Confocal Laser Endomicroscopy In The Morphological And Functional Diagnosis Of Colonic Mucosal Lesions

Backgrounds and aims:The intestinal barrier consists of intestinal epithelium, paracellular junction and secretive barrier. The intestinal barrier plays a crucial role in the maintenance of homeostasis and regulation of immune activity. Barrier dysfunction in the pathogenesis of several gastrointestinal diseases have been drawing much attentions, such as inflammatory bowel disease (IBD), Celiac disease. And recently the functional GI disorders including irritable bowel syndrome (IBS), which was conventionally believed to be lack of organic alterations, are also found to be associated with barrier dysfunction.Studies on barrier dysfunction inevitably involve assessment by endoscopy and gastrointestinal permeability test. Currently the hiding active inflammation could not be easily detected by using conventional white lighe endoscopy. It was reported that 30% of patients appearing normal under conventional endoscopy were revealed as having acute active inflammation by histopathology. And the hiding acute active inflammation was proved to be an independent risk factor to predict relapse. In the area of intestinal permeability tests, both the invasive and non-invasive methods are not easy to carry out. The gold standard invasive method, Ussing chamber, is in need of dozens of fresh biopsy samples, which is difficult to carry out on human. Consequently the non-invasive assessment of urine markers after oral intaking. The markers used in the intestinal permeability test should meet 3 conditions:1, non or low toxic; 2, not absortable in normal conditions; 3, quantitatively detectable in urine. The most commonly used markers are saccharides. The non-invasive tests rely on oral administration and subsequent urinary collection of probes that can selectively characterize permeability from different regions fo the gut. In many studies, a triple sugar test involving administration of 3 saccharides was used to assess small intestinal permeability and colonic permeability. Traditionally, the lactulose to mannitol ratio is used to express small intestinal permeability while 24 hours of sucralose excretion is used to express colonic permeability. The other common probe is (51)Cr-EDTA. This probe is capable of whole gut permeability assessment. However, its radioactivity and requirement of temporary storage limit the applications.Confocal laser endomicroscopy (CLE) is an endoscopic device providing in vivo microscopic morphology during ongoing endoscopic procedures. The mostly applied staining agent in CLE is fluorescein sodium, which is intravenously applied and distributed immediately throughout mucosal tissues. In our previous study on classification of inflammation activity in ulcerative (UC) by CLE, the fluorecein leakage into lumen (FLIL) was associated with active inflammation in UC. One possible explanation of FLIL is increased colonic permeability, since fluorecein sodium and sucralose have the similar molecular mass (376.27 and 397.64, respectively). The primary goal of this study is aimed to validate the association between FLIL in colonic mucosa and colonic permeability.MethodsPartⅠFrom June 1,2008, to April 30,2009, we recruited consecutive patients previously diagnosed as having UC who visited the outpatient department of Qilu Hospital for colonoscopy surveillance. After being informed about the purpose of the study, those who were willing to choose CLE instead of conventional colonoscopy were included in the study. The CLE procedures did not differ from those of conventional colonoscopy, except for the additional evaluation of mucosal inflammation in the distal colon, including the sigmoid colon and rectum, by the Baron Score. After Baron endoscopy scores for inflammation were recorded, the distal tip of the endoscope was placed gently onto the observed mucosa with the endomicroscopy mode turned on. At least 2 to 3 Z-stacks of images (scanning from the superficial to the deep layer of targeted mucosa) were obtained, and the endoscopists evaluated the crypt architecture simultaneously. The crypt architecture was classified into 4 types. Types A and B are considered normal and chronic inflammation respectively, and types C and D indicate acute inflammation. Microvascular alterations were evaluated by real-time CLE assessment of inflammation activity as well. In addition to crypt architecture and microvascular alterations, we introduced a new marker, fluorescein leakage into lumen (FLIL), to define acute inflammation seen on CLE. Images of observed mucosa were stored digitally on laser discs for further evaluation, In the end, a targeted biopsy was performed for histological analysis. Objective post-CLE analysis involved evaluation of all Z-stacks of images of the observed mucosa for number of crypts per CLE image (CPCI) and fluorescein density of CLE images (FDCI). CPCI was considered the objective variable to show the reliability of the objective real-time crypt architecture analysis, as was FDCI for FLIL. Three pathologists (CJZ, WQH, HC) from 2 independent hospitals evaluated the slices for histology assessment of inflammation in UC according to the Geboes Index. The scale includes 6 grades:structural (architectural changes), chronic inflammatory infiltrates, lamina propria neutrophils and eosinophils, neutrophils in epithelium, crypt destruction, and erosion or ulceration. Each grade is divided into 4 