Complications and Quality of Life Following Urinary Diversion After Cystectomy

Transcription

1 EAU Update Series EAU Update Series 3 (2005) Complications and Quality of Life Following Urinary Diversion After Cystectomy Elmar W. Gerharz a, *, Alexander Roosen a, Wiking Månsson b a Department of Urology, Bavarian Julius-Maximilians-University Medical School, Oberdürrbacher Strasse 6, Würzburg, Germany b Department of Urology, University of Lund, Sweden Abstract Objective: The aim of this review is to outline specific surgical complications, metabolic consequences and quality of life (QOL) following urinary diversion in patients undergoing radical cystectomy. Methods: Based on a comprehensive literature search (MEDLINE) the published knowledge on urinary diversion was examined regarding the research question with an emphasis on contemporary cystectomy series. Results: Despite the fact, that urinary diversion is commonplace in these days and the existing literature is rather extensive, the vast majority of assumptions are based on low level evidence (retrospective, single-institutional case series with small sample sizes and short-term followup). There are few randomized trials in this field. Early and late surgical complications following radical cystectomy have decreased over the past three decades for both incontinent and continent diversion, but are still significant. While metabolic disturbances are common after continent forms of urinary diversion, the problems can be minimized in most cases. Most reports testify that QOL is high after cystectomy irrespective of type of urinary diversion although urinary and sexual problems are common. Conclusion: Careful patient selection, strict adherence to proper surgical technique and appropriate life-long follow-up are of paramount importance in the successful management of patients undergoing radical cystectomy for bladder cancer. # 2005 Elsevier B.V. All rights reserved. Keywords: Radical cystectomy; Urinary diversion; Urinary reservoirs, continent; Complications; Metabolic consequences; Quality of life 1. Introduction At first sight, it seems inevitable that there should be a link between complications of any kind and quality of life (QOL) after urinary diversion. However, a second look may reveal that these two concepts overlap to a certain degree, but are very different in principle. It may be both surprising and disappointing, that traditional objective surgical complications are not necessarily reflected by subjective changes in QOL with potential clinical relevance. In intestinal urinary diversion after cystectomy for malignant disease, one may arguably distinguish * Corresponding author. Tel ; Fax: address: Gerharz_E@klinik.uni-wuerzburg.de (E.W. Gerharz). between functional consequences of radical pelvic surgery (risk of incontinence and sexual dysfunction), surgical complications (e.g. ureterointestinal and stomal stenosis, reservoir rupture, insufficiency of the constructed continence mechanism) and metabolic consequences of artificially combining two extremely different organ-systems. One would assume that the nature, severity, duration and, above all, perception of a complication/consequence determine the effect on QOL. A major but silent complication like upper urinary tract obstruction with subsequent deterioration of renal function may not have the same impact as incontinence, as it does not interfere with daily life unless it is detected and communicated at follow-up. While night time incontinence may be more frequent in patients with neobladders than in patients with ileal conduit diversion, /$ see front matter # 2005 Elsevier B.V. All rights reserved. doi: /j.euus

2 E.W. Gerharz et al. / EAU Update Series 3 (2005) the same phenomenon may cause greater distress in the latter [1]; the overall QOL may still be high in both groups. While the inability to insert a catheter into a continent cutaneous reservoir may result in acute panic, it is doubtful whether this life-threatening problem has a lasting impact on QOL once it has been solved appropriately. As most QOL studies in urinary reconstuction were conducted to compare different forms of diversion rather than to evaluate the effect of particular complications on QOL, this review will deal with these issues separately. 2. Diversion-related surgical complications Despite the fact that ileal conduit diversion still is the most commonly performed type of reconstruction in conjunction with cystectomy, most reports on early and late complications after urinary diversion during the last two decades have focused on those that follow continent reconstruction. Specific surgical complications can be broadly divided into those related to reservoir, efferent (stoma, continence mechanism) and afferent limb (uretero-intestinal anastomosis). The different types of urinary diversion have unique characteristics predisposing to certain surgical complications as well as similarities related to intestinal surgery [2]. In several large series with various length of follow-up, the early and late reoperation rates for complications following continent cutaneous and orthotopic urinary diversion have been reported to be 3 7% and 13 30%, respectively [3]. For review of ileal conduit complications see [4]. Comparison between series of cystectomy with ileal conduit diversion, continent cutaneous diversion or orthotopic bladder substitution must be viewed cautiously as the patient characteristics differ. However, some series suggest that there are no major differences with regard to early complications between these groups. The most frequent long-term complications after conduit diversion are stomal/peristomal problems, parastomal hernia, conduit stenosis and upper tract deterioration. The incidence of these correlate with length of follow-up [4]. Peristomal dermatitis is almost always caused by urine trapped between the appliance and the skin. Stomal and parastomal complications reached 50% in some old series and even recently published series present figures around 30%. Peristomal hernia is seen in 5 15%. They are rather large than small and although the majority of patients are asymptomatic some need surgery. High recurrence requiring reoperation is seen. For first-time parastomal hernia repairs, stoma relocation is probably superior to fascial repair. We have seen infections with erosions and fistulas using synthetic mesh, which usually is employed in recurrent hernia repair. Newer techniques with incision placed lateral and far away from the stoma with closure of the fascial defect and using mesh material as onlay have been reported to give good results. Conduit stenosis is a condition affecting ileal conduits. It has never been described in colonic conduits. The whole, or part of the conduit, is transformed into a thick-walled tube without peristaltic activity. The pathogenesis of this disorder, which manifests late after diversion, is obscure. The clinical picture is colicky flank pain and/or fever and is produced