Keratin 19 (KRT19) as a circulating biomarker to measure disease burden and guide treatment in urothelial cancer.

Abstract

4591 Background: Circulating tumor DNA (ctDNA) has emerged as a valuable biomarker in urothelial cancer but remains limited by cost and turnaround time. Keratin 19 (KRT19) is a cytoskeletal protein that is highly and specifically expressed in urothelial tumor tissue in The Cancer Genome Atlas (TCGA) urothelial cancer cohort. Based on this tumor specific expression, we hypothesized that circulating KRT19 levels could serve as a rapid, inexpensive measure of disease burden to guide systemic therapy for urothelial cancer. Methods: Two patient cohorts from completed clinical trials were analyzed. HCRN GU16-257, a phase 2 trial of gemcitabine, cisplatin, and nivolumab as organ-sparing treatment for localized muscle-invasive bladder cancer, included 76 patients evaluable for KRT19 with median follow-up of 30 months. HCRN GU14-182, a phase 2 switch maintenance trial of pembrolizumab versus placebo following first-line chemotherapy for metastatic urothelial cancer, included 46 patients evaluable for KRT19 in the pembrolizumab arm with median follow-up of 12.9 months. Serum KRT19 levels were quantified using Olink proteomics technology alongside 91 additional proteins. Cox proportional hazards models assessed associations between KRT19 expression and clinical outcomes, with adjustment for multiple comparisons using false discovery rate correction. Results: In the GU16-257 cohort, among 92 peripheral blood analytes tested at baseline (C1D1), KRT19 demonstrated the strongest association with overall survival. Elevated baseline KRT19 was independently associated with inferior overall survival (HR 2.62, adjusted p=0.016) and metastasis-free survival (HR 3.01, adjusted p=0.001). A decrease in KRT19 from C1D1 to C3D1 was associated with improved overall survival (HR 0.59, unadjusted p=0.024) and metastasis-free survival (HR 0.60, unadjusted p=0.041). In GU14-182, elevated baseline KRT19 was associated with inferior overall survival (HR 2.11, adjusted p=0.034) and progression-free survival (HR 1.57, unadjusted p=0.024). Optimal KRT19 cutpoints enabled significant risk stratification for mortality and disease progression in both cohorts. Analysis of an independent peripheral blood single-cell RNA sequencing cohort from patients with urothelial cancer confirmed epithelial cells as the predominant source of circulating KRT19. Conclusions: KRT19 represents a promising, readily measurable serum biomarker with specificity for urothelial cancer. Both baseline levels and on-treatment changes correlate with disease progression and survival. Further validation of KRT19 as a tool to assess disease burden and guide treatment decisions in urothelial cancer is warranted.