# Integrative Analysis of GATA3 Expression and Variants as Prognostic Biomarkers in Urothelial Cancer

## TL;DR

This study explores how GATA3 gene expression and genetic variants affect the prognosis of urothelial cancer, suggesting GATA3 could be a useful biomarker for predicting patient outcomes.

## Contribution

The study integrates genomics, pathology, and machine learning to identify GATA3 as a potential prognostic biomarker in urothelial cancer.

## Key findings

- The rs1244159 G allele is associated with reduced urothelial cancer risk and higher GATA3 expression.
- High GATA3 expression predicts improved overall survival in urothelial cancer patients.
- Age, chemotherapy, and GATA3 expression are top predictors of survival identified by machine learning.

## Abstract

GATA3 is a transcription factor involved in urothelial differentiation and is widely used as a diagnostic marker for urothelial carcinoma (UC). Although loss of GATA3 expression has been linked to more aggressive disease, its prognostic significance remains uncertain. Genetic variation within the GATA3 locus, particularly rs1244159, may influence protein expression and clinical outcomes. We conducted a case control study in Taiwan including 461 UC cases and 586 controls genotyped for four GATA3 SNPs. GATA3 expression was assessed via immunohistochemistry (IHC) in 98 tumor tissues. Logistic regression and Kaplan–Meier analyses were used to evaluate SNP associations and survival outcomes. An XGBoost-based machine learning model with SHAP (SHapley Additive exPlanations) was applied to rank survival predictors. The rs1244159 G allele was associated with a significantly reduced UC risk (adjusted OR = 0.48, p = 0.0231) and higher GATA3 expression (p = 0.0173). High GATA3 expression predicted improved overall survival (p = 0.0092), particularly among G allele carriers (p = 0.0071). SHAP analysis identified age, chemotherapy, and GATA3 expression as the top predictors of survival, consistent with Cox regression results. In conclusion, our integrative analysis suggests that the rs1244159 G allele modulates GATA3 expression and influences UC prognosis. Combining genomics, pathology, and machine learning, GATA3 may serve as a clinically useful biomarker for risk stratification and outcome prediction in UC.