The principal lymphatic drainage of the breast (Figure 26.1) is to the axillary nodes lying between the second and third intercostal spaces. Additional drainage occurs to the supraclavicular nodes through the pectoralis major and to the internal mammary chain adjacent to the sternum.

Figure 26.1 (A) The lymph nodes of the axilla. (B) Diagram of the principal pathways of lymphatic drainage of the breast. These follow the venous drainage of the breast – to the axilla and to the internal mammary chain.

(Reproduced with permission from Ellis, Clinical Anatomy, 5th edn, Blackwells, 1975)

Pathology

Epidemiology

Worldwide, breast cancer is the commonest form of malignancy in women. It accounts for 12% of all cancers, 18% of all female cancers, 10% of all cancer deaths and 20–25% of all female cancer deaths. In England, there are 41 000 new cases per year and, in the UK, over 12 000 women die of the disease per year. The incidence is highest in developed countries and low but rising in less developed countries and in Japan. The risk of a woman developing the disease at some stage of her life is 1 in 12. The UK has the highest age standardized incidence and mortality for breast cancer in the world.

Female breast cancer is rare below the age of 30. The incidence rises rapidly during the reproductive years. After the age of 50, the incidence rises at a lower rate. The cumulative incidence in women in Europe and North America is approximately 2.7% by the age of 55, 5% by the age of 65 and 7.7% by the age of 75. Male breast cancer is rare, accounting for only 1% of breast cancer.

Encouragingly, the mortality of breast cancer is declining in the UK and USA and in much of western Europe. However, the mortality from breast cancer in central and eastern Europe and SE Asia continues to climb.

Aetiology

The aetiology of breast cancer is not fully understood, but a number of predisposing factors have been identified.

Genetic factors

High-risk mutations

It is estimated that up to 10% of breast cancers have a genetic basis. The sisters and daughters of a woman with breast cancer have a threefold increased risk of developing the disease. The risk of breast cancer for a woman whose sister is affected is doubled. Women who have a first-degree relative with premenopausal or bilateral breast cancer are at particularly high risk. At least five germline mutations predispose to breast cancer. These include mutations in BRCA1, BRCA2, p53, PTEN and ATM. BRCA1 and BRCA2 mutations confer the highest level of risk of breast cancer and also predispose to ovarian cancer. Germline mutations in p53 give rise to the Li Fraumeni syndrome (childhood sarcomas, brain tumours and early onset breast cancer) and mutations in PTEN to Cowden’s disease.

The mode of inheritance is usually autosomal dominant with incomplete penetrance. It is important to note that the genetic susceptibility can be passed on by both males and females. Most cases of breast cancer with a genetic basis will have occurred by the age of 65. Those at substantial risk include women with:

• three first- or second-degree relatives with breast or ovarian cancer

• one first-degree relative with bilateral breast cancer

• two first- or second-degree relatives with breast cancer diagnosed under the age of 60 or ovarian cancer at any age on the same side of the family.

(First-degree relatives are mother, sister or daughter. Second-degree relatives are grandmother, granddaughter, aunt or niece.)

Low risk polymorphisms

Much of the risk of breast cancer may be conferred by low risk polymorphisms. Most of the polymorphisms relate to the alteration of the metabolism of steroid hormones or carcinogenic compounds.

Acquired

Ductal and lobular carcinoma-in-situ

Premalignant in situ carcinoma may occur confined to the lobules (lobular carcinoma-in-situ (LCIS)) or ducts (ductal carcinoma-in-situ (DCIS)) without evidence of penetration on light microscopy of the basement membrane. DCIS occurs in 25% of breast cancers. With the advent of breast screening, the diagnosis of DCIS has increased three to four times and now accounts for 15% of cases detected by mammography. Ninety percent of DCIS is confined to one segment of the breast. DCIS is a heterogeneous disease based on histology, genetic alterations and clinical course. Inactivation of the tumour suppression gene, e-cadherin, is associated with the development of LCIS. Loss of the tumour suppressor gene, p53, is associated with the development of poorly differentiated DCIS. Most of the genetic alterations seen in DCIS are also seen in invasive cancer which often accompanies DCIS. Progression from DCIS to invasive is associated with the same degree of differentiation (i.e. poorly differentiated DCIS tends to progress to poorly differentiated invasive cancer).

