Endorectal Heal Pullthrough Operation with lleal Reservoir after Total Colectomy Bernard T. Ferrari, MD, Los Angeles, Eric W. Fonkalsrud,
MD, Los Angeles,
California California
T1.e most commonly performed operation for ulcerative colitis, familial polyposis, and certain other nonmalignant diseases of the colon and rectum is tot.al proctocolectomy with permanent cutaneous ileostomy [I]. Although the patient is usually cured of his basic disease by this extensive operation, the res,ultant unfavorable aspects of the incontinent cu.;aneous ileostomy are being investigated to develop a more acceptable method of controlling ileost.omy output subsequent to colectomy. Moreover, complications after proctectomy occur sufficiently often to be a matter of serious concern. For example, abnormalities of bladder function stemming from injury to the autonomic nerves and muscular support in the pelvis frequently accompany proctectomy. An occasional male patient experiences postoperative impotence caused by interruption of the parasympathetic nerve supply to the genitalia [2,3]. Proctectomy, particularly when performed for inflammatory disease of the rectum, is often followed by perineal drainage that may persist for periods of many months, and in some cases even years, becoming a source of continuing discomfort, annoyance, and embarrassment. ;:t is generally believed that ulcerative colitis is a disease originating in the rectal and colonic mucosa and that the muscularis is involved only secondarily in advanced stages. Total colectomy and removal of the rectal mucosa with preservation of the rectal muscle, a technic originally developed by Soave [4] for the treatment of Hirschsprung’s disease, may therefore be adequate to cure the basic disease. Despite numerous operative technics to restore intestinal continuity after colectomy for ulcerative colitis, achievement of satisfactory fecal continence after ileoanal anastomosis has been rare. Indeed, perianal inflammation has been so severe that most patients
have urged abandoning the procedure in favor of conversion to cutaneous ileostomy 151. Rectal mucosal stripping operations combined with ileal pullthrough and ileoanal anastomosis, which have been evaluated in dogs and in occasional patients, are still under intensive study [6-81. However, the inability of such patients to retain fecal contents for more than a few hours at a time has limited the procedure’s clinical application. The present study summarizes our experience with seven dogs that underwent rectal mucosal stripping and endorectal ileal pullthrough with construction of an intraabdominal ileal reservoir immediately above the rectum. This procedure provides storage of fecal contents to decrease the frequency of bowel movements while still retaining fecal continence by preserving the anal-rectal sphincter mechanism.
Rect\al m cuff Rectoslgmold mucosa From the Division of Pediatric Surgery, UCLA School of Medicine, Los Angeles, California. Reprint requests should be addressed to Eric W. Fonkalsrud. MD, Department of Surgery, Division of Pediatric Surgery, UCLA School of Medicine, Los Anqeles, California 90024. Presented at the Forty-Ninth Annual Meeting of the Pacific Coast Surgical Association, Newport Beach, California, February 19-22. 1978.
Volume 136, July 1978
Figure 1. Rectal muscularis is circumcised 1 to 2 cm above peritoneal reflection. Mucosa is separated from muscularis by blunt dissection with dilute epinephrine-soaked pledget. Specific vessels are cauterized. Dissection is carried to within 1 cm of dentate line circumferential/y.
113
Ferrari and Fonkalsrud
Figure 2. The rectal mucosai tube is ligated at the level of the peritoneal reflection and the proximal sigmoid colon is removed (insert). After the anus has been dilated, a large Kelly clamp inserted up the rectum grasps the mucosa at its proximal end to evert it outside the anus. 1
Figure 3. The everted rectal mucosa is transected 7 to 10 mm above the dentate ilne. Quadrant sutures are placed to facilitate orientation for the subsequent anastomosis.
Material and Methods Seven conditioned adult mongrel dogs weighing between 17 and 23 kg were placed on complete bowel rest with no
oral intake for three days before operation. During that time each dog received 1 liter of lactated Ringer’s solution with 10 per cent dextrose parenterally each day. Kefline (cephalothin, sodium), 1 gm daily in divided doses, was given 6 hours preoperatively and during the first seven days after surgery. Each dog was placed under general anesthesia with pentothal sodium assisted by endotracheal ventilation. An enema of normal saline (400 cc) was given under anesthesia. A silicone rubber (Broviac) catheter (Evergreen Products, Inc, Medina, WA) was placed into the external jugular vein to facilitate long-term postoperative parenteral alimentation. Through a midline abdominal incision, the colon was removed from a point 2 to 3 cm proximal to the ileocecal valve distally to the sigmoid colon. The rectal muscle was then incised circumferentially at a point approximately 2 cm superior to the peritoneal reflection. (Figure 1.) The rectal mucosa was then separated from the rectal muscle by a combination of blunt and sharp dissection facilitated by a fine gauze pledget moistened with dilute epinephrine solution. Individual vessels between the mucosa and muscularis were cauterized. After finger dilatation of the anus, the upper end of the rectal mucosa was ligated and then everted through the anus. (Figures 2 and 3.) The colon proximal to the peritoneal reflection was excised. The terminal ileum was oversewn loosely and then brought through the rectal muscle canal and out through an incision in the distal rectal mucosa and then to the anus. The mesentery of the terminal ileum was mobilized sufficiently to permit advancement of the ileum down to the anus while still retaining good blood supply.
114
(Figure 4.) The full thickness of the ileum was then anastomosed to the rectal musosa, 7 to 10 mm above the dentate line with interrupted sutures of Dexone. (Figure 5.) The redundant rectal mucosa was resected during the procedure. The sutures were placed in such a manner as to include a segment of rectal muscle as well as the distal end of the rectal mucosa. The ileum just above the peritoneal reflection was sutured loosely to the open end of the rectal muscle, and a drain was placed in the space between the rectum and ileal pullthrough segment and brought through the abdominal wall. The ileal reservoir was then constructed by flexing the terminal ileum above the peritoneal reflection in an S shape. (Figure 6.) By incising the antimesenteric border of the three segments of the S-shaped loop over a distance of approximately 6 cm each, three suture lines were placed in such a manner as to construct a single reservoir. An inner row of continuous Dexon was used to approximate the mucosa, and an outer row of interrupted seromuscular silk sutures was placed. (Figure 7.) The ileal reservoir was thus directly continuous with both the proximal and distal ileum, without the presence of a side pouch. (Figure 8.)
Results Each of the seven dogs survived at least two weeks Five dogs lived longer than four weeks; one dog remains alive and well seven weeks and another eight months postoperatively. Two dogs were killed at two and two and a half months after operation to evaluate the anatomic and pathologic after the operation.
