ASCRS Ch48 - Crohn's Disease: Surgical Management
Chapter 48 of the ASCRS Textbook of Colon and Rectal Surgery, 4th edition (Lisa M. Cannon and Alessandro Fichera). Book pages 813-830 (PDF pages 806-823). A general surgical-management chapter for Crohn disease, covering indications, preoperative optimization, anastomotic technique, and location-specific operative considerations — heavily overlapping with this wiki's existing Sabiston-sourced "Surgical treatment" and "Specific problems" content on Crohn Disease, but adding substantial genuinely new trial-level and epidemiologic detail throughout. All content below is textbook/secondary-level unless a specific primary trial is named.
Changing trends and introduction
Surgery does not cure Crohn's; goals are relieving symptomatic complications (obstruction, fistula), improving quality of life, preserving small bowel, and minimizing treatment interruptions. The need for surgery based on time from diagnosis has declined markedly across eras — in a population-based cohort diagnosed 1955-1989, 73% of patients overall required surgery (44% at 1 year, 61% at 5 years, 71% at 10 years); in a cohort diagnosed 2003-2004 followed to 2011, only 29% required surgery (14.6% at 1 year, 24.6% at 5 years, 28.5% at 7 years). The largest drop in need for surgery predates the introduction of biologics and appears more reflective of increased corticosteroid use; improved diagnostic modalities have also led to earlier diagnosis/initiation of therapy. Both trends have likely decreased the rate of complicated surgical disease. IBD "surgical home" (SMH) specialty medical-home models with dedicated IBD-focused colorectal surgeons ("IBDologists") have been shown to reduce unplanned care/disease activity and increase quality of life. Quality of life improves as early as 2 weeks postoperatively after intestinal resection and remains high long-term; patients generally are satisfied and often wish they had had surgery earlier in their disease course.
Indications for surgery (Table 48.1)
Failure of medical management, pediatric growth retardation, bowel obstruction, free perforation, penetrating disease/fistula/phlegmon/abscess, cancer/dysplasia, toxic colitis, bleeding. Free perforation, toxic colitis, and major hemorrhage are true surgical emergencies; the rest require individualized nonemergent decision-making. "Failure of medical therapy" carries multiple meanings: inability to achieve symptom control despite aggressive therapy and being transitioned through available agents; ability to achieve control but with unacceptable side effects; pediatric growth retardation as its own failure signal. Steroid-refractory = active disease despite prednisolone up to 1 mg/kg/day for 4 weeks; steroid-dependent = unable to reduce below prednisolone 10 mg/day equivalent without reactivation, or relapse within 3 months of stopping steroids.
Bowel obstruction
About one-quarter of Crohn's surgery is secondary to obstructive symptomatology. Untreated/poorly controlled disease causes progressive transmural injury; histology of a Crohn's stricture shows muscularis mucosa/muscularis propria thickening with fibrotic change and increased submucosal volume/density. No antifibrotic therapies for stricturing disease currently exist. Once obstruction is diagnosed, the relative contribution of fibrosis vs. inflammation is assessed via CT or MR enterography (current standard for the small intestine). CT enterography findings of inflammation: mucosal hyperenhancement, mesenteric fat stranding ("comb sign"), mesenteric hypervascularity — but CT enterography is not successful at identifying degree of fibrosis, and upstream dilation is not reliable at distinguishing inflammation from fibrosis (mesenteric hypervascularity was the only CT finding predictive of fibrosis on multivariate analysis). MRI findings of inflammation: T2 hypersignal, mucosal enhancement, presence of ulceration, blurred margins; a homogeneous enhancement pattern and the percent enhancement gain over time can discriminate severe fibrosis deposition — though most lesions show a mixed inflammation/fibrosis pattern. Inflammatory stenoses are likely to respond to medical therapy; fibrotic strictures typically require surgery. For localized ileocecal obstruction, surgery is indicated if the patient fails a trial of medical management (bowel rest + IV corticosteroids); upfront surgery is indicated if imaging suggests marked fibrosis with low inflammation. Retrospective series predating the biologic era found over half of ileocecectomy patients never require another surgery.
