To assess the long-term results in patients treated using a modified ureterosigmoidostomy (Mainz II).
To assess the long-term results in patients treated using a modified ureterosigmoidostomy (Mainz II).
Between 1994 and 1999, 17 patients had their lower urinary tract reconstructed by a ureterosigmoidostomy, modified by reconfiguring the rectum to make a low-pressure reservoir (Mainz II). All patients were followed on a standard protocol. Data were extracted from the database and from a review of the case-notes. In 12 patients the procedure was with a radical cystectomy for carcinoma. Five had a failed conventional ureterosigmoidostomy for bladder exstrophy and therefore proceeded to a Mainz II. The data on continence and complications were retrieved for a retrospective analysis; the mean (range) follow-up was 6.4 (4–8.6) years.
Ten of those with bladder cancer and one in the revision group were continent. Two patients in the revision group had sufficiently severe nocturnal incontinence to require conversion to a colonic conduit. Seven of the 17 patients had hyperchloraemic acidosis, one had pyelonephritis and one had renal stones. There were no anastomotic neoplasms.
The Mainz II has a good outcome if used as the primary procedure. In patients with an existing ureterosigmoidostomy who are incontinent, detubularization of the rectosigmoid alone is unlikely to restore continence.
In 1852 Simon first described the diversion of urine into the rectum  but problems associated with this procedure, e.g. infection, anastomotic neoplasms and hyperchloraemic metabolic acidosis, caused a decline in its use, in favour of the ileal conduit  and later the orthotopic bladder. However, there remain many young patients who have had few complications and are most satisfied with their conventional ureterosigmoidostomy diversion. If the traditional complications could be overcome, evacuation of urine and stool through the rectum would be very acceptable to patients.
Fisch et al. and Kock et al. revived the ureterosigmoidostomy by introducing modifications which enable a reduction of storage pressure by rectal detubularization. Our early results and technique of the modified ureterosigmoidostomy presented in 1998  were encouraging; we now assess the long-term outcomes.
Since 1994, 31 patients have had a modified ureterosigmoidostomy; we assessed the long-term outcomes in patients 4–8.6 years after surgery (operated between 1994 and 1999). All patients were followed using a standard protocol for lower tract replacement . Data were extracted from the resulting database and by a retrospective analysis of their medical notes for 17 patients (four women and 13 men; mean age 49.7 years, range 23–76).
The patients were divided into two groups (Table 1); those in group A had had a radical cystectomy and modified ureterosigmoidostomy (Mainz II) for bladder carcinoma. These patients had a Mainz II as a primary procedure. Patients in group B had initially had a conventional ureterosigmoidostomy for bladder exstrophy but subsequently had become incontinent with or without other complications; their diversion was therefore changed to a Mainz II.
|Variable||Group A||Group B||Total|
|Mean (range) age, years||55.4 (45–76)||45 (23–62)||49.7 (23–76)|
The modified ureterosigmoidostomy technique (Mainz II) was described by Fisch et al.; we previously and successfully reproduced this technique and reported the short-term follow-up results . Briefly, after cystectomy, the colon lies in the operating field, conveniently next to the cut ureters; 20 cm of lower sigmoid and upper rectum are identified and the anterior aspect opened longitudinally. The posterior wall is then closed in an inverted ‘U’ form and the ureters anastomosed at convenient sites in submucosal tunnels (Fig. 1).
The outcomes assessed were continence rates, incidence of hyperchloraemic metabolic acidosis, pyelonephritis, stones and anastomotic neoplasms. Continence was defined as being totally dry; incontinence was divided further into daytime, nocturnal and total incontinence. Hyperchloraemic acidosis was diagnosed by regular blood analysis (serum chloride, bicarbonate, negative base excess), and treated with oral bicarbonate supplements (600–2400 mg twice daily). The data were analysed descriptively because there were few patients.
