Capsule Endoscopy

Number: 0588

Table Of Contents

Policy

Scope of Policy

This Clinical Policy Bulletin addresses capsule endoscopy.

1. Medical Necessity

   Aetna considers considers capsule endoscopy (e.g., Endocapsule) medically necessary for any of the following indications:

   1. For evaluation of locoregional carcinoid tumors of the small bowel in persons with carcinoid syndrome; or
   2. For evaluation of persons with celiac disease with a positive serology who are unable to undergo esophagogastroduodenoscopy (EGD) (e.g., medically unstable, presence of known or suspected perforated viscus) with biopsy; or
   3. For re-evaluation of persons with celiac disease who remain symptomatic despite treatment and there is no suspected or confirmed gastro-intestinal (GI) obstruction, stricture, or fistulae; or
   4. For initial diagnosis in persons with suspected Crohn's disease (abdominal pain or diarrhea, plus 1 or more signs of inflammation [e.g., fever, elevated white blood cell count, elevated erythrocyte sedimentation rate, elevated C reactive protein], or bleeding) without evidence of disease on conventional diagnostic tests, including small-bowel follow-through or abdominal CT scan/CT enterography and upper and lower endoscopy (esophago-gastro-duodenoscopy (EGD) and colonoscopy); or
   5. For re-evaluation of persons with Crohn's disease who remain symptomatic despite treatment and there is no suspected or confirmed gastro-intestinal obstruction, stricture, or fistulae; or
   6. For investigating suspected small intestinal bleeding in persons with objective evidence of recurrent, obscure gastro-intestinal bleeding (e.g., persistent or recurrent iron-deficiency anemia and/or persistent or recurrent positive fecal occult blood test, or visible bleeding) who have had upper and lower gastro-intestinal endoscopies within the past 12 months (EGD and colonoscopy) that have failed to identify a bleeding source; or
   7. For surveillance of small intestinal tumors in persons with Lynch syndrome, Peutz-Jeghers syndrome and other polyposis syndromes affecting the small bowel; or
   8. For screening or surveillance of esophageal varices in cirrhotic persons with significantly compromised liver function (i.e., Child-Pugh score of Class B or greater) or other situations where a standard upper endoscopy with sedation or anesthesia is contraindicated.

2. Experimental, Investigational, or Unproven

   Aetna considers capsule endoscopy (CE) experimental, investigational, or unproven for any of the following (not an all-inclusive list) because the effectiveness of this approach for these indications has not been established:

   1. As a screening test (other than esophageal varices)
   2. As an initial test in diagnosing GI bleeding
   3. CE of the intestine for evaluating abdominal pain unless one or more of the criteria from Section I are met
   4. In persons with known or suspected GI obstruction, strictures, or fistulas based on the clinical picture or pre-procedure testing and profile
   5. In persons with cardiac pacemakers or other implanted electro-medical devices
   6. In persons with dysphagia or other swallowing disorders
   7. In colorectal cancer screening
   8. In confirming pathology identified by other diagnostic means
   9. In detecting gastric varices
   10. In detecting hookworms
   11. In detecting colorectal polyps
   12. In diagnosing and evaluating mucosal inflammation in ulcerative colitis
   13. In diagnosing intestinal graft versus host disease
   14. In diagnosing of Takayasu’s arteritis
   15. In evaluating diseases involving the esophagus other than esophageal varices
   16. In evaluating intussusception
   17. In evaluating the colon 
   18. In evaluating the stomach
   19. In follow-up of persons with known small bowel disease other than Crohn's disease
   20. In identifying occult primary malignancies (e.g., primary site in individuals with metastatic melanoma)
   21. In investigating duodenal lymphocytosis, small bowel neoplasm, or suspected irritable bowel syndrome
   22. In planning for radiation therapy
   23. In staging portal hypertensive gastropathy
   24. Magnetic-assisted capsule endoscopy (e.g., the NaviCam MCCE System) for upper GI tract screening and detection of esophageal varices and Barrett’s esophagus (BE) 
   25. Repeat use to verify the effectiveness of surgery
   26. Use of the Agile patency capsule for evaluating patency of the gastrointestinal tract before wireless capsule endoscopy, and for all other indications
   27. Use of the Cytosponge capsule and the Esophageal String Test for diagnosis of esophageal pathology (e.g., eosinophilic esophagitis and esophageal varices) 
   28. Use of the Cytosponge capsule for screening of Barrett's esophagus (see also CPB 0728 - Barrett's Esophagus)
   29. Use of artificial intelligence in reviewing colon CE images; and for the management of inflammatory bowel disease
   30. Video CE for the diagnosis of gastro-intestinal graft-versus-host disease.

3. Related Policies

   1. CPB 0728 - Barrett's Esophagus

Background

Capsule endoscopy (CE), also known as wireless capsule endoscopy or video capsule endoscopy, is a noninvasive, diagnostic procedure that is designed to visualize the esophagus, stomach, small bowel or colon. To perform this procedure, a small digestible capsule (approximately the size of a large vitamin) containing a video camera and a light source is swallowed. The camera takes multiple pictures per second and sends electronic signals wirelessly to a data recorder worn around the individual’s waist. The data is then downloaded into a computer program that captures the images to be analyzed by a physician. The capsule is typically excreted naturally by the body within eight to 72 hours after ingestion.

Currently, capsule endoscopy is only utilized for diagnostic purposes; individuals who require a biopsy or therapeutic intervention must then undergo a conventional endoscopic procedure. Examples of FDA approved CE devices include, but may not be limited to:

• PillCam COLON 2 is intended for use only in individuals who had an incomplete optical colonoscopy with adequate preparation and a complete evaluation of the colon was not technically possible.)
• PillCam ESO is intended for the visualization of the esophageal mucosa to detect abnormalities such as gastroesophageal reflux disease (GERD), Barrett’s esophagus or varices.
• PillCam Express is a disposable device that is purportedly for the transendoscopic delivery of the PillCam SB device in individuals who are unable to ingest the PillCam capsule or are known to have slow gastric emptying time.
• PillCam patency system is Intended for use as an accessory to capsule endoscopy to verify small bowel patency or suspected strictures.
• Pill Cam SB 2 and 3 and the Olympus small intestine endoscope capsule are intended for the visualization of the small intestinal mucosa.
• PillCam UGI capsule endoscopy system is intended for visualization of the upper gastrointestinal tract (esophagus, stomach, duodenum) in hemodynamically stable individuals who are at least 18 years of age.

According to guidelines from the American Gastroenterological Association (2001), the standard for diagnosing the source of small intestinal bleeding is push enteroscopy, in which a 4-foot long tube out-fitted with a small video camera is inserted down the esophagus, through the stomach and into the first third of the small intestine. In many cases, a definitive diagnosis cannot be made because the imaging tools cannot reach far enough into the digestive tract to find the problem. Radiologic examination of the small bowel with barium (enteroclysis) may be uncomfortable, time-consuming, and is incapable of detecting completely flat lesions of the small intestine (e.g., arteriovenous malformations).

In August 2001, the U. S. Food and Drug Administration (FDA) cleared for marketing a swallowable capsule containing a small camera that snaps pictures twice a second as it passes through the small intestine. The FDA classified the capsule, called the Given Diagnostic Imaging System (Given Imaging Ltd., Yoqneam, Israel), as a Class II device that is subject only to general regulatory controls. The capsule, marketed as the PillCam SB (previously marketed as M2A^™), has a clear end that allows the camera to view the lining of the small intestine. In addition to the camera, the wireless capsule, about the size of a large vitamin pill, contains a lighting system and a transmitter that will send images from inside the intestine to video monitors, allowing doctors to detect sources of bleeding in the small intestine. FDA cleared the device for use along with, not as a replacement for, other endoscopic and radiological evaluations of the small intestine. The capsule was not studied in the large intestine.

When swallowed, the device travels down the digestive tract at about the same speed as food, propelled by peristalsis, and takes 2 to 3 hours to pass through. Once the device reaches the colon, things slow down, and the disposable device is eliminated like any solid waste within a few days.

The downside to this technology is that the images may not match fiber-optic endoscopes for detail, and concerns have been raised that the camera's view may be obscured by bubbly saliva or green bile. The capsule cannot be stopped or steered to collect close-up details of the small intestine's millions of interior wrinkles where ailments often occur. Nor is it fitted with surgical tools like a conventional endoscope to take biopsies or treat bleeding lesions or remove polyps. If a lesion requiring invasive therapy is found on capsule endoscopy, then the patient will need to undergo surgery with intra-operative endoscopy. In addition, if an abnormality is seen on capsule endoscopy, there is no good way to define its location within the small intestine. Fleischer (2002) has noted that, with capsule endoscopy, “the pylorus is usually seen, and in many patients the ileocecal valve can be demonstrated, but apart from a rough estimate linked to 'time beyond the pylorus' or 'time in front of the ileocecal valve', specific localization is not possible.”

By contrast, push enteroscopy has the advantages of being able to perform biopsies and offer therapy. If capsule endoscopy is performed without a prior push enteroscopy, a push enteroscopy will still need to be performed in most cases since a negative capsule endoscopy may not exclude a lesion, and a lesion observed on capsule examination may be within reach of the enteroscope (Faigel and Fennerty, 2002).

In a study submitted to the FDA, the Given Imaging Diagnostic System detected physical abnormalities in 12 of 20 patients with suspected small intestinal disorders, while push enteroscopy detected physical abnormalities in 7 of 20 patients. All patients included in the trial had previously undergone gastrointestinal endoscopies and radiological procedures to identify the source of their small intestinal disorders, without a conclusive diagnosis. In total, 14 lesions were detected in 13 of the 20 patients participating in the clinical trials using either the Given Imaging Diagnostic System, push enteroscopy or surgical techniques. The Given Imaging Diagnostic System detected 12 of the 14 lesions, while push enteroscopy detected 7 of 14. The investigators also noted that the Given system was able to identify sources of bleeding in 5 cases that were beyond the reach of the traditional enteroscope.

Costamagna et al (2002) compared the performance of capsule endoscopy to upper gastrointestinal barium radiography series with small bowel follow through in 20 patients, including 13 patients with obscure gastrointestinal bleeding, 3 patients with suspected Crohn's disease, 1 patient with suspected sarcoma recurrence, 1 patient with diarrhea, 1 patient with familial adenomatous polyposis, and 1 patient with small intestine polyposis. The rates of a “diagnostic” test were higher for capsule endoscopy (45 %) than for barium examination (27 %), although no test was performed to determine whether this difference was statistically significant. Among the subset of 13 patients with obscure gastrointestinal bleeding, the rates of a diagnostic test were statistically significantly higher for capsule endoscopy (30 %) than for the barium study (5 %); however, the study does not describe how this statistical analysis was performed.

This study has been criticized on several grounds (see, e.g., Faigel and Fennerty, 2002). First, the small heterogenous population included in this study makes it difficult to discern the role of this new technology in clinical practice. Second, the study does not evaluate all relevant competing technologies; specifically, the study does not examine how capsule endoscopy performs in comparison to enteroclysis or push enteroscopy; the latter may have been a more appropriate endoscopic standard for comparison. Third, the study chose to report on “diagnostic yield” because no gold standard study was performed; diagnostic yield cannot differentiate true from false positives or true from false negatives. Two studies reported higher diagnostic yields with capsule endoscopy than push enteroscopy in small groups of patients with chronic gastrointestinal bleeding. Lewis and Swain (2002) reported on the results of a pilot study of capsule endoscopy and push enteroscopy in 21 adult patients with obscure gastrointestinal bleeding whose source was not uncovered with esophago-gastro-duodenoscopy (EGD), colonoscopy or small bowel follow through. Capsule endoscopy was able to identify a bleeding source in 11 patients (55 %), whereas push enteroscopy was able to identify a bleeding source in 6 patients (30 %) (p = 0.0625). In Germany, Ell et al (2002) reported on a comparison of capsule endoscopy to push enteroscopy in 32 patients with chronic gastrointestinal bleeding. Push enteroscopy revealed definite bleeding sites in 9 patients (28 %), including angiodysplasia in 7 patients, small intestine cancer in 1 patient, and lymphoma in 1 patient. Capsule endoscopy detected definite bleeding sites in 21 patients (66 %), including angiodysplasia in 17 patients, malignant stenoses in 2 patients, and inflammatory small-intestine disease in 2 patients. Questionable bleeding sources were seen on push enteroscopy in 3 additional patients (9 %) and using capsule endoscopy in an additional 7 patients (22 %).

Much of the clinical evidence on capsule endoscopy has been presented in the form of abstracts rather than as peer-reviewed published clinical studies. As no study has compared capsule endoscopy to surgical enteroscopy or some other reliable external criterion (i.e., gold standard), the sensitivity, specificity, and predictive values of capsule endoscopy are unknown. In addition, no study has reported on the effect of capsule endoscopy on resolution of bleeding or other relevant clinical outcomes.

In the acute setting, capsule endoscopy is not a substitute for tagged red cell scintigraphy or angiography, because capsule endoscopy takes 8 hours to complete with the results generally not available until the following day.

The BlueCross BlueShield Technology Evaluation Center (2003) evaluated the evidence supporting the use of capsule endoscopy for diseases of the small intestine other than obscure gastrointestinal bleeding. The assessment identified no randomized controlled clinical studies of capsule endoscopy for these indications. The assessment identified 3 published studies (Fireman et al, 2003; Herrerias et al, 2003; Eliakim et al, 2003), involving a total of 58 patients, that prospectively examined the use of capsule endoscopy for initial diagnosis of suspected Crohn's disease when all conventional diagnostic tests, including small-bowel follow-through, have failed to reveal bowel lesions suggestive of Crohn's disease. An additional 41 patients were included in 2 abstract reports and case reports (Sant'anna et al, 2003; Bloom et al, 2003; Costamanga et al, 2003; Chong et al, 2003; Liangpunsakul et al, 2003). The assessment concluded that “[t]hese studies provide consistent evidence that wireless capsule endoscopy may demonstrate small-bowel lesions suggestive of Crohn's disease in a significant proportion of patients ranging from 43 to 71 % when all other conventional tests have been negative. Furthermore, patients in these studies diagnosed with Crohn's disease by wireless capsule endoscopy were reported to improve after treatment for Crohn's disease, which represents an improvement in health outcomes.”

The assessment did not find sufficient evidence to support the use of capsule endoscopy for other indications, including initial diagnosis of irritable bowel syndrome, celiac sprue, small bowel neoplasm, or intestinal polyposis syndrome, or follow-up of persons with known small bowel diseases. The assessment identified 1 published study, involving 20 patients, that examined the diagnostic yield of capsule endoscopy in persons with suspected irritable bowel syndrome, but none of the subjects had significant findings on capsule endoscopy (Bardan, 2003). The assessment found that the evidence for all remaining indications was limited to abstracts and case reports.

An assessment by the National Institute for Clinical Excellence (2004) found adequate evidence to support the use of capsule endoscopy, but that “[c]linicians should consider the use of other investigations prior to wireless capsule endoscopy …” The assessment noted that the main indication for this procedure is obscure gastrointestinal bleeding, which is defined as bleeding of unknown origin that persists or recurs after a negative initial endoscopy. The assessment noted that capsule endoscopy has also been used in the diagnosis and evaluation of Crohn's disease. The assessment noted that some studies have reported a higher diagnostic yield (proportion of patients identified with an apparent abnormality) than the comparator test. The assessment noted, however, in most cases, patients had undergone extensive prior investigations, which would be likely to decrease the apparent diagnostic yield for the comparator procedures. The assessment stated that “[i]t was not possible to determine the relative diagnostic performance (ability to detect correctly both the presence and absence of disease) of wireless capsule endoscopy compared with alternative conventional diagnostic tests” in the assessment of obscure gastrointestinal bleeding. Similarly, with respect to diagnosis of Crohn's disease, the assessment found that the available evidence “is not of sufficient quantity and quality to determine the relative diagnostic performance of wireless capsule endoscopy compared with alternative conventional diagnostic tests in diagnosing unselected patients with suspected Crohn's disease.”

The American Gastroenterological Association position statement on obscure gastro-intestinal bleeding (OGIB) (Raju et al, 2007) stated that patients with occult gastro-intestinal (GI) blood loss and iron deficiency anemia and negative workup on EGD and colonoscopy need comprehensive evaluation, including capsule endoscopy to identify an intestinal bleeding lesion.

An assessment by the Belgian Health Care Knowledge Center (Poelmans et al, 2006) recommended capsule endoscopy in patients with obscure GI bleeding “when a previous ileocolonoscopy and esophagogastroduodenoscopy were negative.” The assessment found that, “[a]t present, the available evidence is not of sufficient quantity and quality to determine the relative diagnostic performance of CE compared with alternative conventional diagnostic tests in diagnosing patients with CD [Crohn's disease], intestinal polyposis and celiac disease. No conclusions can be made as to whether CE is an effective alternative to other tests. Further research is warranted to determine the place of CE in the management algorithm of OGIB [obscure GI bleeding] and on other potential indications for CE such as CD, intestinal polyposis and celiac disease.”

The main limitations of capsule endoscopy in the assessment of small bowel Crohn’s disease are the lack of uniform criteria for diagnosing Crohn’s disease, inability to allow for tissue acquisition or therapeutic intervention, and the risk for capsule retention. Capsule retention in Crohn’s disease patients resulting from underlying small bowel strictures is a major concern. Retained capsules may require surgery in patients that may otherwise have not required surgery. A preingestion radiologic study (computed tomography [CT] or small bowel follow through) is recommended because asymptomatic strictures occur in a substantial proportion of patients with Crohn’s disease. Patients with obstructive symptoms or with endoscopic and radiographic evidence of small bowel narrowing in the setting of Crohn’s disease should not undergo capsule endoscopy.

The American College of Radiology appropriateness criteria for Crohn’s disease (2008) state that wireless capsule endoscopy is likely to play an increasing role in the early diagnosis of Crohn’s disease. However, because of a 5 percent incidence of capsule retention proximal to unsuspected strictures, imaging studies, such as small-bowel follow through, are likely to remain an important screening tool prior to capsule endoscopy examinations. The guidelines state that CT is the initial imaging technique of choice in suspected Crohn’s disease complications, and shows considerable promise in initial diagnosis and estimation of disease severity. Other guidelines similarly recommend reserving capsule endoscopy to patients with suspected Crohn’s disease and negative workup, where strictures have been excluded (see, e.g., World Gastroenterology Organization Global Guidelines, 2009; British Society of Gastroenterology, 2008).  There is a lack of prospective studies demonstrating that use of capsule endoscopy for followup of persons with established Crohn’s disease alters management such that clinical outcomes are improved. Evidence has focused on diagnostic yield. A retrospective study by Long, et al. (2011) reported on changes in management after capsule endoscopy; however, the study does not report on whether clinical outcomes are improved. The European Consensus Statement that was cited states that capsule endoscopy “has a potential role” in assessment of mucosal healing after drug therapy.” It should be noted that this recommendation is based upon low quality evidence, evidence level 4: “Case–control study, poor or nonindependent reference standard”; and a grade C recommendation based upon these studies. The guidelines do not endorse a routine role for capsule endoscopy for followup; rather capsule endoscopy “should only be considered if ileocolonoscopy is contraindicated or unsuccessful,” a recommendation that is based upon limited evidence.

Capsule endoscopy has been used in detecting carcinoid tumors of the small intestine. Guidelines from the National Comprehensive Cancer Network (NCCN, 2008) recommend the use of an Octreoscan for persons who present with carcinoid syndrome to determine tumor location and extent. Appendiceal tumors require abdominopelvic CT. Bronchoscopy, upper gastrointestinal barium swallow with small bowel follow through as indicated, colonoscopy and gastroscopy as indicated to identify the primary site. An MRI of the lung, mediastinum and head, and a CT scan of the chest, abdomen and pelvis may also be helpful, depending on the possible site. For work-up of carcinoid tumors of the small bowel, the NCCN guidelines recommend an initial evaluation with an Octreoscan and abdominopelvic CT scan.  For persons with locoregional disease, additional workup is recommended with a GI series with small bowel follow-through as indicated. Enteroclysis or capsule endoscopy are considered optional tests for work-up of locoregional disease of the small bowel.

