Technique

Young’s initial description (1919) of excising a portion of the bladder neck and significantly tightening the bladder neck over a silver probe was modified by Dees (1949), who extended the length of excised tissue through the trigone. Leadbetter (1964) followed by elevating the ureters off the trigone, placing them in a more cephalad position on the bladder floor. This allowed tubularization of the trigone and further enhanced lengthening of the urethra. A detailed description and illustrations are found in Chapter 124.

Results

Reports of success with the Young-Dees-Leadbetter bladder neck reconstruction in children with neurogenic sphincter dysfunction are limited, not only in the number of patients but also in overall improvement. Tanagho (1981) and Leadbetter (1985) independently reviewed their long-term results and showed minimal success in individuals with neurogenic dysfunction. They speculated that the lack of success was due to a lack of muscle tone and activity in the wrapped muscle related to the underlying neurogenic problem. Many patients in the early series did not undergo bladder augmentation, possibly compromising the continence achieved. Contrary to the results reported in exstrophy patients, the majority of individuals with neurogenic deficiency of the bladder neck will require bladder augmentation and intermittent catheterization. Sidi and colleagues (1987b) documented a 4-hour continence interval in 7 of 11 patients after such repair, although 10 required catheterization and 9 required augmentation. Five of the 7 needed reoperation to achieve continence. This small series represents one of the more recent long-term results of the Young-Dees-Leadbetter reconstruction in children with neurogenic dysfunction because it has largely fallen out of favor. In an attempt to enhance the Young-Dees-Leadbetter procedure, Mitchell and Rink (1983) described the addition of external support and compression achieved through the placement of a silicone sheath around the reconstructed bladder neck. This was somewhat done to establish a plane for future placement of an artificial sphincter cuff, if necessary. In place, it seemed to improve the function of the repair by either improving coaptation or by maintaining the proximal repair in a better anatomic position. Unfortunately, most of the thicker Silastic sheaths eventually eroded and disrupted the repairs (Kropp et al, 1993). Quimby and colleagues (1996) used a thinner Silastic sheath without the same risk of erosion. They also wrapped omentum between the repair and Silastic wrap. The authors reported better results for continence and thought that placement of an artificial sphincter cuff was much easier when needed.

Donnahoo and coworkers (1999) reviewed one of the largest series of the repair used in treatment of neurogenic incontinence (38 children, 25 of whom were girls). A primary repair was performed in 24 children, a secondary procedure in 6, and a primary repair in conjunction with a silicone sheath in 8. Partial continence was achieved after the initial repair in 26 children (68%). All children with the silicone sheath were initially continent, but erosion occurred in 5. More critically, 35 (92%) of the children required augmentation cystoplasty to become continent. The authors found that although continence could be achieved with this technique, it was at the expense of augmentation cystoplasty and multiple procedures. Their results are similar to those reported by Cole and associates (2003).

Technique

The bladder neck is exposed by clearing fatty tissue overlying the bladder neck and the lateral endopelvic fascia. An incision is made within the endopelvic fascia for approximately 2 cm. The junction between the bladder neck and proximal urethra can be identified by placing a transurethral catheter into the bladder and gently pulling down on the catheter to lodge the balloon at the bladder neck. By blunt dissection, a plane between the posterior bladder neck and vagina in girls or rectal wall in boys is developed. The proper plane may be more easily developed from the cul-de-sac by dissecting behind the bladder and ureters from above (Lottmann et al, 1999; Badiola et al, 2000). If the landmarks are not easily defined, as in a secondary repair, the dissection becomes difficult. It may be appropriate to open the bladder to help prevent inadvertent dissection into the urethra or posterior structures. Storm and associates (2008) reported on a robotic-assisted laparoscopic approach for posterior bladder neck dissection and placement of the sling. This may become more appealing as robotic technology expands into bladder reconstruction. Dik and colleagues (2003) proposed a transvaginal approach, eliminating the need to open the bladder or to dissect between the bladder neck and anterior vagina.

When fascial tissue is used the technique is based on that described by McGuire and Lytton (1978) for stress urinary incontinence. Rectus abdominis fascia 1 cm in width and an appropriate length is harvested. This fascia can be taken in either vertical or horizontal fashion, depending on the initial skin incision. Fascia from other sites has been used in a similar fashion but requires a second incision. Often poor nutritional states and prior abdominal procedures limit the availability of rectus fascia, resulting in the use of alternative material such as small intestinal submucosa, cadaveric tissue, biodegradable scaffolds, and bladder wall (Colvert et al, 2002; Misseri et al, 2005; Albouy et al, 2007). All are generally secured to the anterior rectus fascia on either side. Autologous fascial tissue has been used, combining the benefits of a compressive wrap and suspension of the proximal urethra and bladder neck. Several variations of fascial placement and configuration have been described (Woodside and Borden, 1982; McGuire et al, 1986; Elder, 1990; Perez et al, 1996a; Bugg and Joseph, 2003; Dik et al, 2003). When fascial slings and wraps are used for neurogenic sphincter incontinence there is not as much concern for making a wrap or sling that is too tight because most patients are preferentially managed with CIC.

Results

Fascial slings have been used more extensively and with better results in girls with neurogenic sphincter incompetence, although some success has recently been reported in boys. Overall long-term success with fascial slings in the neurogenic population has varied greatly from 40% to 100% (Kryger et al, 2000; Castellan et al, 2005; Misseri et al, 2005; Snodgrass et al, 2007). A variation thought to contribute to a higher success includes a circumferential fascial wrap around the bladder neck. A circumferential wrap may equalize the compressive pressure over a greater surface area of bladder neck and posterior urethra (Walker et al, 1995; Strang et al, 2006). With a 360-degree wrap, simultaneous suspension has also been applied (Bugg and Joseph, 2003). Success rates have varied so much that it is difficult to determine whether any modification of the sling accounts for an increase in continence. Most patients who have undergone a fascial sling or wrap have also had simultaneous bladder augmentation. Success of the sling, as with most repairs, appears to be improved with augmentation cystoplasty in this patient population by almost all reports (Castellan et al, 2005). Perez and associates (1996a) reviewed the outcome of sling cystourethropexy in 39 children, 15 of whom were boys. One of four techniques was performed. When postoperative continence was evaluated on the basis of age, sex, underlying diagnosis, preoperative urodynamics, surgical technique, and enterocystoplasty, only concomitant enterocystoplasty was predictive of a successful outcome.

Contrary to the Silastic sheath, fascial sling erosion rarely occurs. Gormley and coworkers (1994) reported a revision rate with fascial slings of 15%. Placement of a fascial sling does not eliminate the possibility of later placement of an artificial urinary sphincter (Decter, 1993; Barthold et al, 1999). It is not unreasonable to consider placement of a fascial sling in a child with a marginally competent bladder neck and posterior urethra if the child is undergoing augmentation cystoplasty and already requires intermittent catheterization.

Technique

Endoscopic exposure is used to localize the proximal urethra and bladder neck. Injection can be done directly through the working channel of the endoscope, often one with an offset lens system. Ideal placement of the material is in a subepithelial space, mobilizing the epithelium toward the lumen of the bladder neck. When it is completed in a circumferential fashion, adequate epithelial coaptation may occur that effectively raises outlet resistance. Alternatively, periurethral injection in women with a long needle placed from the perineum or through a suprapubic approach has been used. Dean and associates (2007) advocate simultaneous antegrade and retrograde endoscopic injection. Evidence is lacking whether the exact approach affects success, but accurate placement is important and transurethral injection generally preferred.

Results

The durability and success of bladder neck and proximal urethral injection remain in doubt for the pediatric population, particularly those with neurogenic dysfunction. True continence, as defined by a 4-hour dry period between voidings or catheterization, has been reported to be at most 64% and has ranged as low as 5% (Leonard et al, 1990a; Capozza et al, 1995; Bomalaski et al, 1996; Perez et al, 1996b; Sundaram et al, 1997; Silveri et al, 1998; Guys et al, 2001, 2006; Godbole et al, 2003; Halachmi et al, 2004; Lottmann et al, 2006; Dean et al, 2007). Several factors play a role in the outcome, one of which is a history of any previous operative bladder neck repair. Success is enhanced by elevation of the epithelium of the bladder neck, which may be compromised by scarring from previous operative procedures. The concept of a minimally invasive operation used to enhance a marginal result gained from a more formal bladder neck repair is enticing; unfortunately, the data are lacking to show that bladder neck injection is of lasting value in that setting.

Sundaram and colleagues (1997) reported on the efficacy and durability of glutaraldehyde cross-linked bovine collagen in 20 children, 12 of whom had neurogenic sphincter dysfunction. More than half of the children required two or three independent injections. Success was achieved in only 1 patient (5%), who was considered dry; 5 had some improvement, and 10 had either no change or transient improvement of only 2 to 90 days. Thus collagen therapy only delayed the ultimate need for bladder neck reconstruction.

