ASCRS Ch49 - Ulcerative Colitis: Surgical Management
Chapter 49 of the ASCRS Textbook of Colon and Rectal Surgery, 4th edition (Karen N. Zaghiyan and Phillip R. Fleshner). Book pages 835-850 (PDF pages 828-843). A detailed operative/decision-making chapter for UC surgery — heavy overlap with this wiki's existing Ulcerative Colitis "Indications for surgery"/"Surgical options" content and Ileal Pouch-Anal Anastomosis (IPAA)'s construction content, but adds substantial new operative-technique and controversy-level detail. All content is textbook/secondary-level unless a specific primary trial is named.
Indications for surgery
Approximately 30% of UC patients will undergo surgery in their lifetime; most surgery is elective, but 10% require emergent surgery. Elective indications: medically refractory colitis (~70% of the overall surgical cohort), complications or side effects of medications, extraintestinal manifestations, growth retardation in children, dysplasia, or cancer. A survey of UC patients having surgery found over 50% felt they should have undergone surgery at an earlier time-point — supporting early surgical evaluation/shared decision-making rather than exhausting all medical options first.
Colorectal cancer risk: overall CRC rate in UC patients is 3.7%, increasing with disease duration — 2% at 10 years, 8% at 20 years, and 18% at 30 years after disease onset. Young age at diagnosis, pancolitis, family history of CRC, and PSC further increase risk. Surveillance colonoscopy every 1-2 years (chromoendoscopy or high-definition with random quadrant biopsies every 10cm) is recommended beyond 8 years of disease. High-grade dysplasia (any distribution) is an indication for colectomy. The SCENIC guidelines (based on very-low-quality evidence) distinguish visible/invisible and polypoid/non-polypoid dysplasia: endoscopically resected polypoid dysplastic lesions may undergo surveillance colonoscopy rather than colectomy; endoscopic surveillance of non-polypoid (flat) dysplasia remains conditional/controversial vs. other guidelines recommending surgical referral; SCENIC has also challenged the routine use of random quadrant biopsies. Historical data (older, low-definition-endoscopy era) found colorectal cancer in the colectomy specimen in 22% (18/81) of patients with invisible low-grade dysplasia and 32-42% with invisible high-grade dysplasia — rates that supported colectomy even for low-grade dysplasia — but SCENIC argues these high rates may be an artifact of missed lesions in the pre-high-definition-endoscopy era: only 10% of biopsies showing dysplasia are invisible in the current high-definition era vs. 87% previously. Thus, current SCENIC recommendations for invisible dysplasia (confirmed by a GI pathologist) are referral to an IBD center experienced in chromoendoscopy/high-definition colonoscopy for intensive screening rather than automatic colectomy, unless no visible lesion is found on repeat exam (in which case high-grade dysplasia still prompts colectomy, while low-grade is often offered surveillance, more readily referred to surgery if multifocal).
Emergent indications: acute severe UC (ASUC) not responding to medical therapy, sepsis, toxic megacolon, perforation, or severe bleeding (perforation/severe bleeding are less common). ASUC occurs in ~25% of UC patients requiring hospital admission, diagnosed per modified Truelove and Witts criteria (≥6 bloody stools/day plus one of HR>90, temp>37.8°C, Hgb<10.5g/dL, or ESR>30mm/hr). Initial treatment is IV corticosteroids plus supportive care; infectious mimics (C. difficile, CMV) must be excluded. 30-40% have partial/no response to initial steroids. In the pre-biologic era, steroid-refractory ASUC meant urgent colectomy; current standard of care adds inpatient infliximab or cyclosporine as medical rescue, though colectomy rates remain high (13-25% in-hospital, ~50% at 1 year). In a study of 270 ASUC patients (2002-2017), a multivariable logistic regression model found prior thiopurine or anti-TNF treatment (HR 3.86), C. difficile infection (HR 3.73), CRP >30mg/L (HR 3.06), and albumin <3.0g/dL (HR 2.67) independently predicted colectomy; a risk score (each factor = 1 point) found cumulative 1-year colectomy risk of 0%, 9%, 11%, 51%, and 100% for scores of 0, 1, 2, 3, or 4 respectively. Despite these tools, the threshold for surgery remains high — waiting until all medical options are exhausted increases surgical morbidity (>50%) and in-hospital mortality (8%). The chapter authors advocate early surgical evaluation for hospitalized ASUC patients, framing surgery as an alternative to medical management rather than a last resort.
