Nonrefluxing or Permanent

Colon Conduit for Temporary Urinary Diversion in Children By W. Hardy

Hendren

I

N THE PAST 15 yr it has become generally appreciated that vesicoureteral reflux is the most common cause for pyelonephritis in childhood. Indications for surgical correction of reflux have been defined, and it is clear that surgical correction of reflux in appropriate cases can stop the tendency to pyelonephritis and further renal damage. Since 1950, when Bricker’ described the ileal conduit as a means of urinary diversion after pelvic evisceration for cancer, this operation has found wide use in all age groups. In children it has been used as a means for urinary diversion for neurogenic bladder, for urinary tracts with severe damage from obstructive disease, and for the occasional youngster with cancer requiring cystectomy. The classic ileal conduit operation allows free reflux to the kidneys from the bowel conduit which empties on the abdominal wall. Indeed, free reflux on “loopogram” study has been considered reassuring that the ureters are draining freely into the loop. Since bacilluria is present frequently in ileal conduits with apparently satisfactory drainage. it is not surprising that there is often low-grade bacilluria with upper-tract changes of pyelonephritis, stone formation, and gradual deterioration, just as one sees In children with vesicoureteral reflux. Several recent reports ’ 5 have emphasized that in large numbers of children followed for a decade or longer with ileal conduits a considerable proportion have infection and deterioration. In a series of 150 children with these diversions, Smith4 found upper tract changes in lo”,,, whose upper tracts were normal at the time of diversion. Our own experience has been similar. Although some children do extremely well with an ileal diversion, others develop gradual dilatation of the upper tracts without apparent obstruction at the stoma or the ureteroileal anastomosis. Others have persistent bacilluria. For these reasons in March of 1971, we abandoned performing the classic ileal conduit operation in children, in favor of constructing a conduit of sigmoid colon, with nonrefluxing ureterosigmoid anastomoses. This paper will describe an experience with 21 cases. The indications for operation were: three new cases of bladder exstrophy; I:! patients with bladder exstrophy previously diverted with an ileal loop: three new patients with neurogenic bladder; two with neurogenic bladders previously diverted by ileal loop; and one patient who had this type of diversion at the time of anterior pelvic exenteration for a large sarcoma arising in his prostate.

From the Division of Pediatric 5Lrger.v. Massachusetts General Hospital. and Deparrment o/ Surgerv. Harvard Medical School, Bosron. Mass. Presented before the 23rd Annual Meeting of the American Academv of Pediatrics. Surgical Secrion, San Francisco, Calif. October 19-22, 1974. {BAddress for reprint requests: W. Hard.v Hendren. M.D., Professor of Surgery and Director oJ Pediatric, Surger,v. Massachuserts General Hospital. Bosron. Mass. 021 14. 10 1975 hv Grune & Stratton. Inc. Journal

of Pediofric

Surgery,

Vol. 10, No. 3 (June),

1975

38:

W. HARDY

382

HENDREN

Distal end of

aortic

bifurcation Colon to colon anastomosis

Fig. 1. (A) A segment of sigmoid colon is selected, with oppropriote blood supply. The mesentery of the conduit should be maintained as brood os possible, incising the mesentery only enough to give sufficient mobility to anchor the conduit at the desired location. The bowel segment should be of ample length to allow for some shortening from contraction of the toenio, especially when it is contemplated thot the conduit will later be token down from the abdominal wall, resecting the stoma, and rejoining it end to side into the colon. (B) The proximal

