Clinical-histological and genetic features of rare variants of melanoma.

Melanoma is traditionally classified into a limited number of common clinicopathologic subtypes; however, routine clinical and histopathologic practice reveals a broad spectrum of rare variants and unusual morphologic patterns that pose significant diagnostic and therapeutic challenges. These uncommon melanomas may closely mimic benign melanocytic proliferations or non-melanocytic neoplasms, present with subtle or atypical clinical and dermoscopic features, and exhibit distinctive biologic behavior and treatment responses. Consequently, their clinical impact is disproportionate to their low incidence. This narrative review provides an updated and integrated overview of rare primary melanoma subtypes, including desmoplastic melanoma, nevoid melanoma, Spitz melanoma, melanoma ex blue nevus, and primary dermal melanoma, as well as uncommon histologic variants such as rhabdoid, metaplastic (osteogenic and chondrogenic), myxoid, small cell, balloon/clear cell, signet-ring cell, and plexiform melanoma. In addition, rare combined cutaneous tumors containing both melanomatous and epithelial components are discussed. For each entity, we summarize current evidence on epidemiology, clinical and dermoscopic presentation, histopathologic features, immunohistochemical profiles, and molecular alterations, with particular emphasis on key diagnostic pitfalls and differential diagnoses. Advances in molecular pathology have refined the classification of these rare melanomas, enabling more precise integration of morphology, immunophenotype, and genomic data, with direct implications for prognosis, staging, and therapeutic decision-making. Finally, we review available data on management strategies, including surgery, sentinel lymph node biopsy, radiotherapy, targeted therapy, and immunotherapy. By synthesizing current knowledge and highlighting practical diagnostic and therapeutic considerations, this review aims to assist pathologists and clinicians in navigating the "grey zones" of melanoma diagnosis, ultimately supporting more accurate classification and more personalized patient care.