To the Editor: Recently, we observed a polypoid lesion with surface ulceration (Fig. 1) on the right cheek of a 78-year-old man. The tumor was an intradermal neoplasm which involved only the superficial dermis and which was associated with a lateral collarette. The lesion consisted of large polygonal clear cells with bizarre nuclei (Fig. 1). Immunostains for pan-keratin, S100 protein, HMB45, and C-KIT were negative. The neoplastic cells showed diffuse cytoplasmic staining for vimentin, CD68, muscle-specific actin, α1-antitrypsin, and diffuse membranous staining for CD99 and CD10 (Fig. 2C–F). Mitoses were rare (less than 1 × 25 hpf). Proliferation index valuated with Mib-1 antibody was 25%. periodic acid-Schiff stain failed to demonstrate glycogen in tumor cells. Our histological diagnosis was “Clear cell cutaneous neoplasm suggestive for atypical fibroxanthoma (AFX) or distinctive clear cell mesenchymal neoplasm (DCCMN) with atypical features.” Dermal clear cell tumors comprised primarily clear cell adnexal lesions, balloon cell melanocytic lesions, metastatic clear cell carcinomas, clear cell AFXs, and DCCMNs. Histologically clear cell AFX showed a proliferative pattern similar to DCCMN. The differential diagnosis between these tumors is problematic. DCCMN was first described by Lazar and Fletcher1 as a dermal tumor composed of an homogeneous population of clear cells believed to be of mesenchymal origin. All 5 cases were present in the lower extremities. One case exhibited atypical features in the form of nuclear pleomorphism and mitotic activity. Immunohistochemical staining showed the neoplastic cells to be consistently positive for NKI-C3 and to be variably positive for CD68 and vimentin. McAlhany and LeBoit2 described 5 similar lesions with strong positivity for NK1-C3 and nuclear MiTF staining. CD68 and S100 protein were variable positive. Four cases showed focal HMB-45 expression. These authors2 believed that DCCMN might be related to “PEComas”. Later, Gavino et al3 described one case of DCCMN arising in the scalp with similar histological features to case 5 described in Lazar and Fletcher'series.1 Gavino et al3 considered these atypical features to be sufficient to justify a new pathological entity defined “atypical DCCMN”. It is evident that 2 cases are too few to permit the definition of a new histopathological entity. Gavino et al3 believed that these neoplasms could be distinguished from clear cell AFX might because “atypical DCCMN” showed to stain positively with NKI-C3 and vimentin and failed to stain with the histiocytic (CD68 and CD163), melanocytic, epithelial, and lymphoid markers. Positivity for NKI-C3, CD68, and vimentin and negativity for melanocytic, epithelial, and lymphoid markers are considered typical immunohistochemical markers for DCCMN by Lazar and Fletcher.1 The presence of NKI-C3 in 11 of 12 AFX reported by Chan et al,4 the positivity for HMB-45 and S100 protein in 4 of 5 cases described by McAlhany and LeBoit,2 the negativity for CD68 in the atypical DCCMN of Gavino et al,3 the HMB-45 expression in a clear cell AFX,5 the variable expression of muscle immunohistochemical markers in reported clear cell AFXs (Table 1)5–12 demonstrated that immunohistochemical pattern should not be used as diagnostic criteria in the differentiation between atypical AFX and DCCMN. The precise line of differentiation and the histogenesis of clear cells remain uncertain. The histological pattern may be indistinguishable. Clinical presentation (location in sun-damaged areas of the head and neck in an elderly individual) may support a histopathological diagnosis of clear cell AFX, but this tumor has been reported in the extremities. Consequently the origin site should not be considered as a definitive diagnostic criteria. It is possible that 2 neoplasms may be an unique pathological entity and future studies may elucidate the different immunohistochemical pattern. We prefer the denomination of “clear cell cutaneous neoplasm suggestive for AFX or DCCMN with atypical features”. We specify in the comment of histological report that clear cell adnexal tumors, balloon melanocytic lesions, clear cell metastatic carcinomas, and xantomatous lesions has been excluded. The immunohistochemical pattern, the presence of atypical features, and the number of mitosis per hpf are reported. The lesion should be considered of unknown biological potential.1FIGURE 1: Scanning magnification shows an exophytic polypoid lesion with surface ulceration (insert), composed by neoplastic cells of varying sizes with clear to reticulated cytoplasm and enlarged, vesicular nuclei with prominent nucleoli. Markedly atypical neoplastic cells are scattered throughout the nodule.FIGURE 2: Immunohistochemical stains show diffuse cytoplasmic positivity for vimentin (C), muscle specific actin (D), and diffuse membranous staining for CD99 (E), and CD10 (F).TABLE 1: Review of Immunohistochemical Pattern of Clear Cell Atypical Fibroxantoma