Microcystic Adnexal Carcinoma

Abstract

† Since its initial description, microcystic adnexal carcinoma (MAC) of the skin has been controversial. In particular, it features keratin production of the type seen in some pilar neoplasms, and has been thought to pursue partial follicular differentiation. Diagnostically, MAC may be difficult to separate from desmoplastic trichoepithelioma (DTE) in superficial biopsy specimens. We studied 12 MACs, 22 malignant eccrine acrospiromas, 7 sudoriferous syringometaplasias, 6 syringomas, 5 DTEs, and 40 other benign pilar neoplasms immuno-histochemically. Paraffin sections and antibodies to "hard" (pilar) keratins, epithelial membrane antigen (EMA), carcinoembryonic antigen (CEA), Leu-M1, and S 100 protein were employed. The MACs exhibited reactivity for hard keratin subclasses AE13 and AE14, EMA, CEA, and Leu-M1. Desmoplastic trichoepitheliomas expressed positivity for AE14, EMA, and Leu-M1 focally, but lacked the other specified markers. Syringomas and malignant acrospiromas displayed EMA, CEA, and AE14 reactivity, and 5 syringometaplastic lesions were AE14-reactive. Benign pilar tumors aside from DTEs were reactive only for AE13, AE14, or both. These data indicate that MAC exhibits an immunophenotype that is a "hybrid" of those seen in pure sweat glandular and follicular neoplasms, and suggest that it may indeed show combined pilar and sudoriferous differentiation. Based on these results, it also appears that immunohistochemical analysis may be useful in the diagnostic separation of MAC and DTE. (Arch Dermatol. 1990; 126:189-194)