Very few histologic reports describe normal melanocytes of the nail unit. Previous studies predominantly address the distal nail matrix melanocytes; we found no review of nail-bed melanocytes in the literature. The proximal nail matrix melanocytes are difficult to identify; the cells cannot be identified by L-DOPA staining. More recently, their scarcity was confirmed by immunohistochemistry with a large panel of antibodies directed against melanocytes. We wished to detect the proximal nail matrix dormant melanocytes and compare their density and distribution with that of the other melanocytes in the distal matrix and nail bed and to establish criteria of normality that may help clarify the pathologic features of benign nevoid melanonychia in the nails of whites. A panel of five monoclonal antibodies (MoAbs), including HMB45 and TRP1 directed against antigens localized in early melanosomal vesicles, was investigated in frozen sections of six nail specimens from whites. Both vertical and horizontal sections were assessed to determine the presence of dormant melanocytes. Results showed that the proximal nail matrix melanocytes were clearly identified with MoAbs HMB45 and tyrosinase-related protein-1 (TRP-1). By contrast, melanocytes stained by MoAb against tyrosinase and L-DOPA reaction were evident, especially in the distal matrix. With MoAb TRP-1, the epithelial sheets showed counts of approximately 217+/-84/mm2 in the proximal matrix and of 132+/-34/mm2 in the distal matrix; the nail bed counts were only 45+/-25/mm2. The split epithelial sheets had 103+/-17/mm2 L-DOPA-positive melanocytes in the distal third of the matrix, but only a few of them were detected in the proximal matrix and none were noted in the nail bed. We clearly identified proximal nail melanocytes using MoAb HMB45 and TRP1. The total number of matrix melanocytes can be estimated as approximately 217/mm2. In proximal matrix, the dormant melanocytes compartment was predominant. In the distal matrix, two compartments were identified: a functionally differentiated and a dormant compartment. Contrary to classical opinion, longitudinal melanonychia originates more frequently in the distal matrix, not secondary to the larger melanocyte density but because only the distal matrix contains an active melanin synthesis compartment. Furthermore, the superficial distribution of proximal nail melanocytes in vertical sections showed a histologic feature that may simulate the pagetoid pattern of melanoma in situ.

Previous studies of the density of melanocytes in the normal nail bed have had conflicting results. This is unfortunate because knowing the normal values might help the difficult distinction between a benign subungual melanotic macule and an early melanoma in situ. Five specimens of normal nail unit were analyzed. On hematoxylin and eosin-stained sections the melanocytes were undetectable. We defined the melanocyte count (MC) as the number of melanocytes per 1-mm stretch of nail epithelium. The mean MC for nail matrix was 6.86 with a range of 4-14. The melanocytes were irregularly scattered in the basal and suprabasilar layer of the matrix epithelium. Abundant and uneven cytoplasmic dendrites were focally observed in the matrix. The MC for the nail bed ranged from 0 to 5 with a mean of 0.43. The melanocytes were restricted to the basal layer with thin cytoplasmic dendrites. Two cases showed a complete absence of melanocytes in the nail bed. In the ventral portion of the proximal nail fold, called the eponychium, the MC ranged between 0 and 5/mm with a mean of 2.27/mm. In conclusion, we discovered foci in normal nail beds, in which the melanocytic density can be relatively high and reach the level seen in the matrix. HMB45 is more sensitive than Microphtalmia-associated transcription factor (MITF) for the evaluation of intraepithelial melanocytic density of the nail unit. If MITF is used alone in the nail bed, its weak sensitivity may result in a false-negative interpretation and may be wrongly reassuring in the evaluation of early melanomas. On hematoxylin and eosin sections, basal and suprabasal nail keratinocytes are sometimes crowded, showing oval or elongated dark-staining nucleus and a clear cytoplasm and mimics a melanocytic proliferation. On HMB45 or Melan A staining, the morphology and the distribution of the dendrites of matrical melanocytes can mimic the dendritic pattern usually described in acral melanoma. Therefore, the interpretation of nail melanocytic atypia must be prudent.

An intrathyroid primary epithelial spindle-cell tumor with mucous cysts is described in a 9-year-old child. Histologically, this well-circumscribed tumor exhibited a nodular pattern, a prominent spindle cell component with minimal pleomorphism, and well-differentiated mucinous glands within fibrous bands. The spindle cells demonstrated diffuse immunopositivity for cytokeratin and vimentin. Electron microscopy of tissue sections demonstrated that cells contained bundles of cytoplasmic tonofilaments and numerous desmosomes. The light and electron microscopic features and immunohistochemical profile of this tumor were similar to those of recently described thyroid tumors that have been called "SETTLE" tumors (i.e., spindle epithelial tumor with thymus-like differentiation). These uncommon tumors can be considered intrathyroid thymoblastomas and must be regarded as potentially malignant lesions.

This proof-of-concept study shows that infrared microimaging could help in the diagnosis of primary cutaneous melanoma.