Defining and Grading Melanocytic Atypia on Reflectance Confocal Microscopy: A Survey of Confocalists in the United States and Abroad.

To the Editor: Traditionally, histopathology has been the gold standard for identifying and grading dysplastic nevi. The Duke grading criteria are a set of architectural and cytologic features objectively outlining atypia in histopathology of melanocytic nevi, proposed and validated 2 decades ago and widely used today.1 However, a population-based study by Lott et al2 found that of pigmented lesions biopsied, 83.1% are benign or mildly dysplastic (requiring no reexcision). In vivo reflectance confocal microscopy (RCM) stands to bridge the gap between clinical dermatology and histopathology, improving clinical decision making. Melanocytic tumors are one major category of skin lesions that are particularly well suited for noninvasive diagnosis by RCM, owing to the strong endogenous contrast that melanin provides.3 Although the accuracy of this modality has been widely demonstrated for melanoma,4 in vivo RCM may also be used to discriminate between benign nevi and dysplastic nevi, and between dysplastic nevi and melanoma (in situ and invasive).5,6 In this way, RCM can be used as a standalone in vivo tool to guide management decisions to prevent unnecessary biopsies of benign nevi, as well as augment traditional histopathology to help differentiate between a melanoma in situ and a severely dysplastic nevus.5,6 Confocal features correlating with the Duke criteria were introduced by Pellacani et al in 2012.5 RCM demonstrated success in noninvasively predicting histopathologic dysplasia based on architecture and cytology. Although RCM features directly corresponding to histological parameters for architectural atypia were defined, the correlates for cytology were not as specific. Features representing cytologic atypia on histopathology—absence of round/oval and euchromatic nuclei, nuclei greater in size than basal keratinocytic nuclei, and cell diameter greater than twice the size of a basal keratinocyte nucleus—broadly correlated with “atypical cells at the dermoepidermal junction (DEJ)” on RCM.5 Generally, the authors described atypical cells at the DEJ as either roundish or dendritic and having a bright cytoplasm, dark nucleus, clearly defined outline, and irregularity in size and shape (Fig. 1).5 In addition, the grading of cytologic atypia involves “no atypical cells at DEJ” for absent/mild atypia, “mildly atypical cells at DEJ” for moderate atypia, and “moderately to severely atypical cells at DEJ” for marked atypia.5 Because RCM is an evolving specialty where readers have varying degrees of experience, we suspected that confocal readers’ threshold for identifying and grading melanocytic atypia would also vary. We surveyed 57 confocal readers in the United States, Europe, South America, and Australia to determine how they define, grade, and manage dysplastic nevi using RCM. We received 28 responses (11 in the United States and 17 outside the United States), for a response rate of 49%. Training background included dermatology, pathology, and general practice, with some having completed fellowships in dermatopathology, procedural dermatology, and reflectance confocal microscopy. Experience reading confocal images ranged from 1 year to 15 years (mean 6.89 years) (Fig. 2A), and the readers’ monthly confocal case load ranged from 1 to 200 (mean 43.11 cases per month) (Fig. 2B). The majority of respondents used both cytology and architecture equally when determining melanocytic atypia (71.4%). Most considered “size, number, and pleomorphism equally” to determine cytologic atypia (89.3%), and used “cell diameter larger than twice the size of a basal-layer keratinocyte nucleus” as the threshold for atypical size (89.3%). However, respondents could not agree on a minimum number of atypical cells per 1 mm2 to be considered atypical (Fig. 3). In fact, the most common type of answer (35.7%) did not