Principles of Endoscopic Therapy in Inflammatory Bowel Diseases

List of Abbreviations








Adipose-derived mesenchymal stem cells


Anti-outer membrane porin C


The American Society for Gastrointestinal Endoscopy


Colitis-associated neoplasia


Caspase recruitment domain-containing protein 15


Crohn’s disease


Controlled radial expansion


Computed tomography enterography


Endoscopic balloon dilation


Entero-cutaneous fistula


Extracellular matrix


Entero-enteric fistula


Electrolyte solution


Endoscopic mucosal resection


Endoscopic stricturotomy


Endoscopic submucosal dissection


Examination under anesthesia


Anti- Saccharomyces cervisiae




High-grade dysplasia


Human immunodeficiency virus


Inflammatory bowel disease


Ileocolonic anastomosis


Ileocolonic resection




Ileal pouch-anal anastomosis


Ileorectal anastomosis


Ileosigmoid anastomosis


Ileosigmoid fistula


Low-grade dysplasia


Ligation of intersphincteric fistula tract


Matrix metalloproteinases


Magnetic resonance enterography


Nucleotide-binding oligomerization domain-containing protein 2/caspase recruitment domain-containing protein 15


Nonsteroidal antiinflammatory drug




Procedure-associated complications


Primary sclerosing cholangitis


Pouch-vaginal fistula


Rectal-bladder fistula


Rectal-vaginal fistula


Tumor necrosis factor


Tumor necrosis factor superfamily-15




Ulcerative colitis


Dr. Bo Shen is supported by the Ed and Joey Story Endowed Chair.


Stricture and fistula are common complications in Crohn’s disease (CD), which have been considered as the natural consequence of bowel inflammation. The disease phenotype of CD is classified as nonstricturing/nonpenetrating (B1), stricturing (B2), or penetrating (B3), based on the Montreal Classification. A population-based study showed that 81% of the patients had nonstricturing/nonpenetrating disease, 4.6% had stricturing disease, and 14.0% had penetrating disease, at diagnosis. However, a majority of patients with inflammatory CD progresses into stricturing and/or penetrating complication over time. Reported cumulative risk for the development of either stricture or fistula ranged from 34% to 52% at 5 years, and 40%–70% at 10 years after diagnosis. Therefore, the longer the duration of the disease is, the higher risk the patient has for complications.

A combined approach with medical, endoscopic, and surgical therapy is often needed for the management of CD-associated strictures. The inflammatory component of a stricture may benefit from the treatment with antiinflammatory medications. In contrast, the mechanical nature of fibrostenotic stricture is not amenable to the medical therapy. In fact, patients with predominantly stricturing CD were excluded in the majority of published randomized controlled trials in antitumor necrosis factor (TNF) or antiintegrin biologics, as reviewed by our group. On the other hand, surgical intervention, including bowel resection with anastomosis and strictureplasty, is effective in the treatment of CD strictures, and it is often associated with a high risk of operative complications and postoperative disease recurrence.

Management of fistulizing CD has been challenging; endoscopically, medically, or surgically. Patients with perianal fistulae may respond to therapy with antibiotics, immunomodulators, anti-TNF, anti-integrin, or antiinterleukin (IL) biologics. In contrast, medical therapy has a limited role in epithelialized hollow-organ-to-hollow-organ fistulae, such as rectal-bladder fistula or rectal-vaginal fistula (RVF). Surgical options for CD-associated fistulae range from simple seton or mushroom catheter placement, fistulotomy, fistula plug, mucosal or muscle flaps, and ligation of the intersphincteric fistula tract, to invasive resection of diseased bowel and strictureplasty with or without fecal diversion. However, surgical procedures are invasive and technically challenging and carry a risk of postoperative complications and disease recurrence, particularly in those with complex fistula or anal fistula with rectal involvement by CD.

Endoscopy also plays an important role in the diagnosis, differential diagnosis, disease monitoring, and dysplasia surveillance in patients with ulcerative colitis (UC). With advances in imaging technology, we are able to detect more dysplastic lesions during surveillance colonoscopy, which has led an attempt to remove some dysplastic lesions, endoscopically. However, the underlying inflammatory and fibrosis process in IBD make the endoscopic treatment of IBD-associated dysplastic lesion challenging.

The majority of patients and a significant number of patients with UC would eventually require surgery. Disease recurrence and postsurgical complications are common, including stricture, fistula, and anastomotic leak, and abscess. Those complications may be amenable to endoscopic therapy.

