BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION In the context

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION In the context of IBD, the term mucosal healing refers to endoscopic assessment of disease activity and is usually taken to reﬂect resolution of visible ulcers. Endoscopic mucosal healing is a defined endpoint in the treatment of IBD with a better clinical outcome (steroid-free remission, lower rate of hospitalization and surgery). However, a proportion of patients do not enter clinical remission despite healing of the mucosa. Neurath M et al. Gut 2012 Rutgeerts P et al. Gastroenterology 2012 Allez M et al. World J Gastroenterol 2010 Van Assche G et al. Curr Drug Targets 2010

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION IBD patients often report symptoms referable to the GI tract, without objective evidence of ongoing disease activity, despite mucosal healing, with or without normalization of fecal calprotectin… The exact causes remain uncertain…(sub-clinical inflammation? IBS? ) The overall prevalence of symptoms suggestive of IBS, when data were pooled from 13 studies, including 1703 patients with IBD, was almost 40%, with up to 1/3 judged to be in clinical remission. Halpin S et al. Am J Gastroenterol 2012

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION IBD pts with ongoing bowel symptoms had significant impairment of intestinal permeability versus those in clinical remission or healthy controls. Impaired epithelial barrier function has been considered to be important in the pathogenesis of IBD, IBS, but whether impaired epithelial barrier function and increased intestinal permeability underlies ongoing symptoms in IBD has not been adequately addressed. Martínez C et al. Gut 2013 Mc. Guckin MA, et al. Inflamm Bowel Dis 2009

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION Watson grade Kiesslich R et al. Gut 2012

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION Confocal laser endomicroscopy (CLE) was introduced in 2004, aimed top obtain real time in vivo histology during ongoing endoscopy. Provides subcellular resolution with magnification up to 1000 -fold. The technique is based on tissue illumination with a blue laser light after topical or systemic application of fluorescence agents. Lee G Lim et al. IBD 2014 Tontini GE et al. JCC 2014

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION Detecting more neoplasia in surveillance colonoscopies (combination with chromoendoscopy) Several different potential applications of CLE in IBD Differentiate DALM from adenoma (accuracy 97%) sporadic Detect microscopic changes associated with active CD with high accuracy Using fluorescent monoclonal anti-TNF antibodies may predict therapeutic responses to therapy Tontini GE et al. JCC 2014

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION To quantify the severity of barrier dysfunction, a new quantitative numerical score was developed, named the Confocal Leak Score (CLS). continuous grade of severity assessment • Previously demonstrated for celiac disease. • The Confocal Celiac Score demonstrated excellent correlation with the histologic Marsh classification. Kiesslich R et al. Gut 2012 Leong RW et al. Gastroenterology 2008 Yang MY et al. J Gastroenterol Hepatol 2013 Leong RW et al. J Gastroenterol Hepatol 2012

Primary aim: to examine the association of small intestinal permeability as detected by CLE and ongoing symptomatic IBD patients who have achieved mucosal healing. Secondary aim: to examine the association between symptom severity and small intestinal permeability as detected by CLE.

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION STUDY DESIGN • Prospective study on IBD subjects vs age- and sex-matched controls • IBD subjects that demonstrated endoscopic mucosal healing were assessed by imaging of their intestinal mucosal barrier to determine whether impaired intestinal permeability correlated with bowel symptoms. • The 2 patient-reportable outcomes of interest were number of diarrheal stools passed per day and abdominal pain.

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION STUDY DESIGN • ENROLLMENT and INCLUSION CRITERIA: Between May 2009 and September 2015, consecutive patients (18– 70 years of age) with established IBD for at least 12 months’ duration, who required a colonoscopy for clinical indications, were enrolled for e. CLE. • Age- and sex-matched controls were recruited in parallel (excluded if they presented with regular diarrhea or abdominal pain or were subsequently diagnosed with a gastrointestinal disease after the e. CLE)

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION STUDY DESIGN • EXCLUSION CRITERIA: Known IBS, celiac disease, intestinal resection surgery (apart from limited ileal resection for CD), pregnancy or breastfeeding, renal disease, diabetes mellitus, decompensated liver disease, regular use of NSAID or known allergy to fluorescein.

