Pantea Hashemi, MD; Grace F. Kao, MD; Thomas Konia, MD; Lisa C. Kauffman, MD; Christine C. Tam, MD; Bahram Sina, MD
Dr. Hashemi is from the Department of Dermatology, New York Medical College, New York. Dr. Kao is from VA Maryland Health Care System, Baltimore. Dr. Konia is from University of California at Davis Medical Center, Sacramento. Dr. Kauffman is in private practice, Washington, DC. Dr. Tam is from the Department of Dermatology, University of Cincinnati, Ohio. Dr. Sina is from the Department of Dermatology, University of Maryland Medical Center, Baltimore.
The authors report no conflict of interest.
Correspondence: Pantea Hashemi, MD, 1901 1st Ave, New York, NY 10029 (panteahashemi@gmail.com).
Figure 4. Histopathologic examination of clinically unaffected skin of the periumbilical region revealed Toker cells with round nuclei and pale cytoplasm (arrows) scattered in the lower epidermis (H&E, original magnification ×400).
Toker cells, which were first described in 1970,11 have been recognized as precursors to EMPD.12-14 Toker cells are intraepithelial cells with clear to pale-staining cytoplasm that are smaller in size than Paget cells but larger than neighboring keratinocytes. They are found in approximately 10% of normal nipples.11,14 Toker cells show vesicular chromatin, whereas Paget cells are hyperchromatic with prominent nucleoli.12 Willman et al15 examined 11 vulvectomies for the presence of Toker cells in association with mammarylike glands of the vulva. They demonstrated the presence of Toker cells in 4 (36%) of the samples.15 Additionally, Van der Putte et al16 observed Toker cells in an areolar lesion in a 47-year-old woman with MPD without underlying adenocarcinoma, suggesting that cases of MPD and EMPD confined to epithelial cells may be derived from Toker cells. Toker cells have been associated in the pathogenesis of 2 other benign entities, including clear cell papulosis17,18 and cutaneous hamartoma with pagetoid cells.19
Primary EMPD has the potential to develop in several regions of the skin that contain apocrine glands. Our patient presented with persistent genital lesions for many years without any concerns of axillary disease. Interestingly, biopsies from normal-appearing skin in the periumbilical region revealed clear cells (Figure 4). Likewise, Toker cells have been described in biopsies from clinically unaffected skin of the axillae in some cases where EMPD was identified in the genital regions.3,5 Genital lesions were the main clinical presentation and preceded axillary involvement in most instances.3,5,9 Therefore, biopsies from uninvolved apocrine sites (ie, axillary and periumbilical skin) should be considered in these patients. Based on our observation, we speculate that multicentric primary EMPD starts with Toker cell hyperplasia with a propensity to evolve into neoplasia in sites with apocrine or mammarylike glands.
Primary and secondary EMPD cannot be distinguished by histopathology and immunohistochemistry has a limited role.20,21 However, immunostaining with CDX2 (a caudal-type homeobox protein) might be helpful in differentiating primary EMPD from secondary EMPD extending from underlying anorectal adenocarcinomas.22 Primary EMPD has been treated with surgical excision.2,3,8 Despite the high recurrence rate after surgery, the prognosis for localized primary EMPD disease has been favorable.
Conclusion Our case suggests that Toker cell hyperplasia is a precursor to primary EMPD. In patients with established EMPD, other apocrine gland–bearing areas should be examined for multicentric disease. Lastly, the clinical course in our patient supports the hypothesis that primary multicentric EMPD has a favorable outcome. The Table lists the characteristics of MPD and EMPD and describes their development with respect to Toker cells. More studies are required to further outline the cytologic characteristics of Toker cells and to distinguish primary EMPD from secondary EMPD.
