1. Introduction
Bladder cancer is among the most prevalent malignancies of the urinary tract and represents a major global health burden. It ranks as the 10th most common cancer worldwide and is especially prominent in men, with a significantly higher incidence compared to women. Despite advances in diagnosis and treatment, muscle-invasive bladder cancer (MIBC) continues to carry a high risk of disease progression and mortality [
1].
RC with urinary diversion remains the gold standard surgical treatment for patients with MIBC. While this procedure offers oncological control, it is associated with substantial physiological alterations, particularly affecting renal function over time [
2]. As life expectancy increases, long-term surveillance of the upper urinary tract in this patient population has become increasingly critical, yet clinical guidelines for such monitoring remain imprecise. However, concerns regarding long-term effects on the upper urinary tract have gained increasing attention. In particular, progressive changes in renal function and imaging-based anatomical alterations—such as hydronephrosis, reduced parenchymal thickness, and renal volume (RV) loss—raise critical questions regarding follow-up strategies [
3].
The exact pathophysiology of upper urinary tract deterioration following Bricker ileal conduit diversion is multifactorial. It typically involves a combination of chronic low-pressure vesico-ureteral reflux, subclinical ascending infections, and anastomotic stricture [
2]. Ureteroenteric strictures, in particular, occur in up to 3–15% of patients and are predominantly driven by ureteral ischemia, tension at the anastomosis, or extensive periureteral dissection [
4]. If these subtle morphological and functional changes are not detected early, chronic obstruction can lead to irreversible parenchymal atrophy and end-stage renal disease, underscoring the critical need for reliable, early radiological predictive markers.
This study investigates how the upper urinary system is affected in the long term after radical cystectomy and Bricker urinary diversion, based on objective radiological parameters. Furthermore, it explores whether every radiological alteration warrants intervention, whether thresholds can be proposed for action, and whether the anatomical trajectory of the left urinary tract predisposes it to greater impairment. By clarifying these issues, this study aims to contribute to evidence-based postoperative surveillance and decision-making.
3. Results
A total of 120 patients were evaluated in the study. Since radiological measurements were taken separately for each kidney, 240 renal units were evaluated.
Table 1 provides a comprehensive overview of their baseline characteristics, clinical findings, laboratory values, pathological features, and treatment details.
Hydronephrosis requiring intervention was observed in 12.5% (n = 15) of patients. Among these, 46.7% underwent antegrade double J stenting, and 53.3% required percutaneous nephrostomy. Acute kidney injury was the leading indication for intervention (46.7%), followed by renal colic pain (20%), tumor-related compression (20%), and ureterolithiasis (13.3%).
The median creatinine level increased from 0.9 mg/dL preoperatively to 1.1 mg/dL at 24 months, while the eGFR remained stable at a median of 60 mL/min/1.73 m2 throughout follow-up. Patients who underwent intervention had significantly higher creatinine levels at all time points (p < 0.05) and lower eGFR at baseline, 6 months, and 12 months compared to those without intervention.
Quantitative analysis of longitudinal changes demonstrated a progressive reduction in renal parenchymal thickness and volume over the 24-month follow-up period (
Table 2). The median parenchymal thickness of the right kidney decreased by 3.3% at 6 months and remained relatively stable thereafter, whereas the left kidney exhibited a more pronounced reduction, reaching 5.7% by month 24 (
Figure 3a). In terms of renal volume, the right kidney demonstrated a cumulative reduction of 5.3% at 24 months, while the left kidney volume declined more significantly, with a total decrease of 16.3% (
Figure 3b). These findings suggest an asymmetric pattern of renal deterioration following Bricker ileal conduit diversion, with the left kidney being more affected over time.
Hydronephrosis severity (SFU grading) showed no major progression in most patients. At 24 months, Grade 0–1 hydronephrosis was observed in 96.9% of right kidneys and 91.7% of left kidneys. Only 2.1% of right kidneys and 6.3% of left kidneys developed Grade 4 hydronephrosis.
Significant differences in renal parameters were observed between patients with and without hydronephrosis interventions. At the 12-month mark, the right renal pelvis AP diameter and the left kidney AP diameter to parenchymal thickness ratio were significantly higher in the intervention group (p = 0.008 and p = 0.039, respectively).
