Claudin-4 shows superior specificity for mesothelioma vs non–small-cell lung carcinoma compared with MOC-31 and Ber-EP4

Abstract

The distinction of malignant mesothelioma from non-small-cell lung carcinoma (NSCLC) usually requires immunohistochemistry, but some broad-spectrum carcinoma markers stain mesotheliomas, and it remains unclear which broad-spectrum markers are most valuable for distinguishing these malignancies. Here, we directly compared the sensitivity and specificity of three broad-spectrum carcinoma markers, claudin-4, Ber-EP4, and MOC-31, for distinguishing NSCLC from mesothelioma. Immunohistochemistry was performed on tissue microarrays containing 68 epithelioid mesotheliomas, 31 sarcomatoid mesotheliomas, and 147 non-small-cell lung cancers (53 adenocarcinomas, 60 squamous cell carcinomas, 13 large-cell carcinomas, and 21 sarcomatoid carcinomas). For adenocarcinoma, squamous cell carcinoma, and large-cell carcinoma, claudin-4 staining was present in 103 of 126 cases (82%), MOC-31 staining was present in 112 of 126 cases (89%), and Ber-EP4 staining was present in 113 of 126 cases (90%); these values were not statistically different. Claudin-4 stained 0 of 68 (0%), MOC-31 stained 22 of 68 (32%), and Ber-EP4 stained 24 of 68 (35%) epithelioid mesotheliomas; thus, the specificities for NSCLC versus epithelioid mesothelioma were 100%, 68%, and 65%, respectively. Claudin-4 staining was present in 7 of 21 (33%), MOC-31 staining was present in 8 of 21 (38%), and Ber-EP4 staining was present in 5 of 21 (24%) sarcomatoid carcinomas. All three markers were negative in 12 of 21 (57%) sarcomatoid carcinomas. Sarcomatoid mesotheliomas were not stained with any of these markers. We conclude that claudin-4 has considerably greater specificity and comparable sensitivity to MOC-31 and Ber-EP4 for distinguishing NSCLC from epithelioid malignant mesothelioma. The use of all three markers may be necessary for sarcomatoid neoplasms, given their limited sensitivity.