The outcome measures were compared between group A and group B for the 4–8.6 years of follow-up after the Mainz II. In group A 10 patients were totally dry, compared to one in group B; both incontinent patients in group A were wet only during the day but they wore no pads and were wet 2–4 times per week. In contrast, the four in group B were wet mainly during the night, but one of them also had daytime symptoms (Table 1). This patient's symptoms were severe enough to warrant constructing a colonic conduit. Another patient from group B also had a colonic conversion for incontinence.
All other outcomes from the two groups are also summarized in Table 1; the seven patients with hyperchloraemic acidosis required medical treatment with oral sodium bicarbonate, monitored by regular blood tests. One patient had right-sided pyelonephritis at 10 months after the procedure, which resolved with conservative antibiotic management, and there were three further episodes. No anastomotic neoplasms were found during the follow-up; there was one renal stone but no reservoir stones.
Two patients died from their metastatic disease but terminal care was not compromised by their diversion; a large Foley catheter was used to drain the rectum when patients were immobile.
The much-cited paper by Simon  describes one case of ureterosigmoidostomy and the patient's postmortem report; with such a provenance it is surprising that the operation was ever used again. Because of its high storage pressures (200 cmH2O ) incontinence and pyelonephritis are common. Subsequently, it was reported that neoplasia at the anastomosis between the ureter and colon occurred in 24% of patients after 20 years . It is ironic that it was largely replaced by the ileal conduit just as the management of the complications was becoming understood .
However, in the senior author's clinic there is a group of patients who are very satisfied with a conventional ureterosigmoidostomy. Some form of anastomotic neoplasia has occurred in 20%; all of these patients have opted for a local excision of the neoplasm and re-formation of a ureterosigmoidostomy. It would seem that even when patients experience that which could be considered the worst complication, a good ureterosigmoidostomy appears to be better than any of the alternatives.
More recently, detubularization of the rectum has enabled a low-pressure reservoir to be formed , which has led to a decrease in the associated morbidity [3,11]; the follow-up, mean basal pressure in the pouch and contraction value in these two studies (for 20  and 47  patients) were 9–36 months, 19 cmH2O and 19.1, and 1–20 months, 24 cmH2O and 35, respectively. The modified ureterosigmoidostomy (Mainz II) when used as a primary procedure (group A) rendered 10 of 12 patients totally dry, compared with only one of five in group B (conversion to Mainz II). Incontinence is devastating, as the effluent is feculant and malodorous, and is socially damaging. Indeed, two patients from group B required conversion to an incontinent urinary diversion (colonic) for social reasons.
In group A the incontinence was described as stress-related and occurred at 1–4 years after surgery; it is difficult to explain these two cases. Possibly neurological damage at surgery caused progressive damage to the pelvic floor and its innervation.
In group B the incontinence was nocturnal, suggesting a sphincter problem, also seen in orthotopic neobladders. In each case it was presumed that the preoperative incontinence was caused by the high reservoir pressure, because no other cause was obvious. After surgery one patient was completely dry and two were improved; in the two who continued to have unacceptable incontinence, no other cause was found on investigating the sphincters and pelvic floor. In patients with a pre-existing conventional ureterosigmoidostomy that is incontinent, anorectal function should be investigated thoroughly; ‘rectodynamics’ and pelvic floor electromyography are the minimum requirement. If conversion to a Mainz II is selected the prognosis for continence should be guarded.
Comparing our long-term results with other studies of modified ureterosigmoidostomy (Table 2) [12–22] is difficult as a literature review showed a maximum follow-up of 4 years, while in the present study the minimum was 4 years. The early results of the present series  showed all to be continent in group A at 3 months and half of those in group B. Other published rates of continence vary for the early follow-up, at 91–100%, in keeping with our original series. However, from our experience it would seem that incontinence can develop after some years of continence, for no apparent reason.