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION DATA COLLECTION ü Disease phenotype, extent of disease, duration of disease, smoking status, and NSAID use. ü Symptoms prior to the e. CLE bowel preparation were recorded in a diary. ü CDAI symptomatic if ≥ 150; partial Mayo score symptomatic if ≥ 2 ü Symptoms were further subclassified as diarrhea or abdominal pain (the last from 0– 3, corresponding with nil, mild, moderate to severe, and averaged over 7 days). ü Serum inflammatory markers: ESR (upper normal limit, 16 mm. Hg) and CRP (upper normal limit, 6 mg/L)

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION DATA COLLECTION ü Endoscopic activity: CDEIS and Mayo endoscopic sub-score ü Mucosal healing (ileocolonoscopy for all subjects) defined as a CDEIS of 0 for CD, or Mayo 0 -1 for UC ü Histologic biopsies had to demonstrate quiescent disease for those meeting the definition of mucosal healing defined as absence of epithelial breach (ulceration, erosions) and inflammation (cryptitis, crypt abscess, neutrophilic infiltration). ü All symptomatic patients had abdominal imaging such as MRI within the prior 12 months, stool examination to exclude infective causes

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION e. CLE ü Pentax EC-3870 FK, Tokyo, Japan; conscious sedation with propofol delivered by an anesthesiologist. ü 1000 -fold magnification of the intestinal mucosa provided ü Inflamed segments were avoided so local inflammation did not influence the assessment of intestinal permeability

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION e. CLE ü To reduce sampling errors, 5 macroscopically normal (non-inflamed) segments of the terminal ileum were selected for e. CLE, with a minimum of 50 confocal images obtained at each site and stored in a digital database. ü At each of the 5 terminal ileum sites, incremental imaging from the surface down toward the lamina propria was performed to capture a variety of mucosal depths. ü 5 m. L of intravenous fluorescein sodium was administered at increments of 1 ml for each terminal ileal site.

BACKGROUND METHODS RESULTS CONCLUSION CLE: INTESTINAL PERMEABILITY 1. cell junction enhancement (representing fluorescein enhancement between epithelial cells secondary to loss of tight junction protein integrity); (B) 2. fluorescein leak (an efflux of fluorescein from the submucosal space into the gut lumen through epithelial breaks) (C) 3. cell drop out (actively shedding enterocyte(s) with epithelial gap(s) often accompanied with fluorescein leak) (D) CONTROL DISCUSSION

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION CLE: FEATURES OF INTESTINAL PERMEABILITY (CLS) ü The presence or absence of impaired intestinal permeability was assessed in each image, defined as the findings of 1 or more features of cell junction enhancement, fluorescein leak, or cell drop out, or none of the 3 features, respectively. ü To calculate the CLS, all stored e. CLE images for each patient were analyzed independently by a researcher blinded to both patient disease category and endoscopic findings from stored images after the procedure (images excluded if either less than one third of the image contained villi or if presence of artifacts). ü The final CLS is derived from the number of images with 1 or more of the features, divided by the total number of terminal ileal confocal images per patient, multiplied by 100, giving a score of 0 (absence of barrier dysfunction) to 100 (complete barrier dysfunction).

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION CLE: FEATURES OF INTESTINAL PERMEABILITY ü A comparison between CLS against the Watson grades (I, II, and III, in order of increasing severity) had been made ü A set of 30 high-quality images were randomly selected and graded by both methods (testing for discordance, non-parametric correlation)

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION STATISTICAL ANALYSIS ü Sample size calculation was based on 21%-39% of IBD patients in remission having IBS-like symptoms. ü Assuming a 1. 5 - fold increase in permeability comparing symptomatic- vs asymptomatic subjects in mucosal healing with a 2 -sided significance of. 05 and a power of 80%, a total of 32 subjects will be required at an enrollment ratio of 3 asymptomatic subjects to every 1 symptomatic IBD subjects.

