Optimizing Ki-67 and E-Cadherin Thresholds for Improved Grade and Stage Classification in Urothelial Bladder Cancer

Area of Science:

• Oncology
• Pathology
• Biomarker Research

Background:

• Bladder cancer presents significant heterogeneity, necessitating precise differentiation between non-muscle-invasive (NMIBC) and muscle-invasive bladder cancer (MIBC).
• Accurate grading into low-grade (LG) and high-grade (HG) tumors is critical for guiding clinical management strategies.
• Standardized cut-off values for immunohistochemical (IHC) markers like p53, Ki-67, and E-cadherin, and their combined predictive utility, require further clarification.

Purpose of the Study:

• To determine optimal cut-off values for p53, Ki-67, and E-cadherin in bladder cancer.
• To assess the individual and combined discriminatory power of these biomarkers for tumor stage (NMIBC vs. MIBC) and grade (LG vs. HG).
• To evaluate the potential of biomarker panels for improved pathological stratification.

Main Methods:

• Retrospective analysis of 568 bladder cancer cases.
• Quantification of p53, Ki-67, and E-cadherin expression using immunohistochemistry (IHC).
• ROC curve analysis with Youden's J criterion and 10-fold cross-validation for cut-off determination; logistic regression models for performance evaluation.

Main Results:

• Ki-67 demonstrated the strongest single-marker performance for predicting MIBC (AUC 0.842) and HG disease (AUC 0.813) with cut-offs ≥40% and ≥30%, respectively.
• p53 and E-cadherin (inverted scale) showed moderate to acceptable discrimination for stage and grade.
• The combination of Ki-67 and E-cadherin yielded the highest performance (AUCs 0.851 for MIBC, 0.838 for HG), showing modest improvement over single markers.

Conclusions:

• Ki-67 serves as the most potent single biomarker for identifying invasive and high-grade bladder cancers.
• E-cadherin provides valuable complementary information, enhancing diagnostic accuracy when used with Ki-67.
• A two-biomarker panel of Ki-67 and E-cadherin, with defined cut-offs, represents a promising and practical tool for superior pathological stratification in bladder cancer.