Endoscopic balloon dilation: an alternative that can delay or prevent surgery. Ideal candidate: single short-segment (<4cm) fibrostenotic stricture or an anastomotic stricture, without associated fistula/abscess and without marked angulation. Dilation diameter should be ≥14mm; dilation to 16-18mm is associated with less frequent follow-up re-dilation; dilating small bowel strictures beyond 20mm may increase perforation/bleeding risk. Multifocal stricturing disease is an independent risk factor for clinical failure of balloon dilation. In a recent pooled analysis (Bettenworth), 1463 patients underwent over 3000 balloon dilation procedures: overall technical success 89%; anastomotic strictures comprised 62% of procedures; at 2-year follow-up, three-quarters of patients had required re-dilation and half had undergone surgical resection; major adverse events (perforation, bleeding, sepsis) occurred in 2.8% of patients.
Perforation
Rare in Crohn's; typically presents as localized abscess rather than free perforation. If associated with small bowel strictures, perforation is usually immediately proximal to a completely obstructing stricture (relative closed-loop obstruction) — best treated with resection and primary anastomosis if condition allows. When perforation occurs from an obstructing colonic stricture, the site is more commonly the cecum, best treated with total abdominal colectomy and ileostomy (distal resection extent tailored to whether distal colon/rectum are disease-free). Toxic colitis leading to perforation is managed with total abdominal colectomy and ileostomy.
Penetrating disease: fistula and abscess formation
Approximately 11-16% of adult patients have penetrating intestinal disease manifestations. Risk factors include serologic/genetic markers and tobacco. Hirten characterized relative fistula frequency by location: 29% enterocolonic, 18-24% enteroenteric, 6-16% enterocutaneous, 4-9% rectovaginal, 2-8% enterovesical, and rarely enterosalpingeal/enterogastric — originating site is usually the ileocecal region/terminal ileum. Only symptomatic fistulas require intervention; the targeted bowel segment requires resection, but the recipient organ/bowel can often be primarily repaired or undergo small patch excision/transverse closure. Penetration to the retroperitoneum can cause a psoas abscess (special mention: prone to epithelialize and become a recalcitrant recurrent- abscess source, usually requiring surgery; after resection of diseased bowel, the psoas abscess cavity may be unroofed/curetted, with an omental pedicle flap facilitating healing).
Small abscesses/phlegmons should initially be treated with antibiotics; phlegmon associated with active inflammatory disease can safely add corticosteroids. In Felder's series of 24 patients with Crohn's + palpable inflammatory mass treated with high-dose corticosteroids: two-thirds resolved completely, and in the remaining third the mass reduced by >50% — but 58% of patients still required resection for persistence/recurrence of symptoms, most performed electively.
Intra-abdominal abscess: accessible abscesses >3cm average diameter should undergo percutaneous drainage (PCD) plus antibiotics; technical success of PCD is >90%. A 2016 meta-analysis (Clancy; 6 studies, 333 patients) compared primary surgical resection (55%, n=184) vs. PCD alone (45%, n=149): significantly higher rate of recurrent abscess in the PCD group (OR 6.54); pooled proportion of PCD-alone patients requiring subsequent surgery was 70.7%; among patients whose abscess resolved with drainage, 17.9% required surgery for recurrence. Up to 30% of patients with successful PCD may be able to avoid surgery altogether per this meta-analysis. The ongoing MICA trial (GETAID) is prospectively examining predictive factors of anti-TNF response in luminal Crohn's complicated by abscess. Müller-Wille found preoperative PCD (used in 48% of patients with spontaneous intra- abdominal abscess, average 37 days before surgery) significantly lowered postoperative septic complications (25% vs. 69% without preoperative PCD). Zhang (288 patients, 180 with penetrating behavior including 54 with intra-abdominal sepsis) found intra-abdominal sepsis specifically — not penetrating behavior alone — was associated with worse postoperative outcome (more septic complications, superficial SSI, stoma formation). Given that penetrating disease is a recurrence risk factor, prophylactic therapy after resection is typically recommended.
