Redo pouches

Redo Pouches: Salvaging of Failed Ileal Pouch-Anal Anastomoses G i l b e r t o P o g g i o l i , M . D . , F l o r i a n o M a r c h e t t i , M.D., S i m o n e t t a S e l l e r i , M.D., Silvio Laureti, M.D., Luca Stocchi, M.D., Giuseppe Gozzetti, M.D., F.A.C.S.

From the Clinica Chirurgica II, University of Bologna, Bologna, Italy From October 1, 1984 to December 31, 1991 at t h e Clinica Chirurgica II of the University of Bologna, 140 patients submitted to ileal pouch-anal anastomosis for ulcerative colitis (UC) and familial adenomatous polyposis (FAP). Nineteen patients (13.5 percent) d e v e l o p e d septic complications. Of these, 11 patients (7.8 percent) had pelvic sepsis. Eight patients required further surgical intervention. Five patients underwent the redo p o u c h procedure. Another redo p o u c h was performed in a patient who had previously, in another hospital, had an ileal pouch-anal anastomosis placed and then removed because of ischemic necrosis of the reservoir. No deaths are reported in the reoperated patients. Currently, five of the six patients who underwent the redo p o u c h p r o c e d u r e have a well-functioning ileoanal anastomosis. The redo p o u c h p r o c e d u r e should always be attempted prior to the establishment of pelvic fibrosis. [Key words: Ileal pouchanal anastomosis; Pelvic sepsis; Redo pouch]

MATERIALS A N D M E T H O D S From October 1, 1984 to December 31, 1991, 140 patients submitted to an ileoanal procedure: 122 for UC and 18 for FAP. In all patients with FAP, mucosectomy and a hand-sewn anastomosis were carried out. Of the remaining 122 patients who underwent surgery for UC, 56 had a hand-sewn anastomosis fashioned and 68 had a stapled anastomosis. Nineteen patients (13.5 percent) developed septic complications. Of these, eight (5.7 percent) had intra-abdominal sepsis located well above the sacral promontory. All patients recovered, seven with surgery and one with antibiotic therapy. Of the 11 patients with pelvic sepsis (7.8 percent), three were successfully treated with antibiotics and/or lavages. The remaining eight patients (5.7 percent) had pelvic sepsis complicated with a fistula of the reservoir and required further surgical intervention. Four patients submitted to a redo pouch procedure, in three cases successfully with a well-functioning pouch. One case was later converted to a permanent ileostomy. The remaining patients with complicated pelvic sepsis were repeatedly treated with surgery and eventually developed pelvic fibrosis. All of them were converted to a permanent ileostomy. In one case, it was only temporarily possible to close the ileostomy because of the poor functional results that ultimately compelled the authors to remove the reservoir. Overall, five patients (3.5 percent) had their reservoir removed.

Poggioli G, Marchetti F, Selleri S, Laureti S, Stocchi L, Gozzetti G. Redo pouches: salvaging of failed ileal pouchanal anastomoses. Dis Colon Rectum 1 9 9 3 ; 3 6 : 4 9 2 - 4 9 6 .

I leoanal anastomosis is currently the surgical treatment carried out worldwide for ulcerative colitis (UC) and familial adenomatous polyposis (FAP). The operation radically cures both diseases and provides excellent quality of life. Despite this, the procedure has a higher rate of surgical complications than do proctocolectomy with permanent ileostomy and ileorectal anastomosis. In the literature, the rate of complications decreases as surgical experience improves. Nevertheless, pelvic sepsis remains a dreadful complication because of its evolution into pelvic fibrosis, which leads to ileal pouch failure. The presence of complicated pelvic sepsis does not mandate the removal of the reservoir and the placement of a permanent ileostomy. It is possible in some cases to reconstruct the reservoir and perform the so-called "redo pouch" procedure. The authors report on six redo pouches. Their personal experience and a review of the literature define the possibilities and limits of this procedure.

Redo Pouches: Personal Histories

Case 1. Approximately two weeks postoperatively, this 23-year-old female presented with intermittent cramps and fever. A contrast x-ray showed a rotation of the pouch around its longitudinal axis. That in turn brought about ischemia and a fistula of the upper part of the pouch. We proceeded to the resection of this part and then carried out a small J-pouch with the ileum proximal

Address reprint requests to Dr. Gozzetti: Clinica Chirurgica II, Policlinico S. Orsola, Via Massarenti 9, 40138 Bologna, Italy.

