Comparison of Outcomes of a Ureteroileocecal Appendicostomy Based Urinary Reservoir In Radiated And Non-Radiated Patients

Abstract

Purpose

The ureteroileocecal appendicostomy (UIA) reservoir is designed to potentially decrease the morbidity of continent diversion, particularly in previously radiated patients. We report our experience with this reservoir to compare complication among radiated (RT) and non-radiated (NR) patients.

Materials and Methods

Fifty-two consecutive patients who received the UIA diversion between 3/2001 and 1/2008 were evaluated. Outcomes were analyzed according to whether patients received radiation therapy to the pelvis. Complications were reported as early (within 90 days of surgery) or late.

Results

Overall, 29 patients received radiation therapy. The incidence of early complications requiring operative intervention was 14%, affecting 2 patients (9%) in the NR group and 5 (17%) in the RT group (p=0.68). All but one reoperative procedure was performed to revise the stoma. Early urinary tract infections occurred in 17% of NR and 28% or RT patients (p=0.51). The most common late complication was stomal stenosis, which required outpatient dilatation; this occurred in 21% of patients, 4 (17%) in the NR and 7 (24%) in the RT group (p=0.74). Late ureteral complications requiring intervention were reported in 15% and occurred in 2 of 44 renal units (5%) in the NR and 6 of 56 (11%) in the RT group (p=0.21).

Conclusions

The ureteroileocecal appendicostomy is a safe and effective technique for providing continent cutaneous urinary diversion in heavily radiated patients. Complication rates did not significantly differ between radiated and non-radiated patients and appear improved compared to previously reported literature.

Introduction

Advances in urinary tract reconstruction have provided a variety of options for urinary diversion to patients who require cystectomy. Presently, patients who undergo radical cystectomy, anterior pelvic exenteration, or total pelvic exenteration for tumors of the pelvic soft tissues or genitourinary, gynecologic, or colorectal systems may be eligible to receive a continent diversion directly connected to the remnant urethra or to the skin as a cutaneous reservoir. Even in the present era of orthotopic reconstruction, the continent cutaneous reservoir (CCR) remains important in providing an appliance-free diversion alternative in patients who are not candidates for urethral preservation.

Early experience with CCR’s demonstrated they could be safely created at the time of cystectomy and prospective contemporary studies that have meticulously documented the complications associated with radical cystectomy and urinary diversion note that CCR’s are not associated with a higher rate of severe complications¹. Longer term studies have documented that the most frequent pouch-related complications associated with CCR’s are related to the stoma (difficulty catheterizing, incontinence) and the ureteroenteric connections. Prior pelvic radiation was identified as a risk factor for experiencing more frequent pouch related complications following CCR formation^2–5.

Numerous variations in technique are employed in the construction of CCRs, all of which share the basic components of a detubularized bowel reservoir and a cutaneous catheterizable stoma that resists urine flow to maintain continence. The appendix as a catheterizable continence mechanism was first described by Mitrofanoff et al. in the 1980s⁶ and later described as part of a modification of the Mainz pouch that employed a flap valve continence mechanism created by tunneling the appendix within the tinea of the cecum.⁷ Significant experience with this design demonstrated a high degree of continence leading to decreased need for major revisionary procedures⁸; however, concern was raised with the use of the appendix in patients who had received high doses of radiation⁴. Subsequent variations in the Mainz appendix–based reservoirs consisted of preservation of the terminal ileum such that a tubularized ureteral substitution segment could be fashioned.^{9, 10}

In an effort to minimize overall urinary reservoir–related complications in all patients undergoing CCR formation, particularly in heavily irradiated patients, we have employed a modified ileocecal appendicostomy reservoir that utilizes a submucosally tunneled appendiceal valve for continence and a tubularized terminal ileal segment as a ureteral substitute to allow for a more proximal ureteroenteric anastomosis. In the present study we compare our experience with our modified UIA reservoir between previously radiated and non-radiated patients.

Material and Methods

Following institutional review board approval, we identified 52 consecutive patients who received the UIA diversion between 3/2001 and 1/2008 by a single surgeon (BHB). The procedures performed before reconstruction included either radical cystectomy or anterior pelvic or total pelvic exenteration. Twenty-nine patients received prior or intraoperative radiation therapy and 23 patients underwent surgery without prior radiation.

