31 Small Intestinal Diseases Beyond the Duodenum
Jonathan A. Leighton and Lucinda A. Harris
Endoscopically the small intestine has been a challenging part of the gastrointestinal (GI) tract to investigate. Advances in video capsule endoscopy (VCE) and the development of deep enteroscopy (single-balloon enteroscopy [SBE] and double-balloon enteroscopy [DBE]) have introduced new endoscopic modalities which can be used in the diagnosis and sometimes treatment of these diseases. ▶Table 31.1 presents an overview of categorization of small intestinal diseases used in this chapter.
Medication (e.g., olmesartan, valsartan)
Medications (NSAIDs, potassium, 6-mercaptopurine)
NSAIDs, nonsteroidal anti-inflammatory drugs.
The primary indications for small bowel endoscopy are evaluation of GI bleeding, identifying small bowel Crohn’s disease, diagnosing small bowel tumors, and evaluation of malabsorption syndromes, particularly celiac disease. Somewhat less frequently small bowel endoscopy plays a role in identifying certain infections, congenital lesions, and miscellaneous conditions. This chapter summarizes the important diagnostic and therapeutic features of endoscopy in these areas.
31.2 Suspected Small Bowel Bleeding
Historically, bleeding in the small bowel has been difficult to diagnose and treat because of relative inaccessibility to the deep small bowel. However, with the development of newer modalities such as capsule endoscopy (CE) and deep enteroscopy, the ability to diagnose and treat bleeding lesions throughout the small bowel has improved considerably. In fact, because the ability to examine the small bowel has improved significantly, the term “obscure gastrointestinal bleeding” has evolved and the term now used is “suspected small bowel bleeding.” 1 The term “obscure gastrointestinal bleeding” is reserved for truly obscure causes beyond the small bowel, such as vasculitis and hemobilia. It is important to remember that small bowel bleeding can be overt and present with melena or hematochezia or can be occult and present with iron deficiency anemia. Potential etiologies can usually be classified as vascular, inflammatory, or neoplastic (▶Table 31.1). The type of lesion correlates to some extent with age but not gender or ethnicity (▶Table 31.2). Vascular etiologies are the most common, with angioectasia being the primary lesion identified (▶Fig. 31.1). Risk factors include advancing age, aortic stenosis, chronic renal failure, and left ventricular assist devices (LVADs). 1 Inflammatory lesions of the small bowel are most often related to Crohn’s disease (CD) but nonsteroidal anti-inflammatory drug (NSAID) ulcers should also be considered. The pathognomonic lesion is the diaphragm-like strictures and these can develop as multiple circumferential membranous structures after years of NSAID use. It is likely that most NSAID-induced injury is subclinical. Symptoms can include iron deficiency anemia, frank bleeding from ulcers, hypoalbuminemia, malabsorption, abdominal pain, and obstruction. Typical NSAID lesions on CE show circumferential symmetric ulcerating rings with usually normal intervening mucosa. They are often difficult to see with cross-sectional imaging
31.2.1 Diagnostic Approaches
In terms of diagnostic approach in patients with suspected small bowel bleeding, it is first important to document objective evidence of GI bleeding whenever possible, especially in those patients presenting with iron deficiency anemia. Hematologic or malabsorptive causes should always be considered in the differential diagnosis before embarking on an extensive evaluation of the small bowel. In addition, upper and lower gastrointestinal etiologies should be sufficiently ruled out, because studies have shown missed lesions in these areas. Previous studies have revealed potential bleeding sources within reach of upper endoscopy or colonoscopy in 21 to 25% of patients undergoing a small bowel evaluation. 2 , 3 “Second look endoscopy” is recommended whenever there is any doubt, before proceeding with a small bowel evaluation. In the case of repeat upper endoscopy, push enteroscopy should be strongly considered to evaluate the duodenal sweep and proximal jejunum. 4 Once this is completed and no source is identified, it is reasonable to evaluate the small bowel.