To further address renal functional changes and identify independent predictors of hydronephrosis intervention, serial renal function parameters and multivariable logistic regression analyses were performed. Potential multicollinearity among the longitudinal measurements and primary confounders (diabetes mellitus, hypertension) was rigorously evaluated using variance inflation factors (VIFs). All included variables demonstrated highly acceptable VIF values ranging between 1.00 and 1.12 (VIF < 3), indicating no significant multicollinearity. In the multivariable model incorporating bilateral AP diameter to parenchymal thickness ratios across preoperative and 6-, 12-, and 24-month time points, the 12-month left renal ratio emerged as a significant independent predictor of intervention (95% CI 1.74–13.50, adjusted OR = 4.84,
p = 0.003), while the 12-month right renal ratio and the preoperative left ratio demonstrated additional independent associations (
p < 0.05). ROC curve analysis demonstrated a promising apparent discriminatory performance of the 12-month left ratio (AUC 0.82). To account for potential overfitting due to the limited number of events, internal validation was performed via bootstrap resampling (B = 1000). The average model optimism was estimated to be 0.018, yielding an optimism-corrected AUC of 0.805. While these internally validated findings suggest a potential exploratory value, the small event size necessitates a cautious clinical interpretation. An optimal cutoff of 0.65 yielded 69% sensitivity and 88% specificity for predicting the need for intervention (
Figure 4). These findings validate that higher AP diameter to parenchymal thickness ratios, particularly at 12 months, are independently associated with subsequent hydronephrosis intervention and justify their use as predictive markers in this population.
When stratified by tumor stage, patients with T3–T4 disease exhibited significantly lower parenchymal thickness and renal volume in both kidneys at all time points compared to those with T0–T2 disease (p < 0.05). Similar reductions were observed in patients with lymph node metastasis and those who received adjuvant or neoadjuvant chemotherapy.
4. Discussion
This study aimed to evaluate the long-term impact of RC and Bricker ileal conduit urinary diversion on the upper urinary tract using objective radiological parameters. The most striking findings were the measurable decrease in renal parenchymal thickness and renal volume in both kidneys over a 24-month follow-up period, with significantly more deterioration observed on the left side. Moreover, while hydronephrosis was relatively common during follow-up, it did not universally necessitate intervention. Notably, patients who required intervention had significantly higher renal pelvic AP diameter to parenchymal thickness ratios, suggesting that this index may hold clinical value in identifying cases requiring active treatment.
A point worthy of clinical discussion is our utilization of the SFU grading system. Although universally recognized and highly reproducible for cross-study comparisons, the SFU system was inherently developed for pediatric and fetal populations. In adult patients who have undergone radical cystectomy and Bricker diversion, chronic non-obstructive upper tract dilatation due to altered reservoir compliance or low-pressure reflux is common. This physiological adaptation can reduce the clinical specificity of SFU grading when trying to pinpoint true mechanical obstruction. It is precisely this clinical limitation that drove the core objective of our study: because static visual grading is often insufficient in adults, we focused on quantitative, continuous geometric measurements, specifically the AP diameter-to-parenchymal thickness ratio, to directly integrate physical tissue loss into the clinical equation.
From a methodological standpoint, the retrospective design of this study limited the ability to assess certain causal relationships. Nonetheless, its strength lies in the objective and quantitative evaluation of renal changes based on serial CT measurements, including renal volume, parenchymal thickness, and hydronephrosis grading. The bilateral analysis of renal units and incorporation of percent changes over time further strengthened the ability to detect subtle yet clinically meaningful renal deterioration.
Clinically, these findings emphasize that upper urinary tract deterioration post-RC is not uniform and may be significantly influenced by anatomical factors related to surgical technique, especially concerning the left ureter. The requirement for greater dissection and transposition of the left ureter under the sigmoid mesentery likely contributes to the higher incidence of hydronephrosis and more pronounced renal deterioration on that side. Furthermore, our findings suggest that radiological findings alone may not always predict clinical outcomes, as not all patients with hydronephrosis required intervention, and many maintained adequate renal function during follow-up.
Yang et al. evaluated 249 patients who underwent either rectosigmoid (RS) pouch or ileal conduit diversion and reported a progression in mean serum creatinine levels from 1.02 mg/dL to 1.18 mg/dL over a 5-year follow-up period [
10]. Similarly, in our cohort, the median preoperative serum creatinine level was 0.9 mg/dL, which increased to 1.1 mg/dL at the 24-month postoperative follow-up. Regarding eGFR, although no statistically significant change was observed in median eGFR values during follow-up, a notable decline was detected in minimum eGFR values. This finding suggests that, while most patients maintained stable renal function, those who exhibited deterioration experienced a progressive decline. In line with our observations, Fujiwara et al. investigated early and intermediate renal function outcomes following RC with ileal conduit diversion and found that mean eGFR values decreased from 69.6 mL/min/1.73 m
2 at baseline to 60.8 mL/min/1.73 m
2 at 1 year postoperatively [
11]. Taken together, these results indicate that, although overall renal function remains stable in the majority of patients after radical cystectomy and ileal conduit urinary diversion, a subset of patients is susceptible to progressive renal impairment, emphasizing the need for individualized renal monitoring strategies.