Most cases present with non-palpable calficifications. Although mammography may detect over 80% of cases of DCIS, it commonly underestimates the extent of the disease. Magnetic resonance imaging (MRI) is more effective at detecting multifocality. Core biopsy is able to establish a diagnosis in 90% of screen-detected cases. DCIS is associated with a substantial risk of progression to invasive carcinoma, with a mean delay of about 7 years. There are two principal histological groups: comedo and non-comedo type. The most common non-comedo types are solid, cribriform, papillary and micropapillary.

LCIS is associated with an increased risk of tumour in both breasts, particularly infiltrating ductal carcinoma.

Histology

A lump in the breast may be benign or malignant. Benign lesions include cysts, fibroadenomas and papillomas. Malignant tumours mainly arise from the glandular epithelium (adenocarcinomas). Breast cancers are classified as of no special type or of special type. The majority (80%) are of no special type. The histological types of breast cancer are shown in Table 26.1.

Table 26.1 Classification of invasive breast cancer

1.	No special type
2.	Special type
	a. Tubular
	b. Mucoid
	c. Cribriform
	d. Papillary
	e. Medullary
	f. Classic lobular

Invasive cancers have traditionally been classified by their microscopic appearance and by histological grade. Grading is based on nuclear pleomorphism, degree of glandular formation and frequency of mitoses. Grading yields useful prognostic information; both disease-free survival and overall survival are shorter for high-grade cancers.

If in excess of 25% of the main tumour mass contains non-invasive (in situ) disease and in situ disease is present in the adjacent breast tissue, the tumour is described as having an extensive in situ component. Such patients are at risk of developing invasive cancer elsewhere in the breast and are not suitable for breast conservation.

Lymphatic or vascular invasion within the tumour confers an increased risk of local and distant recurrence.

Inflammatory carcinomas are typified by an enlarged warm breast, often associated with an ill-defined underlying mass. Histologically, there is infiltration of the subdermal lymphatics. Prognosis is poor. In contrast, medullary carcinoma is slow growing and has a much better prognosis.

Lobular invasive carcinomas are often bilateral (40%) and multicentric. Paget’s disease of the nipple is commonly associated with an underlying ductal adenocarcinoma.

Diagnosis

The patient may notice a lump herself on routine or casual self-examination. Sometimes it is detected by her general examination as part of clinical examination for other reasons or as a result of routine examination in hospital. With the introduction of the breast-screening programme in the UK, asymptomatic breast cancer is being more frequently diagnosed among women routinely screened for breast cancer between the ages of 50 and 69. In 1998, 75% of women (1.2 million) aged 50–64 were screened, detecting 7000 cancers, a yield of 0.6% (6 per 1000). With the extension of the breast-screening program to older women up to the age of 69, the number of breast cancers detected through breast screening is likely to rise. Patients should be referred to a specialist breast unit with multidisciplinary management from surgeon, oncologist, radiologist, pathologist and breast care nurse.

Clinical assessment

A full history is required, including details of breast-related symptoms, particularly breast pain, nipple discharge, changes noted in the skin (erythema, dimpling) or shape of the breast, indrawing or distortion of the nipple, axillary lumps and systemic symptoms of weight loss, anorexia, nausea, vomiting, bone pain, breathlessness, headache or motor or sensory disturbance. A full menstrual history should be taken including onset of menarche, menopause, parity, age of first pregnancy, breast/bottle feeding and use of the contraceptive pill and hormone replacement therapy.

Mammography

Breast screening by mammography (Figure 26.2) in the UK is recommended to women between the ages of 50 and 69 every 3 years. In the UK, two views of the breast are obtained. In the countries where it has been widely applied (e.g. in Sweden), its use has been shown to reduce the mortality of breast cancer by 30%.

Figure 26.2 Patient undergoing mammography.