NONREFLUXING

COLON

CONDUIT

TECHNIQUE

383

FOR OPERATION

Technical aspects of the operation and its variations, are shown in Figs. l-6. Preoperatively the bowel is thoroughly prepared by cleansing enemas. liquid diet for 3-4 days, and neomycin bk mouth and by rectal irrigation 1 day preoperatively. Just before laparotomy a large rectal tube is inserted into the rectum, and attached to a long tube through which the circulating nurse can irrigate and aspirate neomycin solution. in case the colon is found not ideally prepared. When the colon is filled with neomycin solution by this arrangement, the surgeon can empty it and any residual contents by milking the intestine downward manually. A segment of sigmoid is selected (Fig. 1A) with appropriate blood supply. Generally, the splenlc Rexure is mobilized to permit end-to-end restoration of continuity without tension of descending colon to rectosigmoid. For this anastomosis we prefer two layers of interrupted arterial silk sutures on atraumatic needles. No problem has been encountered with these anastomoses. The proximal end of the sigmoid conduit is closed in two inverting layers, using chromic catgut, running the first layer and interrupting the second. Nonabsorbable suture material is not used for this closure, lest a suture work its way into the lumen of the conduit and provide a nidus for stone formation. The length of sigmoid should be amply long. In constructing ileal conduits for permanent urinary diversion, it has been a routine practice to make as short a loop as possible, to minimize the absorptive surface of bowel, particularly in patients with reduced renal function. The colon conduit has a tendency to contract in length as one is performing these operations. and so it is safer to select a fairly generous length initially, planning to trim off what is not needed when the stoma is fashioned. We recall one case in which the loop was cut too short in an obese patient where length was not sufficient to put the stoma at an ideal level on the abdominal wall. In those patients in whom the plan will be to eventually take down the stoma to anastomose the conduit back into the colon stream (i.e., patients with bladder exstrophy, or after cystectomy for cancer) the conduit is deliberately made a little longer than in a patient for whom the conduit IS deemed a permanent diversion. This is done in order to allow resection of the stoma before joining the conduit to the colon. In resecting the stoma, it is important to know the location of the ureteral orifices. For this reason closure of the seromuscular tunnels is performed with fine nonabsorbable sutures which will remain visible in the future. As an additional precaution one can actually look into the conduit with a panendoscope pre- or intraoperatively to determine the location of the ureteral orifices, sometimes with the help of indigocarmine given intravenously. Also one can look down into the colon loop, after resecting the stoma, by placing small retractors inside the lumen of the conduit, to directly visualize the ureteral orifices. The conduit is then rotated clockwise 180 degrees as shown in Fig. 1B. so that the proximal. closed end comes to lie at the aortic bifurcation. It is anchored there with several sutures. In one case the sigmoid colon was not suitable and so we used a segment of transverse colon based on its blood supply of midcolic artery. In some patients who had undergone a prior ileal loop, the ureters were not long enough to reach the level of the aortic bifurcation, and so the end of the conduit was anchored at a higher level, sometimes close to the lower pole of a kidney. In Fig. IC is shown the usual anatomic relationship after fashioning the conduit, anchoring it in place, reestablishing continuity of the left colon, and mobilizing the ureters which pass close to the base of the conduit. In mobilizing the ureters, their full length should be preserved, ligating them flush with the bladder, and carefully preserving all of the periureteral adventitia which is so important with respect to maintaining a good blood supply to the ureter. A long tunnel is then created in which to lay the ureter to provide a nonrefluxing type of anastomosis. First, saline is infiltrated through a 25gauge needle into the seromuscular layer of a taenia (Fig. 2A): this facilitates making a longitudinal incision without entering the mucosa. The lateral edge of this incision is then undermined, bluntly dissecting away the colonic mucosa from the overlying

end of the conduit is closed with two layers of chromic catgut. The conduit is rotated 180’ clockwise and is anchored at the aortic bifurcation above the sacral promentory. If one or both ureters are short, the point at which the base of the loop is anchored may vary, to be at a higher level or in a retroperitoneal gutter. (C) Colon continuity is reestablished, usually taking down the splenic Aexure to accomplish anartomosis without tension. Mesenteric traps are closed. The ureters lie adjacent to the base of the conduit. In mobilizing the ureters, the periureteral advenWia is preserved to insure a good blood supply.

W.

HARDY

A

End of ureter to colon mucosa

Mucosa

C

6-O

lacking

knots outside

suture

See legend facing page.

HENDREN

NONREFLUXING

COLON CONDUll

385

Fig. 2. (A) Saline is injected with a 25.gauge needle to facilitate raising the lateral seromuscular flap without entry into the underlying mucosa. (B) The lateral seromusculor flap is elevated, bringing into view the small blood vessels running from setomuscular layer to the underlying mucosa. This dissection is performed bluntly, and can be facilitated with “peanut” sponge dissection. (C) The ureter is brought through a separate opening in the seromuscular layer and placed beneath the bridging vessels, so that it lies in a well-supported area of the bowel wall. It is anastomosed end to side with the colon mucosa. (D) Details of anastomosis: (1) Ureter is cut straight across, not spatulated, leaving a piece attached to avoid handling it with forceps. (2) The first three sutures of fine chromic catgut are placed, before tying them, to insure accurate approximation. Small forceps spread the lumen of the ureter, without grasping its edge, to minimize trauma. (3) Additional sutures are placed, tying each one as the anastomosis proceeds. (4 and 5) The final three sutures are placed and tied with knots on the outside. (E) Completed conduit with submucosal ureteral tunnels 4-5 cm long. If one anticipates later resecting the stoma to join the conduit to colon, nonabsorbable sutures should be used to close the seromuscular tunnels, for this facilitates their identification at subsequent laparotomy. (F) Completed conduit. No attempt is made to close the space on either side of the conduit, but mesenteric

traps are closed. seromuscular layer. The medial flap is not undermined. for if that were done from both sides at the same level of the colon (since both ureters will enter at the same level), necrosis of the medlal flap would likely occur. Separation of the mucosa from the seromuscular layer proceeds remarkably easily using a combination of blunt dissection, and small “peanut” gauze sponges. The vessels bridging from the bowel wall to the mucosa can be preserved with care, and a tunnel can be made beneath them in which to place the ureter(C). This maneuver places the ureter near the mesentery of the colon conduit where it is possibly better supported. (Although we have placed the ureter beneath the vessels, as shown in Fig. 2C, in most cases, in some it has been laid on top of those bridging vessels, requiring a lesser degree of mobilization of the lateral flap. I do not know whether one position is indeed better than the other but have thought that. theoretically, beneath the vessels might be better). The ureter is anastomosed end-to-side to the mucosa at the end of the tunnel. Details are shown in Fig. 2D. With a suture in the tip of the ureter, it is partially transected, leaving a bit attached for traction. This avoids possible injury to the delicate wall of the ureter from handling it with forceps. It is not necessary to spatulate the ureter, since, in our experience, a carefully performed mucosa-to-mucosa anastomosis with appropriately fine interrupted suture material will not stenose. The anastomosis is performed with 5-O and 6-O chromic catgut, placing the first three sutures before tying any of them, to give very accurate approximation of the ureter to the colon.