Submucosal bladder neck injection of bovine dermal collagen was used by Perez and coworkers (1996b) in 32 patients. Continence was achieved after a single injection in only 20% of the children with neurogenic dysfunction. Complications were limited to febrile urinary infections, one episode of urinary retention, and worsening incontinence in 2 patients. The authors concluded that even though their success was limited the low morbidity and ease of placement justified submucosal injection in selected children.

Guys and associates (2006) treated 49 children with polydimethylsiloxane, 41 of whom had neurogenic bladder neck and sphincter incontinence. The level of continence continued to deteriorate through 18 months, then stabilized. At an average follow-up of 73 months, success was achieved in 16 (33%) and 7 (14%) were improved. Godbole and colleagues (2003), Halachmi and associates (2004), and Dyer and coworkers (2007) independently came to the conclusion that regardless of the material injected (polytetrafluoroethylene, collagen, polydimethylsiloxane, dextranomer/hyaluronic acid), short-term success is not long lasting.

Kitchens and colleagues (2007) evaluated the use of dextranomer/hyaluronic acid to correct incontinence in patients who had previously undergone bladder neck reconstruction. They achieved success or improvement in 70% of select patients at 17 months, indicating this may be an alternative to other salvage therapy.

Overall, the cost of the bulking agents can be excessive and there does not appear to be any financial benefit over a formal repair (Kryger et al, 2000). At present, bulking agents play a limited role for increasing outlet resistance and should be reserved for a select group of patients. The exact criteria that define that group have not been established. Patients with marginal native outflow resistance are probably better candidates than are those with minimal preoperative function.

Technique

Placement of the cuff should be at the level of the bladder neck in all female patients and prepubertal boys (Fig. 129–3). It is also the most desirable and effective location in pubertal boys and adult men with neurogenic sphincter incompetence. The bulbar urethra can be used as an alternative site in adult men with mature spongiosum. Levesque and associates (1996) have indicated that age is not a factor in placement of the cuff around the bladder neck. They found that children do not outgrow the artificial urinary sphincter as they progress through puberty and that replacement of the cuff is not routinely necessary. The artificial urinary sphincter cuff can be positioned around intestinal segments used in total urinary reconstruction but is more susceptible to erosion there. Several authors have described the successful placement of the cuff around a bowel segment, particularly when omentum is interposed between the cuff and the segment (Burbige et al, 1987; Weston et al, 1991; Light et al, 1995).

Figure 129–3 Artificial urinary sphincter placement in children. A, The cuff is placed around the bladder neck in prepubertal children. Dissection may be facilitated by palpation of a urethral catheter and balloon or by a posterior approach. If dissection is difficult, the bladder can be opened. B, The reservoir is placed beneath the rectus muscle. C, A tunnel is then made into the scrotum or labia for positioning of the pump on the same side as the reservoir. Tubing connections are made ventral to the rectus fascia, and the device is initially left deactivated.

Placement of an artificial sphincter in children is the same as that described for adults. Development of the proper plane for the cuff is virtually identical to that described for a fascial sling. The cuff should be sized snugly, but not tightly, around the bladder neck. Obviously, a sterile environment is critical in considering placement of the artificial urinary sphincter to avoid infection. For that reason, preoperative antibiotics are a necessity, and confirmation of sterile urine is required. With those precautions there is freedom to open the bladder in dissecting around the bladder neck. This will often facilitate the dissection and ensure proper placement. The AS800 model has a locking mechanism in the pump that permits the artificial urinary sphincter to be deactivated and activated without a second operative procedure. Experience has shown that leaving the unit deactivated with the cuff deflated after placement allows formation of a pseudocapsule around the cuff and decreases the risk of erosion (Furlow, 1981; Sidi et al, 1984). The noncycled artificial urinary sphincter occasionally may provide enough resistance for continence, eliminating the disadvantages of activation (Herndon et al, 2004a).

Results

There are substantial short- and long-term data regarding continence after placement of the artificial urinary sphincter, with the largest series presented by Herndon and associates (2003). In investigation of continence, it must be placed in context of the cost experienced by the patient defined by mechanical malfunctions resulting in secondary operative procedures and more catastrophic complications, such as device infection and erosion. Dramatic improvement regarding the need for secondary procedures has occurred with the technical refinements in the device. Ten- to 15-year long-term follow-up of the artificial urinary sphincter in children has been reported (Levesque et al, 1996; Kryger et al, 1999, 2001; Castera et al, 2001; Hafez et al, 2002; Herndon et al, 2003; Lopez Pereira et al, 2006; Ruiz et al, 2006). All groups report an impressive continence rate of 80% and a functioning sphincter in 95% of patients. These reports are consistent with older series in children reporting continence rates of 75% to 90% and a functioning sphincter in 85% to 97% (Nurse and Mundy, 1988; Gonzalez et al, 1989a; Bosco et al, 1991; O’Flynn and Thomas, 1991; Aprikian et al, 1992; Singh and Thomas, 1996; Simeoni et al, 1996; Catti et al, 2008). Herndon and colleagues (2003) presented the most comprehensive long-term data. They achieved overall continence in 86% of 142 patients with an average follow-up of 10 years. Age at implementation does not appear to affect continence (Kryger et al, 2001).

Whereas the artificial urinary sphincter is one of the few surgically created continence mechanisms that does not negatively affect spontaneous voiding, intermittent catheterization remains an important adjunct in approximately 75% of children with neurogenic sphincter incompetence and can be performed successfully through the cuff (Diokno and Sonda, 1981; Gonzalez et al, 1995; Levesque et al, 1996; Kryger, 1999, 2001; Castera et al, 2001; Hafez et al, 2002; Herndon et al, 2003). As boys approach puberty, spontaneous voiding may become progressively inadequate. It has been speculated that growth of the prostate causes an increase in native outlet resistance. Kaefer and associates (1997a) evaluated increases in cuff size to facilitate spontaneous voiding in boys. In their limited series, they did not find that up-sizing restored the ability to void spontaneously. Jumper and colleagues (1990) reported on prostatic development and sexual function in pubertal boys with spinal dysraphism who had been treated with the artificial urinary sphincter. They found that the artificial sphincter did not alter sexual development, prostatic growth, or morphologic features.

Herndon and associates (2003) reported device malfunction in 64% with the pre-AS800 model and 30% with the AS800. Sphincter erosion was similar for the pre-AS800 and AS800, occurring at 19% and 16%, respectively, in their experience. Fastidious attention to detail and sterile technique diminish the risk of infection but do not eliminate it. When infection occurs without erosion, the unit can be removed and later replaced (Nurse and Mundy, 1988). Infections are minimized by sterilization of the urine preoperatively, meticulous cleaning of the wound site, preoperative bowel preparation, perioperative parenteral antibiotics, and copious antibiotic wound irrigation. Newer cuff design and a 6-week delay in activation of the device help formation of a thickened pseudocapsule that substantially decreases bladder neck and proximal urethral erosion. Kryger and associates (1999) indicated that erosion can be virtually eliminated when the cuff is placed as the primary treatment for bladder neck incompetence. They and others (Aliabadi and Gonzalez, 1990; Gonzalez et al, 1995; Simeoni et al, 1996; Levesque et al, 1996; Castera et al, 2001; Hafez et al, 2002; Herndon et al, 2003) noted that the risk of erosion substantially increases after previous failed repairs but this has not been the experience of others (Ruiz et al, 2006). Identification of the correct plane between the bladder neck and vagina in female patients or rectum in male patients preserves the vascularity of the bladder neck and proximal urethra and may decrease the rate of erosion (Aliabadi and Gonzalez, 1990). Initial exposure through a posterior bladder approach as described by Lottmann and coworkers (1999) may be helpful. Shankar and colleagues (2001) suggested that there is an advantage to exposure of the bladder neck with a transperitoneal approach by decreasing the potential of bleeding from the prostatic venous plexus and improving visualization of the rectal wall.

Levesque and associates (1996) evaluated the long-term outcome of the artificial urinary sphincter on the basis of date of insertion and location of the placement. Before 1985 the artificial urinary sphincter had been inserted in 36 children. Between 1985 and 1990, an additional 18 children underwent placement. In the original group, 24 of the 36 sphincters were in place and 22 functional; 12 had required at least one revision. The mean survival of the device was 12.5 years. Success rates at 5 and 10 years were 75% and 72%, respectively. In the group implanted after 1985, 78% retained a functioning sphincter. The overall continence rate in both groups was 59%; sphincter survival probability at 10 years was approximately 70%. There was no difference found between failure rates in males and females with the exception that female patients who had previously undergone bladder neck surgery were more likely to suffer an erosion. The ability to void independently without the use of intermittent catheterization was retained in 36 children (67%). Those findings are supported by contemporary series (Levesque et al, 1996; Kryger, 1999, 2001; Castera et al, 2001; Hafez et al, 2002; Herndon et al, 2003; Lopez Pereira et al, 2006; Ruiz et al, 2006; Catti et al, 2008).