Cancer and dysplasia
Crohn's patients have increased risk of both intestinal and non-intestinal cancers vs. the general population. Colorectal cancer is the cause of 1 in every 12 IBD-patient deaths. Colitis-associated cancer shares many molecular alterations with sporadic CRC but with different timing/frequency: loss of APC function occurs early in sporadic CRC but late in colitis-associated cancer; loss of p53 function occurs early in colitis-associated cancer (opposite of the sporadic sequence). Cancer development in chronic colitis is accelerated by inflammatory activity. Mean age at diagnosis is 51.5 years — about 20 years earlier than the general population. Risk is also associated with disease duration: cumulative risk of colorectal cancer is 2.9% at 10 years, 5.6% at 20 years, and 8.3% after 30 years with the disease. With equivalent disease duration, the risk of cancer in Crohn's appears lower than in ulcerative colitis, but this may be informed by differences in disease distribution rather than a true biological difference.
Colorectal strictures in Crohn's carry a particularly increased cancer risk. Yamazaki analyzed 132 patients/175 strictures identified 1959-1980: 10 malignant strictures were identified in 9 patients (3 with ileocolic disease, 6 with colonic disease); frequency of cancer in patients with a stricture was 6.8%; all malignant strictures were short-segment. A separate Hungarian cohort of 640 patients over a 30-year period, including 62 patients with ileocolic or colonic strictures, observed a 6.5% rate of colorectal cancer in stricture patients — and all four stricture-associated colorectal cancer patients in this cohort were male smokers. Surgical management of colitis-associated dysplasia in Crohn's follows the same principles as ulcerative colitis.
Toxic colitis
Up to 50% of toxic colitis cases from IBD may be attributed to Crohn colitis. Severe colitis is defined per Truelove and Witt: ≥6 bloody bowel movements daily plus one of temperature >37.8°C, heart rate >90bpm, anemia (hemoglobin <10.5 g/dL), or ESR >30mm/h — combined with imaging showing colonic dilation and a disturbed/absent haustral pattern, constituting toxic megacolon. Surgery for clinical deterioration or rescue-therapy failure is total abdominal colectomy with end ileostomy, with or without mucous fistula.
Bleeding
Acute severe GI hemorrhage is a rare Crohn's complication (incidence 1-2%). Bleeding does not correlate with disease activity and can be duodenal, jejunoileal, ileocolic, or colonic. In contrast to ulcerative colitis (where hemorrhage is due to widespread mucosal ulceration), Crohn's hemorrhage is more often from a focal erosion into an intestinal vessel or occasionally an inflammatory pseudopolyp — making source localization critical to minimize unnecessary resection. Initial management follows standard GI-bleeding resuscitation with nasogastric lavage for source localization. Up to 30% of Crohn's patients with hemorrhage have a bleeding duodenal ulcer, and positive gastric lavage should trigger upper endoscopy. Colonoscopy within 24 hours of bleeding successfully locates the source 60-78% of the time; endoscopic control is not often successful but localizes the bleed to guide resection if surgery is required. CT angiography and traditional mesenteric angiography can also localize bleeding; when angiography localizes the site, methylene blue injection can stain a 10-40cm segment of involved bowel to allow ready identification during operative exploration.
Surgical considerations (general philosophy)
Crohn's cannot be cured surgically, so surgery — like medical treatment — should be palliative. Emphasis has swung from margin-negative resection to bowel-sparing approaches based on a landmark Fazio RCT: 152 patients undergoing ileocolic resection were randomized to a 2cm (limited) vs. 12cm (extended) proximal line of resection from the macroscopically involved area — no difference in recurrence rates, and recurrence did not increase when microscopic disease was present at the resection margins. Complete extirpation of microscopic disease is not the goal of surgery, as it does not produce cure and is frequently counterproductive; the aims are relief of complications and symptoms with bowel preservation. Nonresectional techniques (strictureplasty) avoid excessive small bowel loss; in isolated Crohn's colitis, especially if multifocal or associated with perianal disease, a more aggressive approach is often indicated.
Nutritional support and total parenteral nutrition
Exclusive enteral nutrition (EEN) has been shown to induce remission in pediatric Crohn's and to be as effective as systemic corticosteroids for inducing remission — a recent meta-analysis found intestinal healing significantly more likely with EEN vs. corticosteroids (OR 4.5, 95% CI 1.64-12.32). In adults, EEN is not as effective for inducing remission but may be useful for maintaining remission in patients with quiescent disease. Preoperative nutritional support has been evaluated primarily for optimization: in a recent Chinese study, 59.0% of screened patients were deemed at risk for malnutrition (likely an underestimate of the surgical-candidate population, since patients who have failed medical management are probably at even higher risk). Patients with serious nutritional deficits — weight loss >10% in the last 3-6 months, BMI <18.5 kg/m², or albumin <30 g/L — benefit from intensive enteral or parenteral nutritional support, reducing risk of surgical site infection and postoperative septic complications. Malnutrition is an independent risk factor for postoperative morbidity and mortality irrespective of immunosuppressive/biologic therapy status. Duration of preoperative nutritional support depends on operative urgency and GI-tract suitability for enteral administration.