.^

“”

..I

.-

“”

,. .__

.

“. .”

Fig. 3. (A) Tapering of the dilated ureter. Its longitudinal intramural blood supply should be preserved; This usually is located on the medial aspect of the ureter. A tapered ureter is like a pedicle flap of any tissue, i.e., its base must not be too narrow with respect to its length. If the ureter is made too narrow, its blood supply will be in jeopardy, and ischemic necrosis will result. A soft, malleable plastic catheter is used to drain the ureter for lo-12 days. A no. 5 or 8 plastic feeding catheter is well suited for this. (3) Transureteroureterostomy when a ureter is too short (used in three cores). (C) Subsequent implantation of conduit into the colon by endto-side technique, using two layers of chromic catgut (running technique for the inner layer, to insure a watertight anastomosis, and interrupted sutures for the seromusculor layer).

NONREFLUXING

Additional

COLON CONDUIT

interrupted

sutures

are added

387

until the anastomosis

is nearly

completed.

The last three

sutures are placed so that the knots will be on the outside when they are tied. A drainage stent catheter is not used unless the ureter is tapered (as shown in Fig. 3A). The seromuscular layer is then closed with interrupted sutures, and the second ureter is joined to the conduit (ZE). The two ureters lie parallel about 2 cm apart. entering at the same level of the conduit. The trap beneath the mesosigmoid is closed, and a stoma is fashioned. In a patient previously not operated upon, this is generally in the left abdomen on the belt line. In a patient with a prior ileal conduit it may be at the site of a previous ileal stoma. If a ureter is dilated. it is longitudinally resected. just as in a megaureter repair6 as shown in Fig. 3A. A no. 5 or no. 8 plastic feeding catheter is used for temporary drainage for lo--l2 days until this longitudinal suture line is healed. In our 21 patients, II ureters were tapered in nine patients. It is important to preserve the vascularity of the ureter by not making it too narrow. No problems

were encountered

as a result of ureteral

tapering.

In patients with a previous ileal conduit, one or both ureters may be too short to accomplish the usual conduit construction shown in Fig. 2. An alternative is anastomosis of one ureter to the conduit. with transureteroureterostomy to accomplish drainage of the other ureter, as shown in Fig. 3B. If the ureter to be joined to the conduit is somewhat short. the conduit can be placed moderately high in the posterior peritoneal gutter to accomplish a ureterosigmoid anastomosis of adequate length without tension. Transureteroureterostomy was used for one ureter in three cases, Implantation of the conduit at a later stage into the colon is appropriate in some cases (Fig. 3C). That has been accomplished in nine of these 21 patients, now 6-27 mo postoperative. The duration between construction of the conduit and its anastomosis to the colon was 9 mo in one, but a year or more in the other eight (average 20 mo). In new bladder-exstrophy patients, the conduit should not be joined to the colon until they are old enough to have a reasonable chance for learning control of liquid by rectum. In patients whose conduits include ureteral tapering, transureteroureterostomy, etc., a longer term of follow-up is desirable to make certain that all is well in terms of drainage before risking implantation of the conduit into the unsterile colon. We anticipate anastomosing the conduit in several others of these patients in the future (only in patients with normal anal sphincter and satisfactory upper tracts). An alternative method of ureterosigmoidostomy is shown in Fig. 4. devised by Mathisen of Oslo, Norway. This has found great favor among surgeons of Europe as an excellent method of ureterosigmoid diversion. In this series, we used it for five ureters, in circumstances when the ureter did not seem well suited for the usual long ureterosigmoid tunnel (it was too short. or thick walled). A nipple is fashioned which hangs in the lumen of the conduit, and prevents reflux. In our experience It is best to fashion the periureteral flap of colon a little at a time as the anastomosls progresses, in order not to make it too narrow; this flap must be amply wide to wrap around the ureter without compressing it. After the nipple is constructed, closure is carried beyond the base of the ureterocolic junction to effect closure of the colon wall. Of the five cases in which a Mathisen nipple was used, one patient had a solitary kidney, three had transureteroureterostomy of the other ureter, and one had a usual tunnel implant of the other ureter which had better length and smaller caliber. COMPLICATIONS

AND

FOLLOW-UP

OBSERVATIONS

Complications Intestinal obstruction occurred in four cases, each from simple adhesions. In three this occurred after the colon conduit procedure and in one after the subsequent operation to join the conduit to the colon. There were no other serious problems. No ureters were obstructed. Upper tracts which were normal preoperatively, have remained normal. Upper tracts which were dilated preoperatively have improved. Loopograms Retrograde filling of the conduits in 13 patients showed no low-pressure reflux in any case (postoperative loopograms have not yet been done on the other

W. HARDY

388

Fig. 4.

Mathiren

method of ureterosigmoidortomy,

used in five cases in which

not well suited for tunneling. All three cases of transureteral patients in whom the other ureter was managed by the Mathiren

ureterostomy were nipple anastomorir.