Upper urinary tract changes including hydronephrosis have been reported to occur in up to 15% of children after placement of the artificial urinary sphincter (Light and Pietro, 1986; Churchill et al, 1987; Gonzalez et al, 1995; Levesque et al, 1996; Kryger et al, 1999). In extreme cases, renal insufficiency has resulted. It is now recognized that occlusion of the bladder neck in children with neurogenic sphincter incompetence can result in the unmasking or development of detrusor hostility manifested by a decrease in bladder compliance or increase in detrusor hyperreflexia (Bauer et al, 1986). Careful preoperative urodynamic assessment helps identify only some of the children who are at risk (Kronner et al, 1998b). When hostile bladder characteristics are found preoperatively, anticholinergic medications can be beneficial for hyperreflexic contractions but augmentation cystoplasty is usually required for diminished compliance. Churchill and associates (1987) showed that favorable parameters can be maintained after placement of the artificial urinary sphincter; however, close observation is still recommended in any child undergoing bladder neck reconstruction to identify any early deterioration in bladder dynamics before upper tract changes.

Some children undergoing sphincter placement need bladder augmentation as well, and the timing of the two procedures may be questioned because of the concern for artificial urinary sphincter infection. Light and colleagues (1995) reported a 50% infection rate with simultaneous augmentation compared with 9.5% when the procedures were staged. On the contrary, a contemporary review by Miller and coworkers (1998) found that infection necessitating removal of the device occurred in only 2 of 29 such patients (7%). This low rate is similar to that noted by others (Strawbridge et al, 1989; Gonzalez et al, 1989b). Several reports have evaluated various factors and found that the intestinal segment selected for augmentation appeared to be the only parameter affecting results; gastric augmentation was the least offensive regarding infection (Ganesan et al, 1993; Miller et al, 1998; Holmes et al, 2001). Gonzalez and associates (2002) reported an alternative technique using a seromuscular colocystoplasty and simultaneous placement of the artificial urinary sphincter. They achieved continence in 89% without the need for additional procedures and no deterioration of the upper urinary tract.

The AS800 is the subject of most reviews when the artificial urinary sphincter is discussed, but alternative devices have been reported (Lima et al, 1996; de O Vilar et al, 2004). An alternative device is a one-piece adjustable cuff connected to an injection port for inflation. The injection port is placed subcutaneously and made available for percutaneous access. This allows for adjusting the pressure within the cuff in order to achieve continence. It is particularly convenient for individuals with limited ability to actively pump the AS800. The periurethral constrictor is wider and more fluid than the AS800. It can be deactivated and reactivated by tapping the subcutaneous port.

The ultimate benefits of the artificial urinary sphincter lie in its ability to achieve a high rate of continence while maintaining the potential for spontaneous voiding. For practical purposes, when intermittent catheterization is required along with augmentation cystoplasty, use of native tissue for continence eliminates the long-term concern for infection or erosion and the risk of mechanical failure.

Technique

A Foley catheter is placed intravesically and the bladder filled to capacity. The bladder is exposed through either a midline or low transverse abdominal incision. The bladder neck is then identified with application of gentle catheter traction. A 6 × 2-cm rectangular flap based on the bladder neck and urethra is then isolated. Stay sutures are placed, and the flap is mobilized in continuity with the proximal urethra. The detrusor musculature at the bladder neck is then divided, separating the bladder and urethra, or the muscle may be left intact at the 5- and 7-o’clock positions. The anterior vaginal wall is exposed in girls; the seminal vesicles are exposed in boys. The rectangular strip based on the urethra is tubularized posteriorly around the urethral catheter with a continuous absorbable suture. The distal portion of the tubularized strip should be approximated in an interrupted fashion to facilitate excision of excessive tissue without jeopardizing the suture line. A capacious submucosal tunnel through the trigone is then developed posteriorly for the proximal neourethra (Fig. 129–4). A wide tunnel is required to prevent kinking at the level of the bladder neck, which will impede catheterization. It is important to eliminate dead space at the entrance of the urethra into the bladder, and this can be accomplished by placement of lateral anchoring sutures in the region of the bladder neck. The detrusor tube must be pulled straight through the tunnel without curve or deviation to facilitate catheterization. Waters and colleagues (1997) and Kropp (1999) have not found it necessary to reimplant all ureters in a cephalic location; they now typically reimplant only refluxing ureters (Kropp, 1999). When the bladder is closed, the lateral wings in the region of the bladder neck are approximated and incorporate adventitia of the tubularized urethra. This enhances a watertight closure and is continued for 2 to 3 cm anteriorly, often up to the area of augmentation. The tubularized neourethra should be long enough to reach the true lumen of the bladder, where it is exposed to pressure as an effective flap valve.

Figure 129–4 Kropp anterior detrusor tube. A, An anterior flap of bladder (2 cm wide × 5 to 6 cm long) is developed in continuity with the urethra. B, The anterior flap is tubularized over a catheter. A tunnel beneath the mucosa of the trigone is made between the ureteral orifices for the length of the tube. C, The tubularized flap is brought through the tunnel. D, The detrusor tube is secured to the floor of the trigone with interrupted absorbable suture in a straight course.

Because of the difficulties with catheterization, modifications of the Kropp bladder neck procedure have been described. Belman and Kaplan (1989) suggested a simplified approach. They harvested a rectangular strip from the anterior bladder wall similar to that described by Kropp. The lateral and posterior muscles at the bladder wall are not incised, however, and the proximal urethra and bladder are not separated. The flap is tubularized over an 8-Fr catheter. The epithelium on the floor of the bladder is incised, contrary to the tunnel made by Kropp. The tube is placed within the trough with the proximal meatus secured on the floor of the bladder. The epithelial edges of the trough are then secured to the lateral aspect of the tube. As in the initial description, the suture line for tubularization of the urethra is posterior against trigonal muscle. Closure of the bladder begins with reapproximation of the lateral walls of the bladder to the tube until the bladder edges meet. The remaining portion of the bladder is covered by an augmentation. Regardless of any modification to the Kropp bladder neck procedure, catheterization potentially remains problematic and therefore the patient should be prepared for a catheterizable abdominal-vesicle channel.

Results

Snodgrass (1997) examined the results in 23 children, 22 of whom had neurogenic sphincter incompetence and noted continence in more than 90% of the children. The most common complication was difficult catheterization, particularly in boys. Less than half of the boys in Snodgrass’ series catheterize through the native urethra; the majority do so through an abdominal wall stoma. Postoperative vesicoureteral reflux was identified in 9 of 18 children, and Snodgrass speculated that this was due to lateral retraction of the ureters. The recommendation was made to leave the posterior bladder wall open and flat in receiving the augmentation to prevent this distortion. Kropp (1999) has not had as much problem with catheterizations in his patients and likewise has achieved a high rate of continence without a high incidence of new reflux.

Some patients with an effective flap valve mechanism constructed by urethral lengthening will virtually never leak per urethra. This potentially puts them at risk for upper tract deterioration or bladder rupture, particularly if they do not or cannot catheterize reliably. Snodgrass (1997) thought that his modification was beneficial in that it resulted in a shorter intravesical tunnel for the neourethra, allowing leakage per urethra with overfilling.

Technique

An anterior, full-thickness bladder wall flap 5 × 1 cm is mobilized to the bladder neck with 0.1 cm of its epithelial edges excised to avoid overlapping suture lines. Two parallel incisions down to the level of the muscle are made in the mucosa of the trigone from the native bladder neck. The anterior flap is secured to the midline strip of trigone mucosa to form a tube of lengthened urethra with absorbable suture. The muscle on either side of the posterior epithelial strip may be incised superficially to provide an edge to which the muscle of the anterior flap may be approximated with a second layer of suture in an effort to avoid fistula (Fig. 129–5). The more lateral mucosa of the trigone is mobilized and closed over the midline urethra. Distally, the muscle of the bladder neck is wrapped fairly tightly around the urethra with closure. More proximally, the lengthened urethra should extend well into the lumen of the bladder.

Figure 129–5 Pippi-Salle anterior detrusor tube onlay. A, An anterior detrusor tube (1 cm × 4 to 5 cm) is mobilized to the level of the bladder neck. B, A central strip on the floor of the bladder is developed by incision of the epithelium on either side. C, The anterior detrusor flap is secured to the strip in a two-layer fashion, approximating the epithelium with the first layer and muscle with the second. D, The lateral mucosa of the trigone is mobilized and secured over the lengthened urethra. Distally, the bladder is closed to itself and should incorporate a portion of the anterior detrusor flap to ensure a watertight seal. E, After distal closure the remaining portion of the bladder is left open if augmentation cystoplasty is necessary.

Results

Initial complications of this procedure included persistent incontinence, urethrovesical fistula, and partial necrosis of the intravesical neourethra. Widening the base of the urethra at the level of the bladder neck may decrease these problems. Children who have previously undergone bladder surgery can have a secondary Salle repair if the anterior bladder flap is lateralized slightly to avoid any old midline suture line and increased potential for scarring or necrosis.