Overview of operative considerations
Minimally invasive surgery (MIS): MIS in Crohn's poses real/perceived challenges (inflammatory mass/phlegmon increasing bowel-injury risk during laparoscopic manipulation; thick inflamed mesentery difficult to divide with a vessel-sealing device with limited tactile feedback), but evidence has accumulated favoring a tailored MIS approach. Milsom and Maartense both conducted prospective RCTs of MIS ileocecectomy: improved postop pulmonary function, morbidity, and length of stay vs. open; long-term follow-up of these trial populations supported improved body-image perception/cosmesis and decreased risk of bowel obstruction/hernia. Evidence does not support the concern that diminished tactile feedback leads to missed strictures or increased disease recurrence from incomplete resection. Robotic MIS is technically feasible with reduced morbidity/LOS vs. open (like laparoscopy), but longer operative times; conversion rate is lower than laparoscopy in some series; overall no clear demonstrated advantage of robotic over laparoscopic MIS yet. Intracorporeal anastomosis (avoiding a foreshortened mesentery preventing extracorporeal delivery) is enticing and may reduce SSI rate; using a small (<4cm) transverse extraction incision with intracorporeal anastomosis may decrease/eliminate incisional hernia (Heimann). Single-incision laparoscopic surgery (SILS, one incision + platform with 5mm camera + two working ports) has small studies supporting safety in Crohn's but no clear benefit over conventional laparoscopy. Robotic-assisted strictureplasty (Tou) and a robotic approach to stricturing disease (Scaringi, a nonresectional intracorporeal side-to-side isoperistaltic anastomosis) have both been described; further study needed.
Enhanced recovery pathways (ERPs): very little disease-specific data supports ERP application in IBD specifically. A recent review identified only about a dozen English-language ERP studies that included any proportion of IBD patients, and only 28.9% of the total patients across those studies had a stated surgical indication of IBD; most studies lacked important IBD-specific demographic information (biologic therapy, steroids, immune modulation status). Still, available evidence to date suggests ERP application to IBD colorectal surgery is safe and likely leads to decreased LOS without an increased readmission/morbidity rate. IBD is a known independent risk factor for prolonged LOS and should be factored into postoperative recovery expectations.
Perioperative medical management — steroids: concern over postoperative adrenal insufficiency/crisis has traditionally driven liberal perioperative stress-dose steroid use in steroid-treated/dependent patients, but much of the evidence around dosing/duration/indications for steroid supplementation is poorly supported and anecdotal (Truong). Dose and duration of steroid therapy do not correlate with degree/duration of HPA-axis suppression; recovery of HPA-axis function after cessation can be as short as 2 days and as long as 1 year, the basis for recommending stress-dose steroids for patients who have required steroid therapy within the past year. Because perioperative high-dose steroids are associated with impaired wound healing, reducing/omitting stress-dose steroids in Crohn's surgery is desirable. In a small pilot study, Zaghiyan did not administer any perioperative steroids to IBD patients who had received steroids within the year but were not on steroids at the time of surgery — all cases of hypotension/bradycardia/tachycardia spontaneously resolved without needing fluid bolus, vasopressor, or steroid administration. Further support suggests steroid-treated patients can simply be maintained on their usual preoperative steroid dose perioperatively; patients who have been treated with steroids in the past year probably do not need precautionary perioperative steroid supplementation at all, and high "stress-dose" steroids are unnecessary and may increase perioperative risk. There is a dose-dependent relationship between steroid use and infectious complications — highest risk in patients on >40mg prednisolone or equivalent. If surgery is elective/semi-elective, an attempt to wean steroids should be undertaken with the goal of being off steroids for 1 week prior to surgery; if complete cessation isn't possible, weaning to the lowest possible dose (target <20mg prednisolone-equivalent) is recommended.