Toxic megacolon combines ASUC with radiologic colonic dilation (total or segmental). A dilated colon without toxicity signs may be given an initial trial of conservative management (bowel rest, serial abdominal exams); signs of sepsis (fever, tachycardia, progressive abdominal pain) are indications for urgent colectomy.
Preoperative planning
Enterostomal therapist consultation for ostomy site marking/preoperative counseling; preoperative small bowel evaluation (if not recently done) to exclude small bowel inflammation and confirm the UC diagnosis. A steroid taper should be considered as tolerated to minimize perioperative dose. Except for documented adrenal insufficiency, a perioperative corticosteroid stress dose is not recommended and may actually increase infectious complications — patients with prolonged steroid exposure should instead be maintained on their preoperative steroid dose with a taper on discharge. Preoperative nutritional optimization (dietician referral for weight loss/hypoalbuminemia) and prehabilitation (preoperative oral nutritional supplementation ± exercise program) have been suggested to improve postoperative recovery, though colorectal-IBD-specific data is limited. Oral antibiotics combined with mechanical bowel preparation is recommended to minimize postoperative infectious complications.
Ileal pouch-anal anastomosis (IPAA)
Described by Parks and Nichols in 1978; standard operation for medically refractory disease, cancer, or dysplasia. May be one-, two-, or three-stage. A near-total proctocolectomy with a stapled or hand-sewn ileal pouch-anal anastomosis. Laparoscopic approach is preferred electively (reduced minor complications, shorter hospital stay short-term; reduced adhesions and improved female fertility long-term). Surgeon should use the approach safest for their hands/skill level regardless of platform (open/straight lap/hand-assisted/robotic) — technical steps are similar across platforms.
Operative technique (laparoscopic, as described): orogastric tube decompression; three trocars (11mm umbilical camera port, 5mm suprapubic, 12mm at the future ileostomy site — reusable for specimen extraction). Small bowel evaluated for Crohn's disease and abdomen explored for perforation evidence. Standard abdominal colectomy (close bowel mesenteric dissection, ileocolic artery preservation) from right to left: lateral attachments taken down, hepatic flexure mobilized, mesenteric window distal to ileocolic artery incised, transverse mesocolon/gastrocolic ligament divided, lesser sac entered, splenic flexure taken down, mesentery divided to the pelvic brim. Decide to proceed with/abort proctectomy and IPAA based on assessment of mesenteric length for pouch reach at this point. Specimen extracted through the future ileostomy site (12mm trocar site enlarged); ileum transected flush with cecum with a GIA stapler to prevent staple-line blowout during extraction. If the colon/mesentery is too thick to safely extract intact, a small Pfannenstiel incision can be made. End ileostomy matured with a transanal drain to decompress the rectal staple line; in malnourished patients, an alternative is a subcutaneous stump or mucous fistula for the abdominal stump.
If proceeding with IPAA at the same setting: rectal dissection to the pelvic floor (laparoscopically, or via Pfannenstiel-assisted open proctectomy) — either total mesorectal excision plane or an alternative intramesorectal/close rectal dissection (popularized by vessel sealers) to further reduce autonomic nerve injury risk. Superior rectal artery divided with vessel sealer; posterior dissection first (relatively bloodless), then anterior (to rectoprostatic fascia in men / rectovaginal septum in women, level of prostate/mid-vagina), then lateral stalks divided close to the rectum, posteriorly past the coccyx. Double-stapled anastomosis: rectum stapled/transected at the puborectalis level with an articulating endo-GIA or right-angle linear stapler, leaving a 1-2cm rectal mucosal cuff.