Salle and coworkers (1997) found that continence was achieved in 12 of 17 patients (70%) for more than 4 hours. Catheterization difficulties occurred in only 3 of 17 children, 1 of whom subsequently underwent an appendicovesicostomy. Fistula formation at the base of the flap between the proximal, intravesical urethra and bladder occurred in 2 children and resulted in recurrent incontinence. This problem appears to be diminished by making a wider base to the flap and generously trimming the epithelial edges. Jawaheer and Rangecroft (1999) reported a diurnal continence rate of 61% for 3 hours or longer with the Salle procedure. However, only 44% of their patients were dry through the night. Less trouble with catheterization has occurred relative to the Kropp technique, and it rarely remains a problem. Continence rates have not been quite as high in most series (Rink et al, 1994; Mouriquand et al, 1995; Cole et al, 2003).

Canales and coworkers (2006) capitalized on both the Kropp and Pippi-Salle techniques. They evaluated a miniature intravesical urethral lengthening. The primary difference relates to a diminished tunnel (3 cm) and radius of the anterior detrusor tube (8 Fr) with the intent of enhancing resistance, maintaining a popoff mechanism, limiting the use of bladder tissue, and avoiding the need for ureteral reimplantation. At follow-up of 2.5 years, 8 of 9 patients were dry, had normal upper tracts, and had an average leak point pressure of 71 cm H₂O.

Technique

A segment of ileum at least 15 to 20 cm proximal to the ileocecal valve should be selected. The distal portion of terminal ileum is unique from a physiologic standpoint and should be avoided. The isolated segment should be 20 to 40 cm in length, depending on the patient’s size, native bladder capacity, and desired final capacity. With short ureters, an extra tail of isoperistaltic ileum can be useful to reach the foreshortened ureters. This requires construction of an ileal nipple valve to prevent reflux as in the Kock or hemi-Kock pouch. This type of construction may require up to 60 cm of small intestine. The segment to be used should have an adequate mesentery to reach the native bladder without tension.

After selection of the appropriate segment, the mesentery is cleared from the bowel at either end for a short distance to make a window. The bowel is divided at those ends, and a hand-sewn ileoileostomy or stapled anastomosis is performed. The mesenteric window at the bowel anastomosis is closed to prevent an internal hernia. The harvested ileal segment is irrigated clear with 0.25% neomycin solution and opened on its antimesenteric border (Fig. 129–8A). The ileum is folded in a U shape most commonly, although longer segments can be folded further into an S or W configuration. The ileum is anastomosed to itself with running absorbable sutures (Fig. 129–8B). The suture line should approximate the full thickness of ileum to ileum while inverting the mucosa. If the bladder was not opened previously, it is incised in a sagittal plane. The anastomosis of the ileum to native bladder is easily done when it is started posteriorly. The anastomosis may be done in a one- or two-layer fashion, always using absorbable suture and inverting the mucosa to the lumen (Fig. 129–8C). Permanent sutures should never be used for any cystoplasty because they serve as a nidus for stone formation.

Figure 129–8 A, Ileocystoplasty. A 20- to 40-cm segment of ileum at least 15 cm from the ileocecal valve is removed and opened on its antimesenteric border. Ileoileostomy reconstitutes the bowel. B, The opened ileal segment should be reconfigured. This can be done in a U, S, or W configuration. It can be further folded as a cup patch. C, The reconfigured ileal segment is anastomosed widely to the native bladder.

A suprapubic tube is brought out through the native bladder when possible and secured. The anterior aspect of the anastomosis is then completed. A drain is placed near the bladder and brought out of the pelvis through a separate stab incision. It should be removed promptly if it is not draining urine, particularly in neurogenic patients with a ventriculoperitoneal shunt. The wound is thoroughly irrigated and the abdomen is closed in layers.

Technique

The isolated ileocecal segment is irrigated clear with neomycin solution and opened on its antimesenteric border through the ileocecal valve for its entire length. In the typical ileocecal augmentation the ileal and cecal segments are of equivalent length such that the borders of the open segment can be anastomosed and then folded on themselves to form a cup cystoplasty (Fig. 129–9B). The anastomosis of the reconfigured segments is done in either a one- or two-layer closure with absorbable suture. The opening should be left large enough to provide a wide anastomosis to the bivalved bladder. If more volume is necessary, the ileal segment can be significantly longer, allowing it to be folded before anastomosis to the cecum. The Mainz ileocecocystoplasty uses an ileal segment twice the length of the cecal segment. The opened edge of the cecal portion is anastomosed to the first portion of the ileal segment. The first and second portions of the ileal segment are next approximated. The compound ileocecal patch is then anastomosed to the bladder (Fig. 129–10). The mesenteric window is closed, and a suprapubic tube is placed through the native bladder and secured through the abdominal wall.

Figure 129–10 The Mainz ileocecocystoplasty. A and B, The ileal segment is twice the length of the cecal segment. C and D, It is opened on the antimesenteric border. E and F, The ureters can be reimplanted into the opened cecal segment if necessary. G and H, The ileocecal segment is anastomosed to the native bladder.

(From Thuroff JW, et al. In: King LR, Stone AR, Webster GD, editors. Bladder reconstruction and continent urinary diversion. Chicago: Year Book; 1987.)

Technique

Fifteen to 20 cm of sigmoid colon is identified and mobilized. The mesentery is transilluminated to identify the vascular arcade to the segment. After identification of this blood supply the surgeon must ensure that the segment can reach the bladder without tension. If so, the bowel segment is divided between clamps and a colocolostomy performed (Fig. 129–11A). The remainder of the abdominal cavity is carefully packed to prevent contamination from the open sigmoid segment. Detubularization and reconfiguration are done in a fashion determined by the surgeon’s preference. The sigmoid patch is anastomosed to the bivalved bladder in a manner similar to that previously described for ileocystoplasty. Again, a large suprapubic tube is brought out through the native bladder and secured to the bladder and skin exit sites. Drains are placed as previously noted.

Figure 129–11 Sigmoid cystoplasty. A, A sigmoid segment of adequate length is removed from the gastrointestinal tract, and a colocolostomy is performed. B, In the Mitchell technique the two opened ends are closed. The antimesenteric border is incised, and the segment is anastomosed to the bivalved bladder. It may be rotated 180 degrees to allow an easy fit. C, The opened sigmoid segment can be reconfigured into a U or S configuration, which may lower pressure.

Technique with Use of Gastric Antrum

Leong and Ong (1972) described the use of the entire gastric antrum with a small rim of body for bladder replacement. With their technique the left gastroepiploic artery is always used as a vascular pedicle. If the right gastroepiploic artery is dominant and the left vessel ends high of the greater curvature, a strip of body along the greater curvature from the left gastroepiploic artery to the antrum is maintained and provides adequate blood supply (Leong, 1988). Continuity of the upper gastrointestinal tract is restored by a Billroth I gastroduodenostomy.

Technique with Use of Gastric Body

A gastric wedge based on the midportion of the greater curvature may be used (Adams et al, 1988) (Fig. 129–12A). The gastric segment used in this technique is made up mainly of body and consequently has a higher concentration of acid-producing cells. The right or left gastroepiploic artery may be used as a vascular pedicle to this segment. The right artery is commonly dominant and thus more frequently used. The wedge-shaped segment of stomach includes both the anterior and posterior wall. The segment used may be 10 to 20 cm along the greater curvature, depending on the patient’s age and size as well as the needed volume (Fig. 129–12B). The incision into the stomach is stopped just short of the lesser curvature to avoid injury to branches of the vagus nerve controlling the gastric outlet. Branches of the left gastric artery just cephalad to the apex of this incision are suture ligated in situ before incision to avoid significant bleeding. Parallel atraumatic bowel clamps are placed on either side of the gastric incisions to avoid excessive bleeding or spillage of gastric contents. Alternatively, the stomach may be incised by use of a gastrointestinal stapling device that places a double row of staples on each side of the incision (Mitchell et al, 1992). The staple lines, however, must eventually be excised. The native stomach is closed in two layers with permanent sutures on the outer seromuscular layer.

Figure 129–12 Gastrocystoplasty with use of the body. A, Gastric segment of body is mobilized on the right gastroepiploic artery. The left vessel may also be used; neither vessel as a pedicle should be free floating through the peritoneum. B, A longer gastric segment along the greater curvature with wider apex provides more surface area for augmentation. C, The gastric segment is anastomosed to the bivalved bladder with the mucosa inverted. (© Indiana University Medical Illustration Department.)

