The demand for specialist treatment of patients with breast disease is increasing for several reasons. Firstly, the subject has become more complex, and an integrated approach involving not only surgeons but also radiologists, radiotherapists, pathologists, and medical oncologists is necessary. Secondly, the scope of the subject has increased dramatically. Not long ago the majority of surgeons who dealt with breast disease perceived their role as little more than differentiating benign conditions from malignancy and treating the latter. Patients now demand specific management of benign disorders which were previously ignored. Increased understanding of the relationship between benign and malignant disease has increased awareness of conditions previously dismissed as ‘benign changes’, mastitis or fibroadenosis. Finally, the development of new therapeutic and diagnostic methods such as screening, breast conservation, and adjuvant chemotherapy has increased the complexity of the management of breast cancer itself.

Many institutions have now developed specialized breast disease units, housing surgeons with a special interest and training.

The breasts are modified sweat glands in that they are embryologically derived from a downward growth of ectoderm into the underlying mesenchyme. The first stage of development occurs at 6 or 8 weeks of gestation, when two strips of thickened ectoderm, the mammary ridges, grow in a line extending from the embryonal axilla to the inguinal region. In many animals breasts develop along the whole length of this ridge, but in humans true breast tissue occurs only in the pectoral region.

Branching epithelial cords appear as 15 to 20 buds, which eventually become lactiferous ducts and associated alveoli. Each cord becomes surrounded by a connective tissue stroma of mesenchymally derived fat, connective tissue, and vascular tissue. These cords form the basis of the segmental pattern of the adult breast.

Towards the end of gestation the lactiferous ducts become canalized and open on to a pit in the epidermis. At the same time mesenchymal proliferation beneath the epidermis allows nipple development; failure to do so results in inversion.

Lobular development within the breast occurs particularly at puberty, with the ultimate development of 15 to 20 lobes, each of which drains into a single lactiferous duct; true secretory alveoli develop only during pregnancy and lactation.

Few changes occur in the early years of childhood. However, at the age of 10 there is growth of mammary tissue beneath the areola. This produces a characteristic protuberance on the chest wall called the breast bud or mound. True nipple development occurs at about the age of 12, followed 2 or 3 years later by further subareolar growth and formation of the bulk of the breast tissue. Finally, there is areolar recession, resulting in the classical shape of the adult breast. These changes at puberty are a result of the action of follicle stimulating hormone and luteinizing hormone, produced by hypothalmic stimulation of the pituitary gland on oestrogen production from the ovary.

Although the adult breast varies greatly in size its base is fairly consistent anatomically, extending from the second to the sixth rib in the midclavicular line and overlying the pectoralis major, serratus anterior, and external oblique muscles. Medially, the breasts reach the sternal edge and laterally the mid-axillary line. The pyramidally shaped axillary tail extends into the axilla.

The fascial covering of the breast is of importance with respect to surgical technique. As it develops from the skin the breast is invested with superficial fascia which divides into two layers. The anterior layer provides a plane of dissection subcutaneously between the relatively small subcutaneous fat lobules and the larger lobules of mammary fat. The posterior layer of superficial fascia abuts against the deep fascia derived from the pectoralis major and serratus anterior, thus producing a potential space. This retromammary space is easily located surgically (Fig. 1) 795.

Between the two layers of superficial fascia there are condensations of fibrous tissue. These form the suspensory ligaments of Cooper which both divide the breast into lobes and act as a supportive framework.

The breast has a rich vascular supply from the internal mammary artery and from the thoracoacromial, subscapular, and lateral thoracic branches of the axillary artery. The main supply is via the second perforating branch of the internal mammary and lateral thoracic arteries; the former should be preserved during subcutaneous mastectomy.

The regional venous drainage forms a rich subareolar plexus which drains via the intercostal, internal mammary, and axillary veins.

The lymphatic drainage of the breast has obvious implications in the management of breast cancer and has been well documented in an elegant series of studies by Turner-Warwick.

Approximately 75 per cent of the lymphatic vessels drain to the 30 or so regional lymph nodes in front of and below the axillary vein. These nodes can conveniently be subdivided into three main groups according to their relationship with the pectoralis minor muscle; nodes at level 1 lie below the muscle, level 2 lymph glands lie behind it, and those of level 3 are in the apex of the axilla above the muscle. The majority of lymph drains from nodes at level 1 sequentially to those at levels 2 and 3; a small amount drains to the subscapular and interpectoral (Rotter's) groups of nodes. Extension of tumour to these latter nodes may imply retrograde spread and heavy involvement of axillary nodes by metastases (Fig. 2) 796.

A small amount of lymph drains from the superior aspect of the breast directly to the apical nodes. These pathways account in part for the well-documented, but relatively unusual, findings of nodal disease at level 3 but negative nodes in the lower axillary chain.

About 25 per cent of lymph drains to the internal mammary nodes in the second, third and fourth intercostal spaces. Such drainage is from the whole breast, although the majority of these pathways arise in its medial half. Some lymph drains to the opposite breast and down the rectus sheath. Drainage to the contralateral axilla and to intercostal nodes via the pectoral fascia is, however, minor. Similarly, there are few direct lymphatic communications between the breast and the lymph nodes above the axillary vein.

The distribution of major lymphatics accompanies the blood supply. The original description by Sappey in 1885 of a subareolar plexus which receives lymph centripetally and which then redirects it to the axilla has been discounted.

The nervous supply to the breast is primarily by sensory and sympathetic nerves. Little parasympathetic innervation has been demonstrated. The nipple has a rich sensory supply whereas the majority of sympathetic innervation is to breast parenchyma.

The microscopic anatomy of the breast is a complex subject that has been the source of much controversy and confusion. Basically, there is a system of major ducts arranged in a segmental and radial pattern. These lead to the secretory component of the breast—the terminal ductal–lobular unit (TDLV). The breast is thus subdivided into lobes, although these are not precisely defined anatomically. The interlobar fascia (the most prominent of which are Cooper's ligaments) is dense and fibrous, whereas the periductal and intralobular connective tissue is much more loose and vascular. The connective tissue round the lobule is particularly loose, allowing expansion during pregnancy and lactation (Fig. 3) 797.

The ‘terminal ductule’ of the terminal ductal-lobular unit has two components (Fig. 4) 798. One lies outside the lobule and is known as the extralobular terminal ductule (ETD), whereas the other lies within the lobule (intralobular terminal ductule, ITD). The terminal ramifications of the latter form the secretory units of the breast (terminal ductules or acini), in a manner morphologically similar to the terminal units of respiratory alveoli in the lung.

The whole subject is complex because of difficulties with nomenclature. The term ‘terminal duct’ has been used for both the smallest epithelial unit in the lobule and the largest duct that opens on to the nipple surface. The same term has been used interchangeably with ‘tertiary duct’, ‘ductule’, and ‘terminal ductule’. The problem is compounded by some authors wishing to retain the term ‘acinous’ only for those terminal units during breast feeding.

Histologically, that part of the duct system adjacent to the skin of the nipple is lined with stratified squamous epithelium. However, there is a sudden change to a double layer of columnar or cuboidal epithelium which characterizes the remainder of the duct system. The terminal ductules (acini) can undergo secretory changes and therefore have a role of both transport and secretion.

Between the epithelial cell layers and the basement membrane is a network of myoepithelial cells. These respond to oxytocin and are responsible for milk ejection during lactation.

Despite widespread anecdotal belief to the contrary there is little evidence of substantial histological changes occurring in the breast throughout the different phases of the menstrual cycle. Retention of fluid may occur during the luteal phase, but this does not appear to be associated with morphological change.

The main histological change that occurs in the breast during pregnancy is lobular-alveolar growth and the development of new secretory units. This gives rise to the characteristic microscopic description of ‘adenosis of pregnancy’. (Fig. 5) 799. It is characterized histologically by alveolar dilatation and conversion of the resting two-layer epithelium to a monolayer, which demonstrates secretory changes within it. Colostrum formation, capillary growth, vascular engorgement, and myoepithelial cell hypertrophy are also apparent as pregnancy progresses. There is a doubling of breast weight from about 200 to 400 g, much of which is due to fluid retention. These changes occur under the influence of increased levels of luteal and placental sex hormones, placental lactogens, and chorionic gondotrophins.

Although some colostrum is formed within the breast during pregnancy true milk is not produced until about 2 days after parturition. This occurs due to the high postpartum levels of prolactin, which are maintained despite falling levels of other sex hormones which normally inhibit lactogenesis. During lactation the alveoli are distended with milk, the cells become cuboidal in shape and there is a resultant diminution of the intralobular space.