by upper urinary tract obstruction. Treatment is by removal of the conduit or partial resection with or without ureteric reimplantation Uretero-intestinal stenosis (afferent limb) All types of anastomosis, refluxing or non-refluxing, can predispose to certain specific complications. It is generally thought that non-refluxing techniques are associated with a higher rate of uretero-enteric stricture. To date, however, studies evaluating the different anastomotic techniques that are well designed, prospective, and randomized, with appropriate numbers and long term follow-up have not been performed. Meanwhile there is an ongoing intense debate on the pros and cons of refluxing versus antirefluxing anastomosis especially with regard to orthotopic substitution. For conduit diversion, most often a direct end-toside refluxing anastomosis is used. In the series by Madersbacher of ileal conduit patients followed for at least 5 years, 10% of the patients developed stenosis of the ureterointestinal anastomosis [4]. A large series with over 350 patients who received a Hautmann neobladder with a Le Duc anti-refluxing anastomosis had an uretero-enteric stricture rate of 9.3% at average follow-up of 54 months [5]. In a series of 166 Kock ileal neobladders where the ureteroenteric anastomosis was performed with a refluxing end-to-end Wallace technique, the stricture rate was only 0.6% with follow-up of 2.7 years [6]. Abol-Enein and Ghoneim reviewed over 300 patients who received a Hautmann neobladder with an anti-refluxing serous lined extramural tunnel technique and reported a 3.8% rate of stricture with unspecified follow-up [7]. A recent series examined 130 patients who received either a Studer or Hautmann neobladder with a refluxing end to side Bricker anastomosis performed and found only 3 uretero-enteric strictures at a mean fol-

3 158 E.W. Gerharz et al. / EAU Update Series 3 (2005) low-up of 20 months [8]. Thus, the rate of ureteroenteric stricture has varied greatly amongst the large neobladder series that have incorporated both refluxing and non-refluxing anastomosis. Although the majority of strictures involving the uretero-intestinal anastomosis occur within one to two years following construction of the urinary diversion, late onset has been reported emphasizing the need for long-term follow-up. The majority of patients with obstruction are asymptomatic and diagnosed at the time of routine radiographic follow-up. The primary cause of uretero-intestinal stenosis is ischemia. When dissecting the ureters at the time of radical cystectomy, it is important to avoid devascularization of the distal ureteric segment which is to be reimplanted. It is also important not to angulate the ureter during reconstruction as that can also lead to postoperative obstruction. The EAU guidelines from 2004 recommend that follow-up is by ultrasonography and plain film. However, ultrasonography can never be a substitute for IVP or diuresis renography as obstruction can be present without gross dilatation and vice versa. In addition, ultrasonography is user-dependent. Primary treatment of ureteral strictures is usually endoscopic. Stenting, balloon dilatation and endoscopic incision have been employed with varying success in the management of uretero-intestinal stricture. Cold-knife techniques are thought to avoid potential thermal injury with resultant ischemia, fibrosis, and restricture from hot-knife. If endourological techniques fail in the correction of uretero-intestinal strictures, then open surgery may be necessary. The success rate of open revision of uretero-intestinal strictures with careful mucosa-to-mucosa realignment is approximately 90% Reservoir rupture or perforation Review of the literature surrounding rupture or perforation of urinary reservoirs made from bowel indicates that this complication is perhaps not as rare as commonly perceived [9]. Spontaneous rupture or perforation during catheterization has been observed after bladder augmentation, orthotopic ileal and ileocolonic bladder replacement, and in continent cutaneous reservoirs 4 weeks to more than five years after surgery. In a Scandinavian questionnaire survey 20 episodes of perforation of ileal and colonic pouches were documented in 18 patients, representing an incidence of 1.5% [10]. The patients present clinically with abdominal pain and distension, sepsis, ileus, fever, oliguria and peritoneal irritation. Ruptured augmented bladders share a common urodynamic feature of high leak point pressure of the urethra, with sensory and mechanical tolerance of high filling pressure. Other risk factors may include catheter trauma during intermittent self-catheterisation, urinary retention due to mucus retention or noncompliance with the catheterisation regimen, chronic infection, and decreased sensation of bladder filling. Desgrandchamps et al. [11] found acute or chronic overdistension, intraperitoneal adhesions and chronic ischaemic changes to be aetiological factors in five cases of rupture in patients with orthotopic ileal bladder replacement. The diagnosis can be made on low pressure cystography, although failure to demonstrate extravasation is not unusual, and endoscopy. CT is probably superior. Aggressive surgical treatment consists of immediate exploration, primary repair of the defect, drainage of the perivesical space, suprapubic cysto/pouchostomy and broad spectrum antibiotics [9]. Several cases of successful conservative management with continuous bladder drainage and antibiotic treatment for three weeks have been reported [12]. In clinical practice it seems that the diagnosis was often not considered, and treatment was significantly delayed in many cases of reservoir rupture. It is therefore essential to endorse awareness of this life-threatening complication. A practical suggestion to help alert physicians to the possibility of a ruptured urinary reconstruction is that these patients should carry a medical card stating the type of reservoir they have along with their special circumstances [9]. Sudden onset of abdominal pain in a patient with continent cutaneous diversion or orthotopic bladder substitution should be considered pouch rupture until proven otherwise Stomal stenosis/impaired catheterization after continent cutaneous diversion The growing number of techniques described for achieving continence in urinary reconstruction indicates that a universally applicable procedure with low complication rates has not yet evolved. Among the different techniques, however, the versatile Mitrofanoff principle [13] has reached significant popularity. While continence using the appendix, once established, is durable with late onset failure of 1 2% at about 4 years (see below) the notorious drawback of that technique is impaired catheterization due to a gradually stenosing stoma with a true incidence between 8 and 28% depending on the length of follow-up [14]. It has been assumed that Mitrofanoff channels only remain patent because of regular catheterization.