Branches of the gastroepiploic artery to the antrum on the right or to the high corpus on the left are divided to provide mobilization of the gastroepiploic pedicle. For the eventual pedicle to be long enough to reach the bladder, the appropriate segment may be higher on the greater curvature if the right vessel is used as a pedicle or lower if it is based on the left. The vascular pedicle, with omentum, should not be free floating through the abdomen. The segment and pedicle may be passed through windows in the transverse mesocolon and mesentery of the distal ileum and carefully secured to the posterior peritoneum. Despite careful consideration for an adequate pedicle length, on occasion the gastric segment initially may not reach the bladder without tension. Either gastroepiploic artery may be mobilized closer to its origin for further length. The first few branches from the gastroepiploic artery to the isolated gastric segment may also be divided. Because of the rich submucosal arterial plexus in the stomach, devascularization of the isolated segment does not result. Rarely, it may be necessary to approximate some of the isolated gastric segment to itself in one corner. The gastric segment should be approximated to the native bladder with one or two layers of absorbable sutures, with care taken to invert the mucosa (Fig. 129–12C).

Raz and colleagues (1993) and Lockhart and associates (1993) have described the use of a much longer, more narrow segment of stomach based along the greater curvature. Use of this segment, which includes both body and antrum, somewhat narrows the lumen of the stomach along its entire length except at the fundus and pylorus. Raz and colleagues have isolated this segment with a gastrointestinal stapler so that the native stomach is never open. The segment used in both of these series is similar to that first described by Sinaiko (1956), the first surgeon to use stomach in bladder replacement. Postoperative bladder and gastric drainage is no different from that described for intestinal cystoplasty. Histamine-2 (H₂) blockers are often given in the early postoperative period to promote healing (Rink et al, 2000).

Chloride Absorption and Acidosis

The first recognized metabolic complication related to storage of urine within intestinal segments was the occasional development of hyperchloremic metabolic acidosis after ureterosigmoidostomy (Ferris and Odel, 1950). Patients with this metabolic derangement were noted to have fatigue, weakness, anorexia, and polydipsia. Koch and McDougal (1985) demonstrated the mechanisms by which acid is absorbed from urine in contact with intestinal mucosa. Resorption in the form of ammonium results in chronic acid loading. Patients with normal renal function are generally able to handle the resorbed load of chloride and acid without frank acidosis. Mitchell and Piser (1987) noted that essentially every patient after augmentation with an intestinal segment had an increase in serum chloride concentration and a decrease in serum bicarbonate level, although full acidosis was rare if renal function was normal. Similar trends of increased serum chloride and decreased bicarbonate have been noted with ileal conduits (Malek et al, 1971) and continent urinary reservoirs (Allen et al, 1985; McDougal, 1986; Ashken, 1987; Thuroff et al, 1987; Boyd et al, 1989). More severe acidosis and electrolyte disturbances requiring treatment have been reported despite normal renal function (Schmidt et al, 1973; Whitmore and Gittes, 1983). Such derangements may be debilitating to the patient if they are not recognized and treated, and death of patients has been reported (Heidler et al, 1979). Hall and colleagues (1991) noted that there is an increase in the urinary acid load with wasting of bone buffers even in the absence of frank acidosis. Such wasting may result in bone demineralization and could potentially cause retarded growth in children after augmentation cystoplasty (Abes et al, 2003; Hafez et al, 2003; Vajda et al, 2003). Patients with acidosis should receive bicarbonate therapy. Whether all patients with intestine in the lower urinary tract might benefit from supplemental bicarbonate is controversial. Nurse and Mundy (1989) have suggested that arterial blood gas values may be more sensitive than serum bicarbonate or chloride levels for detecting acidosis. Stein and associates (1998) thought that measurements of arterial blood gas for base deficit allowed early treatment of acidosis and avoidance of bone demineralization. In severe cases of acidosis, chloride transport can be blocked with chlorpromazine and nicotinic acid.

Although jejunum is rarely used for bladder reconstruction, storage of urine in this segment results in a unique metabolic pattern of hyponatremic, hypochloremic, and hyperkalemic metabolic acidosis. The problem is often associated with significant hypovolemia.

Gastric mucosa is a barrier to chloride and acid resorption and, in fact, secretes hydrochloric acid (Piser et al, 1987). This difference was the primary factor in the initial consideration of stomach for use in the urinary tract. This secretory nature was shown to be of benefit in azotemic animals during acid loading (Piser et al, 1987; Kennedy et al, 1988). Serum chloride does decrease and serum bicarbonate increases slightly after gastrocystoplasty, whether antrum or body is used in patients with normal and impaired renal function (Adams et al, 1988; Ganesan et al, 1991; Kurzrock et al, 1998). In 21 patients with renal insufficiency, serum bicarbonate improved in all patients except one after gastrocystoplasty, and many patients requiring oral bicarbonate therapy before cystoplasty did not do so after gastrocystoplasty (Ganesan et al, 1991). A similar benefit has been noted in a smaller group of patients with renal failure (Sheldon and Gilbert, 1991).

Patient Growth

Delayed or slowed growth in some children after intestinal cystoplasty has previously been recognized (Wagstaff et al, 1991, 1992; Mundy and Nurse, 1992). A delay in linear growth was noted in 20% of almost 200 pediatric patients without any gross biochemical abnormalities. However, no control patients were included in the series, and body habitus and growth are difficult to measure and predict in children with myelodysplasia. Such patients make up the majority in most series of augmentation cystoplasty. Gros and colleagues (2000) evaluated growth in exstrophy patients. Patients requiring augmentation were matched retrospectively with a similar group not requiring bladder augmentation. Hydronephrosis was not mentioned in the patients but was unlikely with a diagnosis of exstrophy. Only 1 patient was noted to have acidosis in either group. Other factors that might affect growth, such as urinary tract infection, were not controlled. Of 17 patients with adequate measurements before and after augmentation cystoplasty, 14, or 82%, had a decline in percentile height postoperatively. The decline corresponded to a 1.5-inch decrease in expected height. The pattern of growth was significantly different between patients with and without augmentation in the series. That series is small, and no evaluation of familial growth patterns or ultimate height was possible; however, the findings are worrisome, particularly because Feng and coworkers (2002) noted similar differences in exstrophy patients. In the absence of any serum abnormalities the exact mechanism of delayed growth was not evident, although it seems likely to be related to chloride absorption and subclinical acidosis (Koch and McDougal, 1988; Bushinsky, 1989; Hochstetler et al, 1997). It should be noted that Taskinen and coworkers (2008) found no adverse effect on longitudinal growth after bladder augmentation in the exstrophy population. Three recent series did show the effect of bowel cystoplasty on bone mineral density in some patients (Abes et al, 2003; Hafez et al, 2003; Vajda et al, 2003), although Mingin and associates (2002) noted no such changes. One must be careful to determine whether any such changes are due to augmentation cystoplasty or the underlying pathologic process (Boylu et al, 2006; Taskinen et al, 2007). Better analysis of subtle metabolic alterations after enterocystoplasty may establish better understanding of the effect on growth, minimize changes, or aid in early treatment to avoid the complication (Brkovic et al, 2004).

Alkalosis

The secretory nature of gastric mucosa may at times be detrimental to the patient and can result in two unique complications of gastrocystoplasty. Severe episodes of hypokalemic, hypochloremic metabolic alkalosis that followed acute gastrointestinal illnesses have been noted in 5 of 37 patients after gastrocystoplasty (Hollensbe et al, 1992). The episodes were significant enough to require hospitalization in all cases and were recurrent in 2 patients. Three of the 5 patients suffering the complication had renal insufficiency and would not have been good candidates for augmentation with other segments because of acidosis. Ganesan and associates (1991) noted similar episodes of alkalosis in 5 of 21 patients with renal insufficiency after gastrocystoplasty. Those patients with the primary indication for consideration of gastrocystoplasty may be the ones at greatest risk for this unusual complication. It has been proposed that the alkalosis results from ongoing chloride loss from the gastric segment in the bladder in the presence of decreased oral intake. McDougal (1992a) suggested that the decreased ability to excrete bicarbonate from an impaired kidney may compound the problem. Gosalbez and associates (1993b) demonstrated persistently increased fractional excretion of chloride despite profound hypochloremia, suggesting that inappropriate gastric secretion is likely to be the primary mediator. One patient in their series eventually required resection of three fourths of the gastric segment in the bladder because of recurrent problems with alkalosis, and several required therapy with H₂ blockers or H⁺,K⁺ pump inhibitors.

All patients and families should be made aware of this potential problem because it has been reported to occur intermittently in between 3% and 24% of patients. A composite reservoir of stomach and ileum or colon may provide a more metabolically neutral reservoir (McLaughlin et al, 1995; Austin et al, 1997, 1999, 2001), although they have typically been constructed in only complex circumstances. Duel and associates (1996) have used stomach and colon together to advantage in a staged fashion for oncology patients; patients initially diverted with a colon conduit after cystectomy had a composite reservoir constructed later after cure from tumor was ensured.

