- Email: [email protected]
SURGICAL OPTIONS IN ULCERATIVE COLITIS
NEW AND CONTROVERSIAL ISSUES IN THE MANAGEMENT OF COLORECTAL DISEASES
0039-6109/97 $0.00
+ .20
SURGICAL OPTIONS IN ULCERATIVE COLITIS Ridzuan Farouk, MCh, FRCSGlasg, FRCSEd, and John H. Pemberton, MD
Ileal pouch-anal anastomosis (IPAA) has become an established surgical option for the treatment of chronic ulcerative colitis (CUC) and familial adenomatous polyposis (FAP). Much has been written about the procedure, and it is not the intention of this article to reiterate what has been published or to review the literature comprehensively. Instead, this article outlines controversies around which debate still centers. Specifically, the following issues are discussed:
1. Which pouch design is best? 2. Is there a role for defunctioning ileostomy? 3. When should the anal transition zone be excised? 4. When should a stapled anastomosis be avoided? 5. What is the influence of finding a cancer of the colon on the decision to create a pouch? WHICH POUCH DESIGN IS BEST?
The design of the pouch advocated by the earliest proponents of IPAA was a three-limbed reservoir shaped like an S. This pouch was hand sewnz7The advent of the stapler, together with ingenious thinking, has resulted in a multitude of options for constructing pouches, including both hand-sewn and stapled techniques. The variety of pouch configurations now includes the J,38 H,7 W,24 and K." A central concern with the S pouch has always been difficulty with stool evacuation, which had often required self-catheterization especially during the early experience. A long efferent limb was usually blamed for this problem because as the limb was subsequently shortened, so was the incidence of evacua-
From the Mayo Graduate School of Medicine and Division of Colon and Rectal Surgery, Mayo Clinic, and Mayo Foundation, Rochester, Minnesota
SURGICAL CLINICS OF NORTH AMERICA
-
VOLUME 77 * NUMBER 1 FEBRUARY 1997
85
86
FAROUK & PEMBERTON
tion difficulties reduced. The J configuration appears to overcome this problem, but its smaller reservoir size is theoretically associated with reduced early compliance and increased bowel f r e q u e n ~ y The . ~ ~ W and K pouches are larger than the J pouch and do not have the associated evacuatory problems of the S 24 A comparative study on the manovolumetric characteristics and function in patients with the S-, J-, and K-pouch found that, despite use of equal lengths of terminal ileum to construct the pouch with similar initial pouch volumes (120 to 130 mL), the S pouch and the K pouch expanded more favorably, with a maximal volume at 1 year of 420 and 410 mL, respectively, compared with 305 mL for the J pouch." In practice, however, the overall functional results have differed little among the three pouch designs. The W pouch has been described as having the greatest volume for any given length of ileumz4and therefore, theoretically at least, provides greater pouch volume than other configurations. Merely increasing the length of ileum used for construction of the J pouch also improves pouch volume. This translates into a comparable functional outcome for J and W configurations when equal lengths of ileum are A recent comprehensive physiologic assessment comparing W and J pouches has shown that when equal lengths of ileum are used, the manometric, sensory, and compliance data are similar, the only difference being a higher tolerated filling pressure before the sense of urgency occurred in W pouches.* Much investigation has attended these speculations on pouch configuration and its relation to functional outcomes. These studies have failed, however, to address other important issues such as the role of the anal sphincters, stool consistency, altered motility patterns, and the capacity of the pelvis to accommodate the reservoir. In conclusion, it appears there is little difference in functional outcome between the S, W, J, K, and H reservoirs if approximately 30 to 40 cm of terminal ileum are used for reservoir construction. When the pouch has difficulty reaching the anal canal for the anastomosis, the S configuration and its efferent limb have the advantage of achieving better length, thus facilitating an anastomosis without tensi0n.3~ The relative ease in constructing a stapled J pouch, combined with comparable functional outcomes to more complex pouch designs, has made this the favored option at the Mayo Clinic. THE ROLE OF A DEFUNCTIONING ILEOSTOMY The most feared complication of IPAA is pelvic sepsis, and therefore a defunctioning ileostomy after pouch construction is usually performed to minimize its occurrence. Whereas pelvic sepsis complicates 6% of patients undergoing IPAA at the Mayo Clinic,3l the rates reported in the literature vary between 0% and fully 25y0.~. 8, 39 Moreover, disturbingly high rates of pelvic sepsis have been reported in patients undergoing a one-stage procedure (no ileostomy).6 Although the incidence of pelvic sepsis is low, when it occurs it is responsible for a significant number of the failed pouches in the Mayo series. Galandiuk and colleagues8 reported 114 patients who had pouch-related complications requiring reoperation. Of these, 59 (52%) had septic complications (Table 1). Fifty-nine percent of the intra-abdominal sepsis group required further surgery, with 34% (10 patients) eventually undergoing pouch excision. An additional seven patients at follow-up had a diverting stoma. Five of the 30 patients with perianal sepsis (17%) subsequently underwent pouch excision, and a further five had a diverting stoma at the time of review. Protagonists of defunctioning
SURGICAL OPTIONS IN ULCERATIVE COLITIS
87
Table 1. POSTOPERATIVE POUCH FUNCTION AFTER SALVAGE OPERATION FOR
SEPSIS Presenting Problem
Outcome
lleostomy closed Anal pouch function satisfactory Anal pouch function poor lleostomy present Pouch in situ Pouch excised
Perianal FistuldAbscess I n = 30 Datients)
Intra-abdominal FistuldAbscess In = 29 Datients)
45% 21%
35% 7%
17% 17%
24% 34%
From Galandiuk S, Scott NA, Dozois RR, et al: Ileal pouch-anal anastomosis: Reoperation for pouch related complications. Ann Surg 212:446454, 1990; with permission.
ileostomies also argue that diverting stomas allow the anal sphincter and ileal mucosa to recover before restoration of intestinal continuity, and that patients have a short-lived experience of a stoma to fully appreciate the ultimate benefit of IPAA.3,39 Use of a loop ileostomy does not appear to fully protect the patient from pelvic sepsis9 It is nonetheless much easier to manage a patient with sepsis if an ileostomy is in place.9Of Mayo patients who required laparotomy to control sepsis, 41% lost the pouch ultimately and only 29% ever recovered ileoanal function. However, if no reoperation was required, 92% of patients with sepsis eventually had a functioning pouch.” This is a sobering observation. There is little doubt that by avoiding an ileostomy altogether, complications of closure are not an issue, and only one hospital admission is required. Low complication rates with good functional outcome have previously been reported with such an approach.”, 36 At the Mayo Clinic, the clinical and functional outcomes of 37 patients who had a one-stage IPAA without a defunctioning ileostomy have previously been reported,’ and the group was age- and sexmatched with 37 patients who had a defunctioning ileostomy during the same period. The results of this study are summarized in Tables 2 and 3. The incidence of complications was higher when a defunctioning stoma was not used (22% versus 11%,respectively). One-stage IPAA was shown to be a reasonable option
Table 2. POSTOPERATIVE COMPLICATIONS AFTER IPAA: DEFUNCTIONING
ILEOSTOMY VERSUS NO ILEOSTOMY
No lleostomy (37 patients)
Presacral abscess Anastomotic dehiscence Pouch bleeding Fever Jejunal perforation Late presacral abscess lleostomy at follow-up
-
2 5 1 1 1
lleostomy (37 patients)
3 1 2 3
Data from Galandiuk S, Wolff BG, Dozois RR, Beart RW Jr: Ileal pouch-anal anastomosis without ileostomy. Dis Colon Rectum 34:870-873, 1991.
