Diagnostic Accuracy of Liquid-Based Biomarkers for Detecting and Risk Stratifying Upper Tract Urothelial Cancers: A Diagnostic Test Systematic Review by the EAU Guidelines UTUC Panel.

OBJECTIVES

Liquid-based biomarkers (LBBs), derived from urine and blood, are gaining attention in the management of upper tract urothelial carcinoma (UTUC) as potential tools to overcome limitations of current diagnostic and risk stratification pathways. This review evaluates emerging LBBs, focusing on diagnostic accuracy and their role in predicting high-grade and ≥pT2 disease.

EVIDENCE ACQUISTION: A systematic search of MEDLINE, Embase, and the Cochrane CENTRAL (via Ovid) (January 2010-July 2024) identified studies reporting diagnostic performance of LBBs for UTUC, excluding urine cytology. Data on sensitivity, specificity, positive predictive value, negative predictive value, and area under the curve were extracted. A narrative synthesis was performed.

EVIDENCE SYNTHESIS: Thirty-two studies met the inclusion criteria. Among bladder-derived urine biomarkers, RNA-based panels (an eight-gene panel: sensitivity 86-92%, specificity 87-93%) (low certainty) and DNA methylation panels-GDF15, TMEFF2, and VIM (sensitivity 91%, specificity 100%), Bladder Care Index (sensitivity 96%, specificity 88%), NRN1 (sensitivity 92-96%, specificity 92-95%) and ONECUT (sensitivity 89%, specificity 94%) and protein-based panels (BTA-STAT, BTA, and survivin) showed favourable diagnostic profiles (very low certainty). Upper-tract urine assays, EpiCheck (sensitivity 65-83%, specificity 79-83%), demonstrated modest accuracy (very low certainty for sensitivity; low certainty for specificity) with strong performance for high-grade disease. Among blood biomarkers, circulating tumour DNA (ctDNA) with plasma copy number burden >6.5 achieved high accuracy for predicting ≥pT2 and high-grade disease (sensitivity 71%, specificity 94%) (low certainty).

CONCLUSIONS

RNA-based, DNA methylation-based, protein-based, and ctDNA-based assays showed promise for UTUC diagnosis and risk stratification. However, the current evidence is limited, with few validated studies. Large prospective cohorts and biomarker-driven interventional trials are needed before clinical integration.