(Courtesy of the Rotherham Breast Screening service)

Features suggestive of malignancy are small microcalcifications, stellate opacities with ‘legs’ extending into the surrounding tissues (Figures 26.3 and 26.4) or distortion of architecture. Mammography may also show enlarged nodes in the axilla. About 15% of cancers are not detected by mammography and nearly 4% are neither palpable nor visible on mammography.

Figure 26.3 Mammogram showing three spiculated tumours in the breast.

(Courtesy of Dr R. Peck, Sheffield)

Figure 26.4 MRI scan showing recurrent cancer in the breast.

(Courtesy of Mr M. Dixon, Edinburgh Breast Unit)

MRI scanning (Figure 26.4) may have an important role in the assessment of (a) the local extent of the primary tumour and of multifocality in younger women where the density of the breast is often a limiting factor to the resolution of mammography and (b) recurrent disease in the irradiated breast 18 months or more after radiotherapy has been completed. MRI may show 10–30% additional invasive cancer in patients presenting with a unifocal lesion. However, MRI may lead to the overestimate of the size of lesions, potentially resulting in unnecessary mastectomies.

Breast ultrasound

Ultrasound of the breast has an important role in helping to distinguish benign from malignant lesions, particularly when mammography is normal or equivocal. Malignant tumours tend to show an abnormal echo pattern (Figure 26.5). Colour Doppler ultrasound may show changes caused by increased tumour vasculature both in primary tumours and in lymph nodes.

Figure 26.5 Ultrasound of the breast showing indistinct outline of carcinoma (right) compared to well-defined margins of benign fibroadenoma (left).

(Courtesy of Professor M. Dixon, Edinburgh Breast Unit)

Obtaining a histological diagnosis

It is important to obtain a histological diagnosis of breast cancer to confirm suspicious clinical or radiological features. If the tumour is palpable (usually 1 cm or larger), fine needle aspiration cytology (FNAC) is recommended as the initial method of confirming the diagnosis.

Core biopsy under local anaesthetic for palpable lesions or stereotactic image-guided core biopsy for impalpable lesions is becoming increasingly popular as an adjunct to FNAC.

Staging

Staging is important to assess the local, regional and metastatic spread of breast cancer since management may differ significantly depending on the extent of the disease. Staging involves clinical, radiological and laboratory assessment. The simplest staging system which is still in use is shown in Table 26.2. However, the TNM classification of the International Union Against Cancer (Table 26.3) has gained widespread acceptance.

Table 26.2 Clinical staging of breast cancer

Stage	Clinical findings
I	Freely movable (on underlying muscle). No suspicious nodes
II	As stage I but mobile axillary node(s) on the same side
III	Primary more extensive than stage I, e.g. skin invaded wide of the primary mass or fixation to muscle. Axillary nodes, if present, are fixed; or supraclavicular nodes involved
IV	Extension beyond the ipsilateral chest wall area, e.g. opposite breast or axilla; or distant metastases

Table 26.3 TNM classification of breast cancer

Stage	Clinical findings
Primary tumour
Tis	Carcinoma-in-situ
T0	No demonstrable tumour in the breast
T1	Tumour less than 2 cm in greatest dimension confined to the breast
T1a	Tumour 0.5 cm or less in maximum dimension
T1b	Tumour more than 0.5 cm but not more than 1 cm in greatest dimension
T1c	Tumour more than 1 cm but not more than 2 cm in greatest dimension
T2	Tumour more than 2 cm but less than 5 cm in greatest dimension
T3	Tumour more than 5 cm in its greatest dimension
T4	Tumour of any size with direct extension to chest wall or skin
T4a	Fixation to chest wall
T4b	Oedema, infiltration or ulceration of the skin of the breast
T4c	Both of above

Regional lymph nodes
N0	No palpable nodes
N1	Mobile ipsilateral nodes
N2	Fixed ipsilateral nodes, fixed to each other or to other structures
N3	Ipsilateral internal mammary nodes

Distant metastases
M0	No distant metastases
M1	Distant metastases including skin involvement beyond the breast area and supraclavicular nodes

Staging investigations

For patients with T1–2, N0, M0 disease, a full blood count, liver biochemistry and chest radiograph are recommended. For patients with (a) T3 or T4 tumours, (b) N1–3 or MI, or (c) T0–2N0 disease with symptoms which could be due to metastatic disease (e.g. unexplained bone pain or weight loss), a liver ultrasound and a bone scan are recommended.