Capsule endoscopy may also be useful for identifying celiac disease of the small intestine in persons with positive serologies where previous intestinal biopsies have been negative. Rondonotti et al (2007) found capsule endoscopy comparable to EGD for the diagnosis of celiac disease when there are overt villous changes. Consecutive patients with signs and symptoms suggestive of celiac disease and positive anti-gliadin and/or anti-endomysial and/or anti-tissue transglutaminase antibodies underwent upper gastrointestinal endoscopy and capsule endoscopy. Duodenal biopsies were classified according to modified Marsh's criteria. Capsule findings were evaluated for the presence of lesions compatible with celiac disease (scalloping of duodenal folds, fissures, flat mucosa, and mosaic appearance). Duodenal histology was normal in 11 and compatible with celiac disease in 32 of 43 patients studied. Using duodenal histology as the gold standard, the performance characteristics of capsule endoscopy for the diagnosis of celiac disease were: sensitivity 87.5 % (95 % confidence interval [CI]: 76.1 to 98.9 %), specificity 90.9 % (95 % CI: 81.0 to 100 %), positive predictive value 96.5 % (95 % CI: 90.1 to 100 %), negative predictive value 71.4 % (95 % CI:  55.8 to 87 %), positive and negative likelihood ratios 9.6 and 0.14, respectively. Eighteen patients had mucosal changes extending beyond the duodenum, involving the entire small bowel in three. These patients tended to have more severe symptoms, but the difference was not statistically significant. Interobserver agreement for the diagnosis of celiac disease by capsule endoscopy ranged between 79.2 and 94.4 %; kappa values ranged between 0.56 and 0.87. The authors concluded that capsule endoscopy shows good sensitivity and excellent specificity for the detection of villous atrophy in patients with suspected celiac disease.

Capsule endoscopy can be used for surveillance in persons with Peutz-Jeghers syndrome and other intestinal polyposis syndromes. An assessment of capsule endoscopy for the surveillance of persons with Peutz-Jeghers syndrome for the Australian Medical Service Advisory Committee (MSAC, 2007) recommended capsule endoscopy, performed no more than once in any two year period, for small bowel surveillance in patients diagnosed with Peutz-Jeghers syndrome. The assessment concluded that capsule endoscopy is a safe, well-tolerated procedure compared with small bowel surveillance by barium follow-through. The assessment stated that the small body of literature published on the clinical effectiveness of capsule endoscopy in small bowel surveillance of Peutz-Jeghers syndrome "limits the scope of analysis that can be performed to assess this technology." The MSAC found, however, that capsule endoscopy has changed management of persons with Peutz-Jeghers syndrome in situations where x-ray examinations have produced false negative results. The MSAC also found that capsule endoscopy for small bowel surveillance of Peutz-Jeghers syndrome is likely to be as effective and as cost-effective as small intestine x-ray.

Capsule endoscopy is also being investigated for detecting esophageal pathology. Given Imaging Ltd. (Yoqneam, Israel) received marketing clearance from the FDA in November 2004 for its Given Diagnostic System with PillCam ESO video capsule for imaging the esophagus. The PillCam ESO is being marketed for the diagnosis and evaluation of diseases of the esophagus such as gastroesophageal reflux disease (GERD), erosive esophagitis and Barrett's esophagus (BE), a pre-cancerous condition. The FDA classified the PillCam ESO video capsule as a Class II device that is subject only to general regulatory controls.

The PillCam ESO is the same size as the PillCam SB (11 x 26 mm); however, miniaturization of electronics has enabled the PillCam ESO capsule to include 2 video cameras, 1 at each end of the capsule. Each imager captures 2 images per second, totaling 4 images per second. The esophageal transit time of the capsule is brief (less than 5 seconds) when patients ingest the capsule with water in the upright position. The transit time may be lengthened by having the patient ingest the capsule lying horizontally, which may allow visualization of the squamocolumnar junction.

In a feasibility study, Eliakim et al (2004) compared the PillCam ESO to conventional upper endoscopy as the gold standard for detection of esophageal pathologies in patients with suspected disorders of the esophagus (n = 17). Esophageal pathology was found in 12 of the patients by conventional upper endoscopy and with the PillCam ESO. An additional pathology that was found with the PillCam ESO was considered a false-positive. The authors concluded that this pilot study provides evidence that the esophageal capsule is an accurate, convenient, safe and well-tolerated method to screen patients for significant esophageal disorders; however, the authors stated that further, large-scale studies are necessary to fully assess this diagnostic tool.

A multi-center prospective study by the same investigator group, Eliakim et al (2005) compared the PillCam ESO to conventional upper endoscopy in patients with chronic GERD (n =  93) and Barrett's esophagus (n = 13). The PillCam ESO identified esophageal abnormalities in 61 of the 66 patients with positive esophageal findings (sensitivity, 92 %; specificity, 95 %). The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the PillCam ESO for Barrett esophagus were 97 %, 99 %, 97 %, and 99 %, respectively, and for esophagitis 89 %, 99 %, 97 %, and 94 %, respectively. The authors reported no adverse events related to the PillCamESO during the 2-week follow-up period and concluded that it is a convenient and sensitive method for visualization of esophageal mucosal pathology and may provide an effective method to evaluate patients for esophageal disease. The authors reported that future generations of esophageal capsules with higher frame speed are in clinical trials.

The PillCam ESO is also being investigated for use in evaluating patients with esophageal varices.  Potential advantages of capsule endoscopy over EGD is the ability to avoid sedation in patients with liver cirrhosis, and the ability to perform capsule endoscopy during the office visit.  In a pilot study of 32 patients with cirrhosis, the PillCam ESO was compared with EGD in detecting esophageal varices and portal hypertensive gastropathy. A total of 23 patients had esophageal varices at both EGD and PillCam ESO evaluation (Eisen et al, 2006). The overall concordance between PillCam ESO and EGD was 96.9 % for the diagnosis of esophageal varices and 90.6 % for portal hypertensive gastropathy. 

De Franchis et al (2007) reported on a multi-center clinical trial comparing capsule endoscopy to EGD in detecting esophageal varices. Patients who were undergoing clinically indicated EGD for screening or surveillance of esophageal varices were asked to undergo capsule endoscopy prior to the EGD. EGD was performed within 48 hours of capsule endoscopy. A second investigator read each capsule endoscopy study, blinded to patient history and EGD results. A total of 285 patients underwent capsule endoscopy and EGD, 61 % of whom underwent the procedures for screening, and the remainder for surveillance of known esophageal varices. Sensitivity, specificity, positive predictive value and negative predictive value for capsule endoscopy compared to EGD were 86.7 %, 88.4 %, 92.9 %, and 79.1 %, respectively. Overall agreement was 87.3 % (95 % CI: 83 % to 91 %). There was complete agreement on varices grade in 82 % of cases. In 3 cases, capsule endoscopy did not detect esophageal varices that were considered medium/large on EGD, and EGD did not detect 1 case of medium esophageal varices seen on capsule endoscopy. In differentiating between 2 patient management alternatives (i.e., large varices which requires treatment and small varices or no varices which requires monitoring), sensitivity, specificity, positive predictive value and negative predictive value for capsule endoscopy compared to EGD were 84.6 %, 96.1 %, 89.2 % and 94.3 %, respectively. The overall agreement of treatment decisions based on esophageal varices size was 93 %.

Commenting on the study by de Franchis et al, Zaman (2008) stated that although the overall performance of esophageal capsule endoscopy was good, the study’s primary endpoint was not met -- capsule endoscopy was not equivalent to EGD for detecting varices. Zaman concluded that EGD should therefore continue to be the first-line modality for this application. However, capsule endoscopy should be considered an alternative modality if EGD is contraindicated because of concerns regarding safety or tolerance.

Lapalus and associates (2006) reported on a study comparing EGD and PillCam ESO in evaluating portal hypertension in 21 patients with cirrhosis. The PillCam ESO accurately assessed the presence or absence of esophageal varices in 17 of 20 patients (85 %). The 3 patients in whom there was a discrepancy between the 2 procedures were diagnosed with grade 1 varices on EGD and no varices on esophageal capsule endoscopy. The sensitivity of capsule endoscopy for detecting esophageal varices in comparison with EGD as the gold standard was 81.25 % (13 of 16), with a 100 % positive predictive value, a specificity of 100 % (12 of 12), and a negative predictive value of 57.1 % (4 of 7). In evaluating the stomach, 1 patient presented with gastric varices that were diagnosed with both EGD and capsule endoscopy. Portal hypertension gastropathy was diagnosed with EGD in 16 of 21 patients and with capsule endoscopy in 13 of 20 patients. The 4 patients in whom there was a discrepancy were diagnosed as having gastropathy on EGD but not on capsule endoscopy in 3 cases, or as having gastropathy on capsule endoscopy but not on EGD in 1 case.

Rubin et al (2011) reported on a randomized controlled clinical trial that found that use of esophageal capsule endoscopy to risk stratify emergency room patients with upper gastrointestinal bleeding (UGIB) significantly reduced time to emergent EGD and therapeutic intervention. A total of 24 patients with a history of UGIB within 48 hours of admission to the ER were randomized to esophageal capsule endoscopy versus standard clinical assessment. Esophageal capsule endoscopy was read real-time at the bedside and later reviewed after download. Positive capule endoscopy findings included coffee grounds, blood clot, red blood, or a bleeding lesion. Capsule endoscopy positive patients underwent EGD within 6 hours. Control patients and capsule endoscopy negative patients underwent EGD within 24 hours. Seven of 12 patients were capsule endoscopy positive. All 7 had confirmatory stigmata at EGD. Of the 5 capsule endoscopy negative patients, 4 had no stigmata at EGD and 1 was not endoscoped due to comorbidities. The actual lesion was visualized at capsule endoscopy in 4 of 12 patients during live view and in an additional 2 patients after download (6/12). Time to endoscopy in the capsule endoscopy positive group was significantly shorter than control patients (2.5 versus 8.9 hours, p = 0.029). There was no mortality. Blood transfusion requirement and length of stay were not significantly different in the 2 groups. Bjorkman (2010) stated that this was a very small study that had multiple limitations, including the potential for missing duodenal lesions, given that the capsule reached the duodenum in only 7 patients (58 %). The commentator also noted that many positive findings of capsule endoscopy (active bleeding, coffee grounds, and clots) can be identified with simple gastric aspiration. Bjorkman (2010) stated that the cost-effectiveness of esophageal capsule endoscopy in this setting is unclear.

Guidelines on the prevention and management of gastroesophageal varices and variceal hemorrhage in cirrhosis from the American Association for the Study of Liver Diseases states the frequency of surveillance endoscopies in patients with no or smal varices depends upon their natural history (Garcia-Tsao et al, 2007). Upper endoscopy should be performed once the diagnosis is established. In patients with compensated cirrhosis who have no varices on screening endoscopy, upper endoscopy should be repeated in 2 to 3 year intervals. In those who have small varices, upper endoscopy should be repeated in 1 to 2 years. In the presence of decompensated cirrhosis, upper endoscopy should be repeated at yearly intervals.

An assessment by the American Society for Gastrointestinal Endoscopy (ASGE, 2006) found “[t]o date there are limited published data on capsule endoscopy of the esophagus…These preliminary data show an excellent diagnostic yield in cases of erosive esophagitis, Barrett's esophagus, and esophageal varices.”

The American College of Gastroenterology (ACG) (1999) recommends that patients with long-standing GERD symptoms, particularly those 50 years of age and older, undergo an upper endoscopy for evaluating the mucosa for esophagitis. Approximately 20 % of U.S. adults have symptoms of GERD at least once a week; however, a subgroup of patients with GERD develops severe complications that include erosive esophagitis, stricture formation, Barrett's esophagus, and adenocarcinoma of the esophagus. The ACG state that “patients with chronic GERD symptoms are those most likely to have Barrett's esophagus and should undergo upper endoscopy.” ACG guidelines state that the diagnosis of Barrett's esophagus requires biopsy to determine whether intestinal metaplasia is present. Tissue acquisition can be performed during conventional endoscopy for biopsy.

Thus, capsule endoscopy cannot completely replace conventional endoscopy for the evaluation of diseases involving the esophagus and its clinical value as a screening technique for suspected Barrett's esophagus remains unclear. The clinical effectiveness of the PillCam ESO in screening GERD patients for suspected Barrett's esophagus to direct appropriate patients for endoscopic biopsy needs to be demonstrated by large-scale clinical trials published in the peer-reviewed medical literature.

In a prospective, multi-center, blinded study, Lin and colleagues (2007) evaluated the accuracy of esophageal capsule endoscopy (ECE) for the diagnosis of Barrett's esophagus. Major outcome measures included sensitivity, specificity, as well as positive and negative predictive values of ECE for Barrett's esophagus by using EGD results, with histological confirmation as the criterion standard. A total of 96 subjects were enrolled, of whom 90 (94 %) completed the study, including 66 screening and 24 surveillance patients. Esophageal capsule endoscopy was 67 % sensitive and 84 % specific for identifying Barrett's esophagus, diagnosing 14 of 21 cases of biopsy-confirmed Barrett's esophagus. Positive and negative predictive values were 22 % and 98 %, respectively (calculated for screening patients only). Sensitivity for short- and long-segment Barrett's esophagus was similar. The authors concluded that the findings of this study showed that ECE had only moderate sensitivity and specificity for identifying Barrett's esophagus. They noted that ECE in its present form is unsuitable as a primary screening tool for Barrett's esophagus; however, ECE may be used in patients unwilling to undergo EGD.

Johnson (2007) noted that the findings by Lin et al (2007) are contradictory to the favorable results from the study by Eliakim et al (2005). This discrepancy is surprising because some earlier studies had been carried out with capsules that captured only 4 frames per second rather than the 14 frames per second captured by the capsule that was used in the present study. In the validation study, experts who were aware of both the endoscopy and the capsule findings adjudicated final diagnoses; as the current study did not include this protocol, its results could reflect "real-world" use more accurately. Although the convenience, safety, and patients’ tolerance of CE make it an attractive tool for esophageal imaging, at present, this device probably cannot be relied on for the one-time screening to exclude Barrett's esophagus in patients with chronic GERD.

Johnson (2008) commented that 2 recent studies (citing Sharma et al, 2008; Quershi et al, 2008) have demonstrated that esophageal capsule endoscopy is unreliable for detection of Barrett's esophagus (BE) and has high interobserver variability, particularly with respect to short -segment BE (SSBE). Johnson (2008) concluded that these studies demonstrate that esophageal capsule endoscopy in its current form is not an adequate tool for screening for BE.

Sharma et al (2008) reported on a prospective trial involving 100 patients who had GERD symptoms or were under surveillance for BE. All patients underwent esophageal capsule endoscopy, followed by standard upper endoscopy, which was considered the gold standard. The esophageal capsule endoscopy findings were evaluated by investigators who were blinded to the endoscopic findings. BE was confirmed histologically in 45 of the 94 patients who completed the study. Esophageal capsule endoscopy demonstrated a sensitivity of 78 %, a specificity of 75 %, a positive predictive value of 74 %, and a negative predictive value of 79 %. The corresponding accuracy for diagnosing erosive esophagitis was 50 %, 90 %, 56 %, and 88 %, respectively.

Quershi et al (2008) reported on a prospective trial involving patients with short-segment BE (SSBE). A total of 20 patients with biopsy-proven SSBE underwent capsule endoscopy of the esophagus; images were subsequently reviewed by 2 experts who had no knowledge about the purpose of the study. Eighteen patients completed the study. BE was identified or suspected in 44 % of these patients by 1 observer but in only 16 % by the second observer. The Z-line was identified in all 18 patients by both observers, but there was agreement in only 6 as to whether it was normal or irregular.

American College of Gastroenterology guidelines do not support the use of esophageal capsule endoscopy for Barrett's esophagus (Wang et al, 2008).  

Capsule endoscopy has not been proven to be of value in detecting conditions in the colon. The major technical limitations of capsule colonoscopy are its requirement for highly effective bowel preparation and the limited frame speed of the current version. The colon is not well visualized with capsule endoscopy because stool obscures the visualization of the colonic mucosa. Visualization of the colon is more difficult than the small intestine because of its slower transit time and larger diameter; it is possible for the camera to miss suspicious areas of the colon simply by being pointed in the wrong direction. An American Cancer Society position statement (Levin et al, 2003) has concluded that there is no evidence to support the use of capsule endoscopy for detecting colorectal polyps or cancers.

Rex (2008) commented that capsule colonoscopy is similar to computed tomography colonography in having required polyp-size thresholds for referral for polypectomy. Rex stated that, from a cost-effectiveness standpoint, capsule colonoscopy is dominated by colonoscopy when equal adherence is assumed (citing Hassan et al, 2008). For capsule colonoscopy, improvements in adherence could overcome deficiencies in effectiveness (Rex, 2008). However, there currently is no good evidence of improved compliance with capsule colonoscopy. Thus, although capsule colonoscopy could improve adherence, actual demonstrations of whether and how much improvement could be expected are needed.

The Canadian Agency for Drugs and Technologies in Health (CADTH)'s report on capsule colonoscopy/PillCam Colon (Tran, 2007) stated that there is limited evidence on the use of this technology in imaging the colon. Two small, methodologically flawed pilot studies found that for patients with positive findings (i.e., abnormalities detected), the rates of detection with the PillCam Colon capsule were similar to those obtained with conventional colonoscopy. Larger, multi-center trials that compare CE with colonoscopy are needed. The evidence to support the use of CE in screening for colorectal cancer is also lacking.

In a prospective, multi-center study, Van Gossum and colleagues (2009) compared CE with optical colonoscopy (the standard for comparison) in a cohort of patients with known or suspected colonic disease for the detection of colorectal polyps or cancer. Patients underwent an adapted colon preparation, and colon cleanliness was graded from poor to excellent. These investigators computed the sensitivity and specificity of CE for polyps, advanced adenoma, and cancer. A total of 328 patients (mean age of 58.6 years) were included in the study. The capsule was excreted within 10 hours after ingestion and before the end of the lifetime of the battery in 92.8 % of the patients. The sensitivity and specificity of CE for detecting polyps that were 6 mm in size or bigger were 64 % (95 % CI: 59 to 72) and 84 % (95 % CI: 81 to 87), respectively, and for detecting advanced adenoma, the sensitivity and specificity were 73 % (95 % CI: 61 to 83) and 79 % (95 % CI: 77 to 81), respectively. Of 19 cancers detected by colonoscopy, 14 were detected by CE (sensitivity, 74 %; 95 % CI: 52 to 88). For all lesions, the sensitivity of CE was higher in patients with good or excellent colon cleanliness than in those with fair or poor colon cleanliness. Mild-to-moderate adverse events were reported in 26 patients (7.9 %) and were mostly related to the colon preparation. The authors concluded that the use of CE of the colon allows visualization of the colonic mucosa in most patients, but its sensitivity for detecting colonic lesions is low as compared with the use of optical colonoscopy.

In an editorial that accompanied the afore-mentioned article, Bretthauer (2009) stated that colonoscopy, CT colonography and colon CE (CCE) should be tested in randomized, comparative trials that allow valid and precise quantification of their effect on colorectal cancer (CRC) incidence and mortality.

Sieg et al (2009) stated that CRC screening with colonoscopy was introduced into the National Cancer Prevention Program in Germany in 2002. As compliance for screening is low (around 3 % per year), CCE could be an alternative approach. In this study, feasibility and performance of CCE were evaluated in comparison with colonoscopy in ambulatory patients with special attention to a short colon transit time. Colon CE was prospectively tested in ambulatory patients enrolled for colonoscopy who presented for screening or with positive fecal occult blood test. Study subjects underwent colon preparation and ingested the capsule in the morning. Colonoscopy was performed after excretion of the capsule. Colonoscopy and CCE were performed by independent physicians who were blinded to the results. A total of 38 patients were included. One patient was excluded because the capsule remained in the stomach during the entire period of examination. Another patient had limited time and the procedure had to be stopped when the capsule was still in the transverse colon. Thus, these investigators reported the results of 36 patients (30 men and 6 women; mean age of 56 years, range of 23 to 73 years) who successfully completed CCE and the conventional colonoscopy examination. The capsule was excreted within 6 hours in 84 % of the patients (median transit time 4.5 hours). If oral sodium phosphate was excluded from the preparation, the colon transit time increased to a median of 8.25 hours. In total, 7 of 11 small polyps (less than 6 mm) detected by colonoscopy were identified by CCE. One small polyp detected by CCE was not identified by colonoscopy. In this series, no large polyps were found. One CRC was detected by both methods. The mean rates of colon cleanliness (range from 1 = excellent to 4 = poor) in the cecum (2.1), transverse colon (1.6), and in the descending colon (1.5) were significantly better than in the rectosigmoid colon (2.6), and the overall mean rate during colonoscopy was significantly better than during CCE. No adverse effects occurred. The authors concluded that CCE appears to be a promising new modality for colonic evaluation and may increase compliance with CRC screening. To achieve a short colon transit time, sodium phosphate seems to be a necessary adjunct during preparation. The short transit time is a prerequisite to abandon the delay mode of the capsule. With an undelayed PillCam COLON capsule, a "pan-enteric" examination of the gastrointestinal tract would be possible. They stated that further studies are needed to improve the cleanliness, especially in the rectum and to evaluate the method as a potential screening tool.

In a retrospective case series, Triantafyllou and colleagues (2009) evaluated if PillCam colon capsule (PCC) endoscopy can complete colon examination after failure of conventional colonoscopy to visualize the cecum. By using PCC, these investigators studied 12 patients who had incomplete colonoscopy -- 6 patients had an obstructing tumor of the left side of the colon; and in 6 cases, there were technical difficulties to complete colonoscopy. PillCam colon capsule endoscopy visualized the rectum in 1 case. The capsule did not reach the site where colonoscopy stopped in 6 of the 12 cases, i.e., 3 left sited tumors and 3 with technical difficulties. Moreover, in 1 of the 3 cases in which the capsule passed the site where colonoscopy stopped, poor bowel preparation precluded the accurate examination of the colon. Four patients underwent a third colon examination (3 barium enemas and 1 virtual CT colonoscopy). There were no adverse events related to PCC endoscopy. The authors concluded that in this retrospective case series of patients with incomplete colonoscopy, PCC endoscopy did not always satisfactorily examine the colon.

