Introduction

Celiac disease is a common multisystem autoimmune disorder affecting approximately 1% of the population, which is triggered by the ingestion of gluten (a protein component of wheat, rye, and barley) in genetically susceptible individuals. Screening for celiac disease is available via serologic testing, with excellent reported sensitivity and specificity. Antibodies available for celiac disease screening include anti-tissue transglutaminase (TTG) immunoglobulin A (IgA) antibody, anti-endomysial (EMA) IgA antibody, and the more recently developed anti–deamidated gliadin peptide IgA/IgG antibodies. The gold standard for the diagnosis of celiac disease still remains small intestinal biopsy performed during upper endoscopy. Features of celiac disease appreciated on conventional upper endoscopy include a mosaic mucosal pattern, scalloping, villous atrophy, and flattening of folds. In patients with partial villous atrophy, however, the small intestine may seem endoscopically normal and biopsies should be obtained regardless of the appearance if celiac disease is suspected. Histologic features suggestive of celiac disease include increased intraepithelial lymphocytes, crypt hyperplasia, and villous atrophy. Marsh developed an initial grading system to classify the histologic changes associated with celiac disease and this was later refined by Oberhuber. These histologic features are not unique to celiac disease and can be found in other conditions, such as tropical sprue and autoimmune enteropathy.

Small bowel biopsy during upper endoscopy has several limitations. Upper endoscopy is an invasive procedure and sedation is routinely administered. Patients often require an escort home and miss work. During routine upper endoscopy, a very small portion of the small bowel is examined. Celiac disease may be patchy, and biopsies may fail to sample areas demonstrating disease. Some physicians may not obtain the suggested 4 to 6 biopsies during upper endoscopy when evaluating patients for suspected celiac disease. Interpretation of histology may also be problematic because pathologists in some settings, particularly in commercial laboratories or community hospitals, may fail to recognize the features of celiac disease. Histologic samples may also be poorly oriented, and the interpretation may be hindered.

Video capsule endoscopy: technology and administration

Video capsule endoscopy (VCE) provides an alternative means to conventional upper endoscopy to visualize the small intestine. VCE was first introduced in 2001, and there are several systems currently available worldwide: PillCam SB2, Given Imaging, Yoqneam, Israel; Endo Capsule, Olympus America Inc, Center Valley, Pennsylvania; OMOM, Jinshan Science and Technology, Chongqing, China; and MiroCam, IntroMedic, Seoul, South Korea. Currently, the PillCam SB2 and Endo Capsule endoscopes are available in the United States. The capsule endoscopes typically measure 11 × 26 mm and contain a battery, lens, light-emitting diodes (LEDs), and transmitter. The camera lens is surrounded by 6 LEDs that illuminate the mucosal surface of the gastrointestinal tract during the capsule passage. The capsules typically obtain 2 pictures every second and transmit images to a recording device worn by patients. The study is typically 8 hours, but an extended 12-hour capsule has recently been introduced. When the battery on the capsule endoscope is discharged, the study is complete. Images are subsequently downloaded to a workstation and reviewed. A real-time viewer is also available to assess the capsule location while the study is being conducted.

Before VCE, bowel purgatives or other preparations are not routinely recommended. Patients are often instructed to ingest a clear-liquid diet for 1 day and fast after midnight the night before the examination. A prospective randomized trial of 150 patients demonstrated similar VCE completion rates, yield, and image quality with and without the use of purgatives and prokinetics. Conversely, a recent meta-analysis suggested that purgatives before VCE resulted in enhanced mucosal visualization and slightly increased the yield when compared with clear-liquid preparations. In VCE studies of patients with celiac disease, the use of purgatives may be helpful to improve the visualization of the mucosa. Simethicone reduces intestinal bubbles and enhances mucosal visualization during VCE. In patients with celiac disease whereby mucosal visualization is paramount, the authors’ center has been using a 2 L PEG–based bowel purgative preparation the night before the examination and a 60-mg simethicone dose administered with the capsule endoscope.

Video capsule endoscopy: technology and administration

Video capsule endoscopy (VCE) provides an alternative means to conventional upper endoscopy to visualize the small intestine. VCE was first introduced in 2001, and there are several systems currently available worldwide: PillCam SB2, Given Imaging, Yoqneam, Israel; Endo Capsule, Olympus America Inc, Center Valley, Pennsylvania; OMOM, Jinshan Science and Technology, Chongqing, China; and MiroCam, IntroMedic, Seoul, South Korea. Currently, the PillCam SB2 and Endo Capsule endoscopes are available in the United States. The capsule endoscopes typically measure 11 × 26 mm and contain a battery, lens, light-emitting diodes (LEDs), and transmitter. The camera lens is surrounded by 6 LEDs that illuminate the mucosal surface of the gastrointestinal tract during the capsule passage. The capsules typically obtain 2 pictures every second and transmit images to a recording device worn by patients. The study is typically 8 hours, but an extended 12-hour capsule has recently been introduced. When the battery on the capsule endoscope is discharged, the study is complete. Images are subsequently downloaded to a workstation and reviewed. A real-time viewer is also available to assess the capsule location while the study is being conducted.

