Abstract

Distinguishing melanocytic pseudonests encountered in lichenoid dermatoses or lichenoid keratoses from melanoma in situ (MIS) with brisk lichenoid inflammation can prove challenging. We compared PRAME immunohistochemistry for 22 cases of lichenoid keratoses or lichenoid dermatoses with 25 cases of MIS with lichenoid inflammation in a retrospective case-control study and established optimal diagnostic cut-off points for interpretation. PRAME immunohistochemical reactivity was evaluated by two methods: categorical (0: no staining, 1+: 1-25%, 2+:26-50%, 3+: 51-75%, 4+: 76-100%) and numerical (moderate to strongly staining melanocytic nuclei per linear millimeter of biopsy) fashion. Twenty cases diagnosed as melanocytic pseudonests stained 0 or 1+ (n=20, 91%) while twenty-one cases of inflamed MIS labeled 2+, 3+ or 4+ (n=21, 84%). All cases with a score of 4+ were inflamed MIS. The mean counts of PRAME-positive melanocytes per linear mm of biopsy were significantly different: 0.5+/-0.9 for lichenoid melanocytic pseudonests and 10.0+/-11.6 for inflamed MIS (T test, p less than 0.0001). Receiver operating characteristic curve (ROC) calculations found an area under the curve of 0.883, demonstrating good sensitivity and specificity for MIS. A cut-off of PRAME-positive melanocyte counts over 0.5 cell per linear mm showed a sensitivity of 80.0% and specificity of 77.3%. A higher cut-off of 1.7 cells per mm, designed to optimize test specificity, resulted in a sensitivity of 72% but specificity of 91%. Our findings corroborate that optimal cut-offs for PRAME immunoreactivity beyond simple binary staining in lichenoid dermatoses and keratoses may prove a useful ancillary method to distinguish melanocytic pseudonests from inflamed melanoma in situ.

Financial Disclosure:
No current or relevant financial relationships exist.