Epidemiology, risk factors and prevention strategies

Age and sex

Breast cancer is very uncommon in men, being about 1% of the female rate. Age is the strongest risk factor for breast cancer in women. The incidence of breast cancer increases with age ( Fig. 5.3 ), being uncommon in younger women (in 2015–2017 annual rate ∼0.1/thousand aged 25–29 years) and much more prevalent in older women (e.g. annual rate 4.2/1000 aged 65–69 years). In the United Kingdom, almost half (48% in 2015–2017) of all invasive breast cancers are diagnosed in women aged 65 years or older.

Geographical variation

Breast cancer rates are broadly similar around the world in premenopausal women, but there are striking differences in postmenopausal rates between regions. Particularly, rates in Caucasian populations in North America, Western Europe and Australia are substantially higher than for most other regions ( Fig. 5.1 ). The GLOBOCAN study estimates that cumulative average risk to age 84 years is approximately 10% in Northern America, Western Europe and Australia and New Zealand, compared with slightly more than 6% in central and eastern Europe and South America and 4.6% in Eastern Asia. These differences between regions were more pronounced in the past, and for many countries are diminishing fast. This is likely to mostly relate to environmental or lifestyle factors. Studies of migrants from Asian countries to the United States show that the rates of breast cancer assume the rate in the host country within one or two generations ( Fig. 5.4 ). These observations, and data from twin studies, show that environmental and lifestyle factors are of greater importance than inherited genetic factors for breast cancer incidence.

Age at menarche and menopause

Women who are exposed to endogenous oestrogen over a longer time period due to early menarche and late menopause have an increased risk of developing breast cancer. Breast cancer risk increases by approximately 5% for each year younger at age of menarche. Premenopausal women aged 45–54 years have a ∼40% greater risk of breast cancer than postmenopausal women of the same age ( Table 5.1 ).

Age at first pregnancy

Nulliparity and late age at first birth both increase the lifetime incidence of breast cancer ( Table 5.1 ). The risk of breast cancer in a woman who is nulliparous or who has had her first child after the age of 30 is about twice that of a woman who has had her first child before the age of 20. The highest-risk groups are those who have a first child after the age of 35; these women appear to be at even higher risk than nulliparous women. An early age at birth of a second child further reduces the risk of breast cancer.

Breast density

Breast density is one of the most important risk factors, with the largest population attributable risk of any known risk factor for breast cancer ( Table 5.1 ). Although density decreases with age, the increased risk of high breast density is apparent in both pre- and postmenopausal women. Studies have shown that density is reduced with use of tamoxifen, but increased in users of hormone replacement therapy (HRT), and there is some (albeit limited) evidence that it might be a mediator of changes in risk associated with these interventions. There is also some evidence of a reduction in breast density associated with aromatase inhibitors (AIs), but overall less evidence than for tamoxifen. One reason is that it is more difficult to observe changes in breast density in postmenopausal women than premenopausal women because on average they have less mammographic density, and AIs are only generally recommended for postmenopausal women. Breast density is substantially higher in women with atypical hyperplasia. Over the past few years, the techniques to measure breast density have changed and are now more objective and quantitative. A widely used density classification is the Breast Imaging Reporting and Data Systems (BIRADS), which categorises breast density into four distinct risk categories. This qualitative system was not developed to quantify risk, but to indicate density of the tissue. Other techniques to measure breast density include two-dimensional (area-based methods) and three-dimensional (volumetric density) methods.

Previous benign breast disease

Benign breast disease is an important risk factor for breast cancer. Benign breast disease in the absence of epithelial hyperplasia does not carry any increased risk. Simple hyperplasia increases risk by roughly 60%, whereas atypical hyperplasia increases the risk of developing breast cancer about 4-fold. ^, Atypical hyperplasia has been shown to be an intermediate endpoint in breast cancer progression. When atypical hyperplasia is present it is likely that risk is not further modified by family history, so the risk factors are not independent. The age at diagnosis of benign breast disease appears to modify the risk. Women aged 45 years or younger with atypical hyperplasia appear to have a greater relative risk of developing breast cancer compared to women aged 55 years or older. Women with, complex fibroadenomas, duct papillomas, sclerosing adenosis and moderate or florid epithelial hyperplasia have a slightly higher risk of breast cancer (also approximately 60% increase) than the general population. However, the increased risk for these types of benign breast diseases has only been confirmed when such lesions are excised and examined pathologicalically.

