There is no readily available medical literature on the definition of risk and most useful information is in industry or technology-based publications. Fischhoff and Watson (1984) outlined in their essay the controversies around defining risk. They stated that ‘risk is a focal topic in the management of many activities and technologies’. For that management to be successful, an explicit and accepted definition of the term ‘risk’ is essential. Although this essay is on risk in general with regard to day-to-day industry and technology, it is applicable to assessing the risk for breast cancer or any other disease process because they further explain that ‘the choice of definition can affect the outcome of policy debates and allocation of resources’. They also highlight the technical distinction between objective and subjective risk, the former being a product of scientific research, public health statistics, experimental studies, epidemiological surveys and probabilistic risk analysis. In contrast, subjective risk is the public perception of that research. They also highlight the effects of comparing different scientists which can affect the relative riskiness of technologies. This is clearly illustrated in this book when different authors assess the relative risk of ADH, which ranges from 4 to 13 times that of the general public depending on which publication one reads (Dupont and Page 1985; Palli et al. 1991). However, in this paper Fischhoff and Watson do not provide the reader the definition of risk but highlight various factors to consider in assessing risk for any situation, which is also applicable to breast cancer. The factors the authors evaluated include the following:

There is no definition of risk in the National Cancer Institute online dictionary of medical words. The online Oxford Dictionaries (British and World English) define risk as ‘a situation involving danger’. A better definition of risk is provided in online business dictionary which provides ‘a probability or threat of damage, injury, liability, loss or any other negative occurrence that is caused by external or internal vulnerabilities, and that may be avoided through pre-emptive action’ (Business Dictionary). The dictionary provides several examples of risks including that on cancer which is described as the consequence and probability of a hazardous event or phenomenon, i.e. the risk of developing cancer is estimated to be an incremental probability of developing cancer over a lifetime as a result of exposure to potential carcinogens.

Risk assessment for breast cancer is divided into relative risk and absolute risk. The National Cancer Institute Dictionary of Medical Terms (NCI Dictionary of Cancer Terms) defines relative risk as ‘a measure of risk of a certain event happening in one group compared to the risk of the same event happening in another group’. In cancer research, relative risk is used in prospective (forward looking) studies such as cohort studies and clinical trials. A relative risk of one means there is no difference between two groups in terms of their risk of cancer based on whether or not they are exposed to a certain substance or factor or how they respond to two treatments being compared. A relative risk of greater than one or of less than one usually means that being exposed to a certain substance or factor either increases (relative risk greater than one) or decreases (relative risk less than one) the risk of cancer or that the treatments being compared do not have the same effect, also called the ‘risk ratio’.

Most relative risk estimates are from longitudinal studies and have been derived under the assumptions that the woman’s relative risk remains constant over time. However, it is possible that the relative risk of individual patients varies with age or the time since the original biopsy. To illustrate this point, Dupont and Page (1985, 1989) demonstrated that the relative risk of developing breast cancer in women with atypical hyperplasia and proliferative disease without atypia (PDWA) is greatest during the first 10 years of follow-up. Women with PDWA who remain disease-free after 10 years will no longer be at risk of breast cancer when compared with women of a similar age group without PDWA. Similarly, the relative risk of breast cancer in women with atypical hyperplasia is halved if they remain cancer-free for 10 years following the initial biopsy. This supports the notion that atypical hyperplasia may not be an obligate precursor of breast cancer and that the lesion may progress to breast cancer, remain static or regress over a period of time. However, the absolute risk of women with atypical hyperplasia remained constant over a period of 17 years of follow-up (Dupont and Page 1985). The clinical significance of this time-dependent analysis is that, depending on other factors, it will allow women with atypical hyperplasia to be closely followed up during the first 10 years of diagnosis. If the woman has not developed cancer at 10 years or more during follow-up, she can be reassured that her risk is now closer to that of the general population than it was at the time of the diagnosis of atypical hyperplasia, although the absolute risk per year has remained the same.

The National Cancer Institute Dictionary of Cancer Terms (NCI Dictionary of Cancer Terms) defines absolute risk as ‘a measure of risk of a certain event happening’. In cancer research, it is the likelihood that a person who is free of a specific type of cancer at a given age will develop that cancer over a certain period of time. For example, a woman of 35 years of age with no known risk factors for breast cancer has an absolute risk of developing breast cancer over a lifetime of 90 years of about 13.5 % meaning one out of every seven women will develop breast cancer by the age of 90.

Absolute risk is more informative than relative risk and can be determined over time intervals of 10–20 years when the corresponding relative risk estimates have been accurately determined (Dupont and Plummer 1996). To the patient, it is more important to know that the absolute risk of developing breast cancer if one had ADH is 10 % over a period of 10–15 years than to be told the relative risk of developing cancer is four times that of the general population. Although absolute values give a more accurate assessment of risk, they require lengthy follow-ups of large numbers of patients with similar risks (Dupont and Page 1989). For this reason, patient management is usually based on estimation of relative risk.

