The primary reconstruction of bladder exstrophy in infancy is an accepted procedure. The staged approaches of functional closure and single-stage repair focus on creating a secure abdominal wall closure, preserving renal function and urinary continence, and reconstructing the penis in boys . Since 1983, we have used a single-stage repair in primary cases, including symphyseal re-approximation with no osteotomy, functional bladder neck plasty, an antireflux procedure and epispadias repair . However, the management of failed bladder exstrophy closure in adolescence and adulthood remains controversial. Patients may present with severe bladder abnormalities, pathological upper urinary tract, genital prolapse in women, and a rigid, wide pubic diastasis. After initial functional bladder neck reconstruction, persistent inadequate outlet resistance, small bladder capacity or decreased bladder compliance are some of the reasons for the failure of reconstructed bladder exstrophy [3,4]. One of the most important reasons for failure is a wide pubic diastasis . In previous reports, failed bladder exstrophy closure in children was corrected in a radical single-stage reconstruction with osteotomy, bladder neck reconstruction and augmentation, combined with a Mitrofanoff channel [3–5]. There are fewer reports on attempted functional reconstruction of failed bladder exstrophy in adolescence and adulthood . Because they have a prolonged history of incontinence, severe bladder abnormalities or upper urinary tract problems, urinary diversion is mostly carried out in these patients, with no attempt at bladder reconstruction. We report the management of four cases of primary failed bladder exstrophy closure in adolescent and adult patients.
Between 1995 and 1997, one adolescent (17 years old) and three adults (19–35 years old) were referred with previous failed bladder exstrophy closure. Three male patients had undergone multiple attempts at bladder reconstruction. The bladder was finally closed but the cosmetic and functional outcome was unsatisfactory. All remained incontinent, and the epispadiac penis was open or the tubularized urethra had multiple fistulae and strictures, necessitating radical surgical revision. All patients had undergone posterior iliac osteotomy. A 22-year-old woman with a previous cystectomy and ureterosigmoidostomy had an excessive residual pubic diastasis with uterine prolapse, uncorrected genitalia and upper urinary tract dilatation.
All patients were evaluated before surgery using spiral CT (shaded-surface display) of the pelvis to assess the residual deformity ( Fig. 1). The reconstructive procedure was preceded by a lateral transverse Salter-like osteotomy 6–8 days before the urological procedure . A Cohen-like crossed supratrigonal ureteric reimplantation was performed in three patients to cure VUR and to create space for bladder neck surgery. The bladder neck was reconstructed using the Erlangen bladder-neck plasty . The small capacity of the bladder remnants required augmentation using tubularized ileum in these three patients, but in one only a clam ileocystoplasty was necessary. The genitalia were reconstructed, with epispadias repair, using a modified Young-Cantwell technique during the same anaesthesia. Only in the 22-year-old woman with previous ureterosigmoidostomy was conversion to a sigma-rectum pouch (Mainz pouch II ) carried out, with ureteric reimplantation using the Abol-Enein technique  to resolve the bilateral obstruction. The antegrade fixation of the uterus, genital reconstruction and umbilicoplasty were carried out simultaneously.
All four patients had successful bladder closure at a follow-up of 2–4 years ( Fig. 2 and Fig. 3); there was no dehiscence of the closure, or bladder prolapse. The individual distances between the pubic rami before and after surgery are shown in Table 1. There were no wound infections and no patient had a persistent limp. Upper urinary tract dilatation improved in one patient and there were no changes in the three with initially normal upper tracts. Currently only the patient with the Mainz pouch II requires medication for metabolic acidosis. All patients remain dry day and night using regular CISC. Epispadias repair, e.g. genital reconstruction, resulted in good cosmesis in all patients, as assessed by observation and the patient's opinion. No additional cosmetic surgery was required and there were no urethral fistulae. The urethral reconstruction was good in two patients, but there was a urethral stricture in one, which resolved after urethrotomy on two occasions. There was no problem with excess mucus production and to date no bladder stones have developed. Bacteriuria developed occasionally but there were no episodes of acute febrile infection. There were no orthopaedic or anaesthetic complications of the procedure.