Weight and height

A 10% increase in body mass index (BMI) is associated with a 40% increase in the risk of breast cancer in postmenopausal women, ^, whereas among premenopausal women it is associated with a slightly reduced incidence. Furthermore, studies have shown that the risk is higher if the extra fat is around the waist of the woman’s body. Studies also suggest that weight gain between the ages of 20 and 40 increases cancer risk, independently of current BMI.

A woman’s height has been associated with a small but consistently observed increased breast cancer risk (relative risk increases by 17% for every 10 cm in height), independently of BMI. Taller women have an increase in risk of both premenopausal and postmenopausal breast cancer compared to shorter women. It is not clear how height might affect breast cancer risk but it is believed that interactions of genetics, nutrition and hormonal levels play an important role. One possible explanation is that the hormones that affect a woman’s height may also cause an increase in the amount of ductal tissue in the breast. Most breast tumours arise from this tissue so more breast ducts would lead to increased susceptibility to breast cancer.

Physical activity

Numerous observational studies have shown that moderate physical activity is associated with a modest reduced risk of developing breast cancer, but the achieved risk reduction has varied widely in different studies. ^, Although a lifetime of regular, moderate physical activity is thought to be of greatest benefit, women who increase their exercise regime after menopause might also experience a reduced risk compared to inactive women. In general, most studies suggest that 30–60 minutes per day of moderate to rigorous physical activity is associated with a reduction in breast cancer risk.

Diet and alcohol

Although there is a correlation between the incidence of breast cancer and dietary fat intake at the population level, as well as in observational studies, it has been difficult to design and run randomised intervention studies to show this at an individual level. This is partly because studies need to be very large in order to show an effect and lifestyle and dietary changes are notoriously difficult to maintain. The best evidence demonstrating the impact of low-fat diet on breast cancer risk was the Women’s Health Initiative (WHI) trial, which randomised nearly 50 000 women. This was widely reported as a negative result, but the findings did actually indicate a reduction in breast cancer incidence of approximately 9% (hazard ratio, 0.91; 95% confidence interval [CI], 0.83–1.01) after 8 years of follow-up. Subsequently, results from long-term follow-up showed an approximately 20% reduction in risk of breast cancer mortality after 19 years in the intervention arm ( P = 0.02). Overall, a healthy, balanced diet and maintaining a healthy weight are believed to lower the risk of breast cancer. Studies report a consistent but small positive relationship with alcohol intake. The risk increases by about 7% for one drink per day (10 g) and it appears that it is unrelated to the type of alcohol (beer, wine, spirits) ( Table 5.1 ).

Breastfeeding

An overview of epidemiological studies found that the relative risk of breast cancer decreased by ∼4.3% for every 12 months of regular breastfeeding. In general, it is believed that the longer a woman breastfeeds the more she is protected from developing breast cancer. This is a small effect for most parts of the developed world, but can be substantial in less industrialised countries, where women tend to have more children and breastfeed each child for longer.

Exogenous hormones

The Million Women Study (MWS) cohort and the WHI ^, randomised trial both reported that breast cancer risk was increased for current users of combined HRT preparations compared to those on oestrogen-only preparations ( Table 5.1 , Fig. 5.5 ). Following the report of the WHI study, a steep decrease in HRT use and in incidence for women aged 50 or older has been observed in the USA, and similar trends have been seen in other countries.

Data from the WHI study furthermore suggested that the breast cancers diagnosed in women on HRT were larger and more likely to be node-positive, possibly because HRT makes them harder to visualise on mammograms. An individual participant meta-analysis of more than 100 000 women who developed breast cancer reported an increased risk of breast cancer in users of all types of menopausal hormone therapy (MHT), being stronger for oestrogen-receptor-positive than negative tumours. Risk increased with longer use and was attenuated by obesity. After ending MHT some excess risk of breast cancer persisted, depending on length of use (little risk for <1 year use).

Family history

A woman’s risk of developing breast cancer is increased if she has a family history of the disease even when specific genes cannot be identified. The average relative risk of developing breast cancer is estimated to be approximately ∼1.8 if a first-degree relative (mother, sister, or daughter) has had breast cancer and ∼2.9 for two relatives ( Table 5.1 ). The relative risks for a woman are higher when the relative’s cancer occurred at a young age and when the woman herself is young. For example, a woman whose sister developed breast cancer aged 30–39 has a cumulative risk of approximately 10% of developing the disease herself by age 65, but that risk is about 5% if the sister was aged 50 years or older at diagnosis.