The American Journal of Surgery
Endorectal lleal Pullthrough
.e-
stitches through rectal mucosa
Figure 5. The full thickness of the terminal ileum is anastomosed to the rectal mucosa 7 to 10 mm above the dentate fine with interrupted Dexon sutures which incorporate a segment of recta/ muscle. The upper end of the rectal muscle is attached to the ileum circumferential/y to prevent herniation. Figure 4. The termfnal ileum is oversewn loosely and then brought through the rectal muscle canal down to the anus for anastomosis.
cha:nges in the reservoir and rectum. One dog was sacrificed because of weight loss and apparent pain associated with defecation one month postoperatively. Two dogs died between the second and third pos.toperative week because of intraabdominal soiklge and sepsis from suture line leakage of the reservoir. IXach of the dogs that survived four weeks or longer showed clear evidence of rectal continence within two to three weeks subsequent to operation. Each had soft or semisolid stools four to six times daily. Each dog assumed a squatting position for separate defecation and experienced no anal incontinence between defecations. None developed anal excoriation or inflammation. Each of the dogs lost weight, in some cases up to 25 per cent of the preoperative level by the end of the fourth postoperative week; however, thereafter all of the dogs stabilized or gained weight with supplemental high caloric, low residue oral
Volume
136, July 1978
feedings to which MetamuciP was added to provide fecal bulk. Roentgenography of the lower gastrointestinal tract of three dogs who survived four weeks or longer showed that the ileal reservoir was intact, without evidence of obstruction at either the afferent or efferent ends. The average intracavitary volume was approximately 100 cc at the time of construction, and this increased to 245 cc between the fourth and sixth week after operation. The ileum proximal to the reservoir showed minimal evidence of distension. (Figure 9.) Histologic evaluation of the ileal pouch revealed mild inflammation but no significant edema of the mucosa or submucosa. Gross and microscopic evaluation of the segment of the ileum at the anastomosis to the rectal mucosa showed good healing with mild inflammatron. (Figure 10.) There was no fluid or abscess formation between the ileum and rectal muscle in any of the dogs that survived beyond four weeks. Stimulation of the perianal skin evoked a normal anal sphincteric contraction. Rectal examination in
115
Ferrari and Fonkalsrud
Figure 6. The iieai reservoir is constructed by flexing the terminal ileum above the peritoneal reflection in an Sshaped manner. The antimesenteric border of the opposing segments of ileum are incised over a 6 to 8 cm distance and a large side-to-side two-layer anastomosis is constructed.
the awake animal indicated a normal anal sphincter mechanism that could be dilated with the finger. None of the dogs developed a stricture at the anastomosis between the pullthrough segment of ileum and the rectal mucosa above the dentate line. The postoperative convalescence of the four long-term surviving dogs was facilitated by pursuing a longer period of postoperative intravenous hyperalimentation extending up to two weeks. Comments
Most physicians believe that patients suffering from chronic ulcerative colitis who have not achieved permanent remission with medical therapy and who continue to be symptomatic for many years will benefit from cblectomy. Although reports by Aylett [9] have indicated that ileorectal anastomosis may be suitable for as many as 95 per cent of patients with ulcerative colitis, experience with rectum-preserving operations in the United States has been dismal, with more than three fourths of patients requiring sub-
116
Figure 7. Three suture lines are necessary to complete the S-shaped anastomosis for construction of the iieai reservoir.
sequent proctectomy even when the rectum has been diverted with a mucous fistula [IO]. In contrast to cancer of the rectum, in which removal of all layers of the intestine is required for optimal treatment, ulcerative colitis is primarily a disease of the mucosa, and removal of the rectal mucosa when there is only mild to moderate involvement should be curative. The feasibility of removing the rectal mucosa without disturbing the rectal muscle has been demonstrated experimentally, and the technic has found wide clinical application when combined with a colon pullthrough operation for Hirschsprung’s disease and high imperforate anus malformations. Similar rectal mucosal stripping operations, combined with ileal pullthrough and ileoanal anastomosis, have been evaluated in dogs and in occasional patients and are still under intensive study [7,8]. The inability of these patients, however, to retain fecal contents for more than a short period of time has limited the procedure’s clinical application. Nonetheless, this technic has
The American Journal of Surgery
Endorectal Heal Pullthrough
Figure 8. The ileal reservoir is continuous with the proximal and distal ileum.
been successful in preserving continence by avoiding injury to the puborectalis sling, the internal and external anal sphincters, and the neurogenic supply to the rectum, genitalia, and urinary tract. Patient dissatisfaction with the standard cutaneous ileostomy after colectomy for ulcerative colitis has prompted a search for technics to obviate the need for continuous wearing of ileostomy appliances. Experimental and clinical experience with the Kock ile.4 reservoir has been sufficiently favorable to justify further clinical study [II-131, although several limitations have been cited. The operation described in the present report utilizes a modification of the basic concept of the Kock pouch, namely, that of establishing an ileal reservoir near the anus. The S-shaped construction of the reservoir, omitting an inverted “nipple” to keep intestinal contents from escaping, has apparently prevented some of the distension that usually occurs with the Kock reservoir. Nonetheless, the S-shaped reservoir’s capacity of at lezist 250 cc makes it feasible for storage of ileal contents without significant discomfort for several hours. The reservoir may be constructed as large as desired; however, the experience thus far suggests that suture lines of approximately 6 to 8 cm may be optimal. Preservation of the neurogenic sensory mechanism at the dentate line and the lower 1 cm of rectal mucosa appears to increase the likelihood of maintaining an almost normal sensation of rectal fullness and urge
Volume 136, July 1979
Figure 9. Roentgenogram of ileal reservoir in dog containing barium taken one month postoperatively. There is no dilatation of the proximal ileum.
to defecate.