J-pouch construction: terminal ileum aligned in a J configuration; both limbs 15-25cm (length guided by deepest pelvic reach — apex must reach beyond the symphysis pubis for tension-free anastomosis). Selective division of mesenteric vessels to the apex, and superficial anterior/ posterior mesenteric incisions plus mobilization up to/anterior to the duodenum, are lengthening maneuvers. Pouch constructed with continuous absorbable suture or a stapling device; air-tested for leaks; delivered to the pelvis with mesentery oriented straight to the retroperitoneum (not tethered caudally). For a double-stapled anastomosis: circular stapler anvil secured in the pouch apex via purse-string, advanced transanally through the transverse rectal staple line and connected/fired under direct visualization by the abdominal operator (ensuring no adjacent tissue entrapment); integrity confirmed by transanal air insufflation. Alternatively, distal mucosectomy with hand-sewn IPAA (Lone Star retractor, dilute epinephrine submucosal injection, mucosectomy leaving a short cuff of denuded rectal muscle, pouch pulled through and hand-sewn full-thickness to the dentate line/ internal sphincter). A proximal defunctioning loop ileostomy is typically created regardless of anastomotic method; pelvic drains placed. Anti-adhesion barrier around the stoma/incision may reduce postoperative adhesion incidence/severity in open cases. Enhanced-recovery elements (early diet advancement, ambulation, early urinary catheter removal) are safe in this population and improve outcomes/LOS. Ileostomy output can be high (more proximal than a traditional terminal ileostomy) — hydration counseling ± antidiarrheal medication. Ileostomy closed 6-8 weeks later after confirming healed suture lines endoscopically, normal digital exam, and a normal contrast study (no leak, fistula, sinus).
Controversies
One-, two-, or three-stage IPAA: in well-nourished patients with tension-free anastomosis, a single-stage IPAA (omitting diverting ileostomy) has been considered — proponents cite high rates of ileostomy-related complications (43% overall, including 23% obstruction, 25% dehydration, 29% ileostomy-closure-operation complications). Level 1 evidence for single-stage safety is lacking, but retrospective series of selective ileostomy omission suggest similar complication rates/long-term function vs. diverted patients. In a study of 317 diverted vs. 670 undiverted pouches: pouch leaks occurred in 13.7% (diverted) vs. 13.6% (undiverted) — similar rates — but diverted patients with a leak more often needed unplanned OR trips (12% vs. 45% of leak patients, p<0.01) while undiverted leak patients more often needed an unplanned ostomy within 200 weeks (27% vs. 60% of leak patients, p<0.01); pouch salvage rates over total follow-up were similar (74% vs. 78%). A separate 4031-patient evaluation found higher reoperation-for-diverting-ileostomy rates without diversion at time of IPAA (48%) vs. with diversion (12%, p<0.0001), but 5- and 10-year pouch survival was similar (99% vs. 97%, and 88% vs. 87%). These comparisons are biased since omitting diversion is only considered in the healthiest, most straightforward patients.
Many UC patients are not appropriate candidates for upfront (1-stage or 2-stage) IPAA. A three- stage IPAA (initial total abdominal colectomy + end ileostomy, then completion proctectomy + IPAA + diverting loop ileostomy months later, then ileostomy closure) suits hospitalized/urgent- colectomy patients and those with preoperative malnutrition, high-dose steroids, obesity, cancer, female patients desiring pregnancy, or diagnostic uncertainty (IBD-unclassified). Preoperative corticosteroids >20mg/day and preoperative hypoalbuminemia (<3g/dL) are independent risk factors for postoperative pouch-related infectious complications, and in the authors' opinion are indications for initial colectomy with staged IPAA. Preoperative biologic exposure has been extensively debated (conflicting retrospective data, confounded by concomitant corticosteroid use); the recent prospective multicenter PUCCINI study (955 IBD patients undergoing abdominal surgery) found any infection (19% vs. 20%) and SSI (12% vs. 13%) similar with/without preoperative anti-TNF exposure — no robust data yet suggests an independent surgical-morbidity impact for anti-TNF, vedolizumab, or ustekinumab, so preoperative biologic treatment alone does not appear to mandate staged colectomy/ IPAA. Obesity is an independent predictor of pouch abandonment (visceral fat limiting mesenteric pouch reach) and carries higher overall-complication and anastomotic-leak risk — preoperative weight loss counseling (target BMI <30) is recommended, with a staged approach for those who cannot achieve this preoperatively. Patients with colon cancer may be better served by a staged approach (delayed proctectomy/IPAA after systemic chemotherapy, avoiding a scenario where pelvic infectious complications delay chemotherapy). Patients desiring near-term pregnancy may prefer a staged approach (attempt childbearing after colectomy/end ileostomy, before completion proctectomy/IPAA, to minimize infertility risk from pelvic dissection). In IBD-unclassified patients, staging allows pathologic evaluation of the colectomy specimen to guide subsequent surgical decision-making.
