Sir James Paget first described Paget disease (PD) of the breast in 1874. He reported a chronic eczematous disease on the skin of the nipple and the areola in 15 women, with an associated intraductal carcinoma of the underlying mammary gland. In 1881, Thin illustrated the first histologic description of PD.
Mammary PD occurs almost exclusively in women; involvement of the male breast is rarely reported.  Patients with Paget disease frequently present with a chronic, eczematous rash on the nipple and adjacent areolar skin. Proper recognition of this disorder is required to initiate an appropriate workup (eg, skin biopsy) for differentiating it from other benign inflammatory dermatoses and for detecting an underlying breast carcinoma.
A similar disease involving the skin of female and male external genitalia (ie, vulva, glans penis) is known as extramammary Paget disease.  The histologic features of mammary PD and extramammary PD are similar; however, the histogenesis and the pathogenesis are different.
Pathophysiology and Etiology
The pathogenesis of mammary PD and the origin of Paget cells were once controversial. It is now widely accepted that mammary PD is always associated with an underlying carcinoma of the breast. By thorough histologic examination, Muir documented intraepidermal extension of malignant ductal epithelial cells through the lactiferous ducts and ductules into the epidermis (epidermotropism).  The findings are the basis of the epidermotropic theory of mammary PD.
Malignant epithelial (Paget) cells infiltrate and proliferate in the epidermis, causing thickening of the nipple and the areolar skin. These tumorous epithelial cells are derived from luminal lactiferous ductal epithelium of the breast tissue (see the image below).
The cells possess microscopic features of glandular cells. Paget cells and underlying ductal carcinoma cells have been shown to be positive for the oncogene HER2/neu, suggesting common genetic alterations for both the epidermal and breast tumor cells.
It has been speculated that Paget cells may derive from glandular stem cells or epidermal Toker cells (clear cells of the nipple epithelium).  Toker cells have been found in about 10% of normal nipples and rarely in supernumerary nipples on the mild line and apocrine bearing areas.  Like Paget cells of both mammary and extramammary sites, Toker cells contain prominent clear (vacuolated) cytoplasm, and they are considered benign counterparts of Paget cells.
Mammary Paget cells are malignant epithelial cells derived from underlying ductal adenocarcinoma of the breast that invade into the skin of nipple and areolar areas. Toker cells are benign and may sometimes proliferate, resulting in a condition known as clear cell papulosis. 
Kuan et al  reported on the immunohistochemical expression of apomucin MUC1, MUC2, and MUC5AC in PD and concluded that both epidermal Paget cells and underlying ductal carcinoma exhibit the same phenotypic apomucins that are also expressed by the Toker cells.
Morandi et al  reported that the chromosomal alterations seen in Paget cells (eg, loss of heterozygosity and mitochondrial DNA displacement loop sequence analysis) are different from those seen in underlying breast carcinoma cells. They suggested that the epidermal Paget cells are genetically different from those of breast carcinoma.
A concept of a “collision” between the neoplastic lesion of mammary PD and the underlying carcinoma is therefore presented. However, this concept is based on only a few cases. Further studies are necessary to arrive at a more definitive conclusion concerning collision lesions.
Paget cells often express cell markers that mimic those of the underlying breast carcinoma, including glandular epithelial cell markers (eg, low-molecular-weight cytokeratins or cellular adhesion molecule [CAM] 5.2).
They also express tumor markers, including carcinoembryonic antigen (CEA); Ca 15-3 (milk fat globule protein); oncogenes (TP53,c-erb B-2; about 85% of cases tested for c-erb B-2 are positive); and other cell markers, such as epithelial membrane antigen (EMA; all cases tested are EMA positive) and gross cystic disease fluid protein (GCDFP-15) found in tumor cells of ductal carcinoma of the breast.
Paget cells also share similar immunohistochemical characteristics with eccrine and apocrine sweat gland epithelium. Paget cells are periodic acid-Schiff (PAS) positive and diastase resistant; and they are Alcian blue positive at pH 2.5 in 32% and 18%, respectively.
Advances in immunohistochemistry further define the histogenesis of Paget cells. Cytokeratin 7 (CK7) has been proposed as a specific and nearly 100% sensitive marker for mammary PD; whereas cytokeratin 20 (CK20) is negative in mammary Paget cells, about 33% of cases of extramammary PD express this marker. However, there is a rare CK7-negative variant of Paget disease.  If in doubt, staining positive for GATA3 might be diagnostically beneficial.
The mechanism by which large neoplastic cells of glandular origin (ie, the Paget cells) infiltrate and spread to the overlying epidermal layers of the nipple and adjacent areola is induced by a mobility factor (heregulin-alpha) that acts through the HER2/neu receptor. 