Eight of nine women who develop breast cancer do not have an affected first-degree relative.

Data from twin studies estimate that approximately a quarter of variation in risk is explained by heritable factors. It is estimated that 15–20% of the heritable risk is accounted for by mutations in BRCA1 and BRCA2 , 14% by common low-risk alleles (single nucleotide polymorphism [SNPs]) and ∼5% by uncommon genes such as CHEK2 , ATM and PALB2 . The BRIDGES study based on over 60 000 cases and >50 000 controls has confirmed that several other moderate risk genes including BARD1 , RAD51C and RAD51D are associated with an increased risk of oestrogen receptor negative (ER−) breast cancer. Breast cancer susceptibility is generally inherited as an autosomal dominant trait with varying penetrance. This means that it can be transmitted through either parent and that some family members may transmit the abnormal gene without developing cancer themselves. Recent genome-wide association studies (GWASs) have now discovered several hundred common, but low-penetrance, SNPs.

BRCA1 and 2 genes

BRCA1 and BRCA2 ( BRCA1 – breast cancer 1 and BRCA2 – breast cancer 2 genes) are estimated to account for over 80% of highly penetrant inherited breast cancer (population frequency of approximately 0.1–0.3%; Table 5.2 ). It appears that the vast majority of families with highly penetrant breast and ovarian cancer are linked to BRCA1 or BRCA2. Recent analysis based on patients from genetics clinics suggests that average cumulative lifetime risk to age 85 years is ∼72% for breast and ∼44% for ovarian cancer in BRCA1 and ∼69% and ∼17%, respectively, for BRCA2 . There has been some controversy over the true lifetime risk associated with mutations in BRCA1/2 , however, with some population-based studies apparently showing lower risks as low as 30–40%, but more large-scale studies similar to the levels quoted earlier. This may be due to selection criteria used in different studies; there is evidence that family history also modifies risk in BRCA1 / 2 carriers, where women with more than first- or second-degree affected relatives having twice the risk (hazard) of breast cancer than those with no family history. BRCA1-related breast tumours are more frequently grade 3 and ER− or triple negative and can also have an atypical medullary-like histology.

Other genes

The TP53 gene is also known to predispose to early breast cancer, and it is possible to look for germline mutations, which account for over 70% of cases of the Li Fraumeni syndrome. The risk of breast cancer in women aged 30 years or younger is higher than for BRCA1 and mutation carriers also have a very substantially increased risk of brain tumours, sarcomas and other malignancies. The overall impact of Li Fraumeni syndrome on breast cancer incidence is, however, small due to its rarity ( Table 5.2 ).

Carriers of mutations in ataxia telangiectasia ( ATM ), and the CHEK2 c.1100del-C and other truncating mutations are thought to be at a 2–3-fold risk of breast cancer. ^, These moderate-risk gene mutations are the most common gene mutations associated with breast cancer, but are still relatively rare genes (population frequency 0.3–0.6%; Table 5.2 ). PALB2 has a 4–7-fold increase that is likely to decrease with age and is rarer (population frequency 0.1–0.2%). CDH1 , a gene encoding for the E-cadherin adhesion protein, was initially known as the main susceptibility gene for gastric cancer of the diffuse type. CDH1 is of special interest as mutations are always associated with invasive lobular carcinoma, but never invasive ductal carcinoma. The risk of invasive lobular cancer in mutation carriers is high and should therefore undergo intensive breast cancer screening. However, it is even more rare being present in 0.02% of controls in the study shown in Table 5.2 . Other genes associated with breast cancer include BARD1 , RAD51C , RAD51D and PTEN ( Table 5.2 ), but even combined these are rare (population frequency ∼0.2%) . Thus, a panel of BRCA1 , BRCA2 , ATM , CHEK2 and PALB2 is likely to convey most information and the BOADICEA (or CanRisk) risk-assessment model has been extended to incorporate genetic information from this panel and may be a valuable option for counselling women who have undergone genetic testing for breast cancer risk assessment.

In addition to these moderate–high-risk genes, several hundred more common SNPs have now been identified by Genome Wide Association Studies, principally from the Breast Cancer Association Consortium. Most of these have population frequencies of >5%, and individual SNPs have relative risks of order <1.1 per allele. However, they are largely independent and may be combined through a so-called polygenic risk score (PRS). Such genetic risk scores have been shown to provide much more useful information in stratifying breast cancer risk across the population, alone and in combination with other risk factors.