In a meta-analysis evaluating the accuracy of CCE in detecting colon polyps, Rokkas et al (2010) concluded that CCE is a feasible alternative method for colon investigation, including screening for polyps and CRC, patients with incomplete colonoscopy, those with contraindications for conventional colonoscopy, or those unwilling to undergo colonoscopy because of its perceived inconvenience and discomfort. They stated that however, larger, multi-center, well-designed trials are needed to further establish the role of CCE in the evaluation of the large bowel in health and disease.

Spada et al (2010) noted that the PillCam colon capsule endoscopy (PCCE) represents one of the newest diagnostic, endoscopic technology designed to explore the colon. It is a non-invasive, swallowing colonoscope that is able to explore the colon without requiring sedation, nor radiation. The colon capsule measures 31 mm x 11 mm. It has dual cameras that enable to take pictures from both ends at a frame rate of 4 frames per second. The goal of PCCE is to substitute the conventional colonoscopy in the diagnosis of colonic diseases and to discriminate patients who deserve a conventional colonoscopy. The authors stated that although the preliminary results available in literature are encouraging, further studies are needed to confirm and increase the accuracy parameters and to select a more tolerable and effective regimen of preparation. In a review on alternatives to colonoscopy and their limitations, Chaput and co-workers (2010) stated that the PillCam for the small intestine has been adapted to study of the colon. Results of CCE studies are promising.

Capsule endoscopy is contraindicated in patients with known or suspected gastrointestinal obstruction, strictures, or fistulae. The available literature indicates that an upper gastrointestinal series should be performed prior to capsule endoscopy if the patient is suspected of having intestinal obstruction. In a review on contraindications to capsule endoscopy, Storch and Barkin (2006) stated that the only true remaining contraindications to capsule endoscopy are obstruction/pseudo-obstruction and pregnancy.

The Agile Patency System (Given Imaging, Ltd.) was cleared by the FDA through the 510k process for determining the presence of obstructions or strictures in the gastrointestinal tract through a dissolvable capsule. It supposedly will give physicians confidence that a PillCam video capsule will pass freely in a patient with known or suspected strictures. suspected strictures. Currently, there is insufficient evidence on the performance of this patency system as a technique to evaluate patients with known or suspected strictures prior to using the wireless capsule endoscopy system.

Spada et al (2007) evaluated the safety of the Pillcam in patients with known or suspected radiological stricture, previously tested for small bowel patency using the Given Patency capsule. A total of 27 patients (16 females, mean age of 44.2 years) with known or suspected intestinal stricture were enrolled prospectively: 24 had Crohn's disease, 2 had adhesive syndrome, and 1 had a suspected ischemic stricture. Patients underwent the Patency capsule test. In patients in whom the Patency capsule was excreted intact within 72 hours post-ingestion without occurrence of any adverse event, video capsule endoscopy (VCE) was performed to assess the presence of strictures or other gastrointestinal pathologies. The following parameters were evaluated: transit time of Patency capsules and/or tags from ingestion to excretion, condition of the Patency capsule at excretion, transit time of the Pillcam capsule, the ability of Pillcam capsule to detect intestinal strictures and small bowel pathologies, any adverse events. A total of 25 patients (92.6 %) retrieved the Patency capsule in the stools; 6 patients complained of abdominal pain, 4 of whom excreted a non-intact capsule. Hospitalization was required in 1 (4.3 %) patient with Crohn's disease due to occlusive syndrome. Fifteen patients (65.3 %) excreted an intact Patency capsule after a mean transit time of 25.6 hours without any adverse events. These 15 patients underwent the VCE successfully. The authors concluded that passage of an intact Patency capsule across a small bowel stricture provides direct evidence of functional patency of the gut lumen and allows a safe VCE. Intestinal strictures should not be considered an absolute contraindication for VCE.

Herrerias and colleagues (2008) assessed the ability of the Agile Patency System to help physicians identify which patients with known strictures may safely undergo CE. A total of 106 patients ingested the patency capsule. Fifty-nine (56 %) excreted it intact and subsequently underwent CE. There were no cases of capsule retention. Significant findings on CE were found in 24 (41 %). There were 3 severe adverse events. The authors concluded that these findings suggested that the Agile Patency System is a useful tool for physicians to use before CE in patients with strictures to avoid retention. This group of patients may have a high yield of clinically significant findings at CE. This capsule may determine whether patients who have a contraindication to CE may safely undergo CE and obtain useful diagnostic information.

The Agile Patency System has been reported to cause obstruction requiring urgent intervention. There is currently insufficient evidence from well-designed studies to support the use of the Agile Patency System. In addition, the American Society for Gastrointestinal Endoscopy's Technology Status Evaluation Report on wireless capsule endoscopy (Mishkin et al, 2006) stated that "[t]here is limited information about the new patency capsule. While it is intended to assess the passage of a capsule in patients at risk for intestinal stenosis, there have been reported cases that have required hospital admissions, augmentation of medical therapy, and even surgery. As a result, improvements to the system are being implemented before it can be approved in the United States".

In a randomized, controlled trial, Keller and colleagues (2010) assessed the safety and effectiveness of remote magnetic manipulation of a modified capsule endoscope (magnetic maneuverable capsule [MMC]; Given Imaging Ltd, Yoqneam, Israel) in the esophagus of healthy humans. A total of 10 healthy volunteers swallowed a conventional capsule (ESO2; Given Imaging) and a capsule endoscope with magnetic material, the MMC, which is activated by a thermal switch, in random order (1 week apart). An external magnetic paddle (EMP; Given Imaging) was used to manipulate the MMC within the esophageal lumen. MMC responsiveness was evaluated on a screen showing the MMC film in real time. Main outcome measurements included safety and tolerability of the procedure (questionnaire), responsiveness of the MMC to the EMP, esophageal transit time, and visualization of the Z-line. No adverse events occurred apart from mild retro-sternal pressure (n = 5). The ability to rotate the MMC around its longitudinal axis and to tilt it by defined movements of the EMP was clearly demonstrated in 9 volunteers. Esophageal transit time was highly variable for both capsules (MMC, 111 to 1,514 seconds; ESO2, 47 to 1,474 seconds), but the MMC stayed longer in the esophagus in 8 participants (p < 0.01). Visualization of the Z-line was more efficient with the ESO2 (inspection of 73 % +/- 18 % of the circumference versus 33 % +/- 27 %, p = 0.01). The authors concluded that remote control of the MMC in the esophagus of healthy volunteers is safe and feasible, but higher magnetic forces may be needed.

In an open clinical trial, Keller et al (2011) evaluated the safety and effectiveness of manipulation of a MMC in the human stomach by using a hand-held external magnet. A total of 10 healthy volunteers swallowed the MMC and sherbet powder for gastric distention. An EMP-2 was used to manipulate the MMC within the stomach. Responsiveness of the MMC was evaluated on a screen showing the MMC film in real time. Main outcome measurements included safety and tolerability (questionnaire), gastric residence time of the MMC, its responsiveness to the EMP-2, and visualization of gastric mucosa. There were no adverse events. The MMC was always clearly attracted by the EMP-2 and responded to its movements. It remained in the stomach for 39 +/- 24 minutes. In 7 subjects, both the cardia and the pylorus were inspected and 75 % or more of the gastric mucosa was visualized (greater than or equal to 50 % in all of the remaining subjects). A learning curve was clearly recognizable (identification of MMC localization, intended movements). The authors concluded that remote control of the MMC in the stomach of healthy volunteers using a hand-held magnet is safe and feasible. Responsiveness of the MMC was excellent, and visualization of the gastric mucosa was good, although not yet complete, in the majority of subjects. They stated that the system appeared to be clinically valuable and should be developed further.

Gastineau and colleagues (2012) evaluated the contribution of CE in managing risk of further obstructive complications.  These researchers performed a retrospective analysis of 27 children who received at least 1 capsule endoscopy.  Peutz-Jeghers syndrome was diagnosed based on the presence of an STK11 gene mutation or on the association of a hamartoma with 2 of 3 criteria (family history, mucocutaneous pigmentation, small bowel polyposis).  A total of 37 CEs were performed in 27 patients.  The median age at first endoscopy was 11.4 years (range of 5.4 to 20.9).  Jejunal polyps were found in 72 % and ileal polyps in 55 % of capsules.  The original recommendations were followed 20/30 times.  Three gastroscopies, 4 colonoscopies, 7 double balloon enteroscopies and 1 intra-operative enteroscopy were performed after the capsules.  These procedures revealed jejunal polyps in 9/9 cases (8/9 resected) and ileal polyps in 3/5 (all resected).  One intussusception occurred 8.4 months after the CE and required surgical resection.  The authors concluded that CE is easily feasible in Peutz Jeghers syndrome, but the practice of systematic and repeated procedures needs to be validated prospectively.

Minaya et al (2012) stated that intussusception in adult patients accounted for less than 5 % of all intussusceptions. It occurs when a segment of intestine invaginates into itself.  These investigators reported a case of ileo-colic intussusception in an adult caused by a giant ileal lipoma.  Intussusceptions can be classified as ileo-colic, ileo-cecal, colo-colic and ileo-ileal.  Most are due to neoplasms (60 % malign and 24 to 40 % benign).  In the colon, the possibility of malignancy is higher than in small intestine.  Lipomas are the most common benign mesenchymal intestinal tumors, accounting for less than 5 % of all gastro-intestinal tumors.  They are more frequent in colon than small intestine.  Small lipomas (less than 2 cm) are usually asymptomatic.  Larger lesions may produce symptoms such as abdominal pain, obstruction or intussusception.  Lipomas can be diagnosed with endoscopy, CE, barium enemas, CT and ultrasonography.  The authors concluded that intussusceptions in adults is a rare condition, most of them are caused by a malign neoplasms followed by benign neoplasms; CT and ultrasonography are useful for diagnosis.  Capsule endoscopy was not mentioned as a management tool.

Wiener-Carrillo et al (2014) noted that intussusception in adult patients represented 5 % of all intussusceptions and 1 to 5% of bowel obstructions in adults.  In contrast to pediatric patients, 90 % of the time, in adults, it's caused by well-established pathologic mechanisms, such as carcinoma, polyps, diverticula, Meckel diverticula, stenosis, or benign neoplasms.  Small intestine intussusceptions are more frequent, but colonic intussusceptions are caused 50 % of the time by malignant neoplasms, especially adenocarcinoma.  These investigators presented the case of a 70-year old woman, with no relevant familial history, who presented with a 3-day symptomatology consisting of epigastric, colic, diffuse, abdominal pain of moderate intensity, which progressed till reaching a severe intensity, also referring abdominal distension, nausea, and gastrointestinal-content vomits.  In adult patients, the exact mechanism of intussusception is unknown in 8 to 20 % of the cases, however, secondary intussusception can occur with any lesion of the intestinal wall or any irritant factor in its lumen that alters normal peristaltic activity and that could serve as a trigger to start an intussusception of one bowel segment over another the most common site is the small intestine.  The authors concluded that intussusception represents an unusual problem in adult patients; it requires a high clinical suspicion, mainly as a differential diagnosis in patients with intestinal obstruction, and it clinically presents as a subacute or chronic illness.  They stated that CT represents the most useful diagnostic tool.  Capsule endoscopy was not mentioned as a management tool.

The American College of Radiology’s clinical practice guideline on “Suspected small-bowel obstruction” (Katz et al, 2013) noted that ultrasonography has proven useful in evaluating intussusception, mid-gut volvulus, and other causes of suspected small-bowel obstruction.  Moreover, CE was not mentioned as a management tool.  Furthermore, an UpToDate reviews on “Intussusception in children” (Kitagawa and Miqdady, 2014) and “Epidemiology, clinical features, and diagnosis of mechanical small bowel obstruction in adults” (Bordeianou and Yeh, 2014) do not mention the use of capsule endoscopy as a management tool.

Xue et al (2015) evaluated the diagnostic yield of small-bowel CE (SBCE) in patients with unexplained chronic abdominal pain.  These investigators performed a retrospective review of publications reporting the diagnostic yield of SBCE in patients with unexplained chronic abdominal pain and calculated the overall diagnostic yield (DY).  Two investigators independently searched studies from databases and analyzed the results.  A total of 1,520 patients from 21 studies were included.  Per-patient DY, with 95 % CI, was evaluated by a random-effect model.  Clear categorical analysis also was performed.  The pooled DY of SBCE in patients with unexplained chronic abdominal pain was 20.9 % (95 % CI: 15.9 % to 25.9 %), with high heterogeneity (I(2) = 80.0 %; p < 0.001).  Inflammatory lesions were the most common (78.3 %) positive findings, followed by tumors (9.0 %).  The authors concluded that SBCE provided a non-invasive diagnostic tool for patients with unexplained chronic abdominal pain, but the DY was limited (20.9 %).

On behalf of the European Society of Gastrointestinal Endoscopy (ESGE), Pennazio et al (2015) addressed the roles of small-bowel CE and device-assisted enteroscopy for diagnosis and treatment of small-bowel disorders.  One of the main recommendations was that ESGE strongly recommends against the use of small-bowel CE for suspected celiac disease; but suggests that CE could be used in patients unwilling or unable to undergo conventional endoscopy (strong recommendation, low quality evidence). 

Frennette and colleagues (2009) examined the utility of ECE in the diagnosis and grading of esophageal varices.  Cirrhotic patients who were undergoing EGD for variceal screening or surveillance underwent CE.  Two separate blinded investigators read each CE for the following results: variceal grade, need for treatment with variceal banding or prophylaxis with beta-blocker therapy, degree of portal hypertensive gastropathy, and gastric varices.  A total of 50 patients underwent both capsule and EGD; 48 had both procedures on the same day, and 2 patients had CE within 72 hrs of EGD.  The accuracy of CE to decide on the need for prophylaxis was 74 %, with sensitivity of 63 % and specificity of 82 %.  Inter-rater agreement was moderate (kappa = 0.56).  Agreement between EGD and CE on grade of varices was 0.53 (moderate).  Inter-rater reliability was good (kappa = 0.77).  In diagnosis of portal hypertensive gastropathy, accuracy was 57 %, with sensitivity of 96 % and specificity of 17 %.  Two patients had gastric varices seen on EGD, 1 of which was seen on CE.  There were no complications from CE.  The authors concluded that CE has a limited role in deciding which patients would benefit from EGD with banding or beta-blocker therapy.  They stated that more research is needed to assess accuracy for staging esophageal varices, portal hypertensive gastropathy, and the detection of gastric varices.

In a Cochrane review, Colli et al (2014) stated that current guidelines recommend performance of EGD at the time of diagnosis of hepatic cirrhosis to screen for esophageal varices.  These guidelines require people to undergo an unpleasant invasive procedure repeatedly with its attendant risks, despite the fact that 50 % of the people do not have identifiable esophageal varices 10 years after the initial diagnosis of cirrhosis.  Video CE is a non-invasive test proposed as an alternative method for the diagnosis of esophageal varices.  These investigators determined the diagnostic accuracy of CE for the diagnosis of esophageal varices in children or adults with chronic liver disease or portal vein thrombosis, irrespective of the etiology.  They investigated the accuracy of CE as triage or replacement of EGD.  These investigators searched the Cochrane Hepato-Biliary Group Diagnostic Test Accuracy Studies Register (October 2013), MEDLINE (Ovid SP) (1950 to October 2013), EMBASE (Ovid SP) (1980 to October 2013), ACP Journal Club (Ovid SP) (1991 to October 2013), Database of Abstracts of Reviews of Effects (DARE) (Ovid SP) (3rd quarter), Health Technology Assessment (HTA) (Ovid SP) (3rd quarter), NHS Economic Evaluation Database (NHSEED) (Ovid SP) (3rd quarter), and Science Citation Index Expanded (SCI-EXPANDED) (ISI Web of Knowledge) (1955 to October 2013).  They applied no language or document type restrictions.  Studies that evaluated the diagnostic accuracy of CE for the diagnosis of esophageal varices using EGD as the reference standard in children or adults of any age, with chronic liver disease or portal vein thrombosis were selected for analysis.  These researchers followed the available guidelines provided in the Cochrane Handbook for Diagnostic Test of Accuracy Reviews.  They calculated the pooled estimates of sensitivity and specificity using the bi-variate model due to the absence of a negative correlation in the receiver operating characteristic (ROC) space and of a threshold effect.  The search identified 16 eligible studies, in which only adults with cirrhosis were included.  In 1 study, people with portal thrombosis were also included.  The authors classified most of the studies at high risk of bias for the 'Participants selection' and the 'Flow and timing' domains.  One study assessed the accuracy of CE for the diagnosis of large (high-risk) esophageal varices.  In the remaining 15 studies that assessed the accuracy of CE for the diagnosis of esophageal varices of any size in people with cirrhosis, 936 participants were included; the pooled estimate of sensitivity was 84.8 % (95 % CI: 77.3 % to 90.2 %) and of specificity 84.3 % (95 % CI: 73.1 % to 91.4 %).  Eight of these studies included people with suspected varices or people with already diagnosed or even treated varices, or both, introducing a selection bias.  Seven studies including only people with suspected but unknown varices were at low risk of bias; the pooled estimate of sensitivity was 79.7 % (95 % CI: 73.1 % to 85.0 %) and of specificity 86.1 % (95 % CI: 64.5 % to 95.5 %).  Six studies assessed the diagnostic accuracy of CE for the diagnosis of large esophageal varices, associated with a higher risk of bleeding; the pooled sensitivity was 73.7 % (95 % CI: 52.4 % to 87.7 %) and of specificity 90.5 % (95 % CI: 84.1 % to 94.4 %).  Two studies also evaluated the presence of red marks, which are another marker of high risk of bleeding; the estimates of sensitivity and specificity varied widely.  Two studies obtained similar results with the use of a modified device as index test (string capsule).  Due to the absence of data, these researchers could not perform all planned subgroup analyses.  Inter-observer agreement in the interpretation of CE results and any adverse event attributable to CE were poorly assessed and reported.  Only 4 studies evaluated the inter-observer agreement in the interpretation of CE results: the concordance was moderate.  The participants' preferences for CE or EGD were reported differently but seemed in favor of CE in 9 of 10 studies.  In 10 studies, participants reported some minor discomfort on swallowing the capsule.  Only 1 study identified other significant adverse events, including impaction of the capsule due to previously unidentified esophageal strictures in 2 participants.  No adverse events were reported as a consequence of the reference standard.  The authors concluded that they cannot support the use of CE as a triage test in adults with cirrhosis, administered before EGD, despite the low incidence of adverse events and participant reports of being better-tolerated.  Thus, they stated that that EGD cannot be replaced by CE for the detection of esophageal varices in adults with cirrhosis.  Moreover, they found no data assessing CE in children and in people with portal thrombosis.

In an evidence-based analysis, Health Quality Ontario (2015) evaluated the diagnostic accuracy and safety of CCE for the detection of colorectal polyps among adult patients with signs or symptoms of CRC or with increased risk of CRC, and compared CCE with alternative procedures. They performed a literature search using Ovid Medline, Ovid Medline In-Process and Other Non-Indexed Citations, Ovid Embase, the Wiley Cochrane Library, and the Centre for Reviews and Dissemination database, for studies published between 2006 and 2014.  Data on diagnostic accuracy and safety were abstracted from included studies.  Quality of evidence was assessed using Grading of Recommendations Assessment, Development, and Evaluation (GRADE).  The search yielded 2,189 citations; 5 studies, all of which evaluated PillCam COLON 2 (PCC2), met the inclusion criteria.  The per-patient sensitivity and specificity for detecting colorectal polyps were meta-analyzed.  Colon capsule endoscopy, using PCC2, had a pooled sensitivity and specificity of 87 % (95 % CI: 77 % to 93 %) and 76 % (95 % CI: 60 % to 87 %), respectively, for the detection of a colorectal polyp at least 6-mm in size (GRADE: very low).  PCC2 had a pooled sensitivity and specificity of 89 % (95 % CI: 77 % to 95 %) and 91 % (95 % CI: 86 % to 95 %), respectively, for the detection of a colorectal polyp at least 10-mm in size (GRADE: low).  One study directly compared PCC2 with CT colonography and found no statistically significant difference in accuracy (GRADE: low).  Few adverse events were reported with PCC2; 3.9 % of patients (95 % CI: 2.4 % to 6.5 %) experienced adverse effects related to bowel preparation.  Capsule retention was the most serious adverse event and occurred in 0.8 % of patients (95 % CI: 0.2 % to 2.4 %) (GRADE: very low).  The authors concluded that in adult patients with signs, symptoms, or increased risk of CRC, there is low-quality evidence that CCE using the PCC2 device has good sensitivity and specificity for detecting colorectal polyps.  They stated that low-quality evidence does not show a difference in accuracy between CCE and CT colonography; and there is very low-quality evidence that PCC2 has a good safety profile with few adverse events; capsule retention is the most serious complication.

