Endoscopic ultrasound of the sigmoid colon for the differentiation of Crohn`s disease from unspecific, infectious colitis and healthy controls — a prospective, blinded, comparative study

M. Ellrichmann*, P. Wietzke-Braun*, J. Bethge, L. Wintermeyer, S. Nikolaus, A. Arlt,
T. Khbacher, P. Sergeev, S. Schreiber, A. Fritscher-Ravens
*These authors contributed equally to this work.
Interdisciplinary Endoscopy, Medical Department I, University of Schleswig Holstein, Campus Kiel
Kiel, Germany

Introduction:
In Crohn`s disease (CD) colonoscopy is one of the diagnostic priorities, a method of imaging that is limited to the mucosal surface. Endoscopic ultrasound (EUS) might be the best modality for the evaluation of transmural changes, but in the lower GI-tract is limited to the rectum, where rigid probes are used.

Aim:
To evaluate the capability of flexible EUS of the colon wall to differentiate active CD (aCD), CD in remission (rCD) from unspecific infectious colitis (IC) and healthy controls (HC) in a prospective, blinded, comparative study.

Methods:
37 consecutive patients with CD (aCD: n=19; rCD: n=18), 13 with IC and 50 HC without pathology on screening colonoscopy were included. After IRB approval for the use of a forward-viewing radial echoendoscope (Pentax, Japan) in the colon, the layers of the bowel wall, and possible locoregional lymphnodes were evaluated in the sigmoid. All measurements were repeated three times at various areas of the inspected sigmoid to acknowledge possible inhomogenous spread of disease. The examiners were blinded to the diagnosis and status of disease. C-reactive protein (CRP) was determined and macroscopic, endoscopic pathologies were quantified according to the SES-CD-score.

Results:
Total wall thickness (TWT) of HC and rCD was similarly small (1.8±0.04mm, 1.7±0.09mm) and defined to be the normal size. TWT in aCD (2.8±0.27mm) and IC (4.3±0.49mm) were statistically significantly different to the norm (p<0.0001) but not when compared to each other (p>0.05). Patients with IC had a significantly increased mucosal thickening (MucosaIC =2.7±0.24mm), when compared to all others but most importantly when compared to aCD (MucosaaCD=1.3±0.17mm; p<0.001). Patients with aCD had a significant thickening of the submucosa (SubmucosaaCD=2.0±0.22mm) when compared to the others especially when compared to IC (SubmucosaIC=1.3±0.14mm; p<0.001). Paracolonic echopoor lymphnodes were present in 73.7 % (14/19) of aCD patients but in none of the others. In patients with CD significant correlation was observed between macroscopic findings as indicated by SES-CD-Score and TWT (p<0.0001; r=0.75); in contrast, CRP did not correlate with TWT (p=0.24; r=0.48). Using the following criteria: TWT, thickening of the mucosa/submucosa, paracolonic echopoor lymphnodes the sensitivity was 92 % for the differentiation of aCD from IC, sensitivity was 97 % when aCD and IC were compared to the norm.

Conclusion:
Increased TWT of the sigmoid colon has a high positive predictive value for the presence of aCD and IC when compared to HC. EUS can reliably differentiate aCD from IC, when the size of the mucosa, submucosa and presence of paracolonic lymphnodes are evaluated. In patients with aCD EUS of the sigmoid can quantify the level of colonic inflammation and may have the potential to become a prognostic marker of therapeutic success in the future.