Patients previously classified with “obscure gastrointestinal hemorrhage” should now be classified as “suspected small bowel bleeding” according to the 2015 American College of Gastroenterology guidelines. This article provides algorithms for how to manage patients with suspected small bowel bleeding, including utilization of second-look endoscopy and/or colonoscopy, video capsule endoscopy, computed tomographic enterography, magnetic resonance enterography, angiography, and deep enteroscopy.
Key points
- •
Video capsule endoscopy (VCE) should be the third diagnostic test after routine upper and lower endoscopic examinations.
- •
Computed tomographic enterography or magnetic resonance enterography (MRE) examinations should be performed in patients with suspected small bowel strictures from nonsteroidal anti-inflammatory agents, inflammatory bowel disease, prior radiation, or after normal VCE examinations.
- •
Patients with abnormal findings on small bowel imaging should be treated using deep enteroscopy, angiography, or surgery.
- •
For patients with ongoing bleeding or anemia and negative testing should undergo repeated testing with VCE, enteroscopy, or other testing until a source is identified.
- •
The algorithms for patients with overt and/or occult suspected small bowel bleeding are equivalent.
Introduction
Formerly classified as “obscure [gastrointestinal] GI bleeding,” this diagnostic term has been replaced by “suspected small bowel bleeding” or “small bowel bleeding” in the recent 2015 American College of Gastroenterology (ACG) guidelines. The reason for this change in classification is because most bleeding sources outside the upper or lower digestive tracts are found within the small intestine on further investigation with video capsule endoscopy (VCE), deep enteroscopy, computed tomography enterography (CTE), or magnetic resonance enterography (MRE) examinations. This article provides clinicians with algorithms for how to manage patients with suspected small bowel bleeding.
An initial algorithm for patients with obscure GI bleeding developed at the International Conference on Capsule Endoscopy (ICCE) conference in 2005 was published by Pennazio and colleagues and is shown in Fig. 1 . An important recommendation of all algorithms to date has been the consideration for second-look endoscopic examinations before proceeding with VCE. The rationale for this recommendation has that second-look examinations, particularly upper endoscopic examinations, have been associated with diagnostic yields ranging from 3% to 60%, as shown in Table 1 . These numbers were derived from patients undergoing push enteroscopy, VCE, or deep enteroscopy. The most common lesions found on repeat examinations were angiodysplastic lesions. The algorithms presented in this article were recently published in the updated 2015 ACG guideline “Management of Small Bowel Bleeding.”
| Author, Year | Modality | Number of Subjects, Diagnostic Yield (%) | EGD or Colonoscopy (Colon), Yield (%) | Most Common Findings |
|---|---|---|---|---|
| Zaman & Katon, 1998 | PE | 95 (41%) | EGD: 25 (64%) |
|
| Descamps et al, 1999 | PE | 233 (53%) | EGD: 25 (10%) | Cameron ulcerations (N = 6) GU (N = 5) DU (N = 1) GJ anastomosis (N = 1) Gastric angiodysplasia (N = 5) Duodenal angiodysplasia (N = 4) Duodenal tumor (N = 1) Hemosuccus pancreaticus (N = 2) |
| Lara et al, 2005 | PE | 35 (56%) | EGD: 10 (59%) | GU (N = 2) Esophageal varices (N = 2) DU (N = 1) Gastric remnant (N = 1) Gastric AVM (N = 2) Cameron ulcerations (N = 2) |
| Fry et al, 2009 | DBE | 107 (65%) | EGD: 13 (12%) Colon: 12 (11%) |
|
| Van Turenhout et al, 2010 | VCE | 592 (49%) | EGD: 32 (17%) Colon: 8 (4%) |
|
| Lorenceau-Savale et al, 2010 | VCE | 65 (57%) | EGD or colon: 8/35 (23%) |
|
| Robinson et al, 2011 | VCE | 707 (40%) | EGD: 22 (3%) Colon: 6 (1%) |
|
Summary Algorithm
As shown in Fig. 1 , patients may present with overt (melena or hematochezia) or occult (iron-deficiency anemia with or without heme-positive stools) suspected small bowel bleeding. The first step that the clinician needs to take is whether to repeat upper and or lower endoscopic examinations. This decision is based on a variety of factors, including the date of the prior examination and the quality of the prior examination, whether there was blood present in the upper tract, and the quality of the bowel preparation. If the patient did not have an examination of the upper tract within the preceding 2 to 3 months and presents with symptoms suggesting of upper GI bleeding, it is not unreasonable to repeat an upper endoscopy. Data regarding the diagnostic yield associated with repeat upper examinations show diagnostic yields ranging from 25% to 60% (see Table 1 ). The most commonly missed lesions are angiodysplastic lesions and ulcerations.
