To the Editor: A 50-year-old White man was diagnosed with invasive extramammary Paget disease (EMPD) on the left inguinal fold. Excisional biopsy revealed an intraepithelial (in situ) and subjacent intradermal/subcutaneous neoplasm composed of single, nests and cords of cells containing pale staining/vacuolated/foamy cytoplasm and large vesicular nuclei (Figs. 1A–C). Focal glandular/pseudoglandular differentiation was noted. There was intracellular cytoplasmic/interstitial positivity for PAS and Alcian blue (Fig. 1D). By immunohistochemistry, tumor cells were strongly reactive for AE1/AE3, Cam 5.2, CK7, CEA, and GATA-3, with focal nuclear reactivity for estrogen receptor and strong (3+) membranous expression for Her-2/neu. Tumor cells were negative for p63, D2-40, CK20, villin, CDX2, TTF-1, prostate-specific antigen, and progesterone receptor, in addition to S-100, HMB-45, SOX10, and MITF. The tumor showed diffuse intracytoplasmic grainy positivity for melan-A (A103 clone) (Figs. 2A–F).FIGURE 1.: In situ and invasive components of EMPD (A–C). Tumor cells show cytoplasmic positivity for Alcian blue (A, B, original magnification ×40; (C) original magnification ×100; and (D) original magnification ×200).FIGURE 2.: By immunohistochemistry, tumor cells are positive for CK7 (A) and negative for p63 (B) and S-100 (C). There is diffuse intracytoplasmic grainy immunoreactivity for melan-A (A103 clone) (D–F) (A–D, original magnification ×40; (E) original magnification ×100; and (F) original magnification ×200).Melan-A (A103 clone) and MART-1 (melanoma antigen recognized by T cells) (M2-7C10 clone) antibodies recognize the same gene product, a melanosome-associated protein that is involved in melanosome biogenesis by ensuring the stability of GPR143.1 This protein plays a vital role in the expression, stability, trafficking, and processing of melanocyte protein PMEL, which is critical to the formation of stage II melanosomes.1 Both antibodies are widely used in dermatopathology and expressed by normal melanocytes, nevi, and melanoma.2 Melan-A and MART-1 staining are also well-known observations in clear cell sarcoma and other nonmelanocytic tumors that contain premelanosomes, including perivascular epithelioid cell tumor (PEComa), angiomyolipoma, lymphangioleiomyomatosis, and sugar tumor of the lung.1,2 Furthermore, melan-A but not MART-1 expression has been documented in steroid-synthesizing normal cells and their respective benign and malignant tumors of the adrenal cortex and gonads (which lack premelanosomes).1 These findings suggest a lower specificity for premelanosomes of the melan-A antibody as compared to MART-1. Melan-A positivity [in addition to expression of other melanogenic markers (ie, HMB45)] has been reported in melanotic Xp11-associated tumor of the gastrointestinal and genitourinary tracts.3 With reference to cutaneous pathology, reactivity for both melan-A and MART-1 antibodies has been reported in pigmented (melanotic) and nonpigmented variants of diffuse neurofibroma, in addition to melanotic schwannomas.4,5 Melan-A and MART-1 expression was identified in 2 cases of atypical fibroxanthoma and one case of undifferentiated pleomorphic sarcoma of the skin.1 Staining was reported as multifocal rather than diffuse and almost exclusively within large multinucleated tumor cells, although one case showed immunoreactivity within mononucleated epithelioid cells.1 Smith-Zagone et al6 also identified melan-A expression (in addition to HMB-45) limited to large, multinucleated cells with vacuolated cytoplasm in atypical fibroxanthoma. More recently, immunohistochemical expression of melanocytic markers (melan-A, HMB45, and S-100) has been reported in cases of mammary Paget disease and EMPD.7,8 Eftimie et al7 highlighted aberrant melan-A expression in the pagetoid (in situ) cells in EMPD. Our case of EMPD demonstrated diffuse reactivity for melan-A in both intraepithelial and invasive tumor cells. Cross-reactivity of melan-A antibody with an unrecognized epitope within tumor cells has been proposed as the underlying pathogenetic mechanism in nonmelanocytic skin neoplasms in previous reports.1,4–8 Of note, melan-A antibody cross-reactivity has also been identified in molluscum contagiosum bodies.9 This phenomenon likely accounts for the finding of melan-A reactivity in our case of EMPD. Another possibility is that immunoreactive melanosomes/melanin granules produced by stimulated melanocytes in the epidermis might be transported into the cytoplasm of adjacent in situ EMPD cells, although this is unlikely to explain the diffuse melan-A positivity noted throughout the invasive component of our tumor. “False-positive” melan-A expression in EMPD demonstrates the need for caution in microscopic interpretation and for using multiple immunohistochemical stains (including other markers of melanocytic differentiation) to corroborate clinical and histopathological features and ensure accurate diagnosis. This report adds to the growing literature of aberrant melan-A expression in cutaneous nonmelanocytic tumors.