6-mercaptopurine/azathioprine alone does not increase complication rates, and concurrent steroid+6-MP/AZA use does not further elevate complication risk beyond corticosteroids alone. No literature suggests increased complications with methotrexate. Discontinuing immunomodulator therapy prior to surgery appears unnecessary; these medications are typically held the morning of surgery and resumed per the gastroenterology treatment plan.
Perioperative medical management — biologics (Table 48.2 risk factors for postoperative septic complications): corticosteroid use, malnutrition/hypoalbuminemia, anemia/acute blood loss, emergency surgery, anti-TNFα therapy, vedolizumab therapy, penetrating disease/fistula/intraoperative abscess, recurrent disease, smoking. Anti-TNFα pharmacokinetics complicate interpretation of the evidence: one study found 50% of patients on anti-TNFα at time of surgery had no detectable drug levels immediately preoperatively, and complications do not correlate with anti-TNFα serum trough levels. Several heterogeneous retrospective/prospective studies both support and refute the hypothesis that anti-TNFα therapy significantly increases postoperative complications. A reasonable elective strategy is to delay surgery by 4 weeks (2 half-lives) from the last anti-TNFα dose; if not possible, consider temporary diversion if the patient has ≥2 other risk factors. Vedolizumab (anti-α4β7 integrin, approved since 2015) literature is similarly conflicted, though it has been associated with increased postoperative surgical-site infection risk in some studies — a reasonable strategy is to delay surgery by 6 weeks (2 half-lives) from the last dose, or consider temporary diversion if additional risk factors are present. Ustekinumab (anti-IL-12/23) does not appear to increase risk of postoperative septic complications.
Anastomotic type
Comparison of anastomotic techniques is based on safety (leak rate) and risk of recurrence, though surgeon experience with a given technique is a hard-to-quantify confounder. Limited level-1 evidence exists.
Side-to-side vs. end-to-end (non-Kono-S): Muñoz-Juárez's first case-controlled comparison (138 patients, evenly divided stapled side-to-side vs. hand-sewn end-to-end): clinical recurrence in 24% (side-to-side) vs. 57% (end-to-end); cumulative surgical-recurrence rates at 5 years were 11% (side-to-side) vs. 20% (end-to-end), p=0.017. A 2007 meta-analysis (8 studies, 661 patients, 712 anastomoses comparing end-to-end [53.8%] vs. other configurations including stapled side-to-side [46.2%, vast majority]) found no significant differences in overall complications, anastomotic recurrence, or surgical anastomotic recurrence — confirmed by a subsequent Cochrane review (Choy). Two more recent systematic reviews demonstrated no difference in anastomotic leak rates between side-to-side and end-to-end configurations; on surgical recurrence, Guo found no difference while Feng reported side-to-side superiority — the reviews' authors caution these should be interpreted carefully given the retrospective nature of most included studies. The CAST trial (2009, first RCT comparing anastomotic type, 139 patients, mean follow-up 11.9 months): endoscopic recurrence 37.9% (side-to-side) vs. 42.5% (end-to-end), p=0.55; symptomatic recurrence 22.7% vs. 21.9%, p=0.92 — no significant difference. A 2013 German RCT (primary endpoint: recurrence rate, stapled side-to-side vs. hand-sewn end-to-end; secondary endpoint: early postop complications) was terminated early for insufficient patient recruitment; no difference in postop complications, LOS, or surgery duration between the two techniques.
Kono-S anastomosis: developed 2011 by Kono — a hand-sewn antimesenteric functional end-to-end anastomosis intended to reduce surgical recurrence by preserving mesenteric vascularization/ innervation and creating a posterior supporting column maintaining 3D bowel structure (see Anastomotic Construction for full step-by-step technique). In Kono's original paper, 69 CD patients undergoing Kono-S anastomosis were compared with a historical cohort of 73 conventionally- anastomosed CD patients: significantly lower endoscopic recurrence in the Kono-S group at 5 years (0% vs. 15%, P=0.0013). Two subsequent large multicenter studies plus one prospective RCT have evaluated Kono-S further: Kono reported only 2 surgical anastomotic recurrences in 65 Kono-S patients (5- and 10-year surgery-free survival 98.6%); Shimada reported a surgical recurrence rate of 3.4% (Kono-S) vs. 24.4% (conventional end-to-end), with a higher risk of anastomotic leak in the end-to-end group (17.3% vs. 5.1%) — Kono-S had a significantly lower risk of 1-year anastomotic surgical recurrence (OR 0.14), and the 5-year surgery-free survival at the anastomosis site was significantly higher with Kono-S (95.0% vs. 81.3%, P<0.001). The first published RCT confirmed early reports of lower endoscopic recurrence rates, reduced severity of endoscopic scores, and lower rate of clinical recurrence favoring Kono-S. Overall, no definitive evidence yet supports the superior safety of one anastomotic technique over another, but the promising Kono-S results warrant further study.