The ability of the ileorectal pullthrough segment to discriminate gas from fecal content, however, has not been determined. It is nevertheless significant that in the present study the dogs in which the ileal reservoir was constructed above t,he rectal pullthrough segment of ileum have not experienced the severe anal excoriation and skin breakdown that was observed previously in cases of ileoanal anastomosis without rectal preservation [ 71. Technically important features include the fact that caution must be exercised to avoid injury to the vascular supply to the terminal ileum, which during manipula.tion to bring the ileum down to the anus, may be placed under slight tension. It also appears that the segment of rectal muscle through which the ileum is pulled should not exceed 6 to 8 cm lest a
117
Ferrari and Fonkalsrud
Figure 70. Sagiffal section from ileum above ileorectal anastomosis demonstrating mild inflammation and minimal scarring with beginning conversion of ileal mucosa to that with colonic features. ( Magnification X 66; reduced 50 per cent. )
partial obstruction develop from compression by the rectum. Care must also be taken to avoid injury and kinking of the ileal mesentery during the construction of the reservoir. As noted in the illustrations, the mesentery of each of the three segments of the “Sshaped reservoir” are fashioned outward. Although it was not feasible to construct a temporary protective ileostomy proximal to the ileal reservoir in the experiments with dogs, this approach is strongly recommended in a clinical setting in which the care of a cutaneous ileostomy is much easier to manage satisfactorily. It is of interest that the bowel habits of the dogs with the ilea reservoir were not significantly different from those of dogs in a previous study in our laboratory that had an ileoendorectal pullthrough procedure without the reservoir [7]. Nonetheless, with the addition of LomotiP and Metamucil to the diet, the dogs with the reservoir were able to develop semisolid stools, experienced less frequent defecation, and appeared to have more advanced warning of the urge to defecate than did dogs without the reservoir. Mild inflammation of the terminal ileum and the pouch may be related to retention of fecal contents and the response to a change in fecal flora. Current studies are under way to evaluate the bacterial flora of the reservoir and to determine whether the presence of the ileal reservoir permits more water absorption than does the endorectal ileal pullthrough without reservoir. Based upon the favorable results in these experiments with dogs, a fourteen year old boy with a four year history of unremitting severe ulcerative colitis complicated by severe diabetes underwent total colectomy, with removal of the rectal mucosa down to within 1 cm of the dentate line. The terminal ileum 118
Figure If. Barium roentgenogram showing the ileal reservoir three months postoperatively in a fourteen year old boy with ulcerative colitis.
was brought through the rectal muscle pouch and anastomosed to the low rectum in the manner described for the dogs. The ileal reservoir was constructed in the same way, after which the ileum was divided approximately 8 cm proximal to the reservoir with the distal end oversewn. The proximal ileum was brought to the right lower abdominal wall as a cutaneous ileostomy. The patient recovered from the operative procedure without complications and was advanced to a regular diet. A contrast roentgenogram taken of the lower gastrointestinal tract three months after operation showed a large ileal reservoir calcu350 cc of contrast lated to contain approximately medium. (Figure 11.) The ileum both proximal and distal to the reservoir was normally dilated without obstruction. Three and a half months after the initial operation, proctoscopic examination revealed no structure of the ileoanal anastomosis, and therefore, the proximal diverting ileostomy was closed. During the ensuing weeks, the patient experienced in excess of 15 to 20 stools per day initially, which gradually The American Journal
of Surgery
Endorectal
decreased
as a result
oral diet with Metamucil
of an intermittent
low residue
and Lomotil.
and with Although
elemental
administration ileitis
of
in the bowel distal to the reservoir, it was successfully managed by temporary bowel rest, antibiotics, and a course of prednisone. The patient has an intact anal sphincter mechanism and does not experience fecal soiling. Moreover, he is able to identify clearly an urge to defecate well before he must eliminate. During the ensuing weeks the patient has learned to hold back the urge and to defecate slowly to make better use of the ileal reservoir. There has been no evidence of perianal excoriation or ulceration. He has experienced none of the abdominal cramping pain that has been noted in previous patients who have had ileoendorectal pullthrough operations without a reservoir. Summary Ileoanal anastomosis through an extramucosal rectal muscular tube with construction of a proximal S-shaped ileal reservoir was performed on seven dogs in an attempt to reestablish fecal continence. Five of the dogs survived more than four weeks and experienced fecal continence, although they had frequent stools. The use of Lomotil and Metamucil has assisted in decreasing the frequency of defecation, and also in developing semisolid stools. None of the dogs developed perianal excoriation or ulceration. The ileal reservoir volume averaged 250 ml in the longterm surviving dogs. It is anticipated that this operative technic may have clinical application in selected patients with ulcerative colitis and other diseases primarily involving the rectal mucosa, since the neurogenic and muscular sphincteric mechanism necessary for continence can be preserved. Clinical experience with the ileal reservoir and endorectal pullthrough in one patient has been encouraging. References 1. Sirinek KR, Tetirick CE, Thomford NR, Pace WG: Total proctocolectomy and ileostomy. Arch Surg 11’2: 518, 1977. 2. Donovan MJ, O’Hara MD: Sexual function following surgery for ulcerative colitis. N Engl J Med 262: 719, 1960. 3. David LP, Jelenko C: Sexual function after abdominal perineal resection. South Med J 68: 422, 1975. 4. Soave F: A new surgical technique for treatment of Hirschsprung’s disease. Surgery 56: 1007, 1964. 5. Ravitch MM: Anal ileostomy with sphincter preservation in patients requiring total colectomy for benign conditions. Surgery 24: 170, 1948. 6. Devine J. Webb R: Resection of the rectal mucosa, colectomy, and anal ileostomy with normal continence. Surg Gynecol Obstet92: 437, 1951. 7. Levin P, Fonkalsrud EW: Fecal continence following colectomy
Volume 136, July 1976
8.
developed 9.
10.
11.
12.
13.
lleal Pullthrough
and ileoanal anastomosis through extra mucosal rectal tube. J Sur!g Res 8: 234, 1968. Martin L., LeCoultre C, Schubert WK: Total colectomy and mucosal proctectomy with preservation of continence in ulcerative colitis. Ann Surg 186: 477, 1977. Aylett SO: Three hundred cases of diffuse ulcerative colitis treated by total colectomy and ileorectal anastomosis. 6r Med./ 1: 1001, 1966. Binder SC, Miller HH, Deterling RA Jr: Fate of the retained rectum after subtotal colectomy for inflammatory disease of the colon. Am JSurg 131: 201, 1976. Kock NC? Intra-abdominal “reservoir” in patients with permanent Ileostomy: preliminary observations on a procedure resulting in fecal “continence” in five ileosiomy patients. Arch Surg 99: 223, 1969. Beahrs OH, Kelly KA, Adson MA, et al: lleostomy with ileal reservoir rather than ileostomy alone. Ann Surg 179: 634, 1974. Handelsman JC, Fishbein RH: Permanent ileostomy without extemal appliance: Koch internal reservoir (pouch) operation. Johns Hopkins MedJ 138: 161. 1976.