Following cessation of breast feeding there is gradual return to the non-pregnant state. This process may take as little as 3 months, but in some individuals may take much longer. This process, known as post-lactational involution, is characterized histologically by lymphocytic infiltration and hyalinization of the lobules (Fig. 6) 800. There is, however, little or no reduction in the number of ducts and lobules present.

Involutional changes occur from about the age of 35, with regression of glandular tissue and its replacement by fat and fibrosis. Before the age of 50 this process is characterized by loss of some lobular tissue; in older women progression of this process results in the almost complete replacement of lobular tissue by collagen and fat. The end result is that although major duct systems are visible few lobules are seen. This is in contrast to post-lactational involution, which is characterized by minimal loss of lobular units.

Changes at the menopause have two important clinical implications. Firstly, fat infiltration of the breast produces the low density appearance of breast parenchyma seen on mammography, and thus makes this technique more successful in older woman. Secondly aberrations of this involutional change may explain some of the benign disorders that occur in this age group.

It is not the purpose of the present discussion to elaborate on the advantages and disadvantages of breast feeding. However, breast milk from a healthy mother allows normal development of an infant for 6 months or longer without need for any supplementation. In many underdeveloped countries breast feeding provides a major source of nutrition for the infant for up to 2 years. Breast feeding will continue despite a poor nutritional state of the mother.

Breast milk has a relatively low protein content; that which is present is in the form of casein and lactoglobulin. Carbohydrate is present as lactose, and fat as triglyceride. Human breast milk is rich in vitamins C and D and is a source of IgA. It is therefore important in helping provide an immune mechanism for the newborn.

A number of drugs are excluded in breast milk; this should always be taken into account when prescribing for breast feeding women. Of particular importance are alcohol, barbiturates, anticoagulants, tetracycline, and metronidazole.

Until recently benign disorders of the breast were regarded as relatively unimportant: far more attention was focused on breast cancer. This has resulted in many patients with benign breast disease receiving rather scant attention from clinicians, and there has been relatively little academic investigation into this complex subject. Benign breast disease has also suffered from the major disadvantage of a hopelessly confusing terminology, inadequate classification, and poor correlation between clinical, radiological, and pathological features.

During the past decade there has been increasing interest in benign breast disease for a number of reasons. Firstly, patients demand investigation and treatment for symptoms of benign breast disease. This has, in turn, increased the number of women referred to specialist breast disease units; these have participated in scientific studies on the classification and treatment of their condition. Secondly there is the question of premalignant disorders and histological features which may imply an increased risk of breast cancer. Increasing understanding of these conditions may prove important in understanding the pathogenesis of breast cancer and in defining high risk groups in whom regular surveillance may be beneficial. Finally, the recently introduced breast screening programmes are likely to present pathologists and clinicians with as yet ill-defined histological entities which may be of importance in understanding the development of invasive cancer and its eventual treatment.

There is no completely satisfactory classification of benign breast disease. Previous attempts have been based on a number of different factors such as clinical symptoms, patient age, histological features, or that part of the secretory system in which the abnormality has arisen. They all have inherent disadvantages. Firstly, there is poor correlation between clinical, pathological, and radiological features in any particular case. Secondly, benign breast disorders encompass a wide spectrum of clinicopathological features ranging from near normality to severe disease. Finally, the breast must be regarded as a physiologically dynamic structure in which cyclical variations are superimposed on changes of development and involution throughout the woman's life. These physiological changes may themselves be so extensive that they may fall outside what is regarded as the normal spectrum. The histological features of an individual abnormality must therefore be evaluated within the context of a wide range of normality.

It has therefore been suggested that the broad concept of benign breast disease should be reconsidered. Many so-called diseases of the breast might now be regarded more accurately as disorders that are based on aberrations of the processes of development, cyclical change and involution (ANDI). This does not mean that benign breast disease does not occur, but that the term should be reserved for disorders of such severity that they are frankly abnormal. This concept is, of course, rather ill-defined and will depend on interpretation and perception by both patient and clinician.

Aberrations of normal development and involution can account for many, if not all, benign breast disorders. A simplified system based on the various stages of physiological change (development, cyclical change, pregnancy, lactation, and involution) is shown in Table 1 264. It follows that most conditions listed under ‘benign breast disorders’ can be regarded as minor aberrations or normal development or involution. Many patients with these conditions require reassurance rather than specific treatment and explanation that they do not have a disease.

This approach is well demonstrated in patients with cyclical mastalgia and nodularity. The vast majority of premenopausal women experience a degree of breast discomfort and increasing nodularity prior to menstruation. For most, this amounts to little more than an inconvenience and is regarded as a normal physiological process. About 2 or 3 per cent of women are referred to a clinic with cyclical mastalgia, the symptoms of which are more severe, with distressing discomfort lasting a week or more. In the past all such patients, with either mild or severe symptoms, were described as suffering from ‘fibrocystic disease’ although there is little histological evidence of either fibrosis or cyst formation. Despite exhaustive investigation there is little evidence of any specific histological abnormality in women with cyclical breast pain, and if abnormal microscopic features are observed they correlate poorly with clinical features. Cyclical mastalgia and nodularity, like other benign breast conditions, must be regarded in their broadest terms.

Despite the complexity of its classification there are relatively few presenting symptoms of benign breast disease. Symptoms fall into three main groups: breast pain, lumps, and disorders of the nipple and periareolar region. Infection of the breast causes further symptoms. As has been previously stated, attempts to correlate pathological and radiological findings with clinical features are a cause of much confusion and should be discouraged. A working clinical classification of benign breast disease is shown in Table 2 265, although it must be stressed that this does not imply whether the process is physiological or pathological, and in no way does it attempt to correlate symptoms with pathological features.

Finally it must be remembered that the reason for many referrals is anxiety on the part of either the patient or her doctor. The accounts for the increasing number of patients seen with non-breast conditions, including skin rashes and chest wall pain. It is an important function of any breast clinic not only to treat symptoms but to allay the patient's fears of breast cancer.

Pain is the most common reason for referral to a breast clinic and accounts for up to 50 per cent of patients seen. It is, however, the least understood of all breast symptoms, and the one whose management causes most controversy.

Literature on breast pain is often anecdotal, uncontrolled, and of poor quality. Nomenclature poses a major problem: a number of different terms such as mastitis, mastodynia, and mazoplasia, have been used to describe breast pain. Mastalgia has also been correlated with specific histological criteria, resulting in its description as ‘fibrocystic disease’, although this has lost favour for reasons described above.

It must be stressed that mastalgia is a symptom and does not imply any specific pathological process, any more than does pain in other sites of the body.

Attempts to classify breast pain are surprisingly recent. There are two distinct group of patients with these symptoms (Table 2) 265. One group of patients have symptoms which bear a definite relationship to the menstrual cycle (cyclical mastalgia); in the remainder there is no such correlation (non-cyclical mastalgia). Non-cyclical mastalgia has recently been reclassified to distinguish pain in the breast from that originating in surrounding tissues such as the chest wall.

This is the most common type of breast pain, accounting for 40 per cent of all cases referred to a breast clinic. There is an important correlation with the menstrual cycle, with discomfort lasting for a varying period of time prior to menstruation. Because of this cyclical relationship to the menstrual cycle mastalgia is generally a condition of premenopausal women, who present at a median age of about 35 years. Characteristically, the features of cyclical mastalgia wax and wane. Episodes of discomfort may last for some months; there may then be years of freedom until symptoms begin again.

The pain of cyclical mastalgia is frequently, but not always, bilateral and is usually located in the upper outer quadrants. It is poorly localized and may radiate across the chest wall into the axillae or down the inside of the arm. The breasts are frequently described as being ‘heavy’ as if pregnant and many patients describe marked nodularity at the time of the discomfort.

There is a wide spectrum of symptoms in patients with cyclical mastalgia. The majority have only mild discomfort lasting 2 or 3 days prior to menstruation, and they are not unduly concerned by their symptoms. Such individuals are therefore best classified as having a breast ‘disorder’ (ANDI), rather than a disease. The small minority of women who have severe symptoms lasting throughout the cycle with relief only during menstruation are those to whom the term ‘disease’ may be applied.

There are no mammographic or pathological characteristics of cyclical mastalgia: indeed this lack of correlation between clinical, radiological and histological findings is one of the major characteristics of the condition. The mammograms shown in Fig. 7 801 are from patients with severe breast pain. In one there is almost complete replacement with fat, giving a translucent appearance, whereas the other is dense and nodular (Wolfe DY pattern).

The fact that symptoms of cyclical mastalgia correlate with the menstrual cycle implies a hormonal aetiology. Early investigations suggested that hormonal imbalance was the cause, the fundamental abnormality being relative hyperoestrogenism due to either increased oestrogen secretion or deficient progestogen production. However, the vast majority of studies have failed to demonstrate either abnormality in women with mastalgia.