Eliakim and co-workers (2009) stated that a 2nd-generation capsule endoscopy system, using the PillCam Colon 2, was developed to increase sensitivity for colorectal polyp detection compared with the 1st-generation system.  They reported the  performance of this new system.  In a 5-center, feasibility study, 2nd-generation CE was prospectively compared with conventional colonoscopy as gold standard for the detection of colorectal polyps and other colonic disease, in a cohort of patients scheduled for colonoscopy and having known or suspected colonic disease.  Colonoscopy was independently performed within 10 hours after capsule ingestion.  Capsule-positive but colonoscopy-negative cases were counted as false-positive.  A total of 104 patients (mean age of 49.8 years) were enrolled; data from 98 were analyzed.  Patient rate for polyps of any size was 44 %, 53 % of these patients having adenomas.  No adverse events (AEs) related to either procedure were reported.  The capsule sensitivity for the detection of patients with polyps greater than or equal to 6 mm was 89 % (95 % CI: 70 to 97) and for those with polyps greater than or equal to 10 mm it was 88 % (95 % CI: 56 to 98), with specificities of 76 % (95 % CI: 72 to 78) and 89 % (95 % CI: 86 to 90), respectively.  Both polyps missed by colonoscopy and mismatch in polyp size by study definition lowered specificity.  Overall colon cleanliness for CE was adequate in 78 % of patients (95 % CI: 68 to 86).  The authors concluded that the new 2nd-generation CCE was a safe and effective method for visualizing the colon and detecting colonic lesions.  These researchers stated that sensitivity and specificity for detecting colorectal polyps appeared to be very good, suggesting a potential for improved accuracy compared with the 1st-generation system; further prospective and comparative studies are needed.

An accompanying commentary noted that, although the sensitivity appears to be substantially improved compared to that of the first-generation colon capsule endoscopy, specificity remains disappointingly low. The commentators posited that the low specificity could be due to suboptimal sensitivity of the reference standard colonoscopy, or an imperfect polyp-matching algorithm, rather than “false positive” colon capsule endoscopy results (Spada, et al., 2010).  The authors replied that future studies should rectify these limitations by designing a statistical algorithm which will allow the above deficiencies to be corrected (Eliakim & Adler, 2010). The authors noted that clinical trials with complementary adjudication may be helpful too.

Spada and associates (2011) stated that CCE represents a non-invasive technology that allows visualization of the colon without requiring sedation and air insufflation.  A 2nd-generation CCE system (PillCam Colon 2) (CCE-2) was developed to increase sensitivity for colorectal polyp detection compared with the 1st-generation system.  In a prospective, multi-center trial including 8 European sites, these researchers evaluated the feasibility, accuracy, and safety of CCE-2 in a head-to-head comparison with colonoscopy.  This study involved 117 patients (mean age of 60 years); data from 109 patients were analyzed.  CCE-2 was prospectively compared with conventional colonoscopy as the criterion standard for the detection of colorectal polyps that were greater than or equal to 6 mm or masses in a cohort of patients at average or increased risk of CRC.  Colonoscopy was independently performed within 10 hours after capsule ingestion or on the next day.  CCE-2 sensitivity and specificity for detecting patients with polyps greater than or equal to 6 mm and greater than or equal to 10 mm were assessed.  Capsule-positive but colonoscopy-negative cases were counted as false positive.  Capsule excretion rate, level of bowel preparation, and rate of AEs also were assessed.  Per-patient CCE-2 sensitivity for polyps greater than or equal to 6 mm and greater than or equal to 10 mm was 84 % and 88 %, with specificities of 64 % and 95 %, respectively.  All 3 invasive carcinomas were detected by CCE-2.  The capsule excretion rate was 88 % within 10 hours.  Overall colon cleanliness for CCE-2 was adequate in 81 % of patients.  The authors concluded that in this European, multi-center study, CCE-2 appeared to have a high sensitivity for the detection of clinically relevant polypoid lesions, and it might be considered an adequate tool for colorectal imaging.  The main drawbacks of this study were not unblinding the CCE-2 results at colonoscopy; heterogenous patient population; and non-consecutive patients.

An accompanying commentary (Dominitz and Ko, 2011) noted that, unlike prior CT colonography studies and a prior multicenter study of the CCE-2, the test characteristics for detection of advanced adenomas were not reported. In addition, the investigators did not attempt to match polyp location, and therefore could not report per-polyp sensitivity and specificity. The commentators noted that the primary endpoint choice of sensitivity and specificity for polyps ≥6 mm is predicated on the notion that smaller polyps are very unlikely to be clinically meaningful, which is a matter of ongoing controversy. If all polyps (including diminutive polyps) were to be included, the per-patient sensitivity for large polyps certainly would increase, although the specificity would be expected to fall. The investigators' approach to matching polyps was based on size within 50% of the estimated size during colonoscopy. Although software for measuring polyps during CCE-2 procedures was used, the accuracy of this software as well as the accuracy of colonoscopic size determination is not well-understood. The authors suggest that the majority of false-positive CCE-2 cases were associated with a polyp seen during colonoscopy, although the size mismatch exceeded their prespecified criteria. The accuracy of procedures done with the CCE-2 is certain to be operator dependent; it remains to be seen whether results will improve with additional operator experience, whether wide-spread adoption would produce wide variation in results, or whether both outcomes would occur. The commentators also observed that most study participants were undergoing colonoscopy because of symptoms or for surveillance. The commentator explaind that this may have increased the estimated sensitivity over what would have been found in a screening population.  The commentators also noted that, in this study, colonoscopy was performed within 10 hours of capsule ingestion in order to avoid the need for a second bowel preparation. Therefore, the CCE-2 recordings were not reviewed before colonoscopy. In clinical practice, either the images would need to be interpreted very quickly or patients would need to undergo a second bowel preparation, which may limit its acceptability. Also, the bowel preparation was reported to be adequate for CCE-2 interpretation in only 81% of cases (compared to 92% of colonoscopies). In addition to inadequate bowel preparation, 6% of CCE-2 studies were excluded from the analysis for a variety of technical problems.

Pioche and colleagues (2012) noted that in France, approximately 5 % of colonoscopies are incomplete or temporarily contraindicated.  In a prospective study, these researchers tested the diagnostic yield of CCE in these patients.  This trial was carried out in 17 French centers, inclusion criteria were colonoscopy failure or general disease that excluded colonoscopy with anesthesia.  Patients underwent CCE using the 1st-generation PillCam Colon capsule.  The main end-point was CCE diagnostic yield, defined as identification of a colorectal lesion that directly explained symptoms or necessitated a diagnostic or therapeutic examination.  A secondary objective was to test a simplified Movi-Prep colon cleansing.  Follow-up to identify missed symptomatic cancer was scheduled.  CCE showed positive findings in 36 patients (diagnostic yield of 33.6 %), among whom 23 subsequently underwent therapeutic intervention.  Among 64 patients with negative capsule findings, 9 had a complementary procedure showing adenomas in only 1 case; CCE was incomplete in 7/107 patients.  Colonoscopy was done in 1 patient to retrieve a capsule retained in the left colon, and sigmoidoscopy in 11 because the rectum was not reached.  No CRC was diagnosed during the follow-up period.  Colon cleansing with MoviPrep was rated good or excellent in 75.9 % of cases.  The authors concluded that this study showed the feasibility and the usefulness of CCE in the situation of colonoscopy failure or contraindication.  The colon capsule modality should be tested against other available approaches, such as virtual colonoscopy or repeat colonoscopy by an expert (Pioche, et al., 2012; Commentaire, 2012).

In a prospective, single-center study, Negreanu and co-workers (2013) examined the feasibility, accuracy and acceptability of PillCam Colon 2 in detection of significant lesions in patients at risk with CRC, unable or unwilling to perform colonoscopy.  This trial used the 2nd-generation of PillCam Colon capsule.  In all patients the readers were instructed to review the entire CCE examination using Rapid 7 software and additionally to note significant extra-colonic findings.  Colonic significant findings were described according to ESGE guidelines; CCE procedure completion rate, level of bowel preparation and rate of adverse events (AEs) were assessed.  A total of 70 patients at risk of CRC were enrolled in the study.  In 3 patients the procedure failed because the capsule was not functioning when entered the colon.  PillCam Colon 2 showed positive findings in 23 (34 %, 95 % CI: 21.6 % to 44.1 %) of the remaining 67 patients; 6 patients were diagnosed with tumors: 4 with colon cancers, 1 with gastric cancer and 1 with a small bowel cancer.  The capsule findings were confirmed after surgery in all these patients.  The capsule excretion rate in 12 hours was 77 % with 54 patients having a complete examination.  The rectum was not explored during CCE procedure, in 16 patients (23 %, 95 % CI: 13.7 % to 34.1 %).  Every patient accepted CCE as an alternative exploration tool and 65/70 (93 %) agreed to have another future control by CCE.  No complications were reported during or after CCE examination.  The authors concluded that the PillCam Colon 2 appeared to be effective for the detection of clinically relevant lesions with great acceptability rate, and it might be considered as a useful tool for colorectal imaging in patients unable or unwilling to undergo colonoscopy.  Moreover, these researchers stated that further studies are needed to validate the best approach to these patients.

This pilot study employed 2nd-generation of PillCam CE to detect colon cancers as well as other tumors in the GI tract.  Although case controlled studies are ultimately needed to demonstrate the sensitivity and specificity of PillCam CE, this pilot study indicated that PillCam CE is feasible and of high acceptance by patients.  This study suggested that PillCam CE may eventually be used for population-wide colon cancer screening.  It should be noted that this was a descriptive study on a new generation colon capsule.  Since no comparison with the gold standard technique (colonoscopy) was made, specificity and sensitivity of the method could not be assessed.

Rondonotti and colleagues (2014) stated that computed tomographic colonography (CTC) is a reliable option for screening subjects who are unable or unwilling to undergo optical colonoscopy (OC).  A colon capsule (PillCam Colon2 [CC2]) has shown promising results in detecting polyps larger than 6 mm.  These researchers compared the accuracy of CC2 and CTC in identifying individuals with at least 1 polyp greater than 6 mm and subjects' attitude toward the procedures.  A total of 50 individuals (mean age of 59.2 ± 5.8 y; 58 % men) with positive results from the immunochemical fecal occult blood test (iFOBT-positive) underwent CC2, CTC, and OC.  The unblinded colonoscopy, integrating OC, CTC, and CC2 results, was used as the reference standard.  In a per-patient analysis, the accuracy of CC2 and CTC were assessed for individuals with at least 1 polyp 6 mm or larger.  Individuals were asked to choose which procedure they would be willing to repeat between CTC and CC2.  The combination of OC, CTC, and CC2 identified 16 cases with at least 1 polyp 6 mm or larger (reference standard); CTC identified the polyps with 88.2 % sensitivity, 84.8 % specificity, a 3.0 positive likelihood ratio (PLR), and a 0.07 negative likelihood ratio (NLR); CC2 identified the polyps with 88.2 % sensitivity, 87.8 % specificity, a 3.75 PLR, and a 0.06 NLR; 39 subjects (78 %) said they preferred CC2 to CTC.  The authors concluded that CC2 and CTC detected polyps 6 mm and larger with high levels of accuracy; these techniques were effective in selecting iFOBT-positive individuals who did not need to be referred for colonoscopy; CC2 appeared to be better-tolerated than CTC, and could be a reliable alternative to CTC for iFOBT-positive individuals who were unable or unwilling to undergo OC.

Rex and co-workers (2015) noted that CE is a minimally invasive imaging method.  In a prospective study,  these investigators determined the accuracy of this technology in detecting polyps 6 mm or larger in an average-risk screening population.  Asymptomatic subjects (n = 884) underwent capsule colonoscopy followed by conventional colonoscopy (the reference) several weeks later, with an endoscopist blinded to capsule results, at 10 centers in the United States and 6 centers in Israel from June 2011 through April 2012.  An unblinded colonoscopy was performed on subjects found to have lesions 6 mm or larger by capsule but not conventional colonoscopy.  Among the 884 subjects enrolled, 695 (79 %) were included in the analysis of capsule performance for all polyps.  There were 77 exclusions (9 %) for inadequate cleansing and whole-colon capsule transit time fewer than 40 mins, 45 exclusions (5 %) before capsule ingestion, 15 exclusions (2 %) after ingestion and before colonoscopy, and 15 exclusions (2 %) for site termination.  Capsule colonoscopy identified subjects with 1 or more polyps 6 mm or larger with 81 % sensitivity (95 % CI: 77 % to 84 %) and 93 % specificity (95 % CI: 91 % to 95 %), and polyps 10 mm or larger with 80 % sensitivity (95 % CI: 74 % to 86 %) and 97 % specificity (95 % CI: 96 % to 98 %).  Capsule colonoscopy identified subjects with 1 or more conventional adenomas 6 mm or larger with 88 % sensitivity (95 % CI: 82 % to 93 %) and 82 % specificity (95 % CI: 80 % to 83 %), and 10 mm or larger with 92 % sensitivity (95 % CI: 82 % to 97 %) and 95 % specificity (95 %  CI: 94 % to 95 %).  Sessile serrated polyps and hyperplastic polyps accounted for 26 % and 37 %, respectively, of false-negative findings from capsule analyses.  The authors concluded that in an average-risk screening population, technically adequate capsule colonoscopy identified individuals with 1 or more conventional adenomas 6 mm or larger with 88 % sensitivity and 82 % specificity.  They stated that capsule performance appeared adequate for patients who could not undergo colonoscopy or who had incomplete colonoscopies; additional studies are needed to improve capsule detection of serrated lesions.

An accompanying commentary (Tierney, 2015) stated that this study reported a marked increase in specificity relative to prior studies of CCE, particularly for polyps ≥6 mm. The authors of the study reviewed potential reasons for this difference, including variations in the experience or training of the capsule readers, but the commentator noted that a major methodologic difference between these studies likely contributing to this high specificity is the generous size-matching criteria employed in the current study. The commentator noted the FDA's concerns over false-positive and false negative results from CCE. The FDA summary of the clinical data from this trial employed a different, presumably more restrictive, size matching criteria, resulting in a per patient sensitivity and specificity for ≥6 mm of approximately 71% and 81%, respectively. The commentator noted a number of logistical issues to implementing CCE, including the need for a complete bowel preparation plus the additional burden of the booster doses and possibly additional laxatives when capsule excretion is delayed. Capsule colonoscopy is also more susceptible to the bowel preparation compromising quality relative to conventional colonoscopy owing to the inability to wash and evacuate residual colonic contents. The procedure can involve a significant time commitment the day of the procedure with monitoring of capsule location in order to time the administration of booster preparation doses. Third, there are a significant number of technical failures. By contrast, CT colonography rarely has technical failures. Finally, the duration of the study and subsequent delayed interpretation does not allow for a colonoscopy to be performed on the same day. 

Spada and associates (2015) stated that in case of incomplete colonoscopy (IC), several radiologic methods have traditionally been used, but more recently, CE was also shown to be accurate.  In a prospective, cohort trial, these investigators compared CCE and CTC in patients with IC.  Consecutive patients with a previous IC underwent CCE and CTC followed by colonoscopy in case of positive findings on either test (polyps/mass lesions greater than or equal to 6 mm).  Clinical follow-up was performed in the other cases to rule out missed cancer; CTC was performed after colon capsule excretion or 10 to 12 hours post-ingestion.  Since the gold standard colonoscopy was performed only in positive cases, diagnostic yield and PPVs of CCE and CTC were used as study end-points.  A total of 100 patients were enrolled; CCE and CTC were able to achieve complete colonic evaluation in 98 % of cases.  In a per-patient analysis for polyps greater than or equal to 6 mm, CCE detected 24 patients (24.5 %) and CTC 12 patients (12.2 %).  The relative sensitivity of CCE compared to CTC was 2.0 (95 % CI: 1.34 to 2.98), indicating a significant increase in sensitivity for lesions greater than or equal to 6 mm.  Of larger polyps (greater than or equal to 10 mm), these values were 5.1 % for CCE and 3.1 % for CTC (relative sensitivity: 1.67 (95 % CI: 0.69 to 4.00)); PPVs for polyps greater than or equal to 6 mm and greater than or equal to 10 mm were 96 % and 85.7 %, and 83.3 % and 100 % for CCE and CTC, respectively.  No missed cancer occurred at clinical follow-up of a mean of 20 months.  The authors concluded that CCE and CTC were of comparable efficacy in completing colon evaluation after IC; the overall diagnostic yield of colon capsule was superior to CTC.

Nogales and colleagues (2017) stated that CCE is an alternative approach for the examination of the colon in patients who refuse colonoscopy or after IC.  In a prospective, multi-center study involving 10 Spanish hospitals, these researchers determined the frequency of complete colonoscopy after IC, the diagnostic yield of CCE, the therapeutic impact of lesions found in CCE, the level of colon cleanliness and the safety of the procedure.  Consecutive outpatients aged greater than or equal to 18 years with previous IC were invited to participate.  The latest version of the CCE device, PillCam COLON 2 (CCE-2), was administered to all patients according to the protocol.  The study population comprised 96 patients.  The most frequent cause of IC was the inability to move past a loop using standard maneuvers (75/96 patients, 78 %).  Complete visualization of the colon was obtained with CCE-2 in 69 patients (71.9 %).  Of the 27 patients in whom the CCE-2 did not reach the hemorrhoidal plexus, it passed the colonic segment explored with the previous colonoscopy in 20 cases; thus, it could be inferred that a combined approach (CCE-2 plus colonoscopy) enabled complete visualization of the colonic mucosa in 92.7 % of patients.  CCE-2 revealed new lesions in 58 patients (60.4 %).  Polyps were the most frequent finding (41 patients; 42.7 % of the total number of patients).  In 43 of the 58 patients (44.8 % of the total number of patients), the new lesions observed led to modification of therapy, which included a new colonoscopy for polyp resection or surgery in patients with colonic neoplasm.  The authors concluded that CCE-2 was a suitable diagnostic procedure that could lead to more frequent diagnosis of significant colonic lesions after IC.

An accompanying commentary (Mascarenhas-Saraiva, 2017) noted that there are several alternatives to CCE for persons with an incomplete colonoscopy. Patients may undergo computerized tomography colonography, or may undergo a repeat colonoscopy at a tertiary referral center with dedicated endoscopes  – pediatric, variable-stiffness, balloon-assisted enteroscopes or colonoscopes, robotic colonoscopes  –  and/or expert endoscopists using special insertion tricks (such as monitoring progression under the guidance of magnetic imaging or fluoroscopy, water immersion technique, carbon dioxide insufflation). The commentator stated that the application of these techniques allows a complete colonoscopic examination in most patients. The commentator concluded "while the core of evidence opens a door for CCE in patients with incomplete colonoscopy, larger multicenter, randomized studies are needed."

Alvarez-Urturi et al (2017) noted that individuals with a family history of CRC have an increased risk of CRC.  In a prospective, multi-center study, these researchers evaluated the diagnostic yield of CCE in the detection of lesions and also 2 different colon preparations.   This trial was designed to assess CCE diagnostic yield in a cohort of asymptomatic individuals with a family history of CRC; CCE and colonoscopy were performed on the same day by 2 endoscopists who were blinded to the results of the other procedure.  A total of 53 participants were enrolled.  The sensitivity, specificity, PPV, and NPV of CCE for detecting advanced adenomas were 100 %, 98 %, 67 %, and 100 %.  Sensitivity, specificity, PPV, and NPV of CCE for the diagnosis of individuals with polyps were 87 %, 97 %, 93 %, and 88 %, respectively.  CCE identified 100 % of individuals with significant or advanced lesions.  Overall cleanliness was adequate by 60.7 % of them.  The PEG-ascorbic boost appeared to improve colon cleanliness, with similar colonic transit time.  The authors concluded that CCE was a promising tool, but it has to be considered as an alternative technique in this population in order to reduce the number of colonoscopies performed.  These researchers stated that more studies are needed to understand appropriate screening follow-up intervals and optimize the bowel preparation regimen.

The main drawback of this study was that it was performed with a 1st-generation capsule endoscopy.  Recently, a 2nd-generation capsule, Pillcam Colon 2, has been developed, which has higher sensitivity and obtains better results in the detection of colonic lesions.  The main studies published with 1st-generation CCE reported sensitivities of 63 to 88 % and specificities of 64 to 94 % in the detection of colonic lesions with high NPVs.  Two recent meta-analyses with 626 patients and 837 patients, respectively, found sensitivities for significant polyps of 69 to 76 %, with specificities of 86 % and 82 %.  In this study, CCE detected 4 polyps that were missed on colonoscopy and repeat colonoscopy confirmed only 1 of them as a true positive result.  The remainder was confirmed as false positives of CCE.  Assessment of polyp size could also lead to confusion.  In this respect, the 2nd-generation capsule is an important new advance that allows measurement of polyp size, which will probably decrease the number of false-positive results but not make them disappear completely.