492

Vol. 36, No. 5

REDO POUCHES IN FAILED ILEOANAL ANASTOMOSES

493

to the reservoir (Figs. 1A and B). Finally, a new anastomosis between the new pouch and the remaining old one anastomosed to the anus was fashioned (Figs. 2A and B). Five years postoperatively, the patient is in good health and the pouch is functioning well. Case 2. This 42-year-old female developed post-

Figure 2. A. Resection of the upper part of the reservoir and construction of a small, new pouch. Pouch-to-pouch anastomosis. B. Postoperative x-ray.

Figure 1. A. Rotation of the pouch on its longitudinal axis. B. Defecography with dilatation of the upper part of the reservoir rotated on its longitudinal axis.

operative fistulas from the afferent limb to the pouch that evolved into fibrosis and, in turn, into a stricture. The patient was reoperated on and submitted to resection of the afferent limb and the upper part of the reservoir. Finally, a new end-toend anastomosis was carried out. Pelvic fibrosis, which was already established at the time of the first surgery, and a new anastomotic leak after the last procedure led to the removal of the pouch. Case 3. The third patient, a 38-year-old male, developed a postoperative pouch-urethral fistula (Fig. 3). Conservative treatment with an indwelling Foley catheter was started for some months. After an apparent recovery of the fistula, the patient had to undergo surgery. With an abdominoperineal approach (Fig. 4), a fistulectomy was carried out along with urethroplasty with the o m e n t u m as well as resection of the terminal part of the reservoir. The procedure was then completed with an anastomosis between the remaining pouch and the anal canal (Fig. 5). Three years postoperatively, the patient is in excellent condition with a well-functioning pouch. Case 4. This 28-year-old female, who had previ-

494

POGGIOLI ET AL

Dis Colon Rectum, May 1993

Figure 3. Pouch-urethral fistula (urethrography).

Figure 5. Pouch-anal reanastomosis.

Figure 4. Omentoplasty of the urethra. ously submitted to a stapled ileoanal pouch, developed a pelvic abscess with multiple fistulas of the reservoir that mandated the removal of the pouch. A new J-pouch was carried out using the small bowel, and a pouch-anal anastomosis was handsewn. After two years, she is in good condition. In the following two cases of redo pouch, the indications for surgery were other than pelvic sepsis. Case 1. This 28-year-old male, who had submitted to ileoanal anastomosis at another hospital, developed ischemic necrosis of the reservoir and approximately 40 cm of the ileum proximal to it. The patient was then reoperated on and the reservoir resected above the levators' plane. A terminal ileostomy was fashioned. After a difficult postoperative period characterized by diffuse abdominal sepsis and further complicated by renal failure that required repeated dialysis sessions, the patient came to our observation. Since the sphincteric function appeared to be well maintained, as proven by.anal manometry, we carried out an ile0anal

anastomosis utilizing the remaining ileum. In this case it was particularly difficult to take the ileum down to the pelvis in order to reach the anus without tension. A small S-pouch of about 9 cm in length w a s - t h e n assembled, and a hand-sewn pouch-anal anastomosis was carried out. Currently, the patient is in good condition and is waiting for closure of the ileostomy. Case 2. The last redo pouch was carried out for a very unusual indication in an 18-year-old male. In fact, the patient had previously undergone a hand-sewn ileoanal anastomosis for FAP. After closure of the loop ileostomy, he complained of difficulties in emptying the pouch. The patient appeared to be suffering from a semiobstructing syndrome, which presented with evacuations of very small amounts of feces several times a day (a sort of evacuation by regurgitation). A barium enema of the reservoir revealed impressive dilatation of the reservoir that reached the transverse umbilical line. Scintigraphic defecography using a watery paste of hydrosilicate of aluminum magnesium at a concentration of 7.5 percent marked with Tc99 showed a postevacuation residual that amounted to 65 percent of the administered contrast medium.

Vol. 36, No. 5

REDO POUCHES IN FAILED ILEOANAL ANASTOMOSES

Consequently, a surgical reduction of the reservoir took place. At surgery the bottom of the ileal pouch appeared to extend to the aortic bifurcation. This reduction, carried out using a stapler (Fig. 6), amounted to about 10 cm of the distal part of the reservoir, so that a new and smaller J-shaped reservoir remained. Three years postoperatively, the patient has four bowel movements per day. DISCUSSION For many years the high rate of surgical complications has limited the universal recognition of the ileoanal pouch procedure as the therapy of choice for UC and FAP. In fact, even in the larger series, the complication rate is higher than those of the ileorectal anastomosis and the total proctocolectomy with permanent ileostomy. 1' 2 Nevertheless, the centers with greater experience have complication rates that are remarkably less and appear to decrease in a linear way compared with the number of procedures Carried out. Dividing our patients into subgroups of 10, we observed a decrease in incidence of pelvic sepsis

Figure 6. A stapler reduction of a large pouch.