More recently, abnormal prolactin secretion has been incriminated as an aetiological factor in cyclical mastalgia. Although both random and basal levels of prolactin are normal in women with cyclical mastalgia, there is some evidence of impaired hypothalamic control of the release of this hormone in patients with severe symptoms. It should be appreciated, however, that the control of prolactin release is extremely complicated and that our current knowledge of its physiology is rather rudimentary.

The belief that cyclical mastalgia has a hormonal basis resulted in the suggestion that there would be an associated effect on fluid retention. However, despite a widespread belief that breast pain is due to water retention this has never been scientifically confirmed.

Other aetiological factors, including excessive caffeine ingestion or inadequate essential fatty acid intake, have been suggested. The latter is of particular interest as there is good evidence that essential fatty acid supplements can reduce the symptoms of cyclical mastalgia. The actual mechanism is unclear, but it may relate to a resulting deficit in prostaglandin E1 deficiency and a subsequent enhanced effect of prolactin on the breast.

Psychoneurosis has been widely incriminated as an important factor. However, there is no evidence of excessive anxiety, depression, or phobia in these patients when they are evaluated against appropriate control groups.

More than 80 per cent of women attending a breast clinic with cyclical mastalgia require no treatment other than simple reassurance, particularly that such symptoms do not imply any form of neoplastic process. Fear of cancer drives many women to seek specific advice about breast pain, and the importance of such reassurance cannot be over-emphasized.

About 5 to 10 per cent of patients with cyclical mastalgia experience pain despite all reassurance. For those patients specific drug therapy may be considered. There is a sound theoretical basis for use of most agents which have been tried, apart from the fact that no constant physiological or pathological changes have been identified in this condition. Furthermore, no drug satisfies the criteria of being universally effective,free from side-effects, and freely available for use in patients suffering from benign breast disease. A large number of studies evaluating the efficacy of these drugs have been performed. However, because of the placebo effect of such treatment the results of many studies are inherently flawed, and reliance can only be placed on prospective, randomized, placebo controlled trials. A further problem of many studies is that they do not take into account the natural history of mastalgia. As a result a false impression of benefit may occur merely from natural remission, such as occurs in pregnancy or at the menopause.

Table 3 266 shows some of the agents widely used for the treatment of cyclical mastalgia, their possible modes of action, and common side-effects. Their overall efficacy is shown in Table 4 267.

Diuretics have been widely prescribed, although there is little rational basis for their use, and it is widely believed that much of their efficacy is due to a placebo effect. Similarly, there is no rationale for using antibiotics. The concept of relative hyperoestrogenism as a result of luteal deficiency has stimulated the evaluation of progestogens in cyclical mastalgia; the results have been generally disappointing in placebo controlled trials. Other widely adopted treatments of cyclical mastalgia are also of dubious value. Reduction in caffeine intake or administration of vitamins A or B&sub6; have failed to show any effect on cyclical mastalgia. Administration or oral contraceptives may reduce symptoms of cyclical breast pain, but total success can by no means be guaranteed.

Relative hyperoestrogenism can also be treated with the anti-oestrogen drug, tamoxifen. This was shown to be of benefit in a single randomized trial, but as yet it only has a licence for use in malignant disease. Pain relief is provided by 10 mg tamoxifen daily; there seems little benefit in increasing the dose. Side-effects such as weight gain and hot flushes are troublesome and as it is contraindicated in pregnancy appropriate contraception is necessary.

Danazol probably acts as an antigonadotrophin by its action on the pituitary–ovarian axis. At a dose of 200 to 400 mg daily it depresses production of follicle stimulating hormone and luteinizing hormone and ovarian function. It significantly reduces breast pain, but is associated with side-effects in 20 per cent of patients. These include weight gain, acne, amenorrhoea, masculinization with hirsuitism, voice change, and reduction of breast size. Adequate non-hormonal contraception is necessary.

The suggestion of abnormal prolactin levels in patients with cyclical mastalgia, and the possibility that prolactin stimulates glandular breast tissue, led to hopes that the specific prolactin lowering agent, bromocriptine, would be useful in the treatment of cyclical mastalgia. This, a dopamine agonist, significantly reduced symptoms of cyclical mastalgia in benign breast disease. However, as is the case with danazol, its use has been curtailed by sideeffects—such as nausea, postural hypotension, vomiting, and dizziness, which occur in up to 20 per cent of patients.

The suggestion that breast pain may be secondary to a deficit of essential fatty acids has led to its treatment with evening primrose oil, a mixture of linoleic and linolenic acids. Randomized studies have shown that it is effective and well tolerated in patients with cyclical breast pain. It is regarded as a natural homeopathic substance by many patients.

Patients with cyclical mastalgia should be treated initially with evening primrose oil, followed by danazol for patients refractory to treatment. Bromocriptine is a third choice, with activity similar to that of evening primrose oil but with a significant incidence of side-effects. The response rates tend to be lower for second and third lines of treatment. Tamoxifen has the drawback that it is not strictly registered for use in benign disease.

Responses to treatment are relatively short lived, usually of the order of 6 months. It is therefore our policy to treat for 3 months and then to see whether relapse occurs on cessation of therapy. Any relapse is an indication for restarting treatment, perhaps at a lower dose than originally used, or for a change in therapy if the initial response has been poor. Treatment is particularly difficult in young women, in whom mastalgia is often resistant to treatment, whose potential for breast pain may span several decades, and whose fertility must be considered. Bromocriptine and danazol are potentially teratogenic and require the use of a barrier form of contraception, as they interfere with oral contraceptives. Many younger women also dislike the amenorrhoea induced by tamoxifen and danazol.

Occasionally women with a long history of mastalgia unresponsive to all medical treatment demand consideration of mastectomy to release them from their discomfort. Although occasional patients may benefit from subcutaneous mastectomy, the general impression is that it should be avoided if at all possible.

Non-cyclical mastalgia (Table 2) 265 is even less well defined than its cyclical counterpart. It occurs in both pre- and post menopausal women, with a median age of presentation of 45. As well as having no close relationship to the menstrual cycle, non-cyclical mastalgia tends to be more chronic, unilateral, and located in the medial quadrants of the breast or the periareolar regions. It is not associated with lumpiness to the same degree as cyclical mastalgia and the pain is frequently described as burning or dragging rather than being a heavy feeling. The mastalgia is sometimes well localized; ‘trigger spot zone’ has been used in these individuals.

Attempts to classify non-cyclical mastalgia have been compromised by the dubious inclusion of other conditions that may cause breast pain. The best such example is that of duct ectasia/periductal mastitis. The inclusion of this condition was based on the mammographic appearances of many patients with non-cyclical mastalgia, which showed widespread coarse calcification throughout the substance of the breast. This is also a feature of duct ectasia/periductal mastitis. However, the use of this term for patients with non-cyclical mastalgia has fallen into disfavour because of the principle of not mixing symptomatology with pathology, and because of the lack of evidence that pathological changes of duct ectasia correlate with breast pain.

Up to 50 per cent of patients with non-cyclical mastalgia have pain that arises not from the breast but from surrounding musculoskeletal structures (Table 2) 265. A careful history and examination will identify such patients who, unlike those with true non-cyclical mastalgia, can be provided with relatively simple and effective therapy.

The aetiology of non-cyclical breast pain is unclear. However, some factors relating to cyclical mastalgia may also relate to non-cyclical breast pain.

The management of true non-cyclical mastalgia is unsatisfactory. Many principles relating to the treatment of cyclical mastalgia may be applied to non-cyclical breast pain. However, overall response rates to various drug therapies are only about 50 per cent of those observed in patients receiving treatment for cyclical mastalgia. Both bromocriptine and evening primrose oil have shown particularly disappointing results. Response rates to the various drugs improve if patients with true non-cyclical mastalgia are differentiated from those with musculoskeletal pain.

Some success has been ascribed to surgical excision of ‘trigger spot zones’, but this approach has not been widely adopted.

This has previously been included as a cause of non-cyclical mastalgia, but has recently been demonstrated to be a separate entity. It is the most common cause of apparent pain in the breast originating from other sites. Musculoskeletal pain is often unilateral and localized along the lateral chest wall or the costochondral junction (Tietz's syndrome). Injection of local anaesthetic or steroids into the affected area produces good relief of symptoms.

This may be found microscopically either as a single entity or in association with other abnormalities. It may be a minimal histological change or a macroscopically obvious entity. It is classified as an aberration of normal development and involution; microscopically it is characterized by proliferation of terminal duct lobules, myoepithelial cell proliferation, increased number of acini, and fibrous stromal change (Fig. 8) 802. Multifocal and nodular types are described. Either may be painful and have been documented as a cause of mastalgia.