Hussey and associates (2018) stated that same-day CCE immediately following incomplete OC would have a number of advantages for patients, while also presenting unique procedural challenges including the effect of sedation on capsule propulsion and patient tolerance of protracted preparation and fasting.  In an observational, prospective, single-center study, these investigators examined the efficacy of same-day CCE after incomplete OC in an unselected patient cohort.  Patients with an incomplete OC for any reason other than obstruction or inadequate bowel preparation were recruited; CCE was performed after a minimum of a 1-hour fast.  Once the patient was fully alert, intravenous metoclopramide was administered after capsule ingestion when possible, and a standard CCE booster protocol was then followed.  Relevant clinical information was recorded; CCE completion rates, findings and their impact, and AEs were noted.  A total of 50 patients were recruited, mean age of 57 years and 66 % (n = 32) were women; 76 % (n = 38) of CCEs were complete; however, full colonic views were obtained in 84 % (n = 42) of cases.  Patients greater than 50 years of age were 5 times more likely to have an incomplete CCE and there was also a trend towards known co-morbidities associated with hypomobility having reduced excretion rates.  Overall diagnostic yield for CCE in the unexplored segments was 74 % (n = 37), with 26 % (n = 13) of patients requiring significant changes in management based on CCE findings.  The overall incremental yield was 38 %; CCE findings were normal 26 % (n = 13), polyps 38 % (n = 19), inflammation 22 % (n = 11), diverticular disease 25 % (n = 12), angiodysplasia 3 % (n = 1) and cancer 3 % (n = 1).  Significant small bowel findings were found in 3 (6 %) cases, including CD and a neuroendocrine tumor.  A major AE occurred in 1 patient (2 %), related to capsule retention.  The authors concluded that same-day CCE was a viable alternative means to evaluate unexplored segments of the colon after incomplete OC in selected patients.

In a prospective, multi-center study, Baltes and co-workers (2018) examined the ability of PillCamColon2 to visualize colonic segments missed by incomplete OC and evaluated the diagnostic yield.  This trial  included 81 patients from 9 centers who underwent 2nd-generation CCE following incomplete OC performed by an experienced gastroenterologist (more than 1,000 colonoscopies); patients with stenosis were excluded.  According to patient preferences, CCE was performed the following day (protocol A) after staying on clear liquids and 0.75 L Moviprep in the morning or within 30 days after new split-dose Moviprep (protocol B).  Boosts consisted of 0.75 L and 0.25 L Moviprep, and phospho-soda was given as a rescue if the capsule was not excreted after 7 hours.  A total of 74 patients were analyzed (51 % of them in group A; 49 % in group B).  Bowel cleansing was adequate in 67 % of cases, and CCE could visualize colonic segments missed by incomplete colonoscopy in 90 % of patients under protocol A and 97 % of patients under protocol B (p = 0.35, n.s.).  Significant polyps including adenocarcinoma were detected in 24 % of cases.  Detection rates for all polyps and significant polyps per patient were similar in both protocols.  Polyps were found predominantly in the right colon (86 %) in segments that were not reached by OC.  Extra-colonic findings such as reflux esophagitis, suspected BE, upper GI-bleeding, gastric polyps, gastric erosions and angiectasia were detected in 8 patients.  PillCamColon2 capsule was retained in the ileum of 1 patient (1.4 %) without symptoms and removed during an uneventful resection for unknown CD that was diagnosed as the cause of anemia, which was the indication for colonoscopy; CCE was well-tolerated; 1 patient suffered from self-limiting vomiting after consuming the phospho-soda.  The authors concluded that 2nd-generation CCE using a low-volume preparation was useful after incomplete OC, and it allowed for the detection of additional relevant findings, but cleansing efficiency could be improved.

The authors stated that the drawbacks of this study included patients could choose between preparation protocols for CCE without randomization; the area reached by the colonoscopy was described, but tattooing was only optional; and long-term follow-up was not part of the present study.

Kobaek-Larsen et al (2018) determined the polyp detection rate and per-patient sensitivity for polyps greater than 9 mm of CCE compared with colonoscopy as well as the diagnostic accuracy of CCE.  Individuals who had a positive immunochemical fecal occult blood test during screening had investigator blinded CCE and colonoscopy.  Participants underwent repeat endoscopy if significant lesions detected by CCE were considered to have been missed by colonoscopy.  There were 253 participants; the polyp detection rate was significantly higher in CCE compared with colonoscopy (p = 0.02).  The per-patient sensitivity for greater than 9 mm polyps for CCE and colonoscopy was 87 % (95 % CI: 83 % to 91 %) and 88 % (95 % CI: 84 % to 92 %) respectively.  In participants with complete CCE and colonoscopy examinations (n = 126), per-patient sensitivity of greater than 9 mm polyps in CCE (97 %; 95 % CI: 94 % to 100 %) was superior to colonoscopy (89 %; 95 % CI: 84 % to 94 %).  A complete capsule endoscopy examination (n = 134) could detect patients with intermediate or greater risk (according to the European guidelines) with an accuracy, sensitivity, specificity and positivity rate of 79 %, 93 %, 69 % and 58 % respectively, using a cut-off of at least 1 polyp of greater than 10 mm or more than 2 polyps.  The authors concluded that CCE was superior to colonoscopy in polyp detection rate and per-patient sensitivity to greater than 9 mm polyps, but only in complete CCE examinations. The rate of incomplete CCE examinations must be improved.

Parodi and co-workers (2018) noted that CCE has been recognized as an alternative for CRC screening in average-risk people.  These researchers prospectively evaluated the accuracy of CCE as a screening tool in 1st-degree relatives (FDRs) of people with CRC by using OC with segmental unblinding as the reference standard.  Consecutive patients admitted with a CRC diagnosis (index cases) were prospectively evaluated and invited to contact their FDRs.  Available FDRs were invited to undergo CCE and OC on the following day, with segmental unblinding of CCE results.  Sensitivity, specificity, and PPV/NPV of CCE were assessed for detecting patients with any polyp greater than or equal to 6 mm and greater than or equal to 10 mm.  A total of 177 FDRs (median age of 57.0 years, 54.8 % women) identified from 211 index cases were included.  Both CCE and OC were completed in all the included FDRs.  Overall, CCE identified 51 of 56 FDRs with polyps greater than or equal to 6 mm (sensitivity 91 %; 95 % CI: 81 to 96) and correctly classified as negative 107 of 121 participants without lesions greater than or equal to 6 mm (specificity 88 %; 95 % CI: 81 to 93).  Per-patient PPV and NPV for greater than or equal to 6 mm lesions were 78 % (95 % CI: 67 to 87) and 95 % (95 % CI: 90 to 98), respectively.  CCE detected 24 of 27 patients with polyps greater than or equal to 10 mm and correctly classified as negative 142 of 150 patients, corresponding to 89 % sensitivity and 95 % specificity.  Post-CCE referral rates to colonoscopy were 37 % and 18 %, respectively.  The authors concluded that CCE was an accurate method to screen FDRs of patients with CRC and could be offered as an alternative to those who decline or are unfit for colonoscopy screening.

Chronic Kidney Disease

Docherty et al (2015) stated that there are only few reports on the DY of SBCE in patients with chronic kidney disease (CKD).  In a retrospective study, these investigators reported their SBCE experience in patients with CKD.  Case notes of patients with low estimated glomerular filtration rate (eGFR) who underwent SBCE (March 2005 to August 2012) for anemia and/or OGIB were retrieved and abstracted.  Severity of CKD was defined according to Renal Association recommendations as: Stage 3 (eGFR: 30 to 59); Stage 4 (eGFR: 15 to 29); and Stage 5 (eGFR less than 15 or on dialysis).  In the afore-mentioned period, 69 patients with CKD [Stage 3: 65/69 (92.8 %), Stage 4 or 5:4/69 (7.2 %)] had SBCE; 51/65 (78.5 %) patients with Stage 3 CKD had SBCE due to unexplained anemia and/or OGIB [43 (66.1 %) and 8 (12.3 %), respectively].  In 25/51 (49 %), the SBCE was normal and in 17/51 (33.3 %) showed small-bowel angiectasias.  Other findings were active bleeding (n = 2), fold edema (n = 2), ileal erosions (n = 1), adenocarcinoma (n = 1), and inconclusive/videos not available (n = 3).  All patients (n = 4) with CKD Stage 4 or 5 were referred due to unexplained anemia; 3/4 (75 %) had angiectasias and 1 normal SBCE.  Fecal calprotectin (FC) was measured in 12 patients with CKD Stage 3 and unexplained anemia prior to their SBCE; no significant small-bowel inflammation was found in this subgroup.  The authors concluded that SBCE has limited DY in CKD patients referred for unexplained anemia.  Sinister SB pathology is rare, while the most common finding is angiectasias.  Furthermore, FC measurement prior to SBCE -- in this cohort of patients -- is not associated with increased DY.

Evaluation of Mucosal Inflammation in Crohn's Disease / Ulcerative Colitis

Shi and colleagues (2015) stated that evaluation of mucosal inflammation is important in the management of patients with ulcerative colitis (UC). Colon capsule endoscopy has recently been shown to be effective in colorectal polyp detection.  However, its role in the evaluation of mucosal inflammation in UC is unclear.  This systematic review aimed to clarify the state of the art with an evidence-based summary of current studies on the utility of CCE in UC.  The overall results showed that the accuracy of CCE for assessment of mucosal inflammation in UC appeared to be comparable with that of colonoscopy.  Moreover, the authors stated that long-term follow-up studies with larger sample size are needed to further validate the utility of CCE in the management of UC subjects in clinical practice.

Furthermore, UpToDate reviews on “Clinical manifestations, diagnosis, and prognosis of ulcerative colitis in adults” (Peppercorn and Kane, 2016), “Management of mild to moderate ulcerative colitis in adults” (MacDermott, 2016), and “Management of severe ulcerative colitis in adults” (Peppercorn and Farrell, 2016) do not mention capsule endoscopy as a management tool.

Eliakim an colleagues (2020) stated that capsule endoscopy is an important modality for monitoring of CD.  Recently, a novel panenteric capsule, PillCam Crohn’s (Medtronic, Minneapolis, MN), was approved for use.  No quantitative index of inflammation for this method is currently available.  This sub-study of a prospective randomized controlled Comprehensive individUalized pRoactive ThErapy of CD trial (CURE-CD), which aimed to compare the correlation and reliability of the novel PillCam Crohn's score with the existing small bowel capsule Lewis inflammatory score.  The study cohort included CD patients in remission who were evaluated with PillCam Crohn's.  Each result was independently reviewed by 2 experienced readers.  Inflammation was scored in all studies using Lewis inflammatory score and PillCam Crohn's score (comprised of a sum of scores for most common and most severe lesions multiplied by percentage of segmental involvement + stricture score).  A total of 54 PillCam Crohn's studies from 41 patients were included.  The median Lewis inflammatory score was 225 for both readers.  The median PillCam Crohn's score was 6 (0 to 14) and 4 (3 to 15) for readers 1 and 2, respectively.  There was a high inter-rater reliability coefficient between the 2 readers for Lewis inflammatory and PillCam Crohn's score (0.9, p < 0.0001 for both).  The correlation between PillCam Crohn's score and fecal calprotectin was stronger than for Lewis inflammatory score (r = 0.32 and 0.54, respectively, p = 0.001 for both).  The authors concluded that the novel panenteric PillCam Crohn's capsule score correlated well with the Lewis inflammatory score, had excellent reliability, and may be potentially more accurate in estimation of the panenteric inflammatory burden in CD patients.  These researchers stated that PillCam Crohn's usability, responsiveness to change, and predictive accuracy merits further prospective evaluation.

The authors stated that this study had several drawbacks.  Primarily, the study cohort was comprised of patients in clinical remission, with a relatively low degree of mucosal inflammation.  Furthermore, the proportion of patients with colonic disease was quite low.  The number of subjects with repeated PillCam Crohn’s examinations was quite low.  Nevertheless, the main objective of this study was to describe and evaluate the reliability of the novel PillCam Crohn’s index; its performance for repeated evaluations will need to be evaluated in further detail in subsequent studies.  An additional drawback arose from the different methods of segmentation used by the small bowel capsules and the PillCam Crohn's.  The Lewis score, designed for small bowel capsules, splits the small bowel into 3 equal segments using transit time.  In contrast, PillCam Crohn’s software approximated anatomical segmentation.  The created tertiles did not necessarily overlap, potentially creating a measurable difference in individual tertile scores obtained by 2 systems.  However these investigators believed that this was a very minor limitation for all practical purposes, and this was confirmed by strong correlations of the total scores reported in this study.  One important drawback of the PillCam Crohn’s capsule was the need for vigorous colonic preparation, similar to the regimen used for colonic capsules.  Although the small bowel capsule is a very use-friendly modality, extensive preparation used for the PillCam Crohn’s may potentially hamper the patient’s willingness to undergo and repeat the procedure.  In this study, colonic preparation was poor and precluded reading the images in only 1 patient.  Moreover, in the current study these researchers did not require colonic preparation in patients without colonic involvement on the initial examination; the main reason for this decision was the attempt to improve the patient experience.  However, in the future these researchers should consider evaluating a simpler and less rigorous cleansing protocol for the PillCam Crohn's capsule endoscope.

Fecal Occult Blood Testing (FOBT) as a Screening Tool for Small-Bowel Capsule Endoscopy

Yung and colleagues (2017) noted that fecal occult blood testing (FOBT) has been suggested as a potential screening tool for small-bowel CE.  These researchers conducted a meta-analysis of studies correlating FOBT and CE findings to examine the predictive value of positive FOBT for CE findings.  PubMed and Embase searches were carried out; sensitivity, specificity and diagnostic odds ratios (DORs) were calculated.  A total of 6 studies were identified; 4 used fecal immunochemical testing (FIT), 1 used FIT and guaiac FOBT, 1 used hemoglobin/haptoglobin complex testing (Hb/Hpt); 5 of the 6 studies were suitable for statistical analysis.  For all positive FOBT, sensitivity for small-bowel findings was 0.60 (95 % CI: 0.50 to 0.69), specificity was 0.72 (95 % CI: 0.52 to 0.86), and DOR was 3.96 (95 % CI: 1.50 to 10.4).  For the 4 studies using only FIT, sensitivity was 0.48 (95 % CI: 0.36 to 0.61), specificity was 0.60 (95 % CI: 0.42 to 0.76), and DOR was 1.41 (95 % CI: 0.72 to 2.75).  The authors concluded that although a number of modalities have been suggested for screening small-bowel CE referrals, none of them, including FOBT, offered a comprehensive solution.  They stated that further investigation is needed to refine screening methods (e.g., combining other fecal or serum markers) for the selection of patients for small-bowel CE.

Detection of Hookworm

He and co-workers (2018) noted that as one of the most common human helminths, hookworm is a leading cause of maternal and child morbidity, which seriously threatens human health.  Recently, wireless capsule endoscopy (WCE) has been applied to automatic hookworm detection.  Unfortunately, it remains a challenging task.  In recent years, deep convolutional neural network (CNN) has demonstrated impressive performance in various image and video analysis tasks.  In this paper, a novel deep hookworm detection framework is proposed for WCE images, which simultaneously models visual appearances and tubular patterns of hookworms.  This was the 1st deep learning framework specifically designed for hookworm detection in WCE images.  Two CNN networks, namely edge extraction network and hookworm classification network, were integrated in the proposed framework, which avoided the edge feature caching and sped up the classification.  Two edge pooling layers were introduced to integrate the tubular regions induced from edge extraction network and the feature maps from hookworm classification network, leading to enhanced feature maps emphasizing the tubular regions.  Experiments have been conducted on one of the largest WCE datasets with WCE images, which demonstrated the effectiveness of the proposed hookworm detection framework.  It significantly out-performed the state-of-the-art approaches.  The authors concluded that the high sensitivity and accuracy of the proposed method in detecting hookworms showed its potential for clinical application.

Diagnosis of Iron Deficiency Anemia

Yung and colleagues (2017) stated that recent data implied young patients (aged less than or equal to 50 years) undergoing SBCE for iron deficiency anemia (IDA) showed higher diagnostic yield (DY) for sinister pathology. In  a retrospective, multi-center study (2010-2015), these investigators examined DY of CE in a large cohort of young IDA patients, and evaluated factors predicting significant SB pathology.  Consecutive, young patients from 18 centers/12 countries, with negative bi-directional GI endoscopy undergoing SBCE for IDA were included in this analysis.  Exclusion criteria: previous/ongoing obscure-overt GI bleeding; age of less than 19 or greater than 50 years; co-morbidities associated with IDA.  Data retrieved: SBCE indications; prior investigations; medications; SBCE findings; final diagnosis.  Clinical and laboratory data were analyzed by multi-variate logistic regression.  Data on 389 young IDA patients were retrieved.  In total, 169 (43.4 %) were excluded due to incomplete clinical data; data from 220 (122 women/98 men; mean age of 40.5 ± 8.6 years) patients were analyzed.  Some 71 patients had at least 1 clinically significant SBCE finding (DY: 32.3 %).  They were divided into 2 groups: neoplastic pathology (10/220; 4.5 %), and non-neoplastic but clinically significant pathology (61/220; 27.7 %).  The most common significant but non-neoplastic pathologies were angioectasias (22/61) and CD (15/61).  On multi-variate analysis, weight loss and lower mean corpuscular volume(MCV) were associated with significant SB pathology (OR: 3.87; 95 % CI: 1.3 to 11.3; p = 0.01; and OR: 0.96; 95 % CI: 0.92 to 0.99; p = 0.03; respectively).  The model also demonstrated association between use of anti-platelets and significant SB pathology, although due to the small number of patients, definitive conclusions could not be drawn.  The author concluded that in IDA patients less than or equal to 50 years with negative bi-directional GI endoscopy, overall DY of SBCE for clinically significant findings was 32.3 %.  Some 5 % of this cohort was diagnosed with SB neoplasia; lower MCV or weight loss were associated with higher DY for SB pathology.

The authors stated that drawbacks of this study included its retrospective study design, meaning that clinical data were incomplete for several patients (almost 50 % in this cohort).  This could have led to some over-estimation of these findings.  These researchers have attempted to minimize this possible effect using multi-variate analysis as detailed.  Secondly, many of our centers were high-volume or tertiary referral centers, which would therefore have taken a disproportionate number of complex patients or those suspected of having sinister pathology.  Finally, this study used MCV as a marker of iron deficiency in anemic patients, although limitations exist to the use of MCV to quantify iron deficiency.  Other red cell indices such as mean cell hemoglobin (MCH) (i.e. markers of hypochromia rather than microcytosis) may correlate better with severity of IDA than MCV.  Current guidelines state that MCV alone is not enough to make a diagnosis of IDA and other parameters, namely ferritin, should be used to assess iron status, as ferritin correlates well with total body iron stores and is a better marker of iron deficiency; low MCV occurs only in the later stages of iron deficiency.  Data on ferritin were not available for all the patients in this cohort, and MCV was used in this study due to its widespread use and availability.  Both markers are less reliable in elderly and/or hospitalized patient populations, and in several other co-morbidities (e.g., inflammation and anemia of chronic disease) but may be more reliable in the younger group that overall has a lower rate of co-morbidities.

Diagnosis of Takayasu’s Arteritis

Olano and Cohen (2018) reported the case of an asymptomatic 62-year old woman who presented for a routine annual examination and was found to have elevated sedimentation rate and C-reactive protein (CRP).  Physical examination was normal except for a difference in blood pressure between arms (120/74 on the right and 110/76 on the left); CT scan demonstrated an aortitis, and positron emission tomography (PET)-CT scan confirmed the presence of active inflammatory disease.  She received prednisone 40 mg daily for 30 days, but discontinued the treatment because of multiple side effects.  One year later, the patient presented with severe intermittent abdominal pain, which was worst when lying flat.  She had anorexia and lost several pounds.  Results from ultrasonography and basic laboratory studies were unremarkable other than elevation of acute phase reactants and anemia.  Upper GI endoscopy revealed gastritis.  Colonoscopy showed patchy submucosal hemorrhagic lesions in the right colon.  Capsule endoscopy showed diffuse and continuous involvement of the jejunal mucosa, with edema, reddish patches, and erosions.  Ischemic manifestation of a large vessel vasculitis (Takayasu’s arteritis) was suspected.  The authors stated that to their knowledge, this was the first report of Takayasu’s arteritis diagnosed by video capsule endoscopy.

An UpToDate review on “Clinical features and diagnosis of Takayasu arteritis” (Merkel, 2018) does not mention capsule endoscopy as a diagnostic tool.