88
FAROUK & PEMBERTON
Table 3. POSTOPERATIVE FUNCTIONAL RESULTS: DEFUNCTIONING ILEOSTOMY (n = 37) VERSUS NO ILEOSTOMY (n = 37) Night
Day
No lleostomy Continence (%) Perfect Spotting Mean stool frequency
79 21 5.0
lleostomy
No lleostomy
94 3 6.0
70 21 1.o
lleostomy
35 52 1 .o
Data from Galandiuk S, Wolff BG, Dozois RR, Beart RW Jr: Ileal pouch-anal anastomosis without ileostomy. Dis Colon Rectum 34:870-873,1991.
if patients were in good health and were not on chronic steroid therapy and if the IPAA could be constructed without tension or technical difficulty. A subsequent randomized trial of 45 CUC patients who were randomized to defunctioning stoma (n = 23) or no stoma (n = 22) showed no differences in rates of complications between the groups.'O The patient group was highly selected, however, in that no patients were taking steroids. Although the authors found no increased risk of pelvic sepsis in patients without an ileostomy, the study was notable for the rather high rate of ileostomy-related complications (52%). A similarly high rate of ileostomy-related complications was described by Winslett and associatesNin a prospective evaluation of 34 patients who had a loop ileostomy. Formation and closure of the ileostomy were associated with morbidity rates of 41% and 30%, respectively. Obviously, there are savings related to shorter operating time and hospital stays in patients having no diverting ileo~tomy.~ A reasonable approach to this dilemma is to use a defunctioning ileostomy in those patients receiving steroid treatment at the time of surgery and in patients who are nutritionally compromised or undergoing an urgent operation. Additionally, if there are concerns about pouch blood supply or anastomotic tension, a diverting stoma is almost mandatory. Using these criteria, 56 of 1800 patients undergoing IPAA have had a one-stage procedure performed at the Mayo Clinic between 1980 and 1996 (unpublished data). WHEN SHOULD THE ANAL TRANSITION ZONE BE EXCISED?
The majority of patients who undergo a pouch procedure have an average of six or seven pouch evacuations per 24 Fecal incontinence is reported in 4% of patients, with nocturnal spotting occurring in 25%.= Stapled anastomosis to the top of the anal canal without excision of the anal transition zone (ATZ) has, in nonrandomized trials, been associated with better functional outcome.'2,16On the other hand, preservation of the ATZ leaves a small amount of rectal mucosa intact.' The debate is drawn as follows: Leaving the ATZ intact improves functional outcome but leaves rectal mucosa at risk of continued inflammation and neoplastic change. CUC and FAP are diseases of the colonic and rectal mucosa. The IPAA procedure is intended to remove all diseased mucosa, thereby eliminating the risks of neoplasia and recurrent mucosal disease. This view has been opposed
SURGICAL OPTIONS IN ULCERATIVE COLITIS
89
by some who argue that the risk of neoplasia in retained segments of rectal mucosa is infinitesimally small and is offset by improved functional outcome. Endoanal mucosectomy has been associated with poorer functional probably because the anal canal is necessarily dilated during the operation. In practice, however, the ability to discriminate gas from feces and the overall functional outcomes have not been shown to correlate with the presence or absence of the ATZ.2 In a study of 50 patients with CUC who had undergone proctocolectomy for their disease, columnar epithelium was found to extend through 50% of the length of the ATZ in 75% of specimens and through 75% of the length in 46% and to approximate the dentate line in 9%.' The degree of inflammation affecting the mucosa was significant compared with that of a control group of specimens from patients with rectal cancer. The study concluded that the ATZ is composed of transitional epithelium and columnar epithelium. Although the transitional epithelium may be free of inflammation, the columnar epithelium is usually inflamed. Therefore, preserving the ATZ would have preserved a focus of disease in most patients with CUC.', 4, 35, 37 Although the risk of malignancy in patients with long-standing CUC is increased but small, in patients with FAP the malignancy risk is 100% by the age of 60 years. The ATZ should therefore be excised in all FAP patients. Because of the risks of leaving diseased mucosa behind and the risks of malignant degeneration, at the Mayo Clinic all patients with FAP undergoing IPAA have the ATZ excised. In patients with CUC, if the ATZ is not excised, life-long surveillance is advised. A focus of disease may still be present even after mucosectomy. We found that in patients who undergo ATZ excision, up to 7% have residual microscopic columnar epithelium.', 13, 2h The clinical significance of this finding remains unknown. The risk of malignant degeneration of such nests 2fl, 34. 37 of rectal mucosa is admittedly small but WHEN SHOULD A STAPLED ANASTOMOSIS NOT BE USED?
Stapling the ileum directly to the anal canal would destroy the internal anal sphincter. Thus, the stapled anastomosis is instead constructed at the top of the anal canal and the ATZ is preserved. This leaves a small remnant of columnar epithelium with the risk of continuing inflammation as well as a risk of future malignant transformation.' Much of the debate about whether to staple the anastomosis or not has really evolved because of the debate about functional outcome. Does preserving the ATZ enhance continence after the IPAA? Stapled anastomosis in nonrandomized trials has been equated with better outcome, which in turn is attributed to less injury to the anal sphincters, with preservation of the ATZ and hence anal sensory discrimination, and with preservation of the rectoanal inhibitory reflex.I2, 14, 16, 17, 23, 3fl However, randomized trial by Seow-Choen et a132comparing endoanal mucosectomy and hand-sewn anastomosis with stapled anastomosis has not shown any difference in functional outcome between the two methods. These findings were not entirely surprising and indirectly confirmed those of other authors: Horgan et all5 described a reduction in anal pressures after low anterior resection, which was in turn attributed to dilatation of the anal sphincters by simply inserting stapling guns into the anus. Read and Readz9 found no deterioration of continence during rectal infusion of saline when anal canal sensation was reduced by the use of topical anesthetic ointments. The retention or recovery of the rectoanal inhibitory reflex, reported in some
90
FAROUK & PEMBERTON
resembles the observations of others in which there was no attempt to excise the most distal cuff of rectum.21Retention of the rectoanal inhibitory reflex is therefore being attained by leaving mucosa at risk of persistent disease as well as malignant degeneration. These clinical impressions have been supported by a limited amount of anorectal physiology data that are available in the literature. Bartolo and Duthiez reported no difference in functional outcome based on the presence or absence of the ATZ. Instead, patient age, pouch compliance, and abnormal pouch motility patterns were more relevant in determining functional outcome. Anastomotic stricturing may occur with either hand-sewn or stapled techniques and affects up to 30% of patient^.^ There appears to be a higher incidence with the double-stapling technique, however, particularly with the use of smaller-caliber stapling guns. However, these strictures are thin, easily dilated, and transient in nature. Forty-one patients at the Mayo Clinic have been randomized to doublestapled (20 patients) or hand-sewn (21 patients) techniques (unpublished data). In the stapled group, 1.5 to 2.0 cm of the ATZ was preserved, whereas a complete mucosectomy was performed in the hand-sewn group. Twenty of these patients subsequently underwent ambulatory manometry (11 stapled, 9 hand-sewn) 6 months after closure of the defunctioning ileostomy. Mean anal canal pressures and the number of episodes when pouch pressure exceeded anal pressure were similar between the two groups (Table 4). We found that parameters of anal canal and pouch motor function did not differ among patients randomized to double-stapled versus hand-sewn IPAA. Thus, stapling and mucosal preservation appear to confer little if any short-term functional advantage. The final decision as to when to staple without mucosectomy and when to do a mucosectomy and hand sew the anastomosis must not be confused with what is safest and best for the patient. Patients at risk of malignancy should have a mucosectomy to completely eliminate this risk. This should include those patients with CUC with evidence of mucosal dysplasia as well as all patients with FAP. WHAT IS THE INFLUENCE OF FINDING A CANCER ON THE DECISION TO CREATE A POUCH?