Another key implication is the potential value of using specific radiological indices such as the renal pelvis AP diameter-to-parenchymal thickness ratio. This parameter showed consistent association with the need for intervention and may serve as an early indicator of clinically significant hydronephrosis, even in patients with otherwise stable serum creatinine or eGFR values. To our knowledge, no previous work has applied multivariable logistic regression to systematically evaluate radiographic predictors of clinically significant upper-tract obstruction in patients undergoing Bricker ileal conduit diversion. By integrating longitudinal CT-based measurements with objective clinical outcomes, the present study demonstrates that the renal pelvis AP diameter to parenchymal thickness ratio, particularly at 12 months, functions as an independent determinant of whether a renal unit ultimately requires intervention. This distinction is clinically important, as postoperative hydronephrosis after ileal conduit diversion is common but does not, in itself, mandate treatment unless accompanied by acute kidney injury, flank pain, or stone-related obstruction. Our findings therefore support the development of CT-based decision thresholds that can help differentiate benign postoperative dilatation from obstruction warranting decompression. The identification of a ratio threshold with high discriminatory accuracy provides a practical radiologic parameter that may guide surveillance strategies and refine intervention criteria in this patient population.
The findings also offer a nuanced view regarding the impact of perioperative chemotherapy. Patients receiving neoadjuvant or adjuvant chemotherapy experienced more pronounced radiological changes, particularly in renal pelvic dilation and parenchymal thinning, suggesting a potential additive nephrotoxic effect. These data underline the importance of close upper tract monitoring in patients receiving systemic therapies, especially when combined with anatomical factors that may compromise ureteral drainage.
Given that our study period spanned over a decade (2010–2024), potential temporal variations in clinical practice must be considered. However, all surgeries were performed at a single high volume tertiary academic center by senior staff urologists adhering to standard institutional protocols. Crucially, the core Bricker technique, the routine placement of standard 6Fr ureteral stents, and the postoperative stent retention period (21 days) remained homogenous throughout the study period. While the utilization of neoadjuvant chemotherapy evolved over the decade in accordance with guideline updates, this shift was explicitly analyzed as a distinct clinical variable. By stratifying for chemotherapy exposure, we were able to isolate its specific impact independent of the surgical methodology.
This study has several limitations that warrant careful consideration. First, its retrospective, single-center design restricts broad generalizability. Second, our strict selection criteria led to substantial attrition of the initial surgical population (from 537 down to 120 patients), introducing unavoidable selection and survivorship biases. Patients who dropped out of systematic imaging follow-ups may clinically represent a high-risk group involving early oncologic recurrence, severe complications, or death. Consequently, our final analyzed cohort likely reflects a healthier, more stable subpopulation, which may understate the true incidence and scale of long-term upper urinary tract deterioration. Furthermore, the baseline exclusion of prior CKD biases our long-term functional survival data toward more optimistic outcomes. Third, although we adjusted for major comorbidities (diabetes, hypertension) and clear clinical events (sepsis, acute kidney injury), retrospective models cannot fully account for unmeasured confounders, such as subclinical recurrent urinary tract infections, patient hydration habits, or mild metabolic acidosis fluctuations. Finally, our follow-up period was limited to 24 months. While a 2-year timeline is robust for identifying early-to-intermediate structural shifts and early failure patterns, it is too short to capture late-onset benign strictures or chart lifetime renal functional trajectories, which typically require 5 to 10 years of longitudinal tracking. Renal function assessment relied on serum creatinine and eGFR without functional imaging, potentially limiting the evaluation of segmental renal function.
Moving forward, the clinical application of the proposed AP diameter to parenchymal thickness ratio could be significantly enhanced by the integration of artificial intelligence (AI) and automated 3D radiomic segmentations. While manual CT measurements and the ellipsoid volume formula provide practical clinic-based estimations, AI-driven automated volumetric assessments could eliminate inter-reader variability and offer more precise real-time surveillance of parenchymal loss. Future prospective, multi-center studies should not only externally validate our 0.65 threshold, but also incorporate functional dynamic imaging (e.g., MAG3 renography) and novel serum renal biomarkers alongside these structural geometric indices. Such multimodal approaches will ultimately facilitate the development of personalized, risk-stratified follow-up algorithms for patients undergoing radical cystectomy.