Between March 2003 and June 2015, 651 CE procedures were performed for various indications. Patients were excluded from the analysis if they had other SB diseases such as infectious disease, ischemia, vasculitis, tumor, or lymphoma. Finally, we retrospectively analyzed the medical records of patients who underwent CE in 57 cases of sCD and 14 cases of eCD. CE was not indicated for patients in whom obstructive symptoms or obstructive lesions were observed on radiological studies. Among all patients undergoing CE, 74.7% (53/71) underwent SBFT and/or abdominal CT for evaluating SB strictures. This study was approved by Soonchunhyang University Hospital Institutional Ethics Review Board (2015-07-025) and waived for informed consents.

Criteria used to define sCD were the presence of at least 2 of the following: abdominal pain or diarrhea, iron deficiency anemia, elevated ESR or CRP levels, hypoalbuminemia, extraintestinal manifestations, and family history of IBD. eCD was defined as previously diagnosed cases of CD that fulfilled established diagnostic criteria (presence of symptoms and a combination of endoscopic, pathologic, radiologic, and/or laboratory abnormalities).1617 Indications for CE were categorized as follows: (1) initial diagnosis, (2) differential diagnosis, (3) assessment of disease extent and activity, (4) assessment of MH, and (5) identification of postoperative recurrence. Performance of CE in patients with sCD who met the disease criteria was considered the “initial diagnosis” indication for CE. In cases in which evidence of CD, tuberculosis (TB), or other inflammatory diseases was observed in radiological findings obtained prior to CE, the performance of CE indicated “differential diagnosis of CD.” In the present study, the diagnostic criteria used to diagnose CD on the basis of the CE images obtained were the presence of >3 SB ulcerations (in the absence of history of NSAID ingestion) or >10 aphthous or erosive lesions showing either continuous or segment-like distribution.1819 More frequently than CD, TB presents a patulous ileocecal valve, transverse ulcers, pseudopolyps, or cicatricial change. TB was considered in cases wherein CE revealed these findings.20 NSAID-induced enteropathy was diagnosed when CE revealed lesions such as reddened folds, denuded areas, red spots, mucosal breaks, or diaphragmatic ulcers, with a history of NSAID ingestion.21 The phenotype was classified in accordance with the Montreal classification system. The final diagnosis was determined after integrating clinical manifestations, laboratory tests, and disease course based on CE findings. Conservative treatment included nil per os (NPO), total parenteral nutrition, transfusion, or iron replacement. Specific treatment included the addition or increased dosage of anti-inflammatory drugs, steroids, immunomodulators (IMs), or anti-TB medications.

CE was performed using either the PillCam™ SB1/SB2 (Given Imaging, Yokneam, Israel) or the MiroCam™ (Intro-Medic Co., Seoul, Korea). Prior to the procedure, patients underwent bowel preparation (using polyethylene glycol 1-2 L, Colyte™; Taejun Pharmacy, Seoul, Korea) and were kept NPO for at least 8 hours. Abdominal radiography was performed to confirm gastric retention of the capsule 2 hours after swallowing either the PillCam or the MiroCam. All images were reviewed by 2 board-certified endoscopists. The quality of CE images was classified as acceptable (excellent, good, or fair) or unacceptable (poor).22

Statistical analysis was performed using PASW Statistics version 18.0 for Windows (SPSS Inc., Chicago, IL, USA). The results are presented as mean±SD or number (%). Continuous variables were analyzed using a two-tailed Student t-test or Fisher exact test. Categorical variables were analyzed using the chi-square test. A P-value of 0.05 was considered statistically significant.

The mean ages of the patients in the sCD and eCD groups were 45.8±15.9 and 37.4±17.0 years, respectively. Analysis of sex showed that 61.4% of patients in the sCD group and 78.6% of patients in the eCD group were male. In both groups, CE was performed in two-thirds of patients within 2 years of diagnosis of CD (sCD [85.9%] vs. eCD [64.2%]). A higher percentage of patients in the sCD group had a positive history of administration of NSAIDs, anticoagulants, or antiplatelet agents than in the eCD group (21.0% vs. 7.1%). At the time of CE, 71.4% of patients (10/14) in the eCD group were receiving treatment with only 5-aminosalicylic acid (5-ASA), and 28.6% (4/14) were receiving treatment with 5-ASA and steroids. A history of abdominal surgery was more common in the eCD group than in the sCD group (1.8% vs. 35.7%, P=0.001). No significant differences were observed between the sCD and eCD groups with regard to the baseline laboratory characteristics listed in Table 1.

In the sCD group, the most common indication for CE for CD was initial diagnosis (54.4%), followed by differential diagnosis (45.6%). Twenty-five of 57 patients with sCD (43.8%) presented with overt obscure gastrointestinal bleeding. Only 11.5% of patients in the eCD group (14/122) underwent CE. Although 3 of 14 patients with eCD received SB and/or colon resection prior to CE owing to CD, all CEs in patients with eCD were performed for the assessment of disease activity. No significant differences were observed in the percentage of patients in the sCD and eCD groups with regard to the type of CE, quality of images, arrival rate of capsule at the cecum, and SB transit time. Capsule retention was only observed in 1 patient (7.1%) in the eCD group. The capsule was removed endoscopically via double-balloon enteroscopy, and an IM was added to the patient's treatment regimen. Table 2 displays the CE-related data.

In the sCD and eCD groups, the combined jejunal and ileal area represented the most common site of involvement (37.3% and 53.8%, respectively). Variable ulcerations were the most common findings in sCD (41.2%) and eCD (57.1%). Typical longitudinal ulcers were present in 23.5% and 14.3% of patients with sCD and eCD, respectively. The proportions of 14 patients with eCD with B1, B2, and B3 disease at diagnosis were 71.4%, 28.4%, and 7.1%, respectively. The proportions of patients with L1, L2, L3, L4, and L3+L4 disease were 28.6%, 28.6%, 14.3%, 7.1%, and 21.4%, respectively. Following CE, the location of CD was modified in 64.2% of patients (from L1 to L1+L4, n=4; from L2 to L2+L4, n=3; from L3 to L3+L4, n=2). The concordance rates between CE and SBFT/abdominal CT were 36.8%/20.0% and 33.3%/16.7% in the sCD and eCD groups, respectively; these findings were not significant. Table 3 summarizes the endoscopic findings including location of disease.

Of the 57 patients with sCD, 28.1% (16/57) were diagnosed with CD on the basis of CE findings; other diagnoses included tuberculous enteritis, NSAID-induced enteropathy, and other enteritis (Fig. 1), which were identified in 7.0%, 5.3%, and 17.5% of patients, respectively (Fig. 2). The overall diagnostic yield of CE was 59.7% (34/57).

The majority of patients in the sCD group (70.2%) were treated conservatively, while specific medications were initiated in 28.1% (n=14, 5-ASA; n=1, 5-ASA and steroid; n=1, anti-TB medication), and potentially disease-causing drugs (NSAIDs) were terminated for 1.8% of patients. In the eCD group, therapy was retained for 57.1% of patients (n=3, 5-ASA) or specific medication was added and/or increased (n=4, addition of steroid; n=1, dose increase of 5-ASA and addition of steroid), while 14.2% of patients were referred for surgery owing to obstructive symptoms following CE. Nevertheless, there were no dominant obstructive lesions on SBFT. The performance of surgery after CE was not associated with CE retention in any case. The final pathology revealed chronic transmural inflammation with granuloma, consistent with CD. Therapeutic strategies were changed for 70.2% of patients (40/57) in the sCD group and 50% of patients (7/14) in the eCD group on the basis of CE findings. Treatment results are summarized in Table 4.