Routine annual imaging does not aid in diagnosing new functional complications more than five years after RC with ileal conduit

Radical cystectomy (RC) with ileal conduit is a complex procedure known for its high morbidity. During the first five years of follow-up, regular oncological monitoring through imaging is advised, which also allows for the diagnosis of functional complications [1]. Cumulative long-term functional complication rates are reported to be as high as 94% over a period of 15 years [2]. Therefore, guidelines recommend continuing functional follow-up after the completion of oncological follow-up surveillance [3].

However, the value of annual imaging during functional follow-up remains unclear [4]. Limited literature is available regarding functional follow-up beyond five years [2, 5, 6]. As a result, there are no clear recommendations for appropriate follow-up after five years. This study investigates the utility of routine annual imaging in detecting functional complications in patients more than five years after RC with ileal conduit.

Patients with bladder cancer who underwent RC with ileal conduit between 2011 and 2018, in a high-volume comprehensive cancer center, and completed at least five years of follow-up were retrospectively included. Patients were excluded if follow-up was not continued when oncological follow-up ended, had progressive disease within the first 5 years after RC, or were diagnosed with uretero-enteric stenosis (UES) within the first five years after RC. UES was defined as obstruction on diuretic renal scintigraphy combined with deteriorating renal function. Before inclusion approval from the NKI-AvL Institutional Review Board (IRB) (approval number: IRBd23-102) was obtained.

Patient and treatment characteristics, along with follow-up details were collected from patients’ medical records up to September 2024 using Castor (v2024.1.0.3). All functional follow-up data beyond five years after RC was collected. Functional follow-up consisted of annual clinical assessments, laboratory evaluations (e.g., eGFR (eGFR was calculated using the CKD-EPI equation, based on serum creatinine values) and serum creatinine), and imaging of the abdomen or urinary tract (ultrasound or CT, as determined by the treating urologist). Certified radiologists reviewed routine annual imaging for evidence of new or progressive hydronephrosis compared to prior imaging. When indicated, patients underwent additional evaluation with diuretic renal scintigraphy and/or CT scanning, based on the urologist’s judgement of clinical relevance.

Our objective was to assess: 1) the incidence of new or progressive hydronephrosis which occurred beyond 5 years after RC and 2) the incidence of imaging-diagnosed complications i.e. UES and urolithiasis which occurred beyond 5 years after RC.

Outcomes were analyzed using descriptive statistics. Data analyses were performed using R‑Statistics (v4.4.1) [7].

Eighty-two of the 469 screened patients met our inclusion criteria (Fig. 1). Patient baseline characteristics are shown in Tab. 1. The mean duration of follow-up was 7.3 years (IQR 6–9) after RC. In total, 273 imaging studies were made during functional follow-up in the period after 5 years, comprising 202 ultrasounds and 71 CTs (Tab. 2).

Afbeelding vergroten

New or progressive hydronephrosis (which occurred beyond 5 years after RC) was identified in six patients (7.3%) during routine annual imaging, with one event per patient. In half of these cases (n = 3), additional imaging was performed. Of the three patients who underwent additional imaging, one was diagnosed with a renal pelvis tumor, which was not treated due to significant comorbidities. In the other two cases, a diuretic renal scintigraphy was performed. One showed delayed drainage without signs of obstruction and the other showed no abnormalities.

In the remaining patients, no further imaging was performed due to either the absence of clinical consequences related to significant comorbidity (n = 2), or, in one case, the absence of symptoms and preserved renal function, for which conservative management was chosen. In the latter case, hydronephrosis resolved spontaneously without intervention (Tab. 3).

Notably, the only case of UES (1.2%) (which occurred beyond 5 years after RC) observed in this study was not detected by routine annual imaging. No known risk factors, such as cardiovascular disease or prior radiotherapy, were present. The patient underwent an open procedure with a Bricker anastomosis. Hydronephrosis on the left side developed within the first year after radical cystectomy, prompting a diuretic renal scintigraphy that showed no signs of obstruction. Annual imaging thereafter showed no progression of hydronephrosis. In the seventh year of follow-up a diuretic renal scintigraphy was performed based on clinical symptoms (recurrent urinary tract infections (UTI) and a decline in renal function (creatinine 121 µmol/L, previously < 107 µmol/L)), which showed obstruction of flow and a non-functioning left kidney. No UTIs occurred thereafter and conservative management was pursued.

There was no urolithiasis observed in this study on either annual routine imaging or clinically indicated imaging.

