Abstract
BACKGROUND: Lentigo maligna (LM) and lentigo maligna melanoma (LMM) are difficult to manage because of their subclinical extension and ill-defined margins, especially on chronically sun-damaged facial skin. Reflectance confocal microscopy (RCM), a non-invasive imaging technique with near-histological resolution, has emerged as a valuable adjunct for diagnosis, presurgical margin mapping, and postoperative surveillance. OBJECTIVE: To systematically review current evidence on the role of RCM in LM and LMM management. METHODS: A systematic review was performed in accordance with PRISMA 2020 guidelines. PubMed, Embase, and Scopus were searched up to July 2025. Eligible studies assessed RCM for diagnostic accuracy, presurgical mapping, surgical outcomes, or follow-up. Findings were synthesized qualitatively. RESULTS: Twenty-nine studies were included. Across diverse study designs, RCM demonstrated higher diagnostic accuracy than dermoscopy for LM, with reported sensitivities ranging from 85 % to 100 % and specificities from 71 % to 97 %. In comparative studies and meta-analyses, RCM showed superior specificity compared with dermoscopy and improved discrimination between LM and early invasive LMM. Presurgical RCM margin mapping was associated with higher first-stage clearance rates (up to 63.6 % vs. 34.7 % with standard assessment), fewer surgical stages (mean 1.1-1.2 vs. 1.7-1.9), narrower excision margins, and lower local recurrence rates (approximately 2-5 % vs. up to 13 % in non-RCM series). RCM was also effective for monitoring non-surgical therapies and detecting early recurrences within surgical scars. Major barriers to widespread adoption include cost, limited availability, training requirements, and the absence of standardized protocols. CONCLUSIONS: RCM strengthens the management of LM and LMM by improving diagnostic confidence, enabling precise margin mapping, and supporting long-term surveillance. Broader adoption will depend on standardization of mapping techniques, integration into clinical workflows, and validation of artificial intelligence-assisted interpretation.