Mammary Tumors With Sebaceous Differentiation in Dogs

Abstract

Editor: We read with particular interest the 2007 article by Dr Chang et al, which describes the first case of canine mammary carcinoma with sebaceous differentiation. The authors compared the histologic features of the canine tumors with those of human cases. Since then, new and distinctive cases of this mammary carcinoma variant have been reported. Additionally, we sometimes encounter canine mammary gland carcinomas with sebaceous components among our own diagnostic cases. The aim of this letter is to draw the attention of Veterinary Pathology readers to the morphologic variation in reported mammary tumors with sebaceous components to complement the review by Dr Chang et al. Although it has been documented once in the veterinary literature, we believe that these additional data will contribute to the understanding of this peculiar mammary carcinoma subtype. Although the number of well-documented cases is still limited, the World Health Organization classification for human breast tumors now recognizes sebaceous carcinoma as a distinct subtype of invasive breast carcinoma. However, the presence of sebaceous differentiation and other neoplastic components vary markedly among the reported cases, which makes the classification criteria of mammary sebaceous carcinoma ambiguous. A sebaceous component has been described in mammary adenoid cystic carcinomas, invasive ductal carcinoma (some with in situ components of cribriform and comedocarcinoma types), and intraductal breast papilloma with squamous/apocrine metaplasia. Other authors have described tumors with a predominance of sebaceous cells (> 50%), with or without basal cells, and have thus classified them as primary sebaceous carcinomas of the breast. Dr Chang and colleagues were prudent in classifying their canine case as a mammary carcinoma with sebaceous differentiation because, apparently, some cells of the typical ductal carcinoma component had undergone sebaceous differentiation. However, because extensive solid multilobulated neoplastic tissue composed of basal and sebaceous cells was also observed, they probably could have used the term mammary sebaceous carcinoma due to the predominance of sebocytic cells in the neoplasm. The morphologic characteristics (and the degree of cellular atypia) of the nonsebaceous components in these mammary carcinomas may be quite variable. Thus, we believe that the term sebaceous carcinoma is not suitable for all of these lesions and that care must be taken when evaluating mammary tumors with sebaceous differentiation because the nonsebaceous component may be quite aggressive. Histochemistry is helpful in the diagnosis of these neoplasms. The lipid vacuoles in the sebocytic cells are not stained by periodic acid–Schiff, mucicarmine, or alcian blue but are stained by oil red O. Additionally, immunohistochemistry to detect adipophilin may be performed to support the diagnosis. Whereas the oil red O technique is not reliable for detection of lipid in paraffin-embedded tissue, the antibodies to adipophilin will bind sebaceous cells, even in tissues that were routinely processed with the lipid solvents used in paraffin embedding. The origin of the sebaceous component in these mammary carcinomas is unknown. Some authors suggest that basaloid cells of the tumors can differentiate toward cutaneous adnexal structures or that local stem cells might give rise to sebocytes.