495

from 30 percent in the first 20 patients to 0 percent in the last 20 operations. Overall, pelvic sepsis occurred in 7.8 percent of the 140 cases and led to the conversion of the ileoanal anastomosis into a permanent ileostomy in 3.5 percent. Finally, if we eliminate the data of the first 20 patients, we can see how, in the remaining 120 patients, pelvic sepsis decreased to 3.1 percent while the need to convert to an ileostomy fell to 0.7 percent and the success rate of these patients rose to 99.3 percent. In the initial group of 20 patients (the data are mentioned only to emphasize these reports without a statistical comparison), there is a 30 percent incidence of pelvic sepsis, and the percentage of removal of the reservoir is 15 percent. Reconstruction of the reservoir restores pouch function and appears to be more successful than repeated toilet cleaning and drainage of the pelvis, which may resolve the sepsis but inevitably leads to pelvic fibrosis. In our experience, eight patients with pelvic sepsis were reoperated on. Three of the four patients who underwent a redo pouch operation for pelvic sepsis had their reservoir salvaged, as did the two patients who underwent surgery for the other conditions mentioned above. The other four patients with pelvic sepsis underwent minimal procedures to cure the sepsis; yet, in all cases, removal of the pouch was inevitable because of pelvic fibrosis. In the group of eight patients who presented with pelvic sepsis, six had UC and two had FAP. Redo pouch operations were carried out in three cases of UC and in one case of FAP. Of the five patients who had their pouch removed and underwent conversion to a permanent ileostomy, four (80 percent) had UC. The one patient who had FAP and underwent conversion to a permanent ileostomy had undergone an ileorectal anastomosis nine years previously. This is in accordance with what was reported by Dozois. 3 That report showed that removal of the reservoir in patients with UC was 3.2 percent, vs. 0 percent in patients affected by FAP. The rate of reoperation for pelvic sepsis complications was 6 percent in patients with UC and 0 percent in those with FAP. In a recent article, Galandiuk and colleagues 4 reported on a series of nine redo pouches and three conversions to a different type of pouch. Five of these procedures were carried out because of the presence of sepsis and fistulas. The remaining cases were included in a heterogeneous group

496

POGGIOLI E T AL

called "nonsatisfactory results." It is n o t e w o r t h y that 29 patients with a b d o m i n a l sepsis c o m p l i c a t e d with fistulas submitted to reoperation. Of these 29, only five patients submitted to r e d o pouch. Thirtyfour p e r c e n t of this group u n d e r w e n t conversion to a p e r m a n e n t ileostomy. This article 4 leads to the same c o n c l u s i o n that we r e a c h e d in our e x p e r i e n c e . Reservoir-salvaging surgery should not be a i m e d only at controlling sepsis but should be m o r e aggressive w h e n the situation is strongly c o m p r o m i s e d . A r e d o p o u c h p r o c e d u r e should be a t t e m p t e d prior to the establishment of pelvic fibrosis. Despite the fact that Scott e t al. 5 have stated that pelvic sepsis does not affect anal continence, we believe that this is true o n l y before the consolidation of pelvic fibrosis makes it impossible for "the reservoir to function as a reservoir."

Dis Colon Rectum, May 1993

REFERENCES 1. Metcalf AM, Dozois RR, Kelly KA, Beart RWJr, Wolff BG. Ileal "J" pouch-anal anastomosis: clinical outcome. Ann Surg 1985;202:735-9. 2. Williams NS, Johnston D. The current state of mucosal proctectomy and ileo-anal anastomosis in the surgical treatment of ulcerative colitis and familial polyposis. R J Surg 1985;72:159-68. 3. Dozois RR. Pelvic and perianastomotic Complications after ileoanal anastomosis. Perspect Colon Rectal Surg 1988;1:113-21. 4. Galandiuk S, Scott NA, Dozois RR, et al. Ileal pouchanal anastomosis: reoperation for pouch-related complications. Ann Surg 1990;4:446-54. 5. Scott NA, Dozois RR, Beart RW Jr, Pemberton JH, Wolff BG, Ilstrup DM. Postoperative intra-abdominal and pelvic sepsis complicating ileal pouch-anal anastomosis. Int J Colorectal Dis 1988;3:149-52.