Discussion Alan Morgan (Seattle, WA): The authors have made an important contribution to the surgical treatment of ulcerative colitis and other nonmalignant diseases of the colon and rectum. Surgeons and patients will await follow-up results with interest, and the patients almost demand them, for in no other field of surgery do patients play such an integral role in the selection of their operation. Credit must be given to Drs. Ravitch [5] and Sabiston of the Hunterian Laboratory at the Johns Hopkins Hospital for the idea of an endorectal pullthrough and ileoanal anastomosis. They believed that only the mucosa need be removed in benign conditions and they stripped the mucosa up from below. The Soave [4] operation, stripping out the rectal mucosa in Hirschprung’s disease from above, as the authors did, came eighteen years later, in 1966. Those who care for children will know that this is really not too difficult, especially when the rectum has been defunctioned. But is the stripping out difficult in ulcerative colitis? Does the diseased, friable mucosa come out easily? I suspect that a number of patients with a severely involved rectum will not be suitable for the operation even if it is staged. I hope that more patients are suitable for this new procedure than are suitable for Aylett’s ileorectal anastomosis [9]. It could be that a good safe stripping is limited by the same factors that limit a good ileorect,al anastomosis in the hands of the majority of surgeons. Therefo:re, many will find this an attractive concept, particularly those who are not happy with the Koch pouch, for some of the technical aspects which can defeat the pouch, particularly the intussuscepting nipple, are not necessary. The patient has a good continence mechanism already made by nat.ure waiting to control the pullthrough. The colon and rectal surgeon is concerned about the disease in the mucosa, and the pediatric surgeon is concerned about the problems inherent in the pullthrough operations, namely structures at the suture line just above the dentate line and at the upper end of the rectal muscular cuff and abscess formation in the cuff. Staging t.he proce-
119
Ferrari and Fonkalsrud
dure will reduce the risk of the latter, and certainly we all agree that these procedures should be staged. William B. Hutchinson (Seattle, WA): A successful ileoanal anastomosis was performed in 1947 on a twentyfive year old female with chronic ulcerative colitis and was reported before this Association in Coronado in 1951. So successful was the result that after the death of her first husband she married a second. She had normal sexual desires and function, the ability to differentiate gas and feces, and normal bladder function. She was able to constrict on the examining finger and control her feces very well. This took about a three month training period. I therefore operated on three other patients with chronic ulcerative colitis, and all of these ended in total failure, necessitating reestablishment of their anal ileostomy to an abdominal ileostomy. They were able to control their sphincters, having good contraction of their sphincter mechanism, but were totally unable to differentiate between gas and feces, were incontinent of stool, and were tremendously excoriated. Not one of these patients had any disability in anal contraction. They had no fistulas and no infections to account for the incontinence. Is it possible that in the dogs shown here, the horizontal posture of the dog as opposed to the perpendicular posture of man made a difference in the results of their continence? This was an excellent presentation, and I hope the authors are successful in overcoming the problem of incontinence. I can assure you that there is nothing more intolerable than an incontinent anal ileostomy with massive excoriation in an emotional nervous patient. Wiley F. Barker (Los Angeles, CA): I would like to present some soft information from the same population of patients which may, at least in a subliminal way, have led Dr. Fonkalsrud to perform this procedure. Over the last twenty years we have carried out about twenty-five ileorectal anastomoses, and most of them have failed. In about fifteen of these patients we constructed an inverted 9 loop, that is, we carried out side-to-end ileorectal anastomosis 15 cm from the divided end of the ileum, and then performed end-to-side anastomosis between the end of the ileum and the side of the ileum, 15 cm back from the anastomosis, thus creating an “inverted 9.” Most of these patients had a good functional result with this as a pouch, but all but one have ultimately failed primarily because of the recurrence at the distal anastomosis. As reported elsewhere, we thought that the procedure using the rectum with its intact mucosa for an ileorectal anastomosis failed commonly in a period of time inversely proportional to the duration and severity of the disease, rather than because of the actual pedigree of the disease. I also believe that the common failure of the usual ileoanal anastomosis has been due to the imposition of what amounts to intestinal obstruction on the small bowel if the anal sphincter is truly competent. I believe that Drs. Ferrari and Fonkalsrud have indeed introduced a means of first, eliminating the primary site of recurrence as well as second, relieving the practically obstructing effect of the
120
sphincter blunts the I predict but it will
by this damping effect of the reservoir which onrush of intestinal peristalsis. that there will be problems with this operation, be a useful operation in the future.
Donald M. Gallagher (San Francisco, CA): I too would like to share the enthusiasm about this operation, but I suggest caution. First, there was an implication that the operation being discussed is similar to the Kock continent ileostomy. Continence in the Kock procedure is dependent upon the ileostomy nipple acting as a mechanical obstruction to the reservoir. The obstruction of the nipple is overcome only when a catheter is passed through the ileostomy to the reservoir. The reservoir is important, but it is the construction of the nipple that determines the success or failure of the Kock procedure. The human rectum does not have the ability to act as an obstructing ring for a protracted period of time. Second, recently in doing low anterior anastomoses with the endoanal stapler in dogs I have been impressed by the difference of the muscular coat of the dog’s rectum compared with that of humans. In the dog the muscle thickness of the rectum is at least fivefold greater than in humans. Also, in the dog there is ease of complete exposure of the rectum, which is not true in man. Thus, the dog is not applicable as a surgical model for the type of surgery that might be done in humans. Eric W. Fonkalsrud (closing): As indicated, the described procedure is easy to perform in a dog with a normal rectum, but is difficult in a patient who has mucosal ulcerations and severe inflammation. The permanent cutaneous ileostomy has been frustrating to many young patients who may have to live with an appliance for as long as fifty to sixty years. The procedure described is not applicable for patients with Crohn’s disease since the ileal reservoir, just as a Kock pouch, may develop recurrent inflammation. If the reservoir does not function adequately, only approximately 39 cm of ileum might have to be removed, leaving the patient with an ileostomy as would be the case had the reservoir not been constructed. One of the key features of the operation is careful preservation of the blood supply to the terminal ileum during construction of the reservoir. The nipple is the key feature for preservation of continence in the Kock pouch; however, the patient who has an intact anorectal sphincter will experience the sensation of fullness and will be able to defecate when desired without a nipple. We have applied the operation to one patient with ulcerative colitis. The ileum was closed proximal to the pouch and a diverting ileostomy was constructed. The ileostomy was closed two and a half months later, and the patient is currently continent and able to distinguish air from fluid when he defecates, an advance over our previous experience with endorectal ileal pullthrough without a reservoir. Nonetheless, much more experience with the procedure’s technical features is necessary before further clinical application can be recommended.