Although ileal pouch-anal anastomosis has been aimed primarily at eradicating mucosal disease in patients with CUC and prophylactically removing the Table 4. A COMPARISON OF ANAL POUCH PHYSIOLOGY OUTCOME BETWEEN HAND-SEWN AND STAPLED IPAA PATIENTS Pressures (mm Hg)
Mean anal canal pressures Pressure pouch > pressure anal canal (episodedhr)
Stapled Awake 107
Asleep 96 0 (0-12)
Hand-sewn
P Value
Awake Asleep 119 121 1(0-5)
Awake Asleep 0.6 0.4 0.5
SURGICAL OPTIONS IN ULCERATIVE COLITIS
91
neoplastic risk in patients with FAP, there has been an increasing experience with CUC or FAP patients presenting with cancer and undergoing IPAA. Because some patients need adjuvant treatment for advanced cancer, we wished to determine if IPAA should be performed in patients with concomitant cancer and CUC or FAP. Between 1980 and 1995, we identified 150 patients with cancer complicating CUC or FAP presented to the Mayo Clinic, Rochester (unpublished data). Seventy-four of these patients had an IPAA. There were 81 tumors present in these 74 patients. The majority of these were of a relatively early stage, with just 13 Astler-Coller stage C, or Cz tumors. The mean age of these patients was only 32 years, with a range of 15 to 60 years. The mean follow-up was 6 years. There were 11 deaths. Nine were due to systemic progression of the cancer, and 1 patient had local failure with systemic progression of disease. Just 1 patient died of sepsis-related complications 4 months postoperatively. In 8 patients, the ileostomy was permanent; 6 had progression of their disease, and 2 had irreversible pouch injury secondary to radiation therapy. In addition, 5 pouches were excised-2 patients for cancer-related complications, 2 for ischemia, and 1 for intractable pouchitis. Twenty-eight percent of patients have minor nocturnal spotting, 21% require use of constipating medications, and the mean number of pouch evacuations per 24 hours is 6 (range 4 to 13). These functional data are identical to the long-term outcome reported in patients with CUC without colorectal cancer.22 The rate of pouch failure was nearly double that of patients undergoing IPAA for CUC alone.,, Virtually all of the failures (mortality, permanent ileostomy, or pouch excision) occurred in patients with progression of their cancer or secondary to radiation therapy injury (12 of 15 patients). On the basis of these findings, patients with advanced disease (Astler-Coller stage C, or C,) or those who may require adjuvant radiation therapy should not be considered for an IPAA. Instead, an anus-sparing procedure such as ileostomy and preservation of a short rectal stump should be performed and a pouch reconsidered 2 to 3 years later if there is no evidence of radiation injury or recurrent disease. Alternatively, proctocolectomy with a Brooke ileostomy is an excellent choice for these complex patients. If the presence of a cancer is known preoperatively, the patient can be counseled about the possibility of a long-term ileostomy. Unsuspected cancers discovered only during surgery pose a separate problem, however. Before proceeding with pouch construction, the colectomy specimen must be opened and inspected for any suspicious mucosal lesions and frozen-section analysis obtained. If a cancer is found under these circumstances, the decision to proceed with pouch formation must take into consideration location of the cancer (colon versus rectum), the likelihood of transmural penetration, and the possibility of nodal involvement. If a rectal cancer penetrates transmurally or has lymph node metastases or if a colon cancer has nodal involvement, pouch construction should be deferred until adjuvant therapy has been completed and a period of at least 2 years has elapsed to ensure that the patient is free of cancer. SUMMARY
The length of ileum used rather than pouch configuration per se is related to eventual functional outcome: A pouch constructed from 15- to 20-cm limbs is ideal. One-stage procedures appear feasible in those patients who are not
92
FAROUK & PEMBERTON
malnourished or taking steroids and in whom a tension-free anastomosis may be achieved. Because most of o u r patients do not satisfy these criteria, singlestage IPAA is rarely used a t the Mayo Clinic. The decision t o excise the ATZ should relate to the risk of developing subsequent neoplasia. All patients w i t h FAP should have a mucosectomy performed. Patients w i t h C U C who do not have a mucosectomy should have life-long surveillance. Indeed, an argument can be m a d e that all patients should undergo surveillance after IPAA. The decision t o staple the anastomosis impacts little on eventual functional outcome b u t does preserve the ATZ with the attendant risk of recurrent disease, polyps, and neoplasia. When cancer is a presenting feature, the decision to perform IPAA should be based on the stage of the tumor and the subsequent need for radiation therapy. Patients with early-stage tumors not requiring adjuvant radiation therapy attain long-term function comparable to that of patients who have had an IPAA for benign disease.