Patients were selected to undergo the UIA diversion after preoperative counseling and assessment of motivation, manual dexterity, and adequate renal function (serum creatinine <2.5 mg/dL). The UIA pouch was the continent reservoir design of choice and performed in all patients who had a patent appendix of adequate length and caliber identified at the time of surgery. Post-operative follow-up occurred at 3–4-month intervals for the first 2 years, then every 6 months thereafter. Postoperative evaluations included physical exam, serum electrolyte analysis, and radiographic examinations. The radiographic imaging studies included a CT scan at 3–4 months postoperatively and every 6–12 months thereafter.

Early complications were defined as those that occurred within 90 days after surgery; late complications were defined those that occurred >90 days after surgery. The focus of this report is on infectious, stoma, pouch, and ureterointestinal complications. The reported complications were graded according to the extent of intervention required for management. Any complication resulting from the reservoir that required a percutaneous or surgical intervention, prolonged hospitalization, or re-hospitalization was categorized as major; all other complications were categorized as minor. Data were analyzed using Stata 8.2 (Stata Corp., College Station, TX) with Fischer’s exact test used to determine p values.

Surgical Technique

Construction of the reservoir has previously been reported and consists of four components: the creation of the continence mechanism, detubularization and formation of the reservoir, completion of the ureteroileal anastomoses, and maturation of the appendiceal stoma.⁹ Several surgical details require highlighting. The length of terminal ileum left contiguous with the cecum was typically 12–15 cm; this may, however, be increased to provide the desired length needed to serve as a partial or total ureteral substitution segment. The total length of right colon segment used for the reservoir segment was 20 cm.

The appendix was initially evaluated to determine the adequacy of its length and lumen. A luminal diameter that could accommodate a 12–14 F catheter was considered adequate. Full-thickness flaps on either side a of a 4-cm incision along the anterior tinea of the cecum were developed to create the submucosal tunnel. The appendix was secured into position using seromuscular sutures placed into either side of the cecal wall flaps through the openings made in the first 3–4 windows of Deaver in the mesoappendix.

Non-pelvic portions of the ureters were used for the ureterointestinal connections, particularly in patients with prior radiation therapy, to ensure that a non-radiated segment was utilized for the anastomosis. Both ureters were stented with 8.5 F catheters that were temporarily externalized following surgery.

The preferential site for placement of the stoma is at the umbilicus; if, however, the umbilical area was deemed unsuitable, an alternative site was chosen in the right lower quadrant.

Results

Overall, 52 patients were included in the study. Twenty-nine patients had received radiation therapy, 18 (62%) of them male and 11 (38%) female. Twenty-three patients were non-radiated, 15 (65%) of them male and 8 (35%) female. Of those who received radiation therapy, 5 (17%) received intraoperative only, 10 (34%) preoperative only, 13 (45%) both preoperative and intraoperative, and 1 (3%) postoperative radiation therapy. All patients underwent surgical extirpation of pelvic malignancies; clinical characteristics and primary tumor types are outlined in Table 1. The non-radiated group consisted of patients with primary bladder cancer undergoing radical cystectomy or anterior pelvic exenteration in 100% of cases. All female non-radiated patients underwent anterior exenteration while males underwent radical cystoprostatectomy. Of the radiated group, 17 patients (59%) underwent total pelvic exenteration with diverting colostomy or ileostomy, while none of the non-radiated patients required bowel diversion.

Table 1.

Patient clinical characteristics and primary tumor histology

Characteristic	Nonirradiated	Irradiated
Av age	57.1	53.9
No. radiation type:
Preop	0	10
Intraop	0	5
Preop + intraop	0	13
Postop	0	1
No. surgical procedure:
Anterior pelvic extenteration	14	9
Radical cystoprostatectomy	9	3
Total pelvic exenteration	0	17
No. gynecologic malignancy site:	0	16
Cervix	—	6
Uterus	—	5
Vagina	—	3
Vulva	—	2
No. other malignancy site:
Prostate	0	7
Bladder	23	0
Sarcoma	0	5
Rectum	0	1

The mean follow-up, hospital stay, and intraoperative blood loss for each of the groups are shown in Table 2. Mean follow-up was 26.3 months (standard deviation [SD] 18.8; 95% confidence interval [CI] 18.1, 34.7) for the non-radiated group and 23.7 months (SD 18.9; 95% CI 16.4, 38.9) for the radiated group (p=0.626). Forty-five patients had more than 6 months follow-up. Thirty-six patients were alive at last follow-up: 19 of the 23 non-radiated patients (83%) and 17 of the 29 radiated patients (59%).

Table 2.