Small Bowel Arterio-Venous Malformations

Singh and colleagues (2019) noted that small bowel arterio-venous malformations (AVMs) pose a bleeding risk and have traditionally been diagnosed by invasive enteroscopic procedures in patients with hereditary hemorrhagic telangiectasia (HHT).  Capsule endoscopy is emerging as a safe and non-invasive alternative for small intestinal evaluation, but its diagnostic yield and utility in diagnosing small bowel AVMs in HHT patients are under-studied.  These investigators examined the utility of CE for diagnosing AVMs in HHT patients.  They carried out a meta-analysis and systematic review of the literature on CE in HHT patients identified in the PubMed, EMBASE, Scopus, and Cochrane databases from inception to March 2018.  Summary effects were estimated using a random effects model.  After applying exclusion criteria, 5 studies (n = 124 patients) were eligible for meta-analysis.  The pooled diagnostic yield for visualization of small bowel AVMs by CE was 77.0  % (95 % CI: 65.8 - to 85.4  %, p  <  0.001).  The authors concluded that despite its limitations, this meta-analysis revealed that CE is a safe, non-invasive, and inexpensive diagnostic modality to diagnose and map AVMs along the GI tract in patients with HHT.  Although no studies compared the diagnostic yield of CE versus invasive enteroscopy to diagnose AVMs in this population, these findings suggested that CE is a useful alternative to invasive enteroscopy for diagnosing AVMs in patients with HHT.  Studies in the general population suggested that therapies started after CE increased hemoglobin levels and reduced risk of rebleeding, transfusion requirements, the number of GI procedures performed, and the duration of hospitalization.  Hence, when AVMs are identified, endoscopic, surgical, and medical AVM therapies can be considered and decisions can be made according to the mapped locations of AVMs.  Due to the heightened risk of GI bleeding and mortality in HHT patients and current guidelines recommending against treating non-bleeding small bowel AVMs in asymptomatic patients in the general population, additional studies should focus on whether there is clinical benefit to early treatment of asymptomatic AVMs in HHT patients.  If there is demonstrable benefit, CE may not only serve as an important diagnostic modality in HHT patients but may also become an important screening tool to prevent morbidity and mortality associated with GI bleeding.

The authors stated that this study had several limitations.  Hereditary hemorrhagic telangiectasia is relatively rare, so only a limited number of studies had examined use of CE in HHT patients with suspected small bowel AVMs and this was reflected through a broad prediction interval despite a large pooled effect size.  Furthermore, all of the available studies were small, prospective and observational and there were no randomized clinical trials.  Based on the eligibility criteria, the study by Ingrosso et al (2004) was included even though the CE views were frequently suboptimal, which was likely to have affected the overall diagnostic yield.  Together, these factors may have contributed to effect size heterogeneity; indeed, exclusion of the Ingrosso et al (2004) study increased homogeneity.  Although exhaustive and concerted efforts were made, these researchers did not find any unpublished studies, and their meta-analysis may have suffered from publication bias.  Lastly, many of the enrolled patients did not have active GI bleeding and because CE has the highest diagnostic yield during active bleeding, the diagnostic yield in these studies may have been under-estimated.

Cytosponge for Diagnosis of Esophageal Pathology / Screening of Barrett's Esophagus

Iqbal and colleagues (2018) noted that esophageal adenocarcinoma is an increasingly common cause of morbidity and mortality in developed countries.  Most cases are considered the consequence of chronic GERD, with subsequent Barrett's metaplasia and dysplasia.  Because progression from Barrett's metaplasia to cancer occurs over many years, endoscopic screening and surveillance programs have been established, albeit with little or no consideration for cost-effectiveness.  As an alternative to the expensive and resource-demanding endoscopic surveillance, the Cytosponge has been developed to sample the esophageal mucosa efficiently.  The device is a compressed mesh sponge encapsulated in an ingestible gelatin pill attached to a string.  After swallowing, the capsule dissolves allowing the sponge to expand in the stomach.  As it is pulled out, cells are collected from the esophago-gastric junction and throughout the esophagus.  The cellular samples can be analyzed by cytology, immunohistochemistry, and molecular markers.  These researchers conducted a systematic review of all recent relevant studies to aid in defining the role of this novel technology, including studies of screening and surveillance of BE, esophageal squamous dysplasia detection, detection of eosinophilic esophagitis, and evaluation of benign esophageal diseases.  With the major limitation that most studies were performed by a single investigative group that developed the technology, the device yielded overall impressive results against the endoscopy/biopsy gold standard.  Patient acceptability was high.  The authors concluded that if these promising early results are validated by other investigators in other populations, the Cytosponge represents an important new advance in the detection of esophageal pathology that could potentially decrease the burden of endoscopic esophageal sampling.

Pech (2019) stated that BE is a risk factor for esophageal adenocarcinoma (EAC).  However, screening for BE is difficult since it is not yet clear who should be screened and which screening method is cost-effective.  Screening methods could be upper endoscopy at the time of the first screening colonoscopy, trans-nasal endoscopy, ECE, or the Cytosponge.  In order to prevent the development of BE or its neoplastic progression, there are modifiable risk factors like obesity or smoking that can be influenced.  In addition, several drugs like proton pump inhibitors (PPIs), aspirin, non-steroidal anti-inflammatory drugs (NSAIDs) and statins have shown promising effects in mostly observational studies.  However, data from prospective randomized trials are scarce in order to draw final conclusions.

In a pilot study, Xu and co-workers (2019) examined the feasibility of combined screening for upper GI adenocarcinoma risk by serology and Cytosponge testing.  Blood samples from 56 patients with BE and 202 non-BE controls who previously took part in a trial assessing the accuracy of the Cytosponge test for BE were assessed for serum pepsinogen (PG) 1 and 2, gastrin-17, trefoil factor 3 (TFF3) and Helicobacter pylori (H. pylori) infection.  PG1 was pathological (less than 50 ng/ml) in 26 patients (10.1 %), none of whom had BE (p < 0.001).  Smoking and drinking had no influence on these results.  Pathological PG1 was associated with stomach pain (p = 0.029), disruption of sleep (p = 0.027) and disruption of diet by symptoms (p = 0.019).  Serum TFF3 was not associated with any clinical parameter.  The authors concluded that assessment of serum PG1 could be combined with a the Cytosponge test to identify additional patients requiring endoscopy.

The authors stated that this study was limited by the fact that no standard gastric biopsies were obtained on the study cohort to confirm gastric pathology.  The identification of patients with (advanced) pre-neoplastic conditions of the stomach and subsequent endoscopic surveillance of these mucosal changes remains currently the best option for gastric cancer prevention and early detection.  These researchers stated that prospective validation is needed to examine if the combination of a serum markers for gastric pathology with the minimally invasive Cytosponge test for BE could be a potentially complimentary strategy to identify patients at risk for upper GI adenocarcinoma.

Sanghi and Thota (2019) stated that BE is the precursor lesion for EAC.  Screening and surveillance of BE are undertaken with the goal of earlier detection and lowering the mortality from EAC.  The widely used technique is standard EGD with biopsies per the Seattle protocol for screening and surveillance of BE.  Surveillance intervals vary depending on the degree of dysplasia with endoscopic eradication therapy confined to patients with BE and confirmed dysplasia.  These researchers presented various novel techniques for screening of BE such as unsedated trans-nasal endoscopy (uTNE), the Cytosponge with TFF3, balloon cytology, ECE, liquid biopsy, electronic nose, and oral microbiome.  In addition, advanced imaging techniques such as narrow-band imaging (NBI), dye-based chromoendoscopy, confocal laser endomicroscopy, volumetric laser endomicroscopy (VLE), and wide-area trans-epithelial sampling with computer-assisted three-dimensional (WATS3D) analysis developed for better detection of dysplasia were also reviewed.  The authors concluded that although there are clear-cut guidelines on the screening and surveillance of BE, the current strategies are inadequate as more than 90 % of patients diagnosed with EAC do not have a prior diagnosis of BE.  The alternative non-endoscopic methods of screening that are in development may make screening available to the wider population while reducing the costs.  Based on current evidence, uTNE is suitable for mass screening for BE.  The Cytosponge and Esocheck non-endoscopic balloon are being validated in larger studies before they can be implemented for clinical use.  For better detection of dysplasia during surveillance, NBI is useful for recognition of subtle lesions as it is fast, easy to use, and accurate.  Furthermore, VLE and WATS3D are commercially available and have high dysplasia detection rates and have potential for future use.  These researchers stated that further studies are needed to examine their efficacy in decreasing EAC mortality rates and also to develop better biomarker panels for risk stratification of BE patients.

Weismuller and colleagues (2020) noted that although showing an increasing incidence over the past 20 years, EAC remains a rather uncommon cancer (i.e., compared to colorectal and gastric cancer).  Once detected, the prognosis of this cancer entity is still very poor.  Hence, in spite of some unfavorable prerequisites to systematic screening, the development of a screening concept for BE and EAC appeared worthwhile.  Presently, screening for BE and EAC is based on conventional endoscopy, mostly conducted individually in patients with reflux complaints (GERD).  Biopsies are taken obligatorily and are the centerpiece of diagnosis and scheduling of surveillance.  So far, endoscopy is the diagnostic gold standard, but it is expensive and obviously lacks effectiveness -- 8 of 10 cases of EAC are not detected in endoscopic screening (and surveillance), but by an opportunistic 1st-time endoscopy.  Thus, new methods for BE/EAC screening are strongly desirable.  Research must be concentrated to favor procedures applicable for a screening of the population in a primary care setting.  For that, the 1st step needs to consist of identifying a subgroup of people "at risk", which in a 2nd step can be risk assessed and followed-up by endoscopy and biopsy.  From all screening variants, which have been actually tested in clinical application and experimental research, biomarker-based techniques appeared to be most promising.  Among those under the aspect of costs and practicability, the Cytosponge, in addition with a panel of biomarkers, appeared to be promising in clinical trials.

The Esophageal String Test for Diagnosis of Esophageal Pathology

Ramirez and colleagues (2005) examined the feasibility, safety, accuracy, and acceptability of "string-capsule endoscopy (SCE)" in the evaluation of esophageal varices.  Strings were attached to the WCE device to allow its controlled movement up and down the esophagus.  Time of recording and discomfort associated with the procedure was documented.  Patient's preference compared to conventional EGD was recorded.  An independent endoscopist blinded to EGD diagnoses assessed the diagnostic accuracy of pictures obtained.  A total of 30 patients with clinical liver cirrhosis (mean age of 54.4 years; mean MELD score of 12.5, and mean Child-Pugh score of 6.3) were enrolled; 19 for surveillance and 11 for screening purposes.  The procedure was safe (no strings were disrupted and no capsule was lost).  The mean recording time was 5.8 mins (2.9 to 8.7), the accuracy was 96.7 %, and discomfort was minimal.  The majority (83.3 %) of patients preferred SCE to EGD.  The authors concluded that SCE was feasible, safe, accurate, highly acceptable, and preferred by cirrhotic patients undergoing screening/surveillance of esophageal varices.  The technique may prove to be more cost-effective than conventional EGD.

Stipho and associates (2012) noted that EGD is the gold standard for the screening and surveillance of esophageal varices.  A less invasive, safer and sedation-less alternative procedure is needed.  These investigators evaluated the sensitivity, specificity, PPV, NPV as well as the beyond the chance agreement (kappa index) of SCE in the diagnosis of esophageal varices.  This study included cirrhotic patients who underwent SCE and EGD for screening and surveillance purposes.  Varices were graded at EGD and SCE as small, medium and large.  Descriptors at SCE were added as follows: PLUS, for the presence of red wale signs or, MINUS for their absence, irrespective of the estimated variceal size.  Clinically significant varices were defined by their size (medium/large at either EGD or SCE) and/or, the PLUS descriptor irrespective of the estimated size at SCE.  Sensitivity, specificity, PPV, NPV, accuracy and kappa index were calculated.  Procedure time, procedure-related discomfort and patient's preference were documented.  A total of 100 patients (33 for screening and 67 for surveillance) were enrolled.  The sensitivity and specificity of SCE for clinically significant varices when using the PLUS/MINUS descriptors were 82 % and 90 %, respectively with a PPV of 84 % and NPV of 89 % and a kappa of 0.73.  The authors concluded that SCE had an acceptable sensitivity and specificity for the diagnosis of clinically significant esophageal varices; however, the lack of air insufflation hampered its correlation with the grading used with EGD.

Furuta and co-workers (2013) hypothesized that measurements of luminal eosinophil-derived proteins would correlate with esophageal mucosal inflammation in children with eosinophilic esophagitis (EoE).  The Enterotest diagnostic device was used to develop an esophageal string test (EST) as a minimally invasive clinical device.  EST samples and esophageal mucosal biopsies were obtained from children undergoing upper endoscopy for clinically defined indications.  Eosinophil-derived proteins including eosinophil secondary granule proteins (major basic protein-1, eosinophil-derived neurotoxin, eosinophil cationic protein, eosinophil peroxidase) and Charcot-Leyden crystal protein/galectin-10 were measured by ELISA in luminal effluents eluted from ESTs and extracts of mucosal biopsies.  ESTs were performed in 41 children with active EoE (n = 14), EoE in remission (n = 8), GERD (n = 4) and controls with normal esophagus (n = 15).  EST measurement of eosinophil-derived protein biomarkers significantly distinguished between children with active EoE, treated EoE in remission, GERD and normal esophagus.  Levels of luminal eosinophil-derived proteins in EST samples significantly correlated with peak and mean esophageal eosinophils / high-power field (HPF), eosinophil peroxidase indices and levels of the same eosinophil-derived proteins in extracts of esophageal biopsies.  The authors concluded that the presence of eosinophil-derived proteins in luminal secretions was reflective of mucosal inflammation in children with EoE.  They stated that EST was a novel, minimally invasive device for measuring esophageal eosinophilic inflammation in children with EoE.  Moreover, these researchers stated that although they did not suggest that the EST would replace endoscopy and biopsy as a critical tool for analyses of EoE, it certainly has the potential to significantly improve the evaluation and treatment of patients with EoE who may require repeated assessments of their esophageal mucosae.  In addition, as future biomarkers are identified and validated for EoE, ESTs may be able to differentiate patient phenotypes that are more pre-disposed to complications or responsive to various treatments.

Colli and associates (2014) noted that current guidelines recommend performance of EGD at the time of diagnosis of hepatic cirrhosis to screen for esophageal varices.  These guidelines require people to undergo an unpleasant invasive procedure repeatedly with its attendant risks, despite the fact that 50 % of the people do not have identifiable esophageal varices 10 years after the initial diagnosis of cirrhosis.  Video capsule endoscopy is a non-invasive test proposed as an alternative method for the diagnosis of esophageal varices.  Ina Cochrane review, these investigators determined the diagnostic accuracy of capsule endoscopy for the diagnosis of esophageal varices in children or adults with chronic liver disease or portal vein thrombosis, irrespective of the etiology; they investigated the accuracy of capsule endoscopy as triage or replacement of EGD.  These investigators identified 16 eligible studies, in which only adults with cirrhosis were included.  In 1 study, individuals with portal thrombosis were also included.  They classified most of the studies at high-risk of bias for the “Participants selection” and the “Flow and timing” domains.  One study assessed the accuracy of capsule endoscopy for the diagnosis of large (high-risk) esophageal varices.  In the remaining 15 studies that assessed the accuracy of capsule endoscopy for the diagnosis of esophageal varices of any size in people with cirrhosis, 936 participants were included; the pooled estimate of sensitivity was 84.8 % (95 % CI: 77.3 % to 90.2 %) and of specificity 84.3 % (95 % CI: 73.1 % to 91.4 %). Eight of these studies included people with suspected varices or people with already diagnosed or even treated varices, or both, introducing a selection bias; 7 studies including only people with suspected but unknown varices were at low-risk of bias; the pooled estimate of sensitivity was 79.7 % (95 % CI: 73.1 % to 85.0 %) and of specificity 86.1 % (95 % CI: 64.5 % to 95.5 %); 6 studies assessed the diagnostic accuracy of capsule endoscopy for the diagnosis of large esophageal varices, associated with a higher-risk of bleeding; the pooled sensitivity was 73.7 % (95 % CI: 52.4 % to 87.7 %) and of specificity 90.5 % (95 % CI: 84.1 % to 94.4 %); 2 studies also evaluated the presence of red marks, which were another marker of high-risk of bleeding; the estimates of sensitivity and specificity varied widely; 2 studies obtained similar results with the use of a modified device as index test (string capsule).  Due to the absence of data, these researchers could not perform all planned subgroup analyses.  Inter-observer agreement in the interpretation of capsule endoscopy results and any adverse event (AE) attributable to capsule endoscopy were poorly assessed and reported.  Only 4 studies evaluated the inter-observer agreement in the interpretation of capsule endoscopy results: the concordance was moderate.  The participants' preferences for capsule endoscopy or EGD were reported differently but appeared to be in favor of capsule endoscopy in 9 of 10 studies.  In 10 studies, participants reported some minor discomfort on swallowing the capsule.  Only 1 study identified other significant AEs, including impaction of the capsule due to previously unidentified esophageal strictures in 2 participants.  No AEs were reported as a consequence of the reference standard.  The authors concluded that they could not support the use of capsule endoscopy as a triage test in adults with cirrhosis, administered before EGD, despite the low incidence of AEs and participant reports of being better-tolerated.  Thus, these researchers stated that that EGD could not be replaced by capsule endoscopy for the detection of esophageal varices in adults with cirrhosis; and they found no data assessing capsule endoscopy in children and in people with portal thrombosis.

In a prospective study, Sacher-Huvelin and colleagues (2015) compared ECE with EGD for the diagnosis of esophageal varices in patients with cirrhosis.  A total of 330 patients with cirrhosis and with no known esophageal varices were enrolled.  Patients underwent ECE first, followed by EGD (gold standard).  The endoscopists who performed EGD were blind to the ECE result.  Patient satisfaction was assessed using a visual analog scale (VAS; maximum score = 100).  A total of 30 patients were excluded from the analysis because they did not undergo any endoscopic examinations.  Patients (mean age of 56 years; 216 men) had mainly alcoholic (45 %) or viral (27 %) cirrhosis.  The diagnostic indices of ECE to diagnose and correctly stage esophageal varices were: sensitivity of 76 % and 64 %, specificity of 91 % and 93 %, PPV of 88 % and 88 %, and NPV of 81 % and 78 %, respectively; ECE patient satisfaction scored significantly higher than EGD (87 ± 22 versus 58 ± 35; p < 0.0001).  The authors concluded that ECE was safe and well-tolerated in patients with liver cirrhosis and suspicion of portal hypertension.  The sensitivity of ECE was not currently sufficient to replace EGD as a first exploration in these patients.  However, due to its excellent specificity and PPV, ECE may have a role in cases of refusal or contraindication to EGD.  These investigators stated that ECE might also improve compliance to endoscopic follow-up and aid important therapeutic decision-making in the prophylaxis of bleeding.

Gora et al (2016) stated that due to the relatively high cost and inconvenience of upper endoscopic biopsy and the rising incidence of esophageal adenocarcinoma, there is currently a need for an improved method for screening for BE.  Ideally, such a test would be applied in the primary care setting and patients referred to endoscopy if the result is suspicious for BE.  Tethered capsule endomicroscopy (TCE) is a recently developed technology that rapidly acquires microscopic images of the entire esophagus in un-sedated subjects.  These researchers presented their first experience with clinical translation and feasibility of TCE in a primary care practice.  The acceptance of the TCE device by the primary care clinical staff and patients showed the potential of this device to be useful as a screening tool for a broader population.  The authors stated that even though the objective of this study was not to measure the diagnostic accuracy of TCE for BE, but to assess its feasibility and acceptability, a limitation was the lack of comparison with endoscopic examination and biopsy.  Furthermore, Implementation of a validated anxiety scale would have also been useful to standardize the results and compare this new technique to others already implemented in the clinic.

McCarty and co-workers (2017) noted that esophageal variceal bleeding is a severe complication of portal hypertension with significant morbidity and mortality.  Although traditional screening and grading of esophageal varices has been performed by EGD, WCE provides a minimally invasive alternative that may improve screening and surveillance compliance.  In a systematic review and meta-analysis, these researchers evaluated the efficacy of WCE for screening and diagnosis of esophageal varices among patients with portal hypertension.  They carried out searches of PubMed, Embase, Web of Science, and the Cochrane Library databases through December 2015.  Bi-variate and hierarchical models were used to compute the pooled sensitivity and specificity, and to plot the summary ROC with summary point and corresponding 95 % CIs.  Bias of included studies was assessed using the quality assessment of diagnostic accuracy studies-2.  A total of 17 studies from 2005 to 2015 were included in this meta-analysis (n = 1,328).  The diagnostic accuracy of wireless capsule endoscopy in the diagnosis of esophageal varices was 90 % [95 % CI: 0.88 to 0.93].  The diagnostic pooled sensitivity and specificity were 83 % (95 % CI: 0.76 to 0.89) and 85 % (95 % CI: 0.75 to 0.91), respectively.  The diagnostic accuracy of wireless capsule endoscopy for the grading of medium-to-large varices was 92 % (95 % CI: 0.90 to 0.94).  The pooled sensitivity and specificity were 72 % (95 % CI: 0.54 to 0.85) and 91 % (95 % CI: 0.86 to 0.94), respectively, for the grading of medium-to-large varices.  The use of capsule demonstrated only mild AEs.  A sensitivity analysis limited to only high quality studies revealed similar results.  The authors concluded that WCE was safe and well-tolerated in patients with liver cirrhosis and suspicion of portal hypertension.  The sensitivity of capsule endoscopy was not currently sufficient to replace EGD as a first exploration in these patients, but given its high accuracy, it may have a role in cases of refusal or contraindication to EGD.  Moreover, these researchers stated that the role for wireless esophageal capsule endoscopy remains limited as this modality lacks any potential therapeutic intervention.  Furthermore, based upon the current quality of overall evidence, more studies are needed to evaluate the use of WCE for screening and grading of esophageal varices.