Mesenteric excision: a recent report from Ireland compared wide excision of the mesentery with conventional closer division in 64 patients: surgical recurrence rates were 40% (conventional) vs. 2.9% (wide excision), p=0.003 — but this study has significant limitations (historical control group with longer follow-up, postoperative medical prophylaxis of recurrence not standardized, no data on anastomotic technique used).
Disease recurrence trends and surveillance
After ileocolic resection with anastomosis, recurrent Crohn's at the anastomotic site is seen on endoscopy in 70-90% of patients within 1 year, with 20-30% requiring additional operations within 5 years. Potential culprits for anastomotic-site recurrence include fecal stasis, microbiome alteration, and local ischemia, among others. Postoperative biologic therapy has been shown effective for preventing recurrence: Regueiro's small prospective RCT of early infliximab (5mg/kg) for 1 year vs. placebo found significantly lower 1-year endoscopic recurrence with infliximab (1 of 11 patients, 9.1%) vs. placebo (11 of 13, 84.6%), P=0.0006. The larger multicenter follow-up PREVENT trial did not meet its primary endpoint of lower clinical recurrence with infliximab, but patients on infliximab had lower endoscopic recurrence scores. Risk-stratification strategies based on clinical/early-colonoscopy findings have been proposed: De Cruz randomized 174 high-risk patients to early colonoscopy (with treatment step-up guided by clinical recurrence risk) vs. standard clinical observation, and found that early-colonoscopy-guided treatment step-up was better than conventional drug therapy alone for postoperative recurrence prevention. Selective therapy adjusted to individual recurrence risk leads to disease control in most patients, although low-risk patients should still undergo monitoring — early remission does not preclude the need for ongoing surveillance.
Operative considerations for specific locations
Gastroduodenal disease: clinically significant foregut Crohn's is rare (0.5-4% of patients); advances in endoscopy have improved detection — 30-50% of patients have macroscopic upper-GI disease, 40-70% have histologically visible upper-GI disease; at least 2/3 of these patients are asymptomatic, and >90% have coexisting more-distal Crohn's disease. Insidious gastritis-like symptoms; early satiety, postprandial pain/emesis, and weight loss suggest stricture, by far the most common gastroduodenal pathology. Fecal calprotectin is not a reliable indicator of gastroduodenal disease. Aphthous erosions, longitudinal ulcers, and a "bamboo joint"-like appearance in the cardia are characteristic of gastric Crohn's; longitudinal ulcers and notch-like erosions of Kerckring folds characterize duodenal disease. Therapy for symptomatic gastroduodenal Crohn's mirrors that of more distal disease plus acid suppression. Surgery for gastroduodenal disease is uncommon (<1% of all Crohn's surgery at tertiary centers); indications are obstruction and fistula. Almost all gastroduodenal fistulas result from penetration of the gastric/duodenal wall by another diseased site (terminal ileum, transverse colon) — takedown requires Kocherization of the duodenum, with the defect repaired primarily in one or two layers (low morbidity); a jejunal serosal patch is used for larger defects. Strictures are the most common indication for gastroduodenal intervention — endoscopic hydrostatic balloon dilation is feasible for short-segment strictures with low perforation risk, but recurrence is the rule and multiple dilations are often required. Surgical options for duodenal stricture are stricturoplasty (Heineke-Mikulicz or Finney, after Kocherization and visual/ palpation assessment; a 20mm Foley balloon may be floated in to check for more distal strictures) or bypass (gastroduodenostomy, gastrojejunostomy, or duodenojejunostomy — preferred with dense scarring, first/fourth-portion duodenal stricturing, or more than two strictures). Bypass procedures have low morbidity, but 25-33% of patients require reoperation for marginal ulceration or disease recurrence; vagotomy does not decrease the rate of marginal ulceration and need not be performed. Pancreaticoduodenectomy should be reserved only for instances of severe ampullary dysfunction and/or cholangitis.