Early complications by patient group

Complication Grade	No. Nonirradiated (%)	No. Irradiated (%)	p Value
Overall	23	29
Stomal:	4 (17)	6 (21)	1.00
1–2	2	1
3–4	2	5
Pouch:	0	2 (7)	0.50
1–2	—	—
3–4	—	2
Ureterointestinal:	0	1/29 (3)	0.89
1–2	—	—
3–4	—	1
Infectious:	4 (17)	8 (28)	0.51
1–2	3	4
3–4	1	4

Appendix/stoma complications

The incidence of early stomal-related complications for the non-radiated group was 17% and 21% for the irradiated group (p=1.00) (Table 3). The incidence of early complications requiring operative intervention was 9% in the NR group and 17% in the RT group (p=0.37). All but one reoperative procedure was to revise the stoma. The cause for the observed appendiceal-skin disruptions was either early traumatic catheterization or a periumbilical wound infections.

Table 3.

Late complications by patient group

Complication Grade	No. Nonirradiated/Total No. (%)	No. Irradiated/Total No. (%)	p Value
Stomal:	5/23 (22)	7/29 (24%)	0.84
1–2	4	6
3–4	1	1
Pouch:	0/23	1/29 (3)	0.37
1–2	—	—
3–4	—	1
Ureterointestinal:	2/44 (5)	6/56 (11)	0.21
1–2	0	0
3–4	2	6
Infectious:	4/23 (17)	4/29 (14)	0.94
1–2	2	3
3–4	2	1

The incidence of late stomal complications was 22% in the non-radiated group, with 2/23 (9%) requiring surgical revision secondary to a traumatic catheterization and 3/23 (17%) requiring dilatation in the ambulatory setting. The incidence of late stomal complications in the irradiated group was 24% (p=0.84) (Table 3). Late stomal complications (n=7) in the irradiated group consisted primarily of difficulty with catheterization (n=6) that were successfully managed with office dilatation; none required surgical revision. Two patients required in-office dilatation and subsequently requested persistent urinary drainage via an indwelling stomal catheter. Four additional patients reported difficulty with catheterization and 2 required a single in-office dilatation.

Ureterointestinal complications

One early ureterointestinal complication was observed (1/100 ureteroileal anastomoses, 1%). and consisted of a perioperative leak, in a previously irradiated patient which resolved within 2 weeks following proximal diversion with nephrostomy tube diversion. The incidence of late ureteroileal anastomotic complications occurred in 2 of 44 renal units (5%) for the non-radiated group and in 6 of 56 renal units (11%) for the radiated group (p=0.21) (Table 3); all of these complications were ureteral strictures requiring antegrade endoscopic management following nephrostomy tube placement. One patient subsequently underwent laproscopic nephrectomy after failed endoscopic management.

Reservoir complications

Two early pouch-related complications were noted in the non-radiated group (Table 3) . One of these complications, which occurred in a previously radiated patient who received total pelvic exenteration, was related to an inadvertent pouch injury that occurred as part of an emergent re-exploration for pelvic bleeding. This resulted in a prolonged urine leak that required bilateral percutaneous nephrostomy tubes. An entero-pouch fistula subsequently developed and eventually the reservoir was converted to an ileal conduit. The second complication was a pouch leak that resolved after percutaneous drainage. A single late pouch complication was identified, which occurred in a radiated patient and consisted of a pouch stone that was managed by percutaneous litholopaxy.

Urinary infections

The incidence of early infectious complications was 17% for the non-radiated group and 28% for the radiated group (p=0.51). Three patients in the irradiated group (10%) required re-admission or had their hospitalization extended because of a urinary tract infection. The incidence of late infectious complications was 17% for the non-radiated group and 14% for the radiated group (p=0.94). Overall, 5% of patients with late infectious complications necessitated hospitalization or intravenous antibiotics.

Discussion

We report our results with a right colon continent cutaneous urinary diversion that utilizes a reliable appendiceal, flap valve mechanism and the terminal ileum as a ureteral substitution segment in both radiated and non-radiated patients. This design incorporates and expands on established principles. While we and others have reported early experiences with this specific reservoir^{3, 10} in either radiated or non-radiated patients, this is the first comparative evaluation of both populations using consecutive patients. Our analysis demonstrated that radiated and non-radiated patients undergoing UIA urinary diversion were similar in their rates of early and late pouch related complications. Our radiated population consisted of patients undergoing more extensive surgical procedures compared to the non-radiated group, with most radiated patients undergoing total pelvic exenteration. Overall surgical morbidity was higher in the total pelvic exenteration group as evidenced by the prolonged hospital stay, largely related to prolonged postoperative ileus and abdominal infectious complications.