Etiology

The etiology of delayed perforations within a bowel segment is unknown. It has been suggested that perforation might be secondary to traumatic catheterization in some cases (Elder et al, 1988; Rushton et al, 1988). Perforation of a bladder not previously augmented has been recognized after CIC (Reisman and Preminger, 1989). It seems unlikely that catheterization trauma is the lone cause in most patients. The location of the perforations has been variable among patients and even in a single patient with multiple perforations. Perforations have occurred after augmentation in patients who did not catheterize at all. Others have suggested that trauma to the bowel because of fixed adhesions resulting in shearing forces with emptying and filling may result in perforation (Elder et al, 1988). Chronic, transmural infection of the bladder wall has also been proposed as a cause. Histologic examination of bowel segments adjacent to areas of perforation has noted necrosis, vascular congestion, hemorrhage, and hemosiderin deposition compatible with chronic bowel wall ischemia (Crane et al, 1991). Chronic overdistention of the bladder might result in such ischemia. Decreased perfusion in dog bowel used for augmentation can be induced experimentally with high intravesical pressure (Essig et al, 1991). Those changes were noted more prominently at the antimesenteric border of the bowel. Chronic ischemia may thus play a significant role in at least some delayed bladder perforations. Anderson and Rickwood (1991) have reported perforations occurring in bladders with significant uninhibited contractions after augmentation, as have others (Pope et al, 1998). High outflow resistance may maintain bladder pressure rather than allow urine leakage and venting of the pressure, potentially increasing ischemia (Martinez del Castillo et al, 2005). Hyperreflexia alone is not likely to be a solitary cause of perforation; the complication was essentially never recognized in the era before bowel detubularization and reconfiguration when persistent pressure contractions were more common after augmentation cystoplasty. Once bowel is reconfigured, however, it may be more susceptible to ischemia if high pressure does persist. Failure to perform CIC on a regular basis when needed to empty may exacerbate such issues (DeFoor et al, 2003a; Martinez del Castillo et al, 2005).

The majority of patients suffering perforations after augmentation cystoplasty have had myelodysplasia. The incidence of perforation has been lower in series of patients with other diagnoses requiring bladder reconstruction (Hendren and Hendren, 1990). The role of neurogenic dysfunction in the etiology of these perforations is unclear. No matter what the etiology there is likely to be some field effect on the entire segment. Once a patient has suffered a spontaneous perforation, the chance of recurrence is significant (Hollensbe et al, 1992; Martinez del Castillo et al, 2005; Metcalfe et al, 2006). One fourth of patients with ruptures in one very large series had a recurrence (Metcalfe et al, 2006). Consideration must eventually be given to removal of the original segment and replacement by another after repeated perforation.

Incidence

Early postoperative leaks from the bowel-to-bowel or bowel-to-bladder anastomoses after augmentation cystoplasty are rare and represent a technical error or problem with early healing. Delayed perforations more commonly occur within the bowel segment itself and represent a problem with long-term storage of urine within an intestinal segment. There may be no particular increased risk of one intestinal segment over another. At Indiana University, perforations were noted in 43 of 500 patients (8.6%) undergoing cystoplasty (Metcalfe et al, 2006) an average of 4.3 years after augmentation. Analysis of this experience suggested that the use of sigmoid colon was the only significantly increased risk. Several other large series of patients with sigmoid cystoplasty have noted a low incidence of delayed perforation (Sidi et al, 1987a; Hendren and Hendren, 1990; Shekarriz et al, 2000). At Children’s Hospital Boston, the incidence of perforation after augmentation was highest in ileum (9.3%) and less frequent in ileocecal, sigmoid, and gastric segments (Bauer et al, 1992).

With inconsistent differences across multiple large series it is unlikely that any given enteric segment is at significantly increased risk for perforation and that likely multiple factors influence the risk for the complication. Pope, Rink, and the Indiana group (1999) still believe strongly from their experience that sigmoid colon has a greater risk for perforation based on a four-time higher rate than ileum there. Some other experienced surgeons think the same (Ransley, 1998). The highest rate of perforation noted has been 16.6% in one relatively small series (Martinez del Castillo et al, 2005); however, it is reasonable to present an expected risk of about 9% to patients and families based on experience at Indiana University (Metcalfe et al, 2006).

Treatment

The standard treatment of spontaneous perforation of the augmented bladder is surgical repair, as it is for intraperitoneal rupture of the bladder after trauma. There are reported series of conservative management for suspected perforation (Slaton and Kropp, 1994). Conservative management including catheter drainage, antibiotics, and serial abdominal examinations was successful in 87% of the patients, although only 2 of the 13 patients with suspected ruptures had x-ray documentation unequivocally identifying a perforation. Even patients who do well with conservative management during the acute episode often require eventual surgical intervention (Pope et al, 1999). Such management may be a consideration in a stable patient with sterile urine. The surgeon should certainly have a low threshold for surgical exploration and repair. The majority of patients with perforations have myelodysplasia and present late in the course of the disease because of impaired sensation. Increasing sepsis and death of the patient may result from a delay in diagnosis or treatment.

Technique

Ureterocystoplasty may be performed through a midline, intraperitoneal incision. This incision provides access to the intestine should mobilization of the ureter for augmentation be unsatisfactory. Bellinger (1993), Dewan and colleagues (1994a), and Reinberg and associates (1995) have shown that ureterocystoplasty can be done through two incisions, remaining completely extraperitoneal. The general technique is the same. A standard nephrectomy is performed with great care to preserve the renal pelvic and upper ureteral blood supply. All adventitia and periureteral tissue is swept from the peritoneum toward the ureter during mobilization to protect the ureteral blood supply. Proximally, this blood supply typically arises medially. As the ureter enters the true pelvis the blood supply arises posterior and laterally. After mobilization of the ureter into the pelvis the bladder is opened in the sagittal plane. Posteriorly, this incision has typically been carried off-center directly into and through the ureteral orifice of the ureter used for cystoplasty. The ureter is not detached from the bladder but is opened longitudinally along its entire length, with care taken to avoid its main blood supply (Fig. 129–15). The incision in the bladder and distal ureter should avoid branches of the superior vesical artery that serves as an important blood supply to the mobilized ureter. The ureter is folded on itself, and the ureter-to-ureter and ureter-to-bladder anastomosis is performed with running absorbable suture. A suprapubic tube is left indwelling through the native bladder for 3 weeks during healing. After cystography documents the absence of leakage, CIC is started. Patients, particularly those without neurogenic dysfunction, may attempt to void spontaneously, although all must prove that they can empty appropriately by checking postvoid residual urine amounts.

Figure 129–15 Ureterocystoplasty. A, After nephrectomy, the bladder is bivalved, with the posterior aspect of the incision carried off the midline to enter the ureteral orifice. The ureter is not detached from the bladder. B, The ureter is opened opposite its main blood supply. C, The ureter is reconfigured as in intestinal cystoplasty before anastomosis to the bladder.

Alternatively, the bladder incision can be stopped approximately 2 cm from the orifice and a similar length of distal ureter left in situ and intact without incision. The resulting small loop of intact ureter does not cause clinical problems or adversely affect the end volume in a significant manner (Adams et al, 1998). This modification of technique is easier and may be safer in that it avoids potential injury to the blood supply of the mobilized ureter near the ureterovesical junction.

Results

Early reports suggested that ureterocystoplasty could be achieved after nephrectomy by bivalving the bladder through the vesicoureteral junction and laying in the open ureter (Mitchell et al, 1992; Bellinger, 1993). Later reports modified and improved the procedure by noting that with meticulous care the vascularity to the entire renal pelvis could be preserved, allowing more tissue for cystoplasty (Churchill et al, 1993; Landau et al, 1994; McKenna and Bauer, 1995; Reinberg et al, 1995). As with intestinal cystoplasty, folding the ureter into a more spherical configuration maximizes the volume that can be achieved. If a massively dilated ureter drains a functioning kidney the distal ureter alone may be used for augmentation; the proximal ureter is either reimplanted into the bladder or anastomosed to the contralateral ureter (Bellinger, 1993).

Numerous series have reported good results after augmentation with use of ureter, some with follow-up as long as 8 years. The upper tracts have remained stable or improved in virtually all patients. Complications have been uncommon, with only a rare early extravasation of urine reported (Churchill et al, 1993). Landau and colleagues (1994) compared age-matched and diagnosis-matched children having ureterocystoplasty or ileocystoplasty. The total mean bladder capacity was 470 mL in the ureterocystoplasty group and 381 mL in the ileocystoplasty group. Bladder volumes at 30 cm H₂O were 413 mL and 380 mL after ureterocystoplasty and ileocystoplasty, respectively. Ureter effectively enhanced both volume and compliance. There was one failure in the group that occurred in a child in whom the distal ureter did not provide enough volume. Work has shown that one dilated ureter typically is enough for cystoplasty (Zubieta et al, 1999).

