supported a causal link with cigarette smoke exposure. The committee recommends, based on currently available scientific knowledge, that fetal birthweight be used as a reliable outcome measure for evaluating the harm reduction potential of specific PREPs. Study designs should include repeated cohort or case- control studies of pregnant women, with an appropriate distribution of exposures to both PREPs and conventional products, and suitable con- trast groups. Concomitant, coordinated toxicological studies should be undertaken to provide biological correlations with clinical outcomes. Such outcomes as fetal birth weight and the incidence of other reproductive and developmental health outcomes (e.g., fertility outcomes, placental complications, gestational age at birth, incidence of sudden infant death syndrome [SIDS], spontaneous abortions) should be considered primary objects of study in order to assess the harm reduction potential of specific PREPs. Findings in pregnant women exposed to PREPs may have value be- yond maternal or fetal outcomes. The nature of adverse effects derived from PREP exposure will likely be determined much sooner in this case than findings on chronic disease outcomes in humans, such as various cancers and cardiovascular disease. Should adverse findings become ap- parent, there may be substantial implications for chronic illnesses among older adults, and coordinated pathogenic studies might allow conclu- sions on new tobacco product outcomes in advance of studies exploring longer âincubation periods.â The committee recommends that further basic research be undertaken to elucidate the components of cigarette smoke that are primarily respon- sible for adverse health outcomes. In order to evaluate the safety of many PREPs, it is important to understand the toxicity of specific smoke
174 CLEARING THE SMOKE components, especially nicotine and CO, on the pathogenesis of intrauter- ine growth retardation, spontaneous abortion, and other health outcomes. In addition, a better understanding of the risks of bupropion SR use by pregnant women (i.e., seizure risk) and the teratogenic effects of nicotine on the central nervous system (CNS) is needed for adequate risk-benefit analysis of the harm reduction potential of these products. Surveillance of Tobacco Use Patterns Among Pregnant Women Central to understanding exposure to tobacco products is continuous population information on usage patterns among pregnant women. This may not be attainable by general population survey methods because of inadequate sample sizes and insufficient representation of various geo- graphic or demographic groups or of the earliest stages of pregnancy. There is a need for surveys devoted specifically to pregnant women in all stages of gestation, irrespective of the receipt of medical care. Survey content should include other known or putative causes of adverse mater- nal or fetal outcomes, as well as detailed product types and usage pat- terns. Recommendations for general population surveillance can be found in Chapter 6 of this report. Biochemical and toxicological exposure measures should be a routine part of surveillance for exposure to conventional products as well as PREPs. These will be necessary to conduct more precise, coordinated toxi- cological studies and also to assess actual exposure rates more accurately. For example, dose may be measured by maternal serum and urine cotinine levels, which have shown reliable correlations with maternal, and conse- quently fetal, tobacco smoke exposure. Self-reported data have been found unreliable, since pregnant women tend to underreport tobacco use be- cause of the stigma attached to smoking. Also, self-reports do not ad- equately account for differences in depth and frequency of puffs among smokers. OTHER HEALTH EFFECTS (SEE SECTION II, CHAPTER 16) Feasibility of Harm Reduction in Therapy Several important diseases and conditions of adults, in addition to cardiovascular diseases, chronic obstructive lung disease and various can- cers, have been associated with tobacco use, includingâbut not limited toâpeptic ulcer disease, poor wound healing, inflammatory bowel dis- ease, rheumatoid arthritis, oral disease, dementia, osteoporosis, ocular disease, diabetes, dermatological disease, schizophrenia, and depression (see Chapter 16). Some of these associations are supported by substantial