In the 2005 algorithm that was proposed as part of the ICCE consensus conference and cited in the 2007 guidelines by the American Gastroenterological Association, angiography is recommended in cases of massive overt bleeding. If the angiography is negative or if patients present with stable overt or occult bleeding, VCE is recommended as the third diagnostic test. The reason for this recommendation (see later discussion) is the ability of VCE to visualize the entire small bowel in more than 80% of cases, particularly now with 12-hour battery life and the noninvasive nature of the test. In a subsequent 2008 meta-analysis, VCE was shown to have a higher diagnostic rate compared with double-balloon enteroscopy. In addition, performance of a deep enteroscopy after initial VCE has been shown to increase the diagnostic yield of the enteroscopy examination.
In the 2005 algorithm and subsequent algorithms published in the 2015 ACG guidelines, the recommendation was to perform a therapeutic procedure if the VCE demonstrated an abnormal finding. Endoscopic therapy of small bowel angioectasia has been shown to reduce subsequent risk of rebleeding. On the other hand, rebleeding rates from angioectasia have been shown to be significant, approaching 30% to 40% per year, particularly for patients who are at risk, including patients with Heyde syndrome or renal failure. In some of these patients with refractory bleeding or for elderly patients who are not candidates for endoscopic intervention, medical therapy with octreotide or thalidomide can be considered.
On the other hand, for the patients with angioectasias who are no longer anemic, observation or treatment with iron has been recommended because 30% to 40% of patients with isolated lesions have been shown to have spontaneous cessation of bleeding without endoscopic therapy.
In the 2015 ACG Guideline, the term obscure GI bleeding was replaced by the term small bowel bleeding, with the term obscure reserved for patients without a source identified on VCE, deep enteroscopy, or enterography examination. The reason for this change in terminology was related to several advancements: (1) the literature on repeat endoscopic examinations of the upper and/or lower tracts demonstrated findings in 30% to 40% of patients, (2) utilization of VCE or deep enteroscopy detected lesions in the small bowel in 50% to 60% of patients, and (3) in patients with normal VCE or enteroscopy examinations, disease is found on CTE or MRE in 40% to 50% of patients.
The algorithm for suspected small bowel bleeding from the 2015 guideline is shown in Fig. 2 . Realizing that VCE retention can occur in approximately 1% to 2% of patients with small bowel bleeding and up to 5% to 10% of patients with suspected or established inflammatory bowel disease (IBD), the new guidelines recommend imaging with CTE or MRE before VCE in patients with possible signs of obstruction. Radiographic imaging should occur in patients with abdominal pain, history of abdominal radiation, heavy use of nonsteroidal antiinflammatory drugs, or history of IBD. Use of the patency capsule, a capsule with a lactose body that dissolves into a radiofrequency identification tag, is another option instead of CTE or MRE for patients with suspected stricture. If the CTE or MRE is negative, then VCE is recommended with specific therapy if findings are identified. Use of intraoperative enteroscopy has been mainly replaced by deep enteroscopy, with the exception of patients with adhesions from prior surgeries that require surgical lysis to facilitate small bowel advancement. In patients with normal VCE examinations, recommendations are to proceed with CTE or MRE to evaluate for submucosal disease instead of deep enteroscopy, unless there is high clinical suspicion for small bowel angioectasia.