HENDREN

a ureter performed

was in

eight patients). lntraconduit pressures were measured during these loopogram studies. No reflux occurred in six patients, even at pressures of over 60 cm of water. In the seven cases in which reflux could be demonstrated, the lowest pressure at which it was seen was 35 cm of water, and the average pressure was 50 cm. Urinar,v Infection There has been no clinically symptomatic urinary infection in any of the 21 patients. Cultures taken from the conduits by catheter technique were either sterile or had low colony counts. There has been no evidence of upper tract change by intravenous pyelography in any patient, but it should be emphasized that the follow-up is relatively short in duration, since the first patient was operated upon by this technique in March of 197 1.

NONREFLUXING

COLON

CONDUIT

380

Patients With Anastomosis of Conduit to Colon In all nine patients whose conduits were later joined to the colon, this was done only after at least 9 mo had passed after construction of the conduit. This time allowed determining that the upper tracts were draining normally, there was no low-pressure reflux, and no tendency to urinary infection. The nine patients whose conduits were joined to the colon have been followed at close intervals to see if there is any evidence of ascending pyelonephritis. The IVP has been stable in each. In none has gas been observed in the upper urinary tracts. None has had unexplained fever, flank pain, or other stigmata of pyelonephritis. It should be pointed out, however, that clinical signs and symptoms are often absent in such patients. Long-term follow-up will be needed. It is not possible to obtain urine for culture in these patients, except by percutaneous needle aspiration of the kidney. This does not seem justified in a clinically well patient. Blood chemistries were obtained on this group of patients; there has been no tendency to significant hyperchloremic acidosis. In this regard we encourage all of these patients to drink a liberal amount of water and to empty the colon every 3 hr on a definite schedule to reduce as much as possible resorption of solute from the colon. A generous flow of dilute urine is additional protection against ascending pyelonephritis, coupled with the presence of an effective valve mechanism at the ureterocolic junction. All patients are cautioned that if gastroenteritis or some other intercurrent illness should prevent satisfactory fluid intake, they must report immediately for temporary intravenous fluid maintenance. If a patient with ureterosigmoidostomy becomes dehydrated, resorption of a concentrated urine from the colon in the face of dehydration can lead rapidly to uremia, even with normal kidneys. Conversely we have not encountered significant hyperchloremic acidosis in any of these patients, or in a large number of previous patients seen with primary ureterosigmoidostomy for exstrophy of the bladder when: (1) the upper tracts were satisfactory, and (2) a reasonable fluid intake was maintained. The serum chemistries in these patients generally shows a slight increase in serum chloride and slight decrease in serum CO2 levels. Average values in these nine patients were sodium 141, potassium 4.2, chloride 107, CO* 22, BUN 24, and creatinine 0.7 meq/liter. Patients with ureterosigmoidostomy absorb chloride, and secondarily lose bicarbonate, which can lead to secondary acidosis with volume retraction. If the serum CO, falls below 20 meq/liter (normal is 22-28 meq/liter), we prescribe bicarbonate in a dosage of 224 meg per kg of body weight (a 600~mg tablet of sodium bicarbonate contains 7.8 meq of bicarbonate).

Stomas With ileal conduit there is a high incidence of stenosis of the stoma and stomatitis. The colon conduit stomas have uniformly been trouble-free. The 12 patients who had lived previously with an ileal stoma all remarked that the colon conduit stoma is cleaner, bleeds less easily, and they do not tend to stenose. None required dilatation or revision.

Control of Liquid Urine by Rectum Children with ureterosigmoidostomy seldom have perfect control of urine by rectum until about age 5 or 6 yr. In this group two 4-yr olds currently have

W.

390

HARDY HENDREN

some daytime accidents but are improving. One 4-yr old has good control. Two 6-yr olds have occasional daytime accidents. The remaining patients, who are 8 yr or older, all have normal daytime control. It is common for these patients to have some leakage at night while asleep, even when they reach adulthood. We suggest, therefore, that they reduce their fluid intake in the evening before bedtime and set an alarm to arise to empty the colon at some point during the night. All of these patients were maintained on antimicrobial therapy (sulfisoxazole or nitrofurantoin) for at least 3 mo postoperatively. ILLUSTRATIVE

CASE REPORTS

Case I (Fig. 5). Sarcoma of Prostate Treated by Anferior Pelvic Exenteration, Temporary Colon Conduit Diversion and Later Anastomosis of the Conduit to the Colon P. C., MGH 172-68-61 was referred at age 20 mo with a palpable lower abdominal mass. Intravenous pyelogram showed normal upper tracts, but displacement of the lower ureters laterally. Films of chest and lung bones were negative for metastases. Cystourethroscopy showed grape-like fronds of tissue in the prostatic urethra and bladder trigone consistent with sarcoma botryoides. After bowel preparation, operation was performed August 9, 1971. The mass was radically resected, removing bladder, prostate, membranous urethra, seminal vesicles vasa, and distal ureters. Included with the specimen were adjacent pelvic lymph nodes, which proved negative on subsequent microscopic examination. A nonrefluxing colon urinary diversion was performed. The patient convalesced uneventfully and was discharged 12 days later. The tumor was typical rhabdomyosarcoma. X-ray treatment was not given, because the tumor was thought to be completely resected grossly, and resection margins were negative for tumor, as well as the lymph nodes. Chemotherapy was given for 2 yr, alternating actinomycin D, vincristine. and cytoxan. Frequent rectal examinations were performed to exclude the presence of a palpable recurrence. At the end of 2 yr he was a well child, with no evidence of recurrent disease. Bowel control was normal, at almost age 4 yr, and so it was elected to anastomose the colon conduit into the rectosigmoid. Prior intravenous pyelography had shown a still normal upper urinary tract. Retrograde filling of the conduit showed no reflux, even with pressures over 60 cm of saline. On September 14, 1973, the cutaneous stoma was resected, anastomosing the conduit to the rectosigmoid. The abdomen was negative for tumor. He was discharged on the 8th postoperative day. Two weeks later intestinal obstruction occurred. This was not relieved by nasogastric decompression, and so lysis of adhesions was performed. Since then he has been an entirely well patient, on no medication, now over 3 yr after anterior exenteration and I yr after diversion of the urinary stream into the colon.