THE SCIENTIFIC BASIS FOR PREP ASSESSMENT 175 scientific evidence, and a causal linkage is likely. These illnesses must ultimately be subjected to the same evaluation of changing risks and out- comes associated with PREPs, because these are common and clinically important conditions, even if they are not as often fatal as cancer, cardio- vascular disease or pulmonary disease. Further, each of the conditions for which the association with tobacco use is substantial also offers the op- portunity to address pathogenic mechanisms related to the varying con- stituents of PREPs, as well as the impact on disease incidence of concomi- tant behaviors and exposures such as alcohol use, various dietary elements, and certain medications. Utility of Preclinical Studies and Short-term Indicators of Clinical Harm Reduction Some of the conditions reviewed in this chapter may be applied as indicators of the general biological effects of new tobacco products. For example, cigarette smoking has been consistently found to be an indepen- dent risk factor for an adverse clinical course of both peptic ulcer disease and wound healing. The effects of smoking on ulcer formation and heal- ing have been clearly described clinically and in animal models (Ma et al., 1999). Peptic ulcers have been found to be larger, slower to heal, and more likely to recur among smokers and to exhibit clinically improved healing upon cessation (Tatsuta et al., 1987). Surgical and traumatic wounds heal more slowly among cigarette smokers (Kwaitkowski et al., 1996; Mosely et al., 1978). The committee recommends that rigorous clinical studies be designed and executed to determine whether variations in ulcer and wound-healing rates are related to various categories of tobacco prod- ucts, including those with claims of harm reduction. This may offer the opportunity to define some clinical outcomes that have clinical relevance in their own right and to identify potential indicators of harm alteration much sooner after the introduction of PREPs than would be possible when evaluating heart disease and cancer. Other candidate diseases for such evaluation might include periodon- tal disease (Bergstrom, 2000; Haber, 1994), Crohnâs disease (Rhodes and Thomas; 1994), and rheumatoid arthritis (Uhlig et al., 1999). Here the outcomes to assess would be the effect of various conventional tobacco products and PREPs on the natural history of these conditions, including intermittancy, progression or regression, and longitudinally collected biomarkers of disease severity. As noted above, PREPs that alter the his- tory and outcomes of these conditions could be further evaluated for specific constituent exposures associated with this altered history. This may lead to a more refined understanding of pathogenic mechanisms as well.
176 CLEARING THE SMOKE Clinical and basic research on intermediate clinical outcomes is also needed. For example, as noted in this chapter, the risk of osteoporosis has also been strongly linked to cigarette smoking. In controlled observa- tional studies, bone mineral density has been found to be significantly lower among cigarette smokers, which contributes to a greater risk of osteoporotic fractures among older populations. While the effects of smok- ing on fracture rates may take a few decades or longer to detect, it is possible that surveillance of bone mineral density among those using PREPs and conventional products may be informative in a shorter time period and, thus, serve to detect important outcomes over an interval in which tobacco policy and clinical preventive interventions may have their greatest effects. Surveillance The committee recommends that selected conditions, as reviewed in this chapter, be part of a comprehensive, population-based surveillance program, outlined in Chapter 6. This will allow determination of the rela- tionship between the use of PREPs and of trends in occurrence for these tobacco-related conditions and assessment on a national basis of whether changes in tobacco product use have an effect on these important health problems. Based on these surveillance findings, more specific population, clinical, and basic research studies can be directed to evaluate PREPs to pursue causal mechanisms and to suggest more effective interventions. REFERENCES ACOG (American College of Obstetricians and Gynecologists). 1997. Educational Bulletin: Smoking and womenâs health. International Journal of Gynecology & Obstetrics 60:71-82. Anthonisen NR, Connett JE, Kiley JP, Altose MD, Bailey WC, Buist AS, Conway WA Jr, Enright PL, Kanner RE, OâHara P, et al. 1994. Effects of smoking intervention and the use of an inhaled anticholinergic bronchodilator on the rate of decline of FEV1. The Lung Health Study. JAMA 272(19):1497-1505. Benowitz NL, Dempsey DA, Goldenberg RL, Hughes JR, Dolan-Mullen P, Ogburn PL, Oncken C, Orleans CT, Slotkin TA, Whiteside HP Jr, Yaffe S. 2000. The use of pharma- cotherapies for smoking cessation during pregnancy. Tob Control 9 Suppl 3(2):III91-94. Benowitz NL, Fitzgerald GA, Wilson M, Zhang Q. 1993. Nicotine effects on eicosanoid formation and hemostatic function: comparison of transdermal nicotine and cigarette smoking. J Am Coll Cardiol 22(4):1159-1167. Benowitz NL, Zevin S, Jacob P 3rd. 1998. Suppression of nicotine intake during ad libitum cigarette smoking by high-dose transdermal nicotine. J Pharmacol Exp Ther 287(3):958- 962. Bergstrom J, Eliasson S, Dock J. 2000. Exposure to tobacco smoking and periodontal health. J Clin Periodontol 27(1):61-68. CDC (Centers for Disease Control and Prevention). 2000. Tobacco use during pregnancy. National Vital Statistics Report 48(3):10-11.
