Smoking and cancer: value in paradox.

An epidemiologic study of endometrial cancer shows that wome who smoke have a lower incidence of this cancer of the lining of the uterus than do nonsmoking women. That doesn't mean, say the researchers, that women should take up the habit -- the specter of smoking-related diseases greatly overshadows the cancer-protective benefit. The value of the study, they say, is in prompting further investigation into what causes the protective effect to see if it can be exploited for prevention.

Samuel M. Lesko of Boston University and researchers from five other U.S. institutions looked at 510 women 30 to 69 years old hospitalized for endometrial cancer and 727 women in the same age range hospitalized with cancers unrelated to cigarette smoking. Current smokers were at 0.7 times the risk of endometrial cancer as nonsmokers, the researchers report in the Sept. 5 NEW ENGLAND JOURNAL OF MEDICINE. The more extreme the habit, the greater the "benefit" -- women currently smoking more than 25 cigarettes a day were at half the risk of women who had never smoked. The effect was seen primarily among postmenopausal women.

"The present findings do not have direct public health importance since cigarettes, overall, have serious deleterious effects," note the researchers. And in an accompanying commentary, Noel S. Weiss of the University of Washington in Seattle observes that while smoking will each year spare 30 of 100,000 women smokers from getting endometrical cancer and six from dying of it, the habit kills about 30 times that many women through other diseases.

The mystery here is how cigarettes exert a beneficial effect. The researchers and Weiss suggest that the common denominator between smoking and a reduced risk of endometrial cancer is the hormone estrogen. Other studies, they note, have shown that cigatette smoking lowers estrogen levels in the body, and high levels of estrogen have been linked to endometrial cancer.

The estrogen connection raises its onw question: How and when do estrogen levels play a role? Lesko and his colleagues point to a 1982 study led by Brian MacMahon at Harvard University showing that during one part of the menstrual cycle women smokers had lower estrogen levels than nonsmokers. But if the cigarette benefit had been mediated by its menstrual cycle effect, counters Weiss, the Boston University study would have shown a substantial risk reduction in premenopausal women. "Just about everything we know suggests that estrogen has a very rapid effect on endometrial cancer," he says. Endometrial cancer incidence goes up within a couple of years in women taking estrogen to reduce the side effects of menopause and goes down within a couple of years of stopping, he notes.

Weiss suspects a counterplay between estrogen and progesterone, another hormone involved in the female reproductive system, as a key factor in endometrial cancer. In premenopausal women progesterone balances estrogen, so the effect of smoking wouldn't be expected to play a role. After menopause, while progesterone production halts, some estrogen is still produced.