Bioinformatic Approach to Identify Potential TGFB2-Dependent and Independent Prognostic Biomarkers for Ovarian...

High-grade serous ovarian carcinoma is the most common and aggressive form of ovarian cancer, accounting for over 60% of cases and nearly 75% of deaths, mainly due to late diagnosis and tumor aggressiveness. Standard treatment is platinum-based chemotherapy with paclitaxel, but relapse is frequent. This study aimed to identify prognostic biomarkers for patients with poor survival outcomes after Taxol treatment using bioinformatics analysis. We examined the effects of TGFB2 mRNA expression and other markers on overall survival in serous ovarian cancer using the TCGA database, applying a multivariate Cox model that included interaction terms to identify TGFB2-dependent and independent prognostic markers, and controlling for age and treatment type. Candidate TGFB2-independent prognostic markers from TCGA were further validated using patient data from the KMplotter database. High TGFB2 mRNA expression emerged as a prognostic biomarker for three potential gene targets (TRPV4, STAU2, and HOXC4) associated with improved OS at low levels of gene target expression, we identified four additional markers (CLIC3, ANPEP/LAP1, RIN2, and EMP1) that exhibited a TGFB2-independent negative correlation between mRNA expression and OS across the full spectrum of gene expression values in the ovarian cancer cohort validated using independent dataset from KMplotter, for Taxol-treated ovarian cancer patients. This study proposes a panel of potential prognostic biomarkers for the treatment of ovarian cancer patients, particularly by leveraging TGFB2-dependent mRNA expression as a significant biomarker, alongside four additional TGFB2-independent prognostic markers, for patients undergoing Taxol-based therapies. Future prospective clinical trials will be required to validate these prognostic markers.