Aims: The aims of this study were to clarify the usefulness of immunohistochemistry in the differential diagnosis of epithelioid mesothelioma with a solid growth pattern [solid epithelioid mesothelioma (SEM)] and poorly differentiated squamous cell carcinoma (PDSCC), and to confirm the validity of a specific type of antibody panel. Additionally, we aimed to clarify the pitfalls of immunohistochemical analyses. Methods and results: Formalin-fixed paraffin-embedded specimens from 36 cases of SEM and 38 cases of PDSCC were immunohistochemically examined for calretinin, podoplanin (D2-40), Wilms’ tumour gene product (WT1), cytokeratin (CK) 5/6, p40, p63, carcinoembryonic antigen (CEA), epithelial-related antigen (MOC31), claudin-4, thyroid transcription factor-1 (TTF-1), and napsin A. WT1 showed the highest diagnostic accuracy (85.1%) as a mesothelial marker, and CEA, p40 and claudin-4 showed higher diagnostic accuracies (95.9%, 94.6%, and 93.2%, respectively) as carcinoma markers. Calretinin (diagnostic accuracy: 75.7%), D2-40 (diagnostic accuracy: 67.6%), CK5/6 (diagnostic accuracy: 63.5%), TTF-1 (diagnostic accuracy: 55.4%) and napsin A (diagnostic accuracy: 52.7%) could not differentiate between SEM and PDSCC. Among these markers, the combination of calretinin and WT1 showed the highest diagnostic accuracy (86.5%) as a positive marker, and the combination of p40 and CEA showed the highest diagnostic accuracy (97.3%) as a negative marker. The combination of CEA and claudin-4 also showed relatively high diagnostic accuracy (94.6%) as a negative marker. Conclusions: We recommend the combination of WT1 and calretinin as a positive maker, and the combination of CEA and claudin-4 as a negative marker, for differential diagnoses of SEM and PDSCC.
