Portal de Periódicos da CAPES

Vulvar Toker Cells: The Long-Awaited Missing Link

2006; Lippincott Williams & Wilkins; Volume: 28; Issue: 1 Linguagem: Inglês

10.1097/01.dad.0000194052.65695.f1

1533-0311

И. Э. Белоусова, Dmitry V. Kazakov, Michal Michal, Saul Suster,

Cancer and Skin Lesions

To the Editor: The histogenesis of extramammary Paget disease (EMPD) has long been a subject of debate and controversy. Nowadays, it is generally agreed that at least two forms of the disease can be delineated, namely primary and secondary, with the latter representing secondary involvement from an underlying carcinoma originating most commonly in the lower gastrointestinal or urinary tract. As for the primary disease, skin adnexa (eccrine or apocrine glands), ectopic mammary glands, or pluripotential germinative cells in the epidermis and other structures have been implicated as a possible source of neoplastic cells. The recent description of vulvar Toker cells in the Journal by Drs. Willman, Golitz, and Fitzpatrick1 offers an attractive analogy with the breast, where a similar condition is referred to as mammary Paget disease (MPD). In this location, over 95% of cases of MPD are associated with an underlying ductal carcinoma,2 and the rest is thought to arise from Toker cells described in 1970.3 The same seems to be true for rare cases of Paget disease involving ectopic breast adjacent to a supernumerary nipple.4 We wonder why the same approach could not be applied for the anogenital area and the vulva in particular. This anatomic site contains structures that were originally variously called ectopic breast tissue, supernumerary breast tissue, and most recently, interpreted as anogenital mammary-like glands (MLG).5,6 These have been shown to be a normal constituent of this area, and lesions involving these structures have been demonstrated to manifest a striking homology with those seen in the breast.6 In this regard, the following question arises: why should EMPD differ from MPD histogenetically, if all anatomic conditions are so similar? Below are some lines to support the analogy. It is known that primary EMPD (without underlying carcinoma of internal organs) affects most often vulvar and perianal areas, where anogenital MLG are found in highest concentrations.6 There have been several documented cases of EMPD with an underlying ductal carcinoma involving anogenital MLG,7-9 and this is also in accordance with our experience (Fig. 1). EMPD has a notorious propensity for recurrences, and this can be partly explained by the presence of neoplastic cells residing in anogenital MLG that remained undetected after the incomplete removal of the superficial part of the neoplasm.FIGURE 1: The specimen came from a 79-year-old woman with a 7-year history of recurrent extramammary Paget disease in the vulva requiring several surgical interventions and, finally, the creation of a neovagina. A, Suprabasal spread and clefts of Paget cells, focally forming discrete glandular structures can be seen in the epidermis. B, In the deeper tissue, a ductal carcinoma of mammary-like glands with a predominantly solid growth pattern was found.Thus, logic dictates that EMPD in the anogenital area could be approached similarly to that in the breast, and accordingly can be classified likewise, with some modifications considering the close vicinity of other organs. EMPD as a primary intraepidermal neoplasm with or without invasion (the possible origin-Toker cells) EMPD secondary to: (a) carcinoma of anogenital mammary-like glands EPMD secondary to: (a) carcinoma of skin appendages; (b) carcinoma of Bartholin glands; (c) carcinoma of other internal organs (ie, anus, rectum, urethra, cervix) This approach would have certain clinical implications. First, it is clear that a superficial biopsy in the affected area where anogenital MLG are found does not suffice to exclude EMPD associated with an underlying carcinoma involving these glands. Second, "clear" epidermal and the lower resection margins for such a neoplasm do not necessarily mean its complete excision, and therefore thorough sampling of the removed tissue is warranted. Unfortunately, we cannot offer a recommendation for how deep pathologic sampling or surgical excision must be performed because the depth to which MLG normally extend has not yet been defined. Until this is done, it is perhaps best to use the approach summarized by Kurman.10 Malignant Paget cells appear to be able to grow deep downward along the excretory ducts of MLG (Fig. 2), and this situation may be under recognized. If this is confirmed in large studies, the validity of the terms EMPD in situ and minimally invasive EMPD may be questioned, because neoplastic cells growing within ducts of MLG (and within eccrine and apocrine ducts,11,12Fig. 3) can reach deeper tissue while remaining clinically occult and pathologically undetected.FIGURE 2: This specimen came from a 47-year-old woman with vulvar Paget disease. The epidermis along the whole surface of the specimen (3 cm) was infiltrated by Paget cells, and there were no foci of invasion of the neoplastic cells through the epidermal basal membrane. A, The margin of the lesion: abnormal mammary-like glands (MLG) can be seen. B, No carcinoma of MLG themselves was found in this case; only atypical ductal hyperplasia as depicted in B, and changes compatible with columnar cell change and oxyphilic (apocrine) metaplasia as depicted in C. C, Note plentiful Paget cells in the glands, confirming that tumor cells can spread both upward and downward in EMPD.FIGURE 3: The same case as depicted in Figure 1. In addition to the ductal carcinoma of mammary-like glands (MLG), excretory ducts of MLG were involved in this case. Furthermore, Paget cells were found in apocrine or eccrine ducts in the same sections as illustrated here. Apocrine and eccrine ducts cannot be distinguished one from another, but both are different from ducts of MLG in that they lack a peripheral layer of myoepithelial cells. With time such a process of cancerization of sweat glands can lead to the appearance of an invasive carcinoma.We fully realize that the proposed scheme is not perfect, and may rather serve for academic purposes. For example, primary EMPD originating from transformed Toker cells with descent into an underlying anogenital MLG can be nearly or fully indistinguishable from that originating deep in the glands and involving the epidermis secondarily. In the breast, rare carcinomas associated with Paget disease have been found to arise from ductal epithelial at or near the squamocolumnar junction,2 and the same can be true for the skin, making the segregation of the categories 1 and 2 less sharp. Further, a carcinoma originating from adnexal structures may theoretically exhibit a pagetoid spread in the epidermis.13 This is said to rarely occur in the breast,2 but in the anogenital areas, according to the experience of others11,12 and our own, adnexa rather serve as a pathway for carcinoma spread (including carcinomas primarily arising in anogenital MLG), and not as its source (Fig. 3). Involvement of pilosebaceous units or eccrine ducts by Paget cells in EMPD has been found in 83% and 61%, respectively.14 In fact, we are not aware of a single case of adnexal carcinoma in the anogenital area in our files, apocrine or eccrine, with a pagetoid spread into the epidermis. Adenocarcinoma of Bartholin gland associated with Paget disease is extremely rare, with 2 cases reported in the literature,15,16 and the Bartholin gland origin of such carcinomas is evident clinically. Thus, point 3a and 3b can be safely "ignored" in most cases in diagnostic dermatopathological practice, leaving a pathologist only with the 3 main possibilities. In closing, we would like to congratulate Drs. Willman, Golitz, and Fitzpatrick for their discovery of Toker cells in the vulva, which have long been the missing link, obviously so important for our understanding of the histogenesis of EMPD. Irena E. Belousova, MD, PhD Dmitry V. Kazakov, MD, PhD† Michal Michal, MD† Saul Suster MD‡ *Department of Dermatology Medical Military Academy Saint-Petersburg, Russia, †Sikl's Department of Pathology Charles University Medical Faculty Hospital Pilsen, Czech Republic, ‡Department of Pathology Ohio State University Medical Center Columbus, OH