All patients with locally recurrent disease following mastectomy (Figures 26.6 and 26.7) or breast-conserving therapy, regional or metastatic recurrence require bone scan and liver ultrasound in addition to full blood count, liver function tests and chest radiograph.

Figure 26.6 Nodular local recurrence on the skin flaps of a mastectomy scar.

Figure 26.7 Widespread nodular recurrence over left chest wall following mastectomy, extending to the other breast.

Treatment of ductal carcinoma-in-situ

With adequate therapy, 99% of patients will survive following treatment. Where the prognosis is good, the morbidities associated with particular treatments (particularly from radiotherapy) need to be carefully weighed against the benefits of treatment.

The Van Nuys Prognostic Index takes account of three factors which influence the risk of local recurrence (margin width, tumour size and pathological subtype). For each factor there is an assigned score of 1 to 3. The size categories are: ≤1.5 cm, 1.6–4 cm and ≥4.1 cm. The margin categories are: ≤1 mm, 1–9 mm and >10 mm. The pathological categories are: low grade without necrosis, low grade with necrosis, and high grade with or without necrosis. Following conservative surgery and radiotherapy, the 10-year actuarial disease-free survival is 100% for a score of 3–4, 77% for score 5–7 and 37% for a score of 8–9.

The rationale for mastectomy or whole breast irradiation as treatment for DCIS is related to the potential for multicentric disease and/or the presence of occult invasive cancer. Multifocal disease in the same quadrant is not unusual in patients with DCIS. Following wide excision and negative margins, 24–43% of patients will have residual DCIS in the same quadrant. Mastectomy remains the treatment of choice for multicentric DCIS and for large unicentric lesions. Recurrence rates after mastectomy are less than 1%. Regular mammography of the contralateral breast should be carried out, since there is an increased rate of contralateral breast cancer of approximately 7 per 1000. If the extent of the lesion is not more than 3–4 cm, then an attempt at conservative surgery may be made, aiming to achieve complete excision. The margins of clearance should be at least 1 cm. For patients with high-grade DCIS, postoperative whole-breast irradiation should be given, since it reduces the risk of local recurrence and of invasive cancer. The National Surgical Adjuvant Breast Project B-17 trial showed after 12 years of follow up that the incidence of both recurrent DCIS and invasive recurrence was reduced to 16% in patients treated by postoperative radiotherapy (50 Gy in 25 fractions over 5 weeks) compared to wide excision alone. For low or intermediate-grade DCIS, the role of radiotherapy is less clear and still the subject of investigation. There is no indication for axillary dissection or irradiation of the peripheral lymphatics in DCIS since the risk of positive axillary nodes is 4% or less. The overall prognosis of DCIS is excellent with in excess of 97% of patients alive and disease free 10 or more years following diagnosis.

Prognostic factors

While there are a large number of biological prognostic factors for breast cancer, none has surpassed the value of assessing the number of histologically involved nodes and tumour size. There is a direct correlation between number of involved nodes and survival (Figure 26.8). Ten-year survival is about 40–65% with one to three positive nodes, and 20–42% for those with 10 or more positive nodes. Ten-year survival is about 65–70% in women with negative nodes. By contrast, in excess of 50% of all women who are axillary node positive die within 10 years despite treatment.

Figure 26.8 Survival by axillary node status.

(Reproduced with permission from Veronesi U (ed.) Baillière’s Clinical Oncology, 1988)

Within any category of nodal status, tumour size is an independent prognostic factor. The decline in survival with increasing size of the primary tumour is shown in Figure 26.9. Less than 30% of patients with stage IIIb disease (T4) are alive at 10 years. Similarly, survival declines with increasing tumour grade. Five-year survival falls from 80% for grade 1 to 25% for grade 3.

Figure 26.9 Age-corrected survival rates by tumour size for patients treated for breast cancer apparently confined to the breast.

