Long-term complications and functionality after continent cutaneous urinary diversion

Radical cystectomy with a urinary diversion is the gold standard treatment for non-metastatic muscle invasive bladder cancer and in some cases high-risk non-muscle invasive bladder cancer [1]. Urinary diversions are also performed for benign indications or for non-urothelial tumours involving the bladder [2]. Patients are often offered a choice between different types of urinary diversions. Therefore, providing accurate information on the functional outcomes of different urinary diversions is crucial in the shared decision-making process. Whereas large studies exist for common diversion types, such as the incontinent ileal conduit and the continent orthotopic neobladder, only limited data exist to guide patients on the expected functional outcomes of continent cutaneous urinary diversion (CCUD). CCUDs are a continent alternative in cases where an orthotopic neobladder is contraindicated, such as tumour involvement of the trigone/bladder neck, the urethra or the prostate [3].

The CCUD is constructed as a low-pressure reservoir with a high capacity, voided by catheterisation through a small cutaneous stoma. Since the first modern CCUD, the Kock pouch, first described in 1969, many modifications have emerged. Currently, the Indiana pouch, which utilises the ascending colon and terminal ileum, is the CCUD performed at our institution [4, 5]. We have previously published short-term complication rates for all patients with bladder cancer undergoing cystectomy at our centre [6]. This study aims to assess the functional outcomes and late complications after the construction of a CCUD at a specialised tertiary centre in Denmark.

Methods

Patients

Patients who received a CCUD were identified from a pre-existing database of all bladder cancer patients having undergone cystectomy at the Copenhagen University Hospital – Rigshospitalet. The study period was defined as January 2010 to December 2023 based on the availability of Electronic Medical Records (EMR). A total of 53 patients were identified. Patients who died before 2017 were excluded due to the unavailability of data after the implementation of a new EMR system in 2017, leaving 46 patients for inclusion. An additional four patients who received a CCUD for other reasons than bladder cancer were identified by procedure codes and were included in the study, totaling 50 patients eligible for analysis.

At our institution, CCUDs are performed as an Indiana pouch. To be eligible for the procedure, patients must demonstrate having the mental and physical capacity for clean intermittent catheterisation (CIC), have no pre-existing colonic disease and be suitable for high-risk surgery.

Surgical procedure

Cystectomy with CCUD was performed as open surgery throughout the entire inclusion period by four surgeons using the same surgical technique. All four were trained onco-urologists with comprehensive experience in reconstructive surgery.

An illustration and a thorough description of the Indiana pouch procedure are provided in Figure 1.

Definition of variables

The Charlson Comorbidity Index (CCI) was evaluated without age adjustment. Due to varying indications for undergoing surgery, “localised tumour” was included in the CCI for cancer patients.

Precise data on alcohol consumption was not available. If present, alcohol overconsumption was documented in the EMR and classified according to the recommendations from the Danish health authorities. Before 2022, overconsumption was defined as > 14 units/week for men and > 7 units/week for women. From 2022, overconsumption was defined as > 10 units/week for all.

Late complications were defined as appearing between 90 days after surgery and the last follow-up and were graded according to the Clavien-Dindo Classification [7]. Time to first continence after catheter removal was defined as the first mention of continence (no reported leakage for a minimum of three consecutive hours, night and day) in the EMR.

Any leakage at six or 12 months was defined as any instance or period where a leak of urine, regardless of quantity, was mentioned in the EMR at six or 12 months.

Patients have a direct number to the department that they may use as needed, and calls are documented. Personal care items for temporary use were supplied by the department and documented in the EMR. In the absence of any record of leakage, patients were presumed to be continent.

Urinary tract infections (UTIs) were defined as symptomatic UTIs requiring treatment. Recurrent UTI was defined as at least two UTIs within six months or three UTIs within one year, in accordance with Danish national guidelines [8].

The preoperative estimated glomerular filtration rate (eGFR) was defined as the last measurement before surgery, and the post-operative eGFR was defined as the latest available measurement. Chronic kidney disease (CKD) stages were defined using the National Kidney Foundation’s Kidney Disease Outcomes Quality Initiative guidelines (K/DOQi): G1-G2 was defined as normal to mildly decreased eGFR > 60 ml/min.; G3a-G3b was mildly to severely decreased eGFR of 59-30 ml/min.; G4 was severely decreased eGFR of 15-25 ml/min. [9].