References 1. Ambroze WL, Pemberton JH, Dozois RR, Carpenter HA: Does retaining the anal transition zone (ATZ) fail to extirpate chronic ulcerative colitis (CUC) after ileal pouch-anal anastomosis (IPAA) [abstract]? Dis Colon Rectum 3420, 1991 2. Bartolo DCC, Duthie GS: Physiological studies following restorative proctocolectomy. In Nicholls J, Bartolo D, Mortensen N (eds): Restorative Proctocolectomy. Oxford, Blackwell Scientific Publications, 1993, pp 83-100 3. Dozois RR, Goldberg SM, Rothenberger DA, et al: Restorative proctocolectomy with ileal reservoir. Int J Colorectal Dis 1:2-19, 1986 4. Emblem R, Bergan A, Larsen S: Straight ileoanal anastomosis with preserved anal mucosa for ulcerative colitis and familial polyposis. Scand J Gastroenterol 23:493-500, 1988 5. Everett WG: Experience of restorative proctocolectomy with ileal reservoir. Br J Surg 7677-87, 1989 6. Fleshman JW, Cohen Z, McLeod RS, et al: The ileal reservoir and ileoanal anastomosis procedure. Factors affecting - technical and functional outcome. Dis Colon Rectum k10-16, 1988 7. Fonkalsrud E W Endorectal ileoanal anastomosis with isoperistaltic ileal reservoir after colectomy and mucosal proctectomy. Ann Surg 199:151-i57, 1984 8. Galandiuk S, Scott NA, Dozois RR, et al: Ileal pouch-anal anastomosis. Reoperation for pouch related complications. Ann Surg 212446-454, 1990 9. Galandiuk S, Wolff BG, Dozois RR, Beart RW Jr: Ileal pouch-anal anastomosis without ileostomy. Dis Colon Rectum 342370473, 1991 10. Grobler SP, Hosie KB, Keighley MR Randomized trial of loop ileostomy in restorative proctocolectomy. Br J Surg 79:903-906, 1992 11. Hallgren T, Fasth S, Nordgren S, et a1 Manovolumetric characteristics and functional results in three different pelvic 'pouch designs. Int J Colorectal Dis 4:156-160, 1989 12. Heald RJ, Allen DR Stapled ileo-anal anastomosis: A technique to avoid mucosal proctectomy in the ileal pouch operation. Br J Surg 73:571-572, 1986 1.3. Heppel J, Kelly KA, Phillips SF, et al: Physiologic aspects of continence after colectomy, mucosal proctocolectomy and endorectal ileo-anal anastomosis. Ann Surg 195:435443, 1982 14. Holdsworth PJ, Johnston D Anal sensation after restorative proctocolectomy and endorectal ileoanal anastomosis. Br J Surg 75:993-996, 1988 15. Horgan PG, O'Connell PR, Shinkwin CA, Kirwan WO: Effect of anterior resection on anal sphincter function. Br J Surg 76:753-756, 1989 16. Johnston D, Holdsworth PJ, Nasmyth DG, et al: Preservation of the entire anal canal in conservative proctocolectomy for ulcerative colitis: A pilot study comparing end-
SURGICAL OPTIONS IN ULCERATIVE COLITIS
93
to-end ileo-anal anastomosis without mucosal resection with mucosal proctectomy and endoanal anastomosis. Br J Surg 74:940-944,1987 17. Keighley MRB: Abdominal mucosectomy reduces the incidence of soiling and sphincter damage after restorative proctocolectomy and J-pouch. Dis Colon Rectum 30:386390, 1988 18. Keighley MRB, Winslett MC, Yoshioka K, Lightwood R Discrimination is not impaired by excision of the anal transition zone after restorative proctocolectomy. Br J Surg 74:1118-1121, 1987 19. Keighley MRB, Yoshioka K, Kmiot WA: Prospective randomized trial to compare the stapled double lumen pouch and the sutured quadrupled pouch for restorative proctocolectomy. Br J Surg 75:1008-1011, 1988 20. King DW, Lubowski DZ, Cook TA: Anal canal mucosa in restorative proctocolectomy for ulcerative colitis. Br J Surg 76:970-972, 1989 21. Landi E, Fianchini A, Landa L, et al: Proctocolectomy and stapled ileo-anal anastomosis without mucosal proctectomy. Int J Colorectal Dis 5:E-154, 1990 22. McIntyre PB, Pemberton JH, Wolff BG, et al: Comparing functional results one year and ten years after ileal pouch-anal anastomosis for chronic ulcerative colitis. Dis Colon Rectum 37:303-307, 1994 23. Nasmyth DG, Williams NS, Johnston D Comparison of the function of triplicated and duplicated ileal reservoirs after mucosal proctectomy and ileoanal anastomosis for ulcerative colitis and adenomatous polyposis. Br J Surg 73:361-366, 1986 24. Nicholls RJ, Lubowski DZ: Restorative proctocolectomy: The four loop (W) reservoir. Br J Surg 74:564-566,1987 25. Nicholls RJ, Pezim ME: Restorative proctocolectomy with ileal reservoir for ulcerative colitis and familial adenomatous polyposis: A comparison of three reservoir designs. Br J Surg 72470474, 1985 26. O'Connell PR, Pemberton JH, Weilano LH, et al: Does rectal mucosa regenerate after ileoanal anastomosis? Dis Colon Rectum 3O:l-5, 1987 27. Parks AG, Nicholls RJ, Belliveau P: Proctocolectomy with ileal reservoir and anal anastomosis. Br J Surg 67533-538, 1980 28. Pemberton JH, Kelly KA, Beart RW Jr, et al: Ileal pouch-anal anastomosis for chronic ulcerative colitis: Long-term results. Ann Surg 206:504-513, 1987 29. Read MG, Read N W Role of anal sensation in preserving continence. Gut 23:345347, 1982 30. Sagar I'M, Holdsworth PJ, Johnston D Correlation between laboratory findings and clinical outcome after restorative proctocolectomy: Serial studies in 20 patients with end-to-end anastomosis. Br J Surg 78:67-70, 1991 31. Scott NA, Dozois RR, Beart RW Jr, et al: Postoperative intra-abdominal and pelvic sepsis complicating ileal pouch-anal anastomosis. Int J Colorectal Dis 3:149-152, 1988 32. Seow-Choen A, Tsunoda A, Nicholls RJ: Prospective randomized trial comparing anal function after hand sewn ileoanal anastomosis with mucosectomy versus stapled ileoanal anastomosis with mucosectomy in restorative proctocolectomy. Br J Surg 78:430434, 1991 33. Smith LE, Friend W, Medwell S: The superior mesenteric artery: The critical factor in pouch pull-through procedure. Dis Colon Rectum 27:741-744, 1984 34. Stern H, Walfisch S, Mullen B, et al: Cancer in the ileoanal reservoir. A new late complication? Gut 31:473475, 1990 35. Sugerman HJ, Newsome HH, DeCosta G, Zfass AM: Stapled ileoanal anastomosis for ulcerative colitis and familial polyposis without a temporary diverting ileostomy. Ann Surg 213:606-619, 1991 36. Sugerman HJ, Newsome HH: Stapled ileoanal anastomosis without a temporary ileostomy. Am J Surg 1675845, 1994 37. Tsunoda A, Talbot IC, Nicholls RJ: Incidence of dysplasia in the anorectal mucosa in patients having restorative proctocolectomy. Br J Surg 77506-508, 1990 38. Utsonomiya J, Iwama T, Imajo M, et al: Total colectomy, mucosal proctectomy and ileoanal anastomosis. Dis Colon Rectum 23:459466, 1980
94
FAROUK & PEMBERTON
39. Williams NS, Johnston D. The current state of mucosal proctectomy and ileo-anal anastomosis in the surgical treatment of ulcerative colitis and adenomatous polyposis. Br J Surg 72:159-168, 1985 40. Winslett MC, Barsoum G, Pringle W, et al: Loop ileostomy after ileal pouch anal anastomosis-is it necessary? Dis Colon Rectum 34:267-270, 1991 Address reprint requests to John H. Pemberton, MD Section of Colon and Rectal Surgery Mayo Clinic 200 First Street SW Rochester, MN 55905
0039-6109/97 $0.00
+ .20
SURGICAL OPTIONS IN ULCERATIVE COLITIS Ridzuan Farouk, MCh, FRCSGlasg, FRCSEd, and John H. Pemberton, MD
Ileal pouch-anal anastomosis (IPAA) has become an established surgical option for the treatment of chronic ulcerative colitis (CUC) and familial adenomatous polyposis (FAP). Much has been written about the procedure, and it is not the intention of this article to reiterate what has been published or to review the literature comprehensively. Instead, this article outlines controversies around which debate still centers. Specifically, the following issues are discussed:
1. Which pouch design is best? 2. Is there a role for defunctioning ileostomy? 3. When should the anal transition zone be excised? 4. When should a stapled anastomosis be avoided? 5. What is the influence of finding a cancer of the colon on the decision to create a pouch? WHICH POUCH DESIGN IS BEST?