The two main endoscopic modalities for evaluating the small bowel include capsule endoscopy and deep enteroscopy. Capsule endoscopy is considered to be the next procedure after second look endoscopy. The diagnostic yield is highest in patients with ongoing overt bleeding as compared to occult bleeding. 5 , 6 , 7 Capsule endoscopy is an ideal screening tool prior to deep enteroscopy. Studies have shown that this approach can suggest the initial route for deep enteroscopy and also increase both the diagnostic and therapeutic yield. 8 , 9 , 10 The diagnostic yield is improved if performed within 72 hours of presentations and significantly decreases after 2 weeks of the initial bleeding. 11 , 12 , 13
Deep enteroscopy consists of SBE and DBE and spiral enteroscopy. Deep enteroscopy methods appear to be similar in terms of diagnostic yield, safety and learning curve. While there are conflicting reports, it is believed that SBE and DBE are equally effective and performance depends on the local expertise. 14 , 15 Performance of deep enteroscopy within 72 hours of presentation may improve the diagnostic yield, and should be done in a timely manner. 7 , 16 , 17
A meta-analysis comparing CE to DBE showed that there was no overall difference in diagnostic yield, however, the diagnostic yield of deep enteroscopy significantly increased after a positive capsule study. 18 Deep enteroscopy can be considered initially for those patients who present with active small bowel bleeding. Total deep enteroscopy using both the oral and anal route should be performed if there is a strong suspicion for a lesion despite previously negative studies.
It is important to remember that capsule endoscopy and deep enteroscopy are complimentary in the evaluation of suspected small bowel bleeding. Both have missed lesions found by the alternative technique. In addition, in those patients with negative studies, multidetector computed tomography (CT) scan imaging, that is, CT enterography or CT angiography, may be complimentary as well. While the overall diagnostic yield is less than capsule endoscopy, it can identify bleeding lesions when other tests are negative. Multidetector CT scanning should also be considered in those patients where capsule retention is a concern. 19 , 20 An algorithm for suspected small bowel bleeding is shown in ▶Fig. 31.2. There have been some reports on pharmacologic provocation challenges in refractory cases of small bowel bleeding. One study looking at provocative angiography when conventional angiography was normal showed good results and low complications rates. 21 Various agents have been used including anticoagulants, thrombolytics, hemodilution agents, and vasodilators. Provocative angiography can be considered in refractory cases when all other diagnostic modalities have been unsuccessful. It has also been noted that a negative capsule has a high negative predictive value.
A negative capsule study is associated with significantly lower rebleeding rates compared with those who have a positive CE. 22 , 23 Therefore, unless there is a high index of suspicion, further small bowel evaluation may be deferred in most patients after a negative CE, unless they present with recurrent bleeding.
31.2.2 Therapeutic Approaches
Although capsule endoscopy has been revolutionary in the evaluation and management of small bowel diseases, it is important to remember that it lacks therapeutic capabilities. The advantage to deep enteroscopy is that it allows for therapeutic intervention in those patients with small bowel bleeding. Deep enteroscopy lends itself in particular to the treatment of vascular lesions that are causing small bowel bleeding. Angioectasias are the most common lesions identified, particularly in the elderly. Rebleeding can occur in approximately one-third of patients, particularly if there are multiple lesions. 24 , 25
In those patients where no definitive source for the bleeding or iron deficiency anemia is identified, and conservative observation is warranted, the patient should be managed with oral or intravenous iron with blood transfusions as needed. If bleeding persists, then further diagnostic evaluations should be considered. A small bowel evaluation should be repeated as appropriate. If work-up is again nondiagnostic, and patient is on anticoagulation and/or antiplatelet therapy, then consideration to stopping one or both medications should be discussed with the appropriate physicians. With regard to angioectasias in particular, the recurrent bleeding rate is approximately 33%, but may be as high as 45%. 24 , 26 Recurrent bleeding from angioectasia correlates with the number of lesions, advanced age, presence of comorbid conditions, and anticoagulant therapy. It is important that if endoscopic therapy is going to be successful, it is likely to require multiple sessions. In some patients, medical therapy will prove to be the mainstay of therapy.