The authors stated that aside from the small, inherent heterogeneity limitations of meta-analyses that exist between studies, a major limitation of this study was the modality of choice.  Only one study, by Sacher-Huvelin et al (2015), used the 2nd generation Pillcam.  All other studies included used the 1st generation modality.  In this French study, the sensitivity of capsule endoscopy in correctly diagnosing and staging esophageal varices was 76 % and 64 %, respectively; the specificity of the test was 91 % and 93 %, demonstrating that although capsule endoscopy was not sufficient at this time for the initial diagnosis of esophageal varices, the high specificity of the test may allow its use in cases where patients were refusing or were unable to undergo EGD.  The accuracy of capsule for diagnosis and grading was 84 % and 81 %, respectively.  This lower value may indicate the pooled accuracy of this study may be over-estimated; thus requiring more studies with the newer generation before a change in guidelines or surveillance recommendations are made.  Additionally, of the included studies, there was a high-risk of bias (8 of the total 17 studies) that may limit generalizability of these results; however, separate analysis excluded these studies was performed and yielded similar results.  Overall, based upon the current quality of evidence, more studies are needed to further evaluate the role of WCE for patients with portal hypertension.

In a review on “New developments in the diagnosis and treatment of eosinophilic esophagitis”, Nhu and Moawad (2019) provided an update on recent clinically relevant advances in the development of diagnostic and therapeutic approaches for EoE.  These investigators stated that current diagnostic and disease monitoring protocols for EoE rely on repetitive endoscopic evaluations and esophageal tissue acquisition for histopathologic analysis.  Recent advancements in EoE diagnosis include endoscopic functional lumen imaging probe (FLIP), trans-nasal endoscopy (TNE), and the emergence of non-invasive diagnostic tools including Cytosponge, esophageal string test, and mucosal impedance probe.

Furthermore, an UpToDate review on “Clinical manifestations and diagnosis of eosinophilic esophagitis” (Bonis, 2019) states that “Other diagnostic tests that have been evaluated but that are not routinely used include functional lumen imaging probe, endoscopic ultrasound, impedance planimetry to measure esophageal pressures and distensibility, mucosal impedance contour analysis that evaluates esophageal mucosal integrity, esophageal manometry, and endoscopic confocal laser microscopy.  The esophageal string test has been developed as a tool to measure esophageal inflammation in patients with esophagitis.  In one study that included 41 children, measurement of eosinophil-derived protein biomarkers distinguished between children with active eosinophilic esophagitis, treated eosinophilic esophagitis in remission, GERD, and those with a normal esophagus”.

Ackerman and colleagues (2019) noted that EoE lacks sensitive and specific peripheral biomarkers.  These researchers hypothesized that levels of EoE-related biomarkers captured using a 1-hour minimally invasive EST would correlate with mucosal eosinophil counts and tissue concentrations of these same biomarkers.  In a prospective, multi-center study, these investigators examined if a 1-hour EST accurately distinguishes active from inactive EoE or a normal esophagus.  Children and adults (age of 7 to 55 years) undergoing a clinically indicated EGD performed an EST with an esophageal dwell time of 1 hour.  Subjects were divided into 3 groups: active EoE, inactive EoE, and normal esophageal mucosa.  Eosinophil-associated protein levels were compared between EST effluents and esophageal biopsy extracts.  Statistical modeling was performed to select biomarkers that best correlated with and predicted eosinophilic inflammation.  A total of 134 subjects (74 children, 60 adults) with active EoE (n = 62), inactive EoE (n = 37), and patient controls with a normal esophagus (n = 35) completed the study.  EST-captured eosinophil-associated biomarkers correlated significantly with peak eosinophils/high-power field, endoscopic visual scoring, and the same proteins extracted from mucosal biopsies.  Statistical modeling, using combined eotaxin-3 and major basic protein-1 concentrations, led to the development of EoE scores that distinguished subjects with active EoE from inactive EoE or normal esophagi; 87 % of children, 95 % of parents, and 92 % of adults preferred the EST over endoscopy if it provided similar information.  The authors concluded that the 1-hour EST accurately distinguished active from inactive EoE in children and adults and may facilitate monitoring of disease activity in a safe and minimally invasive fashion.

The authors concluded that drawbacks of the study included the use of patient controls with an endoscopically and histologically normal esophagus, and that comparisons between subjects with active/inactive EoE were cross-sectional rather than longitudinal in the same subject.  Since the study was not designed to track patients with respect to treatments, this was not fully addressed and would be the focus of future studies.  Limitations of the 1-hour EST also included that it could not be used in patients who are unable to swallow pills or in those with esophageal narrowing or allergy to the gelatin capsule.  A potential confounding variable was that atopic patients may swallow EAPs derived from nasal, pulmonary, or ocular secretions; these secretions may adhere to the EST and increase the EST EAP concentrations.  These researchers have not noted any correlation between self-reported co-morbid allergic disease and increased levels of the EAPs in EST samples.

Colon Capsule Endoscopy for Detection of Colorectal Polyps

Kjolhede and associates (2021) stated that colon capsule endoscopy (CCE) is a technology that might contribute to CRC screening programs as a filter test between FIT and standard colonoscopy.  In a systematic review and meta-analysis, these investigators evaluated the literature for studies examining the diagnostic yield of second-generation CCE (CCE-2) compared with standard colonoscopy.  They carried out a systematic literature search in PubMed, Embase, and Web of Science.  Study characteristics including quality of bowel preparation and completeness of CCE transits were extracted.  Per-patient sensitivity and specificity were extracted for polyps (any size, greater than or equal to 10 mm, greater than o equal to 6 mm) and lesion characteristics.  Meta-analyses of diagnostic yield were performed.  The literature search revealed 1,077 unique papers and 12 studies were included.  Studies involved a total of 2,199 patients, of whom 1,898 were included in analyses.  The rate of patients with adequate bowel preparation varied from 40 % to 100 %.  The rates of complete CCE transit varied from 57 % to 100 %.  The meta-analyses demonstrated that mean (95 % CI) sensitivity, specificity, and DOR were: 0.85 (0.73 to 0.92), 0.85 (0.70 to 0.93), and 30.5 (16.2 to 57.2), respectively, for polyps of any size; 0.87 (0.82 to 0.90), 0.95 (0.92 to 0.97), and 136.0 (70.6 to 262.1), respectively, for polyps greater than or equal to 10 mm; and 0.87 (0.83 to 0.90), 0.88 (0.75 to 0.95), and 51.1 (19.8 to 131.8), respectively, for polyps greater than or equal to 6 mm.  No serious AEs were reported for CCE.  The authors concluded that CCE had high sensitivity and specificity for per-patient polyps compared with standard colonoscopy; however, the relatively high rate of incomplete investigations limited the use of CCE in a CRC screening setting.

Mollers and colleagues (2021) noted that adherence to CRC screening is still unsatisfactory in many countries; thus, limiting prevention of CRC.  Colon capsule endoscopy (CCE) could be an alternative to fecal immunochemical tests or optical colonoscopy for CRC screening, and might enhance adherence in CRC screening.  In a systematic review and meta-analysis, these researchers examined the diagnostic accuracy of CCE compared to optical colonoscopy (OC) as the gold standard, adequacy of bowel preparation regimes and the patient perspective on diagnostic measures.  They carried out a systematic literature search in PubMed, Embase and the Cochrane Register for Clinical Trials.  Pooled estimates for sensitivity, specificity and the DOR with their respective 95 % CI were calculated for studies providing sufficient data.  Of 840 initially identified studies, 13 were included in the systematic review and up to 9 in the meta-analysis.  The pooled sensitivities and specificities for polyps of greater than or equal to 6 mm were 87 % (95 % CI: 83 % to 90 %) and 87 % (95 % CI: 76 % to 93 %) in 8 studies, respectively.  For polyps of greater than or equal to 10 mm, the pooled estimates for sensitivities and specificities were 87 % (95 % CI: 83 % to 90 %) and 95 % (95 % CI: 92 % to 97 %) in 9 studies, respectively.  A patients' perspective was evaluated in 31 % (n = 4) of studies, and no preference of CCE over OC was reported.  Bowel preparation was adequate in 61 % to 92 % of CCE examinations.  The authors concluded that CCE provided high diagnostic accuracy in an adequately cleaned large bowel.  Moreover, these researchers stated that conclusive findings on patient perspectives require further studies to increase acceptance/adherence of CCE for CRC screening with a specific focus on patient perspectives on CCE-2.

The authors stated that a drawback of this meta-analysis was the small number of studies that could be included in the meta-analysis based on their protocol.  Furthermore, the number of newly published clinical trials was very low (n = 4) since the last meta-analysis was published in 2016.  The reasons might range from a poor adaptation of the technology to awaiting the 3rd generation of CCE, the focus on the patient perspective of CCE-2 or the evaluation of the impact of CCE-2 on screening participation.  There is a scarcity of studies among different populations (average risk, FIT/FOBT + or FDR) that prohibits generalization of these findings.  Another drawback of this meta-analysis was the heterogeneity of specificities, which was partially controlled by these researchers’ approach using random-effects models.  Furthermore, the low number of studies reporting the patient perspective and the heterogenous assessment did not allow for a clear conclusion on the patient perspective.  For more extensive data on the patient perspective and bowel preparation, separate reviews focusing on those outcomes might be needed.

Magnetic-Assisted Video Capsule Endoscopy

In a prospective, blinded, multi-center study, Lai and co-workers (2020) compared the safety and feasibility after GI check-up by standing-type magnetically controlled capsule endoscopy (SMCE) and conventional gastroscopy in patients from April 2018 to July 2018.  All patients first underwent SMCE and then subsequently had gastroscopy with intravenous (IV) anesthesia.  These investigators calculated the compliance rates of gastric lesion detection by SMCE using gastroscopy as the standard.  Capsule retention rate, incidence of AEs, and patient satisfaction were documented throughout the study.  A total of 161 patients who completed SMCE and gastroscopy were included in the analysis.  Positive compliance rate among SMCE and gastroscopy was 92.0 % (95 % CI: 80.77 % to 97.78 %).  Negative compliance rate was 95.5 % (95 % CI: 89.8 0 % to 98.52 %).  Moreover, overall compliance rate was 94.41 % 95 % CI: (89.65 % to 97.41 %); 64 pathological outcomes were identified.  Of these 64 outcomes, 50 were detected by both procedures.  The gastroscopy method neglected 7 findings (i.e., 5 erosions, 1 polyp, and 1 ulcer).  Furthermore, SMCE also overlooked 7 lesions (i.e., 1 erosion, 2 polyps, 1 atrophy, and 3 submucosal tumors).  Capsule retention or related AEs were not reported.  The authors concluded that SMCE provided equivalent agreement with gastroscopy and may be useful for screening of gastric illnesses without any anesthesia.  Moreover, these researchers stated that technical modifications as well as trials with larger sample sizes in a high‐risk population are needed to validate these preliminary findings.

The authors stated that this study had several drawbacks.  SMCE examination time was longer than the time needed for gastroscopy (24 mins versus 7 mins); however, in the future, the time needed will be less when the image is automatically analyzed using artificial intelligence (AI).  SMCE requires less time than does a similar product, which requires approximately 30 mins.  Furthermore, SMCE examination time varied from 7 mins to 47 mins.  Reasons for this were as follows: First, the SMCE capsule moved in the liquid by means of rolling and rotating.  Thus, when moving the same distance, the path of the capsule's lens was much longer than that of gastroscopy.  Second, the visual field of the capsule in the motion state changed rapidly, which made manipulation of the capsule more difficult and thus increased operation time.  Furthermore, the structure of each person's stomach cavity is different, resulting in different trajectories of capsule movement.  Third, the discomfort caused by standing may limit the use of SMCE in certain patients, but, in the future, a sitting method will solve this problem.  In addition, upper GI endoscopy generally includes examination of the entire esophagus, stomach, and duodenum.  Thus, SMCE targeted for stomach only may limit its use in clinical practice.  These investigators noted that SMCE system produces approximately 20,000 images per inspection and 30 to 60 mins are needed for a doctor to read them.  With the application of AI, SMCE system has been able to screen 80 to 90 % of similar images, greatly reducing the burden on doctors.  Similar to published studies, doctor's reading time will further shorten with the application of computer‐assisted diagnosis.  Image‐processing technologies have also been applied to CE.  A recent meta‐analysis showed that improved delineation was observed in 89 % of angioectasias and in 45 % of ulcer/erosions using flexible spectral imaging color enhancement.  However, imaging processing technology has not been applied in this study, but it will be implemented in the next generation of SMCE.

Melson and colleagues (2021) noted that VCE enables visualization of the mucosal surface throughout the GI tract in a minimally invasive manner.  Since initial FDA approval in 2001, this technology has been refined to provide superior resolution, increased battery life, and capabilities to view different parts of the GI tract.  VCE has an established, essential role in the evaluation of small-bowel lesions and bleeding.   Potential clinical applications for VCE have expanded to include the evaluation of inflammatory bowel disease (IBD) and screening for colorectal neoplasia in selected patients.  This document is an update of a 2013 ASGE Technology Committee article entitled “Wireless Capsule Endoscopy” and reviews currently available VCE systems and their applications.  These investigators stated that VCE currently serves diagnostic purposes but in the future may be able to provide therapeutic interventions.  Locomotion systems may permit anti-peristaltic capsule movement; alternatively, the application of external magnetic fields can be used to influence capsule movement.   Magnets have been used to guide the capsule; thus, improving visualization of the UGI tract.  In preliminary studies, magnets have shown to aid in trans-gastric passage of the capsule, reducing pyloric transit times.  The ability to reliably control the capsule by magnets to obtain visualization is still a field in development.  Proof-of-principle using magnetic-assisted capsule endoscopy studies have reported accuracy higher than previously reported for conventional VCE for esophageal varices (EV) and BE detection.  Magnetic-assisted capsule endoscopy can achieve longer times for visualization of the distal esophageal mucosa, and the extent to which this makes it a viable alternative to examine BE and presence of varices deserves further study.  Moreover, these researchers stated that VCE has a potential role in evaluating other UGI tract pathology (e.g., varices or BE) but has not yet found a role in routine practice.

Geropoulos and associates (2021) stated that the introduction of magnetically controlled CE overcame the restriction of passive CE movement; thereby, allowing an improved visualization of the GI lumen, where other imaging studies appeared to be unhelpful.  In a systematic review and meta-analysis, these researchers examined the performance of magnetically controlled CE and assessed its potential as a less invasive diagnostic method in the detection of gastric lesions.   They carried out a systematic search in PubMed (Medline), Embase, Google Scholar, Scopus, Who Global Health Library (GHL), Virtual Health Library (VHL), Clinicaltrials.gov, Cochrane Library, and ISI Web of Science databases.  Proportion meta-analyses were carried out to estimate the pooled sensitivity of magnetically controlled CE in the detection of GI lesions.  Among the 3,026 studies that were initially evaluated, 7 studies were finally included, with a total of 916 patients and 745 gastric lesions.  The mean CE examination time was 21.92 ± 8.87 mins.  The pooled overall sensitivity of magnetically controlled CE was 87 % [95 % CI: 84 % to 89 %].  Subgroup analysis showed that the sensitivity of identifying gastric ulcers was 82 % (95 % CI: 71 % to 89 %), gastric polyps was 82 % (95 % CI: 76 % to 87 %), and gastric erosions was 95 % (95 % CI: 86 % to 98 %).  In general, magnetically controlled CE was well-tolerated by the participants with minimal AEs.  The authors concluded that magnetically controlled CE demonstrated an acceptable sensitivity of identifying gastric lesions.  Moreover, these researchers stated that further prospective, comparative studies are needed to evaluate the risks and benefits of this new technique, as well as to determine its role as a replacement for conventional gastroscopy.

Wang et al (2023) stated that magnetically controlled capsule endoscopy (MCCE) is a non-invasive, painless, comfortable, and safe equipment to diagnose GI diseases (GID), partially overcoming the shortcomings of conventional endoscopy and WCE.  With advancements in technology, the main technical parameters of MCCE have continuously been improved, and MCCE has become more intelligent.  In a systematic review, these investigators examined the research progress of MCCE and artificial intelligence (AI) in the diagnosis and treatment of GID.  They carried out a systematic search of PubMed and Embase for published studies on GID detection of MCCE, physical factors related to MCCE imaging quality, the use of AI in aiding MCCE, and its additional functions.  These researchers reviewed the included studies, extracted relevant data, and made comparisons.  MCCE was confirmed to have the same performance as conventional gastroscopy and WCE in detecting common GID, while it lacks research in detecting early gastric cancer (EGC).  The body position and cleanliness of the GI tract are the main factors affecting imaging quality.  The uses of AI in screening intestinal diseases have been comprehensive, while in the detection of common gastric diseases such as ulcers, it has been developed.  MCCE can perform some additional functions, such as observations of drug behavior in the stomach and drug damage to the gastric mucosa.  In addition, it can be improved to perform a biopsy.  The authors concluded that this comprehensive review showed that the MCCE technology has made great progress; however, studies on GID detection and treatment by MCCE were in the primary stage.  Moreover, these researchers stated that it is important to note that the findings regarding the performance of MCCE in detecting common GI lesions were still based on limited evidence; thus, further investigations are needed to confirm the performance of MCCE.

Detection of Occult Primary

Furnari et al (2017) stated that neuroendocrine tumors (NETs) are a heterogeneous group of neoplasms with unclear etiology that may show functioning or non-functioning features.  Primary tumor localization often requires integrated imaging.  The European Neuroendocrine Tumors Society (ENETS) guidelines proposed (WCE as a possible diagnostic tool for NETs, if intestinal origin is suspected.  However, its impact on therapeutic management is debated.  These investigators examined the yield of WCE in detecting intestinal primary tumors in patients showing liver NET metastases when 1st-line investigations are inconclusive.  A total of 24 patients with a histological diagnosis of metastatic NET from liver biopsy and no evidence of primary lesions at 1st-line investigations were prospectively studied in an ENETS-certified tertiary care center.  WCE was requested before explorative laparotomy and intra-operative ultrasound.  The diagnostic yield of WCE was compared to the surgical exploration.  A total of 16 subjects underwent surgery; 11/16 had positive WCE identifying 16 bulging lesions.  Mini laparotomy found 13 NETs in 11/16 patients (9 small bowel, 3 pancreas, 1 bile ducts).  Agreement between WCE and laparotomy was recorded in 9 patients (sensitivity = 75 %; specificity = 37.5 %; PPV = 55 %; NPV = 60 %).  Correspondence assessed per-lesions produced similar results (sensitivity = 70 %; specificity = 25 %; PPV = 44 %; NPV = 50 %). No capsule retentions were recorded.  The authors concluded that WCE is not indicated as 2nd-line investigation for patients with gastro-entero-pancreatic NETs.  In the setting of a referral center, it might provide additional information when conventional investigations are inconclusive regarding the primary site.

Alexandraki et al (2019) stated that neuroendocrine neoplasms (NENs) of unknown primary origin (UPO-NENs) are advanced neoplasms constituting 11 % to 22 % of all NENs that by definition their primary tissue of origin has not been identified with standard diagnostic work-up.  Delineating the primary site of origin of UPO-NENs has important implications for selecting the appropriate treatment and overall prognosis.  The small bowel, followed by the lung and pancreas are the most prevalent primary sites of origin of UPO-NENs that are uncovered during an extensive and prolonged diagnostic work-up; however, a number of UPO-NENs may still remain occult even after prolonged follow-up.  A number of diagnostic algorithms that incorporate histopathological, molecular, imaging (either morphological or functional imaging), and serum biomarkers can help to identify the primary tumor origin.  It is expected that advances in these fields will help reduce significantly the number of UPO-NENs.

Furthermore, NCCN guidelines on “Melanoma: Cutaneous” (Version: 3.2022) and “Occult primary: Cancer of unknown primary [CUP]” (Version 1.2022) have no recommendations for use of capsule endoscopy.

