UTUC originates from the urothelium of the renal pelvis or ureter can seriously threaten a patient’s life [8, 9]. However, the epidemiological and biological characteristics of UTUC in the Chinese population have not been a research focu.Consequently, data are limited, and it remains unclear which factors are associated with bladder cancer recurrence in these patients [10]. Bladder cancer recurrence not only increases the economic burden on UTUC patients, but some patients may even require radical cystectomy, which increases the risk of ureteroenteric anastomotic stricture and severely affects their quality of life [11,12,13,14]. Accurately predicting the prognosis of UTUC patients helps with risk stratification, thereby providing a reference for urologists or oncologists in formulating treatment plans.

The results of the multivariate analysis in this study found that T3-T4 tumor stage, G3 tumor grade, history of bladder cancer, and preoperative ureteroscopy were independent risk factors for postoperative bladder cancer in UTUC patients, while preventive intravesical instillation was an independent protective factor. Wu et al. [15] found that T stage, N stage, and tumor grade were risk factors associated with bladder cancer recurrence after surgery in UTUC patients. Our findings are consistent with theirs; we found that patients with T3-T4 tumor stage had a 4.512-fold higher risk of postoperative bladder cancer recurrence than patients with T1-T2 stage, and patients with G3 grade had a 4.943-fold higher risk than patients with G1-G2 grade. Therefore, for high-grade, high-stage UTUC patients, close follow-up and cystoscopy are necessary. However, the aforementioned study also found that tumors located in the ureter were more prone to bladder cancer recurrence, mainly because the anatomical proximity of ureteral tumors to the bladder, and the mechanical stress caused by higher urine flow rates and greater bladder pressure, promote the easy spread of ureteral tumor cells to the bladder. In contrast, our study did not find a significant effect of tumor location on bladder cancer recurrence, possibly due to the small number of patients with ureteral tumors included, which may have biased the statistical analysis results. Therefore, larger, multicenter studies are needed to clarify this association.

Wang et al. [16] found that a history of bladder cancer is a factor affecting the long-term survival rate of UTUC patients. This may be because a preoperative history of bladder cancer suggests a higher likelihood of tumor cell seeding within the urinary tract during surgery, thereby increasing the risk of subsequent bladder recurrence. Mertens et al. [17] found that ureteroscopy is associated with a high risk of bladder cancer recurrence. Ureteroscopy, with or without biopsy, is commonly used to definitively diagnose UTUC and determine its grade and subsequent risk category. The primary proposed mechanism for this association is cancer cell seeding; ureteroscopy can increase the risk of tumor cell shedding from UTUC lesions and their subsequent seeding in the bladder. Therefore, when clinically diagnosing UTUC, to potentially reduce the risk of bladder recurrence, clinicians might consider prioritizing other reliable imaging methods over ureteroscopy, weighing the diagnostic benefits against the risk of seeding. The results of a study by Fan et al. [18] showed that intravesical instillation can significantly reduce the bladder recurrence rate in UTUC patients undergoing radical nephroureterectomy.Moretto et al. [19]similarly demonstrated that intravesical instillation significantly reduced the bladder recurrence rate at 12 and 24 months after radical nephroureterectomy.Intravesical instillation can reduce the risk of bladder cancer by directly introducing chemotherapy drugs into the bladder, thereby killing residual tumor cells. Additionally, regarding the impact of surgical approach on postoperative recurrence, the study by Cella et al. [20] showed no significant differences between open surgery and robot-assisted radical cystectomy in terms of postoperative complications and perioperative outcomes. Similarly, our study found no significant effect of different surgical approaches on postoperative bladder cancer recurrence.