To our knowledge, this is the first study to evaluate the diagnostic value of annual imaging beyond five years of follow-up after radical cystectomy (RC) with an ileal conduit. While imaging plays a critical role in detecting recurrent disease and early functional complications, its utility in long-term surveillance remains uncertain. In our cohort, no cases of urolithiasis were identified beyond five years after RC. Notably, only one UES (1.2%) (which occurred beyond 5 years after RC) was confirmed by imaging in our cohort. In this patient, routine annual imaging demonstrated no progression of hydronephrosis, yet clinical signs prompted further diagnostic evaluation that ultimately confirmed the UES. Additionally, two patients developed hydronephrosis on routine imaging, accompanied by a corresponding decline in renal function, raising clinical suspicion for UES. However, further imaging was not pursued in these cases, as significant comorbidities precluded intervention, rendering additional diagnostics clinically irrelevant. Including these clinically suspected cases results in a potential UES incidence rate of 3.7% in the period beyond 5 years after RC.

Our low incidence seem to contradict the study by Madersbacher et al. as they reported functional complications in up to 95% of patients with follow-up extending to 15 years post-RC [2].

However, their findings reflect the cumulative incidence of all functional complications, not the incidence of new-onset complications specifically diagnosed through imaging in the long-term. This distinction is critical, as cumulative incidence inherently includes early postoperative events and may therefore overestimate the risk of complications arising later in follow-up.

Notably, Madersbacher et al. reported that 9 of 131 patients developed urolithiasis more than five years after RC, with most presenting symptomatically. However, it remains unclear whether these cases required intervention, or occurred in patients with a prior history of urolithiasis. In contrast, no new cases of urolithiasis were identified in our cohort beyond five years postoperatively. This discrepancy may be partially explained by differing risk profiles, as none of the patients in our cohort had a documented history of urolithiasis.

Similarly, the interpretation of ileal anastomosis-related complications in their study is limited, as Madersbacher et al. did not report the cumulative incidence of ileal anastomosis-related complications—such as UES—as a distinct category. Furthermore, no median time to UES is provided. As a result, it is possible that all 13 reported cases of upper urinary tract stenosis (UES) occurred within the first five years postoperatively, which would align with our findings. The low incidence of UES observed in our cohort supports the notion that most cases are diagnosed early, typically within the initial years following radical cystectomy. This is consistent with previous studies with shorter follow-up durations, which also report that UES commonly presents within the first few postoperative years [6, 8‐11].

Furthermore, earlier studies have shown that the majority of patients with UES present with clinical symptoms [6, 8, 10]. Consistent with these findings, the single radiologically confirmed case of UES in our study was symptomatic and showed no progression of hydronephrosis on annual imaging. The clinically suspected UES cases presented with hydronephrosis on annual imaging, accompanied by either renal function decline or recurrent urinary tract infections. Notably, in one case, hydronephrosis was detected on routine imaging in the absence of any clinical symptoms, and spontaneous resolution was observed on imaging the following year. These findings further support the notion that a symptom-guided approach to imaging may be sufficient during long-term functional follow-up. Given that only one case of UES was radiologically confirmed and no cases of urolithiasis were observed, we are unable to recommend a specific threshold of renal function decline that should prompt additional imaging.

Our study has some limitations. Firstly, its retrospective design introduces the potential for selection bias, which may lead to either overestimation or underestimation of the true incidence. Not all patients completed the full 10-year follow-up, and some were discharged earlier. Those who remained under long-term follow-up may have had a higher risk of developing functional complications, potentially leading to an overestimation of incidence. Conversely, the loss of follow-up for some individuals may have resulted in missed cases of late-onset UES or urolithiasis, thereby risking underestimation.

Secondly, the cohort size is relatively small, and the number of patients with follow-up beyond five years is limited. This is primarily due to the exclusion of patients with oncological progression. However, this exclusion highlights the clinical relevance of our findings, as such patients would typically undergo continued imaging as part of oncological surveillance and may receive additional treatments that could independently affect the incidence of late complications. Lastly, two patients with new or progressive hydronephrosis did not undergo further imaging, as no clinical intervention was deemed necessary in these cases. While the presence of UES cannot be ruled out in these individuals, including them in the analysis would still result in a relatively low incidence rate of UES (3.7%). Moreover, this highlights the limited clinical impact of continued routine annual imaging in an aging population with increasing comorbidities, and further suggests that our cohort may reflect the characteristics and clinical course of the general population treated with RC and ileal conduit.

Our results suggest that new-onset functional complications diagnosed by imaging are rare more than five years after RC with ileal conduit. Moreover, routine annual imaging did not appear to contribute to the detection of these complications. Therefore, we propose that imaging should be performed primarily based on clinical indication. This approach may support a more efficient allocation of healthcare resources and is in line with the principles of value-based care.