Intervention for cancer recurrence was defined as treatment with chemotherapy, radiation therapy, immunotherapy or cancer surgery.

Outcomes

The primary outcome was the rate of continence after cystectomy with CCUD. The secondary outcomes were the rate and cumulative risk of reinterventions, the rate of other late complications and post-operative kidney function.

Statistics

Categorical variables were presented as numbers and percentages, continuous variables as medians and IQR. Outcomes were presented as numbers and percentages of the population and median (IQR) days since surgery. The cumulative risk of reintervention was calculated using a competing-risks analysis, with death before reintervention as the competing risk. The curve was truncated at ten years because only one patient had follow-up data beyond that point. A sub-analysis of patients operated between 2017 and 2023 was performed to test for the effect of survival bias. The reverse Kaplan-Meier was used for calculating the median follow-up. All analyses were performed in SPSS (version 29) and R (version 2024.04.2).

Trial registration: not relevant.

Results

Table 1 presents patient demographics. The median age at surgery was 59 years (IQR: 51-66 years), with equal numbers for males and females. Twelve patients had prior abdominal surgery (24%). The indications for cystectomy with CCUD were non-metastatic bladder cancer in 46 patients (92%), vesico-vaginal fistula in one patient, interstitial cystitis in one patient, melanoma in one patient and radiation-induced cystitis in one patient.

Table 2 outlines the functional outcomes. All patients were able to manage CIC after a median of 23 days (IQR: 20-33 days) after cystectomy. Continence was achieved by 92% of patients at a median of 167 days (IQR: 75-229 days) after cystectomy. The incidence of any leakage was 43% after the first six months and 29% after 12 months.

The preoperative median eGFR was 86 ml/min. (IQR: 71-96 ml/min.), with five patients (10%) in CKD stage G3a-G3b, and with the lowest recorded eGFR being 38 ml/min. At the last post-operative follow-up, the median eGFR was 79 ml/min. (IQR: 57-93 ml/min.), with 14 patients (28%) in CKD stage G3a-G4 and the lowest recorded eGFR being 17 ml/min.

Table 3 shows late complications and reinterventions graded according to the Clavien-Dindo Classification. In total, 38 patients (76%) experienced at least one late complication, of whom nine patients (18% of the total population) had one complication, ten patients (20% of the total population) had two complications and 19 patients (38% of the total population) had three or more complications. The most frequent late complications included treatment of recurrent UTIs at one year (n = 12, 26% of patients with CCUD at one year), diagnosis of incisional hernia (n = 10, 20%), revision of pouch ileal segment (n = 10, 20%) and revision of stomal stenosis (n = 8, 16%).

A total of 33 patients (66%) underwent at least one reintervention, with a median time of 366 days (IQR: 192-581 days) after surgery for the first reintervention. Supplementary Figure 1 shows the cumulative incidence curve of the first reintervention. The absolute cumulative risk of reintervention was 33% (95% CI: 19.8-46.3%) at one year, 71% (95% CI: 57.3-85.3%) at five years and 78% (95% CI: 62.4-93.6%) at ten years. The most frequent indications for reintervention were a suboptimal course of the pouch ileal segment and stomal stenosis. Ten patients (20%) had a reintervention of the ileal segment, and eight patients (16%) had a revision of the stoma due to stenosis. Two patients (4%) opted to convert functioning CCUDs to an ileal conduit at 215 and 1,533 days after the primary surgery due to a change in preference.

A sub-analysis of patients operated after 2017 is shown in Supplementary Table 1, demonstrating a similar continence rate and lower complication rates. The analysis included 33 patients, of whom 30 (91%) became continent. The types of late complications were similar to those in the original population.

Discussion

In this retrospective long-term follow-up study of 50 patients who underwent cystectomy with CCUD (Indiana pouch), we identified a continence rate of 92% and a reintervention rate of 66%. Studies on other CCUDs have found similar continence rates in the 80-95% range [10-12].

