How Accurate Are Capsule Endoscopy and Small Bowel Imaging for Small Intestinal Crohnʼs Disease? Using Total Enteroscopy via Double-Balloon Enteroscopy as the Reference Standard

Abstract

Introduction: The target for treatment in small intestinal Crohn's disease (CD) is mucosal healing. The diagnostic modalities used to not only assess healing, but also diagnose Crohn's disease are small bowel imaging and capsule endoscopy (CE). However, ileocolonoscopy has been the reference standard for their accuracies. The aim of this study was to measure the true accuracy of these modalities by using total enteroscopy (TE) via double balloon enteroscopy (DBE). Methods: Single center, retrospective chart review of patients with suspected and established CD status post TE via DBE and small bowel imaging. CD activity was defined as positive if >1: aphthous lesion, small ulcer, large ulcer, ulcerated mucosa, and/or stenoses. For imaging, CD was defined as positive if presence of wall thickening, increased contrast enhancement, or hyperemia, and/or the presence of stenoses. Results: 2216 DBEs were performed between 2004-2015 with 635 TEs in 361 patients. 75 TEs were achieved in one direction. 61 patients (M40: F21, 51yo) identified with CD (33 suspected and 28 established) underwent 120 DBEs for 64 TEs. 14 DBEs required balloon dilation. 6 DBEs were performed for capsule retrieval. 0 perforations, pancreatitis. 19 patients underwent CE with a median Niv score 6.0. The median CRP was 0.21 mg/dl (0.04-18.5) and median fecal calprotectin 160.6 mcg/g (46.8-1360.0). Because of TE, there was maintenance or escalation of therapy 9, resection/lysis of adhesions 4, IBS 8, de-escalation of therapy 2, symptom improvement post balloon dilation 3, diagnosis of CD 20, CD ruled out 5, and other diagnoses 7(celiac sprue, NSAID enteropathy, carcinoid, Meckel's). For all imaging modalities the sensitivity for CD was 80% vs 67% in CE, yet CE had a PPV of 83%. Specific small bowel imaging, such as CTE/MRE/SBFT, had a similar sensitivity at 79% but lower PPV at 52%. However, in terms of strictures, CTE/MRE/SBFT had a lower sensitivity of 56% vs 86% in CE. Additionally, CE had a 100% specificity for strictures. Conclusion: Although there were fewer CEs in this study, CE had a high PPV overall for CD activity and a high sensitivity for strictures, especially inflammatory strictures, when compared to small bowel imaging. Specifically, imaging was excellent for identifying the presence of inflammation but inaccurate when identifying the degree of inflammation, such as stenoses.Therefore, CE can be an adjunct study to imaging as a safe modality to assess not only accurate disease phenotype, but also healing post treatment.Table 1: Imaging vs. CE for Stricturing Small Bowel Crohn's DiseaseTable 2: Operating Characteristics of Imaging vs. CE: Total Enteroscopy as the Reference StandardFigure 1