Upper small bowel disease (jejunum and ileum, excluding the terminal ileum): affects 3-10% of Crohn's patients; obstruction and sepsis are the two most common surgical indications (massive hemorrhage and carcinoma much less common). The approach has shifted from extensive resection with negative-microscopic-margin intent to resection of only the macroscopically diseased segment and/or bowel-sparing strictureplasty. Patients with jejunoileitis have a particularly high reoperation rate (up to 30%) and are at real risk of short bowel syndrome (reported in 8.5% of cases at 20 years after the index surgery). A significant proportion of small bowel strictures are not identified on preoperative workup, so the entire small bowel should be examined at exploration (calibration sphere or cuffed catheter to assess lumen when dealing with multifocal disease). Fibrostenotic strictures with luminal diameter <20mm are a clear indication for strictureplasty or resection (cutoffs vary by patient/bowel size and prior resections); less critical strictures, especially with prior resections and extensive disease, may not mandate operative treatment in an era of effective medical therapy. Strictures may be marked with metal clips for future reference; measuring remaining bowel length and recording stricture locations in the operative report matters for long-term management.
Strictureplasty techniques: Heineke-Mikulicz is the most commonly performed technique, suited for short-segment (<10cm) chronic strictures — a single longitudinal incision over the antimesenteric border, extending 2cm beyond thickened portions, closed transversely. Finney strictureplasty is used for strictures longer than 10cm but shorter than 25cm. The Jaboulay strictureplasty is also used for medium-sized (>10 and <25cm) strictures, sparing bowel length via a lateral diverticulum but with a resulting blind loop carrying potential stasis/bacterial overgrowth risk — described by Lee (1982) as a shorter "bypass" of a diseased segment. Strictures longer than 25cm treated with Finney/Jaboulay technique risk a functional large blind loop with bacterial overgrowth and short bowel syndrome; for these, Michelassi proposed an isoperistaltic side-to-side strictureplasty for significantly long-segment strictures (>20cm) or multiple short strictures in tandem (avoiding the unsafe multiple- Heineke-Mikulicz approach): the bowel and mesentery are divided at the midpoint of the strictured segment, the two loops are approximated with interrupted seromuscular Lembert sutures, a longitudinal enterotomy is performed on both loops with tapered ends to avoid blind stumps, and the anastomosis is completed with a circumferential luminal suture layer plus an outer layer of interrupted seromuscular Lembert stitches — preserving long segments of bowel with excellent long-term results (one series with 7.5-year follow-up in 20 patients showed radiographic, endoscopic, and histopathologic regression of previously active Crohn's disease with restoration of intestinal function).
Multiple studies confirm safety/efficacy of both short and long strictureplasties, with early postoperative complications (bleeding, sepsis) in 8-15% of cases. Given case reports of cancer arising at strictureplasty sites in long-standing disease, mucosal biopsy and marking the site with a metal clip should be considered, especially in long-standing disease. Reese compared recurrence rates between strictureplasty and resection: surgical recurrence was more likely after strictureplasty (p=0.09), with a significantly longer recurrence-free interval after resection (p=0.01). Overall recurrence rates after strictureplasty have been reported between 18% and 29%, but in one study only 4.6% of recurrences were at the previous strictureplasty site itself.