Table 1. Individual distances between the pubic rami before and after surgery
Advantages and disadvantages
The management of failed bladder exstrophy closure in adolescence and adulthood is a rare but challenging condition, and remains controversial. Individual management is needed to correct the factors that led to the initial failure, and the quality of life and undisturbed body image must be considered, by choosing urinary diversion or functional reconstruction .
Successful initial bladder closure is the most important factor for achieving urinary continence and sufficient bladder capacity . The potential for gaining adequate bladder capacity is usually decreased by each episode of bladder surgery, which is especially true when there are repeated attempts at bladder neck reconstruction . Therefore, any reconstruction procedure in adolescence usually requires augmentation [3,4,12]. None of the present patients required bladder neck closure and a Mitrofanoff channel; all remain dry using regular CISC after augmentation.
In a previous report , two women with primary unclosed exstrophy presenting at 22 and 35 years old needed a cystectomy and urinary diversion because they had long-standing irritation and inflammation of the bladder base. The present female patient had an excessive pubis diastasis, a colpoptosis, uncorrected external genitalia and a dilatation of the upper urinary tract after initial ureterosigmoidostomy. Osteotomy allowed good approximation of the pelvic floor and conversion into a sigma-rectum pouch with ureteric reimplantation resolved the bilateral obstruction.
In adulthood, rigidity of the bony pelvis and the musculoskeletal defect of the pelvic floor are major problems . The exstrophic pelvis has a wide pubic diastasis of 4–14 cm diameter [6,14,15]; not only is there an open pelvic ring, but also a short perimeter . Furthermore, patients with previous posterior osteotomy often have a malrotated asymmetric pelvis ( Fig. 4). Therefore, all patients should undergo preoperative axial CT of the pelvis to assess the residual deformity, especially to detect individual variability in the diastasis of the pelvis, as the pubic widening increases with age, and the variability of the musculoskeletal defects according to the severity of the disease [7,16]. Osteotomy should be specific to the bony deformity. The interpubic distance, an important factor, should be as small as possible to gain continence [1,7]. Women with insufficient pelvic adaptation may have a prolapsed uterus, a condition which may deteriorate in pregnancy . By closing the pelvis, the urethra will be placed within the pelvic ring. A normal urethrovesical angle after bladder neck reconstruction will lead to sufficient urethral suspension. Adaptation of the levator ani and the ischiopubic sling will help the voluntary control of the pelvic floor. With bony approximation the stress on the midline abdominal closure will be lessened. Finally, the healing of the urethra will be independent of any pelvic movement.
Comparison with other methods
At least three different types of osteotomy have been reported previously ; in most patients a bilateral iliac osteotomy was used, but a high morbidity rate and lower efficiency rate were reported . Furthermore, delayed symphysitic migration in patients with a stable interpubic distance and deteriorating pelvic asymmetry with overlapping bone fragments were reported [7,18], as also seen in the present patients. More successful results were reported with lateral transverse osteotomy, a Salter-like innominate osteotomy [6,7,19]. The narrowing of the true pelvis to the smallest anteroposterior pelvic diameter tightens the puborectalis sling and levator well. High success rates are reported for reclosure in patients with initially failed closure, persistent wide pubic diastasis and previous posterior osteotomy . Measurements of the birth canal size after Salter-like osteotomy show a significant decrease in pelvic outlet, with a reduced transverse outlet diameter (intertubero-ischial diameter of 3–3.5 cm) with no narrowing of the inlet, and therefore reduced intrinsic abdominal pressure . Complications include transient femoral nerve palsy, medial displacement of the distal bone fragment, avascular necrosis and pin penetration of the hip . In the present patients, this procedure was successful in all, with no complications. Similar osteotomies with no fixation were also reported . The anterior diagonal osteotomy was suggested as the most effective procedure [18,21]. The pubis is brought together effectively, the pelvic volume increased dramatically and the diastasis corrected symmetrically. Pubic ramus osteotomy can be carried out bilaterally to correct the pelvic diastasis, but unfortunately there are no long-term results .
In summary, osteotomy is beneficial for revising the failed bladder exstrophy closure or bladder neck reconstruction, even if it has been undertaken previously. Lateral transverse osteotomy allows excellent approximation of the pubis even in adolescence and adulthood. Currently, augmentation and CISC are suitable solutions for reliable continence and to preserve renal function. Urinary diversion should remain exceptional in adolescent and adult patients with failed bladder exstrophy reconstruction.