ACS’s Michael J. Thun on Cancer: Progress, But a Long Way to Go
Science Watch Newsletter Interview: July/August 2010 (Page 2 of 3)
Were there other areas of research that you focused on at ACS, or were you doing mostly incidence and mortality research from CPS-II?
Another focus has been tobacco research, specifically the question of whether cigarettes that have reduced “tar” yield, when measured by machine smoking, are any less lethal than high-tar brands. We were surprised to notice that, despite large reductions in machine-measured tar and nicotine yields in cigarettes since the 1950s, lung cancer mortality rates in men and women kept going up.
In an analysis for a National Cancer Institute monograph published in 1997, we looked at lung cancer rates in male and female current smokers and never-smokers. We compared lung cancer risk among current smokers during the first six years of CPS-I, 1959 to 1965, and to the risk among smokers during the first six years of CPS-II, 1982 to 1988. Over this 20- to 25-year period, the age-standardized death from lung cancer doubled in male smokers and increased fivefold in female smokers.
The increase in women was expected, because they started smoking later than men, but the increase in men was unexpected and ran contrary to the idea that reductions in smoking and in the tar and nicotine yield in cigarettes would reduce the risk of smoking for lung cancer.
How do you account for this trend?
Well, the protocol by which smoking machines measured the tar and nicotine yields of cigarettes was developed by the tobacco industry in the 1930s. In the late 1960s, the Federal Trade Commission adopted this machine-measured approach as the basis for the labeling of tar and nicotine yields on cigarettes.
"...the utility of aspirin in prophylaxis for cancer is obviously a function of whether the benefits exceed the risk. With aspirin, the principal risk is bleeding..
What the public-health community did not realize was that some of the design changes introduced by the industry to lower the machine-measured yield of cigarettes could be more easily circumvented by smokers than other changes. For example, the cellulose acetate filters that were added to cigarettes beginning in the early 1950s could not be circumvented without breaking the filter off.
However, this was followed by a series of other changes that diluted the smoke—tiny holes in the filter, for instance—that were easier for smokers to counteract. Whereas the machine took a fixed number of puffs of fixed volume at fixed intervals, smokers could cover up the ventilation holes in the filters with their lips or fingers, take more frequent and deeper puffs, and retain smoke longer in their lungs.
Thus, smokers could extract their accustomed nicotine and tar yield out of supposedly lower-yield cigarettes. Basically these developments fooled the machine, but not the smoker. With Jeffrey Harris of MIT, we reported in the British Medical Journal in 2004 that people who smoked regular tar cigarettes did have a lower risk of lung cancer than people who smoked unfiltered cigarettes, but their risk was no different from people who smoked light or ultralight cigarettes (J.E. Harris, et al., 328[7431]: 72-6, 2004). The lowest rates, of course, were still in never-smokers.
At about the same time, biomarker studies, looking at saliva nicotine compared to the machine-measured yields, found gigantic scatter. In other words, smokers could vary enormously in the amount of nicotine they extracted from the cigarette. So the whole field of compensation got a lot of attention, and the Federal Trade Commission withdrew their approval of machine-measured smoking as a reliable index of tar and nicotine yields. When the Food and Drug Administration was granted regulatory authority over cigarettes, they banned the use of “light” and “ultralight” and other, similar terms to describe these cigarettes.
Did this explain all the increase in lung cancer rates observed?
It certainly accounted for why the expected decrease did not occur. Other factors that may also have affected the pathogenecity of cigarettes have also been identified. The addition of filters and introduction of holes to dilute the smoke motivated smokers to inhale more deeply. This vastly increased the surface area of respiratory epithelium exposed to the smoke.
The industry also introduced blended tobacco, which was a way to avoid having to throw out the stems of the tobacco. By doing that they increased the nitrosamine content of the cigarettes, because the stems release much higher concentrations of nitrosamines than the leaves when they burn. Nitrosamines are known to cause adenocarcinoma of the lung in rodents, even when injected peripherally.
The result was that the incidence of squamous-cell carcinoma and small-cell carcinoma, which occur centrally in the lung and were most strongly associated with the older types of cigarettes, decreased as expected, but the incidence of adenocarcinoma, which occurs peripherally in the lung, was increasing and basically canceling out the other favorable trends.
What do you consider the major lessons from the trends in cancer mortality and incidence over the past two decades?
The recent trends are favorable for reasons that have at least as much to do with prevention as with improvements in treatment. In men, from 1991 to 2006, there was 20.8% decrease in the all-cancer death rate. In women, we had a 12.4% decrease. Forty percent of this decrease in men in reflects the decrease in the lung cancer death rate due to the decrease in smoking.
Per-capita cigarette sales in the U.S. have decreased by almost two-thirds since the Surgeon General’s Report on smoking in 1964. This represents the greatest public health achievement with respect to chronic disease in the last 60 years. But we still have such a very long way to go.
What do you now consider the major environmental causes of cancer other than cigarettes?
Back in 1982, Richard Doll and Richard Peto published estimates on this that have turned to be remarkably robust. They said 30% of cancers were from tobacco and another 30%, give or take, were related to nutrition. We still believe that to be the case.
Recently it’s the obesity epidemic that’s delaying progress in preventing half a dozen common cancers: post-menopausal breast cancer, colorectal cancer—more in men than in women—pancreatic cancer, kidney cancer, and endometrial cancer. Then globally, over the past 10 or 15 years, we’ve recognized the importance of certain chronic infections to cancer: human papillomavirus, hepatitis B and C virus, and Helicobacter, the major cause of stomach cancer.
Page 1 ¦ 2 ¦ 3 [Highly Cited Papers by Michael J. Thun on page 3]
Citing URL: http://sciencewatch.com/inter/aut/2010/10-jul/10junThun/