4 E.W. Gerharz et al. / EAU Update Series 3 (2005) Stomal stenosis has also been described in reservoirs with intussuscepted valves (9%) and plicated tubes (3 6%; Indiana pouch, Lundiana pouch) [3,15]. Although stenosis represents a minor technical problem, the subsequent inability to insert the catheter is a highly distressing complication with potentially fatal consequences in patients in whom the Mitrofanoff channel is the only route of evacuation. In these cases, the reservoir must be immediately emptied percutaneously. Skin stenosis may be dilated, incised or repaired by open surgery carefully removing fibrotic tissue. Some authors describe fewer recurrences with laser treatment than cold incision. All reports so far state that a proportion of such stenoses can be treated conservatively though generally with few figures. Dilatation alone seldom produces a lasting improvement. A high acceptance rate for patients receiving a continent catheterizable stoma was demonstrated by Horowitz et al. All patients who had the potential to catheterize either the Mitrofanoff tube or their native urethra preferred to use the artificial channel [16] Incontinence after continent cutaneous diversion The true incidence is difficult to calculate due to lack of standard terminology and use of non-validated questionnaires in retrospective series. Overall, only 3.2% of patients appear to have problems with incontinence. This rate varied amongst the various continence mechanisms with 5.8% of the patients with intussuscepted ileal nipple valves, 3% of patients with a tunneled appendix, and 0.6% of patients with stapled plicated ileocaecal valves reporting leakage [17]. Incompetent continence mechanisms usually require open revision. This can range from plication of the efferent limb to tapering a segment of ileum along an opened taenia. The technique of intussuscepted ileal nipple valve is elaborate with a long learning curve and the initial enthusiasm has been considerably reduced by reports on high complication rates. There are reports of success with treatment of injectable bulking agents for mild valve incontinence. Additionally, false passage of a catheter can create a small fistula beyond the intussuscepted nipple valve causing leakage. Lastly, problems with the nipple valve including ischemia, fibrosis, or prolapse can also result in incontinence with open surgical revision often required in such cases. In a long-term follow-up Jonsson et al. [18] reported that 31% of patients operated since 1984 had outlet revision, this figure dropping to 21% in those operated since % of surviving patients got a well-functioning reservoir. In a recent report on the Lundiana pouch, being a modification of the Indiana pouch, 6 out of 97 (6%) of the patients needed revision of the outlet due to leakage and all but 4 patients were continent at follow-up [15]. In their series of 118 patients with submucosally embedded in-situ appendix Gerharz et al. [19] report 3 patients with necrosis of the appendix resulting in total incontinence. The appendix had to be replaced by an intussucepted ileal nipple Incontinence after orthotopic reconstruction Patients with orthotopic reconstruction void by simultaneous relaxation of the pelvic floor musculature and raising of intraabdominal pressure by performing a Valsalva maneuver. Objective determination of continence/incontinence is hampered by differences in length of followup, reporting methodology and arbitrary definitions of degree of continence. Continence usually improves within the first 6 to 12 months postoperatively as the compliance of the reservoir increases. Daytime continence is achieved before night-time continence. The overall rate of good or excellent day-time continence (i.e. totally dry or use of one pad per day one year after reconstruction) is approximately 85% to 90% [20]. Daytime continence rates may decrease four to five years postoperatively, partly due to decreasing urethral sphincter tone with advanced age. Some degree of night leakage is common in all types of orthotopic diversion. The majority of series report a prevalence of night time leakage of 20 30%. Ghoneim et al. [21] and El Bahnasawy et al. [22] found similar rates of enuresis varying from 27 50% at increasing follow-up intervals beyond 12 months in male patients with either hemi-kock or Hautmann ileal neobladders. In univariate analysis, incontinent patients had higher pressures, maximal capacity, postvoid residual urine volumes, and rates of positive urine cultures, and lower maximal urethral pressures, flow rates, and compliance compared to continent patients. However, in multivariate analysis only the amplitude of uninhibited contractions and increased post-void residual volume were associated with night time incontinence [23]. Persistent severe incontinence is a difficult clinical problem. In addition to intermittent self-catheterization, therapeutic options are augmentation of the neobladder, peri-urethral collagen injection, placement of a urethral sling or implantation of an artificial urinary sphincter. Single patients may even need removal of the neobladder and cutaneous diversion. Pharmacotherapy is of little or no value Chronic urinary retention/voiding dysfunction after orthotopic reconstruction Chronic urinary retention (so-called hypercontinence ) requires CIC in 4 33% of men and in 0 53%