The pain associated with sclerosing adenosis may be due to perineural invasion, and this may account for some patients with trigger spot zones. The main importance, however, is that macroscopically sclerosing adenosis may have a stellate appearance and may calcify—it may thus mimic carcinoma, both clinically and radiologically. The increased cellularity associated with sclerosing adenosis has been confused with carcinoma histologically, especially when examining frozen sections.

A small number of patients continue to complain of pain after biopsy for benign breast disease. There is no clear reason for this phenomenon, although if the biopsy has previously been performed in an area subject to mastalgia it is likely that the original process will continue and be painful.

Cancer is an uncommon cause of breast pain, although women do occasionally complain of discomfort, which may be more pronounced prior to menstruation, at the site of a tumour. The author has seen one young, intelligent woman complaining solely of breast pain, while failing to notice that the breast has been largely replaced by tumour.

This cause of pain in the breast should be considered in elderly patients in whom no other cause for mastalgia can be found.

Approximately 40 per cent of all patients attending a breast clinic have a benign breast lump (Table 2) 265. In the past there was a tendency to excise all lumps, and an excessive amount of unnecessary surgery was performed for benign disease. The main problem from the woman's point of view is fear that such a lump may be cancer. The clinician must therefore provide a high degree of diagnostic accuracy while at the same time ensuring that an excessive biopsy rate is prevented. It is now easier to exclude cancer with the development of diagnostic aids such as mammography, ultrasonography, and aspiration cytology.

The surgeon in the breast clinic has two important tasks when confronted with a patient with a breast lump. Firstly, he has to decide whether the lump is truly an abnormality or whether it can be regarded as being within the spectrum of normality. Secondary, if the lump is a true abnormality, he has to determine whether it is malignant.

The history is of particular importance. Enquiry should be made into the nature of the lump with respect to its duration, pain, change in size, and relationship to the menstrual cycle. The presence of menstrual irregularity and a previous history of similar problems should be sought. It is also important to enquire into the patient's risk factors for breast cancer, such as her age, number of children, age at first pregnancy, family history, and other potential predisposing factors such as hormone replacement therapy or oral contraceptive use.

Having established the risk or otherwise for breast cancer, the clinical impression must be confirmed by careful examination. Tethering of the skin, distortion of the breast, and nipple retraction are common features of malignancy, but they can also occur as a result of benign change. Mobility of the lump should be assessed: this is characteristic of a fibroadenoma and also quite obvious for cysts. Cancers tend to be more fixed, but are occasionally quite mobile.

Finally the surface of the lump should be assessed. Fibroadenomas have a lumpy, bosselated, surface whereas cysts are usually smooth and tense. Cancerous lumps are usually, but not always, hard. While a cancerous lump is likely to be hard, irregular, and fixed, it is not uncommon to see malignant tumours that are firm, quite regular, and which have a degree of mobility.

If there is any doubt a pathological diagnosis or biopsy is necessary; this should always be undertaken in any woman over the age of 25 with a solid discrete lump. In the majority of patients aspiration cytology is sufficient, but if this is in any way unsatisfactory or, if real doubt remains, then biopsy is mandatory. It is a brave clinician who does not remove a discrete breast lump from a 40-year-old nulliparous woman with a family history of breast cancer, even if fine-needle aspiration cytology is benign and mammography is reassuring.

Fibroadenoma and related tumours are derived from the breast lobule and are characterized by both connective tissue and epithelial proliferation. They encompass a wide spectrum of conditions, ranging from the totally benign simple fibroadenoma to locally invasive and, rarely, frankly malignant tumours. There has been great confusion over their pathogenesis, particularly at the more malignant end of the spectrum. Such tumours are now commonly known as phyllodes tumours, but were previously described as cystosarcoma or phyllodes sarcoma. These descriptions are inappropriate because they imply a totally mesenchymal stromal origin whereas all of these tumours, whether benign, locally invasive, or malignant, also have an epithelial component. The major features of fibroadenomas and associated conditions are summarized in Table 5 268.

Fibroadenomas are benign tumours showing evidence of both connective tissue and epithelial proliferation. They originate from the breast lobule and can be regarded as an aberration of normal lobular development rather than a true tumour. Their origin explains why fibroadenomas are common in young women at the time of lobular development, and why they are occasionally seen in combination with lobular carcinoma. The aetiology of a fibroadenoma is unknown; hypersensitivity to oestrogen within a lobule has been suggested.

The most important pathological aspect of a fibroadenoma is its connective tissue stroma. In the past great importance has been attached as to whether this stroma compresses adjacent ducts to form curved slit-like structures (intracanalicular pattern) or whether it simply surrounds a duct circumferentially (pericanalicular pattern). The fibrous stroma of such fibroadenomas has low cellularity and a regular cytology (Fig. 9) 803,804. Occasionally there is histological evidence of fat, smooth muscle, squamous metaplasia, and infarction. The epithelial proliferation may be quite hyperplastic, but this is of no prognostic importance.

If the fibrous stroma shows a marked increase in both cellularity and atypia then the locally invasive and occasionally metastatic phyllodes tumour should be considered (see below). This entity can be regarded as the extreme end of the disease spectrum, simple fibroadenoma representing the other end.

Most fibroadenomas present in girls aged 16 to 24 years old. However, the use of pathological examination in the diagnosis of breast lumps in older women, the overall median age of presentation is nearer 30. They decrease in incidence after the menopause, when they undergo involution. During this time they may calcify and thus become apparent on mammography.

Fibroadenomas are smooth or slightly lobulated structures, usually measuring about 2 or 3 cm in diameter. With the exception of those adjacent to the nipple they are characteristically mobile. In young girls the term breast mouse is thus aptly applied. With increasing age the degree of mobility lessens because of the restraining effects of surrounding involuting fibrotic tissue. In the elderly they can still present as a small hardened mass which is still quite mobile.

About 10 per cent of all patients have multiple fibroadenomas on presentation,and occasionally one sees young women in whom the whole breast seems to be replaced by fibroadenotic tissue. Others present with multiple recurrent fibroadenomas. This occurs particularly among the black and oriental races.

Up to the age of 25 a clinical diagnosis suffices. Thereafter pathological confirmation is required because of the need to exclude carcinoma. Fine-needle aspiration cytology provides an accurate method of diagnosis (Fig. 10) 805,806 in older women, although hyperplastic epithelial cells can occasionally be mistaken for neoplasia. As fibroadenoma usually presents clinically in younger women, mammography has no place in its routine diagnosis. In older patients fibroadenomas appear as a solitary smooth lesion on radiography, with a density similar to or slightly greater than the surrounding tissue. With increasing age stippled calcification becomes apparent (Fig. 11) 807.

The practice of surgically removing all fibroadenomas has now been condemned because of greater understanding of the natural history of this condition. If fibroadenomas are left untreated most will slowly increase in size from 1 to 3 cm in diameter over a period up to 5 years. The active growth phase is about 6 to 12 months, during which time there is a doubling of size. Thereafter they remain static or may (in up to one-third of cases) gradually become smaller.

In women under the age of 25 routine removal is unnecessary. This conservative approach may be recommended for woman under 35, provided that cytological examination rules out malignancy. This is, however, likely to result in a small number of cancers being missed, and removal of fibroadenomas is generally recommended after the age of 25. Such excision is best done under a general anaesthetic.

Recurrence of a fibroadenoma after removal is not uncommon. This may be due to a number of factors. Firstly, a truly metachronous fibroadenoma may develop. Secondly the original tumour may have been incompletely removed or missed at operation and, finally, it may be the mode of presentation of a previously undiagnosed phyllodes tumour.

Giant fibroadenoma has a bimodal age of presentation at the extremes of reproductive life; those occurring in the younger age group have been described as juvenile fibroadenomas. They occur particularly in the 14 to 18 and the 45 to 50 age groups and are characterized by rapid growth to a large size. Giant fibroadenomas are, by definition, bigger than the common type of fibroadenomas, being at least 4 or 5 cm in diameter, and sometimes achieving a diameter size of 10 cm or more (Fig. 12) 808.

Histologically, giant fibroadenomas contain the typical hypocellular stromal and epithelial components showing varying, though usually mild, degrees of hyperplasia and atypia; mitoses are uncommon. Such features are different from phyllodes tumours, which generally exhibit much more cellularity, pleomorphism, and mitotic activity. However there is some overlap of microscopic appearance between these two conditions (Table 5) 268.

Giant fibroadenomas are more common in black and oriental races. Clinically patients present with pain in the breast associated with a rapid increase in size. On examination the breast is enlarged and the nipple may be displaced. The overlying skin frequently has a characteristic shiny appearance, and dilated veins may be apparent. In extensive, neglected cases skin necrosis can occur.