to 5 subgroups.PartⅡConsecutive patients who were scheduled for colonoscopy from April 1st to June 30th,2009 were informed about the purpose of this study. Those prefering CLE over conventional colonoscopy were enrolled and gave their written informed consents. After successful intubation into cecum,5 ml of 10%fluorescein sodium was intravenously given. The distal tip of CLE was gently placed onto mucosa while laser turned on. Laser volume were fixed on 500 nm and brightness on median of all the procedures. Each procedure included CLE observation per 10 cm of colon during the withdrawl of the colonoscope as well as conventional white light colonoscopy examination. Images of each z-stack were stored in a specific folder. Two biosies were taken from each patient, one was for routine histopathology, and another for transmission electron microscopy. For patients without FLIL, biopsies were taken from the distal colon, and biopsies were taken from the targeted mucosa of those with FLIL. In addition to evaluation of FLIL, analysis of epithelial gaps in colon was also conducted in some patients. Number of gaps were counted by per z-stack of images as reported. We assumed that decreased numbers of gaps under CLE after fluorescein injection indicates epithelial gaps are permeable, and constant gaps despite of fluorescein injection indicates impermeability of epithelial gaps. Two days before CLE procedures, patients fasted overnight and emptied their bladder before drinking a test solution containing 5 g mannitol,10 g lactulose, and 5g sucralose diluted in 100 ml of tap water. Then the patients were asked to collect 24 hours of urine into the sealed and disinfected containers. Urine was then collected for the following 24 hours. The ratio of lactulose and mannitol (L/M) recovery during the first 5 hours was used as an index of small intestinal permeability and the total mass of sucralose excretion (mg) during the 24 h was used as an index of colonic permeability. Samples from mucosa with or without FLIL were sent for transmission electronic microscopy (TEM) analysis.ResultsPartⅠAssessment of crypt architecture and fluorescein leakage with CLE both showed good correlation with histology results (Spearman rho, both P< 0.001). CLE seemed to be more accurate than conventional white-light endoscopy for evaluating macroscopical normal mucosa. More than half of the patients with normal mucosa seen on conventional white-light endoscopy showed acute inflammation on histology, whereas no patients with normal mucosa or with chronic inflammation seen on CLE showed acute inflammation on histology. Assessment of microvascular alterations by CLE showed good correlation with histology findings (P<0.001). On post-CLE objective assessment, subjective architectural classifications were supported by number of crypts per image (P< 0.001) but not fluorescein leakage results by grey scale(P= 0.194).PartⅡFLILs were found in 21 z-stacks of CLE images during 18 procedures. Among the 21 z-stacks, 17 were found in the cecum mucosa, and the rest were in the distal segments including sigmoid colon (3) and rectum(l). Time durations between fluorescein injection and FLIL were from 2 to 11 minutes. Distribution of procedures with FLILs among patients'groups are:7 in UC,8 in IBS and 3 in HC. Chi-square test showed no significant difference of FLIL prevalence among 3 groups (X2=2.326, P=0.312). Though average colonic permeability (parameter:sucralose excretion) is higher in patients with UC than IBS and HC, the differencs were not statistically significant P=0.083). However, patients with FLIL found during CLE have significantly higher average colonic permeability than those without (total sucralose secretion:FLIL,60.64±18.07mg vs. non-FLIL,33.26±8.54mg, P=0.013). There was no significant difference between numbers of epithelial gaps before and after fluorescein injection (Wilcoxon test, P=0.9). In the 12 patients without FLIL under CLE examinations, TEM showed normal basement membrane in 10 patients. In the 18 patients with FLIL, TEM showed normal basement membrane in 6 patients (Contingency Coefficient, value=0.441, P=0.007). The abnormalities of basement membrane includes decreased thickness, breaked and infiltrated by lymphocyte. Patients with abnormal TEM findings had higher colonic permeability (57.66±21.24 vs.42.71±16.77, P=0.04).Conclusions●Confocal laser endomicroscopy is reliable in assessment of colonic mucosal inflammation.●For those with hiding acute active inflammation but appearing normal under conventional white light endoscopy, confocal laser endomicroscopy could provide more accurate diagnosis.●Crypt architecture, fluorescein leakage and microvascular alterations are promising markers for acute inflammation under confocal laser endomicroscopy.●Fluorescein leakage under confocal laser endomicroscopy is associated with increased colonic permeability.●Paracellular pathway, and abnormal basement membrane are more likely to the pathways of fluorescein leakage than epithelial gaps.SignificanceA practical criteria upon classification of inflammatory activity in colonic mucosa by confocal laser endomicrsocopy combining crypt architecture, microvascular alterations and fluorescein leakage was established in this study. Fluorescein leakage into crypt lumen is the first time to be raised in the assessment of colonic permeability. And epithelial gaps were firstly approved to be impermeable in an in vivo manner by confocal laser endomicroscopy.