In terms of medical management of angioectasias, a recent systematic review and meta-analysis again showed that up to 45% of patients may have repeat bleeding despite therapy. 26 Somatostatin analogues, such as octreotide, also showed benefit, while hormonal therapy did not. 26 , 27 One study suggested that thalidomide also decreases the need for blood transfusions and led to an increase in hemoglobin. 28
In most cases, surgery is not needed in the management of patients with small bowel bleeding due to the advent of capsule endoscopy and deep enteroscopy. However, in those patients with persistent and severe suspected small bowel bleeding and negative small bowel imaging, surgery combined with intraoperative enteroscopy should be considered.
31.3 Small Bowel Crohn’s Disease
Crohn’s disease is an inflammatory condition that can affect the entire GI tract. Most commonly, it involves the terminal ileum and colon segmentally, but can involve only the small bowel or only the colon. As with ulcerative colitis, there is no gold standard for diagnosis and a final diagnosis is based on a comprehensive evaluation which includes history, physical examination, laboratory tests, endoscopy, pathology, and radiology. Endoscopy can play a pivotal role in the management of small bowel CD.
Isolated small bowel CD can be even more challenging to diagnose. About one-third of patients may have disease confined to the small bowel. 29 One prospective study suggested that small bowel involvement occurs more frequently than was previously recognized. 30 There are also data to suggest that localization of inflammation changes only minimally over time. 31 We also know that there is a poor correlation between symptom scores and the degree of inflammation in the bowel. 32 Because diagnosis of small bowel CD can be challenging, the average lag time between onset of symptoms and reaching a diagnosis may be as long as 35 months. 33 A comprehensive evaluation of the entire small bowel may be indicated to make a definitive diagnosis, determine disease extent and severity, and/or to evaluate for mucosal healing. The differential diagnosis of inflammation involving the small bowel can be found in ▶Table 31.3.
Immune deficiency related
Cryptogenic multifocal ulcerous stenosing enteritis
In the majority of patients with small bowel CD, the disease is localized to the terminal ileum or duodenum, and thus can be diagnosed with ileocolonoscopy and/or upper endoscopy. In the case of duodenal CD, upper endoscopy and/or push enteroscopy can be used to obtain biopsies and assess for mucosal healing. Colonoscopy with ileoscopy can assess the terminal ileum and perform similar functions. Such a traditional endoscopic evaluation allows for direct mucosal examination and biopsy and can provide valuable information on extent and severity. It can also help differentiate CD from ulcerative colitis in the appropriate setting. 34 However, it is important to realize that for isolated small bowel CD or in those with sparing of the terminal ileum, the diagnosis may be more elusive.
The indications for a more in-depth evaluation of the small bowel in suspected CD would include the clinical or biological suspicion in the absence of lesions at ileocolonoscopy. This is particularly true in patients with negative endoscopic studies, but the presence of “alarm symptoms” such as anemia, weight loss, abdominal pain, diarrhea, and/or extraintestinal manifestations. There is evidence to suggest that a subgroup of patients will have endoscopic skipping of the terminal ileum. 35 In these situations, one must consider complimentary small bowel imaging tests such as capsule endoscopy, deep enteroscopy, and cross-sectional imaging with computed tomography enterography (CTE) or magnetic resonance enterography (MRE). The ideal test to image the small bowel in suspected CD is not known. Studies have suggested a reasonable sensitivity and specificity for CE in the evaluation of patients with suspected CD. There is evidence that CE may be more sensitive in detecting inflammatory lesions in the proximal small bowel compared with CTE or MRE. 36 , 37 An example of a patient with positive findings on capsule endoscopy but negative MRE and ileoscopy is shown in ▶Fig. 31.3. In addition, the negative predictive value of CE for CD is quite high. However, the specificity of CTE for CD compared to CE appears to be better. 38 A suggested algorithm for the evaluation of a patient with suspected CD is shown in ▶Fig. 31.4. Finally, CE and deep enteroscopy may play a role in established CD, especially in determining the extent and severity of disease, as well as providing information on mucosal healing. 39 In particular, lesions identified on capsule endoscopy can be further evaluated and biopsied if necessary. Several studies have demonstrated a significant role for deep enteroscopy in suspected CD. 40 , 41 , 42 These studies suggest that deep enteroscopy can play a role in diagnosis and management and may be complementary to capsule endoscopy and radiology. However, a complete examination may be hampered by previous surgery and active inflammation, as well as an increased risk of perforation.