The design of the pouch advocated by the earliest proponents of IPAA was a three-limbed reservoir shaped like an S. This pouch was hand sewnz7The advent of the stapler, together with ingenious thinking, has resulted in a multitude of options for constructing pouches, including both hand-sewn and stapled techniques. The variety of pouch configurations now includes the J,38 H,7 W,24 and K." A central concern with the S pouch has always been difficulty with stool evacuation, which had often required self-catheterization especially during the early experience. A long efferent limb was usually blamed for this problem because as the limb was subsequently shortened, so was the incidence of evacua-
From the Mayo Graduate School of Medicine and Division of Colon and Rectal Surgery, Mayo Clinic, and Mayo Foundation, Rochester, Minnesota
SURGICAL CLINICS OF NORTH AMERICA
-
VOLUME 77 * NUMBER 1 FEBRUARY 1997
85
86
FAROUK & PEMBERTON
tion difficulties reduced. The J configuration appears to overcome this problem, but its smaller reservoir size is theoretically associated with reduced early compliance and increased bowel f r e q u e n ~ y The . ~ ~ W and K pouches are larger than the J pouch and do not have the associated evacuatory problems of the S 24 A comparative study on the manovolumetric characteristics and function in patients with the S-, J-, and K-pouch found that, despite use of equal lengths of terminal ileum to construct the pouch with similar initial pouch volumes (120 to 130 mL), the S pouch and the K pouch expanded more favorably, with a maximal volume at 1 year of 420 and 410 mL, respectively, compared with 305 mL for the J pouch." In practice, however, the overall functional results have differed little among the three pouch designs. The W pouch has been described as having the greatest volume for any given length of ileumz4and therefore, theoretically at least, provides greater pouch volume than other configurations. Merely increasing the length of ileum used for construction of the J pouch also improves pouch volume. This translates into a comparable functional outcome for J and W configurations when equal lengths of ileum are A recent comprehensive physiologic assessment comparing W and J pouches has shown that when equal lengths of ileum are used, the manometric, sensory, and compliance data are similar, the only difference being a higher tolerated filling pressure before the sense of urgency occurred in W pouches.* Much investigation has attended these speculations on pouch configuration and its relation to functional outcomes. These studies have failed, however, to address other important issues such as the role of the anal sphincters, stool consistency, altered motility patterns, and the capacity of the pelvis to accommodate the reservoir. In conclusion, it appears there is little difference in functional outcome between the S, W, J, K, and H reservoirs if approximately 30 to 40 cm of terminal ileum are used for reservoir construction. When the pouch has difficulty reaching the anal canal for the anastomosis, the S configuration and its efferent limb have the advantage of achieving better length, thus facilitating an anastomosis without tensi0n.3~ The relative ease in constructing a stapled J pouch, combined with comparable functional outcomes to more complex pouch designs, has made this the favored option at the Mayo Clinic. THE ROLE OF A DEFUNCTIONING ILEOSTOMY The most feared complication of IPAA is pelvic sepsis, and therefore a defunctioning ileostomy after pouch construction is usually performed to minimize its occurrence. Whereas pelvic sepsis complicates 6% of patients undergoing IPAA at the Mayo Clinic,3l the rates reported in the literature vary between 0% and fully 25y0.~. 8, 39 Moreover, disturbingly high rates of pelvic sepsis have been reported in patients undergoing a one-stage procedure (no ileostomy).6 Although the incidence of pelvic sepsis is low, when it occurs it is responsible for a significant number of the failed pouches in the Mayo series. Galandiuk and colleagues8 reported 114 patients who had pouch-related complications requiring reoperation. Of these, 59 (52%) had septic complications (Table 1). Fifty-nine percent of the intra-abdominal sepsis group required further surgery, with 34% (10 patients) eventually undergoing pouch excision. An additional seven patients at follow-up had a diverting stoma. Five of the 30 patients with perianal sepsis (17%) subsequently underwent pouch excision, and a further five had a diverting stoma at the time of review. Protagonists of defunctioning
SURGICAL OPTIONS IN ULCERATIVE COLITIS
87
Table 1. POSTOPERATIVE POUCH FUNCTION AFTER SALVAGE OPERATION FOR
SEPSIS Presenting Problem
Outcome
lleostomy closed Anal pouch function satisfactory Anal pouch function poor lleostomy present Pouch in situ Pouch excised
Perianal FistuldAbscess I n = 30 Datients)
Intra-abdominal FistuldAbscess In = 29 Datients)
45% 21%
35% 7%
17% 17%
24% 34%
From Galandiuk S, Scott NA, Dozois RR, et al: Ileal pouch-anal anastomosis: Reoperation for pouch related complications. Ann Surg 212:446454, 1990; with permission.
ileostomies also argue that diverting stomas allow the anal sphincter and ileal mucosa to recover before restoration of intestinal continuity, and that patients have a short-lived experience of a stoma to fully appreciate the ultimate benefit of IPAA.3,39 Use of a loop ileostomy does not appear to fully protect the patient from pelvic sepsis9 It is nonetheless much easier to manage a patient with sepsis if an ileostomy is in place.9Of Mayo patients who required laparotomy to control sepsis, 41% lost the pouch ultimately and only 29% ever recovered ileoanal function. However, if no reoperation was required, 92% of patients with sepsis eventually had a functioning pouch.” This is a sobering observation. There is little doubt that by avoiding an ileostomy altogether, complications of closure are not an issue, and only one hospital admission is required. Low complication rates with good functional outcome have previously been reported with such an approach.”, 36 At the Mayo Clinic, the clinical and functional outcomes of 37 patients who had a one-stage IPAA without a defunctioning ileostomy have previously been reported,’ and the group was age- and sexmatched with 37 patients who had a defunctioning ileostomy during the same period. The results of this study are summarized in Tables 2 and 3. The incidence of complications was higher when a defunctioning stoma was not used (22% versus 11%,respectively). One-stage IPAA was shown to be a reasonable option
Table 2. POSTOPERATIVE COMPLICATIONS AFTER IPAA: DEFUNCTIONING
ILEOSTOMY VERSUS NO ILEOSTOMY
No lleostomy (37 patients)
Presacral abscess Anastomotic dehiscence Pouch bleeding Fever Jejunal perforation Late presacral abscess lleostomy at follow-up
-
2 5 1 1 1
lleostomy (37 patients)
3 1 2 3
Data from Galandiuk S, Wolff BG, Dozois RR, Beart RW Jr: Ileal pouch-anal anastomosis without ileostomy. Dis Colon Rectum 34:870-873, 1991.