The American Journal of Surgery
MD, Los Angeles,
California California
T1.e most commonly performed operation for ulcerative colitis, familial polyposis, and certain other nonmalignant diseases of the colon and rectum is tot.al proctocolectomy with permanent cutaneous ileostomy [I]. Although the patient is usually cured of his basic disease by this extensive operation, the res,ultant unfavorable aspects of the incontinent cu.;aneous ileostomy are being investigated to develop a more acceptable method of controlling ileost.omy output subsequent to colectomy. Moreover, complications after proctectomy occur sufficiently often to be a matter of serious concern. For example, abnormalities of bladder function stemming from injury to the autonomic nerves and muscular support in the pelvis frequently accompany proctectomy. An occasional male patient experiences postoperative impotence caused by interruption of the parasympathetic nerve supply to the genitalia [2,3]. Proctectomy, particularly when performed for inflammatory disease of the rectum, is often followed by perineal drainage that may persist for periods of many months, and in some cases even years, becoming a source of continuing discomfort, annoyance, and embarrassment. ;:t is generally believed that ulcerative colitis is a disease originating in the rectal and colonic mucosa and that the muscularis is involved only secondarily in advanced stages. Total colectomy and removal of the rectal mucosa with preservation of the rectal muscle, a technic originally developed by Soave [4] for the treatment of Hirschsprung’s disease, may therefore be adequate to cure the basic disease. Despite numerous operative technics to restore intestinal continuity after colectomy for ulcerative colitis, achievement of satisfactory fecal continence after ileoanal anastomosis has been rare. Indeed, perianal inflammation has been so severe that most patients
have urged abandoning the procedure in favor of conversion to cutaneous ileostomy 151. Rectal mucosal stripping operations combined with ileal pullthrough and ileoanal anastomosis, which have been evaluated in dogs and in occasional patients, are still under intensive study [6-81. However, the inability of such patients to retain fecal contents for more than a few hours at a time has limited the procedure’s clinical application. The present study summarizes our experience with seven dogs that underwent rectal mucosal stripping and endorectal ileal pullthrough with construction of an intraabdominal ileal reservoir immediately above the rectum. This procedure provides storage of fecal contents to decrease the frequency of bowel movements while still retaining fecal continence by preserving the anal-rectal sphincter mechanism.
Rect\al m cuff Rectoslgmold mucosa From the Division of Pediatric Surgery, UCLA School of Medicine, Los Angeles, California. Reprint requests should be addressed to Eric W. Fonkalsrud. MD, Department of Surgery, Division of Pediatric Surgery, UCLA School of Medicine, Los Anqeles, California 90024. Presented at the Forty-Ninth Annual Meeting of the Pacific Coast Surgical Association, Newport Beach, California, February 19-22. 1978.
Volume 136, July 1978
Figure 1. Rectal muscularis is circumcised 1 to 2 cm above peritoneal reflection. Mucosa is separated from muscularis by blunt dissection with dilute epinephrine-soaked pledget. Specific vessels are cauterized. Dissection is carried to within 1 cm of dentate line circumferential/y.
113
Ferrari and Fonkalsrud
Figure 2. The rectal mucosai tube is ligated at the level of the peritoneal reflection and the proximal sigmoid colon is removed (insert). After the anus has been dilated, a large Kelly clamp inserted up the rectum grasps the mucosa at its proximal end to evert it outside the anus. 1
Figure 3. The everted rectal mucosa is transected 7 to 10 mm above the dentate ilne. Quadrant sutures are placed to facilitate orientation for the subsequent anastomosis.
Material and Methods Seven conditioned adult mongrel dogs weighing between 17 and 23 kg were placed on complete bowel rest with no
oral intake for three days before operation. During that time each dog received 1 liter of lactated Ringer’s solution with 10 per cent dextrose parenterally each day. Kefline (cephalothin, sodium), 1 gm daily in divided doses, was given 6 hours preoperatively and during the first seven days after surgery. Each dog was placed under general anesthesia with pentothal sodium assisted by endotracheal ventilation. An enema of normal saline (400 cc) was given under anesthesia. A silicone rubber (Broviac) catheter (Evergreen Products, Inc, Medina, WA) was placed into the external jugular vein to facilitate long-term postoperative parenteral alimentation. Through a midline abdominal incision, the colon was removed from a point 2 to 3 cm proximal to the ileocecal valve distally to the sigmoid colon. The rectal muscle was then incised circumferentially at a point approximately 2 cm superior to the peritoneal reflection. (Figure 1.) The rectal mucosa was then separated from the rectal muscle by a combination of blunt and sharp dissection facilitated by a fine gauze pledget moistened with dilute epinephrine solution. Individual vessels between the mucosa and muscularis were cauterized. After finger dilatation of the anus, the upper end of the rectal mucosa was ligated and then everted through the anus. (Figures 2 and 3.) The colon proximal to the peritoneal reflection was excised. The terminal ileum was oversewn loosely and then brought through the rectal muscle canal and out through an incision in the distal rectal mucosa and then to the anus. The mesentery of the terminal ileum was mobilized sufficiently to permit advancement of the ileum down to the anus while still retaining good blood supply.
114
(Figure 4.) The full thickness of the ileum was then anastomosed to the rectal musosa, 7 to 10 mm above the dentate line with interrupted sutures of Dexone. (Figure 5.) The redundant rectal mucosa was resected during the procedure. The sutures were placed in such a manner as to include a segment of rectal muscle as well as the distal end of the rectal mucosa. The ileum just above the peritoneal reflection was sutured loosely to the open end of the rectal muscle, and a drain was placed in the space between the rectum and ileal pullthrough segment and brought through the abdominal wall. The ileal reservoir was then constructed by flexing the terminal ileum above the peritoneal reflection in an S shape. (Figure 6.) By incising the antimesenteric border of the three segments of the S-shaped loop over a distance of approximately 6 cm each, three suture lines were placed in such a manner as to construct a single reservoir. An inner row of continuous Dexon was used to approximate the mucosa, and an outer row of interrupted seromuscular silk sutures was placed. (Figure 7.) The ileal reservoir was thus directly continuous with both the proximal and distal ileum, without the presence of a side pouch. (Figure 8.)
Results Each of the seven dogs survived at least two weeks Five dogs lived longer than four weeks; one dog remains alive and well seven weeks and another eight months postoperatively. Two dogs were killed at two and two and a half months after operation to evaluate the anatomic and pathologic after the operation.
