"It was the unanimous and fundamental conclusion of the Tobacco Research Implementation Group that an unequivocal commitment of the NCI to a comprehensive but focused program of research on tobacco use can help to reverse the epidemic of tobacco-related cancers."	Tobacco-related cancers exact an exorbitant toll on the Nation's health. Tobacco use is responsible for over 450,000 deaths and 170,000 cancer deaths each year in the United States-more than 30 percent of all cancer deaths. Worldwide, tobacco use causes an estimated 3 million deaths per year, and this number is expected to rise to 10 million deaths annually by the 2020s or early 2030s. Smoking, the most common form of tobacco use, is a known cause of cancer in eight major sites, many of which are among the most difficult to diagnose and treat. Smoking causes more than 85 percent of all deaths from lung cancer and over 50 percent of all deaths from cancers of the larynx, oral cavity, and esophagus. Further, smoking is a known cause of chronic obstructive lung disease, peripheral vascular disease, coronary heart disease, stroke, and many other serious diseases. Exposure to cigarette smoke in the environment also can cause lung cancer and respiratory symptoms. It is difficult to overstate the health burden of smoking; the annual death toll from smoking exceeds the combined annual death toll from all accidents, suicides, drug use (both licit and illicit), homicides, murders, and AIDS. It is as though two fully loaded jumbo jets crashed every day of the year-with no survivors. Using smokeless tobacco and smoking cigars also results in serious health problems. Smokeless tobacco is causally related to several forms of oral cancer, particularly cancers of the cheek and gum, and causes other noncancerous oral diseases. Cigar use causes cancers of the lung, larynx, esophagus, and oral cavity. Tobacco-related deaths today reflect the cumulative outcome of tobacco use over decades. Both current smokers and former smokers are affected. Reductions in smoking prevalence over the past several decades provide some optimism for the future. Today, just under 25 percent of adults in the United States are cigarette smokers, including those who do not smoke daily (20 percent of the total). These figures are an improvement over smoking prevalence rates in the middle part of this century, when nearly 60 percent of men and 30 percent of women smoked and most smoked every day. Nonetheless, the number of current smokers, and thus future cancers, remains alarmingly high. Forty-six million adults in the United States alone currently smoke cigarettes, and millions more use other forms of tobacco. In contrast to the steady declines in smoking rates through the 1970s and 1980s, smoking among adults over the past few years appears to have leveled off. Moreover, rates of smoking are disproportionately high among certain population groups, such as African-American men. Other areas of concern persist. Smoking among adolescents has been increasing since the early 1990s. A recent report from the CDC shows a 73 percent increase in teenage smoking from 1988 levels. More than 6,000 adolescents try their first cigarette and over 3,000 teens become regular smokers every day. These findings are consistent with previous studies that suggest significant increases in smoking prevalence among adolescents in the U.S. since 1991. Overall, these data show that public health gains observed during the l970s and 1980s are being reversed. Moreover, promising declines in smoking among black teens have reversed. In the early 1990s, fewer than 10 percent of black high school seniors reported smoking in the past 30 days, compared to over 30 percent of white teens. Unfortunately, smoking among black teens has risen over 50 percent in just the past five years. These are alarming trends. Equally troubling is the increasing prevalence of smoking now being reported among younger teens. Should these patterns continue into adulthood, they will have a profound impact on future cancer rates and smoking-related diseases. Trends in the use of other tobacco products, such as smokeless tobacco and cigars, are also of concern. The U.S. Department of Agriculture reports that consumption of moist snuff has surpassed chewing tobacco for the first time in U.S. history. Cigar use, which declined for a period of over 20 years, increased by nearly 50 percent between 1993 and 1997. Consumption of large cigars, which accounts for over two-thirds of all cigars in the U.S., increased by nearly 70 percent over the same time period. National and regional surveys of children have documented high rates of current cigar use among both boys and girls. Cigar use among adolescent males is two to three times higher than their use of smokeless tobacco and is approaching their rate of cigarette smoking. Cigar use also appears to be gaining in popularity among adolescent girls. Moreover, cigar use has gained acceptance among well-educated men and women 18-34 years, a group that had previously eschewed cigarettes. It is not known whether