The main disadvantage to ureterocystoplasty is the limited population of patients with a nonfunctioning kidney draining into a megaureter. McKenna and Bauer (1995) have reported the use of a normal-sized ureter. The ultimate success of ureterocystoplasty with normal ureter requires further follow-up, particularly since Gonzalez (1999) has stated that one fourth of his patients with posterior urethral valves have failed ureterocystoplasty with a dilated ureter because of their huge urinary volume. Husmann and colleagues (2004) noted even worse results if the ureter used was less than 1.5 cm in diameter. Atala and colleagues (1994) presented an experimental technique to slowly dilate a normal ureter for later use. Work continues to develop such a technique that is clinically applicable for patients (Stifelman et al, 1998; Desai et al, 2003).

Techniques and Results

Cartwright and Snow (1989a, 1989b) described an ingenious method to improve bladder compliance and capacity with use of native urothelial tissue. In their procedure, known as autoaugmentation, they excised the detrusor muscle over the dome of the bladder, leaving the mucosa intact to protrude as a wide-mouth diverticulum. Initially, they made a midline incision through the bladder muscle (Fig. 129–16A). With the bladder distended with saline so that mucosa bulged from the incision, the muscle was then mobilized and excised laterally in each direction (Fig. 129–16B). The lateral edges of the detrusor muscle were then secured to the psoas muscle bilaterally to prevent collapse of the diverticulum (Fig. 129–16C). The early experience of these investigators noted improved compliance in most, and increased capacity in some, patients (Cartwright and Snow, 1989a).

Figure 129–16 Autoaugmentation. A, The detrusor is incised. B, Detrusor is stripped and excised from mucosa. C, The detrusor muscle is anchored bilaterally so that the mucosa bulges with bladder filling.

(From Cartwright PC, Snow BW. Bladder autoaugmentation: early clinical experience. J Urol 1989;142:505.)

This procedure has since been modified by a number of surgeons, each providing a different name for the procedure, depending on whether the detrusor muscle is simply incised or excised to make the diverticulum. In an effort to determine if incision or excision provided superior results, Johnson and colleagues (1994) performed 16 vesicomyotomies and 16 vesicomyectomies in rabbits after previously reducing the bladder capacity. Functional bladder capacity in the animals increased by 43.5%, and there was no statistical difference between the two techniques. They then performed vesicomyotomies (incision) in 12 patients with neurogenic bladder dysfunction and demonstrated a mean increase in capacity of 40% and a mean decrease in leak point pressure of 33% (Stothers et al, 1994). They concluded that detrusor excision offered no advantage over incision. All patients demonstrated some increase in capacity (15% to 70%), and no patient in early follow-up clinically deteriorated and required enterocystoplasty.

Detrusorectomy, leaving a small cap of muscle at the dome through which a suprapubic tube can be placed, has been proposed by Landa and Moorehead (1994). They have been concerned that although these procedures usually improve compliance, increase in volume is “modest at best,” a concern shared by others (Snow and Cartwright, 1996; Cartwright and Snow, 1998). In a report of 12 detrusorectomies, five patients were considered to have excellent results, two had acceptable results, and one was lost to follow-up. There were four failures (one hydronephrosis, two persistent incontinence, one worsening renal insufficiency); three patients have undergone gastrocystoplasty or ileocystoplasty (Landa and Moorehead, 1994). Overall, only 52% of patients had a good result with autoaugmentation; 20% had a poor outcome (Snow and Cartwright, 1996). Reoperative enterocystoplasty was not hampered by the prior detrusorectomy. The urothelial diverticulum at the time of augmentation cystoplasty was noted to be thick and fibrous, similar to a leather bag.

There are obvious advantages to autoaugmentation and its variants. Native urothelial tissue is used. It is an extraperitoneal procedure, which shortens operative time and avoids the risks of intestinal surgery and adhesions. Autoaugmentation is compatible with CIC and does not seem to complicate subsequent intestinal cystoplasty when it is necessary.

Complications from the procedures are generally uncommon. Perforation, a major concern after intestinal cystoplasty, has not been reported. Inadvertent opening of the mucosa during the procedure can make subsequent mobilization more difficult and may promote prolonged postoperative extravasation. Such extravasation usually stops with bladder drainage (Landa and Moorehead, 1994; Stothers et al, 1994). Prolonged drainage, however, may result in compromised results from collapse of the diverticulum. If concomitant ureteral reimplantation or bladder neck surgery is necessary, various authors have recommended that it be done first and the bladder then closed before detrusorectomy (Stothers et al, 1994).

Ehrlich and Gershman (1993) first reported laparoscopic myotomy (incision). A laparoscopic approach uses a smaller incision and perhaps shortens postoperative hospitalization; it may make effective fixation of the detrusor muscle in an open fashion to allow good bulging more difficult.

Concerns

The main disadvantage of autoaugmentation is a limited increase in bladder capacity such that adequate preoperative volume may be the most important predictor of success (Landa and Moorehead, 1994). If the maximum capacity and the volume of urine held at 40 cm H₂O are similar the patient may be better served by immediate intestinal cystoplasty. It is of note that some patients have demonstrated clinical improvement after these procedures without a significant change in urodynamics. The reason for the improvement in that setting is unknown.

In most series of autoaugmentation, no matter what the technique, it has been noted in occasional patients or concern has existed at the time of the initial procedure that adequate expansion was not achieved. In most such cases it was elected to proceed with enterocystoplasty immediately at the time (Landa and Moorehead, 1994). These patients are not included in the failure rate of autoaugmentation. The patient and surgeon must be prepared for such an event on occasion. Stohrer’s group (1999) as well as Leng and associates (1999) reported good results with the technique among patients with hyperreflexia. In terms of capacity, even if adequate expansion is achieved initially, there is concern that any improvement may not persist long term (Dewan et al, 1994b). In animals the surface area of the autoaugmentation site was noted to decrease approximately 50% at 12 weeks. Progressive thickening and contracture of the site have been noted because of collagenous infiltrate (Johnson et al, 1994). Milam (2000) has noted that almost half of his adult patients with hyperreflexia who had a good early result after autoaugmentation have failed with longer follow-up. Similar delayed failures have been noted in pediatric patients (Lindley et al, 2003), including one series with very poor results (MacNeily et al, 2003).

At this point, autoaugmentation should probably be considered only in patients who have reasonable capacity but poor compliance due to uninhibited contractions. If a significant increase in capacity is needed, autoaugmentation is unlikely to be as definitive as other techniques.

Techniques and Results

To avoid contracture, a combination of autoaugmentation after detrusorectomy and coverage with a demucosalized enteric segment has now been used. This has been done to potentially preserve the advantages of both procedures. In a similar fashion the combination has been undertaken with both colon and stomach. Buson and colleagues (1994) used reconfigured, demucosalized sigmoid colon placed over the urothelium (seromuscular colocystoplasty lined with urothelium). They and others noted that the intestinal submucosa should be preserved to avoid contracture (Buson et al, 1994; Vates et al, 1997) (Fig. 129–17). This procedure has been performed clinically with early reports of good results in most patients (Gonzalez et al, 1994). Postoperative bladder capacity increased an average of 2.4-fold (139 to 335 mL) in 14 of 16 patients, and end filling pressure decreased from an average of 51.6 to 27.7 cm H₂O. Two patients experienced failure of the procedure and required ileocystoplasty; their urodynamic data were excluded. Two other patients developed an hourglass deformity (Gonzalez et al, 1994). Endoscopic biopsy of the segments has been interesting. Of 10 biopsies performed in the series, in 1 urothelium with islands of colonic mucosa was noted and in 2 others only colon mucosa was found. Removal of all of the enteric mucosa is important when sigmoid is used to prevent mucoceles or overgrowth of intestinal mucosa (Gonzalez et al, 1994; Lutz and Frey, 1994). Dewan and associates (1997) thought that preservation of the submucosa eventually promoted regrowth of bowel mucosa. The interaction of the two different tissues will be interesting to follow. The long-term effects on the urothelium by the seromuscular segment and vice versa are unknown. Work has shown that persistent transitional lining will protect from metabolic problems and mucus production (Denes et al, 1997).

Figure 129–17 Seromuscular enterocystoplasty with use of sigmoid colon. Detrusor incision is performed as in autoaugmentation; however, the bulging mucosa is covered with a demucosalized segment of sigmoid colon.

(From Buson H, Manivel JC, Dayanc M, et al. Seromuscular colocystoplasty lined with urothelium: experimental study. Urology 1994;44:745.)

Dewan and Byard (1993) and Close and associates (2004) alternatively used demucosalized stomach to cover an autoaugmentation, first in sheep and then in patients. Early results showed improved bladder function both clinically and by urodynamics (Horowitz and Mitchell, 1993; Dewan et al, 1994c; Horowitz et al, 1994; Robinson et al, 1994), although long-term results are not as encouraging (Carr et al, 1999). These procedures are technically more demanding than simple augmentation or autoaugmentation and are associated with more blood loss and a longer operative time (Gonzalez et al, 1994; Horowitz et al, 1994). Increased bleeding is particularly true when stomach is used. These urothelium-lined, seromuscular augmentations are theoretically attractive. Thus far, the failure and reoperation rate after such procedures remains higher than that noted for standard enterocystoplasty (Vates et al, 1997; Carr et al, 1999; Shekarriz et al, 2000). The results have been reported with colon (Shekarriz et al, 2000). Those results may be partially attributed to the learning curve with a new, complex procedure. Longer follow-up and more experience are necessary to determine whether the complication rate will decrease with experience or increase because of problems with the combination.