Before VCE, bowel purgatives or other preparations are not routinely recommended. Patients are often instructed to ingest a clear-liquid diet for 1 day and fast after midnight the night before the examination. A prospective randomized trial of 150 patients demonstrated similar VCE completion rates, yield, and image quality with and without the use of purgatives and prokinetics. Conversely, a recent meta-analysis suggested that purgatives before VCE resulted in enhanced mucosal visualization and slightly increased the yield when compared with clear-liquid preparations. In VCE studies of patients with celiac disease, the use of purgatives may be helpful to improve the visualization of the mucosa. Simethicone reduces intestinal bubbles and enhances mucosal visualization during VCE. In patients with celiac disease whereby mucosal visualization is paramount, the authors’ center has been using a 2 L PEG–based bowel purgative preparation the night before the examination and a 60-mg simethicone dose administered with the capsule endoscope.

Indications for VCE in celiac disease

According to an international consensus conference, VCE can be considered as an imaging modality in cases of known or suspected celiac disease in select situations ( Box 1 ). VCE can be performed when a person is unable or unwilling to have a conventional upper endoscopy. Examples include patients with cardiopulmonary instability or history of bleeding disorders. VCE may also be performed in cases of positive serology (TTG or EMA) and normal duodenal histology to visualize more distal portions of the small intestine for features of celiac disease. In patients with known celiac disease who develop warning signs, such as weight loss, bleeding, and/or abdominal pain, VCE can be used to evaluate the intestine. These concerning symptoms may occur on the initial presentation of celiac disease or after an extended period on a gluten-free diet. If there is concern for refractory celiac disease or malignancy, VCE is usually performed in combination with conventional upper endoscopy, colonoscopy, and often radiographic enterography. These studies assist in excluding malignancy, particularly adenocarcinoma and lymphoma. In patients with refractory celiac disease type II whereby a clonal population of cells is present, particular concern exists for enteropathy-associated T-cell lymphoma (EATL). If an abnormal area is detected on VCE, enteroscopy using a device-assisted approach can be performed to sample tissue.

Role of VCE in celiac disease

The use of VCE in celiac disease is an area of active research. It is an appealing imaging modality for patients with celiac disease because it is a relatively noninvasive and safe procedure. Because capsule endoscopes are unable to obtain biopsies, and small bowel histology remains the gold standard for the diagnosis of celiac disease, VCE serves a complementary role to conventional endoscopy. Features of celiac disease visualized during VCE are similar to those seen during conventional upper endoscopy: scalloping of folds, mosaic appearance, fissures, flat mucosa, and stacking of folds ( Fig. 1 ).

Features of celiac disease seen on video capsule endoscopy. ( A ) Mosaic pattern, mucosal fissures; ( B ) scalloping; ( C ) villous atrophy.

VCE examines the whole intestine and can be used to determine the extent of enteropathy. Interestingly, in a study of 38 patients with untreated biopsy-proven celiac disease, VCE detected villous atrophy in 92% of patients, but the extent of enteropathy did not correlate with clinical symptoms. In this study by Murray and colleagues, 59% of patients with untreated celiac disease had extensive enteropathy, 32% had duodenal enteropathy, and less than 1% had jejunal enteropathy alone.