5 160 E.W. Gerharz et al. / EAU Update Series 3 (2005) of females with orthotopic substitution [20,24,25]. Urethral stricture must be ruled out as a cause of incomplete voiding. The reason for the higher rate of voiding dysfunction in women remains largely unclear. Most investigators assume formation of a pouchocele to be the main factor in hypercontinence; in their concept, lack of posterior support of the neobladder leads to angulation and obstruction of the neobladder-urethral junction Recent experiences suggest that reducing reservoir capacity by using approximately 40 cm length of bowel rather than 60 cm may contribute to improved voiding function. 3. Metabolic consequences Ever since Ferris and Odel described the electrolyte pattern of the blood after bilateral ureterosigmoidostomy in the middle of last century [26], the pathophysiology of urinary diversion through intestinal segments has been a challenging exercise. The potential consequences are a result of either a reduction in the absorptive bowel capacity because of functional loss of those segments required for reservoir construction, or of the highly unphysiological chronic exposure of the reconfigured bowel to urine [27] Electolyte and acid-base balance Various electrolyte disturbances can occur based on which segment of the gastro-intestinal tract is chosen for urinary reconstruction: stomach, jejunum, ileum, or colon. By understanding the normal physiologic function of each segment of the gastrointestinal tract, one can predict which abnormalities are likely to develop when a particular segment is used in urinary tract reconstruction [28]. The most prevalent and widely researched consequence of intestinal urinary diversion is chronic metabolic acidosis. Depending on definition, diagnostic modality, reservoir characteristics, renal function and length of follow-up, it has been reported in up to 100% of patients after ureterosigmoidostomy, and bladder substitution or continent diversion using ileal and/or colonic segments. In intestinal urinary diversion metabolic acidosis is mainly due to reabsorption of ammonium chloride and, to a much lesser extent, secretion of bicarbonate. In cross-sectional and longitudinal studies of patients with ileal urethral Kock reservoir all median values for capillary blood ph, carbon dioxide pressure, standard bicarbonate and standard base excess were within the reference interval of the normal population but significantly lower than in control subjects [29]. The values for 44% of patients and none of the controls were consistent with mild metabolic acidosis. Median values of serum sodium, potassium, chloride and anion gap were similar in patients and controls and remained so throughout the observation period. While there were no differences in net urinary acid excretion, urinary ph and excretion of ammonium were significantly higher in the patients with ileal reservoirs. In their evaluation of the metabolic alterations at different levels of renal function in patients with colonic urinary reservoirs, Kristjansson et al. [30] showed that a glomerular filtration rate (GFR) of approximately 55 ml/min/1.73 m 2 is sufficient to compensate adequately for acute and chronic endogenous acid loads. When Racioppi et al. [31] compared acid-base and electrolyte balance in ilecaecal and ileal neobladders, they concluded that length rather than the kind of bowel used for bladder replacement, appears to be important in maintaining homeostasis. Metabolic acidosis can on occasions be life threatening, so it is of great importance that prophylaxis and treatment with oral bicarbonate are commenced early. Such supplementation is usually simple, cheap and without serious side-effects Renal function The preservation of renal function is both the ultimate goal and an essential prerequisite of successful intestinal urinary diversion. This raises the crucial question of whether storage of urine in bowel is inherently damaging to kidney function. Unfortunately, most reports on renal function have used IVP and serum creatinine only, methods which are inadequate for proper assessment. In a recent study [32] of 53 patients with intestinal urinary reservoirs, and a minimum follow-up of ten years, GFR was measured using 51 Cr-ethylene diamine-tetra-acetic acid (EDTA). Mean GFR decreased from 97.9 to 92.9 ml/min/ 1.73 m 2 (P = 0.24). The overall decrease was due to a fall in GFR in 10 patients (19%) whose GFR fell by at least 20%. The causes of renal deterioration were identified in all instances, and included chronic retention and/or infection due to inadequate catheterization and poor compliance, uretero-ileal stenosis and high storage pressure in the reservoir. For the remaining 80% of the patients, the storage of urine in intestinal reservoirs did not change renal function over at least ten years. This is supported by long-term outcomes in 176 patients, who underwent continent urinary diversion using a Kock ileal reservoir between 1975 and 1999 [18]. At the time of investigation, 126 patients were still alive. Serial determination of 51 Cr-EDTA clearance indicated that GFR decreased somewhat with

6 E.W. Gerharz et al. / EAU Update Series 3 (2005) long-term follow-up. The decrease, however, was similar to the known decrease in kidney function with advancing age. According to Granerus and Aurell, GFR decreases by 0.4 ml/min/1.73 m 2 per year before the age of 50 years, and 1.0 ml/min/1.73 m 2 afterwards [33]. Jonsson et al. concluded that continent urinary diversion does not appear to impair kidney function per se. In their evaluation of renal function up to 16 years after conduit or continent caecal urinary diversion, Kristjanson et al. found a moderate drop of GFR in all groups with no significant intergroup differences, and the continent caecal reservoir compared favourably with conduit diversion [34]. There is no evidence to suggest that patients with continent reconstruction fare worse than conduit patients with regard to renal function. It might in fact be that the former patients do better due to less bacterial burden in a pouch than in a conduit and the use of antirefluxing ureterointestinal anastomosis. In the series of ileal conduit by Madersbacher et al. [4], morphological/functional deterioration developed in 27%, most often in the form of hydronephrosis or shrunken kidney. Renal pathology was present in 40% after 5 years, increasing to 80% after 10 years. These data suggest that life-long surveillance of the upper tracts from both the functional and structural aspects is appropriate, and that a fall in GFR beyond that expected with advancing age should prompt a careful search for its cause Bone disease Impaired bone metabolism following urinary diversion through intestinal segments has always been a controversial subject of unclear clinical relevance [35]. It was Boyd who proposed a link between ureterosigmoidostomy, chronic acidosis and metabolic bone disease in 1932 [36]. His assumptions have been