Urinary Diversion: Anastomosis of the Ureters into a Sigmoid Pouch and End-to-Side Sigmoidorectostomy

0022-534 7/85/1333-0391$02.00/0 Vol. 133, March

THE JOURNAL OF UROLOGY

Copyright© 1985 by The Williams & Wilkins Co.

Printed in U.S.A.

URINARY DIVERSION: ANASTOMOSIS OF THE URETERS INTO A SIGMOID POUCH AND END-TO-SIDE SIGMOIDORECTOSTOMY SADAO KAMIDONO, YOSHINORI ODA, GAKU HAMAMI, KOJI HIKOSAKA, NOBUO KATAOKA AND JOJI ISHIGAMI From the Department of Urology, Kobe University School of Medicine, Kobe, Japan

ABSTRACT

From 1978 to 1982 bilateral ureterorectostomy and end-to-side sigmoidorectostomy were done following cystectomy for carcinoma of the bladder in 7 patients. No patient had recurrent pyelonephritis or ureterointestinal obstruction. Rectography showed the absence of rectoureteral reflux of contrast medium but rectosigmoid reflux appeared after injection of more than 150 ml. opaque solution. Adequate alkali therapy was performed in 2 patients with hyperchloremic acidosis. While there were some problems concerning the quality of urinary and fecal control achieved in our patients they were no worse than those of ordinary ureterosigmoidostomy. This operation might be recommended for patients in whom a collection appliance is unacceptable. However, before this surgical procedure can be performed the fact that the end result in terms of fecal urinary continence is unpredictable must be explained thoroughly to the patient. Improvements in surgical techniques, particularly in the prevention of coloureteral reflux and stenosis, and efficient management of metabolic disorders have eliminated many of the problems that had caused the initial disrepute of ureterosigmoidostomy.1-3 In view of the enormous advantage of avoiding a stoma ureterosigmoidostomy has been considered for urinary diversion in a selected group of children.3 In 1977 Goodwin and Scardino suggested that the ureterosigmoidostomy procedure is "due for a renaissance". 1 On the other hand, patients who have undergone the procedure always experience some anal leakage of malodorous fecal urine, usually at night or when passing gas. 4 Furthermore, McConnell and Stewart contended that since the quality of urinary control following the procedure usually was poor it should be abandoned largely in favor of the conduit type procedure. 5 However, interest has been expressed in bilateral ureterorectostomy with end-to-side sigmoidorectostomy (one method of urinary diversion free of an appliance) and about 30 cases have been reported to date. 6 - 11 While Gentil and Shahbender reported that the postoperative clinical course of several patients undergoing this procedure was almost identical to that of patients undergoing ordinary ureterosigmoidostomy, 7 Modelski, 8 and Charghi and associates10 reported that it yielded the same advantages as the Gersuny-Johnson-Lowsley method12 without some of the drawbacks of the latter, and claimed that the procedure was proof against urinary and fecal incontinence. In this regard, on the basis of experimental and clinical observations, Leiter and Brendler also suggested that the complications following classical ureterosigmoidostomy might be eliminated by this operation. 9 Their study inspired us to use this procedure in patients who claimed that for social, emotional or economic reasons a urinary diversion with a wet stoma following total cystectomy was unacceptable. We present the detailed results concerning the quality of urinary and fecal control following this procedure based on our clinical experience. MATERIALS AND METHODS

Operative technique. The operation was performed first on a corpse in 1909 by Descomps. 6 Since this method has been reported in detail previously9• 10 we describe only a few modifications. After cystectomy the rectosigmoid is severed between Accepted for publication October 12, 1984. 391