For the probability of an event to occur, the individual must be exposed to a risk factor. This is another term which is not clearly defined in medical literature. As explained by Dr Brian Burt at the Department of Epidemiology (University of Michigan), the literature is unclear whether risk factors should be truly causal, i.e. a link in the aetiological chain or whether it is more peripherally associated with an outcome (Burt BA Definitions of risk). This statement is pertinent to benign breast lesions where it is not always clear whether the presence of a proliferation will directly metamorphose into breast cancer or it is just a marker of the likelihood that cancer will develop at some point in the woman’s lifetime.

In the dictionary of epidemiology, risk factor is defined as ‘an aspect of personal behaviour or lifestyle, an environmental exposure, or an inborn or inherited characteristic which on the basis of epidemiological evidence is known to be associated with health-related condition(s) considered important to prevent’ (Last 2001). Some examples cited as risk factors for cancer are age, a family history of certain cancers, use of tobacco products, being exposed to radiation or certain chemicals, infection with certain viruses or bacteria and certain genetic changes. This definition is wider and more inclusive when compared to the National Cancer Institute definition which defines risk factor as ‘something that increases the chance of developing a disease’.

An equally informative definition of risk factor is that provided by Dr Beck (1998) which states that ‘a risk factor is an environmental, behavioural or biological factor confirmed by temporal sequence, usually in longitudinal studies, which if present increases the probability of a disease occurring, and if absent reduces the probability’. Risk factors are part of the causal chain. Once disease occurs, removal of a risk factor may not result in a cure. Dr Beck (1998) explains that this definition is more specific as there is

The definition by Dr Beck is clearly applicable for ADH as risk factor for breast cancer in that

If some risk factors in breast cancer are part of the causal chain, this raises the question what is a risk marker? The dictionary of epidemiology (Last 2001) defines a risk marker as ‘an attribute or exposure that increases the probability of occurrence of disease or specified outcome’. This would apply to a lesion such as ALH which some consider to be a marker of rather than a causal factor of invasive breast cancer because the resultant breast cancer and can be either ductal or lobular and affect both breasts. In contrast, ADH is considered to be a causal link in the pathogenesis of breast cancer and is also a modifiable risk factor, i.e. a determinant that can be modified by intervention, thereby reducing the probability of disease (Dictionary of Epidemiology, Last 2001). There is evidence that prescribing prophylactic tamoxifen to women with ADH reduces the risk of developing cancer, however, the same applies to ALH (Chuba et al. 2005).

Assessment of risk factors for breast cancer can be divided into those individuals who have a familial predisposition because they carry a genetic mutation such as BRCA1 and BRCA2 and those without a mutation (Parmagiani et al. 1996). Five to ten percent of breast cancer occurs as a result of genetic predisposition caused by a mutation in a single cancer susceptibility gene which confers considerable risk (Vogel and Bevers 2003). Features of familial breast cancer include

BRCA1 gene and BRCA2 genes which predispose to hereditary breast and ovarian cancer account for 81 % and 14 % of hereditary breast cancer, respectively (Ford et al. 1998). Other genetic syndromes associated with familial breast cancer include Cowden syndrome, Li-Fraumeni syndrome, ataxia-telangiectasia heterozygotes and Peutz–Jeghers syndrome (Vogel and Bevers 2003). The UK NICE recently recommended the use of tamoxifen for chemoprevention in women with high risk of developing cancer due to familial predisposition (June 2013). In addition to the above risk factors the NICE Guidelines include

Hormonal changes play an important role and reduce or increase the risk of breast cancer in an individual. Early menarche (<12 years), late menopause (>55 years), nulliparity or first born child after 30 years and current use of oral contraceptives are associated with increased risk of breast cancer (Vogel and Bevers 2003). The risk is due to the high levels of circulating oestrogens.

Earlier studies published conflicting evidence as to whether hormone replacement therapy (HRT) is associated with an increased risk of breast cancer or not (Dupont et al. 1999). The presence of oestrogen and progesterone receptors in the breast tissue makes it sensitive to circulating hormonal therapy. A meta-analysis of 51 international case control and cohort studies found no appreciable increased risk of breast cancer in women on short-term oestrogen replacement therapy of less than 5 years, (Manson and Martin 2001). However, long-term use of oestrogen replacement therapy (>5 years) was associated with an increased risk of 35 %. A separate study by the Writing Group for Women’s Health Initiative Investigators, reported an increased risk of 26 % in women on combination HRT (2002).