perpetuated and extended to other surgical techniques (ileal ureteral replacement, colocystoplasty) in a number of case reports in the medical, surgical and orthopedic literature in the following decades. Whereas the perpetuated pathophysiological considerations seem conclusive in theory, the role of acidosis and malabsorption is less clear in animal experimentation and, even more so, in the clinical reality of modern continent diversion. It has been proposed that the main skeletal effects may be related not just to impaired absorption of calcium and vitamin D, but also to chronic acidosis. It is said to act through direct physicochemical non-cell-mediated cation-hydrogen exchange at the bone surface (bone as buffer system), stimulation of osteoclastic bone resorption and inhibition of osteoblastic bone formation. The often quoted assumption that acidosis impairs 1-a-hydroxylation of 25-hydroxycholecalciferol in the kidney with consequent deficiency of the biologically active metabolite has been refuted in vitamin D-replete rats and in humans and therefore can be discounted. In patients with continent diversion, however, the role of acidosis and bone disease is less clear. When continent intestinal urinary reservoirs gradually gained popularity at the beginning of the nineties, the issue was revived in a series of clinical reports, which were mainly retrospective. The authors used the terms low-grade, mild, slight and subtle when qualifiying acidosis, without clear definitions. Bone demineralization was found in only three series, and in all instances, it was minor and asymptomatic in nature [37 39]. Despite equally sophisticated methodology, and a follow-up of up to 30 years the other series failed to demonstrate such changes, even in the presence of acidosis [40]. The assumption that colonic reservoirs have a higher risk of developing metabolic bone disease could not be confirmed by clinical data. Out of the 11 clinical series, calcium and phosphate were determined in all, PTH in six, vitamin D and bone-specific alkaline phosphatase in five, osteocalcin in three and calcitonin in one series; apart from elevated osteocalcin after ureterosigmoidostomy and increased alkaline phosphatase and ionized calcium in patients with orthotopic ileal neobladder, all other parameters were normal. Normal PTH and vitamin D levels in all the clinical studies make a contribution of these hormones to bone disease in continent urinary reconstruction extremely unlikely. As early correction of base excess is easy and probably a common policy in patients with intestinal urinary reservoirs, it will be virtually impossible further to study the natural history of bone metabolism after urinary diversion. While there is no need for a bone specific follow up in asymptomatic adults with a normal acid-base balance, particular attention should be paid to children and to all patients with impaired renal function Mucus production Following transposition into the urinary tract, intestinal segments continue to produce mucus and problems related to excessive production do not diminish with time [41]. This mucus can prove troublesome when it interferes with upper urinary tract drainage or emptying of the bladder reservoir. Acute ureteric and urethral obstruction by mucus plugs has been described, in one case with rupture of an ileal neobladder. Moreover, mucus may be an important factor in the genesis of reservoirs stones, possibly acting as a nidus.

7 162 E.W. Gerharz et al. / EAU Update Series 3 (2005) In a recent prospective randomized, double-blind placebo-controlled crossover study [42] the effectiveness of N-acetylcysteine, aspirin and ranitidine in reducing mucus secretion and urine viscosity was determined. Pre- and post-treatment 24-hour urine samples from 12 patients with an ileal conduit and 31 patients with bladder reconstruction were analyzed. A disease-specific questionnaire and a SF-36 quality of life survey were completed. Analysis of questionnaires and laboratory results failed to demonstrate any benefit of taking muco-regulatory agents compared with placebo. It is possible that the mucus may interfere with pregnancy tests. The test was positive in 57% of patients with enterocystoplasty regardless of gender [43]. Patients should be aware of this and not rely on commercial test kits Bacteriuria The incorporation of intestinal segments in the urinary tract favors bacterial growth of the skin flora, anaerobic bacteria, and uropathogenic strains. The bacterial strains growing in the reservoir change spontaneously, indicating colonization rather than infection. The route for infection is ascending; bacteria enter the urethra, or the abdominal stoma, followed by colonization of the reconstructed lower urinary tract [44]. Bacteriuria is common in all kinds of reconstruction, however, urine from neobladder patients with complete emptying is reported to carry bacteria to a lesser extent. Clean intermittent catheterization and residual urine seem to increase the bacterial burden. Symptomatic upper urinary tract infections usually reflect afferent limb problems with reflux or obstruction requiring intervention. Patients with augmentation cystoplasty constitute a distinct subgroup, in which the remaining part of the bladder tissue is an important determinant of urinary tract infection susceptibility. The increased rate of bacteriuria in the reconstructed patients indicates a lack of antibacterial defenses, and the symptomfree state of the patients suggests that only a restricted host response is triggered. The role of the specific and inflammatory antibacterial defenses in the reconstructed lower urinary tract remains largely unknown. Bacterial growth in the urine correlates poorly to increased anti-body titres. Thus, asymptomatic bacteriuria in these patients is generally of no clinical relevance and should not be treated with antibiotics [45] Urolithiasis When urine is stored in intestinal reservoirs its composition is altered through absorption and excretion of ions by the intestinal mucosa. The urine is further changed in its physical composition, by infection and by the mucus [46]. As a consequence, patients with