The American Journal of Surgery
Endorectal lleal Pullthrough
.e-
stitches through rectal mucosa
Figure 5. The full thickness of the terminal ileum is anastomosed to the rectal mucosa 7 to 10 mm above the dentate fine with interrupted Dexon sutures which incorporate a segment of recta/ muscle. The upper end of the rectal muscle is attached to the ileum circumferential/y to prevent herniation. Figure 4. The termfnal ileum is oversewn loosely and then brought through the rectal muscle canal down to the anus for anastomosis.
cha:nges in the reservoir and rectum. One dog was sacrificed because of weight loss and apparent pain associated with defecation one month postoperatively. Two dogs died between the second and third pos.toperative week because of intraabdominal soiklge and sepsis from suture line leakage of the reservoir. IXach of the dogs that survived four weeks or longer showed clear evidence of rectal continence within two to three weeks subsequent to operation. Each had soft or semisolid stools four to six times daily. Each dog assumed a squatting position for separate defecation and experienced no anal incontinence between defecations. None developed anal excoriation or inflammation. Each of the dogs lost weight, in some cases up to 25 per cent of the preoperative level by the end of the fourth postoperative week; however, thereafter all of the dogs stabilized or gained weight with supplemental high caloric, low residue oral
Volume
136, July 1978
feedings to which MetamuciP was added to provide fecal bulk. Roentgenography of the lower gastrointestinal tract of three dogs who survived four weeks or longer showed that the ileal reservoir was intact, without evidence of obstruction at either the afferent or efferent ends. The average intracavitary volume was approximately 100 cc at the time of construction, and this increased to 245 cc between the fourth and sixth week after operation. The ileum proximal to the reservoir showed minimal evidence of distension. (Figure 9.) Histologic evaluation of the ileal pouch revealed mild inflammation but no significant edema of the mucosa or submucosa. Gross and microscopic evaluation of the segment of the ileum at the anastomosis to the rectal mucosa showed good healing with mild inflammatron. (Figure 10.) There was no fluid or abscess formation between the ileum and rectal muscle in any of the dogs that survived beyond four weeks. Stimulation of the perianal skin evoked a normal anal sphincteric contraction. Rectal examination in
115
Ferrari and Fonkalsrud
Figure 6. The iieai reservoir is constructed by flexing the terminal ileum above the peritoneal reflection in an Sshaped manner. The antimesenteric border of the opposing segments of ileum are incised over a 6 to 8 cm distance and a large side-to-side two-layer anastomosis is constructed.
the awake animal indicated a normal anal sphincter mechanism that could be dilated with the finger. None of the dogs developed a stricture at the anastomosis between the pullthrough segment of ileum and the rectal mucosa above the dentate line. The postoperative convalescence of the four long-term surviving dogs was facilitated by pursuing a longer period of postoperative intravenous hyperalimentation extending up to two weeks. Comments
Most physicians believe that patients suffering from chronic ulcerative colitis who have not achieved permanent remission with medical therapy and who continue to be symptomatic for many years will benefit from cblectomy. Although reports by Aylett [9] have indicated that ileorectal anastomosis may be suitable for as many as 95 per cent of patients with ulcerative colitis, experience with rectum-preserving operations in the United States has been dismal, with more than three fourths of patients requiring sub-
116
Figure 7. Three suture lines are necessary to complete the S-shaped anastomosis for construction of the iieai reservoir.
sequent proctectomy even when the rectum has been diverted with a mucous fistula [IO]. In contrast to cancer of the rectum, in which removal of all layers of the intestine is required for optimal treatment, ulcerative colitis is primarily a disease of the mucosa, and removal of the rectal mucosa when there is only mild to moderate involvement should be curative. The feasibility of removing the rectal mucosa without disturbing the rectal muscle has been demonstrated experimentally, and the technic has found wide clinical application when combined with a colon pullthrough operation for Hirschsprung’s disease and high imperforate anus malformations. Similar rectal mucosal stripping operations, combined with ileal pullthrough and ileoanal anastomosis, have been evaluated in dogs and in occasional patients and are still under intensive study [7,8]. The inability of these patients, however, to retain fecal contents for more than a short period of time has limited the procedure’s clinical application. Nonetheless, this technic has
The American Journal of Surgery
Endorectal Heal Pullthrough
Figure 8. The ileal reservoir is continuous with the proximal and distal ileum.
been successful in preserving continence by avoiding injury to the puborectalis sling, the internal and external anal sphincters, and the neurogenic supply to the rectum, genitalia, and urinary tract. Patient dissatisfaction with the standard cutaneous ileostomy after colectomy for ulcerative colitis has prompted a search for technics to obviate the need for continuous wearing of ileostomy appliances. Experimental and clinical experience with the Kock ile.4 reservoir has been sufficiently favorable to justify further clinical study [II-131, although several limitations have been cited. The operation described in the present report utilizes a modification of the basic concept of the Kock pouch, namely, that of establishing an ileal reservoir near the anus. The S-shaped construction of the reservoir, omitting an inverted “nipple” to keep intestinal contents from escaping, has apparently prevented some of the distension that usually occurs with the Kock reservoir. Nonetheless, the S-shaped reservoir’s capacity of at lezist 250 cc makes it feasible for storage of ileal contents without significant discomfort for several hours. The reservoir may be constructed as large as desired; however, the experience thus far suggests that suture lines of approximately 6 to 8 cm may be optimal. Preservation of the neurogenic sensory mechanism at the dentate line and the lower 1 cm of rectal mucosa appears to increase the likelihood of maintaining an almost normal sensation of rectal fullness and urge
Volume 136, July 1979
Figure 9. Roentgenogram of ileal reservoir in dog containing barium taken one month postoperatively. There is no dilatation of the proximal ileum.
to defecate.