Occasionally girls may present with unilateral breast enlargement, and the fact that a mass is the cause is not appreciated. Giant fibroadenomas may be confused with virginal hypertrophy, although the latter is bilateral and not associated with cutaneous or venous changes.

Management is by enucleation through an appropriately cosmetic incision. While this treatment initially results in some discrepancy in breast size the remaining breast tissue expands to virtual normality within a year or two. Wider excision or mastectomy is contraindicated.

Although some giant fibroadenomas can appear somewhat aggressive histologically and may even be confused with phyllodes tumour their clinical behaviour is completely benign. There is no evidence that they recur locally or metastasize.

Phyllodes tumours have been the cause of much misunderstanding and argument, partly related to their varied nomenclature. They have been described as phyllodes sarcoma, cystosarcoma, cystosarcoma phyllodes, and benign cystosarcoma. They have also been equated with giant fibroadenoma, but this is also misleading as it understates the malignant potential of the phyllodes tumour and implies similar histology. Conversely, terms such as cystosarcoma overstate the malignant potential and imply a false correlation with true mesodermally derived sarcomas. Phyllodes tumours show a wide spectrum of activity, varying from an almost benign condition to a locally aggressive, and sometimes metastatic tumour.

Phyllodes tumours appear well circumscribed but are characterized by irregular surface projections that may be cut during surgical excision, predisposing to recurrence. The cut surface is soft, brown in colour, and may exhibit cysts, necrosis, or haemorrhage. Histologically, both epithelial and fibrous stromal elements are present, with the stroma showing hypercellularity, much atypia and numerous mitoses (Fig. 9) 804.

Phyllodes tumours occur in premenopausal women. They are usually seen in the 30 to 50 year age group, but are not uncommon in women aged about 20. They have the features of a common fibroadenoma but can grow rapidly to a large size and may involve much of the breast. The overlying skin may become reddened and, in advanced cases, can become frankly ulcerated. However, a degree of mobility is retained, even by large tumours. Axillary lymphadenopathy is uncommon; if it occurs it indicates an extremely aggressive form of the disease.

Despite the tendency to grow rapidly to a large size it is not uncommon for phyllodes tumours to present as a much smaller mass which is clinically indistinguishable from a simple fibroadenoma.

Phyllodes tumours occurring in young women under the age of 20 are said to represent the benign end of the spectrum of this condition. As such, simple enucleation has been recommended, although the author prefers to excise the area more widely.

Older patients require wider excision with a 1 cm margin of normal breast tissue. Vary large tumours or those with aggressive histology may merit even wider excision, with quadrantectomy or even simple mastectomy and reconstruction for the largest tumours.

Even with an aggressive surgical policy of wide excision approximately 25 per cent of phyllodes tumours recur over a 10-year period. Such local recurrence should be widely excised. If recurrent tumours develop persistently mastectomy with reconstruction should be considered. A major worry with persistently recurrent phyllodes tumours is that they may metastasize, although this is a rare occurrence, being described in less than 5 per cent of patients.

Breast cysts are among the more common reasons for referral to a breast clinic. They are frequently confused with more extensive cyclical nodularity. The description of cyclical nodularity as fibrocystic disease compounds the problem, and results in the false hope that many patients with cyclical nodularity can be treated by simple cyst aspiration.

True breast cysts are very common: up to 7 per cent of women develop a clinical cyst at some time during their lives. Postmortem studies show that a further 20 per cent of women have evidence of subclinical cysts in the breast, although many of these are only 2 or 3 mm in diameter.

As is the case with fibroadenomas, breast cysts can be regarded as an aberration of normal lobular physiology. The specific aetiology of this aberration is unknown, although there is some weak evidence that cyst formation may relate to hyperoestrogenism, such as may result from hormone replacement therapy. The pathogenesis of breast cysts is similarly unclear. Early workers suggested that they might simply be distended ducts or that they may result from cystic lobular involution. During this process lobules develop microcysts which eventually coalesce to become larger cysts; this process is potentiated by obstruction of lobular outflow and replacement of surrounding stroma by fibrous tissue.

More recent investigations have suggested that the aetiology of breast cysts is more complex than previously believed. There appear to be two distinct populations of macrocyst defined by their microscopic appearance, their biochemical profile, and clinical features (Table 6) 269.

Aspirated fluid from simple uncomplicated cysts has a high Na⫀:K⫀ ratio (>3), similar to that found in plasma. The pH of this fluid is less than 7.4 and it is likely that the flat, rather featureless, epithelium of such cysts acts as a simple membrane through which interstitial fluid passively diffuses. These simple cysts tend to be single, not recurrent, and are unlikely to be associated with an increased risk of cancer.

The second type of cyst is lined by apocrine epithelium, characterized by large columnar cells resembling those found in apocrine sweat glands. The Na⫀:K⫀ ratio is less than 3, and similar to that of interstitial fluid. The pH of apocrine cysts is higher than that of simple cysts and their lining membrane secretes substances such as androgen conjugates. These observations suggest that apocrine epithelium actively secretes potassium into the cyst fluid. These cysts tend to recur as the balance between secretion of fluid and its reabsorption is in favour of reaccumulation. They may also be associated with an increased risk of breast cancer.

Other studies have shown that in the early stages of cyst development the microcysts are of the apocrine secretary type. It is only when macrocysts develop that differentiation into simple cysts occurs.

Cysts are classically seen in perimenopausal women between the ages of 45 and 52, although they frequently occur outside this age range, especially in individuals receiving hormone replacement therapy. They are usually single at presentation, but it is not uncommon to see multiple cysts in a breast. In the most extreme examples the whole breast seems to be composed of a number of cysts.

A characteristic of breast cysts is that they suddenly appear, even if they are quite large. The reason for this is that the cyst exists in a flaccid subclinical state prior to its presentation as a lump. The accumulation of a relatively small amount of fluid causes a disproportionate effect on intracystic pressure according to La Place's equation (P = 2T/ r).

Cysts may be uncomfortable and are frequently frankly painful. There may be a vague relationship between discomfort and the menstrual cycle, with increasing pain prior to menstruation, although this is not a pronounced feature.

Cysts are frequently visible. However, their most characteristic feature is their smooth, tense nature on palpation. They have a degree of mobility but this is not as pronounced as that of fibroadenomas. The classic clinical appearances may be masked if the cyst is situated deep in the breast. Normal nodular breast tissue overlying the cyst may hide its classic smooth nature on palpation.

The diagnosis of a simple breast cyst is straightforward as cyst aspiration confirms the diagnosis. The amount of fluid aspirated is variable: it is usually about 6 or 8 ml, although occasionally cysts containing 60 or 80 ml of fluid are encountered. Cyst fluid varies in colour, ranging from pale yellow to almost black; sometimes the aspirate appears translucent whereas on other occasions it is thick and turgid.

Mammography and ultrasonography may aid diagnosis but these investigations are not essential, except as methods of screening these patients for cancer (Fig. 13) 809,810. Little important information is gained from cytological examination of cyst fluid unless it is bloodstained.

Breast cysts were previously treated by surgical excision. Such treatment is no longer recommended as simple aspiration will normally suffice. After aspiration the cyst remains as a lax impalpable structure which may still be seen on mammography. However, there must be no clinical evidence of a mass remaining after aspiration. If a mass does remain after aspiration, further investigation with fine-needle aspiration cytology or biopsy is indicated.

There are two main indications for surgical excision of the cyst. If the aspirate is blood-stained (as long as this is not due to direct trauma from the needle) an intracystic carcinoma may be present. The second indication, which is perhaps more contentious, is cyst recurrence. This may be simply due to inadequate aspiration and further such treatment may be attempted before excision. However, if the cyst recurs rapidly and more than once its excision is recommended.

Patients who develop persistently recurrent cysts throughout their breasts can present a difficult management problem. Recurrence is often at a different site from the presenting cyst. Up to 50 per cent of patients develop further cysts over 5 to 10 years, although the majority of individuals will only have one or two recurrences. However, a small number of women develop recurrences on a regular basis and may attend the breast clinic every 2 or 3 months for cysts to be drained. In the past some of these patients were treated by subcutaneous mastectomy. We now recommend danazol or tamoxifen treatment, although evidence in favour of this management is sparse and there are side-effects and limitations associated with the use of these drugs.

In theory, patients with breast cysts may be at an increased risk for breast cancer. Mammography should be performed on women presenting with cysts, although the yield of occult cancers is low. Patients with recurrent apocrine cysts may benefit from continued mammographic surveillance, although the evidence for their liability to breast cancer is conflicting. Patients with a solitary simple cyst do not require regular mammographic monitoring.

Many patients are referred to breast clinics with a lump that is really a manifestation of cyclical nodularity, but in a more localized and clinically discrete form. It is the mass, rather than the pain, which is the predominant feature. A careful history, however, will frequently reveal that the lump has been present for some time and that its size varies with the menstrual cycle. Many patients will also admit to discomfort or pain in the lump when it is most prominent.

