Survey of keratin 17 (K17) expression in different urothelial malignancies.

458 Background: The current non-invasive urine tests tend to display low sensitivity for the detection of urothelial cancers, especially in low-grade cancer. Recent reports suggest that the URO17 test based on the Keratin 17 (K17) biomarker displays high sensitivity and specificity in the detection of urothelial carcinoma in both low-grade and high-grade cancers and show promise as an effective non-invasive test for bladder cancer. The current study aimed to evaluate K17 expression in different histological categories of urological diseases. Methods: Ninety-five urine specimens were collected from 2020-2021 that represented 19 different histological categories of urothelial diseases. These samples were prepared onto slides and tested with the URO17 Bladder Cancer test (KDx Diagnostics Inc.) on DAKO Link48 autostainer using the manufacturer’s recommended procedures. The stained samples were evaluated by a cytopathologist and scored using five K17-positive urothelial cells as a cut-off. Results: K17 was positive in 20/20 (100%) of the urothelial carcinoma, and 23/27 (85%) of Non-invasive papilloma and urothelial carcinomas. K17 was positive in 100% of carcinoma in situ, papillary urothelial carcinoma, sarcoma, and small cell carcinoma, papillary urothelial carcinoma, urothelial papilloma, inverted papilloma, renal-pelvis urothelial carcinoma, and urothelial carcinoma with squamous differentiation. K17 was also expressed in many cystitis, inflammatory myofibroblastic tumors, and PUNLMP, but not in prostate carcinoma. In 19/22 cases (86%) K17 was mostly positive in urine samples from patients with prior history of bladder cancer, but were histologically determined to be benign. The overall sensitivity of K17 in the cancer group was 88.5% (54/61 positive) with K17 being expressed in both high-grade and low-grade cancers. Of the seven false-negative samples, five were low-grade, one was high-grade, and one was from the inflammatory myofibroblastic tumor. Conclusions: Keratin 17 is highly expressed in almost all cases of urothelial carcinoma including those of papillary and squamous characteristics, confirming the high sensitivity of K17 in detecting all categories of urothelial carcinoma. Keratin 17 was also observed in many benign and premalignant conditions such as cystitis, PUNLMP, and metaplasia suggesting the K17 expression can result from a trauma to urothelium that could lead to activation of K17 in regenerating urothelial cells. Additional studies are ongoing to confirm and expand the utility of the K17 biomarker in bladder cancer.