Commenl Primary diversion of the ureters into the colon has been performed for many years with anterior pelvic exenteration; however, in a young child who is not yet bowel trained, this would be socially undesirable, leading to several years of combined urine and fecal incontinence. Staging the procedure as was done in this child has several advantages: (I) It permits assessment of the child for possible recurrence of tumor in the pelvis before directing the urine flow to the rectum; (2) It delays diversion of the urine to the rectum until an age at which rectal control of liquid is possible; (3) It permits assessment of the ureterosigmoid anastomoses over a long term to make certain that they drain well, and do not have low-pressure reflux before bringing them into contact with the unsterile bowel content. In our experience if there is obstruction or reflux after ureterosigmoidostomy, severe pyelonephritis occurs quickly with renal damage. Should there be a problem with the ureterosigmoid anastomosis in a diverted

NONREFLUXING

COLON CONDUIT

Fig. 5. (Case 1). (A) IVP 19 mo after temporary colon conduit procedure. Note normal upper tracts. No reflux on loopogram until pressure exceeded 60 cm of water. (6) The patient at age 3 yr with colonic stoma, urine draining into bag. Hair loss secondary to chemotherapy which was continued for 2 yr, prophylacticolly. (C) Introvenous pyelogram 13 mo after anastomosis of conduit to colon. Normal upper tracts. Arrow points to conduit, which empties into rectosigmoid. Note concentration of dye in rectum, but some dye reaching the splenic flexure. (D) The patient at 4 l/2 yr, a well child, on no medications. Fig. 6. (Case 2). (A) IVP preoperatively, age 4 l/2 yr, showing severe hydronephrosis. Cystogrom showed massive reflux from her neurogenic bladder. (6) IVP 11 mo after colon conduit operation, which included tapering of the right ureter. Loopogrom showed no reflux, even at pressures over 60 cm of water. (C) The colon conduit stoma. These stomas are larger and more bulky than ileal stomas, but have been uniformly more satisfactory as viewed by physicions, parents, and in particular by the children who have lived with both ileal and colon conduit stomas. There has been no problem with stomatitis, bleeding, or stenosis to date. (D) The patient, age 6 l/2 yr, a well child. She is maintained on low-dose nitrofurantoin. Fig. 7. (Case 3). (A) IVP age 1 yr preoperatively. (B) IVP 10 days after colon conduit operation. No upper tract dilatation. Subsequent examinations remain normal. (C) IVP 7 mo after anostomosir of conduit to colon. Upper tracts remain normal. (D) The patient at age 4 yr, a well child. Bowel control of urine not yet perfect. Receives supplementmy sodium bicarbonate prophylactically.

392

W.

HARDY

HENDREN

conduit, rapid renal destruction is less likely and it would be possible to reoperate to revise the anastomosis. In this series, in fact, there was no problem with any of the anastomoses to date. Heal conduit diversion has been used in many patients with anterior exenteration. We believe that it would be reasonable to reoperate upon some of these patients, convert their drainage to a nonrefluxing colon conduit, and then later anastomose this to the colon. Case 2 (Fig. 6) Neurogenic Bladder Treated With Colon Conduit Urinary Diversion R. M., MGH 183-17-42 was referred at age 5 yr. A myelomeningocele had been closed at birth. She had urinary and fecal incontinence, although the latter was managed with success by a daily enema. Previous intravenous pyelogram showed severe bilateral hydronephrosis. Cystogram showed massive vesicoureteral reflux. A nonrefluxing colon conduit was performed, with tapering of the dilated right ureter. She has had an uneventful postoperative course now 20 mo postoperatively. !VP (Fig. 6B) has shown great improvement. Cultures obtained from the conduit by double-catheter technique are reported as “rare” or “few” gram-negative rods. She is clinically well, but is maintained on low-dose prophylactic antimicrobial therapy in view of the severe upper tract damage present before referral. Her stoma (6C) has given no problems, and she is well adjusted to this form of urinary diversion (6D).

