Utility of D2-40, Cytokeratin 5/6, and High-Molecular-weight Cytokeratin (Clone 34βE12) in Distinguishing Intraductal Spread of Urothelial Carcinoma From Prostatic Stromal Invasion

Intraductal spread of urothelial carcinoma (UC) is not an uncommon finding in bladder cancer that requires appropriate clinical management. The presence of prostatic stromal invasion in non-muscle-invasive bladder cancer upstages the disease, necessitating cisplatin-based neoadjuvant chemotherapy and subsequent cystroprostatectomy. However, the identification of prostatic stromal invasion can be challenging, especially in biopsy and transurethral resection specimens. We assess the utility of D2-40, CK5/6, and high-molecular-weight cytokeratin (HMWCK) immunohistochemistry as an ancillary tool to differentiate prostatic stromal invasion from intraductal UC spread. We reviewed 13 cystoprostatectomies performed for UC with prostatic involvement. The presence of stromal invasion was histologically determined by the presence of circumferential retraction artifact, paradoxical differentiation, complex architecture, and desmoplastic reaction. The areas of interest were subsequently stained with D2-40, CK5/6, and HMWCK (clone 34 beta E12). Four bladder biopsies were used as a control to assess labeling in the benign urothelium. Nine cases had histologic evidence of prostatic stromal invasion (4 transmurally through bladder wall). D2-40 highlighted basal cells in all benign prostatic ducts and was consistently negative in UC, benign urothelium, prostatic adenocarcinoma, and benign luminal prostatic epithelium. D2-40 and CK5/6 performed similarly for intraductal UC, labeling only the basal cell layer with the exception of 1 case with squamous differentiation where CK5/6 exhibited full thickness staining. HMWCK diffusely stained 9 of 10 intraductal UCs without squamous differentiation and 1 intraductal UC with squamous differentiation. All 8 cases of invasive UC without squamous differentiation were negative for D2-40. Seven of these cases had focal CK5/6 and diffuse HMWCK staining. In 1 case of invasive UC with squamous differentiation, all stains were positive. D2-40 is expressed in prostatic basal cells, but it is not expressed in the benign or neoplastic urothelium. D2-40 and CK5/6 effectively highlight the intraductal spread of UC. While invasive UC is negative for D2-40, CK5/6 is usually patchy and localized to the periphery of the tumor nests. HMWCK often demonstrates diffuse staining in both scenarios. However, these stains do not perform well in cases of UC with squamous differentiation. Thus, D2-40 can be used as an ancillary tool to rule out prostatic stromal invasion.