Invited Commentary: Smokeless Tobacco—An Important Contributor to Cancer, but More Work Is Needed

Abstract

In this issue of the Journal, Wyss et al. (Am J Epidemiol. 2016;184(10):703–716) describe the association between use of smokeless tobacco and head and neck cancer in 11 US case-control studies. Despite use by an estimated 300 million people worldwide and prior evidence for a causal association with cancer, these products remain understudied. Data are particularly needed for persons who do not use cigarettes or other smoking tobacco products. With 6,772 cancer cases and 8,375 controls, the current study is larger than previous efforts, allowing evaluation of associations among never cigarette smokers. Importantly, snuff use was positively associated with head and neck cancer, particularly for cancers of the oral cavity, whereas associations were weaker for chewing tobacco. Associations were observed among never cigarette smokers but not among ever cigarette smokers. Yet, despite the large sample size, only 44 cases and 62 controls had used snuff and only 61 cases and 96 controls had used chewing tobacco in the absence of cigarettes, precluding detailed examinations of dose response and cessation. Future studies set in high-prevalence populations with detailed assessment of lifetime use are needed to better understand the cancer risks of exclusive smokeless tobacco use and dual use of smokeless tobacco with other tobacco products, including cigarettes.

An estimated 300 million people worldwide use smokeless tobacco products (1), with nearly 90% of use occurring in Southeast Asia. In the United States, approximately 2.6% of adults (4.8% of men and 0.3% of women) currently use smokeless tobacco products (2). However, substantial geographic variation is observed, with the prevalence of current use ranging from 1.5% in the District of Columbia and Massachusetts to 9.4% in West Virginia (3). Whereas the adult prevalence of cigarette smoking has declined in the United States over time (4), the prevalence of smokeless tobacco use among US adults has been largely stable over the last 10 years (5). Smokeless tobacco use is also common among youth, with recent data indicating that 9.9% of male high school students and 1.2% of female high school students had used it in the past 30 days, compared with 10.6% of male students and 7.9% of female students who smoked cigarettes (6).

Smokeless tobacco contains nicotine as well more than 30 carcinogens, such as tobacco-specific nitrosamines and polycyclic aromatic hydrocarbons (1, 7). The International Agency for Research on Cancer has determined that smokeless tobacco products are carcinogenic to humans and that such products are causally related to cancers of the oral cavity, esophagus, and pancreas (8). Nevertheless, far less information is available on the health effects of smokeless tobacco than is available for cigarette smoking. Perhaps the greatest amount of data is available for cancers of the oral cavity and overall head and neck. Yet even here, there are many remaining research questions. Smokeless tobacco products also differ substantially around the world, varying in chemical composition and in the way that they are used (1), necessitating a range of studies that carefully assess associations with different types of tobacco products. Within the United States and elsewhere, previous studies have often been limited by small sample sizes and have lacked comprehensive assessment of lifetime use patterns. Many smokeless tobacco users in these studies also used cigarettes, which has made it challenging to define the risks of exclusive smokeless tobacco use.

Within this context, the paper by Wyss et al. (9) in this issue of the Journal provides important information on the association of smokeless tobacco products with head and neck cancer. The study is set within the International Head and Neck Cancer Epidemiology Consortium and includes data from 11 US case-control studies, representing 6,772 cancer cases and 8,375 controls. The large sample size of this study provides numerous advantages, including the possibility of investigating associations among never cigarette smokers and within subsites of head and neck cancer. Importantly, among never cigarette smokers, ever users of snuff had higher risks of head and neck cancer (odds ratio (OR) = 1.71, 95% confidence interval (CI): 1.08, 2.70), particularly for oral cavity cancer (OR = 3.01, 95% CI: 1.63, 5.55) (9). Associations for ever use of chewing tobacco among never cigarette smokers were more modest (OR = 1.20, 95% CI: 0.81, 1.77), although again stronger for oral cavity cancer (OR = 1.81, 95% CI: 1.04, 3.17) (9). Such findings make an important contribution to the literature regarding these common but understudied products.

The study was well conducted, and the authors thoroughly harmonized available data. Nevertheless, there are limitations to the analyses as well. Despite the large sample size, among never cigarette smokers only 44 cases and 62 controls had used snuff and only 61 cases and 96 controls had used chewing tobacco (9). Even fewer smokeless tobacco users were available when the authors restricted their analysis to never users of any other tobacco product. These limited numbers precluded robust assessments of dose response, although the authors do provide useful information about associations with duration and frequency per day. Other limitations include an inability to examine current versus former use or years of cessation or to examine a potential dose response among users of either type of smokeless tobacco product, which would have increased the study's statistical power.

Carcinogen content and use patterns vary substantially across types of smokeless tobacco products (1), such that determining the risks for different types of smokeless tobacco products is of substantial public health interest. Although results of the current study suggest stronger associations for snuff than for chewing tobacco, misclassification may have affected these results. For example, the reported prevalence of chewing tobacco was higher than that reported for snuff in 1999–2012 data from the National Health and Nutrition Examination Survey (7), even though market share data indicate that most smokeless tobacco purchased from convenience stores during this time was snuff (10). Future studies could potentially overcome this limitation through detailed exposure assessment, including asking participants to identify the brand of smoking tobacco that they used, and through use of pictures. However, such assessments were unavailable in the studies participating in the consortium.

A substantial number of tobacco users in the United States and elsewhere use more than 1 type of tobacco product (3). Wyss et al. observed associations with smokeless tobacco products among never cigarette smokers but not among ever cigarette smokers (9). These results for dual use of smokeless tobacco and cigarettes should be interpreted cautiously, as participating studies lacked comprehensive assessment of both cigarette use and smokeless tobacco use over the life course and thus provided only limited ability to examine lifetime patterns of exposure to smokeless tobacco and cigarettes among exclusive and dual users.

In summary, the current study makes a valuable contribution to our understanding of the role of smokeless tobacco products in head and neck cancer risk. However, there remain numerous questions that can only be answered with future studies. As Wyss et al. have suggested (9), large studies conducted among groups with higher levels of smokeless tobacco use, such as residents of particular states (3) or members of particular occupations (11), would contribute greatly to our understanding of the disease risks of these products. Such studies should include detailed assessments of the use of smokeless tobacco, cigarettes, and other tobacco products over the life course, as well as collection of urine, blood, and other biological samples that could aid in assessment of biomarkers of carcinogen exposure. Conducting studies in this way would increase our understanding of the health effects of different types of smokeless tobacco products, alone and in combination with other forms of tobacco, as well as better assess whether there is a dose response with use and cessation. Such information is critical to our understanding of the health effects of smokeless and other tobacco products and (with the advent of the US Food and Drug Administration's Center for Tobacco Products) to tobacco regulation.