A variation of this presentation is seen, particularly in teenagers and occasionally in older woman approaching the menopause. A large, diffuse, and frequently uncomfortable swelling develops suddenly, often, but not always, in the upper outer quadrant. Examination discloses a diffuse nodular swelling that may be somewhat tender. This changes usually resolves with the next menstrual cycle. If it persists, biopsy or aspiration cytology is indicated, together with mammography in older women.

As long as malignancy is excluded patients with cyclical nodularity presenting as a breast lump can simply be reassured.

Galactocele is well documented in older texts relating to breast disease, but they are perhaps less common than previously thought. Classically they present as a cyst in a woman who has recently stopped breast feeding, although they occasionally occur during lactation. Aspiration shows breast milk and is usually successful in resolving the problem.

The pathogenesis of galactocele is unclear; it may simply be a pre-existing simple cyst that fills with milk.

Sclerosing adenosis is an uncommon cause of a breast lump. Clinically, it usually presents as a smooth, relatively mobile mass in the 30- to 50-year age group. It is frequently painful and can occasionally be a cause of mastalgia rather than of a mass.

Sclerosing adenosis can be regarded as an abnormality of normal development, characterized by lobular enlargement and distortion associated with fibrous stromal change.

Fat necrosis is another condition that has attracted more attention than it deserves, although it is a frequent cause of diagnostic difficulty in women with breast disease. This arises from delayed diagnosis of cancer in patients with a history of trauma, and the occasional mastectomy performed in patients with fat necrosis thought to be a tumour.

A history of trauma is easily provided by the patient. However, this is a trap for the unwary as many patients with cancer will try to attribute the lump to a previous injury.

By the time the patient presents at the clinic, any external evidence of injury has frequently disappeared. The lump can be small and hard and clinically may easily be confused with a carcinoma. Sometimes there is associated inflammation, tethering, and oedema.

Fine needle aspiration cytology is usually diagnostic (Fig. 14) 811. Areas of fat necrosis are occasionally cystic, and aspiration can partially resolve the condition. However, in our unit we review patients diagnosed as having fat necrosis after 6 weeks, when the lump is excised if it has not disappeared. Mammography must be interpreted with care as fat necrosis can have radiological features similar to those of cancer (Fig. 15) 812.

Lipomas occur quite frequently in the breast, but not to the extent that might have been thought in view of the amount of fat that is present. They produce a soft mass which is best excised if there is any doubt as to its nature. The adenolipoma is a variation of lipoma. It sometimes has a marked fibrotic component and is best regarded as a hamartoma.

The increasing use of antibiotics as a treatment for inflammatory breast conditions occasionally results in a chronic sterile abscess. Treatment is by aspiration, or by open drainage with excision of the wall if recurrence develops.

Patients may present with the impression of a breast lump that may be due either to a normal structure, such as a rib, or a prominent area of breast tissue which may be made more obvious by the defect from a previous biopsy or from weight loss. Fatty replacement at the time of breast involution can also give the impression of a lump: at operation fibrotic fatty tissue is found.

Disorders of the nipple are no less controversial than other aspects of benign breast disease. There are the usual reasons for these difficulties, such as lack of consensus over terminology and poor correlation between clinical features, mammography, and pathological findings. The symptoms of disorders of the nipple and periareolar tissues are discharge, retraction, and the effects of periareolar sepsis (Table 7) 270. There are multiple causes for each of these symptoms but one condition, duct ectasia/periductal mastitis, is of paramount importance.

Duct ectasia and periductal mastitis has been described for more than a century but until recently it has been relatively ignored by both clinicians and pathologists. In the past it has been regarded as part of the spectrum of fibrocystic disease and because of its histological appearance it has been given a variety of names such as plasma cell mastitis, mastitis obliterans, and granulomatous mastitis. The clinical features of duct ectasia/periductal mastitis are extremely varied and, as histological or cytological confirmation is not always possible, diagnosis must sometimes be made on clinical grounds alone.

In 1951 Haagensen suggested that the primary change in patients with duct ectasia/periductal mastitis was simple dilatation of the larger periareolar ducts (duct ectasia, Fig. 16 813). It is unusual for all ducts to become dilated but such changes are frequently bilateral. The dilated ducts fill with a stagnant, thick green or creamy secretion (grumous). These stagnant secretions lead to loss of duct epithelial lining, and associated ulceration can cause further discharge, with bleeding from the nipple. There may also be a chronic inflammatory response (periductal mastitis) in the periareolar tissues because of leakage of the secretions through the damaged duct walls. Periductal mastitis may produce a painful mass or even a frank periareolar abscess; repeated inflammatory processes cause fibrosis and result in nipple retraction.

The various pathogenic processes of duct ectasia explain many of the symptoms associated with this condition. A difficult question to answer, however, is what causes the initial duct dilatation. Suggested possibilities include hormonally induced muscular relaxation of the duct wall, inadequate absorption of secretions, or obstruction of the system by squamous debris. Unfortunately, there is little scientific basis for these suggestions. An alternative theory is that the periareolar inflammation rather than the duct ectasia is the primary pathogenic process. Periareolar inflammation may result in secondary duct dilatation and the consequent discharge, fibrosis, and nipple retraction. This alternative theory explains why younger women tend to suffer the inflammatory complications of duct ectasia, whereas nipple inversion and discharge occur in older age groups.

Duct ectasia/periductal mastitis is, therefore, a complex disorder of uncertain aetiology. The wide variety of clinical symptoms associated with it can be best explained and understood by appreciating that there is more than one process in its pathogenesis.

The more severe symptoms of duct ectasia/periductal mastitis, such as abscess formation, can be regarded as a true benign breast disease. However, in all but its most severe form duct ectasia is best classified as a breast benign disorder originating from the normal process of duct involution with fibrosis.

Nipple discharge is a common symptom presenting to breast clinics (Table 7) 270. The patient may fear cancer and the discharge may in itself be a cause of social embarrassment and annoyance. Treatment is therefore directed not only at diagnosing the cause of the discharge accurately but, if necessary, stopping the discharge itself.

Patients presenting with nipple discharge should be questioned as to whether it has been blood-stained, whether it is unilateral or bilateral, and whether it is associated with a lump. If a lump is present then diagnosis of the cause of the discharge becomes secondary to investigation of the lump.

Although inspection may reveal the source of discharge as one or more ducts, most patients have little visible abnormality. Excessive crusting may occasionally be seen on the nipple: this may simply be the dried products of secretion. Sometimes a skin disorder such as eczema may be found, the associated serous exudate producing the impression of nipple discharge.

Routine palpation of the breast must be followed by firm but gentle pressure around the areola. This will determine whether the discharge is unifocal or multifocal and whether it occurs in one or both breasts. It should also determine the nature of the discharge. Particular attention should be paid as to whether the discharge has the greyish shiny characteristic of human milk (galactorrhoea in the non-pregnant or non-lactating patient), whether it is watery, serous or, as is often the case, if it is of a thick opalescent nature which may vary in colour from cream to almost black. Of paramount importance, however, is whether the discharge contains blood. This may be bright red and fresh or much darker in colour due to the presence of degradation products.

Our own experience in investigating nipple discharge radiologically and with cytology has been disappointing, although other authorities have reported success. Particular features of note on mammography are the intraductal microcalcifications of carcinoma in situ and dilated ducts associated with the secretory granules that indicate duct ectasia (Fig. 17) 814. Ductography may demonstrate the presence of duct papillomas, but this investigation tends to be insensitive.

A number of centres recommend cytological examination of the discharge. This occasionally produces a positive finding although false negative results are common.

There are a number of causes of nipple discharge (Table 7) 270. Unfortunately it is difficult to correlate accurately the nature of the discharge with the cause, although certain features do act as a diagnostic guide. The presence of bright red blood in the discharge, a watery discharge, or drainage from a single duct are all factors that indicate a potentially serious cause and which require thorough investigation.

Discharge of milk is a normal phenomenon during pregnancy and lactation. During pregnancy blood-staining is occasionally observed: this is of no significance and reassurance is all that is required. A milky nipple discharge may occur transiently in the neonate due to transplacental passage of luteal and placental hormones from the mother's circulation.

Galactorrhoea is secretion of milk not related to pregnancy or lactation, although in itself it can be a primary physiological process. Physiological causes include mechanical stimulation of the breast and stress. It may also occur for some years after cessation of breast feeding and may be seen as part of a normal physiological process at the menarche and menopause. Under these physiological circumstances simple reassurance is all that is required.

