Open Access

ARTICLE

A Novel Partial EMT-Associated Transcriptomic Signature for Prognostic Stratification in Ovarian Cancer

Chia-Chia Chao¹, Cheng-Yao Lin^2,3,4, Po-Chun Chen^5,6,7, Wen-Tsung Huang², Teng-Song Weng⁸, Sheng-Yen Hsiao^2,9,*

1 Department of Respiratory Therapy, Fu Jen Catholic University, New Taipei City, Taiwan
2 Division of Hematology-Oncology, Department of Internal Medicine, Chi Mei Medical Center, Liouying, Tainan, Taiwan
3 Department of Senior Welfare and Services, Southern Taiwan University of Science and Technology, Tainan, Taiwan
4 Department of Environmental and Occupational Health, College of Medicine, National Cheng Kung University, Tainan, Taiwan
5 School of Life Science, National Taiwan Normal University, Taipei, Taiwan
6 Translational Medicine Center, Shin-Kong Wu Ho-Su Memorial Hospital, Taipei, Taiwan
7 Department of Medical Research, China Medical University Hospital, China Medical University, Taichung, Taiwan
8 Department of Pharmacy, Chi Mei Medical Center, Liouying, Tainan, Taiwan
9 Department of Nursing, Chung Hwa University of Medical Technology, Tainan, Taiwan

* Corresponding Author: Sheng-Yen Hsiao. Email:

(This article belongs to the Special Issue: Novel Biomarkers and Treatment Strategies in Solid Tumor Diagnosis, Progression, and Prognosis (Ⅱ))

Oncology Research 2026, 34(5), 27 https://doi.org/10.32604/or.2026.074383

Received 10 October 2025; Accepted 29 January 2026; Issue published 22 April 2026

Abstract

Background: Partial epithelial–mesenchymal transition (p-EMT) is a dynamic cellular state associated with metastasis and adverse outcomes in multiple cancers, but its prognostic significance in ovarian cancer remains unclear. This study aimed to develop and validate an ovarian cancer–specific transcriptomic signature based on p-EMT–related genes, and to determine whether this signature can improve prognostic stratification and overall survival prediction across independent cohorts. Methods: A pan-cancer p-EMT gene set was curated from ten published studies. Using transcriptomic and clinical data from TCGA-OV (n = 488), a six-gene p-EMT signature was developed via LASSO regression to generate a patient-specific risk score. The score was integrated with clinical variables to construct a prognostic nomogram and validated in the external GEO cohort GSE140082 (n = 380) and GSE165808 (n = 51). Results: A six-gene p-EMT transcriptomic signature (ADAM9, ANXA8L1, FSTL3, RABAC1, TPM4, and TWIST1) was significantly associated with overall survival (OS) and stratified patients into high- and low-risk groups (adjusted HR = 1.74, p < 0.001). Incorporation with age and FIGO stage in a nomogram improved predictive performance, with AUCs of 0.727, 0.700, and 0.656 at 1-, 3-, and 5-year OS, respectively. External validation in GSE140082 and GSE165808 confirmed model robustness, yielding 3-year AUCs of 0.630 and 0.826, respectively, demonstrating preserved prognostic value across independent cohorts and disease stages. Conclusions: This six-gene p-EMT transcriptomic signature demonstrates prognostic value in ovarian cancer and offers potential for individualized risk stratification and clinical decisionsupport.

---

Keywords

---

Supplementary Material

Supplementary Material File

---