ORIGINAL RESEARCH article
Front. Oncol.
Sec. Genitourinary Oncology
Predicting Postoperative Recurrence in Bladder Cancer by Integrating Circulating Tumor Cells with Clinical Variables Using Machine Learning
Provisionally accepted
- 1Department of Urology, The First Affiliated Hospital of Zhengzhou University, Zhengzhou, China
- 2The First Clinical Medical College, Zhengzhou University, Zhengzhou, China
Select one of your emails
You have multiple emails registered with Frontiers:
Notify me on publication
Please enter your email address:
If you already have an account, please login
You don't have a Frontiers account ? You can register here
Background: Postoperative recurrence is a major challenge in bladder cancer (BC). We aimed to develop a machine learning (ML) model using circulating tumor cells (CTCs) and clinical data to improve recurrence prediction. Materials and Methods: Clinical data and preoperative CTC counts from 137 BC patients were analyzed. Patients were grouped by a CTC cutoff of 4. Multiple ML models were developed and compared, with the best model interpreted using SHAP analysis. Results: High CTC count was significantly associated with advanced T stage, muscle invasion, and high tumor grade (all p < 0.05). Kaplan-Meier analysis showed significantly worse recurrence-free survival (RFS) in the high-CTC group (log-rank p < 0.001). Univariate Cox regression confirmed CTCs as a strong prognostic factor (HR = 4.51, p < 0.001). Among nine ML models, XGBoost achieved the highest predictive performance (AUC = 0.849). SHAP analysis identified CTC count, T stage, and tumor size as the most influential features. Conclusion: CTCs are a strong prognostic marker. Our ML model accurately predicts recurrence and could help guide personalized treatment, pending future validation.
Keywords: Bladder cancer, Bladder cancer1, circulating tumor cells, circulating tumor cells2, machine learning, Machine Learning3, recurrence prediction, Recurrence prediction4
Received: 23 Oct 2025; Accepted: 02 Feb 2026.
Copyright: © 2026 Li, Xu, Ye, Niu, Chen and Qi. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
* Correspondence: Li Qi
Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.