Continent reservoirs have historically been associated with a 20% to 86% complication rate related to reservoir construction, particularly in the setting of prior pelvic irradiation.^{2, 4, 11–13} The most commonly encountered pouch-related complications include those involving the catheterizable limb or stoma and the ureterointestinal anastomosis. Radiation injury to the ureters and small bowel have contributed to the higher rate of postoperative complications seen in previously radiated patients.^{2, 4, 11–13} Radiation injury ultimately results in vascular damage, fibrosis, and poor tissue viability, which compromise the integrity of the ureteroenteric anastomoses. The reported rate of ureteroenteric complications in previously irradiated patients undergoing continent urinary diversion has ranged from 22% to 37%.^{4, 5, 11, 13, 14} The most commonly encountered ureteral complications are strictures and anastomotic leaks.

The use of ileal reservoirs in previously irradiated patients has been associated with urinary leakage rates of 20% to 37.5% and a re-operation rate of 7.5% to 25%.^{2, 4, 5}. Reported complication rates associated with colon pouches in irradiated patients have been as high as 86%,⁴ four-fold higher than the rate observed in non-radiated patients. In those undergoing a modified Indiana pouch reconstruction, a five-fold higher ureterointestinal complication rate was reported in irradiated compared to non-radiated patients.¹⁵ Wammack et al. compared operative outcomes in 36 irradiated patients and 385 non-radiated patients undergoing either a Mainz I pouch or a right colon reservoir with an intussuscepted nipple valve. Pouch-related complications approximated 86% for the irradiated patients versus 23% for the non-radiated group. The most common complications included leakage (with a rate of 25%), stomal stenosis (39%), and ureterointestinal anastomotic strictures (22%).⁴ We did not observe any postoperative stomal incontinence, and our stomal-stenosis rate compared favorably at 23%. Our overall rate of ureterointestinal complications was 7%; 4% in non-radiated and 10% in radiated patients.

The design of the UIA reservoir allows for the substitution of lower ureteral segments with the terminal ileum. Utilizing the terminal ileum segment as a ureteral substitution allows the anastomosis to be created with non-pelvic segments of the ureter. Using a well-vascularized, non-radiated segment of proximal ureter in a tension-free manner is of particular importance in patients who have received prior high doses of pelvic radiation. The lower rate of early and late ureterointestinal complications observed in this report highlights the benefit of using this design in a complex group of previously radiated patients.⁹ In the non-radiated setting, an advantage of this design allows for substituting distal ureteral segments that may require excision secondary to tumor involvement (ie. TCC extending into the lower ureteral segments) and the theoretical advantage of having an isoperistolic limb in place to minimize functional reflux (similar to that observed with the Studer type afferent limb). The increased segment of terminal ileum did not appear to be associated with an increased rate of intestinal complications such as diarrhea; however, long-term vitamin B12 levels were not routinely evaluated during the follow-up period for most patients in this study.

It appears that the use of a flap valve mechanism created by the subserosal tunneling of the appendix in the cecum is superior to the hydraulic valve design for maintaining stomal continence⁸. The complications reported with many of the ileocecal valve-based limbs are largely attributed to the efferent limb design;¹⁶ they include incontinence, stomal stenosis, difficulty with catheterization, and parastomal hernia.⁹ Although the continence rates using the appendix are excellent, the most commonly encountered complication was stomal-related⁸. Fifteen percent of our patients in the radiated group required surgical revision of their stoma during the early postoperative period, as compared to 5% in the non-radiated group. The reason for surgical revision was disruption of the stomal connection to the skin, which resulted from either early traumatic catheterization or from a perioperative wound infection involving the periumbilical region. Late stomal complications consisted exclusively of stomal site narrowing that was managed with office dilatation. The rates of stomal-related complications reported in the literature for the tunneled appendix versus ileal valve appear comparable.^{17, 18} However, the advantage of improved continence of the appendix mechanism is offset by the increased stomal stenosis noted using the appendix stoma .^{8, 17}

Conclusions

The ureteroileocecal appendicostomy is an effective means of providing continent cutaneous diversion of the urinary tract in both non-radiated and previously heavily irradiated patients. We found that the modifications of previous appendix-based stomal mechanisms provided excellent continence with minor pouch-related complications. Additionally, the use of a tubularized ileal segment to allow for the use of a more proximal ureter segment provided an improved performance of the ureterointestinal anastomoses with a lower stenosis rate than previously reported for irradiated patients. Ongoing improvements in surgical techniques for continent urinary diversion can lower the morbidity associated with cutaneous diversion and improve quality of life by minimizing postoperative complications.

Abbreviations

CI

confidence interval

NR

non-radiated

RT

radiation therapy

SD

standard deviation

UIA

ureteroileocecal appendicostomy

CCR

continent cutaneous reservoir