their use of cigars could serve as a gateway to use of other tobacco products. THE NCI's TOBACCO RESEARCH PROGRAM The NCI has played a vital and leading role in the battle against tobacco use. The NCI's leadership, commitment to and involvement in tobacco-related research dates back more than 40 years, and NCI research on smoking and tobacco has spanned a broad spectrum, from basic biology to epidemiology to prevention and treatment. The NCI's initial research efforts focused primarily on epidemiologic studies of the association between tobacco use and cancer and on the basic biology of tobacco-induced cancers. Later research identified hazardous elements in tobacco and tobacco smoke and sought ways of reducing exposure. The NCI's cancer control research in the past 15 years focused on why people smoke and the development of interventions to prevent and stop tobacco use. During most of the 1980s, the NCI's smoking research program emphasized interventions in three broad areas: those delivered through specific intervention channels (e.g., schools, worksites, mass media, health care settings, and community groups); interventions that targeted specific populations (e.g., minority and ethnic populations and women); and groups of tobacco users considered at high risk (e.g., poor smokers and heavy smokers). This research provided a strong foundation of empirical information about tobacco control interventions. Behavioral models of tobacco use cessation were established; characteristics of successful school-based interventions to prevent or delay tobacco use initiation were defined; physician training and office protocols for patient smoking cessation programs were developed; and approaches for utilizing and enhancing self-help interventions were identified. Mass media interventions capable of reaching large numbers of individuals with prevention and cessation messages were developed and strategies for reaching minority, ethnic, and high-risk populations were tested. With this information as the cornerstone, the NCI launched two projects that sought to translate this knowledge about tobacco control interventions into widespread public health impact. These two initiatives-the Community Intervention Trial for Smoking Cessation (COMMIT) and the American Stop Smoking Intervention Study for Cancer Prevention (ASSIST)-are important for their unique designs and because they remain two of the largest community-based smoking control efforts ever attempted. These initiatives are described in the Community and State Intervention Research section. Concurrent with the implementation of these large-scale community-based tobacco control projects, sweeping societal changes in attitudes, public policy, and legal regulation of tobacco products have taken place. Government agencies, professional and voluntary organizations, and academic institutions joined to eliminate tobacco use and exposure to tobacco carcinogens. Public attitudes reflected decreasing acceptance of smoking as a social norm, and public policies restricted the locations where smoking was permitted and limited the access of minors to tobacco products. The U.S. Food and Drug Administration (FDA), with support from the President of the United States, asserted jurisdiction over tobacco products. States' attorneys general sued tobacco firms, and secret files from the tobacco industry and depositions from industry officials exposed the continued promotion of tobacco use by the tobacco industry in full knowledge of the myriad related health effects. These factors converged to create the current climate of societal readiness to tackle the formidable challenge of reducing the health burden of tobacco use. During this time, spectacular advances also were made in understanding the molecular and genetic factors underlying the progression of a normal cell into a cancer cell. Research focusing on the role of tobacco as a trigger for the cell changes that lead to cancer has revealed that different tobacco products and methods of nicotine administration influence the type and quantity of exposure to cancer-inducing agents. Other tobacco product characteristics, such as burning temperatures, additives, and filter composition, have been found to alter these carcinogenic agents as well. Research also has identified three specific classes of agents in tobacco smoke that cause cancer: nitrosamines, aromatic amines, and polycyclic aromatic hydrocarbons. While all the specific sites of action of these tobacco carcinogens are not yet characterized, many of the molecular changes induced in different cancer sites (e.g., head and neck, lung) appear similar, suggesting a common mechanism of epithelial cell damage in these sites. These discoveries open the door to identifying precancerous lesions and biomarkers that may serve as predictors of cancer in these and other sites. Additionally, research on the molecular characteristics of tobacco-induced cancers has identified differences in lung cancer mutation characteristics. For example, lung cancers in women more commonly have estrogen