intestinal urinary reservoirs have an increased propensity to form urinary calculi, predominantly in the reservoir rather than in the upper tracts, and with a high tendency to recur. As most patients will have multiple physical factors such as immobility, need for self-catheterisation and poor urine drainage, it is not certain that an intestinal reservoir is cause of stones on its own. The incidence increases gradually with length of follow-up; mean time to stone formation is between 24 and 45 months (range one month to 10 years). Most reservoir stones are found incidentally, but some patients may present with symptomatic infection or deterioration in urinary continence. The stones are mainly infection-related, with a mixture of struvite, carbonate apatite and ammonium hydrogen urate. Variable amounts of calcium oxalate may occur [46]. Risk factors include the presence of foreign material (e.g. staples), recurrent and chronic infection, abnormalities of the composition of the urine, mucus production, urinary stasis and non-compliance with irrigation and catheterization regimes. Outflow conditions may play a role, as stones formed more commonly after bladder outlet resistance procedures and in patients with catheterizable abdominal wall stomas in a study of 286 children following bladder augmentation [47]. This was confirmed by a recent analysis of 146 adolescents with enterocystoplasty (mean follow-up of 3.4 years): Woodhouse and Lennon found that all patients with an augmented bladder drained by a suprapubic Mitrofanoff channel formed stones at some stage. The incidence of stones in Kock pouches and in reservoirs drained by urethral catheterization was 50% and 9% respectively. As no patient who voided spontaneously formed stones the authors concluded that this complication is strongly related to the lack of downward, gravitational emptying [48]. There are significant differences reported in the rate of stone formation between surgical techniques and type of diversion. In a comparison of Indiana and Kock pouches with a minimum follow-up of 12 months, stones developed in 12.9% of colonic and in 43.1% of ileal reservoirs. More than half of the patients had stones on more than one occasion [49]. In a single institutional series of 193 cutaneous ileocaecal reservoirs stones were observed only if an intussuscepted ileal nipple had been used as continence mechanism (2.8%); in all cases, staples were identified as the nidus. Not a single stone was observed in pouches with submucosally embedded in situ appendix as efferent limb [50]. Contrary, 10% and 12% stone incidence was reported in patients with right colonic pouch for

8 E.W. Gerharz et al. / EAU Update Series 3 (2005) continent diversion and orthotopic substitution, respectively [15]. Stone formation is rarely reported in ileal neobladders [20,25]. Several authors have focused on urine composition. In a Swedish study of 27 patients with continent ileal reservoirs the presence of urease-producing bacteria was associated with stone formation. Urine from those patients tended to have high calcium and low citrate concentrations. After incubation with urease, significantly more and larger particles were observed in the urine from stone formers. There was a strong correlation between urinary calcium content and urinary ph when the urease-induced precipitation commenced [51]. Osther et al. compared biochemical and physicochemical properties of urine from patients with ilealurethral Kock reservoirs, and normal men who served as controls. All subjects had sterile urine at the time of urine collection. While 24-h urinary volume was similar in both groups there were significant differences in most of the other variables. The most striking findings were markedly lower urinary excretion rates of citrate and higher ph in patients compared with controls. These findings were reflected in significantly higher levels of urinary supersaturation with respect to calcium oxalate, calcium phosphate, brushite and magnesium ammonium phosphate [52]. Depending on stone size, trans-stomal pouchoscopic ultrasonic lithotripsy, prone position extracorporeal shock wave lithotripsy, percutaneous approaches and open surgery have been used successfully, alone or in combination, for fragmentation and clearance. Daily irrigations to clear mucus and crystals, as well as complete emptying of the reservoir, may have important roles in preventing stone formation. Oral citrate supplementation and acetohydroxamic acid, a urease inhibitor, have been suggested for prophylaxis. The aetiology of non-infective urolithiasis is multifactorial in patients with ileal resection. With decreased availability of bile salts fat malabsorption occurs. Fatty acids bind with calcium and magnesium to form soaps, resulting in increased levels of free oxalate available for absorption. Moreover, fatty acids directly increase colonic permeability to oxalate Bowel function Removal of bowel segments for the purposes of urinary diversion may have a significant effect on bowel function. The remaining bowel may not be able to fulfil its normal role, resulting in not only malabsorption syndromes, but also dysfunctional defecation. There appears to be a significant risk of developing symptoms of increased bowel frequency following urinary diversion and reconstruction [53]. The effect of functional bowel loss in continent urinary reconstruction is mainly related to bowel type and length. While a considerable length of colon can be excluded from the fecal stream without detrimental nutritional disturbances, resection of the terminal ileum and ileocaecal valve in particular can lead to several well established pathophysiologic problems, including malabsorption of bile acids, maldigestion of fat, mixed secretory-osmotic diarrhoea, acceleration of bowel transit, hypovitaminosis, and an increased propensity to form gallstones and renal tract stones. Although there are many theoretical mechanisms of malabsorption and diarrhoea, the clinical picture may be relatively benign. Ben-Chaim et al. [54] compared 50 patients after ileocaecal resection for urinary reconstruction to 50 patients following left colectomy for cancer. No