Pan-Intestinal (Pan-Enteric) Capsule Endoscopy for Evaluation of Crohn's Disease / Ulcerative Colitis

Bruining et al (2020) stated that Crohn's disease (CD) diagnosis and monitoring remains a great clinical challenge and often requires multiple testing modalities.  Assessing CD activity in the entire GI tract using a pan-enteric capsule endoscopy (PCE) system could be used as an alternative to colonoscopy and cross-sectional imaging.  In a prospective, multi-center study, these researchers examined the accuracy and safety of PCE in CD as compared with ileo-colonoscopy and/or magnetic resonance enterography (MRE).  This trial was carried out in subjects with established CD.  Individuals with proven small bowel patency underwent a standardized bowel preparation, followed by CE ingestion and ileo-colonoscopy either the same or following day.  Ileo-colonoscopy, MRE, and CE interpretations were performed by blinded central readers using validated scoring systems.  The primary endpoint was the overall sensitivity of CE versus MRE and/or ileo-colonoscopy in CD subjects.  This study enrolled 158 subjects from 21 sites in the U.S., Austria, and Israel.  Of those, 99 were included in the analysis.  Imaging modality scores indicated none-to-mild inflammation in the proximal small bowel and colon, but discrepant levels of inflammation in the terminal ileum (TI).  Overall sensitivity for active enteric inflammation (CE versus MRE and/or ileo-colonoscopy) was 94 % versus 100 % (p = 0.125) and specificity was 74 % versus 22 % (p = 0.001).  Sensitivity of CE was superior to MRE for enteric inflammation in the proximal small bowel (97 % versus 71 %, p = 0.021), and similar to MRE and/or ileo-colonoscopy in the terminal ileum and colon (p = 0.500 to 0.625).  There were 7 serious AEs of which 3 were related to the CE device.  The authors concluded that the findings of this study showed the potential of PCE to provide accurate CD assessments; they noted that PCE should be considered for evaluation of in non-stricturing CD.

The authors stated that this study had several drawbacks.  First, the results applied to patients with CD with mild non-stricturing disease.  Bias could exist for the reported performance estimates for all modalities given the mild disease (e.g., patients demonstrating strictures with inflammation were excluded by MRE before CE assessment).  Moderate-to-severe disease was only identified by MRE magnetic resonance index of activity (MaRIA) in the TI.  Second, the low specificity (high discrepancy rate) for MRE in the TI was reported.  This could be the result of altered perfusion (previous ileocecal surgery, n = 22), high MaRIA scores in the absence of convincing wall thickening (e.g., potentially due to near-field artefacts), differences in the identification of ulcerations or intramural edema, and the relatively small number of patients without TI inflammation.  The concept of intramural disease has now been well described with MRE and CT enterography (CTE), and isolated findings of intramural inflammation by MRE with normal ileoscopic findings can exist.  However, in this study, the reference standard in cases of discrepant numerical indices required radiologist review of MRE images, and acknowledgement that the images and preponderance of evidence indicated inflammation was absent.  The use of a consensus panel as the reference standard when discrepancies occurred was to take advantage of the diagnostic advantages of each modality (e.g., mucosal visualization for ileo-colonoscopy, display of the intestinal wall, and peri-enteric tissues for MRE), but should also be acknowledged as a potential limitation given questions regarding validity and reproducibility of this methodology.  Finally, the potential non-specific features of CD on CE and other testing modalities should be noted as a potential limitation.

Valdivia et al (2021) noted that in recent years, PCE with double-headed capsules has been employed to carry out complete, single-sitting exploration of both small bowel and colon in different clinical conditions.  Double-headed capsules for colonic examination (CCE) have been exploited first in this setting, followed by newer generations of capsules (i.e., PillCam Crohn, PCC) specifically engineered for this purpose.  In a systematic review, these researchers examined the feasibility of PCE.  They carried out a comprehensive literature search to identify papers in which CE was specifically used for a PCE of the GI tract.  Data on CE, bowel preparation regimen, rate of cleanliness and completeness, and data on transit times were analyzed.  The primary outcome was to evaluate the feasibility of a whole-gut exploration with CE.  A total of 16 studies including 915 CE procedures with CCE1 (n = 134), CCE2 (n = 357) and PCC (n = 424) were included; 13/16 studies were conducted in the setting of CD.  Cleanliness and completeness rates were acceptable in all studies, ranging from 63.9 % and 68.6 % to 100 %, respectively.  The authors concluded that PCE is a feasible technique, although further structured studies are needed to examine its full potential.

Furthermore, UpToDate reviews on “Clinical manifestations, diagnosis, and prognosis of Crohn disease in adults” (Peppercorn and Kane, 2022) and “Overview of the management of Crohn disease in children and adolescents” (Zitomersky and Bousvaros, 2022) do not mention pan-intestinal (pan-enteric) capsule endoscopy as a management option.

Tamilarasan et al (2022) noted that CE is a non-invasive diagnostic modality enabling real-time video imaging of the GI mucosa.  Pan-enteric CE (PCE) is now able to thoroughly evaluate the entire GI tract, including for IBD.  In a systematic review and meta-analysis, these investigators examined the diagnostic accuracy of PCEs in IBD.  They comprehensively searched electronic databases (Medline, SCOPUS, Embase, and Cochrane Central Register of Controlled Trials) for studies comparing the diagnostic accuracy of PCE with endoscopic evaluation, intestinal ultrasound (US) or MRE.  Data were analyzed by calculating forest plots and the use of the I2 statistic for heterogeneity.  A total of 14 studies were identified, with 7 studies evaluating PCE diagnostic yield in CD and 7 studies in UC.  In CD, there was a trend to superiority of PCE over MRE and colonoscopy with a pooled odds ratio (OR) of 1.25 (95 % CI: 0.85 % to 1.86 %) for the detection of CD.  This translated to an increased diagnostic yield of 5 % and 7 % for PCE compared with MRE and colonoscopy, respectively.  PCEs had a diagnostic sensitivity for the detection of UC of 93.8 % (95 % CI: 87.6 % to 97.0 %) and a specificity of 69.8 % (95 % CI: 38.2 % to 89.6 %).  The authors concluded that PCEs have a comparable diagnostic yield to colonoscopy and MRE in CD.  The major difficulty remains standardization of PCE scoring systems and the lack of transmural assessment.  In UC, PCE has an excellent diagnostic sensitivity and PPV; however, there are limitations to its use including the lack of histologic assessment and poor specificity.  Moreover, these researchers stated that these findings supported the use of PCE in the diagnosis and monitoring of CD, potentially as a single diagnostic or monitoring tool given its pan‐enteric assessment capacity.  For UC, PCE has excellent diagnostic sensitivity but is not suggested to be a replacement for fecal calprotectin or colonoscopy.  These investigators stated that PCE may have a role in UC in certain patient groups, and its use should be tailored to individual clinical circumstances.

Artificial Intelligence in Colon Capsule Endoscopy

Moen et al (2022) noted that the applicability of CCE in daily practice is limited by the accompanying labor-intensive reviewing time and the risk of inter-observer variability.  Automated reviewing of CCE images using artificial intelligence (AI) could be timesaving while providing an objective and reproducible outcome.  In a systematic review, these investigators provided an overview of the available evidence on AI for reviewing colonic mucosa by CCE and to examine the necessary action points for its use in clinical practice.  They carried out a systematic literature search of literature published up to January 2022 using Embase, Web of Science, OVID Medline and Cochrane CENTRAL.  Studies reporting on the use of AI to review 2nd-generation CCE (CCE-2) colonic images were included.  A total of 1,017 studies were evaluated for eligibility, of which 9 were included.  Two studies reported on computed bowel cleansing assessment, 5 studies reported on computed polyp or colorectal neoplasia detection and 2 studies reported on other implications.  Overall, the sensitivity of the proposed AI models was 86.5 % to 95.5 % for bowel cleansing and 47.4 % to 98.1 % for the detection of polyps and colorectal neoplasia.  Two studies performed per-lesion analysis, in addition to per-frame analysis, which improved the sensitivity of polyp or colorectal neoplasia detection to 81.3 % to 98.1 %.  By employing a convolutional neural network (CNN), the highest sensitivity of 98.1 % for polyp detection was found.  The authors concluded that the use of AI for reviewing CCE-2 images is promising.  The highest sensitivity of 98.1 % for polyp detection was achieved by deep learning (DL) with a CNN.  Moreover, these researchers stated that actual implementation of AI for reviewing CCE-2 colonic images is a crucial step in the applicability of CCE in daily clinical practice.  Future studies should preferably focus on CNN, because of its high potential for reaching human-like performance.  In order to reach its implementation, several steps need to be taken.  CNN algorithms need to be optimized and tested with more data, possibly requiring the set-up of a large, international CCE database.  To ensure adequate evaluation of the added value of the AI method, studies should always report the version of the capsule used and the accuracy of their models in terms of sensitivity and specificity.  Furthermore, studies should preferably only use the results from expert CCE readers to test the performance of their AI methods, since the concluded sensitivities and specificities represent the ability of the AI models to reach the same performance levels as these readers.  Besides CNN, which requires an adequate number of coloscopy images, synthetic samples can also be used as AI methods.  Finally, when these gaps and barriers have been overcome, prospective clinical trials have to confirm the accuracy of the optimized CNN models.

Martins et al (2023) stated that device-assisted enteroscopy (DAE) has a significant role in approaching enteric lesions.  Endoscopic observation of ulcers or erosions is frequent and can be associated with many nosological entities, namely CD.  Although the use of AI is growing exponentially in various imaged-based gastroenterology procedures, there is still a lack of evidence of the AI technical feasibility and clinical applicability of DAE.  These researchers developed and tested a multi-brand CNN-based algorithm for automatically detecting ulcers and erosions in DAE.  A retrospective, single-center study was carried out for the development of a CNN, based on a total of 250 DAE examinations.  A total of 6,772 images were used, of which 678 were considered ulcers or erosions after double-validation.  Data were divided into a training and a validation set, the latter being used for the performance assessment of the model.  The primary outcome measures were sensitivity, specificity, accuracy, PPV, NPV, and an area under the curve precision-recall curve (AUC-PR).  Sensitivity, specificity, PPV, and NPV were respectively 88.5 %, 99.7 %, 96.4 %, and 98.9 %.  The algorithm's accuracy was 98.7 %.  The AUC-PR was 1.00.  The CNN processed 293.6 frames/second, enabling AI live application in a real-life clinical setting in DAE.  The authors concluded that this was the 1st study regarding the automatic multi-brand pan-endoscopic detection of ulcers and erosions throughout the digestive tract during DAE, overcoming a relevant interoperability challenge.  These findings highlighted that using a CNN to detect this type of lesion is associated with high overall accuracy.  The development of binary CNN for automatically detecting clinically relevant endoscopic findings and assessing endoscopic inflammatory activity are relevant steps toward AI application in digestive endoscopy, particularly for pan-endoscopic evaluation.  These researchers stated that this study was a continuation of their previous foundational studies regarding AI and DAE.  It suggested that automatic detection of ulcers and erosions using a CNN during DAE is associated with high overall accuracy.  In addition, implementation of this CNN during DAE might contribute to a higher detection rate of ulcerative lesions, given by real-time feedback during the procedure, which in turn may contribute to lower inter-observer variability and a lower false-negative rate.

The authors stated that this study had several drawbacks.  On the one hand, this was a single-center study based on retrospective data, which might be associated with a selection bias.  On the other hand, the number of frames used to develop and test this model was relatively small; therefore, compromising the external validity of these findings.  Although these results look promising, the risk of over-fitting should not be omitted; thus, a final conclusion cannot be reached regarding its applicability in everyday situations.  These investigators stated that multi-center studies are needed, with more extensive and prospective data collection, to ensure appropriate dataset variability.  Moreover, it should be noted that anatomopathological diagnosis of CD remains challenging, since there is a higher proportion of false-negatives when biopsies are taken during enteroscopy.  Clinical and non-endoscopic imaging techniques remain crucial, namely when a strong suspicion exists, despite the microscopic result being negative.  AI applications, namely AI-guided biopsies, during DAE, may increase diagnostic yields of this technique, especially for CD.

Lei et al (2023) stated that AI applications have become widely popular across the healthcare ecosystem; CCE was adopted in the NHS England pilot project following the recent COVID pandemic's impact.  It demonstrated its capability to relieve the national backlog in endoscopy.  As a result, AI-assisted colon capsule video analysis has become gastroenterology's most active research area.  However, with rapid AI advances, mastering these complex machine learning (ML) concepts remains challenging for healthcare professionals.  This forms a barrier for clinicians to take on this new technology and embrace the new era of big data.  These researchers aimed to bridge the knowledge gap between the current CCE system and the future, fully integrated AI system.  The authors concluded that with the current ethical uncertainty and challenges, future randomized, multi-center studies, which validate AI models, should focus on answering the fundamental question of whether AI models can enhance physician performance safely and reliably.  In the end, a robust multi-disciplinary collaboration among physicians, computer scientists, and entrepreneurs is needed to promote AI’s clinical use in medical practice.

Video Capsule Endoscopy for Diagnosis of Gastro-Intestinal Graft-Versus-Host Disease

Varkey et al (2023) stated that the GI tract is the 2nd most involved organ for graft-versus-host disease (GVHD) where involvement of the small intestine is present in 50 % of the cases.  Thus, the use of a non-invasive investigation i.e., video capsule endoscopy (VCE) appeared ideal in the diagnostic work-up; however, this has not been systematically evaluated.  In a systematic review, these investigators examined the safety and effectiveness of VCE, in comparison with conventional endoscopy in patients who received hematopoietic stem cell transplantation (HSCT).  Databases searched were PubMed, Scopus, Embase, and Cochrane CENTRAL.  All databases were searched from their inception date until June 17, 2022.  The search identified 792 publications, of which 8 studies were included in the final analysis comprising of 232 unique patients.  Effectiveness was calculated in comparison with the golden standard i.e., histology.  Risk of bias assessment was carried out using the Quality Assessment of Diagnostic Accuracy Studies-2 tool.  The pooled sensitivity was higher for VCE at 0.77 (95 % CI: 0.60 to 0.89) compared to conventional endoscopy 0.62 (95 % CI: 0.47 to 0.75); however, the difference was not statistically significant (p = 0.155, Q = 2.02).  Similarly, the pooled specificity was higher for VCE at 0.68 (95 % CI: 0.46 to 0.84) than for conventional endoscopy at 0.58 (95 % CI: 0.40 to 0.74) but not statistically significant (p = 0.457, Q = 0.55).  Moreover, concern for AEs such as intestinal obstruction or perforation was not justified since none of the capsules was retained in the small bowel and no perforations occurred in relation to VCE.  The authors concluded that this systematic review of 8 studies consisting of 232 patients undergoing VCE following HSCT refuted the notion that CE is unsafe and leads to obstructive symptoms within this patient cohort.  Moreover, when aptly performed, this approach may yield useful information although its place in diagnosing or dismissing GVHD is yet to be established and requires further studies.  Furthermore, these investigators stated that this analysis was exclusively focused on GVHD, it is still necessary to establish the value of VCE in diagnosing and managing other GI complications following HSCT, especially viral cytomegalovirus (CMV) enteritis, which represent an important differential and a frequent complication following HSCT.  As the viral load surveillance via polymerase chain reaction (PCR) may be used as an adjunct, the role of VCE is perhaps to monitor the treatment response and as a complimentary measure to examine the extent of the disease during the initial diagnosis.  Moreover, early recognition of severe gut GVHD may also help select adequate treatment promptly.

These investigators stated that this review had several drawbacks.  First, this analysis had the shortcomings typical for many systematic reviews, namely a low number of available reports and a small sample size for most of the included studies.  Second, many studies were retrospective (50 % of the studies and 78 % (183/232) of the patients) and endoscopists were likely not blinded to the clinical history; however, VCE readings appeared to have been carried out before the retrieval of histopathology reports.  Third, VCE examinations were mainly focused to the small intestine whereas the biopsies were sampled from either the proximal GI tract, the colon or both.  Fourth, the VCE patients consisted of a selected patient cohort that further lowered the overall study quality.

Artificial Intelligence in Capsule Endoscopy for the Management of Inflammatory Bowel Disease

Ferreira et al (2022) noted that capsule endoscopy is a central element in the management of patients with suspected or known CD.  In 2017, PillCam Crohn's Capsule was introduced and demonstrated to have greater accuracy in the evaluation of extension of disease in these patients.  Artificial intelligence (AI) is expected to enhance the diagnostic accuracy of capsule endoscopy.  In a retrospective, multi-center, pilot study, these researchers developed an AI algorithm for the automatic detection of ulcers and erosions of the small intestine and colon in PillCam Crohn's Capsule images.  A total of 8,085 PillCam Crohn's Capsule images were extracted between 2017 and 2020, comprising 2,855 images of ulcers and 1,975 images of erosions; the remaining images showed normal enteric and colonic mucosa.  This pool of images was subsequently split into training and validation data-sets.  The performance of the network was subsequently evaluated in an independent test set.  The model had an overall sensitivity and specificity of 90.0 % and 96.0 %, respectively.  The precision and accuracy of this model were 97.1 % and 92.4 %, respectively.  In particular, the algorithm detected ulcers with a sensitivity of 83 % and specificity of 98 %, and erosions with sensitivity and specificity of 91 % and 93 %, respectively.  The authors concluded that a deep learning (DL) model capable of automatically detecting ulcers and erosions in PillCam Crohn's Capsule images was developed for the first time.  These investigators stated that these findings paved the way for the development of automatic systems for detection of clinically significant lesions, optimizing the diagnostic performance and efficiency of monitoring CD activity.  The authors stated that this study had 2 main drawbacks.  First, it was a retrospective proof-of-concept (POC) study including a small number of PCC examinations; thus, prospective studies with larger numbers of PCC examinations are needed before this model could be applied to clinical practice.  Second, this tool was tested in still frames; therefore, assessment of its performance using full-length videos is needed before clinical application of these tools.

Pal et al (2024) stated that in spite of rapid growth of AI in digestive endoscopy in lesion detection and characterization, the role of AI in IBD endoscopy is not clearly defined.  In a systematic review, these investigators examined the role of AI in IBD endoscopy and identifying future research areas.  They searched the PubMed and Embase database using keywords ("artificial intelligence" OR "machine learning" OR "computer-aided" OR "convolutional neural network") AND ("inflammatory bowel disease" OR "ulcerative colitis" OR "Crohn's") AND ("endoscopy" or "colonoscopy" or "capsule endoscopy" or "device assisted enteroscopy") between 1975 and September 2023 and identified 62 original articles for detailed review.  Review articles, consensus guidelines, case reports/series, editorials, letter to the editor, non-peer-reviewed pre-prints and conference abstracts were excluded.  The quality of the included studies was assessed using the MI-CLAIM checklist.  The accuracy of AI models (25 studies) to evaluate ulcerative colitis (UC) endoscopic activity ranged between 86.54 % and 94.5 %.  AI-assisted capsule endoscopy reading (12 studies) substantially reduced analyzable images and reading time with excellent accuracy (90.5 % to 99.9 %). AI-assisted analysis of colonoscopic images can aid in differentiating IBD from non-IBD, UC from non-UC, and UC from Crohn's disease (CD) (3 studies) with 72.1 %, 98.3 % and greater than 90 % accuracy, respectively.  AI models based on non-invasive clinical and radiologic parameters could predict endoscopic activity (3 studies).  AI-assisted virtual chromoendoscopy (4 studies) could predict histologic remission and long-term outcomes.  Computer-assisted detection (CADe) of dysplasia (2 studies) is feasible along with AI-based differentiation of high-grade from low-grade IBD neoplasia (79 % accuracy).  AI was effective in linking electronic medical record data (2 studies) with colonoscopic videos to facilitate widespread machine learning (ML).  The authors concluded that AI-assisted IBD endoscopy has the potential to impact clinical management by automated detection and characterization of endoscopic lesions.  Moreover, these researchers stated that prospective, large, multi-center studies and commercially available IBD-specific endoscopic AI algorithms are needed.

Brodersen et al (2024) noted that pan-enteric capsule endoscopy [PCE] is a highly sensitive but time-consuming tool for detecting pathology; AI algorithms might offer a possibility to aid in reviewing and reducing the analysis time of PCE.  In a prospective, blinded, multi-center study, these investigators examined the agreement between PCE assessments aided by AI technology and standard evaluations, in patients suspected of CD.  This trial included patients suspected of CD; they were processed by the deep learning solution AXARO (Augmented Endoscopy, Paris, France).  Based on the image output, 2 observers classified the patient's PCE as normal or suggestive of CD, UC, or cancer.  The primary outcome was per-patient sensitivities and specificities for detecting CD and IBD.  Complete reading of PCE served as the reference standard.  A total of 131 patients' PCEs were analyzed, with a median recording time of 303 mins.  The AXARO frame-work reduced output to a median of 470 images (2.1 %) per patient, and the pooled median review time was 3.2 mins/patient.  For detecting CD, the observers had a sensitivity of 96 % and 92 %, and a specificity of 93 % and 90 %, respectively.  For the detection of IBD, both observers had a sensitivity of 97 % and had a specificity of 91 % and 90 %, respectively.  The NPV was 95 % for CD and 97 % for IBD.  The authors concluded that using the AXARO frame-work lowered the initial review time substantially while maintaining high diagnostic accuracy; suggesting its use as a rapid tool to rule out IBD in PCEs of patients suspected of CD.

References