THE SCIENTIFIC BASIS FOR PREP ASSESSMENT 177 Chalon S, Moreno H Jr, Benowitz NL, Hoffman BB, Blaschke TF. 2000. Nicotine impairs endothelium-dependent dilatation in human veins in vivo. Clin Pharmacol Ther 67(4):391-397. Davis DL, Nielsen MT, eds. 1999. World Agriculture Series; Tobacco Production, Chemistry and Technology. London: Blackwell Science. Eclipse Expert Panel. 2000. A safer cigarette? A comparative study. A consensus report. Inhalation Toxicology 12(Supp 5):1-48. FagerstrÃ¶m, K, Tejding, R, Westin, A, Lunell, E. 1997. Aiding reduction of smoking with nicotine replacement medications: hope for the recalcitrant smoker? Tobacco Control 6: 311-316. Fiore, M, Bailey, W, Cohen, S, Dorfman, S, Goldstein, M, Gritz, E, Heyman, R, Jaen, C, Kottke, T, Lando, H, Mecklenburg, R, Mullen, P, Nett, L, Robinson, L, Stitzer, M, Tommasello, A, Villejo, L, Wewers, M. 2000. Treating Tobacco Use and Dependence. Clinical Practice Guideline. Rockville, MD: U.S. Department of Health and Human Ser- vices. Public Health Service. Gairola CG, Daugherty A. 1999. Acceleration of atherosclerotic plaque formation in ApoE- mice by exposure to tobacco smoke. The Toxicologist 48:297. Gori GB, Lynch CJ. 1985. Analytical cigarette yields as predictors of smoke bioavailability. Regul Toxicol Pharmacol 5(3):314-326. Gynther LM, Hewitt JK, Heath AC, Eaves LJ. 1999. Phenotypic and genetic factors in mo- tives for smoking. Behav Genet 29(5):291-302. Haber J. 1994. Smoking is a major risk factor for periodontitis. Curr Opin Periodontol: 12-8. Heath AC, Kirk KM, Meyer JM, Martin NG. 1999. Genetic and social determinants of initia- tion and age at onset of smoking in Australian twins. Behav Genet 29(6):395-407. Hebel JR, Fox NL, Sexton M. 1988. Dose-response of birth weight to various measures of maternal smoking during pregnancy. J Clin Epidemiol 41(5):483-489. IARC (International Agency for Research on Cancer). 1986. Tobacco Smoking: Monographs on the Evaluation of Carcinogenic Risk of Chemicals to Humans. Vol. 38 ed. Lyon, France: IARC. Ichinose M, Sugiura H, Yamagata S, Koarai A, Shirato K. 2000. Increase in reactive nitrogen species production in chronic obstructive pulmonary disease airways. Am J Respir Crit Care Med 162(2 Pt 1):701-706. Kawachi I, Colditz GA, Stempfer MJ, Willett WC, Manson JE, Rosner B, Hunter DJ, Hennekens CH, Speizer FE. 1997. Smoking cessation and decreased risk of total mor- tality, stroke, and coronary heart disease incidence in women; a prospective cohort study. National Cancer Institute. Changes in Cigarette-Related Disease Risks and Their Implication for Prevention and Control. NCI Smoking and Tobacco Control Monograph 8. Washington, DC: NCI. Pp. 531-565. Kleinman JC, Pierre MB, Madans JH, Land GH, Schramm WF. 1988. The effects of maternal smoking on fetal and infant mortality. Am J Epidemiol 127(2):274-282. Koopmans JR, Slutske WS, Heath AC, Neale MC, Boomsma DI. 1999. The genetics of smok- ing initiation and quantity smoked in Dutch adolescent and young adult twins. Behav Genet 