PRAME immunohistochemistry can distinguish melanocytic pseudonests of lichenoid reactions from melanoma in situ

Abstract

AbstractBackgroundDistinguishing melanocytic pseudonests encountered in lichenoid dermatoses or lichenoid keratoses from melanoma in situ (MIS) with brisk lichenoid inflammation can prove challenging.MethodsWe designed a case–control study to evaluate the accuracy metrics of PRAME immunohistochemistry to distinguish melanocytic pseudonests in lichenoid dermatoses or keratoses from inflamed MIS. Immunostaining for PRAME was performed on paraffin‐embedded formalin‐fixed diagnostic tissue using a rabbit monoclonal antibody to PRAME (Abcam), with a 1:3200 dilution on a Leica Bond detection system.ResultsOur search identified 21 cases of melanocytic pseudonests (n = 21, 46%) encountered in lichenoid dermatoses and 24 cases of inflamed MIS (n = 24, 53%). Each method of evaluating PRAME immunohistochemistry (PRAME+ clusters, PRAME % of melanocytes by four categories and PRAME+ melanocyte counts per linear mm of epidermal basal layer) showed statistically significant differences between the MIS and the pseudonest cohorts (respectively, p < 0.001; p < 0.001; and p < 0.001). Receiver operating characteristics analysis for PRAME+ melanocyte counts per linear mm of epidermal basal layer revealed an area under the curve of 0.9 ± 0.05 (95% confidence interval 0.9–1.0). When determining an optimal cut‐off point for the best Youden index [sensitivity (%) + specificity (%) − 100], the cut‐off of 1.0 PRAME+ melanocytes per linear mm showed a sensitivity of 79.2% and specificity of 85.7% (Youden index 0.65) to distinguish MIS from pseudonests.ConclusionPRAME immunohistochemistry may constitute an additional tool for this challenging differential diagnosis.