The ability of the ileorectal pullthrough segment to discriminate gas from fecal content, however, has not been determined. It is nevertheless significant that in the present study the dogs in which the ileal reservoir was constructed above t,he rectal pullthrough segment of ileum have not experienced the severe anal excoriation and skin breakdown that was observed previously in cases of ileoanal anastomosis without rectal preservation [ 71. Technically important features include the fact that caution must be exercised to avoid injury to the vascular supply to the terminal ileum, which during manipula.tion to bring the ileum down to the anus, may be placed under slight tension. It also appears that the segment of rectal muscle through which the ileum is pulled should not exceed 6 to 8 cm lest a
117
Ferrari and Fonkalsrud
Figure 70. Sagiffal section from ileum above ileorectal anastomosis demonstrating mild inflammation and minimal scarring with beginning conversion of ileal mucosa to that with colonic features. ( Magnification X 66; reduced 50 per cent. )
partial obstruction develop from compression by the rectum. Care must also be taken to avoid injury and kinking of the ileal mesentery during the construction of the reservoir. As noted in the illustrations, the mesentery of each of the three segments of the “Sshaped reservoir” are fashioned outward. Although it was not feasible to construct a temporary protective ileostomy proximal to the ileal reservoir in the experiments with dogs, this approach is strongly recommended in a clinical setting in which the care of a cutaneous ileostomy is much easier to manage satisfactorily. It is of interest that the bowel habits of the dogs with the ilea reservoir were not significantly different from those of dogs in a previous study in our laboratory that had an ileoendorectal pullthrough procedure without the reservoir [7]. Nonetheless, with the addition of LomotiP and Metamucil to the diet, the dogs with the reservoir were able to develop semisolid stools, experienced less frequent defecation, and appeared to have more advanced warning of the urge to defecate than did dogs without the reservoir. Mild inflammation of the terminal ileum and the pouch may be related to retention of fecal contents and the response to a change in fecal flora. Current studies are under way to evaluate the bacterial flora of the reservoir and to determine whether the presence of the ileal reservoir permits more water absorption than does the endorectal ileal pullthrough without reservoir. Based upon the favorable results in these experiments with dogs, a fourteen year old boy with a four year history of unremitting severe ulcerative colitis complicated by severe diabetes underwent total colectomy, with removal of the rectal mucosa down to within 1 cm of the dentate line. The terminal ileum 118
Figure If. Barium roentgenogram showing the ileal reservoir three months postoperatively in a fourteen year old boy with ulcerative colitis.
was brought through the rectal muscle pouch and anastomosed to the low rectum in the manner described for the dogs. The ileal reservoir was constructed in the same way, after which the ileum was divided approximately 8 cm proximal to the reservoir with the distal end oversewn. The proximal ileum was brought to the right lower abdominal wall as a cutaneous ileostomy. The patient recovered from the operative procedure without complications and was advanced to a regular diet. A contrast roentgenogram taken of the lower gastrointestinal tract three months after operation showed a large ileal reservoir calcu350 cc of contrast lated to contain approximately medium. (Figure 11.) The ileum both proximal and distal to the reservoir was normally dilated without obstruction. Three and a half months after the initial operation, proctoscopic examination revealed no structure of the ileoanal anastomosis, and therefore, the proximal diverting ileostomy was closed. During the ensuing weeks, the patient experienced in excess of 15 to 20 stools per day initially, which gradually The American Journal
of Surgery
Endorectal
decreased
as a result
oral diet with Metamucil
of an intermittent
low residue
and Lomotil.
and with Although
elemental
administration ileitis
of
in the bowel distal to the reservoir, it was successfully managed by temporary bowel rest, antibiotics, and a course of prednisone. The patient has an intact anal sphincter mechanism and does not experience fecal soiling. Moreover, he is able to identify clearly an urge to defecate well before he must eliminate. During the ensuing weeks the patient has learned to hold back the urge and to defecate slowly to make better use of the ileal reservoir. There has been no evidence of perianal excoriation or ulceration. He has experienced none of the abdominal cramping pain that has been noted in previous patients who have had ileoendorectal pullthrough operations without a reservoir. Summary Ileoanal anastomosis through an extramucosal rectal muscular tube with construction of a proximal S-shaped ileal reservoir was performed on seven dogs in an attempt to reestablish fecal continence. Five of the dogs survived more than four weeks and experienced fecal continence, although they had frequent stools. The use of Lomotil and Metamucil has assisted in decreasing the frequency of defecation, and also in developing semisolid stools. None of the dogs developed perianal excoriation or ulceration. The ileal reservoir volume averaged 250 ml in the longterm surviving dogs. It is anticipated that this operative technic may have clinical application in selected patients with ulcerative colitis and other diseases primarily involving the rectal mucosa, since the neurogenic and muscular sphincteric mechanism necessary for continence can be preserved. Clinical experience with the ileal reservoir and endorectal pullthrough in one patient has been encouraging. References 1. Sirinek KR, Tetirick CE, Thomford NR, Pace WG: Total proctocolectomy and ileostomy. Arch Surg 11’2: 518, 1977. 2. Donovan MJ, O’Hara MD: Sexual function following surgery for ulcerative colitis. N Engl J Med 262: 719, 1960. 3. David LP, Jelenko C: Sexual function after abdominal perineal resection. South Med J 68: 422, 1975. 4. Soave F: A new surgical technique for treatment of Hirschsprung’s disease. Surgery 56: 1007, 1964. 5. Ravitch MM: Anal ileostomy with sphincter preservation in patients requiring total colectomy for benign conditions. Surgery 24: 170, 1948. 6. Devine J. Webb R: Resection of the rectal mucosa, colectomy, and anal ileostomy with normal continence. Surg Gynecol Obstet92: 437, 1951. 7. Levin P, Fonkalsrud EW: Fecal continence following colectomy
Volume 136, July 1976
8.
developed 9.
10.
11.
12.
13.
lleal Pullthrough
and ileoanal anastomosis through extra mucosal rectal tube. J Sur!g Res 8: 234, 1968. Martin L., LeCoultre C, Schubert WK: Total colectomy and mucosal proctectomy with preservation of continence in ulcerative colitis. Ann Surg 186: 477, 1977. Aylett SO: Three hundred cases of diffuse ulcerative colitis treated by total colectomy and ileorectal anastomosis. 6r Med./ 1: 1001, 1966. Binder SC, Miller HH, Deterling RA Jr: Fate of the retained rectum after subtotal colectomy for inflammatory disease of the colon. Am JSurg 131: 201, 1976. Kock NC? Intra-abdominal “reservoir” in patients with permanent Ileostomy: preliminary observations on a procedure resulting in fecal “continence” in five ileosiomy patients. Arch Surg 99: 223, 1969. Beahrs OH, Kelly KA, Adson MA, et al: lleostomy with ileal reservoir rather than ileostomy alone. Ann Surg 179: 634, 1974. Handelsman JC, Fishbein RH: Permanent ileostomy without extemal appliance: Koch internal reservoir (pouch) operation. Johns Hopkins MedJ 138: 161. 1976.