intestinal clamps at the level of the sacral promontory. The distal sigmoid stump is closed and a rectosigmoid pouch is fixed to the promontory of the sacrum. The ureters are implanted in the rectosigmoid pouch by the open transcolonic method, and the anterior rectosigmoid incision is closed. Subsequently, an end-to-side anastomosis of the sigmoidal stump to the distal segment of the rectum is made at the level of the peritoneal reflection (fig. 1). The sigmoidorectostomy is not extraperitonealized. Patients. Between February 1978 and August 1982, 7 patients underwent this procedure following cystectomy for carcinoma of the bladder (table 1). Laboratory examinations, including blood urea nitrogen (BUN), creatinine, serum chloride, potassium, blood pH, base excess and blood gasses, were monitored closely initially and at 3-month intervals. An excretory urogram (IVP) was performed in each case at 6-month intervals and ascending rectography with 200 to 300 ml. contrast medium was done to detect rectoureteral reflux. Detailed information on the life style of the 7 patients 1 year postoperatively was obtained, especially concerning the frequency of bowel evacuation, anal incontinence and social consequences. Patients received instructions about fluid intake or dietary restrictions. RESULTS

Postoperative complications included wound dehiscence in 1 patient, pelvic abscess in 1, gastric hemorrhage in 1 and a febrile attack (pyelonephritis?) in 1. No fecal-urinary fistulas, ileus or operative mortality was seen in the early postoperative period. The IVP at 1 year was normal in all patients (fig. 2) and there was no clinical evidence of pyelonephritis. No air was observed in the upper urinary tract and in 6 patients the rectogram performed under 40 cm. gravity pressure revealed no rectoureteral reflux. Sigmoid reflux began after injection of 150 ml. contrast medium (fig. 3). The supine renogram (131iodinehippurate) was within normal limits in patients 3 to 7. Patients 1, 4, 6 and 7, in whom sigmoidoscopy was performed 18 months postoperatively, had no malignant changes of the mucosal epithelium in the rectosigmoid pouch. In 2 patients with immediate postoperative hyperchloremic acidosis adequate treatment consisted of alkali administered in the form of sodium bicarbonate. Figure 4 shows the base excess and blood pH measurements of these 2 patients after cessation

392

KAMIDONO AND ASSOCIATES

of alkali therapy. The values for blood gasses and plasma electrolytes in the 5 patients who did not receive alkali therapy always were within approximately normal limits (fig. 4) and none showed any clinical sign of hyperchloremic acidosis, such as weakness, nausea and vomiting. All 7 patients showed normal serum creatinine values but a mild increase in BUN was noted in 4. During the daytime (18 hours) the frequency of bowel evacuation was once every 1.5 to 2.5 hours in 6 patients and once every 3.5 hours in 1. Also, 6 patients were continent, while 1 suffered from occasional daytime incontinence. At night 5 patients awakened less than 3 times per night for bowel evacuation, and of 3 who wet the sheets with fecal urine 2 required incontinence pads (tables 1 and 2). The type of evacuation in 3 patients always was a watery diarrhea, compared to formed stool along with urine in the remaining 4. In addition, 2 patients could pass gas safely (table 2). Despite varying degrees of inconvenience in bowel evacuation all 7 patients were able to return to their previous life styles and all returned to work: 5 (71 per cent) returned to their previous occupation and 2 (29 per cent) performed lighter work.

Presently, 4 of the 7 patients are alive. Two patients died of recurrent tumor at 55 and 23 months, respectively, and 1 was lost to followup 2 years postoperatively. Of the 4 surviving patients 2 are able to hold the fecal urine for 1 hour after a sense of fullness develops, while 2 must evacuate the bowels within 5 minutes. DISCUSSION

A unique feature of this procedure is that, unlike ordinary ureterosigmoidostomy, 2 partially separated reservoirs are con-

FIG. 2. A, preoperative IVP in patient 5 demonstrates normal upper urinary tract. B, IVP 1 year postoperatively shows normal upper urinary tract and filling of rectosigmoid pouch.

,,,,.ureters ......

Rectosigmoid

-+

pouch

FIG. 1. Schematic drawing of completed operative procedure. Ureters are anastomosed into partially isolated rectosigmoid pouch. Proximal sigmoid is reimplanted to rectum. anchoring suture.

*,

TABLE

FIG. 3. Rectogram with injection of 190 ml. opaque solution under 40 cm. gravity pressure shows rectosigmoid reflux but no rectoureteral reflux.