88
FAROUK & PEMBERTON
Table 3. POSTOPERATIVE FUNCTIONAL RESULTS: DEFUNCTIONING ILEOSTOMY (n = 37) VERSUS NO ILEOSTOMY (n = 37) Night
Day
No lleostomy Continence (%) Perfect Spotting Mean stool frequency
79 21 5.0
lleostomy
No lleostomy
94 3 6.0
70 21 1.o
lleostomy
35 52 1 .o
Data from Galandiuk S, Wolff BG, Dozois RR, Beart RW Jr: Ileal pouch-anal anastomosis without ileostomy. Dis Colon Rectum 34:870-873,1991.
if patients were in good health and were not on chronic steroid therapy and if the IPAA could be constructed without tension or technical difficulty. A subsequent randomized trial of 45 CUC patients who were randomized to defunctioning stoma (n = 23) or no stoma (n = 22) showed no differences in rates of complications between the groups.'O The patient group was highly selected, however, in that no patients were taking steroids. Although the authors found no increased risk of pelvic sepsis in patients without an ileostomy, the study was notable for the rather high rate of ileostomy-related complications (52%). A similarly high rate of ileostomy-related complications was described by Winslett and associatesNin a prospective evaluation of 34 patients who had a loop ileostomy. Formation and closure of the ileostomy were associated with morbidity rates of 41% and 30%, respectively. Obviously, there are savings related to shorter operating time and hospital stays in patients having no diverting ileo~tomy.~ A reasonable approach to this dilemma is to use a defunctioning ileostomy in those patients receiving steroid treatment at the time of surgery and in patients who are nutritionally compromised or undergoing an urgent operation. Additionally, if there are concerns about pouch blood supply or anastomotic tension, a diverting stoma is almost mandatory. Using these criteria, 56 of 1800 patients undergoing IPAA have had a one-stage procedure performed at the Mayo Clinic between 1980 and 1996 (unpublished data). WHEN SHOULD THE ANAL TRANSITION ZONE BE EXCISED?
The majority of patients who undergo a pouch procedure have an average of six or seven pouch evacuations per 24 Fecal incontinence is reported in 4% of patients, with nocturnal spotting occurring in 25%.= Stapled anastomosis to the top of the anal canal without excision of the anal transition zone (ATZ) has, in nonrandomized trials, been associated with better functional outcome.'2,16On the other hand, preservation of the ATZ leaves a small amount of rectal mucosa intact.' The debate is drawn as follows: Leaving the ATZ intact improves functional outcome but leaves rectal mucosa at risk of continued inflammation and neoplastic change. CUC and FAP are diseases of the colonic and rectal mucosa. The IPAA procedure is intended to remove all diseased mucosa, thereby eliminating the risks of neoplasia and recurrent mucosal disease. This view has been opposed
SURGICAL OPTIONS IN ULCERATIVE COLITIS
89
by some who argue that the risk of neoplasia in retained segments of rectal mucosa is infinitesimally small and is offset by improved functional outcome. Endoanal mucosectomy has been associated with poorer functional probably because the anal canal is necessarily dilated during the operation. In practice, however, the ability to discriminate gas from feces and the overall functional outcomes have not been shown to correlate with the presence or absence of the ATZ.2 In a study of 50 patients with CUC who had undergone proctocolectomy for their disease, columnar epithelium was found to extend through 50% of the length of the ATZ in 75% of specimens and through 75% of the length in 46% and to approximate the dentate line in 9%.' The degree of inflammation affecting the mucosa was significant compared with that of a control group of specimens from patients with rectal cancer. The study concluded that the ATZ is composed of transitional epithelium and columnar epithelium. Although the transitional epithelium may be free of inflammation, the columnar epithelium is usually inflamed. Therefore, preserving the ATZ would have preserved a focus of disease in most patients with CUC.', 4, 35, 37 Although the risk of malignancy in patients with long-standing CUC is increased but small, in patients with FAP the malignancy risk is 100% by the age of 60 years. The ATZ should therefore be excised in all FAP patients. Because of the risks of leaving diseased mucosa behind and the risks of malignant degeneration, at the Mayo Clinic all patients with FAP undergoing IPAA have the ATZ excised. In patients with CUC, if the ATZ is not excised, life-long surveillance is advised. A focus of disease may still be present even after mucosectomy. We found that in patients who undergo ATZ excision, up to 7% have residual microscopic columnar epithelium.', 13, 2h The clinical significance of this finding remains unknown. The risk of malignant degeneration of such nests 2fl, 34. 37 of rectal mucosa is admittedly small but WHEN SHOULD A STAPLED ANASTOMOSIS NOT BE USED?
Stapling the ileum directly to the anal canal would destroy the internal anal sphincter. Thus, the stapled anastomosis is instead constructed at the top of the anal canal and the ATZ is preserved. This leaves a small remnant of columnar epithelium with the risk of continuing inflammation as well as a risk of future malignant transformation.' Much of the debate about whether to staple the anastomosis or not has really evolved because of the debate about functional outcome. Does preserving the ATZ enhance continence after the IPAA? Stapled anastomosis in nonrandomized trials has been equated with better outcome, which in turn is attributed to less injury to the anal sphincters, with preservation of the ATZ and hence anal sensory discrimination, and with preservation of the rectoanal inhibitory reflex.I2, 14, 16, 17, 23, 3fl However, randomized trial by Seow-Choen et a132comparing endoanal mucosectomy and hand-sewn anastomosis with stapled anastomosis has not shown any difference in functional outcome between the two methods. These findings were not entirely surprising and indirectly confirmed those of other authors: Horgan et all5 described a reduction in anal pressures after low anterior resection, which was in turn attributed to dilatation of the anal sphincters by simply inserting stapling guns into the anus. Read and Readz9 found no deterioration of continence during rectal infusion of saline when anal canal sensation was reduced by the use of topical anesthetic ointments. The retention or recovery of the rectoanal inhibitory reflex, reported in some
90
FAROUK & PEMBERTON
resembles the observations of others in which there was no attempt to excise the most distal cuff of rectum.21Retention of the rectoanal inhibitory reflex is therefore being attained by leaving mucosa at risk of persistent disease as well as malignant degeneration. These clinical impressions have been supported by a limited amount of anorectal physiology data that are available in the literature. Bartolo and Duthiez reported no difference in functional outcome based on the presence or absence of the ATZ. Instead, patient age, pouch compliance, and abnormal pouch motility patterns were more relevant in determining functional outcome. Anastomotic stricturing may occur with either hand-sewn or stapled techniques and affects up to 30% of patient^.^ There appears to be a higher incidence with the double-stapling technique, however, particularly with the use of smaller-caliber stapling guns. However, these strictures are thin, easily dilated, and transient in nature. Forty-one patients at the Mayo Clinic have been randomized to doublestapled (20 patients) or hand-sewn (21 patients) techniques (unpublished data). In the stapled group, 1.5 to 2.0 cm of the ATZ was preserved, whereas a complete mucosectomy was performed in the hand-sewn group. Twenty of these patients subsequently underwent ambulatory manometry (11 stapled, 9 hand-sewn) 6 months after closure of the defunctioning ileostomy. Mean anal canal pressures and the number of episodes when pouch pressure exceeded anal pressure were similar between the two groups (Table 4). We found that parameters of anal canal and pouch motor function did not differ among patients randomized to double-stapled versus hand-sewn IPAA. Thus, stapling and mucosal preservation appear to confer little if any short-term functional advantage. The final decision as to when to staple without mucosectomy and when to do a mucosectomy and hand sew the anastomosis must not be confused with what is safest and best for the patient. Patients at risk of malignancy should have a mucosectomy to completely eliminate this risk. This should include those patients with CUC with evidence of mucosal dysplasia as well as all patients with FAP. WHAT IS THE INFLUENCE OF FINDING A CANCER ON THE DECISION TO CREATE A POUCH?