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION STATISTICAL ANALYSIS ü Non-parametric continuous variables were described as medians and IQR. ü Linear regression univariate and multivariate ü ROC curve and its area under curve was used to designate a CLS cutoff that is sensitive and specific to the presence of ongoing bowel symptoms in the setting of a normal mucosa.

BACKGROUND METHODS RESULTS CONCLUSION 31 UC, 57 CD 16. 3% DISCUSSION

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION

BACKGROUND METHODS RESULTS CONCLUSION CLS IN SYMPTOMATIC AND ASYMPTOMATIC IBD SUBJECTS The median CLS of the IBD cohort in mucosal healing (12. 9, IQR: 6. 7– 19. 4) was significantly higher than controls (5. 9, IQR: 3. 7– 9. 6; P=0. 001). DISCUSSION

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION CLS IN SYMPTOMATIC AND ASYMPTOMATIC IBD SUBJECTS A CLS cutoff of 13. 1 defined ongoing bowel symptoms in IBD with a sensitivity of 95. 2% and specificity of 97. 6%. In this study, 36. 2% of IBD subjects in mucosal healing had intestinal permeability that exceeds this cutoff. The ROC area under curve of the CLS in determining symptoms in the setting of mucosal healing was 0. 88 (95% confidence interval, 0. 79– 0. 98; p <0. 001).

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION INFLAMMATORY MARKERS AND INTESTINAL PERMEABILITY In patients in mucosal healing, neither the median ESR (p 0. 64) or CRP (p 0. 97) differed significantly between symptomatic and asymptomatic subjects. Elevated ESR and/or CRP did not predict for bowel symptoms (OR, 1. 12; p 0. 89). Neither ESR (r. 0. 06; p 0. 75) nor CRP (r. -0. 06; p 0. 70) correlated with CLS.

BACKGROUND METHODS RESULTS CONCLUSION SEVERITY OF SYMPTOMS AND INTESTINAL PERMEABILITY DISCUSSION

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION ASSESSMENT OF INTESTINAL PERMEABILITY WITH CLS AND WATSON SCALE There was consistency and low discordance in the assessment of intestinal permeability (p=0. 45). The 2 scores also correlated significantly (r=0. 40; p=0. 03). Median CLS for Watson grades I, II, and III were 9. 2 (IQR: 2. 7 -15. 4), 15. 3 (IQR: 9. 2 -28. 7), and 15. 4 (IQR: 15. 4 -33. 1), respectively, but did not differ significantly between the 3 levels (p=0. 09).

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION ü CLS was significantly higher in symptomatic than asymptomatic IBD patients (even separately for UC and CD) who are in mucosal healing, supporting author’s hypothesis that intestinal permeability is associated with ongoing symptoms. ü Age, IBD disease duration, smoking status, NSAID use, CRP and ESR were not predictive of CLS on linear regression. ü No significant differences were seen between asymptomatic IBD subjects vs healthy controls, indicating that full recovery of intestinal permeability is possible.

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION ü A CLS value of 13. 1 correlated well with ongoing bowel symptoms despite mucosal healing in IBD. Based on this cutoff, one third of IBD subjects who are in mucosal healing had increased intestinal permeability. ü Higher CLS in symptomatic patients in this study was not attributable to either intestinal or systemic inflammation (no correlation of CLS to either ESR or CRP). ü CLS correlated significantly with diarrheal motions per day but not to abdominal pain. Interestingly, female sex was also an independent predictor of a higher CLS. ü Higher levels of psychological stress and anxiety in symptomatic IBD patients in remission, may explain this findings.

BACKGROUND METHODS RESULTS CONCLUSION DISCUSSION ü CLE: effective tool in the assessment of intestinal permeability beyond its traditional role of obtaining virtual histology. ü Impaired intestinal permeability underlies ongoing bowel symptoms in IBD subjects who are in mucosal healing. The degree of impaired intestinal permeability was predictive of the severity of diarrhea. the ultimate treatment target of IBD should include not only mucosa healing but also recovery of the intestinal mucosal barrier function.