In the past decade, endoscopic therapy has emerged as a valid alternative or a bridging approach to medical and surgical therapy, as it is more effective than medical therapy and less invasive than surgical intervention for mechanical complications of the disease. Endoscopic therapy of IBD is technically demanding, and endoscopist should be medically and mentally prepare the patient as well as himself or herself. It is imperative to identify proper patients, proper lesions, and proper time to endoscopically intervene, to organize a proper team with proper room and equipment settings, and to have a backup and rescue plan in mind.

The Patient

Since endoscopic therapy is considered a new approach for patients with complications in IBD, the patient may not aware the therapeutic aspects and possible safety issues of endoscopic therapy. Detailed procedure process, risks, benefits, and expectation should be fully explained to the patient. In most cases, repeat endoscopic therapy may be needed, and some of the patients may eventually need surgery.

Not all patients with IBD-related strictures or fistula are eligible for endoscopic therapy. Therapeutic endoscopy may be avoided, if possible, in patients with following conditions: (1) malnutrition or severe comorbidities and poor candidates for operative intervention of rescue surgery in case of endoscopy-associated complications; (2) emergency setting; (3) bleeding disorders or concurrent use of anticoagulants; (4) concurrent use of immunosuppressive agents, especially systemic corticosteroids; (5) short gut syndrome; and (6) small bowel transplantation. If the endoscopic therapy is absolutely necessary, extreme precaution should be taken. The decision for pursuing the endoscopic therapy should be made by the multidisciplinary team along with the patient and family.

The detailed information in the disease history of the patient is an important factor for determining whether medical versus endoscopic versus surgical approach or combination is preferred. For example, endoscopic therapy may be offered to a CD patient with a short (≤4 cm) fibrotic stricture, a long history of the disease (e.g., 15+ years) without previous bowel resection, minimum concurrent bowel inflammation, a single recurrent anastomotic stricture, or recurrent anastomotic stricture with multiple prior resections in the area.

General medical conditions in UC patients with stricture or anastomotic sinus in the ileal pouch often is not as unhealthy as those with CD complications. In most cases, endoscopic therapy can be safely and effectively performed in those patients. The presence of concurrent primary sclerosing cholangitis (PSC) in UC and Crohn’s colitis patients with portal hypertension and possible thrombocytopenia or coagulopathy may pose a unique challenge of procedure-associated bleeding.

It is critically important to appreciate anatomy in healthy and diseased gut. Before we deliver any endoscopic therapy, we should have potential “land mines” in mind. The anatomy of human body is also a determining factor for the aggressiveness of endoscopic therapy ( Fig. 10.1 ). For example, stricture at the pylorus, which is considered a safer zone for the risk of perforation, can be treated aggressively. Endoscopic needle-knife stricturotomy has advantages over balloon dilation, as endoscopist can have a control on where to cut and how deep to cut ( Fig. 10.2 ). Additionally, surgical endoscopic procedure-associated perforation in the esophagus or duodenum may carry far more severe consequences than that in distal bowel. Perforation of high ileocolonic anastomosis (ICA), often located at the right upper quadrant, during stricture therapy will have more detrimental consequences than that in low ICA, due to the presence of multiple adjacent organs in the former ( Figs. 10.3 and 10.4 ). Another example is the endoscopic treatment of distal rectum or distal pouch disease. The anterior wall of anal canal, distal rectum, and distal pouch is adjacent to the vagina in female patients and prostate in male patient. In the left lateral decubitus position, the vagina is at the 4–5 o’clock position ( Fig. 10.5 ). Therefore, endoscopic therapy, such as balloon dilation or endoscopic stricturotomy, should stay away from the anterior wall, to prevent iatrogenic trauma and development of vaginal fistula. The treatment of distal bowel disease in patients with a diverting ostomy can be little more “aggressive” as procedure-associated perforation (except for anterior wall perforation of the distal bowel) would bear milder consequences with rare occurrence of pelvic abscess or sepsis ( Fig. 10.6 ).

Figure 10.1

“Safer” and “riskier” zones in therapeutic endoscopy in IBD. (A) The sequence from safer to riskier zones ranges from green to yellow then red ; (B) IBD surgery often alters bowel anatomy. For example, the anastomosis following ileocolonic resection usually located at the right upper quadrant. Iatrogenic perforation from endoscopic treatment for the lesion (such as balloon dilation of stricture) can carry detrimental consequences. IBD , inflammatory bowel disease.

Figure 10.2

Endoscopic stricturotomy in the safer (for risk of perforation) zone (pylorus). (A) Tight stricture at the pylorus from Crohn’s disease; (B) close view of the strictured pylorus; (C and D) needle-knife stricturotomy at the anterior wall, to minimize the injury to the gastroduodenal artery at the posterior wall.