29(6):383-393. Kwiatkowski TC, Hanley Jr EN, Ramp WK. 1996. Cigarette smoking and its orthopedic consequences. Amer J Orthopedics 25(9):590-597. Leach CE, Blair PS, Fleming PJ, Smith IJ, Platt MW, Berry PJ, Golding J. 1999. Epidemiology of SIDS and explained sudden infant deaths. CESDI SUDI Research Group. Pediatrics 104(4):e43. Li CQ, Windsor RA, Perkins L, Goldenberg RL, Lowe JB. 1993. The impact on infant birth weight and gestational age of cotinine-validated smoking reduction during pregnancy. JAMA 269(12):1519-1524.
178 CLEARING THE SMOKE Ma L, Chow JY, Cho CH. 1999. Cigarette smoking delays ulcer healing: role of constitutive nitric oxide synthase in rat stomach. Am J Physiol 276(1 Pt 1):G238-248. Mosely LH, Finseth F, Goody M. 1978. Nicotine and its effect on wound healing. Plastic and Reconstructive Surgery 61(4):570-574. NIH (National Institutes of Health), National Cancer Institute. 1996. The FTC Cigarette Test Method for Determining Tar, Nicotine, and Carbon Monoxide Yields of U.S. Cigarettes. Smok- ing and Tobacco Control Monograph 7. Bethesda, MD: National Institutes of Health. NIH (National Institutes of Health), National Cancer Institute. 1997. Changes in Cigarette- Related Disease Risks and Their Implication for Prevention and Control. Smoking and Tobacco Control Monograph 8. Bethesda, MD: National Institutes of Health. NRC (National Research Council). 1983. Risk Assessment in the Federal Government: Managing the Process. Washington, DC: National Academy Press. Perera FP. 1987. Molecular cancer epidemiology: a new tool in cancer prevention. J Natl Cancer Inst 78(5):887-898. Practico D, Basili S, Vieri M, Cordova C, Violi F, Fitzgerald GA. 1998. Chronic obstructive pulmonary disease is associated with an increase in urinary levels of isoprostane F2alpha-III, an index of oxidant stress. Am J Respir Crit Care Med 158(6):1709-1714. Rennard, S, Daughton, D, Fujita, J, Oehlerking, M, Dobson, J, Stahl, M, Dobson, J, Stahl, M, Robbins, R, Thompson, A. 1990. Short-term smoking reduction is associated with re- duction in measures of lower respiratory tract inflammation in heavy smokers. Euro- pean Respiratory Journal 3(7):752-759. Rhodes J, Thomas GA. 1994. Smoking: good or bad for inflammatory bowel disease? Gastro- enterology 106(3):807-810. Sabha M, Tanus-Santos JE, Toledo JC, Cittadino M, Rocha JC, Moreno H Jr. 2000. Transdermal nicotine mimics the smoking-induced endothelial dysfunction. Clin Pharmacol Ther 68(2):167-174. Shah NR, Bracken MB. 2000. A systematic review and meta-analysis of prospective studies on the association between maternal cigarette smoking and preterm delivery. Am J Obstet and Gynecol 182(2):465-472. Shiffman S, Mason KM, Henningfield JE. 1998. Tobacco dependence treatments: review and prospectus. Annu Rev Public Health 19:335-358. Sopori ML, Kozak W, Savage SM, Geng Y, Soszynski D, Kluger MJ, Perryman EK, Snow GE. 1998. Effect of nicotine on the immune system: possible regulation of immune responses by central and peripheral mechanisms. Psychoneuroendocrinology 