A multicentre study involving over one million women, in which 9,364 women developed cancer and 637 women died from breast cancer, Beral and the Million Women Study Collaborators (2003) observed that current users of HRT were more likely than ‘never users’ to develop cancer (adjusted RR = 1.66, 95 % CI = 1.58−1.75, P < 0.0001) and to die from breast cancer (RR = 1.22, 95 % CI = 1.00−1.48, P = 0.05). The risk of breast cancer was increased much more with combined oestrogen and progesterone preparations than with oestrogen only and other preparations. Long duration of HRT use, up to 10 years, was also associated with an increased risk. Based on these findings, it may be clinically advantageous to assess the parenchymal density of the women before they commence HRT and offer shorter interval screening periods for those with a significant increase in parenchymal density while on HRT. This may detect women with early cancer when this is potentially curable. Selective oestrogen-receptor modulators (SERMs) such as tamoxifen and raloxifene may be an alternative to HRT for women at risk for breast cancer (Vogel et al. 2010).

There is accumulating evidence that alcohol intake is associated with increased risk of breast cancer (Vogel and Bevers 2003; Longnecker 1994; Longnecker et al. 1995; Rosenberg et al. 1993). The following mechanisms have been implicated:

Coutelle and colleagues (2004) demonstrated that oral intake of alcohol in women with alcohol dehydrogenase 1C*1 allele, which rapidly metabolises alcohol to produce increased levels of acetaldehyde have increased levels of serum estradiol. The women with this alcohol dehydrogenase genotype were 1.8 times more at risk of developing breast cancer than women without the genotype (95 % CI 1.431–2.330, p < 0.001). Thus, individuals with this genotype increase their risk of breast cancer if alcohol is consumed in excess, resulting in increased blood oestrogen concentration.

In a previous study, Wright et al. (1999) proposed that metabolism of alcohol produced reactive oxygen species which induce carcinogenic mutations and damage to DNA predisposing to breast cancer. The amount of alcohol consumed is directly proportional to the risk of breast cancer. Drinking 10 g of alcohol (one drink daily) is associated with relative risk of 1.09 when compared to non-drinkers. Increasing alcohol consumption to 30–60 g/day (2–5 drinks) increased the relative risk of breast cancer to 1.41 when compared to non-drinkers (Smith-Warner et al. 1998). Alcohol consumption in individuals with family history of breast cancer increases the RR of developing breast cancer to 2.45 (Vachon et al. 2001). The risk is similar for beer, wine and liquor (Smith-Warner et al. 1998).

Other factors associated with increased risk of breast cancer are old age and high body mass index (BMI), but there is no evidence that diet and exercise influence the occurrence of breast cancer (Vogel and Bevers 2003). As already illustrated in the previous chapters, different types of benign proliferative disease confer varying degrees of risk of subsequent breast cancer and these will be summarised in this chapter.

John Wolfe highlighted the concept of breast cancer risk based on mammographic parenchymal patterns, and most of the publications on this subject were in 1976. Age, pregnancy, thyroid hormone levels, endogenous and exogenous hormones can influence breast parenchymal appearances. Wolfe classified the appearance of the breast parenchyma into four categories irrespective of vasculature, presence or absence of masses, calcifications, parity, family history of breast cancer or age of the patients. He noted four breast parenchymal patterns, which he classified as follows:

The exact histological features of ‘dysplastic’ breast tissue are not clear from Wolfe’s original study (Wolfe 1976a). However, Wellings and Wolfe (1978) carried out a radiological and pathological correlation of tissue from the different parenchymal patterns. This included subgross analysis of the excised breast tissue. The histological features of the different parenchymal patterns were reported as follows:

When Wolfe (1976b) assessed the different radiological parenchymal patterns in relation to the risk of breast cancer, the incidence of detecting cancer in the N1, P1, P2 and DY parenchymal patterns was 0.1, 0.4, 1.7 and 2.2 %, respectively, indicating that dense breast carried the highest risk of harbouring malignancy. Based on these results, the risks of the parenchymal patterns were assessed as follows: N1 = minimal risk, P1 = moderate risk, P2 = significant risk, DY = highest risk.

In a subsequent report, Wolfe (1976c) followed up 995 women who had undergone mammography over a period of 15 years (range 5–15 years). All women had presented symptomatically (age range 30–49 years). The study reported that breasts initially classified as N1, P1 or P2 rarely changed their parenchymal density with age. In contrast, breasts classified as DY mostly regressed to P2, which is still a high-risk category, with some breasts regressing to P1 or N1. These changes occurred between 35 and 50 years. In a similar but separate study, Wolfe et al. (1987) reported that most breast cancers occurred in the P2 breast, with decreasing frequency in the DY, P1 and N1 parenchymal patterns. In this case-controlled study of 160 women who had presented symptomatically (mean age 53 years), the RR of developing breast cancer was 3.3 in women who had P2 or DY patterns when compared with women who had N1/P1 patterns. Black women had a higher risk of 5.2 when compared with white women with an RR of 2.3 for the P2 and DY parenchymal patterns. Women younger than 53 years had a lower RR of 2.8 compared with 3.8 for women older than 53 years if their mammograms exhibited the P2 and DY parenchymal patterns.