significant change of bowel habits was noted in the control group. Among patients with ileocaecal reservoirs, 42% had transient loosening of stools, which gradually resolved within three months after surgery. Twelve months postoperatively, only six patients with solid stools before reconstruction had loose bowel motions, and none suffered diarrhoea. Roth et al. evaluated the risk of diarrhea after urinary diversion interviewing 65 patients with ileal reservoirs and 35 patients with ileocaecal reservoirs [55]. Chronic diarrhoea of greater than 6 months in duration was observed in 11% and 23%, respectively. Symptoms subsided spontaneously in two patients in each group. All but two of the remaining patients responded to treatment with either cholestyramine, loperamide or psyllium. Olofsson et al. determined the effect of urinary diversion with a Kock ileal reservoir on bowel habit, using the SeHCAT (selenium-75 labeled tauroselcholic acid) test to assess bile acid absorption [56]. Out of 75 patients, 13% had 15 or more stools per week, and 15% complained of always having loose stools. All patients with more than 10 defecations per week had a pathological SeHCAT test. Several reports demonstrate a high incidence of bowel dysfunction in patients who have an enterocystoplasty, much higher than in patients with ileal conduit after cystectomy, despite the fact that a similar length of ileum was removed. Another intriguing finding is that the severity of symptoms seems to be related to the condition for which augmentation was performed [57,58]. Altered vitamin metabolism has been the subject of numerous sophisticated clinical studies. A particular emphasis has been put on vitamin B12, as patients may develop irreversible neurologic disease in case of unrecognized deficiency.

9 164 E.W. Gerharz et al. / EAU Update Series 3 (2005) Stein et al. determined the levels of vitamins A, B1, B2, B6, B12, D and E, folic and bile acids, ammonia and intracorpuscular vitamin B12 in 137 patients with ileocaecal reservoirs. The stratification of patients was according to age at operation and length of follow-up. All variables were normal in children, whereas in the 86 adults, mean serum vitamin B12 decreased significantly within four years of the operation. There was no significant decrease in intracorpuscular vitamin B12 [59]. Several comparative studies were performed. While no patient with a neobladder fashioned from ascending or sigmoid colon developed a low vitamin B12 concentration, decreased levels were found in three of 18 patients five to six years after ileocolic bladder substitution, and in three of 22 patients nine months to three years after ileal neobladder construction [60]. Racioppi et al. [61] found a tendency for vitamin B12 levels to fall in patients in whom ileum is used for orthotopic bladder substitution; in their study, resection of the ileocaecal segment, including the ileocaecal junction, did not alter the level of vitamin B12. Pannek et al. [62] compared Kock pouches and ileal neobladders, and concluded that losing no more than 50 cm of terminal ileum seems to be necessary for sufficient vitamin B12 absorption. Assessment of complete cobalamin profiles (serum vitamin B12, methylmalonic acid and homocysteine) recently revealed a much higher incidence of tissue cobalamin deficiency in patients with various types of urinary diversion than measuring the vitamin B12 level alone [63]. Substitution of vitamin B12 is simple and even less expensive than regular determination, and so empirical supplementation should perhaps be considered Risk of malignancy and new tumour formation Patients and their urologists are quite rightly concerned about the possibility of neoplastic change within intestinal segments transposed to the lower urinary tract. There is definite consensus that mixing of faecal and urinary streams by standard ureterosigmoidostomy or it subsequent variants results in an increased risk of development of adenocarcinoma [64]. The first case of malignancy following ureterosigmoidostomy was reported in a French journal in 1929 [65]. Meanwhile, more than 200 cases have been published. Based on several large series the incidence has been estimated to be 3.5% to 19%. When compared to the incidence of primary colorectal cancer in agematched controls, tumours occurred 8 to 550 times more frequently in patients with ureterosigmoidostomy. The majority of these malignancies originate at the uro-enteric anastomosis approximately 15 to 25 years after diversion. Patients with these tumours had a higher stage and an increased mortality (up to 50%) when compared to similar series of primary intestinal tumours. The regular use of screening colonoscopy has shown that adenomas can be found in % of patients approximately six years before the reported mean latency period for carcinoma development (26 years; range 6 54) [66]. The mean latency period for tumour development is significantly shorter if ureterosigmoidostomy had been performed for malignant disease [67]. The risk of secondary malignancy has not been changed by the numerous recent modifications of the technique (sigma-rectum pouch). While tumour development is a well established late complication of ureterosigmoidostomy, the risk of carcinogenesis in urinary diversion via isolated intestinal segments without a mixture of faeces and urine is controversial. Some authors claim that for the majority of patients who undergo procedures involving the transposition of intestinal segments to the urinary tract there is no evidence that the use of such segments intrinsically increases the magnitude of risk for neoplastic change. Because of the long latency period and the relatively short follow-up in many patients with orthotopic bladder substitution and continent cutaneous diversion the true incidence of malignant change is unknown. Up to 2003, 81 tumours have been reported in conduits (n = 18), enterocystoplasies (n = 45), rectal reservoirs (n = 5), neobladders (n = 3), colonic pouches (n = 6) and ileal ureters (n = 4)[67]. The majority of tumours were adenomas and adenocarcinomas, followed by transitional cell carcinomas, carcinoids and one sarcoma. In