Background: Epidemiology and Family History
Breast cancer is the most common non-cutaneous cancer in women, accounting for 20 % of all new cases of cancer. The lifetime risk of breast cancer in the UK is one in nine females, with an annual incidence of <10 per 100,000 women aged <30 years, rising to 300 per 100,000 in women aged over 85 years. Similar, but slightly higher rates are seen in North America. It is rare in men (<1 per 100,000). Breast cancer incidence shows marked geographical variation: it is much less common in Asian than in Caucasian women and less frequent in South America and Spain than in Northern Europe, North America, and Australia. In India, the prevalence is lower (although rising) among most ethnic groups. In North America, the incidence of breast cancer appears to have increased in recent years; most of this increase is due to mammographic detection of ductal carcinoma in situ leading to early diagnosis of minimally invasive ductal carcinoma. It is notable that the rates of breast cancer to age 40 are fairly stable around the globe (Narod 2012a). Nearly all the differences in incidence from country to country occur in women diagnosed above this age, and it is even more noticeable in women diagnosed in the postmenopausal years (Leong et al. 2010), suggesting that nongenetic factors predominate in women diagnosed at older ages. Because the known highly penetrant breast cancer susceptibility genes cannot explain most of the diagnoses of breast cancer in women diagnosed under 40 years of age (they are clearly highly “genetic”), other genetic mechanisms must be at play.
In 1948, Penrose and colleagues observed that breast cancer may have a hereditary basis in some families (Penrose et al. 1948). A genetic influence on breast cancer susceptibility is suggested by twin studies, as the
concordance for breast cancer in identical twins (0.28) is more than twice that of dizygotic twins (0.12). A statistically significant heritable factor of 27 % was observed for breast cancer (95 % confidence interval (CI): 4–41) in the three-country study of 44,788 pairs of twins (Lichtenstein et al. 2000). By contrast, the factor for prostate cancer was 42 %, so clearly these percentages do not reflect the contribution of highly penetrant genes to these diseases.
There is an increased risk of breast cancer among female relatives of breast cancer patients and this increased relative risk (RR) is more pronounced when the index case has bilateral disease or early (premenopausal) age at onset. For patients 55 years of age or older at diagnosis, the RR of breast cancer in a first-degree female relative is approximately 1.8, increasing to 3–4 when more than one first-degree relative is affected. A large study (7,496 women with breast cancer and 7,438 controls) of the risks of breast cancer in women in relation to their family history found that, compared to women with no affected relative, the risk
ratio for breast cancer was 1.80 (99 % CI 1.69–1.91), 2.93 (99 % CI 2.30– 3.64), and 3.90 (2.03–7.49), respectively, for one, two, and three or more affected first-degree relatives (P < 0.0001 each) (Collaborative Group on Hormonal factors in breast cancer 2001). Risk ratios were greatest at young ages, and with younger age at diagnosis in the relative. For women with zero, one, or two or more first-degree affected relatives, the estimated cumulative risk of breast cancer up to age 50 years was 1.7, 3.7, and 8 %, respectively, corresponding to incidences up to age 80 of 7.8, 13.3, and 21.1 %, and death from breast cancer of 2.3, 4.2, and 7.6 %. There was no evidence for an autosomal recessive effect. An important finding was that other factors such as hormones and diet did not vary by family history.
Among women with one affected relative, the cumulative incidence of breast cancer between ages of 20 and 80 was 12.3 % if the relative was over 60, but 16.1 % if under 40. Another key finding was that most breast cancers in women with a family history of breast cancer were likely to occur after the age of 50, even if there were two affected relatives (Collaborative Group on Hormonal factors in breast cancer 2001).
For a woman aged 30 with a mother and a sister affected with breast cancer, the cumulative risk of breast cancer to age 70 has been found to be 17.4 %, significantly lower than the 43 % risk to be expected if all such familial clusters were due to highly penetrant mutations in susceptibility genes (Peto et al. 1996), suggesting that most such familial clusters are not due to these genes.
There is epidemiological evidence for overlap between breast and ovarian cancer susceptibility. The overall age-adjusted RR for ovarian cancer in first- degree relatives of women with breast cancer has been estimated to be 1.7 (and 2.1 for breast cancer), and first-degree relatives of index cases of ovarian cancer have RR of 1.6 and 2.8 for breast cancer and ovarian cancer, respectively. In one study, the RR for breast cancer in first-degree relatives of index cases ascertained because of colon cancer was estimated to be about 5, but in general, later studies of the risk of breast cancer after colorectal cancer, or colorectal cancer after breast cancer, have not supported this finding (Newschaffer et al. 2001). The only gene that might reasonably contribute to both breast and colorectal cancer in this context is CHEK2 (Meijers-Heijboer et al. 2003), but this is probably only clinically relevant in the Netherlands and possibly Finland. It is possible that some “breast/colorectal” families are due to mutations in two genes, with one responsible for breast cancer, and the other for colorectal cancer (Thiffault et al. 2004). It should be noted that recent prospective studies have suggested an up to fourfold increase in breast cancer for patients carrying germline mismatch repair gene mutations (Win et al. 2012).