Galactorrhoea may also be a secondary phenomenon, occurring as a side-effect of drugs that enhance dopamine activity, such as chlorpromazine, haloperidol, metoclopromide, and methyldopa. Hyperprolactinaemia due to a primary prolactin-secreting tumour or from a secondary source, such as bronchogenic cancer, is an uncommon cause of galactorrhoea.

Duct ectasia is a common cause of nipple discharge. Characteristically it causes a multifocal, bilateral discharge which is thick and opalescent and of varying colours. However, the discharge of duct ectasia can be unifocal and frankly blood-stained, particularly in the perimenopausal and older age groups.

Papillomas are one of the more important causes of nipple discharge. The discharge is often from a single duct; it is frequently serous or serosanguinous and is frankly blood-stained in 50 per cent of cases. Papillomas also account for many of the relatively unusual cases of a watery discharge.

Most papillomas are solitary and are not considered to be premalignant. Occasionally two or three papillomas may be found in a single duct, and they are unlikely to have neoplastic potential (Fig. 18) 815. Multiple papillomas, however, especially those occurring in the periphery of the breast and affecting more than one duct, carry an increased risk of malignant change. In one series, 15 or 39 such patients developed carcinoma. These peripheral lesions are more likely to cause a breast mass than a nipple discharge.

As discussed below intraductal carcinoma, and even invasive cancer, can present as nipple discharge. This is usually from a single duct, is frequently watery or serous, and is generally frankly blood-stained.

Cysts may be more common as a cause of nipple discharge than is generally appreciated. Some authorities regard cysts simply as dilated ducts and patients can occasionally produce a discharge simply by compressing an area of cystic change.

Despite all attempts at establishing a diagnosis, even with biopsy, about 10 per cent of patients have no obvious pathological entity ascribed to the discharge. If the discharge has been blood-stained it must be considered that the cause has been missed. However, if the discharge was serous and non-blood-stained simple reassurance with careful follow-up will suffice.

If the nipple discharge is associated with a mass then treatment appropriate for the lump is constituted. If no mass is felt the management depends on the nature of the discharge.

1.Galactorrhoea causes such as mechanical stimulation and ingestion of drugs known to promote galactorrhoea should be excluded. If prolactin levels are normal simple reassurance is all that is required.

2.Discharge located to one duct but not blood-stained. Mammography should be performed in patients older than 35. All patients should undergo microdochetomy and in older women, who no longer wish to breast feed, a good case can be made for major duct excision.

3.Blood-stained discharge from a single duct. Treatment is the same as for a non-blood-stained discharge from a single duct. However, as a greater proportion of patients will have significant pathological findings exploration is even more essential. Bleeding often occurs only once or twice and there are no abnormal findings on firm palpation when the patient presents. Mammography should still be performed in patients over the age of 35, and all patients should be followed up for at least 6 months.

4.Multifocal and non-blood-stained discharge. Mammography should be performed in all women over the age of 35. If this is normal, simple reassurance is all that is necessary. If the discharge is excessive and causes social embarrassment major duct excision can be recommended. The usual cause of such discharge is duct ectasia. Operations for this condition have been associated with a high incidence of infection, and cephradine and metronidazole administration is recommended.

5.Multifocal and blood-stained discharge. Mammography with major duct excision is indicated, although many cases will be due to duct ectasia.

Surgery for nipple discharge remains somewhat controversial. The more conservative approach is microdectomy, which has the advantage that women should be able to breast feed after the operation. It is therefore probably the treatment of choice in younger individuals with a single duct discharge. However, as many patients with a single duct discharge will have a more widespread disorder, simple microdochetomy may be followed by recurrence of symptoms. A more appropriate operation for older women is major duct excision, which also allows easier nipple eversion if required.

The terms nipple inversion and retraction have been used interchangeably, and the distinction is somewhat arbitrary. Nipple inversion describes a congenital failure of eversion during development whereas retraction relates to a secondary process, usually due to duct ectasia or carcinoma.

Congenital nipple inversion is a common problem occurring to a greater or lesser extent in up to 20 per cent of all women. Frank inversion causing difficulty with breast feeding is much less common: a degree of flattening of the nipple does not seem to interfere with the breast feeding process. This is hardly surprising because the nipple itself probably plays a relatively small part in anatomical aspect of suckling, since the infant creates a ‘teat’ from surrounding breast tissue as well as from the nipple itself.

Congenitally inverted nipples tend to be bilateral and patients can generally be reassured that there are few long-term sequelae from this condition. They should certainly be encouraged to breast feed. Women with congenitally inverted nipples have a higher incidence of duct ectasia/periductal mastitis.

Young women with congenitally inverted nipples often request surgical eversion. If possible this should be resisted, as the only satisfactory way of everting the nipple is to divide all the underlying ducts, which will prevent subsequent breast feeding. Furthermore, after such cosmetic surgery the nipples still have a rather flattened appearance rather than being protuberant.

The three main causes of nipple retraction are duct ectasia, carcinoma, and the effects of injudicious previous surgery.

Nipple retraction due to duct ectasia is characterized by bilateral changes and a characteristic linear transverse defect in its early stages. At this time it is possible to evert the nipple but as the process progresses digital eversion becomes more difficult. Other features of duct ectasia, such as a creamy multifocal discharge or stigmata of previous periareolar abscesses, are often present. An earlier belief that these changes often begin during pregnancy has not been confirmed.

Retraction due to carcinoma is characterized by a shorter history and is unilateral. It is frequently associated with a mass. The effects of nipple retraction itself and any associated inflammation can make clinical assessment of the retroareolar area difficult. If there is any doubt, mammography and biopsy by major duct excision should be performed.

Injudicious surgery with inadequate care in reconstituting the breast can result in nipple inversion. This is often associated with an ugly periareolar defect and is difficult to rectify. It should be avoided by due attention to surgical technique.

While discharge and retraction account for the majority of patients presenting with nipple disorders a number of other problems are occasionally encountered. Skin disorders, especially eczema, often affect the nipple and periareolar regions. Eczema must be differentiated from Paget's disease and is characterized by a long intermittent history, bilaterality, and its presence at other cutaneous sites. Features such as itching, serous discharge, and the nature of the periphery of the lesion are of little diagnostic significance. If there is any diagnostic doubt biopsy is required; a blind trial of topical steroids is to be condemned as Paget's disease can resolve temporarily in response to this treatment.

Fibroepithelial polyps present in adolescence and in young women. Excision, if required, can be performed under local anaesthetic. Chronic sebaceous cysts and retention cysts arising from Montgomery's tubercles are occasionally seen.

Nipple adenomas produce an uncomfortable mass beneath the nipple. They can cause ulceration and may be confused with Paget's disease. Simple excision will suffice; the condition is completely benign.

Pain and hypersensitivity in the nipples are difficult to explain. Sometimes this has a cyclical nature and enquiry will show a traumatic cause. Other patients describe hypersensitivity to cold, as in a Raynaud's type of phenomenon.

The majority of breast infections can be subdivided into those occurring during lactation and those which are a complication of duct ectasia/periductal mastitis. They have an entirely different aetiology, pathogenesis, and treatment, and may be considered completely separately.

Lactational breast abscesses occur during breast feeding, are generally somewhat peripherally situated, and are the result of infection by Staphylococcus aureus. Such abscesses tend to occur at the commencement of breast feeding when an inexperienced mother develops cracked nipples. They also occur at weaning when engorgement results from incomplete drainage of breast milk.

Cracked nipples resulting from the trauma of breast feeding are seen in the first week after delivery and again after about 6 months when the child's teeth first develop. There is acute pain in the nipple and examination reveals a linear fissure which may become secondarily infected.

The patient initially complains of discomfort in the breast, followed by painful swelling. The overlying skin may appear red and in extreme cases may undergo necrosis. Constitutional symptoms are common and by the time the patient presents she may have a high fever and be systemically unwell.

If ignored breast abscesses, like those elsewhere, will point and spontaneously discharge on to the skin surface.

If lactational breast infection is seen prior to frank abscess formation, antibiotic treatment alone is often successful. Aspiration should be attempted to ensure that no pus is present. As many cases, especially those occurring in hospital, are due to penicillinase-producing staphylococci treatment must be with a penicillinase-resistant antibiotic such as a second-line penicillin or a cephalosporin. Such treatment will be successful in about 90 per cent of early cases.

If, on aspiration, pus is found or if other systemic features of an abscess are present drainage is necessary. However, some recent authorities have recommended repeated daily aspiration under antibiotic cover rather than formal surgical intervention. If formal drainage is performed, however, further antibiotic therapy is no longer required. The incision used to drain the abscess should limit any cosmetic deficit and allow drainage of pus under the influence of gravity. The majority of surgeons leave the wound open and pack it daily with antiseptic-soaked ribbon gauze, although primary closure under antibiotic cover has also been recommended.

