In light of a growing focus on the issue of lung cancer in never-smokers, it makes sense to try to identify potential causes in this population. Among the leading candidates as a cause of lung cancer in never-smokers is secondhand, or environmental tobacco smoke (ETS) exposure. This potential cause was identified more than 25 years ago in the never-smoking wives of heavily smoking Japanese men (Hirayama Br Med J 1981 abstract) Since then several other studies evaluating the association of ETS with lung cancer have been reported. The Environmental Protection Agency noted in its report from 1992 that there was an increased risk for lung cancer from ETS and estimated that it accounts for approximately 3000 lung cancer deaths in the US. It is fair to say, though, that the methods by which these conclusions were reached were imperfect and have been seriously questioned, particularly by smoking advocates (counter-arguments provided by a strong non-believer here), but also by others who just noted that the science behind the calculations was pretty weak.

One meta-analysis (a collection of data from multiple smaller trials to try to determine a conclusion based on the pooled data asking the same question) reviewed 19 different studies that focused on never-smoking women and estimated an approximately 20% increased risk of developing lung cancer from exposure to ETS. Another report by the International Agency for Research on cancer estimates a 25% increased risk of developing lung cancer from ETS in women, and a 35% increase in men (full review article from J Natl Cancer Inst here). A large European trial (Vineis British Medical Journal full article here) comparing patients who developed cancer with an otherwise similar matched control population included over 123,0000 never or former smokers (3/4 were women) and found a 34% increased risk of lung cancer from ETS in the overall population, but only a 5% increased risk in never-smokers. Overall, there was not a statistically significant difference in risk for developing lung cancer from ETS.

One of the biggest issues is in quantifying ETS exposure and recall bias (not being able to control what people in a study recall or don’t in terms of prior exposures; sometimes they may not even know about prior exposures, such as with radon, which I’ll discuss in the future). One of the issues that has limited our ability to study the association of smoking with cancer is that doctors don’t consistently obtain a good smoking history on patients, even those with potentially smoking-related cancers. And if we don’t reliably collect details on a patient’s direct smoking history, we certainly don’t do a good job ferreting out a history of ETS exposure. That’s part of why trials like SWOG 0424, which I discussed previously, are so important. This national observational study (collecting information at a starting point and on follow-up, but not influencing treatment) compares the genetic differences of lung cancers in men vs. women and smokers vs. never-smokers, are so important, and it includes not only molecular characteristics of tumors and in blood, but also a detailed questionnaire that asks about active and passive smoking history and also exposure to many other potentially relevant carcinogens.

Overall, the studies of ETS suggest a rather modest increased risk of lung cacner that requires large numbers of patients to clarify the statistical significance of that exposure. For the very motivated, there are several other sources on the internet that summarize the data supporting a modestly increased risk for lung cancer, other cancers, and/or other tobacco-associated medical problems available through the following links:

American Cancer Society Secondhand Smoke Summary Page

American Lung Association Secondhand Smoke Fact Sheet

NCI tobacco control monograph (including chapter on the association of ETS with cancer) -- please note, even the chapter alone is 94 pages of dense science summary, so these links are for the highly motivated and/or highly bored only.

Overall, the data suggest that ETS may play a role in development of lung cancer in never-smokers, but we certainly need to look elsewhere for contributing causes. I'll discuss more of those potential factors in coming weeks.