Comment It is distressing that children with myelomeningocele should be permitted to develop this degree of severe.renal damage in this day and age when so much attention has been drawn to the need to follow these children carefully from the time of birth. They require periodic cultures, intravenous pyelograms, and cystograms to make certain that they do not have infection, reflux, and uppertract changes occurring. This youngster’s preoperative films serve as a poignant case in point. Although intermittent catheterization has been very helpful in managing many children with neurogenic bladder in the past few years, we felt that immediate diversion was needed in this child with already advanced upper-tract changes. With recent advances in the development of implantable prostheses to control urinary incontinence, ’ it is possible that permanent diversion will be practiced much less frequently in the future than it is now. Although we have not had an occasion to rejoin one of these nonrefluxing conduits to the bladder, we believe that it may be possible judging from the experience with nonrefluxing ileocecal cystoplasty. Case 3 (Fig. 7). Young Male With Exstrophy of the Bladder 1. B., MGH 169-61-46, was referred as a newborn March, 1970 with a rudimentary and exstrophied bladder. Intravenous pyelogram was normal. At age 1 yr a nonrefluxing colon conduit was performed. Postoperative convalescence was uneventful. Three months later bilateral inguinal hernias were repaired electively. At age I7 mo cystectomy and epispadias repair were performed. A reaction to blood transfusion occurred during this procedure, with temporary oliguria, treated by peritoneal dialysis with success and complete return of normal renal function. A loopogram performed at age 21 yr showed. no reflux at low pressures, but transient retlux into one ureter at pressures over 60 cm of water When bowel control was well established, the colon loop was taken down from the abdominal wall, resecting the stoma, and anastomosing it to the colon August 6, 1973.

NONREFLUXING

COLON

CONDUIT

393

In the 14 mo which have passed since then, he has been a clinically well patient, His intravenous pyelogram remains normal and he is on no urinary antimicrobial therapy. He takes 1 teaspoon four times daily of sodium bicarbonate. His serum values include sodium 130, potassium 4.4. chloride 109. and CO, 22 meq/liter. The rectal control of urine is socially satisfactory, but not yet perfect (he does not wear diapers, but occasionally has slightly damp underwear which must be changed). He attends nursery school and participates in all activities, with his peers who are unaware that he is different from them.

Comment

Although primary closure of the exstrophied bladder has been performed by many surgeons in the past 20 yr, in general we take a dismal view of that procedure, for the really successful results are few and far between. This particular patient would not have been a candidate for that in any event, since his bladder was rudimentary. We have seen many children in whom primary closure was performed, whose upper tracts then sustained serious damage before some diversionary procedure was performed. Thus, in our own cases we have not practiced primary closure of the exstrophied bladder since the chance for success is so small. Many of these patients have been treated by primary ureterosigmoidostomy. If that procedure is performed during early infancy, there is little likelihood of satisfactory bowel control for several years. Combined incontinence with urine and feces is socially disastrous, leading to ostracism by one’s peers. We have seen many examples of this and are convinced that ureterosigmoidostomy should not be done until an age at which bowel control of feces is already established, giving a reasonable expectancy that control of rectal liquid can soon be mastered, In former years nothing was done with these children until about age 4-5 yr, which was equally disturbing psychologically to the children and their parents, having to cope meanwhile with the exstrophied bladder and unrepaired genitalia. Although these children with unrepaired exstrophy did not develop urinary infection, there was considerable pain from the exstrophied bladder, bleeding, and psychological discomfort to the parents and child. We believe that the concept of a temporary colon conduit, with the plan for later anastomosis to the colon, has merit in these patients for the following reasons: (I) the conduit can be done at about I yr of age, providing a way to keep the youngster dry at an early age; (2) the genitalia can then be repaired before the child is aware of his or her deformity. This, further, leads to much better acceptance of the deformity by the parents. The youngster can then develop in a normal way for the next several years, deferring anastomosis of the conduit to the colon until a suitable age. By then there is long-term follow-,up of the ureterosigmoid anastomoses, amply ruling out obstruction or reflux. the two most important factors leading to renal damage when the ureters are brought into contact with the fecal stream. Immediately after ureteral reimplantation into the bladder we have seen vesicoureteral reflux demonstrated in some cases if a cystogram is done within the first 2 or 3 wk postoperatively, while there is still edema and before healing has completed. This can be seen even with technically satisfactory ureteral reimplants, which show no reflux subsequently when cystogram is done several weeks or months later. It would seem likely that the same can occur with a

394

W.

HARDY

HENDREN

primary ureterosigmoid diversion into the colon filled with fecal material, i.e., ascending infection can occur early even with a good ureterosigmoid anastomosis. With the staged conduit method described here, the ureters have long healed and demonstrated their competence before being brought into the unsterile environment of the colon. Case 4 (Fig. 8). Older Girl With Previous Ileal Conduit for Exstrophy of the Bladder C. W., MGH 129-94-06, was first seen as a newborn February 23, 1963, with exstrophy of a tiny, rudimentary bladder. Intravenous pyelogram was normal. On October 19, 1964, at age 20 months, ileal loop urinary diversion was performed. On June lst, 1965, cystectomy and epispadias repair were accomplished: IVP at this time showed some dilatation of the right ureter. It was felt that this might be secondary to stenosis of the stoma, which was small, and so it was revised. Subsequent intravenous pyelograms showed even more dilatation of the right ureter. and the stoma was revised once again, June 28th. 1967. Because the stoma tended to contract, despite two previous revisions, the family was instructed in its daily dilatation. In December of 1970, the intravenous pyelogram showed still some slight dilatation of the right ureter (Fig. 8A). Cultures from the ileal loop showed bacteria to be present on most occasions, and so she was maintained on continuous antimicrobial therapy using nitrofurantoin. On June 18, 1971 the ileal loop was resected, performing a nonrefluxing colon conduit. The right ureter was short, and so the proximal end of the colon conduit was placed in the right gutter, to permit ureterosigmoid anastomosis without tension. The left ureter was of ample length to bring it across the midline and construct a tunnel of satisfactory length. Intravenous pyelogram IO days later showed moderate bilateral dilatation (Fig. 8B). A repeat intravenous pyelogram 6 wk later was completely normal, with no dilatation of either upper tract. Loopogram done 6 mo postoperative (Fig. SC) showed no reflux even when the loop was distended with moderate pressure. On August 12, 1972 the conduit was detached from the abdominal wall and anastomosed to the colon. She has been a completely well patient, on no medication, with complete continence of urine. Recent intravenous pyelogram (8D) now 2 yr later shows normal upper tracts. Serum electrolyte values: sodium 144, potassium 4.1, chloride 106, CO* 22, BUN 16 meq/liter.

