Authors:
Band, PR., Le, ND., Fang, R., Deschamps, M.

Title:
Carcinogenic and endocrine disrupting effects of cigarette smoke and risk of breast cancer.

Journal:
The Lancet, 360(9339):1044, 2002.

Summary:
Tobacco smoke contains over 4000 chemicals including those with carcinogenic, cytotoxic and mutagenic properties. On top of this, some chemicals contained in cigarette smoke are thought to exert an antiestrogenic effect. Women who smoke have been found to have an earlier age at menopause and reduced urinary concentrations of estrogens during the luteal phase of the menstrual cycle. Therefore, some authors have speculated that smoking may be protective against breast cancer through its antiestrogenic effects. Yet, so far the epidemiological findings of smoking and breast cancer risk have been inconsistent. Band et al. hypothesized that the reason for this inconsistency was because of the competing carcinogenic and antiestrogenic effects of cigarette smoke. The objective of this study was to assess both these effects on the risk of breast cancer.

Two specific time windows of smoking initiation were analyzed in order to maximize the probability of a carcinogenic and antiestrogenic effect on breast tissues. The susceptibility of breasts to carcinogens is greatest around the time of menarche. At puberty, the breasts consist mainly of primitive and developing lobules and it is not until the first full-term pregnancy that full differentiation of the breasts occurs. Breast cells that originate from poorly differentiated breast epithelial cells are more prone to neoplastic alteration than well differentiated lobules. Thus, there is an increased susceptibility to chemical carcinogens between menarche and first full-term pregnancy. As well, around the period of menarche, endogenous estrogens are high and the potential for a carcinogenic effect is the most likely. Unless the anti-estrogenic effect was marked at this time, it would have little impact on breast tissue. Conversely, the antiestrogenic effect of smoking would be maximized in women who develop post-menopausal breast cancer and start smoking after their first full-term pregnancy. Endogenous estrogens are lower after menopause and are derived primarily through the aromatization of adrenal androgens in fat tissue. Therefore, estrogen levels are directly related to obesity in post-menopausal women. At this time it is possible that a potential antiestrogenic effect would be noticeable particularly in women with a change in their body mass index.

To conduct this study the authors identified all women who were younger than 75 years of age with breast cancer diagnosed between June 1st, 1988 and June 30th, 1989, and listed on the British Columbia cancer registry. Age-matched controls were randomly selected from the 1989 British Columbia provincial voters list. Subjects were eligible if they were Canadian citizens, lived in British Columbia, and had no previous history of breast cancer for cases and not diagnosed before June 30, 1989 for controls. A questionnaire was sent to participants to gather pertinent demographic, health, reproductive, occupational, smoking and lifestyle history. Conditional logistic regression modeling was used in the analysis with age as the matching variable. Analysis was done separately for pre- and post-menopausal women. The authors assessed several other potentially important factors including: ethnic origin, marital status, education, alcohol consumption, age at menarche and menopause, family history of breast cancer in a first degree relative, history of biopsy for benign breast disease, with and BMI at age 18 to current age, use of birth control pills, use of estrogen replacement therapy, reproductive and breastfeeding history.

There were 1018 cases (68%) that returned a questionnaire who also had histological confirmation of their disease. There was a similar percentage of controls (n=1025) who returned a questionnaire. The authors obtained smoking information for some of the participants who did not return the questionnaire (98 with breast cancer and 59 controls) via telephone. The smoking status was found to not differ between those who returned the questionnaire and those who did not. Results showed that for pre-menopausal women who had at least one pregnancy, smoking initiation less than 5 years from the onset of menarche increased the odds of breast cancer (OR=1.69, CI=1.13-2.51). As well, smoking before the first pregnancy increased the odds of breast cancer (OR=1.47, CI=1.02-2.10) in this group of women. For pre-menopausal women who had never been pregnant the odds of breast cancer was increased for women who smoked 20 or more cigarettes per day (OR=7.08, CI=1.63-30.8) and women who smoked for 20 or more cigarette pack years (OR=7.48, CI=1.59-35.2). Conversely, for post-menopausal women who had been pregnant at least once, the initiation of smoking after the first full-term pregnancy was found to be protective against breast cancer (OR=0.58, CI=0.36-0.91). In addition, a significant protective effect was seen for post-menopausal women who had an increase in their BMI since the age of 18 and initiated smoking after their first full-term pregnancy (OR=0.49, CI=0.27-0.89).

This study's strengths are that it used a population-based cancer registry and controlled for several other factors that may have influenced the results. This study is important because it addresses the biological mechanisms involved in breast development and looks at critical windows of exposure to maximize the chance of discriminating between carcinogenic and anti-estrogenic effects of cigarette smoke. The exposure assessment was done indirectly through a questionnaire which may have limited the findings however, the results are suggestive that cigarette smoke may exert an anti-estrogenic effect in post-menopausal women which warrants further investigation.