Nomograms can be used to calculate the probability of an outcome for individual patients and are of great value in clinical practice [21, 22]. Luo et al. [23] provided a nomogram based on history of bladder cancer, pathological stage, lymphovascular invasion, etc., to predict the probability of extra-urinary tract recurrence in UTUC patients, with model AUCs of 0.793 for both the training and validation sets. Zhang et al. [24] constructed a nomogram prediction model for the overall survival rate of second primary cancer in UTUC patients based on age, tissue type, grade, etc., with a predictive AUC as high as 0.737. Chou et al. [25]developed an early postoperative recurrence prediction model for UTUC patients based on diabetes, tumor necrosis, and pathological T stage, with a predictive AUC of 0.84 and an external validation AUC of 0.76. Internal validation showed that the nomogram model constructed in this study based on risk factors has certain efficacy in predicting postoperative bladder cancer in UTUC patients, with predictive AUCs of 0.864 and 0.831 in the training and validation sets, respectively, which are generally consistent with the aforementioned studies. Furthermore, the calibration curves further confirmed the good calibration of the model, and the DCA curves demonstrated its high net benefit and clinical applicability. Using the optimal cutoff value of 54% from the training set’s ROC curve, the model demonstrates potential clinical utility. At this threshold, for every 100 patients, applying the model to guide interventions would be equivalent to a strategy that successfully identifies 11 additional at-risk patients without increasing harm compared to not intervening. Further external validation confirmed the model’s performance, achieving a high predictive AUC of 0.847. The calibration and DCA curves likewise indicated good model calibration and high clinical utility, respectively. This indicates that the nomogram prediction model constructed in this study has potential value in the clinical setting, facilitating the screening of high-risk patients for bladder cancer and guiding clinical decision-making.

Example of Nomogram Model Application: For example, a UTUC patient with tumor stage T3–T4 would receive a weight of 85.46 points; tumor grade G1–G2 adds 0 points; a history of bladder cancer adds 74.12 points; preoperative ureteroscopy adds 87.59 points; and absence of prophylactic intravesical instillation adds 99.18 points. The total score is 346.35, corresponding to a predicted postoperative bladder cancer risk of 0.81 (81%).For UTUC patients with a predicted risk exceeding 54%, clinicians can use the nomogram results to effectively communicate with the patient. Prophylactic intravesical instillation with mitomycin C or pirarubicin may be initiated promptly after surgery. In addition, regular postoperative follow-up with urine cytology and cystoscopy should be performed to facilitate early detection of potential malignant changes.Patients should also be advised to increase physical activity, adopt healthy eating habits, and avoid the development of an acidic body constitution. Psychological support and interventions may also be provided, along with rehabilitation measures such as bladder function training and pelvic floor muscle exercises.

However, this study still has limitations. Firstly, as a retrospective study, some surgical details and key variable data were missing. For example, the use of access sheaths and ureteral stents during or after ureteroscopy was not considered. In addition, important factors such as the type of ureteroscopy, time from diagnosis to surgery, histologic variants, and time to recurrence were not included in the analysis.Secondly, due to the limitations of single-center data, genetic or molecular biomarkers could not be incorporated into the model.Thirdly, the external validation cohort was relatively small; therefore, further large-scale external validation with larger, multicenter datasets is required. Fourthly, there may be potential selection bias, as the relatively strict exclusion criteria may limit the generalizability of the findings to the broader UTUC population. Fifthly, while the number of variables included in the final model met the EPV criterion, the risk of model overfitting cannot be entirely ruled out. Future prospective studies with larger sample sizes, potentially using methods like LASSO regression analysis, are planned to minimize this risk. Sixthly, the study lacks survival follow-up data and did not include Cox regression analysis to evaluate patient survival time and status.Despite these limitations, the validation performance of the nomogram model in this study in both the training and validation sets confirms the reliability of our findings.

In summary, this study developed and validated a novel nomogram for predicting the risk of bladder cancer recurrence in UTUC patients. This tool can assist clinicians in easily and effectively identifying high-risk patients for bladder cancer, thereby providing valuable support for clinical decision-making, facilitating the development of personalized treatment strategies, and ultimately improving patient outcomes.