WT1 as a myoepithelial marker: a comparative study of breast, cutaneous, and salivary gland lesions.

WT1 immunostain is expressed in various benign and malignant neoplasms, as well as normal myoepithelial cells. WT1 shows differential expression in non-neoplastic, benign, and malignant neoplastic myoepithelial cells of the salivary gland. In this study, WT1 immunostain and other myoepithelial markers were compared to investigate the value of WT1 as a myoepithelial marker, and to delineate the expression profile of WT1 in nonsalivary gland myoepithelial cells. WT1, p63, and calponin immunostains were performed on normal and lesional tissues from the breast (adenosis, sclerosing adenosis, lactating adenoma, nipple adenoma, tubular adenoma, adenomyoepithelioma, and adenoid cystic carcinoma [ACC]), skin (cutaneous mixed tumor, hidradenoma, spiradenoma, and ACC), and salivary gland (pleomorphic adenoma and ACC). The stained slides were digitized and orientated with H&E images and assessed simultaneously using QuPath. A total of 129, 58, and 56 breast, cutaneous, and salivary gland lesions, respectively, were included. There was poor agreement between WT1-p63 and WT1-calponin (κ < 0.1) in all organs, with absence of WT1 expression in normal salivary gland myoepithelium and most ACCs. There were no significant differences in WT1 expression in myoepithelial cells in normal breast tissue and benign breast neoplasms. Compared to pleomorphic adenomas, cutaneous mixed tumors showed lower WT1 expression (P < .001). WT1 is a less sensitive myoepithelial marker than calponin and p63. However, its unique pattern of expression in salivary gland primary for pleomorphic adenomas/cutaneous mixed tumor can favor a diagnosis of benign salivary gland tumors, particularly in small biopsy specimens.

PRAME immunostain expression in sebaceous lesions, cutaneous carcinomas and adnexal structures.

The use of immunostain for PRAME antigen is well established for cutaneous melanolocytic lesions. However, its staining in other cutaneous structures and lesions is under reported. This study assessed PRAME staining in a large cohort of normal skin tissue, sebaceous lesions, and cutaneous carcinomas to better delineate patterns of PRAME immunoreactivity. PRAME immunostaining was performed on sections of sebaceous lesions and tissue microarrays of basal cell carcinomas (BCCs) and squamous cell carcinomas (SCCs). Normal cutaneous adnexal structures were assessed on the sections of sebaceous lesions. For sebaceous lesions and non-lesional sebaceous glands, PRAME immunostaining was assessed for mature, germinative and sebocytes independently. A total of 193 sebaceous lesions, 64 BCCs and 35 SCCs were stained for PRAME immunostain. Staining pattern was predominantly cytoplasmic in normal apocrine glands, germinative sebocytes of sebaceous glands, and hair germs (p<0.001). Lesional sebocytes did not show different staining compared to normal sebaceous glands (p>0.05). Rare nuclear staining was observed in the normal epidermis (0.6%) and junctional melanocytes (4.1%). BCC, SCC and sebaceous carcinoma all showed low levels of PRAME immunoreactivity with variable proportions of cases demonstrating nuclear staining (BCC 59.4%, SCC 37.1%, sebaceous carcinoma 5.3%). PRAME immunostaining is positive in germinative sebocytes, various cutaneous structures and carcinomas. Nuclear staining, identical to melanoma, was observed in normal epidermis, junctional melanocytes, BCCs, SCCs, and sebaceous carcinomas. The pattern of PRAME staining in the skin must be recognised to avoid pitfalls in interpretating PRAME immunostain.