INTRODUCTION: Conventional histopathology has been considered as the gold standard in the diagnosis of cutaneous malignant melanoma, despite the progress of molecular biology and immunohistochemistry. There are many microscopic criteria for diagnosis of melanoma, however there is not a single one that can be useful to define malignancy. AIM: Our purpose is to analyse the criteria considered more important to the diagnosis of melanoma, comparing their presence in benign melanocytic lesions and melanomas. MATERIAL AND METHODS: We studied 33 benign melanocytic lesions (Spitz nevi, 13; Reed nevi, 6; dysplastic nevi, 6; congenital nevi, 3; acquired nevi, 3; combined nevus, 1; recurrent nevus, 1) and 101 extensive/superficial melanomas (25 in situ and 76 invasive up to 2 mm thickness). RESULTS: Some criteria showed high frequency in benign lesions, showing low-specificity, while others had low-positivity in the benign and high-frequency in malignant lesions, consequently high-specificity and greater importance in the melanoma diagnosis. CONCLUSION: The five criteria that presented statistically significant difference after comparison with benign lesions were: 1. linear proliferation of a single layer cells in basal layer; 2. single cells at periphery; 3. cells in the granular layer; 4. extensive pagetoid array; and 5. large, irregular and/or multiple nucleoli . Thin melanomas usually do not show many of important criteria, like lack of cell maturation, necrosis and deep mitoses.
Melanoma; Histopathological criteria; Differential diagnosis; Spitz nevus