The sensitivity and specificity for detecting features of celiac disease has varied in several clinical studies when compared with the gold standard of duodenal histology. In an initial multicenter trial by Rondonotti and colleagues of 43 patients with suspected celiac disease who underwent both upper endoscopy and VCE, VCE had a sensitivity of 87.5%, specificity of 90.9%, positive predictive value of 96.5%, and negative predictive value of 71.4% for the detection of villous atrophy as compared with duodenal histology. In a study of 21 patients with suspected celiac disease (positive EMA) and 23 control patients having both upper endoscopy and VCE, 17 out of 20 patients with suspected celiac disease with villous atrophy on biopsy had abnormal VCE studies. In this study by Hopper, the sensitivity, specificity, positive predictive value, and negative predictive value of VCE as compared with duodenal histology was 85.0%, 100%, 100%, and 88.9%, respectively. When the small bowel mucosa was evaluated in VCE studies of patients with celiac disease, Crohn disease, and irritable bowel syndrome by 3 blind observers, the sensitivity ranged from 90% to 95%, whereas the specificity was only 63.6%. In this study, VCE had only moderate agreement with histologic pattern with a Kappa statistic for observer 1, 2, and 3 of 0.45, 0.49, and 0.51, respectively. In a recent meta-analysis including 166 individuals using VCE in the diagnosis of celiac disease, the overall pooled VCE sensitivity was 89% (95% confidence interval 89%–98%) and the specificity was 95% (95% confidence interval 82%–94%). If suspicion for celiac disease remains following a normal VCE study, upper endoscopy with biopsy should be performed.

The role of VCE in monitoring patients with celiac disease is less clear. Up to 30% of patients with celiac disease may fail to improve on a gluten-free diet alone. In these patients, a systematic approach is used to exclude causes, such as gluten contamination; bacterial overgrowth; fructose/lactose intolerance; microscopic colitis; pancreatic exocrine function; refractory celiac disease; and associated malignancy, such as EATL. VCE has been suggested in the evaluation of these patients combined with dietary counseling, breath testing, upper endoscopy, colonoscopy, radiologic imaging, and enteroscopy when indicated. There is still limited evidence, however, to support the use of VCE in patients with nonresponsive celiac disease. VCE was not included in a recent proposed algorithm for patients with celiac disease with persistent symptoms on a gluten-free diet, but the investigators currently use VCE in their evaluation of patients with suspected refractory celiac disease. In a blinded pilot comparison study of 19 patients with unresponsive celiac disease, 53% had normal VCE studies, 47% demonstrated features of celiac disease, 11% (2 patients) had ulcers, but no small bowel tumors were found. In a study performed at the authors’ center, 47 patients were examined with features of complicated celiac disease: abdominal pain, weight loss, diarrhea, iron deficiency anemia, occult blood positive stool, long-standing untreated celiac disease, or a history of small bowel cancer /adenoma. In this study, VCE detected lesions suspicious for lymphoma in 2 patients with refractory celiac disease type II, a case of carcinoma, a submucosal mass, a stricture, and a case of intussusception. In this study, ulcerations were found in 45% of subjects and villous atrophy was detected in 68% of subjects.

The sensitivity for detecting villous atrophy may vary in patients with treated celiac disease. In an analysis of VCE studies of patients with nonresponsive celiac disease reviewed by a single reader with expertise in celiac disease, 31% of patients had villous atrophy on VCE, but the sensitivity of VCE for detecting villous atrophy was only 56%, whereas specificity was 85%. Nonresponsive celiac disease was defined as persistent or recurrent symptoms, including diarrhea, weight loss, and abdominal pain, after at least 6 months on a gluten-free diet. In this study by Atlas, the presence of ulcerations was associated with the use of nonsteroidal medications; but, importantly, a case of ulcerative jejunitis or lymphoma was detected in a patient with refractory celiac disease type II and adenocarcinomas were reported. There was also weak agreement of capsule endoscopy and histology in the patients with nonresponsive celiac disease (κ = 0.44).

The literature concerning VCE in patients with refractory celiac disease is particularly limited. Daum and colleagues examined 14 patients with refractory celiac disease (7 type I, 7 type II) who had VCE and abdominal imaging performed via computed tomography or magnetic resonance tomography. No abnormalities were detected in the patients with refractory celiac disease type 1 via any method, but VCE and radiologic imaging each detected signs of ulcerative jejunitis or lymphoma in separate patients with refractory celiac disease type II. Not all erosions or ulcerations represent significant abnormalities because these can be found in healthy patients, patients with treated celiac disease, and patients with nonresponsive celiac disease. Although there is limited evidence to support its use, VCE should be considered in patients with refractory celiac disease type II, but it is likely of limited yield in patients with refractory celiac disease type I.

Finally, an international consensus conference recommended VCE as an imaging modality to examine the distal small intestine for features of celiac disease in cases of positive TTG antibodies and normal duodenal histology, but there is little data to support its use. Capsule endoscopy has not been shown to have any added benefit for detecting villous atrophy in patients with positive celiac serology and normal duodenal histology. In this small prospective single-center study by Lidums and colleagues, 8 patients with positive TTG or EMA and normal duodenal histology had normal VCE studies. In addition, only 9 out of 14 patients (64%) had a complete evaluation of the small intestine, whereas the remainder of the patients had incomplete examinations.