In planning whether to use CE, one must consider the risk of capsule retention. Those patients with suspected small bowel bleeding or suspected CD have a low risk of retention under 2%. However, those patients with known CD have a risk of retention ranging from 5 to 13%. Risk factors for retention includes not only CD, but also use of NSAID and a history of abdominal radiation. Therefore, patients with abdominal pain, distension, suspected obstruction, or known CD should undergo cross-sectional imaging and/or a patency capsule examination of the small bowel prior to swallowing the actual capsule. If the evaluation suggests a significant risk of retention, then capsule endoscopy should not be performed.
Up to 30% of patients with CD are prone to developing strictures in the GI tract which can lead to obstruction and abscess formation. While surgery is often indicated, endoscopy can play a role in dilating symptomatic strictures. Proximal duodenal strictures can be dilated with push enteroscopy. For CD involving the jejunum or proximal ileum, deep enteroscopy can be used for dilation of Crohn’s related strictures with very good therapeutic success. 43 , 44 An ideal stricture for endoscopic balloon dilation is one that is 4 cm or less in length and is not complex or angulated. A recent study showed that deep enteroscopy is preferred to MRE for identifying significant strictures. 45 Deep enteroscopy can also be used for the retrieval of retained capsule endoscopes. 46
31.4 Dilation of Small Bowel Stricture
Deep enteroscopy allows for a more detailed examination of the small bowel stricture, to exclude malignancy and assess for possible dilation. Typically, dilation is performed using a through-the-scope balloon system. Dilations should be performed carefully in a gradual stepwise fashion 2 mm at a time up to 13 mm on the initial dilation to avoid perforation. Severe strictures should be dilated initially to 10 mm. The balloon dilator should slowly and carefully be inflated under direct vision for up to 30 to 60 seconds. Up to 50% may require repeat dilation at which time dilations can be increased up to 15 mm cautiously. Balloon dilation should be avoided when active ulcers and/or inflammation is seen at the stricture site. The injection of strictures with corticosteroids is controversial and not routinely recommended. Fully covered self-expandable metal stents in the treatment of Crohn’s strictures is not currently the standard of care.
31.5 Small Bowel Tumors
Although there is approximately 20 ft of small intestine, only about 2% of GI tumors arise in the small intestine. ▶Table 31.4 outlines the most common GI tumors and their relative frequency. Adenocarcinoma, malignant carcinoids, lymphomas, and sarcomas are the most common. 47 In the western world, tumors most commonly occur in the duodenum (3 per 1, 000, 000) and less commonly in the jejunum and ileum (~ 1 per 1, 000, 000). 48 Sarcomas are fairly evenly distributed throughout the small bowel. Adenocarcinomas arising from CD, carcinoids, and lymphomas occur most commonly in the terminal ileum. Multiple tumors are suggestive of familial adenomatous polyposis (adenomas) or hamartomas (Peutz–Jeghers, Cronkhite–Canada, etc.). Tumor incidence is slightly higher in males presenting on average in people in their 60s but starting to rise in patients even in their 30s. There has been a slight rise of incidence of adenocarcinoma and malignant carcinoids, both of these have a higher incidence in African American patients more than in Caucasian patients. 48 , 49 , 50 Patients with carcinoid tumors may additionally present with diarrhea and flushing and have elevations in chromogranin A or 24-hour urinary 5-hydroxyindoleacetic acid (5-HIAA). Additionally, an octreotide scan can help identify carcinoid tumors.
Diagnosing small bowel tumors can be challenging because clinicians can fail to consider small intestinal tumors as a source of patient’s symptoms, particularly when the symptoms are rather vague. Another reason for failure to diagnose these lesions early is that diagnostic modalities, like X-rays were suboptimal in identifying tumors deep in the small bowel. In the past, endoscopy had a limited role in the diagnosis of small bowel tumors, primarily upper endoscopy (with a side-viewing endoscope) could be used to identify ampullary lesions in patients with familial adenomatous polyposis (FAP) syndromes or colonoscopy with terminal ileal intubation might identify an ileal carcinoid, adenocarcinoma, or lymphoma.