1. Summary of patient data-frequency and leakage of bowel evacuation

Daytime

Bladder Ca

Nighttime

Adequate Alkali Therapy

Frequency of Bowel Evacuation

Leakage (feces/urine)

Sleep Disruptions per Night (6 hrs.)

6 yrs.

Yes

7-8

Minimal

0-1

Pt.-Pt.-Sex-Age No.

Stage

1-TT-M-52

pTl (m)

2-KK-M-54 3-TH-M-62 4-HM-M-41 5-SY-M-60

pTis pTis pT2(m) pTl (m)

II II III III

2 yrs.* 4 yrs. 7 mos. t 2 yrs. 1 yr. 11 mos.t

No Yes No No

7-8 8-10 9-11 10-12

No No No No

0-1 2-3 1-2 2-3

6-HO-F-51 7-MO-M-30

pTl (m) pT2 (m)

II III

1 yr. 7 mos. 1 yr. 6 mos.

No No

4-5 7-8

No No

1-2 0-2

* Lost to followup. t Died with tumor.

Grade

Interval After Diversion

Leakage (feces/urine) Usually wets sheets No Minimal No Often wets sheets No No

URINARY DIVERSION WITH SIGMOID POUCH AND SIGM0IDORECT0STO!V1Y BUN

Base Excess (Blood Ga.s)

393

Serum Cl

mEq/S

mEc,,/£

+2

118

-2

110

-4 -6 100

-8 !!6

~~~-
1 1 operation

operation

operation

Serum K

Serum Creatinine

pH

"Old.I

mEq/i

7 .45

5

. •........ ...... • ............... .. . .. ........ .

••P•••••••••··••••"•·•·•·?rHt•••••••••
.·_·.·.··_.··.--. 7.·.·_.· ... ·.·.- .•..•..•..•.•. ::·.••.•.·. ···:...· ·.::\:::.:.:·:.::::: ··:f<:::<:>:<: . ·.. ·... · ;·'· ... .

.·.·-·-

~~~~~-6mos 1 2 4 6 yrs t 1 operlltion

operation

operation

FIG. 4. Blood gas analysis, BUN, serum creatinine and serum electrolyte values in 7 patients postoperatively (shaded area indicates normal range). ~--~. no alkali therapy (patients 2 and 4 to 7). 6.--6., adequate alkali therapy (patients 1 and 3). *, died with tumor. t, lost to followup.

TABLE

2. Problems of bowel evacuation in 1 year postoperatively

Use incontinence pads at night Separate feces/urine Pass flatus safely Diet modification required

7

patients

Yes

No

Occasional

2* 1t 2:j: 1§

4 3 4

1 3 1

6

0

* Patients 1 and 5.

t Patient 2. :j: Patients 2 and 6. § Patient 6. structed: 1 for urine and 1 for fecal matter. 10 Therefore, discussion of the advantages of this surgical procedure is important. Hyperchloremic acidosis occurred in 60 to 80 per cent of the patients undergoing ordinary ureterosigmoidostomy without alkali therapy. 13 In comparison, the incidence of hyperchloremic acidosis with the present procedure was reported to be 33 per cent (2 of 6 patients) by Charghi and associates, 10 0 per cent (0 of 8) by Modelski 8 and 29 per cent (2 of 7) in our study. These good results may have been caused by the partially separated rectosigmoid pouch. However, sigmoid reflux was recognized by ascending rectography9 •10 (fig. 3) and, as in the cases of ordinary ureterosigmoidostomy, 10 some reabsorption of urine was reflected by an increase in BUN (fig. 4). This fact means that it is difficult to expect the same infrequency of electrolyte disturbance and metabolic acidosis with this urinary diversion as in patients with rectosigmoid bladders. 14 Of our 5 patients in whom metabolic acidosis was not recognized 4 reported frequent bowel voiding, which might have helped minimize biochemical complications. 5 The good results of ureterointestinal anastomosis in this surgical procedure 1 • 2 have been suggested to be related to the low intraluminal pressure of the partially separated rectosigmoid pouch. 8 This advantage cannot be anticipated by ureterosigmoidostomies without proximal colostomy. The firm attachment of the rectosigmoid pouch to the sacral promontory, which is performed easily with this method, means that the weight of the pouch is not brought to bear on the anastomosis, so that the method is extremely effective. In ordinary ureterosigmoidostomy the mesosigmoid is tacked to the promontory