and progesterone receptors, and these hormones stimulate cancer growth. Some epidemiologic studies also have provided evidence of differences in the proportions of specific types of lung cancer among men and women and possible differences in risk associated with cigarette smoking. These significant advances over the past several decades in basic and applied research, as well as changes in public attitudes and policies, present an unprecedented opportunity to expand research to gain new insights that will significantly reduce the burden of death and disease caused by tobacco use. But while there is reason for optimism about the potential to exert a positive impact on the public health problem of tobacco use, the challenge is formidable. It is clear that reducing the burden of tobacco use requires novel strategies, borne from new ways of thinking about old problems. A deliberate, focused plan is required and a renewed commitment to ending this public health threat is essential. It is imperative that the NCI focus its research agenda, identifying where research is most needed and how best to prioritize and achieve research objectives that will have the greatest impact on the problem. THE NCI TOBACCO RESEARCH IMPLEMENTATION GROUP To accomplish this goal, the Director of the NCI created the Tobacco Research Implementation Group (TRIG), which includes more than two dozen leading scientists and experts from within the NCI, the National Institute on Drug Abuse (NIDA), and the Office of Behavioral and Social Sciences Research (Office of the Director), and from the extramural research community, as well as representatives of major NCI review and advisory committees. The TRIG was charged with establishing the NCI's tobacco-related cancer research priorities for the next 5 to 7 years. The TRIG began by reviewing the extensive recommendations of four earlier advisory groups.1 Each of these excellent reports produced major recommendations for tobacco control research. While a number of the recommendations from these earlier review groups had already been partially or completely implemented, much remained to be done; no single previous report had considered the entire spectrum of tobacco control research, from basic biological research to dissemination research.2 1 Major reports reviewed included the report of the Behavioral Research in Cancer Control Meeting, published in Preventive Medicine in 1995, the Report of the National Cancer Institute Cancer Prevention Program Review Group (1997), the Report of the National Cancer Institute Cancer Control Review Group (1997), and Taking Action To Reduce Tobacco Use, a report of the National Cancer Policy Board of the Institute of Medicine (1998). 2 For example, a major recommendation to create a tobacco control research branch has been accomplished, and a branch chief has been appointed. An earlier recommendation that also is being implemented is the transfer of the American Stop Smoking Intervention Study (ASSIST) to the Centers for Disease Control and Prevention. In addition, new initiatives have been developed in biobehavioral research, genetic regulation of susceptibility to tobacco-related carcinogenesis, and the prevention of tobacco use among youth. The TRIG also analyzed the FY97 portfolio of tobacco-related research across NCI divisions. This analysis reviewed the balance of Cancer Control Research across topic areas and the distribution of funding across projects and by funding mechanism; it also identified emerging issues from research in progress. Research projects (intramural and extramural) were included if at least 10 percent of the project was related to tobacco. Projects focusing upon treatment of tobacco-related diseases (e.g., interventions with lung cancer patients) were excluded from portfolio analysis. The NCI spent $76.2 million on 175 tobacco-related research projects in FY97. This represents less than 3 percent of the NCI's budget. The disproportionate funding of tobacco research relative to the NCI's total research budget is remarkable given the contribution of tobacco to the total cancer burden (approximately 30 percent of all cancer deaths). Projects were categorized into nine broad research areas, including biobehavior, prevention of nicotine use, treatment of nicotine addiction, state and community interventions, policy, surveillance, basic biology, epidemiology, and chemoprevention. The definitions of these categories may be found at the beginning of each section of the report. More detailed presentation of the portfolio analysis is provided in the Appendix. Epidemiology, treatment of tobacco addiction, and community and state intervention research each accounted for about 20 percent of the total number of tobacco projects funded by the NCI. This picture differed when assessed by the allocation of funding, with community and state intervention research accounting for more than 40 percent of the FY97 funds for tobacco-related research. This category will be reduced as the ASSIST program is transferred from the NCI to the Centers for