The advent of VCE and deep enteroscopy techniques has helped to expand the role of endoscopy in identifying small bowel tumors and the number of tumors and polyps that have been identified has increased. 51 , 52 A recent meta-analysis compared overtube-assisted enteroscopy and VCE and found that these had a high diagnostic concordance rate (80–100%—overall 93%) for small bowel polyps and tumors. 53 This is attributed to the fact that VCE is often used as the initial test to guide the direction of DBE to localize the lesion. It was also noted that overtube endoscopy was more accurate than spiral endoscopy to identify lesions. This was postulated to be the case because rotational advancement enteroscopy is not as effective as balloon-assisted enteroscopy for deep intubation.
It should be noted that both VCE and DBE can still miss lesions and there are diagnostic challenges. For instance, total small bowel enteroscopy cannot always be achieved because of technical issues or because the patient may be unwilling to undergo bidirectional endoscopy. In this meta-analysis total enteroscopy was successful in 85.7% of patients. Other endoscopic challenges include the fact that in polyposis syndromes CE can miss polyps in the periampullary area and the duodenal sweep. 54 CE has been shown though to identify more lesions and lesions that may not be reached by deep enteroscopy. CE can also have a high false-positive rate due to transient bulges that may be identified as lesions when they are not. This study demonstrated that 80% of the lesions missed by DBE were of subepithelial origin (e.g., gastrointestinal stromal tumors [GISTs] and leiomyomas). However, other lesions such as lipomas, carcinoids, and metastatic disease can be submucosal as well. For FAP patients, the miss rate of polyps on CE is not known. However, a small prospective study looking at the detection of duodenal polyps with CE versus upper endoscopy suggested that CE was more accurate in detecting polyps in the third and fourth part of the duodenum. 55 The upper endoscopy was more accurate in polyp detection in the first and second portion of the duodenum, particularly at the ampulla of Vater. Another study of Peutz–Jeghers polyposis syndromes suggested that for more distal polyps (> 15 mm), MRE was a complementary procedure in detecting distal polyps because CE of larger lesions may be limited. 56 There are limited data on the miss rate of CE for single mass lesions. While CE is superior to other modalities for detection of vascular and inflammatory lesions, it does have a significant miss rate for detection of solitary lesions. One study showed a reported miss rate of 19% for neoplasms. 57
Several tips can be useful in trying to distinguish submucosal lesions from bulges on VCE. Submucosal lesions unlike bulges due to adjacent loops of bowel may have an altered vascular pattern over the mucosa or the mucosa will be stretched, making it appear thin and translucent. A mass lesion may also have a certain white or grayish cast. “Bridging folds” are thought to be a pathognomonic feature—where the valvulae conniventes stop at the edge of the mass and form on the other side (▶Fig. 31.5). 58 Additionally on VCE, the images must be examined over several frames. There is also the smooth protruding index on capsule endoscopy (SPICE) score that can be used to identify possible small bowel tumors (▶Table 31.5). There are four features: (1) ill-defined boundary with the surrounding mucosa; (2) diameter larger than its height; (3) visible lumen in the frames in which it appears; and (4) image of the lesion lasting more than 10 minutes. An answer of no to features 1 and 2 is scored one point each and answer of yes to features 3 and 4 is also scored one point each with a maximum score of four. With this index, one can discriminate a mucosal bulge from a mass on VCE and a score greater than 2 has shown 83% sensitivity and 89% specificity for identifying tumors. 59 These are endoscopic criteria—for example, bleeding, irregular surfaces—identified by a consensus group on CE in 2006 to be more likely associated with tumor and therefore likely to improve the diagnosis of tumor during CE. 60
VCE and deep enteroscopy procedures are complimentary procedures with both false-positive and false-negative results. It is unlikely at this time that endoscopy will be the sole means of diagnosing small bowel tumors. The literature demonstrates that CT and MRE are still important modalities in identifying lesions missed by CE. 61 ▶Fig. 31.6 demonstrates a proposed algorithm for diagnosing small bowel tumors. 62