23(2):189- 204. Stallings MC, Hewitt JK, Beresford T, Heath AC, Eaves LJ. 1999. A twin study of drinking and smoking onset and latencies from first use to regular use. Behav Genet 29(6):409-421. Tatsuta M, Iishi H, Okuda S. 1987. Effects of cigarette smoking on the location, healing and recurrence of gastric ulcers. Hepatogastroenterology 34(5):223-228. Thun MJ, Myers DG, Day-Lally C, Namboodiri MM, Calle EE, Flanders WD, Adams SL, Heath CW. 1997. Age and exposure-response relationships between cigarette smoking and premature death in cancer prevention study II. National Cancer Institute. Changes in Cigarette-Related Disease Risks and Their Implication for Prevention and Control. NCI Smoking and Tobacco Control Monograph 8. Washington, DC: NCI. Pp. 383-475. Transdermal Nicotine Study Group. 1991. Transdermal nicotine for smoking cessation. JAMA 266(22):3133-3138. U.S. DHHS (U.S. Department of Health and Human Services). 1983. The Health Consequences of Smoking; Cardiovascular Disease; A Report of the Surgeon General. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Preven- tion.
THE SCIENTIFIC BASIS FOR PREP ASSESSMENT 179 U.S. DHHS (U.S. Department of Health and Human Services). 1988. The Health Consequences of Smoking, Nicotine Addiction; A Report of the Surgeon General. Atlanta, GA: U.S. Depart- ment of Health and Human Services, Centers for Disease Control and Prevention. U.S. DHHS (U.S. Department of Health and Human Services). 1989. Reducing the Health Consequences of Smoking; A Report of the Surgeon General. Washington, DC: U.S. Depart- ment of Health and Human Services, Centers for Disease Control and Prevention. U.S. DHHS (U. S. Department of Health and Human Services). 1990. The Health Benefits of Smoking Cessation; A Report of the Surgeon General. Rockville, MD: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention. Uhlig T, Hagen KB, Kvien TK. 1999. Current tobacco smoking, formal education, and the risk of rheumatoid arthritis. J Rheumatol 26(1):47-54. Vineis P, Porta M. 1996. Causal thinking, biomarkers, and mechanisms of carcinogenesis. J Clin Epidemiol 49(9):951-956. Wang X, Tager IB, Van Vunakis H, Speizer FE, Hanrahan JP. 1997. Maternal smoking dur- ing pregnancy, urine cotinine concentrations, and birth outcomes. A prospective co- hort study. Int J Epidemiol 26(5):978-988. Witschi H, Espiritu I, Maronpot RR, Pinkerton KE, Jones AD. 1997a. The carcinogenic po- tential of the gas phase of environmental tobacco smoke. Carcinogenesis 18(11):2035- 2042. Witschi H, Espiritu I, Peake JL, Wu K, Maronpot RR, Pinkerton KE. 1997b. The carcinoge- nicity of environmental tobacco smoke. Carcinogenesis 18(3):575-586. Witschi H, Espiritu I, Uyeminami D. 1999. Chemoprevention of tobacco smoke-induced lung tumors in A/J strain mice with dietary myo-inositol and dexamethasone. Car- cinogenesis 20(7):1375-1378. Witschi H, Uyeminami D, Moran D, Espiritu I. 2000. Chemoprevention of tobacco-smoke lung carcinogenesis in mice after cessation of smoke exposure. Carcinogenesis 21(5):977- 982.