What is Extraocular Cutaneous Sebaceous Carcinoma In Situ?

Abstract

To the Editor: We thank Drs Kramer and Chen1 for their interest in our article published in this Journal. The authors raise a valid question as to why the term in situ carcinoma cannot be applied to cutaneous sebaceous neoplasms. They also rightfully cite the definition of carcinoma in situ, which is a malignant epithelial neoplasm confined within the epithelium of origin. But the main question is: What is the origin for cutaneous sebaceous carcinoma? Whereas in periorbital sebaceous lesions, it is accepted that sebaceous lesions arise from meibomian glands and glands of Zeis, sebaceous glands elsewhere in the skin practically never appear to give rise to a sebaceous carcinoma. In our files we have over 100 unequivocal extraocular sebaceous carcinomas, and in none of them is there evidence that the tumor has arisen from a preexisting sebaceous gland in a manner analogous, for example, to an invasive squamous carcinoma for which intraepithelial precursor lesions including actinic keratosis and Bowen disease (squamous cell carcinoma in situ) are typically found. We have seen a limited number of intraepidermal sebaceous carcinomas (Fig. 1), not extending beyond the epidermal basement membrane and lacking any association with sebaceous gland. Such lesions qualify as sebaceous carcinoma in situ and their features indicate that they have originated from epidermis, and that the original basement membrane at the dermoepidermal interface is intact. It is our view that the term “sebaceous carcinoma in situ” is valid only for a vanishing minority of sebaceous neoplasms and in most of these it is not possible to determine their precise origin (viz. sebaceous gland vs. epidermis) let alone assessing accurately whether the original basement membrane has been breached. Therefore, we would discourage the use of the term “sebaceous carcinoma in situ” for other sebaceous tumors.FIGURE 1: Sebaceous carcinoma in situ. The tumor is wholly intraepidermal and does not extend beyond the preexisting basement membrane (A). There are more and less differentiated areas (B).Our article2 intended to highlight the existence of a small group of cutaneous sebaceous proliferations wherein the distinction between malignant and benign may be difficult and we accept that opinions on their classification may vary. The current approach in dermatopathology is to classify sebaceous lesions as benign or malignant. However, in sebaceous tumors attempting to classify accurately all lesions into either of these 2 categories may be an oversimplification of the problem. Some sebaceous tumors have a symmetrical silhouette similar to other benign adnexal tumors and they behave in a benign fashion despite having clearly malignant cytologic features. It is important to state that a similar situation can be seen infrequently in other adnexal tumors such as trichoblastomas and pilomatricomas. We disagree with Drs Kramer and Chen that the 5 cases we reported are better classified as sebaceous adenoma. Sebaceous adenoma has completely different architectural and cytologic features. It has a distinctly multilobular architecture with several contiguous lobules usually showing connection to the overlying epidermis and replacing the squamous cells. The lobules vary in shape, sometimes even within a single lesion and there is a pushing, sharply demarcated interface with the adjacent stroma. The lateral-most lobules often show a polarity imparting a wart-like appearance to the lesion at low power. The individual lobules consist of peripheral layers of small basaloid germinative cells (usually more than 2 layers) while more centrally are found maturing sebocytes having multivacuolated cytoplasm and scalloped nuclei. Cytologic atypia and abnormal mitoses are absent. None of our 5 cases had such a multinodular architecture or epidermal connection and in all cases immature germinative cells predominated over mature sebocytes, thus limiting the differential diagnosis to sebaceoma or sebaceous carcinoma. We would also point out that the extreme view of Dr Ackerman as cited by Drs Kramer and Chen that all sebaceous adenomas are in fact carcinomas is not shared by many, us included. Using the criteria outlined above, we have collectively diagnosed over 200 sebaceous adenomas over the last 15 years and in no case did the follow-up suggest a clinically malignant behavior. Dmitry V. Kazakov, MD, PhD Sikl's Department of Pathology, Charles University, Medical Faculty Hospital, Pilsen, Czech Republic Heinz Kutzner, MD Dermatohistopathologische Gemeinschaftspraxis, Friedrichshafen, Germany Dominic V. Spagnolo, MMBS Division of Tissue Pathology, PathWest Laboratory Medicine WA, Nedlands, WA, Australia Arno Rütten, MD Dermatohistopathologische Gemeinschaftspraxis, Friedrichshafen, Germany Petr Mukensnabl, MD Michal Michal, MD Sikl's Department of Pathology, Charles University, Medical Faculty Hospital, Pilsen, Czech Republic