A modified two-stage IPAA (total abdominal colectomy + end ileostomy, then a single final stage combining completion proctectomy + IPAA without diverting ileostomy) has recently been described — proponents argue patients returning in good nutritional health can skip the third stage/diversion- related complications. An initial report (23 modified-two-stage vs. 31 traditional-three-stage patients) found no pouch-related complications requiring stoma creation in the modified group, and lower total cost/hospital stay. A more recent cohort (223 traditional two-stage [restorative proctocolectomy + diverting ileostomy, then closure] vs. 237 modified two-stage) found a 4.7% anastomotic leak rate with modified two-stage vs. 15.7% with traditional two-stage (p<0.01) — promising but limited by patient-selection/generalizability concerns; prospective RCTs are needed.
Optimal pouch design: originally a three-limb "S" pouch (Parks and Nichols); superseded by the two-limb "J" pouch (Utsunomiya) which became standard due to ease of stapled construction. A 2007 meta-analysis of 23 studies found no difference in anastomotic leak, pelvic sepsis, or pouch failure between pouch designs (J/S/W) short-term. Two large long-term meta-analyses found J-pouches have increased stool frequency (average one more bowel movement per 24 hours) vs. other designs, but all other functional outcomes were equivalent.
Type of anastomosis (hand-sewn vs. stapled): original Parks/Nicholls description used complete mucosectomy + hand-sewn anastomosis; stapling became default practice over the following three decades. Several 1990s RCTs found no superiority of either technique (small sample size/single- institution limitations). More recent evidence shows equal long-term functional results between techniques, with consistently lower short-term morbidity after stapled IPAA. Regarding disease relapse/malignancy risk, the largest published series found no higher neoplasia rate in either the anal transition zone or pouch after a stapled procedure, while dysplasia/malignancy at the time of IPAA remains an independent risk factor regardless of technique. Practical recommendation: in patients with colitis and rectal high-grade dysplasia or adenocarcinoma at the time of surgery, mucosectomy and hand-sewn IPAA should be strongly considered; stapled anastomosis can be the first choice in all other circumstances.
Transanal pouch (ta-J-pouch): addresses the technically demanding step of ensuring perpendicular distal rectal transection in laparoscopic IPAA — an oblique angle risks multiple stapler firings (higher leak risk) or inadequate transection leaving a long retained rectal cuff (cuffitis/pouch- evacuation problems). The transanal approach also changes the ileoanal anastomosis from a double- staple technique (with "dog ears" at the sides) to a single stapled line, and allows horizontal-plane dissection of the most distal/curved rectum. Short- and long-term functional data appear to support the transanal approach, though more robust comparative data are still needed.
Ileorectal anastomosis (IRA)
While IPAA is the gold-standard approach, recent IRA series report similar long-term function/QOL. Proponents cite lower technical demand, single-stage feasibility, elimination of pelvic dissection (and associated pelvic sepsis/poor function/sexual-urinary dysfunction/female infertility risk). Technique is otherwise similar to the colectomy portion of IPAA (ileocolic pedicle and superior rectal artery preserved; ileum transected flush with the cecum, rectum transected at the taenia coli splay; 29mm EEA circular-stapled anastomosis; fluorescence angiography for perfusion assessment; diverting ileostomy generally not required in healthy, well-nourished patients with a tension-free, intact anastomosis).
Several studies show IRA safety, with overall complications 24-28% and anastomotic leak rate 3-4%. Long-term failure rate (the most important concern) ranges 18-49%. A recent 343-patient multicenter French retrospective study found treatment with immunosuppressants and anti-TNF before colectomy, and colectomy for severe acute colitis, were independent predictors of IRA failure, while colectomy for severe acute colitis was associated with decreased risk of IRA failure — [note: as summarized in the chapter, "colectomy for severe acute colitis" appears both as a predictor of failure and as associated with decreased failure risk in different clauses; this apparent internal inconsistency is preserved here rather than silently resolved, since it could not be disambiguated from the available chapter text].