Technique

Appropriate preoperative planning is required to position the skin incision to allow adequate mobilization of the appendix to the bladder. In most children this can be achieved through a low midline or a transverse incision. On occasion, the cecum may be high in the abdomen. Mobilization of the ascending colon along the line of Toldt may be required to gain mobilization of the appendix and its mesentery. Cadeddu and Docimo (1999) have used laparoscopy to aid in mobilization of the ascending colon and cecum. Once the cecum has been mobilized the base of the appendix is amputated, leaving a small cuff of cecum with the appendix. Use of the cuff at the stoma may decrease the risk of stenosis. The cecum is closed in a fashion similar to an open appendectomy. If the length of the appendix is marginal, then a greater portion of the cecum can be harvested, which effectively increases the functional length of the appendix (Cromie et al, 1991). A portion may be tubularized with the appendix (Bruce and McRoberts, 1998).

After harvesting, a location is selected for implantation of the appendix into the bladder. The location is based on the length of the appendix, the mobility of the bladder, and the location for the appendiceal stoma. Typically, the distal end of the appendix is tunneled into a posterolateral position within the bladder (see Fig. 129–19).

The base of the appendix is brought to the abdominal wall in a location chosen preoperatively to suit the patient. If a patient wears a supportive brace, stomal positioning must take this into account. The appendix should be brought up to reach the skin without tension, and care must be taken not to twist the pedicle or to occlude it as it passes through the abdominal wall fascia. The base of the appendix often can be hidden within the umbilicus, which allows elimination of a small but obvious abdominal stoma. Because of the small circular diameter of the appendiceal base, stomal stenosis is common; and techniques have been described to prevent this problem. Various flaps have been described as a method for avoiding stomal stenosis (Keating et al, 1993; Kajbafzadeh et al, 1995; Kaefer and Retik, 1997; Frane-Guimond et al, 2006; Berrettini et al, 2008; Landau et al, 2008). The VQZ stoma-plasty and its variations have been described in an attempt to prevent exposing intestinal mucosa from the catheterizable channel and decreasing the risk of stomal stenosis (Frane-Guimond et al, 2006; Berrettini et al, 2008; Landau et al, 2008). Regardless of the technique, stomal stenosis will remain a primary complication. Chronic nighttime catheterization with a stent placed into the channel but not reaching the bladder may further decrease the risk of superficial stenosis (Mickelson et al, 2009). For prevention of kinking and problems with catheterization it is advisable to maintain as short a conduit as possible. Securing the appendix and bladder wall to the peritoneum beneath the fascia will help diminish the problem of conduit kinking with reservoir filling. When appendix or any catheterizable stoma is used, it is advisable to repeatedly catheterize the channel after each step in reconstruction to confirm easy passage. If the catheter does not pass easily into the reservoir, the preceding step needs to be revised. Problems with catheterization by the surgeon during the operative procedure usually result in more difficulty for the patient afterward. It is also beneficial to catheterize at variable reservoir volumes to confirm proper fixation of the reservoir and the absence of any kinking. A 6- to 10-Fr Silastic stent is typically left for 10 to 14 days within the efferent catheterizable channel during the healing process. It is advisable for the surgeon to personally catheterize the efferent limb before allowing the patient or other family members to do so. Catheterization should be repeated at least every 4 hours during the day for reservoir drainage, maintenance of patency, and minimalization of the risk of stomal stenosis.

The appendix may not be available for use in all patients because of previous appendectomy, its location or length, congenital absence, involvement with adhesions, or its use for continence enemas. Histologic abnormalities of the appendix have been reported to occur in as many as 30% of patients (Liebovitch et al, 1992) and to increase with age. They rarely are of enough clinical significance to preclude use (Mulvihill et al, 1983).

Results

Several papers have reviewed large series of patients after appendicovesicostomy (Kaefer et al, 1997b; Cain et al, 1999, Harris et al, 2000; Thomas et al, 2006; Welk et al, 2008). The populations of patients have been typical of most pediatric groups, the majority having neurogenic dysfunction. Inability to use the appendix, other than because it was needed in situ for antegrade continence enemas, has been rare. The results, in terms of continence, have been superb, usually above 95% (Gerharz et al, 1997, 1998; Kaefer and Retik, 1997d; Mor et al, 1997; Suzer et al, 1997; Gosalbez et al, 1998; Cain et al, 1999; Castellan et al, 1999; Liard et al, 2001; Thomas et al, 2006; Welk et al, 2008). Incontinence is a rare event with the Mitrofanoff principle and may result from inadequate length of the flap valve mechanism or persistently elevated reservoir pressure. Urodynamic evaluation is required to evaluate the cause of the incontinence. Injection of collagen or other biomaterials is a possible treatment for inadequate outflow resistance, with success reported up to 50% but long-term durability unknown (Welk et al, 2008). A more formal approach with takedown and revision of the leaking Mitrofanoff valve is usually required (Kaefer and Retik, 1997d). The most common complication has been stomal stenosis, which has generally occurred in 6% to 10% of patients (Thomas et al, 2006; Welk et al, 2008). Such stenosis resulting in difficult catheterization may occur early in the postoperative course and require formal revision (Harris et al, 2000). Another recognized complication has been appendiceal perforation. Most problems with stomal stenosis and creation of a false passage (perforation) occur in the first few years after reconstruction (Thomas et al, 2006). Stricture and necrosis, particularly of cecal extensions of the appendix, have occurred rarely. Abdominal stomas may be associated with a higher risk of reservoir calculi because of the potential for incomplete emptying.

Follow-up in the recent series has averaged approximately 4 years. Patients from Mitrofanoff’s early experience (1980) should now have follow-up of more than 20 years, suggesting that the appendix may provide a durable alternative for pediatric patients with a long life expectancy, a finding also noted by others (Cain et al, 1999; Harris et al, 2000; Liard et al, 2001). Whereas the majority of the experience with the Mitrofanoff technique has been in children, similar outcomes with successful durable continence has been achieved in the adult population (Gowda et al, 2008; Van der Aa et al, 2009).

Technique

Parallel incisions 3 cm apart are made into the anterior bladder and used to form a long rectangular flap. The abdominal wall should be measured to ensure that the strip is long enough to reach the skin without tension. The full-thickness strip is tubularized down to the bladder, typically in two layers. The muscle portion is left broad to come around without tension and provide good blood supply. The mucosa may be trimmed in width before tubularization to avoid redundancy. A strip of mucosa within the bladder, 2 to 3 cm in length and 1.5 cm in width, is incised in a direct line and in continuity with the mobilized bladder tube. The edges of this strip are mobilized until it can be tubularized along its entire length. It may be beneficial to mobilize only one edge over to the other side, which allows one to avoid overlapping suture lines. Casale (1991) originally incised the mucosa transversely at the end of the intravesical strip to be tubularized; Rink and associates (1995b) then suggested that it could be left intact (Fig. 129–21). The bladder mucosa from either side of the tube is then mobilized and closed over the mucosal tube to create a flap valve. More extensive mobilization of the side opposite that mobilized for the inner tube allows closure without overlapping suture lines, which may help avoid fistula formation and incontinence. A soft stent is usually left through the tube for 3 weeks during healing to prevent stenosis. It does tend to close if it is not catheterized regularly and may be even more susceptible than other catheterizable channels to stomal stenosis (Cain et al, 2002; Thomas et al, 2005).

Figure 129–21 Casale continent catheterizable stoma. A, Parallel incisions are made in the bladder dome, forming a full-thickness bladder strip. B, The epithelium of the bladder is then incised for an additional 2.5 cm. The edges of the epithelium are mobilized, allowing tubularization. C, The epithelium is tubularized from the bladder out to the tip of the strip with an absorbable suture. The muscle of the bladder strip is then tubularized with an absorbable suture. D, The lateral edges of the epithelium within the bladder are reapproximated over the tubularized bladder segment with absorbable suture. The bladder is then closed with absorbable suture, incorporating the bladder tube with the initial sutures to prevent kinking.

Results

Continence rates have been good, as with most flap valve mechanisms (Cain et al, 1999, 2002). Stomal stenosis remains a significant problem—45% in the experience at Indiana University (Cain et al, 2002). Skin flaps and avoidance of tension to reach the skin may minimize this risk but not remove it. Advantages include avoidance of an intraperitoneal procedure and bowel anastomosis; the appendix can be reserved for use with enemas. It does use some bladder and decreases capacity, which may not be appropriate for some patients.