Eccrine glands are directly developed from the embryonic epidermis in the early months of fetal development
. They are widely distributed almost everywhere on the skin
. The topographic distribution of adnexal structures also offers insight into the logical classification of adnexal neoplasms
Microscopy alone is insufficient to establish eccrine lineage neoplasm because there are none specific microscopic features
Some problems are related to the classification of sweat gland carcinomas, which are currently classified on the basis of the corresponding classification of benign sweat gland adenoma
. Such an approach, however, poses several problems; for example, (1) some carcinomas have no benign counterpart and do not fit the scheme (ductal carcinoma, adenoid cystic carcinoma, and mucinous carcinoma); (2) poorly differentiated carcinomas can be diagnosed only when a contiguous adenoma is found histologically; (3) histologic classification can be very complicated because adenomas are numerous and their classification is complex; and (4) terminology includes unusual and difficult terms, deriving from the terminology used for adenomas (malignant acrospiroma, porocarcinoma, hidradenocarcinoma, malignant cylindroma, malignant spiradenoma, and syringocystadenocarcinoma)
Recently, a classification of sweat gland carcinomas designed based on the classification of breast carcinomas has been tentatively proposed
Finally, recent studies have classified sweat gland carcinomas into eccrine and apocrine tumors
[8, 9]. In addition, no established authentic criteria are available for differentiation of an eccrine from an apocrine tumor
. Moreover, both eccrine and apocrine forms seem to exist in some categories
[5, 9, 11, 12].
The correct identification of the origin of tumor is of utmost importance for the determination of appropriate therapy and prognosis. Immunohistochemistry has proved to be a useful adjunct for this purpose
The cells in the excretory coil of eccrine gland sweat express positivity for low molecular weight keratin, epithelial membrane antigen (EMA) and carcinoembryonic antigen (CEA), as well as S100 protein in the basal layer only. The myoepithelial cell layer may be highlighted by smooth muscle actin (SMA), p63 and calponin. Acrosyringeal cells stain for high molecular weight keratin and cytokeratin 14. A subpopulation of cells also expresses positivity to bcl-2
. Some eccrine carcinomas are positif to estrogen and progesterone receptors wich has an important clinical implications, as affected patients may be partially treated with hormonal therapy
. The Ki-67 and p53 may be used to differentiate benign from malignant lesions
Clinically, eccrine carcinoma must be considered in the differential diagnosis of patients older than fifty years with long standing tumors in the limbs, including basal cell carcinoma (BCC), Paget’s disease, melanoma, metastatic cancer
, inflammatory, lymphocytic and vascular lesions
The histopathology of mucinous eccrine carcinoma of the skin is analogous to that of its counterpart in the breast. Consequently, it is sometimes difficult to exclude a metastatic disease
[15, 18]. In a comparison between primary sweat gland tumors and metastatic breast carcinoma to skin, Busam et al.
[13, 19] found that the use of antibodies against epidermal growth factor receptor strongly decorated 81% of sweat gland tumors, but only 17% of metastatic breast cancers (P=0.001). There was no significant difference between the skin tumors and metastatic breast carcinoma when antibodies against estrogen and progesterone receptors were used, but with progression of disease, androgen receptors (AR) are preserved with higher frequency than ER/PR in metastatic mammary carcinoma
. In another study comparing benign, malignant primary eccrine and apocrine neoplasms to metastatic breast carcinoma, only 3.5% of these primary adnexal cancers demonstrated HER-2 positivity, whereas 10–23% of the breast carcinomas were positive
[13, 20]. The authors suggested that this test may be useful in distinguishing primary skin cancers from metastatic breast cancers
. Another study showed that primary cutaneous neoplasms stained strongly for p63 and CK 5/6 with high specificity, while CK 7 and 20 were not useful
[15, 21]. When CK 7 was positive in the cutaneous lesions, it exhibited marked focality and a specific pattern, while metastatic breast carcinoma expressed CK 7 diffusely and did not express p63 or CK 5/6. The usefulness of CK 5/6 was again shown by Plumb et al.
[15, 22]. Ivan et al.
[15, 23] presented further evidence of the usefulness of p63 in the diagnosis of adnexal cancer; in their study, none of the examples of metastatic adenocarcinoma to skin stained for p63, whereas virtually all the adnexal carcinomas were positive.
These studies showed that immunohistochemistry does not distinguish cutaneous eccrine tumours from cutaneous metastases of breast carcinoma, in which case clinical and radiological correlation is critical
Other differential diagnoses comprise neoplasms with clear cell differentiation. These include trichilemmal carcinoma, clear cell BCC and clear cell carcinomas metastatic to the skin. With respect to the latter, the dominant diagnostic considerations are metastatic clear cell carcinoma from thyroid gland and lung. The expression of TTF seen in the great majority of these cases is helpful in resolving this dilemma. Some metastatic thyroid cancers also manifest expression of thyroglobulin. Metastatic clear cell carcinoma of renal primary origin manifests a prominent stromal vascularity with hemorrhage and areas of granular necrosis typically evident in biopsy material. Clear cell squamous cell carcinoma of the cervix is a glycogen-rich cancer with intercellular bridge formation characteristic of squamous differentiation