An area of confusion is whether breast feeding should continue after treatment of a breast abscess. There is no place for suppression of lactation. The infant should be encouraged to feed from the contralateral breast while the affected side should be emptied either manually or with a breast pump. Cracked nipples should be washed gently and dried by dabbing rather than by rubbing. If there is evidence of nipple infection a nystatin-containing ointment should be applied topically.

Non-lactational breast abscesses are entirely different from those occurring during breast feeding. They occur in the periareolar tissues, frequently recur, and the infecting organisms (if successfully cultured) are a mixture of bacteroides, anerobic streptococci and enterococci. Such non-lactational breast abscesses are a manifestation of duct ectasia/periductal mastitis and are usually seen in the 30- to 60-year age groups.

Patients with non-lactational breast abscesses often have a history of previous infective episodes in the periareolar region. Such abscesses often begin as a slightly tender periareolar mass. Spontaneous resolution is common but they often progress, with associated reddening of the overlying skin and increasing tenderness until the features of a frank abscess are present. Systemic upset is less pronounced than in patients with lactational abscesses. Scarring and distortion arising from treatment of similar such episodes may be present, and there may be other manifestations of duct ectasia/periductal mastitis, such as nipple retraction.

If a non-lactational abscess is suspected the inflammatory mass should be aspirated and sent for both aerobic and anerobic culture. Initial treatment with metronidazole and flucloxacillin may be successful, although repeated aspiration may be required. Many patients will require drainage, although drainage should be avoided if possible. If drainage is necessary it should be performed through the smallest possible incision.

Definitive treatment requires duct excision and possible nipple eversion. This should be performed under appropriate antibiotic prophylaxis and only when the condition is quiescent. The majority of surgeons leave the wound open and leave it to heal secondarily although there is an increasing vogue for primary closure if local conditions are suitable.

Surgical or spontaneous discharge of an abscess from periductal mastitis may result in a mammary fistula with intermittent drainage of pus or serous fluid on to the areola. This may be superimposed on further episodes of periductal sepsis. Characteristically the breast demonstrates multiple incisions for drainage of previous abscesses, distortion, nipple retraction, and a fistula at the edge of the areola (Fig. 19) 816. The basic cause is, again, duct ectasia/periductal mastitis.

Treatment is by fistulectomy, excision of the offending duct, and nipple eversion if required. If there is extensive periareolar disease or the nipple is grossly retracted a major duct excision should be performed. Most surgeons prefer to leave the wound open but again there is a growing trend for primary closure.

Fistula recurs in about 5 per cent of patients. The cause is variable, but includes continuation of duct ectasia/periductal mastitis in adjacent ducts and persistence of the proximal duct adjacent to the nipple because of an inadequate surgical technique.

Postoperative wound infections are relatively common after breast surgery, especially following surgical treatment of duct ectasia/periductal mastitis. Reasons for this include the relatively poor blood supply to fat, ischaemia resulting from deep sutures, and an accumulation of serum in the wound itself. Prophylactic antibiotics and drainage seem to have little impact in reducing the incidence of such problems; they are best minimized by meticulous technique.

The incidence of wound infection after surgery for duct ectasia/periductal mastitis has been so high in some centres that antibiotic prophylaxis with drugs such as combination metronidazole and flucloxacillin is now widely used.

Breast abscesses occasionally occur in neonates due to infection of milk induced by the transplacental passage of maternal hormones. If antibiotics do not help this condition great care must be taken during surgical drainage as damage to the breast disc at this age may lead to distortion in later life.

Tuberculosis, syphilis, hidradenitis, and pilonidal abscess have all been described as causing inflammation in the breast.

General risk factors for breast cancer are discussed below. Pre-existing benign breast disease is considered to be a risk factor, although the literature on the subject is conflicting. Studies have generally been based on selected populations who have had incomplete follow-up for an inadequate amount of time. A major problem has been lack of consensus between pathologists in ascribing individual terms to the various microscopic features and an obsession with the cancer risk associated with fibrocystic disease. The following terms require definition before further discussion.

This is cellular proliferation (Fig. 20) 817. As far as the breast is concerned the normal ducts are lined by two layers of cells above the basement membrane. Hyperplasia is therefore defined as presence of three or more layers of cells, although individual cells may fill up or protrude into epithelial lined spaces. It is synonymous with the terms papillomatosis and epitheliosis.

This occurs when hyperplastic cells exhibit bizarre or unusual features, either in the pattern of their cellular relationships or in the appearance of the nuclei (Fig. 21) 818. The extent of atypia can be graded from 1 to 3, although this classification is empirical and subject to variation between pathologists.

This is an increase in the number of glandular elements. There is a normal relationship between cells and the basement membrane.

This controversial term has found greater acceptance in the United Kingdom than in the United States. It has been used to describe the solid or semisolid benign epithelial proliferation which is found predominantly in the small ducts, ductules, and lobules. It has been implied that severe forms of epitheliosis amount to carcinoma in situ, although this point of view has been widely criticized. There has been a recent tendency to replace the term epitheliosis with hyperplasia.

This is also a controversial term which has found favour in North America. It has been criticized as it implies a derivation from the term papilloma, and the vast majority of examples bear no resemblance to a papillary pattern. Hyperplasia is being used instead of papillomatosis.

The only study which has used accurate data to assess the risk of cancer in patients with benign breast disease is that of Page et al. The results of this study form the basis of an American Cancer Society consensus statement on the risk of breast cancer in patients with benign disease which was published in 1986 (Table 8) 271. Cancer risk is defined as a liability to develop breast cancer in the ensuing 10 to 20 years, compared with development in age-matched women who have had no breast biopsy. It should be noted that these are not life-time risks.

In their original study, Dupont and Page ascribed a slightly increased cancer risk to patients with sclerosing adenosis. However, other studies have not confirmed this finding and sclerosing adenosis is therefore currently considered as having no increased cancer risk.

The American consensus statement maintained that breast cysts were not associated with an increased cancer risk. Other studies have confirmed this opinion, but there is an increasing body of data suggesting that macroscopic cysts, especially those of the apocrine type, may be associated with an increased risk of breast cancer. Further work is required to clarify this question.

The inclusion of in-situ cancer may be criticized, as this diagnosis implies a neoplastic rather than a benign process. Conversely, multiple papillomatosis, not included in the above discussion, has been clearly demonstrated to be associated with an increased cancer risk.

The relative risk of developing invasive breast cancer following benign disease implies the existence of the classical pathway from normality to invasive cancer: Equation 27

Care must be taken in interpreting this rather crude and perhaps naïve pathway. There is no evidence that the progression from normality to invasive cancer needs to pass through all the above stages. Furthermore once a certain point in the pathway has been reached there does not have to be further progression; in theory, regression may occur.

Dixon JM, Scott WN, Miller WR. Natural history of cystic disease: the importance of cyst type. Br J Surg 1985; 72: 190–2.

Dupont WD, Page DL. Risk factors for breast cancer in women with proliferative breast disease. N Engl J Med 1985; 312: 146–51.

Fentiman IS, Caleffi M, Hamed H, Chaudar MA. Dosage and duration of tamoxifen treatment for mastalgia: a controlled trial. Br J Surg 1988; 75: 845–6.

Haagensen CD. Mammary duct ectasia—a disease the may simulate cancer. Cancer 1951; 4: 749–61.

Hughes LE, Mansel RE, Webster DJT. Aberrations of normal development and involution (ANDI): a new perspective on pathogenesis and nomenclature of benign breast disorders. Lancet 1987; ii: 1316–9.

Hughes LE, Mansel RE, Webster DJT. Benign Disorders and Diseases of the Breast: Concepts and Clinical Management. London: Bailliere Tindall 1989.

Maddox PR, Harrison BJ, Mansel RE, Hughes LE. Non-cyclical mastalgia: an improved classification and treatment. Br J Surg 1989; 76: 901–4.

Mansel RE, Dogliotti L. European multicentre trial of bromocriptine in cyclical mastalgia. Lancet 1990 335: 190–3.

Mansel RE, Wisbey JR, Hughes LE. Controlled trial of the antigonadotrophin danazol in painful nodular benign breast disease. Lancet 1982; 1: 928–31.

Page DL, Anderson TJ. Diagnostic Histopathology of the Breast. Edinburgh: Churchill Livingstone 1987.

Thomas WG, Williamson RN, Davies JD, Webb AJ. The clinical syndrome of mammary duct ectasia. Br J Surg 1982; 69: 423–5.

Turner-Warwick RT. The lymphatics of the breast. Br J Surg 1959; 46: 574–82.

Winchester DP. American College of Pathologists Consensus Statement. The relationship of fibrocystic disease to breast cancer. Am Coll Surg Bull 1986; 71: 29–31.