In a separate study, Saftlas and colleagues (1991) assessed parenchymal density in women undergoing screening mammography. The authors reviewed mammograms that were taken 4 years prior to diagnosis of breast cancer to determine whether increased mammographic density was predictive of breast cancer risk. The median ages of the women were 54 and 53 for the study cases (n = 266) and controls (n = 301), respectively. This study applied planimetry to assess the percentage of mammographic density and this had a better predictive value than Wolfe’s parenchymal patterns. The odds ratios (OR) of breast cancer increased steadily with increasing breast density (P < 0.0001). The breast cancer OR were 1.7 for women with 5–24.9 % parenchymal density, 2.5 for women with 25–44.9 % parenchymal density, 3.8 for women with 45–64.9 % parenchymal density and 4.3 for women with densities over 65 %. These OR were irrespective of other breast cancer risk factors such as family history of cancer or age at first parity.

Byrne et al. (2001) assessed the histological features of benign breast disease and mammographic density of the breast in a case-controlled study of 347 women who developed breast cancer and 410 women without breast cancer. Adjusting for mammographic density, the OR for subsequent cancer for atypical hyperplasia was 2.1. In contrast, adjusting for benign breast histology, the OR for increased breast density of >75 % was 3.8. Women with non-proliferative benign breast disease and increased breast density of >75 % had an OR of 5.8, whereas women with breast density of less than 50 % and atypical hyperplasia had an OR of 4.1.

Previously, Friedenreich and colleagues (2000) assessed the mammographic density of fibroglandular tissue as a risk factor for subsequent malignancy in the screening age group. They assessed women who attended the Alberta Screening Programme in Canada. The women were asymptomatic, aged between 50 and 69 years and attended for breast screening biannually. The aim of the study was to determine whether women with histologically proven benign breast disease had a radiologically identifiable increase in breast tissue density. The study compared 165 biopsy-proven benign diseases with 217 women who had no histological evidence of proliferative breast disease. The proliferative benign disease was classified into multiple papillomas, radial complex sclerosing lesion, nodular adenosis, sclerosing adenosis, usual hyperplasia (moderate or florid), ADH and ALH. The mammographic content of fibroglandular tissue was assessed as follows: (1) fatty, (2) >0 – <25 %, (3) 25–50 %, (4) 50–75 %, (5) >75 %. The women also responded to a questionnaire on demographic factors, which included smoking and dietary habits, menstrual and reproductive history, physical fitness, hormone usage and family history of benign proliferative breast disease. When all the demographic and radiological features were assessed, the authors found that women with increased fibroglandular density on mammography of >25 % had an increase in benign proliferative disease when compared with women with less than 25 % of fibroglandular tissue (OR = 1.91; 95 % CI = 1.24−2.94). Most studies that assessed benign breast disease for the potential risk of malignancy were based on symptomatic patients. The study by Friedenreich and colleagues should pave the way for prospective follow-up studies, which should combine demographics, radiological and histological features of benign proliferative disease to accurately prognosticate the risk factors in the screening age group. In a separate study, Boyd et al. (2000) reported a high prevalence of benign epithelial hyperplasias associated with increased mammographic densities of over 75 %. The benign epithelial hyperplasias were in turn risk factors for malignancy. The findings were comparable to those of Saftlas and colleagues (1991).

The association of breast cancer risk with increased breast tissue density has not been universally supported by other investigators (Mendell et al. 1977; Moskowitz et al. 1980; Arthur et al. 1990). In the UK Trial for Early Detection of Breast Cancer in Nottingham, Arthur and co-workers (1990) examined breast tissue from 119 women with a histological diagnosis of fibrocystic change with epithelial hyperplasia or in situ carcinoma. Women attending routine breast cancer screening were used as controls. The Nottingham study reported that the Wolfe patterns were related to the distribution of fibrous and adipose tissue in the breast interlobular stroma and appeared to have no relationship to the epithelial parenchymal content. Based on these findings, the authors concluded that radiological densities of P2 and DY patterns did not correspond to high-risk epithelial proliferations.

Despite the variation in the reports regarding the significance of breast tissue density as a risk factor for breast cancer, there is a growing evidence that increased breast tissue density may be a risk factor for subsequent malignancy. It is not clear whether the increased risk of malignancy associated with dense breasts is due to masking of small cancer because mammographic interpretation of dense breast is difficult, or whether mammographic density is an independent risk factor (Fig. 16.1). However, there is evidence that including mammographic density as a risk factor minimally improves the predictive accuracy of the Gail Model (Tice et al. 2006).

Sterns and Zee (2000

Stay updated, free articles. Join our Telegram channel

Join