|Study||Year||N||Follow-up, months||Reason for surgery||Continence, %|
|||2001||11||25–60||4 CC, 2 BC, 2 UC, 1 EC, 1 VC||100||91|
|||1996||73||1–34||15 C, 14 Ex, 3 trauma, 1 SU||94.5||98.6|
|||1998||34||5.4–68.1||30 BC, 3 Ex, 1 intractable incontin.||100||97|
|||1996||20||9–36||18 BC, 2 VVF||100||100|
|||2001||12||31.7 (mean)||12 BC||100|
|||2001||60||31 (mean)||54 BC, 2 VVF, 1 UVF, 1 UTB, 1 US, 1 failed USig||98|
|||1992||47||1–68||39 C, 8 benign||89||87|
|Continent catheterizable conversion|
|||2001||46||1–57||11 NB, 7 IO, 28 Ex||86|
|||2000||50||180 (mean 51.6)||31 MM, 6 Ex, 2 PB, 2 PUV, 10 other||98|
|||1997||50||62.4||34 Ex, 23 Nel, 13 C, 4 other congenital||82|
Other complications, e.g. pyelonephritis and stones, are much less common. In a large series of conventional ureterosigmoidostomy, Wear and Barquin  reported that 57% of patients had clinical pyelonephritis at least once. There was only one case in the present series and continuing experience suggests that this complication is rare. Hyperchloraemic acidosis is common (as in any intestinal urinary reservoir ) but is easily managed with sodium bicarbonate. The follow-up in these patients is important; we use a protocol slightly modified from that which we use for all patients having lower urinary tract reconstructions with bowel .
There were no uretero-colonic anastomotic carcinomas in the present series with up to 9 years of follow-up, but the earliest previously recorded case was at 10 years. We would expect the incidence to be the same as that for a conventional ureterosigmoidostomy (about 24% of patients at 20 years ).
The question then arises as to which, if any, reconstruction is the best; it would seem self-evident that to void through the urethra would be the most desirable. However, patients with either an orthotopic neobladder after cystectomy or a reconstructed bladder neck with exstrophy do not void in the accepted sense of the word. They empty a urinary reservoir by a combination of Valsalva manoeuvre, straining and abdominal compression, with greater or lesser success. Day and night continence is 78–89% with an orthotopic neobladder, compared to 10 of 12 continent in the present group A patients. Up to 30% of patients may require clean intermittent self-catheterization after 5 years. Likewise, continent reconstruction of an exstrophy bladder is a surgical triumph which may require many revisions. On present evidence the repair may not be durable, with only a third of patients maintaining their success into their third decade of life .
In patients with bladder cancer the risk of recurrence of cancer in the urethra after cystectomy and formation of an orthotopic neobladder is debatable. In early series of cystectomy and ileal conduit, an incidence of 27% after 10 years was reported . With orthotopic neobladders the risk seems to be lower, at ≈ 2%, but with a shorter follow-up of ≈ 4 years . Patients having a cystectomy will be informed of this risk and, if they find it unacceptable, may find a urethrectomy and Mainz II to be a reasonable alternative. A Mainz II should not be used after pelvic radiotherapy or with a history of anorectal dysfunction (we test the anus by filling the rectum with 500 mL of porridge).
Continent suprapubic diversions, particularly the Mitrofanoff, have been widely used for > 25 years. Continence is at the price of life-long dependence on clean intermittent self-catheterization, but incontinence does occur and with various other complications; the rate of re-operation is 33%. The levels of continence in this group (82–98%) is comparable with 10 of 12 in the present group A; overall, the complication rate seems to be higher than in other forms of reconstruction (Table 2).
The major long-term complication with ureterosigmoidostomy is the risk of anastomotic neoplasia. Because of the long delay in development, it may be ignored in all but the youngest patients with bladder cancer. Interestingly, even patients who have had this condition appear to retain their enthusiasm for rectal voiding compared with the alternatives.
Because the Mainz II requires no appliances (bags or catheters) it is particularly suitable for patients from poor countries. The main complication of hyperchloraemic acidosis is cheap to investigate and treat.
We have had experience of all of the available methods of lower urinary tract replacement. The long-term outcomes of the Mainz II seem to be little different from those of other reconstructions; none are ideal, as each has its own idiosyncratic advantages and complications, and there seems little to choose amongst them. We include the Mainz II in the canon of alternatives offered to patients requiring lower tract reconstruction.