Although ileal pouch-anal anastomosis has been aimed primarily at eradicating mucosal disease in patients with CUC and prophylactically removing the Table 4. A COMPARISON OF ANAL POUCH PHYSIOLOGY OUTCOME BETWEEN HAND-SEWN AND STAPLED IPAA PATIENTS Pressures (mm Hg)
Mean anal canal pressures Pressure pouch > pressure anal canal (episodedhr)
Stapled Awake 107
Asleep 96 0 (0-12)
Hand-sewn
P Value
Awake Asleep 119 121 1(0-5)
Awake Asleep 0.6 0.4 0.5
SURGICAL OPTIONS IN ULCERATIVE COLITIS
91
neoplastic risk in patients with FAP, there has been an increasing experience with CUC or FAP patients presenting with cancer and undergoing IPAA. Because some patients need adjuvant treatment for advanced cancer, we wished to determine if IPAA should be performed in patients with concomitant cancer and CUC or FAP. Between 1980 and 1995, we identified 150 patients with cancer complicating CUC or FAP presented to the Mayo Clinic, Rochester (unpublished data). Seventy-four of these patients had an IPAA. There were 81 tumors present in these 74 patients. The majority of these were of a relatively early stage, with just 13 Astler-Coller stage C, or Cz tumors. The mean age of these patients was only 32 years, with a range of 15 to 60 years. The mean follow-up was 6 years. There were 11 deaths. Nine were due to systemic progression of the cancer, and 1 patient had local failure with systemic progression of disease. Just 1 patient died of sepsis-related complications 4 months postoperatively. In 8 patients, the ileostomy was permanent; 6 had progression of their disease, and 2 had irreversible pouch injury secondary to radiation therapy. In addition, 5 pouches were excised-2 patients for cancer-related complications, 2 for ischemia, and 1 for intractable pouchitis. Twenty-eight percent of patients have minor nocturnal spotting, 21% require use of constipating medications, and the mean number of pouch evacuations per 24 hours is 6 (range 4 to 13). These functional data are identical to the long-term outcome reported in patients with CUC without colorectal cancer.22 The rate of pouch failure was nearly double that of patients undergoing IPAA for CUC alone.,, Virtually all of the failures (mortality, permanent ileostomy, or pouch excision) occurred in patients with progression of their cancer or secondary to radiation therapy injury (12 of 15 patients). On the basis of these findings, patients with advanced disease (Astler-Coller stage C, or C,) or those who may require adjuvant radiation therapy should not be considered for an IPAA. Instead, an anus-sparing procedure such as ileostomy and preservation of a short rectal stump should be performed and a pouch reconsidered 2 to 3 years later if there is no evidence of radiation injury or recurrent disease. Alternatively, proctocolectomy with a Brooke ileostomy is an excellent choice for these complex patients. If the presence of a cancer is known preoperatively, the patient can be counseled about the possibility of a long-term ileostomy. Unsuspected cancers discovered only during surgery pose a separate problem, however. Before proceeding with pouch construction, the colectomy specimen must be opened and inspected for any suspicious mucosal lesions and frozen-section analysis obtained. If a cancer is found under these circumstances, the decision to proceed with pouch formation must take into consideration location of the cancer (colon versus rectum), the likelihood of transmural penetration, and the possibility of nodal involvement. If a rectal cancer penetrates transmurally or has lymph node metastases or if a colon cancer has nodal involvement, pouch construction should be deferred until adjuvant therapy has been completed and a period of at least 2 years has elapsed to ensure that the patient is free of cancer. SUMMARY
The length of ileum used rather than pouch configuration per se is related to eventual functional outcome: A pouch constructed from 15- to 20-cm limbs is ideal. One-stage procedures appear feasible in those patients who are not
92
FAROUK & PEMBERTON
malnourished or taking steroids and in whom a tension-free anastomosis may be achieved. Because most of o u r patients do not satisfy these criteria, singlestage IPAA is rarely used a t the Mayo Clinic. The decision t o excise the ATZ should relate to the risk of developing subsequent neoplasia. All patients w i t h FAP should have a mucosectomy performed. Patients w i t h C U C who do not have a mucosectomy should have life-long surveillance. Indeed, an argument can be m a d e that all patients should undergo surveillance after IPAA. The decision t o staple the anastomosis impacts little on eventual functional outcome b u t does preserve the ATZ with the attendant risk of recurrent disease, polyps, and neoplasia. When cancer is a presenting feature, the decision to perform IPAA should be based on the stage of the tumor and the subsequent need for radiation therapy. Patients with early-stage tumors not requiring adjuvant radiation therapy attain long-term function comparable to that of patients who have had an IPAA for benign disease.