for the same purpose, reducing the mobility of the sigmoid. 3 The success of the ureterointestinal anastomosis in the former procedure depends upon the aforementioned technique, as well as the healthy status of the ureteral and rectosigmoid tissue, and the fact that irradiation has not been performed. 3 Frequent voiding of fecal urine during the day, and sleep disruption and leakage owing to bowel evacuation at night were problems of this procedure. Patient 6, who had the fewest problems among all 7 patients, was troubled by sleep disruption. Considering that a small number of patients treated by this procedure were able to pass gas safely and that the leakage of fecal urine was relatively mild, this method can be considered to yield slight advantages over ordinary ureterosigmoidostomy. 4· 5 However, one must remember that in these types of urinary diversion, the results concerning frequency of bowel evacuation and the success in regard to leakage depend largely upon the fluid intake of the patient and care for the prevention of escape of fecal urine. Of the surviving 4 patients in whom followup was possible, although there were varying degrees of complications, none required a stoma type diversion. Our patients did not want a stoma from the beginning and, perhaps because they are grateful to survive after malignant neoplastic disease, they generally are uncomplaining and seem to be satisfied with the present situation. 4 "5 The incidence of carcinoma of the sigmoid colon in patients with ureterosigmoidostomy is about 280 to 550 times that of the general population. 15 Patients undergoing ureterosigmoidostomy have a 5 per cent risk of colon cancer developing at the implant site, with the shortest interval being 2 years. 16• 17 The threat of carcinoma also is present with this method. Therefore, monthly examination for occult blood is performed and sigmoidoscopy is done if the result is positive. In addition, followup IVPs and barium enemas are necessary .16· 18 Because the rectosigmoid pouch is separated partially examination of the implant site, which is the most common location for the development of carcinoma, is easy. Rabinovitch 19 suggested that the staged ureterosigmoidostomy of Hendren 20 is promising because the anastomotic site is excluded potentially from the fecal stream, thereby reducing the causative factor in tumorigenesis. As a result, one can hope

394

KAMIDONO AND ASSOCIATES

that this method of urinary diversion with a partially separated reservoir may involve a reduced danger for carcinoma to develop. Because of the small number of patients and the relatively short followup, we cannot evaluate fully this method of urinary diversion by anastomosis of ureters in the rectosigmoid pouch. However, our results do indicate that this method should not be abandoned completely and that it is not less effective than ordinary ureterosigmoidostomy. We believe that our results with this method are promising and plan to evaluate it further clinically. REFERENCES 1. Goodwin, W. E. and Scardino, P. T.: Ureterosigmoidostomy. J. Urol., 118: 169, 1977. 2. Hohenfellner, R.: Ureterosigmoidostomy. In: Surgical Pediatric

3. 4. 5. 6. 7. 8. 9. 10. 11.

12.

13.