Figure 10.3

Anastomotic leak at transverse staple line of side-to-side ileocolonic anastomosis in Crohn’s disease. (A) The leak on colonoscopy ( green arrow ); (B and C) perihepatic fluid and gas collections ( red arrows ) and staple line at the anastomosis ( yellow arrow ); (D) artist’s illustration of side-to-side anastomotic leak (leading to entero-cutaneous fistula) and transverse staple line leak (leading to abdominal abscess or peritonitis).

Figure 10.4

Endoscopic dilation of duodenum cap stricture with perforation. (A) The tight stricture; (B and C) iatrogenic perforation ( green arrows ).

Figure 10.5

Location of the vagina in endoscopy-left lateral decubitus position. (A) Artist’s sketch of location of bowel, presacral abscess, or sinus and vagina; (B) MRI pelvis showed the location of the vagina at 4–5 o’clock ( green arrow ) and presacral sinus at 10–11 o’clock ( yellow arrow ). Extreme care should be taken during endoscopic treatment of lesion at right lower quadrant, to avoid iatrogenic injury, such as vaginal trauma with fistula formation.

Figure 10.6

Perforation during endoscopic therapy for a distal rectal stricture in a Hartmann pouch. The patient fully recovered with medical or surgical intervention. (A) Strictured distal rectum due to fecal diversion (embedded picture end ileostomy); (B) endoscopic stricturotomy. The air insufflation of diverted rectum resulting in microperforation in retroperitoneal space (C) and mediastinum (D) ( green arrows ).

The Disease

It is important to understand the mechanism of stricture and fistula formation in IBD ( Chapter 3, Chapter 4 ). In addition to patient’s general health condition, the determining factors for the type and success of endoscopic intervention are the nature of stricture/fistula, the altered anatomy in those with a prior history of surgery, available endoscopy tools, and training and experience of the therapeutic endoscopist and the multidisciplinary team and immediate availability of surgery backup.

The disease process of CD is characterized by the evolution of phenotypes, from inflammatory to fibrostenotic and fistulizing entities. We may use the following phrase to describe the natural history of CD as “no inflammation, no stricture; no stricture, then no fistula; and no fistula, no abscess,” in the majority of patients with bowel stricture and fistula. This notion and clinical observation have led to a rationale of treating the disease aggressively at the time of diagnosis in those at risk for disease progression (such as active smoking, multiple locations of disease, the presence of upper gastrointestinal [GI], rectal, or perianal disease), with the intention to modify the disease course. Once “mechanical” complications such as strictures and fistulae develop, the role of medical therapy is limited. Those “mechanical” complications usually require the treatment in a “mechanical way”, with endoscopic and/or surgical therapies.

Development of Stricture

A stricture is defined as the narrowing of the intestinal lumen with or without prestenotic dilation. In histopathology, a strictured area comprises of fibroblasts, and collagens, intervening acute or chronic inflammatory cells. Strictures are divided into inflammatory, fibrostenotic, and mixed types. The primary strictures, that is, those related to disease, arise from chronic persistent inflammation, regeneration of tissues, and impaired tissue repair in the bowel wall. The inflammatory stricture develops from edema and infiltration of inflammatory cells in the bowel wall, leading to an increase in wall thickness and a significant reduction in luminal area. In contrast, fibrostenotic stricture has been considered as the consequence of overproduction of extracellular matrix (ECM) during chronic inflammation. The tissue injury of the inflamed bowel wall causes the accumulation of mesenchymal cells, which subsequently secret ECM for tissue repair. The persistent and/or severe inflammation of patients with CD or UC often leads to an excessive production of ECM, and ultimately intestinal fibrosis. The CD-associated, primary strictures contain mainly fibrostenotic tissue with nonreversible mechanical nature. There have been concerns that repeat endoscopic therapy with barotrauma, such as balloon dilation, may make the tissue prone to the development of perforation. Therefore, electric incision using endoscopic knife may help minimize the barotrauma and reduce the risk for perforation.

In addition to the disease process of CD or UC, many factors may contribute to the development of stricture in those patients, including smoking, cardiovascular disease, and medications, for example, nonsteroidal antiinflammatory drugs, potassium chloride tablets, oral pancreas enzyme supplements, surgery-related ischemia, or even healing process after medical therapy for bowel inflammation, particularly in anti-TNFα therapy. Genetic factors play an important role in the initiation and progression of CD-associated strictures. The most common genetic variants associated with fibrostenotic CD are nucleotide-binding oligomerization domain-containing protein 2/caspase recruitment domain-containing protein 15 (NOD2/CARD15) mutations. genetic mutations of matrix metalloproteinases (MMP) , and tumor necrosis factor superfamily-15 (TNFSF15) were reported to be associated stricturing CD. Reported serological markers for the prediction of stricture formation include anti-outer membrane porin C, anti- Saccharomyces cervisiae , anti-laminaribioside, anti-chitobioside, and anti-mannobioside.

