Related Experiment Video
Updated: Aug 19, 2025
Competing-Risk Nomogram for Predicting Cancer-Specific Survival in Multiple Primary Colorectal Cancer Patients after Surgery
Published on: September 27, 2024
Development and external validation of a novel nomogram model for predicting postoperative recurrence-free survival
Li Ding1, Xiaobin Deng2, Wentao Xia1
1Department of Urology, the Affiliated Hospital of Xuzhou Medical University, Xuzhou, Jiangsu, China.
Background:
Transurethral resection of the bladder tumor with or without adjuvant intravesical instillation (IVI) has been the standard treatment for non-muscle-invasive bladder cancer (NMIBC), whereas a high percentage of patients still experience local tumor recurrence and disease progression after receiving the standard treatment modalities. Unfortunately, current relevant prediction models for determining the recurrent and progression risk of NMIBC patients are far from impeccable.
Methods:
Clinicopathological characteristics and follow-up information were retrospectively collected from two tertiary medical centers between October 2018 and June 2021. The least absolute shrinkage and selection operator (LASSO) and Cox regression analysis were used to screen potential risk factors affecting recurrence-free survival (RFS) of patients. A nomogram model was established, and the patients were risk-stratified based on the model scores. Both internal and external validation were performed by sampling the model with 1,000 bootstrap resamples.
Results:
The study included 299 patient data obtained from the Affiliated Hospital of Xuzhou Medical University and 117 patient data obtained from the First Affiliated Hospital of Guangxi Medical University. Univariate regression analysis suggested that urine red blood cell count and different tumor invasion locations might be potential predictors of RFS. LASSO-Cox regression confirmed that prior recurrence status, times of IVI, and systemic immune-inflammation index (SII) were independent factors for predicting RFS. The area under the curve for predicting 1-, 2-, and 3-year RFS was 0.835, 0.833, and 0.871, respectively. Based on the risk stratification, patients at high risk of recurrence and progression could be accurately identified. A user-friendly risk calculator based on the model is deposited at https://dl0710.shinyapps.io/nmibc_rfs/.
Conclusion:
Internal and external validation analyses showed that our model had excellent predictive discriminatory ability and stability. The risk calculator can be used for individualized assessment of survival risk in NMIBC patients and can assist in guiding clinical decision-making.
