Various historic factors influence the prior probability of breast cancer. Certain risk factors that have received a great deal of publicity (hormone use, alcohol consumption, obesity, and nulliparity) present a fairly modest relative risk (RR) for breast cancer (RR < 2). A history of breast cancer in a first-degree relative confers an RR of 2.1 or more, and a positive test for a breast cancer susceptibility gene (BRCA), an RR greater than 4. Women with BRCA1 mutations are estimated to have a 65% risk of developing breast cancer by age 70 (American Cancer Society, 2008). Patients with goiters owing to iodine deficiency have a three times greater incidence of breast cancer. For this and other reasons, “subclinical” iodine deficiency manifested by a urinary iodine concentration less than 50 µg per L is thought by some to be a risk factor both for breast cancer and fibrocystic disease (Miller, 2006).

The effect of ionizing radiation depends upon the dose. High doses are harmful, but low doses appear to be beneficial through nonspecific stimulation of biologic repair mechanisms (hormesis). In the Canadian fluoroscopy study, the RR of breast cancer at 10 to 19 cGy (1 cGy = 1 rad) cumulative exposure was 0.66 compared with controls; the RR was 0.85 at 20 to 29 cGy, and it was not significantly higher at 30 to 69 cGy. Beginning between 1930 and 1952 and being followed for up to 50 years, this study involved 31,710 Canadian women examined and treated for tuberculosis with X-ray doses to the chest (Miller et al., 1989). Mammography routines in the 1990s were claimed to deliver as little as 0.2 cGy, apparently for each pair of views of each breast, totaling 0.4 cGy per total examination. Thus, if a woman of 50 began in 1990 to have annual mammograms until age 75, the cumulative dose would be 10 cGy, which is within the optimum hormetic dose range. Therefore, avoiding mammography because of the radiation is not justified (Kauffman, 2003).

Early childbearing has a protective effect. One study estimated that “women having their first child when aged under 18 years have only about one third the breast cancer risk of those whose first birth is delayed until the age of 35 years or more” (MacMahon et al., 1970).

Breast cancer is known to be hormonally sensitive, yet the role of various risk factors involving estrogens—hormonal contraceptives and estrogen-replacement therapy—has been intensely controversial. Evaluation of these factors is confounded by the undisputed enhanced risk of early menarche and late menopause and the likely prolonged use of hormonal contraceptives before the first full-term pregnancy (Fentiman, 2002). In one major textbook, the RRs are listed in the text, but not in a table, as being 1.24 for oral contraceptives and 1.26 for hormone replacement therapy (Bland and Copeland, 2004).

Oral contraceptives have been classified as a class I carcinogen (Cogliano et al., 2005), but conflicting information has been published on the National Cancer Institute’s website (Lanfranchi, 2008).

It is noteworthy that breast cancer incidence (i.e., the incidence of cancers that had grown large enough to be detectable) diminished by 3.5% per year between 2001 and 2004, particularly in women aged 50 to 69; this has been attributed to the rapid decrease in the
use of hormone replacement therapy beginning in 2002 (American Cancer Society, 2008).

The hormones of pregnancy have a far greater proliferative effect on the breast than exogenous hormone treatment. As the first pregnancy comes to term, however, the milk glands mature, accounting for the protective effect. If this process is interrupted before 32 weeks, by premature delivery or abortion, the woman is left with a large increase in the number of immature cells that are vulnerable to carcinogenic factors. Spontaneous abortion in the first trimester appears to have no significant effect on breast cancer incidence. Women often report that they “did not feel pregnant”; the pregnancy fails owing to inability of the fetal-placental unit to produce enough hormones to sustain the pregnancy or to stimulate growth in the mother’s breasts (Lanfranchi, 2008).

At least 38 epidemiologic studies exploring an independent link between induced abortion and breast cancer have been published, 29 of which report risk elevations. There have been two metaanalyses coming to conflicting conclusions (Brind et al., 1996; Collaborative Group on Hormonal Factors in Breast Cancer, 2004).

Brind et al. estimated that in 1996 an excess 5,000 cases of breast cancer were attributable to abortion and that the annual excess would increase by 500 cases each year. Reviewing a 2001 report, Brind stated that abortion could explain the entire rise in breast cancer since the mid-1980s and that the absolute numbers of increased cases fall within the range of the numbers predicted in the 1996 metaanalysis (Malec, 2003). The overall odds ratio was found in that analysis to be 1.3 (95% CI 1.2 to 1.4) for any induced abortion (Brind et al., 1996).

The increasing incidence of breast cancer in eight European countries since 1971 was found to correlate best with induced abortion, of the seven reproductive risk factors. Fertility was also a useful predictor. Using data up to 1997, the model gave accurate predictions through 2004 (Carroll, 2007).

Although an increased risk on the order of 30% is relatively modest in an epidemiologic sense, the absolute number of cases in the population would be large because breast cancer is so common. Moreover, the risk in certain subgroups appears to be quite significant. In one study, women without positive family histories who had obtained abortions before age 18 and gestation between 9 and 24 weeks had an RR of 9.0 (Daling et al., 1994). The study also included 12 cases with a family history of breast cancer in which the women obtained abortions before age 18. No controls free of breast cancer in the study had this history. All of the cases developed breast cancer before age 45. For this group, the study reported an RR of infinity.

Currently, many influential groups deny any possible connection between induced abortions and breast cancer. For some women, however, the existing evidence could influence their decisions, and some courts have ruled that the potential risk should be included in obtaining informed consent. Additionally, physicians at some point could be held liable if they are not sufficiently vigilant in monitoring patients who may be at risk or in advising patients of the risk. Aside from an independent effect of abortion on breast cancer risk, abortion in a nulliparous woman certainly increases the duration of the “susceptibility window”—the length of time between puberty and the first full-term pregnancy— during which the breasts are especially vulnerable to carcinogens.

In summary, the history of risk factors for breast cancer should include the use of hormones and a complete pregnancy history, including the patient’s age when she became pregnant and the outcome of each pregnancy, including the length of gestation.

The absence of risk factors is never decisive. Biopsies must be performed on all suspicious lesions. If the patient herself presents with a lump, ask how long it has been present, whether there is associated cyclic tenderness, and whether it waxes or wanes with the menstrual cycle. A nodule that has been unchanged for more than a year or one that shows cyclic changes is likely to be benign (Grady et al., 1988).

Adequate exposure, of course, is essential, but many women will be uncomfortable if their modesty is not protected. Gynecologists often have the patient wear a gown that opens in the front and then place a sheet on the patient’s lap. The gown can be opened to examine the breasts, then closed to cover them during the rest of the examination. The sheet can be used to cover the pubic area while the abdomen is inspected and then be raised to cover the abdomen when the genitalia are examined. From the viewpoint of the internist or general physician, it is inconvenient to have the gown open in the front for purposes of the general physical examination, especially the chest examination. (It cannot easily be raised because the patient will be sitting on it.) Therefore, some prefer to have the opening in the back. To examine the breasts and auscult the heart, the physician can either lower the gown from the shoulders (when the patient is sitting) or raise it (when the patient is supine), keeping the lower part of the body covered with a sheet. One solution is to use a short gown that opens in front, plus a sheet.

In these litigious days, it is advisable to have a female attendant present during the breast examination, especially if the physician is a man.