Two other studies have assessed long-term outcomes in patients after CCUD construction [5, 13]. A 2016 systematic review, including 519 patients, reported late complications and long-term functional outcomes. Individual cohort sizes ranged from 26 to 125 patients, with follow-up from 12 to 56 months [5]. A retrospective study from 2024 reported long-term outcomes of 33 patients who underwent an Indiana pouch with a median follow-up of 21.5 years [13]. Although not all studies included in the systematic review provided continence rates, both papers presented high continence rates (95% and 94%) that are comparable to our findings [5, 13]. The same two papers presented reintervention rates of 6-69% and 67%, which are also comparable to the 66% observed in our study. Likewise, the study showed pouch calculi rates of 10% and 18%, which are comparable to the 10% observed in our population. Our finding of 4% ureteral anastomosis strictures was low compared with the 3-27% reported in the two other papers [5, 13].

Complications after CCUD have been reported to be higher than for the ileal conduit and orthotopic neobladder. Among 32 articles included in a meta-analysis of patients receiving a urinary diversion, 21 studies covering 33,742 patients reported long-term post-operative complication rates of 62% for patients receiving an ileal conduit and 60% for patients receiving an orthotopic neobladder [14].

In previous studies, the incidence of incontinence for patients who had previous pelvic radiation therapy was high (approximately 25%). This highlights the need for careful preoperative counselling [5, 15, 16]. Only three patients in our cohort had received radiation to the pelvic area before surgery.

In the present study, stone formation was identified in 4% of patients in the upper urinary tract and 10% in the CCUD, whereas previous reports have documented rates of up to 30% [17-19]. A possible explanation is the use of resorbable sutures rather than staples during pouch construction. Two prevalent late complications were related to the efferent limb: stenosis at the skin stoma in the ileal segment itself. Stomal stenosis is reported in 16-38% of cases in the literature, whereas our observed incidence was in the lower end of this range [16, 19]. Revision of the pouch’s ileal segment was reported to occur in as many as 47% of cases in other studies, which is notably higher than the 20% observed in the present study [13, 17].

Among the patients in this study, 18% transitioned from G1-G2 preoperatively to G3a-G4 categories after surgery. Recently, a larger study conducted at our institute investigated the decline in renal function after radical cystectomy with all types of urinary diversions offered at our institute [20]. The study found that the hazard ratio for developing renal insufficiency was 2.4% (95% CI: 1.4-3.9%, p < 0.001) for continent diversion types (orthotopic neobladder and CCUD) compared with the ileal conduit [20].

We found a higher overall rate of late complications than in other cohorts. This might be explained by the completeness of data in the Danish EMR or the fact that we included complications that did not require surgical intervention, such as UTIs, contrary to other studies. Conversely, some of the most common late complications were less prevalent in our cohort than in other studies.

This study does not address short-term complications, as these were previously published from an overlapping database [6]. The study included 729 patients with bladder cancer who had undergone cystectomy with any type of urinary diversion and found no significant difference in the rate of major complications between diversion types within 90 days. Comorbidity, age, and tumour stage are generally lower in patients who receive CCUD, possibly masking an actual higher complication rate. At our institution, patients are counselled that the choice of diversion may increase diversion-related complications.

Limitations

This cohort was treated at a single centre where only selected patients meeting the defined criteria were offered a CCUD, resulting in a selected, small population of relatively healthy individuals. Another limitation was the lack of a prospectively applied standardised definition of leakage and alcohol overconsumption throughout the study period. Generalisability to a wider population may therefore be limited. As inherent to retrospective designs, this study encountered selection and survivorship bias attributable to the inability to include cases where the patient died before 2017. To investigate bias, a sub-analysis was performed, showing similar rates of continence and a lower rate of complications after 2017. This is likely due to the shorter length of follow-up compared with the 2010-2023 population.

Conclusions

In this study, we found that CCUD offered to selected patients produced high rates of three-hour continence and CIC success. Two in every three patients needed at least one intervention within the follow-up period, and one in five patients transitioned from normal kidney function to renal insufficiency. This study adds to information on the benefits and risks of CCUD and may aid physicians and patients in the shared decision-making process when discussing different urinary diversion options.