References 1. Ambroze WL, Pemberton JH, Dozois RR, Carpenter HA: Does retaining the anal transition zone (ATZ) fail to extirpate chronic ulcerative colitis (CUC) after ileal pouch-anal anastomosis (IPAA) [abstract]? Dis Colon Rectum 3420, 1991 2. Bartolo DCC, Duthie GS: Physiological studies following restorative proctocolectomy. In Nicholls J, Bartolo D, Mortensen N (eds): Restorative Proctocolectomy. Oxford, Blackwell Scientific Publications, 1993, pp 83-100 3. Dozois RR, Goldberg SM, Rothenberger DA, et al: Restorative proctocolectomy with ileal reservoir. Int J Colorectal Dis 1:2-19, 1986 4. Emblem R, Bergan A, Larsen S: Straight ileoanal anastomosis with preserved anal mucosa for ulcerative colitis and familial polyposis. Scand J Gastroenterol 23:493-500, 1988 5. Everett WG: Experience of restorative proctocolectomy with ileal reservoir. Br J Surg 7677-87, 1989 6. Fleshman JW, Cohen Z, McLeod RS, et al: The ileal reservoir and ileoanal anastomosis procedure. Factors affecting - technical and functional outcome. Dis Colon Rectum k10-16, 1988 7. Fonkalsrud E W Endorectal ileoanal anastomosis with isoperistaltic ileal reservoir after colectomy and mucosal proctectomy. Ann Surg 199:151-i57, 1984 8. Galandiuk S, Scott NA, Dozois RR, et al: Ileal pouch-anal anastomosis. Reoperation for pouch related complications. Ann Surg 212446-454, 1990 9. Galandiuk S, Wolff BG, Dozois RR, Beart RW Jr: Ileal pouch-anal anastomosis without ileostomy. Dis Colon Rectum 342370473, 1991 10. Grobler SP, Hosie KB, Keighley MR Randomized trial of loop ileostomy in restorative proctocolectomy. Br J Surg 79:903-906, 1992 11. Hallgren T, Fasth S, Nordgren S, et a1 Manovolumetric characteristics and functional results in three different pelvic 'pouch designs. Int J Colorectal Dis 4:156-160, 1989 12. Heald RJ, Allen DR Stapled ileo-anal anastomosis: A technique to avoid mucosal proctectomy in the ileal pouch operation. Br J Surg 73:571-572, 1986 1.3. Heppel J, Kelly KA, Phillips SF, et al: Physiologic aspects of continence after colectomy, mucosal proctocolectomy and endorectal ileo-anal anastomosis. Ann Surg 195:435443, 1982 14. Holdsworth PJ, Johnston D Anal sensation after restorative proctocolectomy and endorectal ileoanal anastomosis. Br J Surg 75:993-996, 1988 15. Horgan PG, O'Connell PR, Shinkwin CA, Kirwan WO: Effect of anterior resection on anal sphincter function. Br J Surg 76:753-756, 1989 16. Johnston D, Holdsworth PJ, Nasmyth DG, et al: Preservation of the entire anal canal in conservative proctocolectomy for ulcerative colitis: A pilot study comparing end-
SURGICAL OPTIONS IN ULCERATIVE COLITIS
93
to-end ileo-anal anastomosis without mucosal resection with mucosal proctectomy and endoanal anastomosis. Br J Surg 74:940-944,1987 17. Keighley MRB: Abdominal mucosectomy reduces the incidence of soiling and sphincter damage after restorative proctocolectomy and J-pouch. Dis Colon Rectum 30:386390, 1988 18. Keighley MRB, Winslett MC, Yoshioka K, Lightwood R Discrimination is not impaired by excision of the anal transition zone after restorative proctocolectomy. Br J Surg 74:1118-1121, 1987 19. Keighley MRB, Yoshioka K, Kmiot WA: Prospective randomized trial to compare the stapled double lumen pouch and the sutured quadrupled pouch for restorative proctocolectomy. Br J Surg 75:1008-1011, 1988 20. King DW, Lubowski DZ, Cook TA: Anal canal mucosa in restorative proctocolectomy for ulcerative colitis. Br J Surg 76:970-972, 1989 21. Landi E, Fianchini A, Landa L, et al: Proctocolectomy and stapled ileo-anal anastomosis without mucosal proctectomy. Int J Colorectal Dis 5:E-154, 1990 22. McIntyre PB, Pemberton JH, Wolff BG, et al: Comparing functional results one year and ten years after ileal pouch-anal anastomosis for chronic ulcerative colitis. Dis Colon Rectum 37:303-307, 1994 23. Nasmyth DG, Williams NS, Johnston D Comparison of the function of triplicated and duplicated ileal reservoirs after mucosal proctectomy and ileoanal anastomosis for ulcerative colitis and adenomatous polyposis. Br J Surg 73:361-366, 1986 24. Nicholls RJ, Lubowski DZ: Restorative proctocolectomy: The four loop (W) reservoir. Br J Surg 74:564-566,1987 25. Nicholls RJ, Pezim ME: Restorative proctocolectomy with ileal reservoir for ulcerative colitis and familial adenomatous polyposis: A comparison of three reservoir designs. Br J Surg 72470474, 1985 26. O'Connell PR, Pemberton JH, Weilano LH, et al: Does rectal mucosa regenerate after ileoanal anastomosis? Dis Colon Rectum 3O:l-5, 1987 27. Parks AG, Nicholls RJ, Belliveau P: Proctocolectomy with ileal reservoir and anal anastomosis. Br J Surg 67533-538, 1980 28. Pemberton JH, Kelly KA, Beart RW Jr, et al: Ileal pouch-anal anastomosis for chronic ulcerative colitis: Long-term results. Ann Surg 206:504-513, 1987 29. Read MG, Read N W Role of anal sensation in preserving continence. Gut 23:345347, 1982 30. Sagar I'M, Holdsworth PJ, Johnston D Correlation between laboratory findings and clinical outcome after restorative proctocolectomy: Serial studies in 20 patients with end-to-end anastomosis. Br J Surg 78:67-70, 1991 31. Scott NA, Dozois RR, Beart RW Jr, et al: Postoperative intra-abdominal and pelvic sepsis complicating ileal pouch-anal anastomosis. Int J Colorectal Dis 3:149-152, 1988 32. Seow-Choen A, Tsunoda A, Nicholls RJ: Prospective randomized trial comparing anal function after hand sewn ileoanal anastomosis with mucosectomy versus stapled ileoanal anastomosis with mucosectomy in restorative proctocolectomy. Br J Surg 78:430434, 1991 33. Smith LE, Friend W, Medwell S: The superior mesenteric artery: The critical factor in pouch pull-through procedure. Dis Colon Rectum 27:741-744, 1984 34. Stern H, Walfisch S, Mullen B, et al: Cancer in the ileoanal reservoir. A new late complication? Gut 31:473475, 1990 35. Sugerman HJ, Newsome HH, DeCosta G, Zfass AM: Stapled ileoanal anastomosis for ulcerative colitis and familial polyposis without a temporary diverting ileostomy. Ann Surg 213:606-619, 1991 36. Sugerman HJ, Newsome HH: Stapled ileoanal anastomosis without a temporary ileostomy. Am J Surg 1675845, 1994 37. Tsunoda A, Talbot IC, Nicholls RJ: Incidence of dysplasia in the anorectal mucosa in patients having restorative proctocolectomy. Br J Surg 77506-508, 1990 38. Utsonomiya J, Iwama T, Imajo M, et al: Total colectomy, mucosal proctectomy and ileoanal anastomosis. Dis Colon Rectum 23:459466, 1980
94
FAROUK & PEMBERTON
39. Williams NS, Johnston D. The current state of mucosal proctectomy and ileo-anal anastomosis in the surgical treatment of ulcerative colitis and adenomatous polyposis. Br J Surg 72:159-168, 1985 40. Winslett MC, Barsoum G, Pringle W, et al: Loop ileostomy after ileal pouch anal anastomosis-is it necessary? Dis Colon Rectum 34:267-270, 1991 Address reprint requests to John H. Pemberton, MD Section of Colon and Rectal Surgery Mayo Clinic 200 First Street SW Rochester, MN 55905