Comment There are many children who have been treated by ileal loop urinary diversion for exstrophy of the bladder who are candidates for this type of staged diversion of the urine flow to the colon. A number of years ago we anastomosed an ileal loop to the colon in two patients and immediate ascending pyelonephritis occurred. The ureters can not be connected to the colon without a nonrefluxing technique if pyelonephritis is to be averted. A segment of actively peristalsing small bowel will not prevent this reflux. One might consider switching primarily from ileal conduit diversion to ureterosigmoidostomy. That has the same potential disadvantages of primary ureterosigmoidostomy, i.e., if a technical problem with anastomosis occurs, severe renal damage can ensue before it is corrected. Further, leakage of an anastomosis in the intact colon can give fatal peritonitis. Also, ureters which are shorter than normal would be more difficult to work with in performing primary ureterosigmoid anastomosis than previously unoperated upon ureters. We believe, therefore, that placing them first into an isolated conduit is safer, even though it requires an additional operation to subsequently anastomose the conduit to the colon. Angulation of the ureter is one of the reported complications when anastomosing ureters at different levels into the intact colon. This has been no problem when dealing with an isolated conduit. In a case not operated upon previously, in which the conduit is anchored at the aortic bifurcation, the ureters

NONREFlUXlNG

COLON

CONDULT

Fig. 8. (Case 4). (A) IVP age 8 yr with ileal conduit. Note moderate dilatation of right ureter and renal pelvis. Two previous ileal stoma revisions. No evidence of ‘ureteroileal obstruction, as proved by free reflux on loopogram study. Colon conduit operation performed age 8 l/2 yr. The right ureter was short, and so base of stoma was anchored near lower pole of right kidney, instead at aortic bifurcation. (6) IVP 10 days after colon conduit operation, to demonstrate degree of temporary upper tract dilatation immediately after tunneling type of ureteral sigmoid anastomosis. In a defunctioned bowel loop this creates no problem. In a colon full of fecal material, this could result in immediate ascending pyelonephritis. Arrows demonstrate right ureteral tunnel. (C) Retrograde filling of loop, tamponading stoma with Foley catheter, showing no reflux, even at high pressure, greater than 60 cm of water. (D) IVP 26 mo after anastomosis of conduit to rectosigmoid. Note delicate upper tracts and complete resolution of dilatation of upper tract on right, which had been persistently dilated for several years with an ileal conduit. (This dilatation disappeared shortly after colon conduit operation was performed.) Small arrows point to long intramural course of right ureter in wall of conduit. Large arrow points to anastomosis of conduit to rectosigmoid. Note concentration of dye in conduit and rectum. This youngster, age 11 l/2 yr, is completely well, on no medication, with normal continence. The quality of her life is much improved, having gotten rid of her external urinary appliance.

are very close by, making fear of angulation.

it particularly

easy to join

them to the colon

without

DISCUSSION

The use of sigmoid colon for a urinary conduit is not new, being reported in H over 20 yr ago by Gross.* Mogg ‘*lo has favored the use of colon conduit

396

W.

HARDY

HENDREN

Fig. 9. Postoperative rocntgenogmms in g-yr-old girl 5 mo after colon conduit diversion for neurogenic bladder secondary to racrol ogeneris. Bilateral tunneling anastomoser were performed with tapering of the lower 3 inches of the left ureter, which was dilated. (A) Introvenous pyelogrom. Arrows point to the intramural segment of both ureters which contain dye ond can be readily identified, (6) High pressure filling of the colon conduit, producing reflux on the left, but not the right. No reflux was seen in any loopogram except with pressures exceeding 35 cm of water.

large number of children for urinary diversion with myelomeningocele. However, these operations were performed with end-to-side anastomosis of the ureter to the colon conduit, without establishing a long tunnel to prevent reflux. Almost 50 yr ago Coffey” emphasized that an intramural tunnel improves the success of ureterosigmoidostomy. Leadbetter ‘* showed that meticulous anastomosis of the end of the ureter to the colonic mucosa, in combination with an intramural tunnel, gives improved results. A large experience with ureterosigmoidostomy by this technique was subsequently reported.13 The Mathisen nipple has also proved effective.‘4,‘5 in dogs and found that Spence et aLI compared ileal and colonic conduits pyelonephritis occurred more often with colonic conduits; however, the anastomoses were of the refluxing type. Recently, Richie et al.‘7.‘8 compared the classic refluxing ileal conduit with a nonrefluxing colon conduit in a series of dogs. An ileal loop was done to divert one ureter and a colon loop to divert the other. Reflux was present in all but one of the ileal diversions, and in none of the colon conduits. Emptying time of the two conduits was the same, as were resting pressures within the bowel segments (3-4 mm of mercury). Peristalsis in the ileal segments occurred at 6 times per minute, with pressures to 25 mm of mercury. Colon peristalsis was measured at one wave per minute with pressures to 10 mm of mercury. When the stomas were occluded, ileal segment pressures