Colonic and rectal disease: operations include total proctocolectomy, total abdominal colectomy with ileostomy or ileorectal anastomosis, and segmental resection. Toxic colitis typically requires total abdominal colectomy with end ileostomy irrespective of rectal involvement. Rectal-stump management options when the rectum has significant inflammation include leaving it intraperitoneally (± rectal tube), or extraperitoneal fixation in the LLQ or above the fascia in the lower midline (so a staple-line leak manifests as an abdominal-wall rather than pelvic abscess), or maturing it as a mucus fistula; a systematic review of rectal-stump management (in ulcerative colitis) found subcutaneous placement associated with the lowest morbidity. Patients with rectal involvement plus colitis are poor candidates for limited resection (no good distal anastomotic target); total proctocolectomy with end ileostomy is typically indicated. Poor nutritional status or high-dose steroids favor a near-total proctocolectomy with an ultralow Hartmann pouch and end ileostomy (deferring completion perineal proctectomy) to avoid high perineal-wound-sepsis risk, or a total abdominal colectomy with end ileostomy and later completion proctectomy. For short-segment Crohn's colitis with rectal sparing, segmental resection is an option — Table 48.3 compares rates across procedures: segmental resection carries 26-55% colon/rectal disease recurrence, 4-14% small bowel disease recurrence, 11-66% further intestinal surgery, and 5-44% permanent stoma; total abdominal colectomy carries 24-66% colon/rectal recurrence, 8-21% small bowel recurrence, 30% further surgery, and 25-50% permanent stoma; total proctocolectomy carries up to 39% small bowel recurrence (colon/rectal recurrence not applicable), 9-32% further surgery, and up to 100% permanent stoma. Only total proctocolectomy with end ileostomy minimizes risk of disease recurrence, though even after this there is up to 39% small bowel recurrence with up to 32% requiring surgical intervention at 10 years. Smoking is associated with need for further intestinal surgery/eventual proctectomy. Perianal disease, young age, and female sex are independent risk factors for disease recurrence and eventual permanent stoma. These recommendations apply to Crohn's colitis being managed with rectal-sparing intent, not to patients with dysplasia.
Ileal pouch-anal anastomosis in Crohn's disease: because Crohn's can be difficult to distinguish from ulcerative colitis, some patients thought to have UC who undergo IPAA are later found to have Crohn's. Older retrospective series of these patients demonstrated high complication/pouch-failure rates (up to 56% required pouch excision, with a further proportion undergoing indefinite diversion). Preoperative features suggestive of Crohn's (subtle perianal disease, discontinuous inflammation) do very poorly with no meaningful symptom-free interval after pouch formation. However, patients with Crohn's who underwent IPAA in these series who did well had similar functional results to UC patients. Two more recent studies suggest lower rates of pouch loss/indefinite diversion: a large prospective Cleveland Clinic series (2013) reported 13.3% pouch-failure rate in Crohn's patients vs. only 5.1% in ulcerative colitis. Li evaluated intentional IPAA and ileorectal anastomosis for Crohn's (noting the two patient populations have distinctly different disease characteristics), reporting a 15.5% indefinite-diversion rate in the IPAA group. Taking this data together, highly selected patients with Crohn's colitis and no perianal disease and no small bowel disease may consider restorative IPAA, provided risk tolerance and shared decision-making priorities of patient and surgeon are aligned.
Ileosigmoid fistula: a common complication of perforating terminal ileal Crohn's — typically the inflamed terminal ileum adheres to an otherwise normal sigmoid colon free of primary Crohn's involvement. Most ileosigmoid fistulas are small, may be asymptomatic, and do not require operative management by themselves; larger fistulas can bypass intestinal contents from the terminal ileum to the distal colon, causing debilitating diarrhea requiring surgery. More than half of ileosigmoid fistulas are not recognized preoperatively despite imaging/endoscopic evaluation, so surgeons should be prepared for this complication in any terminal-ileal Crohn's case. Management: simple division of the fistulous adhesion plus standard resection of the diseased terminal ileum, with the sigmoid defect debrided and simply closed (manages 75% of ileosigmoid fistulas). The remainder require sigmoid colon resection — necessary when the sigmoid is also involved with Crohn's, the fistulous opening is particularly large, there is extensive fibrosis extending along the sigmoid, or the tract enters the sigmoid near the mesentery (difficult to close primarily, often requiring resection and primary anastomosis).
Complex perineal wounds: perianal Crohn's occurs in about one-third of patients with intestinal Crohn's and is covered in the dedicated chapter (see ASCRS Ch47 - Anorectal Crohn's Disease / "Perianal (anorectal) Crohn's disease" on Crohn Disease). Complicated/active rectal disease significantly increases the need for proctectomy; aggressive medical management (antibiotics, biologics) plus drainage of local sepsis is a mainstay in attempting to avoid proctectomy.
Conclusion
Management of Crohn's disease is complex and requires a multidisciplinary team approach. Surgical intervention is reserved for refractory disease or complications. Much work remains to understand how newer medications will shift surgical treatment and whether specific surgical techniques lower the risk of recurrent disease.