Disease Control and Prevention (CDC) in FY99. Research on the treatment of tobacco addiction accounted for 18 percent of the funds expended. When assessed either by the number of projects or the amount of funding, biobehavioral, prevention, and policy research accounted for less than 10 percent of the tobacco portfolio. Surveillance was the most undersupported area in terms of tobacco-specific funding. Research grants (R01) and contracts (N01) represented the most frequently used funding mechanism, accounting for almost three-quarters of the tobacco research funding combined. TRIG members generated research recommendations after reviewing the portfolio analysis. These proposed recommendations provided the foundation for subsequent discussion of research priorities and development of an initial draft of the Tobacco Research Implementation Plan (TRIP). The group then sought input from experts reflecting various research and public health perspectives. Several scientists were provided with the initial draft of the TRIP and invited to present their perspectives to the group. Following this input, TRIG members narrowed and refined the list of research recommendations and, through a consensus-building process, identified and prioritized a core set of tobacco-related cancer research opportunities. These recommendations build on past research accomplishments, are consistent with the recommendations of previous advisory groups, address major gaps in knowledge, and have the potential for great impact on the cancer burden caused by tobacco. The recommendations reflect research across scientific disciplines and levels of focus from the basic structure of the cell to the broad influence of society. Because nicotine addiction plays a critical role in the ongoing use of tobacco products, a better understanding of nicotine addiction is a vital first step toward preventing or stopping tobacco use. This report places a major emphasis on the factors that influence tobacco use and nicotine addiction. The Biobehavioral Model of Nicotine Addiction and Tobacco-Related Cancers provides a framework for conceptualizing the broad spectrum of research factors. Tobacco use and nicotine addiction arise from a complex interplay of social, psychological, and biological factors that interact with genetic vulnerabilities to nicotine addiction. Social influences are broad, including peer and family modeling, tobacco industry marketing, and media influences. Depressed or anxious mood and attention-deficit hyperactivity disorder (ADHD) are examples of psychological factors that affect smoking, from initiation to maintenance to cessation. The effects of these three overarching determinants are mediated by behavioral, neurochemical, and physiological factors to influence tobacco use, dependence, cessation, and relapse in the individual. Not all individuals exposed to tobacco carcinogens develop cancer. This suggests that genetic and other biological factors have a modifying effect on tobacco-induced carcinogenesis. By addressing the behavioral endpoints identified in this model-tobacco use, addiction, cessation, and relapse-the NCI's research agenda has the potential to dramatically reduce tobacco-related cancers. The tobacco research agenda must be based on recognition of the complexity of tobacco use if we are to succeed in prevention and intervention targeting to children, adolescents, and adults. ADDICTION AND TOBACCO-RELATED CANCERS Within the core set of recommendations, the TRIG identified nine unique, overarching research opportunities as the highest priorities, requiring immediate implementation. These opportunities cover the range of tobacco control research from basic biological and basic biobehavioral research to clinical intervention, policy, epidemiology, surveillance research, and support for research infrastructure. The research priorities also emphasize the unique opportunities and challenges of tobacco initiation, use, addiction, and cessation among youth and populations at disproportionate risk. The TRIG emphasized that transdisciplinary approaches and the formation of strategic partnerships in the implementation of this research agenda are critical for success. Moreover, the NCI must collaborate with partners in both the public and private sectors, such as NIDA, CDC, the American Cancer Society, and the Robert Wood Johnson Foundation. For example, NIDA has supported research on the fundamental neurosicence, biobehavioral, and genetic factors underlying addiction to drugs, including nicotine. This report will highlight where particularly fruitful collaborative activities with NIDA and other organizations can be initiated to enhance the NCI's support of research on nicotine addiction. Each section of this report focuses on specific components of the Biobehavioral Model of Nicotine Addiction and Tobacco-Related Cancers, identifies outstanding opportunities and priorities in each area of research, and discusses the potential short- and long-term impact on tobacco-induced cancer that can be realized by pursuing this research.