A second concern is dysplasia/cancer development in the retained rectum. In an 89-patient study (1981), overall rectal-remnant cancer rate was 4.8%, ranging from 0% (<10 years of disease) to 13% (>25 years). Patients with cancer/dysplasia in the colon at colectomy had substantially higher later rectal cancer/precancer risk: 22% (2/9) with mild colonic dysplasia, and 71% (5/7) among surviving patients with colon cancer/precancer at colectomy. A meta-analysis of UC patients undergoing surgery found subsequent CRC risk of 2.1% (rectal stump), 2.4% (IRA), and 0.5% (IPAA) — IRA/rectal stump carried significantly higher risk than IPAA (OR 6.4, 95% CI 4.3-9.5); a history of CRC was the strongest risk factor for subsequent CRC after both IRA (OR 12.8) and IPAA (OR 15.0). While a CRC/high-grade-dysplasia history may make certain populations (ASUC patients naïve to biologics/immunomodulators, indeterminate colitis with rectal sparing, young women desiring fertility preservation) reasonable selective-IRA candidates, rectal compliance (chronic UC can impair it) and surgeon capability with both operations are important additional considerations.
Continent ileostomy
Not primarily advised for patients needing permanent fecal diversion, but a viable option for patients who failed a Brooke ileostomy or are IPAA candidates who cannot have a pouch (rectal cancer, perianal fistulas, poor anal sphincter function, occupations precluding frequent toilet access). Suspected Crohn's disease contraindicates construction (recurrent pouch disease risk may require resecting the ~45cm-of-viable-small-bowel pouch, risking malnutrition). Obesity and age >40 are relative contraindications (increased pouch dysfunction risk). Major complications are common and often require surgical correction (sometimes pouch excision + standard Brooke ileostomy) — only highly motivated, emotionally stable patients should consider this procedure.
Operative technique (Kock pouch): uses the terminal 45-50cm of ileum. Distal 3-5cm forms the outlet; next 18cm forms the nipple valve (via intussusception, stabilized with 3 firings of a GIA stapler without a knife, plus a two-layer anterior closure and circumferential interrupted sutures between outlet and pouch); proximal 30cm forms the pouch (folded into three 10cm limbs with a posterior row of sutures between limbs, an enterotomy along the S-shape, and a second posterior row re-approximating cut edges). Peritonectomy overlying the nipple-valve mesentery on both sides increases adhesion formation during intussusception to help prevent valve slippage. The stoma is matured flush with the skin and anchored to the posterior rectus sheath. A wide plastic tube with large openings is placed for gravity drainage postoperatively, occluded for progressively longer periods starting 10 days after surgery until removable for 8 hours without distress — at which point the pouch is drained by intubation 3 times daily.
Postoperative complications: nipple-valve slippage (most common, occurring in almost 30% of patients) — the intussuscepted segment tends to slide/extrude on its mesenteric aspect, causing difficult catheterization, chronic outflow obstruction, and incontinence; wrapping the valve with prosthetic material prevents slippage but risks unacceptably high parastomal abscess/fistula rates; surgical correction (repairing or recreating the valve from the afferent limb) is virtually always required once slippage occurs. Pouchitis occurs in 25% of patients (second most common complication) — nonspecific reservoir inflammation from stasis/anaerobic overgrowth, presenting with increased ileostomy output, fever, weight loss, and stomal bleeding; diagnosed by history + pouch endoscopy; usually responds to antibiotics and continuous pouch drainage. Intestinal obstruction occurs in ~5%. Fistula occurs in ~10%, most commonly from the pouch itself or the nipple-valve base (dehiscence of suture lines or, rarely, ileostomy tube erosion); base-of-valve fistulas cause incontinence (bypassing the high-pressure zone) and rarely heal without surgery (valve excision, pouch rotation, new continent valve construction from the afferent limb). Despite frequent complications, reported patient satisfaction is high (~90% eventually achieve total continence after one or more procedures), though this must be weighed against the frequency of major reoperation. The Barnett modification (afferent limb used to construct the nipple valve, wrapping a portion of the residual efferent limb around it) was designed to reduce valve slippage/fistula but has no controlled data showing superiority over the standard Kock pouch technique most centers use.