Many risk factors have been associated with breast cancer. These include age, race, previous benign breast disease, previous breast cancer diagnoses, lifestyle habits, reproductive factors, family history and genetic factors, exposure to radiation and environmental factors.
There is now good evidence for the involvement of female sex hormones in the incidence of breast cancer. The risk appears to be related to prolonged cumulative exposure of the milk duct cells to unopposed oestrogens. The rate of cancer rises up to the age of 45 to 50, following which there is a less gradual rise up to the age of 75.
There is a striking variation in incidence in different countries across the world which cannot be explained by genetic make-up alone. For example, there is a striking difference between the incidences of breast cancer in the USA (high) versus Japan (low). This correlates with differences in age at the beginning of menstruation and post-menopausal weight, the heavier post-menopausal women having higher serum levels of oestrogen circulating.
Benign breast disease can marginally increase the risk for breast cancer. Examples of these conditions include atypical lobular hyperplasia, atypical ductal hyperplasia, fibroadenoma, intra-ductal papillomas.
A history of breast cancer in one breast can pre-dispose to development of cancer in the other breast; this can be high for lobular cancer.
Lifestyle factors which may influence a patient’s reproductive history can impact on the development of the breast cancer. An early menarche, late first pregnancy, low parity and a late menopause all increase the risk of breast cancer in an individual woman.
A positive family history of breast cancer, particularly several first degree relatives developing the disease at an early age suggest the genetic origin, and this may account for perhaps up to 10% of all cases. Specific genetic mutations that pre-dispose to breast cancer include BRCA1, BRCA2, P53, ATM and PTEN.
Hormone replacement therapy has recently been the subject of a number of large studies. Long term use of hormone replacement therapy is associated with an increased risk of breast cancer when women were taking a combined preparation for up to five years. This has led to recommendations for more vigilant administration of hormone replacement therapy.
Patients previously exposed to ionising radiation such as survivors of the atom bomb or those patients treated for a condition called Hodgkin’s lymphoma are at increased risk of breast cancer. There is no known link between breast cancer and low levels of radiation such as those used in imaging tests unless there is a possible genetic susceptibility to breast cancer.
One of the most interesting epidemiological features of breast cancer is the striking variation in incidence in different countries across the world. There is a five fold difference in incidence between the incidence in some western countries and low risk areas of the world (see main figure in the Incidence section). These differences cannot be entirely explained by differences in the genetic make up of the populations. For example, there is a striking difference between the incidences of breast cancer in the USA (high) versus Japan (low). This correlates with the differences in age at menarche and postmenopausal weight (heavier post menopausal women having higher serum levels of oestrogens circulating).
Where there is a strong family history in first degree relatives, there are now available tests for the oncogenes that have been found to predispose to this disease. The common ones to be responsible such familial breast cancer are the BRCA-1/ BRCA-2 and the p53 genes and the routine testing for these genes in families with such strong incidences is something for which clinical geneticist counselling is now available.
Amongst patients with a strong family history of breast cancer, there have been several studies looking at the chances of a specific gene being at fault: The Breast Cancer Linkage Consortium recently published a study of 237 families with at least four cases of breast cancer within the family. They found that 52% of families were harbouring the BRCA 1 gene, 32% the BRCA 2 gene and 16% were due to some other genetic predisposition. Almost all families with breast and ovarian cancer represented within the family were associated with one of the BRCA genes (81% BRCA 1 and 14% BRCA 2 and only 5% unidentified). In families with one or more cases of male breast cancer (usually only 1% of the overall cases of breast cancer) there was a greater than 75% chance of the predisposing gene being BRCA 2 gene, which also seemed to be more likely to cause very large percentage of the risk (i.e. to be associated with 6 or more family members to be affected).
There are dilemmas for patients who undergo such testing: On the one hand the patient will know, if she tests positive, that she falls into a category of patient who is at very high risk of developing a breast cancer or ovarian cancer, or, if she has already had one breast cancer, that she is at an 85% risk of developing another, and the whole issue of bilateral prophylactic/insurance mastectomies (or even prophylactic removal ovaries) then comes into discussion.
Next, patients have to consider the issues of having subsequent life insurance premiums loaded against them or mortgages etc – because for sure they will be required to disclose these results in the application forms in the future.
Lastly, they have to consider that all these points will apply to their children. The medical discovery of genetic predisposition genes (oncogenes) and the knowledge by patients of their possession within their genetic code and the whole question of testing is a highly controversial and important one in breast cancer, and many other cancers at present.
Statistical models based on the above data have been researched to estimate a patient's chance of having the BRCA 1 or 2 mutation based on their family and personal medical histories. Computer modelling estimates probabilities of breast cancer risk and one of these is now available via the internet: www.isds.duke.edu/~gp/brcapro.html. Whilst all such modelling contains flaws (biases due to selection) they give an overall band of risk category for patients which can then be further discussed with clinical geneticists if appropriate.