It should be pointed out that colonic stricture in patients with UC should be evaluated for colitis-associated neoplasia (CAN). In addition, benign strictures can also occur in patients with UC, resulting from mucosal and submucosal fibrosis and hyperplasia of muscularis mucosae.

Development of Fistula

A fistula is an abnormal connection between two epithelialized or endothelial surfaces. The common areas connected by fistula in CD are the intestine and cutaneous tissue (perianal fistula and entero-cutaneous fistula [ECF]), intestine to vagina (entero- rectal- or ileal pouch-vaginal fistula), intestine to bladder (entero-bladder fistula), and intestine to intestine (entero-enteric fistula [EEF]).

The pathogenesis of CD-associated fistulae is not fully understood ( Chapter 4 ). It appears that perianal fistula in patients with CD respond to medical therapy best. The current hypothesis holds that fistula outside perianal zone results from the presence of concurrent downstream luminal stricture in the environment prone to the development of transmural inflammation. Multiple studies have shown that fistulae were almost exclusively found in patients with strictures. Subsequent to the formation of stricture, fistula may occur at the bowel segment proximal to the stricture due to increased intraluminal pressure. Often fistula evolves through the muscularis propria along with blood vessels, in which there is minimum mechanical resistance to an increased intraluminal pressure. Herniated mucosa is occasionally found in some of the patients with fistula, suggesting that fistula develops under an increased luminal pressure. Therefore, the mechanical pressure from a concurrent downstream stricture is considered as the main contributing factor to the development of fistula. Once a fistula is formed, persistent pressure of luminal content at its feeding orifice in the setting of transmural inflammatory process may result in abscess (contained perforation) or even free perforation.

In addition to CD per se, diseases including human immunodeficiency virus infection and other sexually transmitted disease, bowel surgery, radiation, and trauma can also cause fistula. CD patients are not immune to those in juries or disorders.

Patients with UC can also develop perianal complications, including perianal fistula. Although any perianal suppurative lesions should raise the suspicion for CD, patients with UC can sometimes anorectal abscess. In a study of 763 patients with UC, 13 developed perianal abscess. The etiology of anorectal abscess in UC is not clear. It appears that patients with severe UC and/or on strong immunosuppression medication may have a high risk. It is possible that transmural inflammation in severe UC may facilitate the development of perianal abscess. In addition, distal rectal inflammation may affect the anal crypts, leading to cryptoglandular fistula or abscess. In addition to perianal fistula, even RVF can occur in patients with UC. RVF in UC is strongly associated with severe disease activity.

Patients with UC and restorative proctocolectomy with IPAA can have various complications, including stricture, fistula, and acute and chronic anastomotic leaks. Those complications can result from de novo CD of the pouch or CD-like condition of the pouch or surgery itself or both. The treatment of the complications from two different categories of etiology is different. For example, endoscopic therapy, including balloon dilation or stricturotomy, is more effective in patients with surgery-associated stricture than CD-associated stricture. Pouch revision surgery is more effective in surgical technique–associated fistula than those with CD-associated fistula. Therefore, the differential diagnosis is important for the management and prediction of treatment response and prognosis. However, differential diagnosis for the cause of stricture, fistula, and anastomotic sinus between CD of the pouch and surgery-associated complication is often difficult. Another example is pouch-vaginal fistula (PVF), which can be related to CD of the pouch, anastomotic leak, iatrogenic trauma from the stapler or cryptoglandular abscess ( Fig. 10.7 ). The presence of following presentations may suggest a diagnosis of CD-related PVF: (1) delayed onset, that is, more than 6–12 months after ileostomy closure for IPAA; (2) fistula opening at the anal canal, rather than at anastomosis or dentate line; (3) inflammation around the fistula opening at the anal canal; and (4) failure to multiple attempts for fistula repair. In the end, a diagnostic trial of anti-TNF agent may be helpful.

Dec 30, 2019 | Posted by in GASTROENTEROLOGY | Comments Off on Principles of Endoscopic Therapy in Inflammatory Bowel Diseases

Full access? Get Clinical Tree

Get Clinical Tree app for offline access