NONREFLUXING

COLON

397

CONDUIT

rose to higher levels, with pressure waves as high as 45 mm of mercury. When the animals were sacrificed 3 mo postoperatively, pyelonephritic changes were: found in 83”,; of kidneys diverted by ileal conduit in contrast to only 7”,, in, those diverted using a nonrefluxing colon conduit. This important laboratory investigative study strongly supports the clinical experience that a nonrefluxing anastomosis of the ureter to the colon is better whether it is into an isolated bowel segment or in contact with the fecal stream, and further it supports the view that an isolated segment of colon with nonrefluxing anastomoses is pre ferred to an ileal segment with free reflux. It should be emphasized that virtually any anastomoses can be made to reflux if undue pressure is used, a:< shown in Fig. 9. SUMMARY

Colon conduit, with nonrefluxing ureterosigmoid anastomoses, was performed upon 21 children in &he past 34 yr for the following indications: (I) previously unoperated exstrophy of the bladder: (2) exstrophy of the bladder with prior ileal conduit; (3) previously unoperated neurogenic bladder; (4) neurogenic bladder with prior ileal conduit; and (5) with anterior pelvic exenteration for sarcoma of the prostate. In some of these patients the conduit will be a permanent diversion; in others it will be anastomosed later to the colon, providing a staged method of ureterosigmoid urinary diversion. Nine patients have undergone the second stage of anastomosis of the conduit to the colon with satisfactory outcome. Colon conduit diversion is a more time-consuming procedure than ileal loop diversion. It can require 668 hr in a patient with a prior ileal diversion, particularly if the ureters require tapering. Our experience with this procedure leads us to the following conclusions: (I) this is a better method than ileal conduit for permanent diversion; (2) it offers an alternative way for treating patients with bladder exstrophy; (3) it is useful in patients with pelvic cancer; and (4) it should be considered for those patients with ileal conduits who are not doing well, i.e., who have infection and renal deterioration. REFERENCES I. Bricker E: Bladder substitution after pelvic evisceration. Surg Clin N Am 30: 15 I I, 1950 2. Retik AB, Perlmutter AD, Gross RE: Cutaneous uretero-ileostomy in children. N Engl J Med 277:217, 1967 3. Ray P. DeDomenico I: Intestinal conduit urinary diversion in children. Br J Ural 44:345, 1972 4. Smrth ED: Follow-up studies on 150 ileal conduits in children. J Pediatr Surg 7: I. 1972 5. Richie JP: Intestinal loop urmary diversion in children. J Urol I I l:687. 1974 6. Hendren WH: Operative repair of megaureter in children. J Urol 101:491, 1969 7. Scott FB. Bradley WE, Timm G: Use of an implantable artificial sphincter in the treat-

ment of urinary incontinence. J Urol ll2:i’S. 1974 8. Gross RE: Urinary and fecal incontinence of neurogenic origin. Chapt. 56 in The Surgery of Infancy and Childhood, Philadelphia, Samnders. 1953, p 740 9. Mogg RA: The treatment of neurogenic incontinence using the colonic conduit. Br J Urol 37:681. 1965 IO. Mogg RA: Some observations on urinary diversion. Ann Royal Coll Surg Engl 46:251. I970 I I. Coffey RC: Transplantation of the ureters into the large intestine. Surg Gynecol Obstet 47:593, 1928 12. Leadbetter WF: Consideration of prob-

398

lems incident to performance of ureteroenterostomy: Report of a technique. J Ural 65:818, 1951 13. Leadbetter WF, Clarke GB: Five years experience with uretero-enterostomy by the “combined technique.” J Ural 73:67, 1954 14. Mathisen W: A new method for ureterointestinal anastomosis. Surg Gynecol Obstet 96: 255, 1953 15. Mathisen W: Clinical and experimental studies on uretercolic anastomosis. A Mono-

W.

HARDY HENDREN

graph published by Universitetsforlaget. Oslo, 1969. 16. Spence B, Esho J, Cass A: Comparison of iliac and colonic conduit urinary diversions in dogs. J Urol 108:712. 1972 17. Richie JP, Skinner DC, Waisman J: The effect of reflux on the development of pyelonephritis in urinary diversion: An experimental study. J Surg Res 16:256, 1974 18. Richie JP, Skinner DG: Urinary diversion: The physiological rationale for nonrefluxing colonic conduits. Br J Urol 1974, In press

Surgical closure of patent ductus arteriosus in the premature infant with respiratory distress.

During the period from February, 1971 to February, 1973, 30 premature infants underwent surgical ligation of patent ductus arteriosus. The gestational...
3MB Sizes 0 Downloads 0 Views