BACKGROUND: The histopathologic pattern of clonal seborrheic keratosis (SK) is quite similar to the nested pattern of pagetoid Bowen's disease [squamous cell carcinoma in situ (SCCIS)], and differentiation between the two can be challenging, especially when only small pieces are available for interpretation. METHODS: Eleven examples of clonal SK and 13 examples of pagetoid SCCIS were examined histopathologically (tabulating necrotic keratinocytes, suprabasal mitoses, infiltrate, parakeratosis housing plump nuclei, crowding of nuclei) and immunohistochemically (using Ki-67, bcl-2, cytokeratin 7 and cytokeratin 10). Sensitivity, specificity, p-values (Fisher's exact test, two-tailed) and positive/negative likelihood ratios (+LR/-LR) were calculated. RESULTS: Significant differences were seen with regard to crowding (p = 0.0009) and mitoses (p = 0.0006); however, only complete absence of necrotic keratinocytes or of crowding appeared to be diagnostically convincing for a diagnosis of clonal SK (-LR < 0.01). Significant differences were also seen with bcl-2 (p = 0.0005) and cytokeratin 10 antibodies (p < 0.00001). Both markers displayed a typical nested pattern in clonal SK, nests being bcl-2-positive and cytokeratin 10-negative. Cytokeratin 10-negative nests were the most convincing criterion for differentiation between clonal SK and pagetoid SCCIS (+LR > 10, -LR < 0.01). CONCLUSIONS: The most reliable marker to distinguish clonal SK from pagetoid SCCIS is cytokeratin 10 when it spares nests. Other criteria that assist in the differential diagnosis are bcl-2 expression, absence of crowding and of mitoses.
BACKGROUND: The histopathologic pattern of clonal seborrheic keratosis (SK) is quite similar to the nested pattern of pagetoid Bowen's disease [squamous cell carcinoma in situ (SCCIS)], and differentiation between the two can be challenging, especially when only small pieces are available for interpretation. METHODS: Eleven examples of clonal SK and 13 examples of pagetoid SCCIS were examined histopathologically (tabulating necrotic keratinocytes, suprabasal mitoses, infiltrate, parakeratosis housing plump nuclei, crowding of nuclei) and immunohistochemically (using Ki-67, bcl-2, cytokeratin 7 and cytokeratin 10). Sensitivity, specificity, p-values (Fisher's exact test, two-tailed) and positive/negative likelihood ratios (+LR/-LR) were calculated. RESULTS: Significant differences were seen with regard to crowding (p = 0.0009) and mitoses (p = 0.0006); however, only complete absence of necrotic keratinocytes or of crowding appeared to be diagnostically convincing for a diagnosis of clonal SK (-LR < 0.01). Significant differences were also seen with bcl-2 (p = 0.0005) and cytokeratin 10 antibodies (p < 0.00001). Both markers displayed a typical nested pattern in clonal SK, nests being bcl-2-positive and cytokeratin 10-negative. Cytokeratin 10-negative nests were the most convincing criterion for differentiation between clonal SK and pagetoid SCCIS (+LR > 10, -LR < 0.01). CONCLUSIONS: The most reliable marker to distinguish clonal SK from pagetoid SCCIS is cytokeratin 10 when it spares nests. Other criteria that assist in the differential diagnosis are bcl-2 expression, absence of crowding and of mitoses.
Authors: Philipp Jansen; Daniel Otero Baguer; Nicole Duschner; Jean Le'Clerc Arrastia; Maximilian Schmidt; Bettina Wiepjes; Dirk Schadendorf; Eva Hadaschik; Peter Maass; Jörg Schaller; Klaus Georg Griewank Journal: Cancers (Basel) Date: 2022-07-20 Impact factor: 6.575