contrast to ureterosigmoidostomy, only half of the tumours developed immediately adjacent to the uretero-intestinal anastomosis, the other tumours were located either in the intestinal or the vesical part of cystoplasty. Tumour latency period was 22 years (0.3 28) in benign underlying conditions and seven years (0.3 18) if the diversion was performed for pelvic malignancy, suggesting that the latter may predispose to a more rapid tumour induction. Mean latency period was also shorter in colonic (11 years; 0.3 to 28) than in ileal reservoirs (20 years; 5 40). Depending on extent and location of the tumour, treatment ranged from local excision of the tumour to partial and total resection of the diversion. In many case reports follow-up is short or not recorded. Eleven patients have died of their secondary malignancy [67]. The pathogenesis of such tumours remains uncertain, but in common with most sporadic malignancies is likely to follow a combination of carcinogenic action

10 E.W. Gerharz et al. / EAU Update Series 3 (2005) and mucosal cellular instability at the anastomotic site leading to sequential changes from dysplasia through to invasive disease [64]. While N-nitrosamines have been blamed in the beginning, a causal link with the tumours could never be established. The levels are the same in all ureterosigmoidostomies; the association could, therefore, be coincidental. From the available data, the initiating event for mutagenesis appears to occur early in the postoperative period. Epithelial proliferation at the healing anastomosis in the presence of carcinogens may be an aetiological factor. Other factors may include the presence of bacteriuria, suture material, catheterization or calculi: these may provoke a chronic inflammatory response that can further promote carcinogenesis [68]. The suggestion from animal studies that prophylaxis with either antibiotics or nitrosamine inhibitors such as vitamin C may be useful lacks any experimental or clinical basis for efficacy. For the early detection of neoplasms, some have proposed that regular endoscopy should be performed annually starting 3 years after surgery for malignant disease and 10 years in patients with benign conditions [67], although this will generate substantial amounts of additional work, and possibly morbidity, for a problem which may in fact be very rare indeed in the total population of patients who have a diversion. Further clarity is required on how best to target endoscopic follow-up of these patients. This is compounded by the significant difficulties in the interpretation of urine cytology after diversion, rendering cytology an ineffective tool in surveillance for tumours related to diversion. 4. Quality of life Some authors claim that quality-of-life considerations were among the main driving forces behind the development of continent urinary diversion [69]. This attribution, though logical, is difficult to substantiate in retrospective. As serious scientific attempts to measure QOL in this particular setting were never made in the early days, it was much more an aprioriassumption that higher QOL was provided through continent reconstruction compared to what was previously standard treatment, i.e. the ileal conduit. This view, which may have prevented the conduction of prospective randomized trials, has been questioned recently, when methodologically sound studies were undertaken correcting for confounding factors such as age and co-morbidity. In a recent evidence-based analysis of QOL studies in patients with urinary diversion [70], it was concluded, that the existing literature does not support the assumption that continent reconstruction provides higher QOL than ileal conduit diversion. By and large, patients with ileal conduit diversion do not seem to fare worse than those with continent reconstruction, suggesting that patients adapt to whatever is required of them. Urinary problems and sexual dysfunction are common with all methods of urinary diversion, but most studies have shown that overall quality of life after cystectomy remains good for most patients. Why do patients with a bag attached to their abdomen for storage of urine seem to do as well as those with an orthotopic reconstruction? There are several plausible reasons for this finding. Firstly, researchers may have asked the wrong questions, i.e. the employed instruments/methods may have been inappropriate to reveal truly existing differences between techniques. In the future, this could be avoided by developing and using questionnaires for patients who have undergone cystectomy due to bladder cancer. Secondly, patients in the referred studies probably were subjected to a careful method-to-patient matching and received ample information concerning surgery and what to expect postoperatively. A high QOL in the majority of patients therefore can be viewed as a retrospective validation of successful patient selection, objecting the nihilistic misinterpretation that urologists should go back to producing ileal conduits on a large scale. Comprehensive patient education, exploration of the pros and cons of the various alternatives, and active patient participation in treatment decisions may be the key to postoperative satisfaction. Thirdly, there may be important QOL related concepts to which insufficient attention was paid. Coping with the new diversion, disease-related social support and confidence in the success of treatment may have a crucial influence on a patient s QOL and may help to interpret and understand results. In addition, human beings tend to adapt to circumstances after a period of adjustment, making timing crucial in studies of preand post-treatment. Especially after a major operation the terms of reference by which QOL is judged may change over time as the patient reassesses his life ( response shift ). There is ample evidence that persons with severe handicap and patients close to death can maintain good quality of life ( well-being paradox ), especially if they are embedded in an intact network of family and friends. 5. Conclusion While in the first generation of papers on continent cutaneous urinary diversion and orthotopic bladder