Utility of a triple antibody cocktail intraurothelial neoplasm-3 (IUN-3-CK20/CD44s/p53) and α-methylacyl-CoA racemase (AMACR) in the distinction of urothelial carcinoma in situ (CIS) and reactive urothelial atypia

Urothelial carcinoma in situ (CIS) is a prognostically and therapeutically significant lesion with considerable morphologic overlap with reactive conditions especially in the setting of prior therapy. Various markers including CK20, CD44s, and p53 have been used as an adjunct in making this distinction; however, the utility of these markers in the posttreatment scenario is not fully established. α-Methylacyl-CoA racemase (AMACR) is a tumor-associated marker that is expressed in a subset of high-grade urothelial carcinomas but has not been studied in CIS. This study was undertaken to evaluate the immunoreactivity of CK20, CD44s, and p53 as a triple antibody cocktail intraurothelial neoplasm-3 (IUN-3) in distinguishing CIS from its mimics and to compare its utility with AMACR in the diagnosis of CIS. A total of 135 specimens (7 benign ureters and 128 bladder biopsies - 28 reactive, 33 posttherapy reactive, 43 CIS, 24 CIS posttherapy) were included in this study. Immunostaining for p53 (brown, nuclear), CD44s (brown, membranous), and CK20 (red, cytoplasmic and membranous) was performed as a cocktail, and the staining pattern was further classified as: malignant (full-thickness CK20 and/or full-thickness p53 with CD44s negativity), reactive/benign (CK20 limited to the umbrella cell layer, p53 negative, and CD44s positivity ranging from basal to full thickness), and indeterminate (CK20 and p53 positive but not full thickness and/or CD44s positive). AMACR staining was performed in 50 cases. Cytoplasmic staining for AMACR was graded as negative (absent to weak focal staining [<5% cells]) and positive (≥5%). The "IUN-3 malignant" pattern was observed in 84% of cases of CIS without a history of prior therapy and in 71% of the cases of CIS with a history of prior therapy. Cases with posttherapy reactive atypia showed an "IUN-3 reactive" pattern in 84% cases and "IUN-3 indeterminate" pattern in 16% of the cases; the IUN-3 malignant pattern was not identified in any of the cases. Benign and reactive urothelium (with and without a history of therapy) showed an IUN-3 reactive pattern and negative AMACR staining in all the cases (100%). AMACR positivity was observed in 78% of nontreated CIS cases and 50% of CIS posttherapy cases. In these cases, the IUN-3 cocktail showed an IUN-3 malignant pattern in 83% of untreated CIS cases and 88% of CIS posttherapy cases. AMACR positivity is a potentially useful marker of CIS. However, the IUN-3 malignant pattern is a more reliable indicator of CIS compared with AMACR, especially in the posttreatment setting. The simultaneous evaluation of all 3 markers (p53, CD44s, and CK20) in a single slide in the form of a cocktail is advantageous, especially in small biopsy specimens.