The diagnostic utility of prame in differential diagnosis of melanocytic lesions and comparative analysis with other immunohistochemical markers

Aim: Melanocytic lesions can present with a wide range of morphological characteristics, making diagnosis difficult. Therefore, identifying new immunohistochemical markers for differentiating benign and malignant lesions plays an important role. Our study aimed to investigate the expression of PRAME antibody, which has gained importance in distinguishing melanoma from benign lesions in recent years, and it's importance in the diagnostic approach with BRAF(VE1), p16, MelanA+Ki67 dual staining, and BAP1 antibodies used in routine practice. Materials and Methods: 27 atypical spitzoid lesions, 12 spitzoid melanoma, 20 conventional melanoma, and 10 dysplastic nevi with high grade dysplasia diagnosed in Ankara University Medical Faculty Pathology Department between 2005-2022 formed the study group. Patients diagnosed with 20 melanocytic nevi, 20 Spitz nevi and 10 dysplastic nevi with low grade dysplasia formed the control group. PRAME, BRAF, BAP1, p16, MelanA, and Ki67 antibodies were applied using the Streptavidin-Biotin Peroxidase method in automatic immunohistochemistry staining devices (Roche-Ventana). Evaluations were made under a light microscope and staining above 75% was accepted as the threshold value for anti-PRAME. Evaluation for the BRAF VE1 antibody was based on the severity of membranous-cytoplasmic staining; Cases with moderate membranous staining (2+) and cases with severe staining were scored as (3+), and cases with 2+ and 3+ staining patterns were considered positive for BRAF V600E . Presence of diffuse nuclear and cytoplasmic staining with p16 (3+), the presence of patchy staining (2+) were accepted as positive for p16 expression. Focal staining (score 1+) or complete absence of nuclear expression (score 0) were evaluated as loss of p16 expression. BAP1 evaluation was evaluated as positive or negative according to nuclear staining. MelanA and Ki67 dual staining was preferred to show the proliferation of melanocytic cells, and Ki67 expression was noted as percentage. The data was analyzed in the "SPSS for Windows 11.5" package program, and p<0.05 was considered statistically significant. Results: When all of the groups were taken, 55.5% of them were female (n=66), and 44.5% were male (n=53). According to the age distribution, which ranged from 1 to 94, the mean age was determined to be 36.84 ±20.16. Considering the PRAME staining percentage, nuclear staining ranged from 0% to 95% in conventional melanomas, with a mean of 58.05% ±38.76; between 0% and 85% in spitzoid melanomas, with a mean 28% ±37.72; between 0% and 50% in ASTs with a mean 2% ±9.60; and between 0% and 30% in Spitz nevi with a mean 3.25% ±8.31; It was discovered to range from 0% to 10% in dysplastic nevi, with an average of 0.75% ±2.44. Melanocytic nevi did not show any nuclear staining. When staining over 75% was considered as positive for PRAME, 12 (60%) of 20 cases in conventional melanoma group and 3 (25%) of 12 cases in spitzoid melanoma group were positive. Nuclear staining with PRAME below the threshold value was seen in 1 of 27 cases (3.7%) in AST, in 3 (15%) of 20 cases in Spitz nevi, and 2 (10%) of 20 cases in dysplastic nevi. A statistically significant difference was found for PRAME staining between conventional melanoma and melanocytic nevi, dysplastic nevi, AST, and Spitz nevi (p<0.05). When staining over 75% was considered as threshold for PRAME, the sensitivity was 60%, and the specificity was 100% in conventional melanomas. For spitzoid melanomas, the sensitivity was 25%, and the specificity was 100%. As a result of ROC analysis, the highest Youden index was found at 55%. When 55% was considered positive for PRAME, sensitivity increased to 70% in conventional melanomas and 33.3% in spitzoid melanomas without change in specificity. There was no statistically significant relationship between PRAME expression and the degree of dysplasia. However, the percentage and the severity of PRAME expression increased with the increasing severity of dysplasia. In AST cases, there was a statistically significant relationship between PRAME expression and Clark level (p<0.05). However, there was no significant correlation between Clark level or Breslow thickness in spitzoid or conventional melanoma cases (p>0.05). We found a statistically significant relationship between TILs and PRAME expression in spitzoid melanoma cases (p<0.05). However, in conventional melanoma cases, no statistically significant association was found (p=0.5). Conclusion: In our study, the absence of staining in melanocytic nevi with PRAME and it's high and diffuse staining in malignant cases suggested that PRAME could be a helpful tool in the differentiation of benign and malignant lesions. Our results also suggested that reducing the used threshold value for PRAME to 55% would increase the value of PRAME in melanoma diagnosis due to the increased sensitivity without a decrease in specificity. Key Words: PRAME, melanoma, Spitz tumor, atypical Spitz tumor, Spitzoid melanoma nevi, histopathology