Normal epidermal keratinocytes produce and release heregulin-alpha. This factor plays a significant role in the pathogenesis of PD. Paget cells express heregulin receptors HER2/neu and coreceptors HER3 and HER4. These receptor complexes on Paget cells bind heregulin-alpha, the mobility factor, resulting in the chemotaxis of breast ductal carcinoma cells. This process, in turn, causes migration and infiltration of Paget cells into the overlying epidermis of the nipple and the areolar skin.
Molecular studies have shown that abnormal plakoglobin (one of the cell-to-cell adhesion proteins) may be involved in the formation of some cases of PD of the breast and the vulva; and reduced expression of E-cadherin (another cell-cell adhesion protein) may have a role in the pathogenesis of extramammary PD. 
In the United States, about 1-4% of female breast carcinoma cases are associated with PD of the nipple, the areola, and the surrounding skin; nearly 100% of mammary PD cases are associated with an underlying carcinoma, either in situ (intraductal, 10%) or infiltrating cancer (90%). Occasional cases of PD have been reported in ectopic breast or supernumerary nipples.  Worldwide, t he exact frequency of mammary PD is unclear.
In female patients with mammary PD, ages range from 24 to 84 years, with a mean age at diagnosis of 55 years; the average age range is 53-59 years. The average patient age is 5-10 years older for patients with mammary PD than for individuals with breast carcinoma. The age of onset in male patients is generally in the fifth and sixth decades, with a range of 48-80 years.
Mammary PD occurs almost exclusively in females; PD of the male breast is extremely rare. The estimated occurrence of male breast carcinoma is only 1% of the rate in females. A rare case of PD in the male breast was recorded after treatment of a prostate carcinoma with estrogen. No racial predisposition is reported for mammary PD.
The prognosis for patients with mammary PD is related to disease stage and appears to be similar to that of women with other types of breast cancer.
Survival is related to the presence or the absence of a palpable breast tumor. When present, the prognosis is the poorest.  In one study, 31 (62%) of 50 patients with mammary PD presented with a detectable breast mass.
One half (50%) of patients with PD presenting with a palpable breast mass have associated axillary lymph node metastasis. Two thirds of patients with axillary node metastasis were reported to have a palpable breast mass, whereas one third of patients with axillary metastasis did not have a palpable mass.
Even in patients with mammary PD and no underlying tumor, 30% may develop an invasive carcinoma at a later date, and 20% of patients already have an associated in situ carcinoma of the breast. However, other reports indicate that no axillary metastases were detected in patients without a palpable breast mass.
Patients with an identifiable associated underlying breast tumor have a survival rate of 38-40% at 5 years and a survival rate of 22-33% at 10 years. The death rate for metastatic breast carcinoma in patients with mammary PD and underlying cancer is 61.3%, with a 10-year cumulative survival rate of 33%.
The reported survival rate of patients with PD without a palpable breast tumor (prior to surgery) ranges from 92% to 94% at 5 years and from 82% to 91% at 10 years.
Mastectomy is the standard treatment of mammary Paget disease. Conservative treatment with preservation of the nipple-areola complex results in a higher rate of recurrence than treatment by mastectomy.
A small number of patients with mammary PD who did not have a palpable breast mass and who had a negative mammogram underwent conservative excision of the involved nipple-areola with a wedge resection of the underlying breast tissue and remained disease free at 10-year follow-up.
Rare cases of recurrent mammary PD have been reported in patients who underwent partial nipple excision; however, no patient developed breast parenchymal recurrence at a mean follow-up of 36 months.
In a study of patients with mammary PD (without clinical or mammographic evidence of breast tumor) who were treated with radiation therapy alone or excision plus radiation, 3 of 20 patients developed recurrent disease in the nipple-areola at a mean follow-up of 7.5 years. Those patients eventually underwent a mastectomy.
According to Osteen,  9 (11.4%) of 79 patients treated by local excision (with or without radiation therapy) developed recurrences.
Patients should be informed that PD of the nipple-areola complex is a rare form of breast cancer that is clinically characterized by eczematous changes in the nipple. Clinical symptoms include erythema, itching, a burning sensation, thickening of the skin, inversion of the nipple, and sometimes a bloody nipple discharge. Usually, symptoms are present for 6 months or more before the detection of an underlying breast cancer. A biopsy should promptly be performed on all suspicious lesions of the nipple area for accurate diagnosis and treatment.
Failure to educate patients about the potential for invasive breast carcinoma associated with mammary PD is a pitfall; without this knowledge, patients may delay further diagnostic mammography examination and possibly delay detection of an underlying breast cancer.