Background: Crohn's disease (CD) and ulcerative colitis (UC) have been associated with altered intestinal barrier function. Moreover, it has been proposed that a defective barrier function is related to risk of relapse in patients with quiescent CD. Fluorescein-aided confocal laser endomicroscopy (CLE) is a novel endoscopic method, which enables real-time in vivo microscopy. Hence, the intestinal barrier function can be assessed as part of endoscopic evaluation of patients with inflammatory bowel disease (IBD) by measuring microerosions and fluorescein leakage into the intestinal lumen. Furthermore, barrier dysfunction can be correlated with biomarkers associated with intestinal barrier impairments. E-cadherin is a key factor for the adherence of epithelial cells and Smad4 is a cofactor in TGF-β signalling, which is compromised in IBD.
Aim: To correlate ileal and colonic CLE parameters with endoscopy and histopathology in IBD. Further, we wanted to correlate these features with risk of relapse and evaluate whether they were reproducible and reversible after intensified medical treatment. We also wanted to analyse, whether Smad4 and E-cadherin mucosal protein expression levels were associated with impairments of intestinal barrier function.
Methods: CLE was performed and correlated to histopathology and endoscopic appearance in two prospective studies in CD (n = 39, controls = 11) and UC patients (n = 22, controls = 7), respectively. In the first study, results were correlated to risk of relapse, whereas the latter assessed the reversibility of CLE features in a longitudinal setting. κ-statistics were used in both studies to assess reproducibility of the CLE findings. Furthermore, ileal biopsy specimens from CD patients and controls were stained by immunohistochemistry for Smad4 and E-cadherin and subsequently correlated to the severity of CD and intestinal barrier impairments.
Results: We found that fluorescein leakage and microerosions in the terminal ileum were significantly associated with CD com-pared to controls (p = 0.005 and p = 0.006, respectively) and that ileal fluorescein leakage and microerosions could predict relapse (log-rank p = 0.003 and p = 0.017, respectively). In UC patients with clinical relapse, an augmented crypt architecture and colonic fluorescein leakage were significantly correlated to the severity of the disease (p = 0.001 and p < 0.001, respectively). After intensified medical treatment, a correlation was found between histopathological progress and improvement of abnormal colonic crypt architecture (rs = 0.35, p = 0.016), but we did not observe a resolution of the intestinal barrier dysfunction (rs = 0.09, p = 0.56). The inter-observer variability of CLE parameters ranged from fair to substantial, while the intra-observer variability was somewhat higher. Smad4 expression (rs = 0.56, p = 0.002), but not E-cadherin (rs = 0.01, p = 0.95), was correlated with the severity of the disease; however, Smad4 expression did not correlate with a defect barrier function.
Conclusions: CLE can visualise crypt alteration and barrier impairments in both CD and UC, which are otherwise undetectable. Further studies are warranted to incorporate CLE in the endoscopic and therapeutic management algorithm for CD and UC possibly refining the definition of mucosal healing. Smad4 expression was correlated with CD as well as disease severity and may serve as a novel treatment target.