Urology. Edited by H. B. Eckstein, R. Hohenfellner and D. I. Williams. Philadelphia: W. B. Saunders Co., sect. 6, chapt. 6.5, pp. 354-361, 1977. Marberger, M. and Straub, E.: Ureterosigmoidostomy in children. In: Urinary Diversion. Edited by M. H. Ashken. Berlin: Springer Verlag, chapt. 3, pp. 59-74, 1982. Macfarlane, M. T., Lattimer, J. K. and Hensle, T. W.: The unheralded hazard ofureterosigmoidostomy. Pediatrics, 64: 668, 1979. McConnell, J. B. and Stewart, W. K.: The long-term management and social consequences of ureterosigmoid anastomosis. Brit. J. Urol., 47: 607, 1975. Descomps, P.: Abouchement ureteral dans le rectum exclu. Ureterorectostomie haute terminate apres sigmo1do-rectostomie basse termino-laterale. Arch. Gen. de Chir., 4: 892, 1909. Gentil, F. and Shahbender, S.: The use of segments of small intestine in surgery. Surg., Gynec. & Obst., 109: 417, 1959. Modelski, W.: The transplantation of the ureters into the partially excluded rectum. J. Ural., 87: 122, 1962. Leiter, E. and Brendler, H.: Method of urinary diversion which preserves continence: description of surgical technique and postoperative electrolyte study. J. Ural., 92: 37, 1964. Charghi, A., Charbonneau, J., Cholette, J. P. and Gauthier, G. E.: A method of urinary diversion by anastomosis of the ureters into a sigmoid pouch. J. Urol., 94: 376, 1965. Hendren, W. H.: Ureterocolic diversion of urine: management of some difficult problems. J. Urol., 129: 719, 1983. Lowsley, 0. S. and Johnson, T. H.: A new operation for creation of an artificial bladder with voluntary control of urine and feces. J. Urol., 73: 83, 1955. McConnell, J. B., Murison, J. and Stewart, W. K.: The role of the colon in the pathogenesis of hyperchloraemic acidosis in ureter-

osigmoid anastomosis. Clin. Sci., 57: 305, 1979. 14. Ghoneim, M. A.: The rectosigmoid bladder for urinary diversion. Brit. J. Urol., 42: 429, 1970. 15. Parsons, C. D., Thomas, M. H. and Garrett, R. A.: Colonic adeno16.

17.

18. 19. 20.

carcinoma: a delayed complication of ureterosigmoidostomy. J. Urol., 118: 31, 1977. Leadbetter, G. W., Jr., Zickerman, P. and Pierce, E.: Ureterosigmoidostomy and carcinoma of the colon. J. Urol., 121: 732, 1979. Sooriyaarachchi, G. S., Johnson, R. 0. and Carbone, P. P.: Neoplasms of the large bowel following ureterosigmoidostomy. Arch. Surg., 112: 1174, 1977. Uehling, D. T., Starling, J. R. and Gilchrist, K. W.: Surveillance colonoscopy after ureterosigmoidostomy. J. Urol., 127: 34, 1982. Rabinovitch, H. H.: Ureterosigmoidostomy in children-revival or demise? J. Ural., 124: 552, 1980. Hendren, W. H.: Exstrophy of the bladder-an alternative method of management. J. Urol., 115: 195, 1976. EDITORIAL COMMENTS

This article is a clear, succinct exposition of an incompletely explored method of continent urinary diversion that certainly is far simpler than a Koch pouch, and may decrease the long-term incidence of pyelonephritis and hyperchloremic acidosis to acceptable levels. Unfortunately, the series is too small and the followup is too short to allow definite conclusions. Hopefully, other investigators will be stimulated to study this interesting method of diversion further. I point out 3 considerations that are important: 1) preoperative radiation may make the procedure exceedingly efficient, 2) the sigmoidorectostomy is not a technically easy anastomosis and 3) the idea of an anastomosis in the immediate proximity of the primary malignancy does not satisfy the demands of classic cancer surgery. E.L. A number of variations upon the basic operation of ureterosigmoidostomy have been proposed ostensibly for the purpose of avoiding or minimizing the traditional shortcomings of this form of diversion. I have never been convinced that any of them actually succeeded in achieving this laudable goal and, indeed, many of them seem only to make matters worse. One of the problems with the isolated rectal pouch as a urinary reservoir is that when the operation is a success, that is when the urine is isolated effectively from the rest of the fecal stream, it often is contained within a volume so restricted that when the segment contracts the patient experiences marked urgency and even urinary incontinence. In 1 patient with urinary incontinence after a successful Heitz-Boyer procedure we found it necessary to incise the septum between the 2 pouches to permit the urine to reflux back into the remainder of the colon to restore urinary continence. T.D.A.