Discussion Alan Morgan (Seattle, WA): The authors have made an important contribution to the surgical treatment of ulcerative colitis and other nonmalignant diseases of the colon and rectum. Surgeons and patients will await follow-up results with interest, and the patients almost demand them, for in no other field of surgery do patients play such an integral role in the selection of their operation. Credit must be given to Drs. Ravitch [5] and Sabiston of the Hunterian Laboratory at the Johns Hopkins Hospital for the idea of an endorectal pullthrough and ileoanal anastomosis. They believed that only the mucosa need be removed in benign conditions and they stripped the mucosa up from below. The Soave [4] operation, stripping out the rectal mucosa in Hirschprung’s disease from above, as the authors did, came eighteen years later, in 1966. Those who care for children will know that this is really not too difficult, especially when the rectum has been defunctioned. But is the stripping out difficult in ulcerative colitis? Does the diseased, friable mucosa come out easily? I suspect that a number of patients with a severely involved rectum will not be suitable for the operation even if it is staged. I hope that more patients are suitable for this new procedure than are suitable for Aylett’s ileorectal anastomosis [9]. It could be that a good safe stripping is limited by the same factors that limit a good ileorect,al anastomosis in the hands of the majority of surgeons. Therefo:re, many will find this an attractive concept, particularly those who are not happy with the Koch pouch, for some of the technical aspects which can defeat the pouch, particularly the intussuscepting nipple, are not necessary. The patient has a good continence mechanism already made by nat.ure waiting to control the pullthrough. The colon and rectal surgeon is concerned about the disease in the mucosa, and the pediatric surgeon is concerned about the problems inherent in the pullthrough operations, namely structures at the suture line just above the dentate line and at the upper end of the rectal muscular cuff and abscess formation in the cuff. Staging t.he proce-
119
Ferrari and Fonkalsrud
dure will reduce the risk of the latter, and certainly we all agree that these procedures should be staged. William B. Hutchinson (Seattle, WA): A successful ileoanal anastomosis was performed in 1947 on a twentyfive year old female with chronic ulcerative colitis and was reported before this Association in Coronado in 1951. So successful was the result that after the death of her first husband she married a second. She had normal sexual desires and function, the ability to differentiate gas and feces, and normal bladder function. She was able to constrict on the examining finger and control her feces very well. This took about a three month training period. I therefore operated on three other patients with chronic ulcerative colitis, and all of these ended in total failure, necessitating reestablishment of their anal ileostomy to an abdominal ileostomy. They were able to control their sphincters, having good contraction of their sphincter mechanism, but were totally unable to differentiate between gas and feces, were incontinent of stool, and were tremendously excoriated. Not one of these patients had any disability in anal contraction. They had no fistulas and no infections to account for the incontinence. Is it possible that in the dogs shown here, the horizontal posture of the dog as opposed to the perpendicular posture of man made a difference in the results of their continence? This was an excellent presentation, and I hope the authors are successful in overcoming the problem of incontinence. I can assure you that there is nothing more intolerable than an incontinent anal ileostomy with massive excoriation in an emotional nervous patient. Wiley F. Barker (Los Angeles, CA): I would like to present some soft information from the same population of patients which may, at least in a subliminal way, have led Dr. Fonkalsrud to perform this procedure. Over the last twenty years we have carried out about twenty-five ileorectal anastomoses, and most of them have failed. In about fifteen of these patients we constructed an inverted 9 loop, that is, we carried out side-to-end ileorectal anastomosis 15 cm from the divided end of the ileum, and then performed end-to-side anastomosis between the end of the ileum and the side of the ileum, 15 cm back from the anastomosis, thus creating an “inverted 9.” Most of these patients had a good functional result with this as a pouch, but all but one have ultimately failed primarily because of the recurrence at the distal anastomosis. As reported elsewhere, we thought that the procedure using the rectum with its intact mucosa for an ileorectal anastomosis failed commonly in a period of time inversely proportional to the duration and severity of the disease, rather than because of the actual pedigree of the disease. I also believe that the common failure of the usual ileoanal anastomosis has been due to the imposition of what amounts to intestinal obstruction on the small bowel if the anal sphincter is truly competent. I believe that Drs. Ferrari and Fonkalsrud have indeed introduced a means of first, eliminating the primary site of recurrence as well as second, relieving the practically obstructing effect of the
120
sphincter blunts the I predict but it will
by this damping effect of the reservoir which onrush of intestinal peristalsis. that there will be problems with this operation, be a useful operation in the future.
Donald M. Gallagher (San Francisco, CA): I too would like to share the enthusiasm about this operation, but I suggest caution. First, there was an implication that the operation being discussed is similar to the Kock continent ileostomy. Continence in the Kock procedure is dependent upon the ileostomy nipple acting as a mechanical obstruction to the reservoir. The obstruction of the nipple is overcome only when a catheter is passed through the ileostomy to the reservoir. The reservoir is important, but it is the construction of the nipple that determines the success or failure of the Kock procedure. The human rectum does not have the ability to act as an obstructing ring for a protracted period of time. Second, recently in doing low anterior anastomoses with the endoanal stapler in dogs I have been impressed by the difference of the muscular coat of the dog’s rectum compared with that of humans. In the dog the muscle thickness of the rectum is at least fivefold greater than in humans. Also, in the dog there is ease of complete exposure of the rectum, which is not true in man. Thus, the dog is not applicable as a surgical model for the type of surgery that might be done in humans. Eric W. Fonkalsrud (closing): As indicated, the described procedure is easy to perform in a dog with a normal rectum, but is difficult in a patient who has mucosal ulcerations and severe inflammation. The permanent cutaneous ileostomy has been frustrating to many young patients who may have to live with an appliance for as long as fifty to sixty years. The procedure described is not applicable for patients with Crohn’s disease since the ileal reservoir, just as a Kock pouch, may develop recurrent inflammation. If the reservoir does not function adequately, only approximately 39 cm of ileum might have to be removed, leaving the patient with an ileostomy as would be the case had the reservoir not been constructed. One of the key features of the operation is careful preservation of the blood supply to the terminal ileum during construction of the reservoir. The nipple is the key feature for preservation of continence in the Kock pouch; however, the patient who has an intact anorectal sphincter will experience the sensation of fullness and will be able to defecate when desired without a nipple. We have applied the operation to one patient with ulcerative colitis. The ileum was closed proximal to the pouch and a diverting ileostomy was constructed. The ileostomy was closed two and a half months later, and the patient is currently continent and able to distinguish air from fluid when he defecates, an advance over our previous experience with endorectal ileal pullthrough without a reservoir. Nonetheless, much more experience with the procedure’s technical features is necessary before further clinical application can be recommended.
The American Journal of Surgery
