Local invasion in endocrine mucin‐producing sweat gland carcinoma: Still an open issue

Abstract

Endocrine mucin-producing sweat gland carcinoma (EMPSGC) is a rare cutaneous adnexal adenocarcinoma with mucinous and neuroendocrine differentiation, defined by a constellation of histopathologic features: solid growth pattern with expansile adjacent nodules on low-power examination; relatively homogenous cytomorphology of neoplastic cells; and presence of a delicate intratumoral vascular network, betraying the papillary growth pattern of this neoplasm. Additional histopathological findings in support of a diagnosis of EMPSGC include production of mucin, occasional spindling of malignant cells, and formation of perivascular pseudorosettes. Based on this combination of features, ever since as its original description EMPSGC has been regarded as the cutaneous counterpart of solid papillary carcinoma of the breast (SPC; previously referred to as “endocrine ductal carcinoma in situ”), a malignant papillary lesion with good prognosis. Immunohistochemical (IHC) data have confirmed the close biological relationship between EMPSGC and SPC, as indicated by identical immunophenotype, including expression of neuroendocrine markers. Regardless of the legitimate question as to whether EMPSGC should be simply termed “cutaneous SPC” (in line with current nomenclature of cutaneous counterparts of mucinous carcinoma and mammary secretory carcinoma of the breast, respectively), we would like to address the critical issue of how to evaluate local invasion in EMPSGC. Multiple studies, including a recent one by Held et al in this journal, appear to support the initial finding by Zembowicz et al: most EMPSGC cases seem to be characterized by IHC loss of peripheral myoepithelial cells, at least around the largest neoplastic nodules, as showed by staining for p63, CK5/6, and/or alpha-smooth muscle actin. As a result, several authors have interpreted this finding as definitive proof that EMPSGC is, in most cases, an invasive malignancy; such a view, which is based on the concept that IHC evidence of preservation of a distinct myoepithelial layer is the ultimate determinant in assessing invasion in apocrine carcinomas, appears to be at odds with the seemingly benign morphology of EMPSGC nodules, characterized by rounded margins with expansile growth. Other authors, including recently Qin et al in this journal, have limited themselves to state that EMPSGC represents an intermediate step in the progression from “eccrine cyst” to invasive mucinous carcinoma, falling short of addressing the issue as to whether EMPSGC itself (a true carcinoma, as its name implies) should be regarded as an invasive neoplasm or not. We would like to point out that IHC staining of myoepithelial cells should always be interpreted with caution because of its well-known, intrinsic shortcomings. As already extensively reviewed by Dewar et al in spite of the numerous IHC markers available for staining of myoepithelial cells in the context of apocrine neoplasms (including muscle-specific actin, smooth-muscle-myosin heavy chain, p63, p40, CD10, calponin, and high-molecular-weight cytokeratins, among others), none has ever showed 100% sensitivity or specificity. Alphasmooth muscle actin, for instance, may cross-react with stromal myofibroblasts (which by contrast are negative for p63 and high-molecular-weight cytokeratins), leading to the false perception of a preserved myoepithelial lining. On the other hand, use of p63 is partially hindered by the fact that, being a nuclear marker, its staining pattern may appear discontinuous, in certain circumstances resulting in the erroneous impression that the myoepithelial lining is absent. Accordingly, multiple IHC markers should always be employed in order to carefully assess the preservation of the myoepithelial layer in difficult cases. Unfortunately, no such ad hoc study, rigorously employing a comprehensive panel of IHC myoepithelial markers, has been conducted to date in EMPSGC. We believe that, in the absence of more cogent scientific evidence, and in light of the close biological similarities between EMPSGC and SPC, the assessment of local invasion in EMPSGC should be performed following the same rationale and criteria as currently done for mammary SPC. Indeed, an analogous debate developed in the last two decades among breast pathologists, as to whether all SPC cases exhibiting IHC loss of peripheral myoepithelial cells in one or more nodules should be classified as invasive carcinoma. This issue was clarified in the fourth edition of the World Health Organization (WHO) classification of breast tumors, in which it is univocally stated that SPC cases presenting as rounded nodular masses should be interpreted (and reported, for staging purposes) as in situ disease, irrespective of IHC loss of myoepithelial cells. By contrast, according to the 2012 WHO classification of breast tumors, SPC should be considered as truly invasive only in presence of infiltrative, jagged neoplastic tongues, unambiguously disrupting the contours of tumor nodules, and inducing a stromal desmoplastic reaction. This approach has been further supported by subsequent clinicopathological studies and has also been adopted for encapsulated papillary carcinoma, another subtype of papillary carcinoma of the breast with rather indolent behavior and favorable prognosis. In our view, Received: 29 October 2019 Revised: 10 November 2019 Accepted: 18 November 2019