(Reproduced from Halnan, Treatment of Cancer, Chapman & Hall, 1982)

Endocrine therapy is based on blocking the effects of oestrogen which stimulates tumour growth or inhibiting its production. Oestrogen passes through the cell membrane and binds the oestrogen receptor (ER). ER alpha, one of the two species of ER binding to ligand, leads to phosphorylation and transcription. Most of the oestrogen is sited in the nucleus but some probably exists in the cell membrane where it may interact with growth factors. Activation of membrane ER by either oestrogen or tamoxifen and the ensuing activation of growth factor signalling may be a cause of tamoxifen resistance in certain patients which overexpress HER2. It is possible that a combination of ER-targeted therapy and growth factor receptor-targeted therapy might reverse tamoxifen resistance. The strategy is being tested in clinical trials. Cross-talk between growth factor receptor pathways and ER might elucidate the relationship between progesterone receptor (PR) expression and response to a variety of hormonal therapies. As a result of an active growth factor kinase cascade, the transcription of the PR gene may be inhibited. Hence, in some tumours, PR negativity may be a surrogate for active growth factor signalling. This would suggest that ER-positive, PR-negative tumours would respond better to aromatase inhibition than to selective oestrogen receptor modulators (SERMs), as in HER2-positive tumours. Indeed, the ATAC trial showed better outcomes in PR-negative tumours with anastrozole than tamoxifen. There is accumulating data that ER-positive, PR-negative tumours are a discrete subset with a tendency to higher growth factor receptor activity, relative resistance to SERMs and greater sensitivity to aromatase inhibitors, such as anastrazole and letrozole.

Oestrogen receptor status is predictive for disease-free and overall survival. Irrespective of stage, ER positivity predicts for longer disease-free (Figure 26.10) and overall survival. Higher recurrence and lower survival rates are found in ER-negative patients. About 60% of ER-positive patients will respond to hormonal manipulation. Progesterone receptor (PgR) status may also help. Oestrogen stimulates PgR production in normal reproductive tissue and in human breast cancer cell lines. The highest response and disease-free survival rate is seen in ER+/PgR+ tumours. Very few tumours are ER−/PgR+, consistent with the production of progesterone receptors being dependent on oestrogen synthesis. Lowest response and disease-free survival rates are seen in ER−/PgR− tumours (Figure 26.11).

Figure 26.10 Survival by axillary node status and oestrogen receptor status in 854 patients: 1 = negative nodes ER+; 2 = negative nodes, ER−; 3 = positive nodes, ER+; 4 = positive nodes ER−.

(Reproduced with permission from Baillière’s Clinical Oncology, International Practice and Research, 1988)

Figure 26.11 Disease-free survival by oestrogen and progesterone receptor status in pathological stage II patients.

(Reproduced with permission from Baillière’s Clinical Oncology, International Practice and Research, 1988)

Of the biological markers of prognosis including p53, cathepsin D, epidermal growth factor receptor and HER2/neu, HER2/neu is the most reproducible. Patients overexpressing HER2/neu have a higher risk of recurrence and shorter survival. There is evidence that tumours that overexpress HER2/neu are relatively resistant to chemotherapy with cyclophosphamide, methotrexate and 5-fluorouracil (5-FU) (CMF) and have greater responsiveness to anthracyclines. In addition, different gene profiles may be found for ER-positive and ER-negative patients.

Gene profiling

More recently, microarray based gene profiling has enabled thousands of genes to be studied in a one tumour. A microarray consists of known and unknown DNA samples on a solid slide (Figure 26.12). The probes may be cDNAs or oligonucleotides of varying length. The sequence hybridized to probes on the array can be fluorescently labelled. Expression signatures based on the expression level of large nunbers of genes can be determined reflecting the properties of the cells studied. There is now preliminary evidence that DNA microarrays may be able to discriminate patients at higher or lower risk of systemic relapse among conventionally ‘low risk’ node-negative patients treated by breast conserving therapy. These findings will have to be validated prospectively in larger data sets before they can be used to determine management for individual patients.