Molecular Imaging And And Functional Imaging Of Lower Gastrointestinal Tract Disease Using Gonfocal Endomicroscopy

Background and aims:Colorectal cancer (CRC) is one of the most common malignant tumors of gastrointestinal tract. The well-accepted pathogenetic framework for colorectal tumorigenesis is an adenoma-carcinoma sequence. Early detection, diagnosis and differential diagnosis of colorectal cancer and precancerous lesions could significantly increase the overall survival rate of patients. Inflammatory bowel disease (IBD) is a kind of nonspecific chronic bowel inflammatory disease with unclear pathogenesis and characterized by intestinal immune system disorder. Although the probability of IBD become colorectal cancer is low, colorectal neoplasia is recognized as the most severe complication of IBD. Early assessment and intervention for IBD is one of the successful prevention strategies, improving the prognosis of IBD patients.To date, endoscopy and pathology are the gold standard of many lower gastrointestinal tract diease, including colorectal neoplasia and inflammatory bowel diease. With the advances in in optical instruments, e.g. confocal laser endomicroscopy (CLE), the early detection of lower gastrointestinal tract diease became possible. Confocal endomicroscopy is new endomicroscopic technique with a1000fold magnification, could observe the gastrointestinal tract in vivo at tissue, cellular and subcellular level, and differetiate normal, hyperplastic and neoplastic mucosa, which is named as "optical biopsy". With CLE, we can investigate the physiological and pathological mechanism gastrointestinal trat in vivo. Molecular imaging and functional imaging using CLE are rapidly advanced subjects of gastrointestinal endoscopy.The specific molecular expression of tumor cells can be detected by molecular imaging. Molecular imaging using CLE realized the breakthrough of molecular imaging from ex vivo to in vivo, combining molecular probe and endomicroscopic technique, detecting the molecular expression of tumor cells in vivo. Epidermal growth factor receptor (EGFR) is one of the molecular markers of tumor angiogenesis. A study has reported in vivo molecular imaging of EGFR in human CRC mice model, firstly achieving in vivo molecular imaging of EGFR in CRC. The study demonstrated that molecular imaging using CLE could differentiate CRC with different EGFR expression. Moreover, molecular imaging was also carried out in fresh CRC tissue by topically applied with fluorescein labelled EGFR antibody. This study provided a theoratical basis for the transition of molecular imaging from animal experiment to clinical use.Based on the morphological and functional diagnostic criteria, functional imaging using CLE could observe the pathophysiological process of tissue and cell in vivo, e.g. cell shedding, cell migration, cell necrosis and apoptosis and tissue perfusion. Cell shedding (gaps) is a potential factor that might cause gastrointestinal epithelial barrier dysfunction. Studies have shown that inflammation could drive the epithelial cells to increased shedding. A study by Kiesslich et al has demonstrated20%of gaps were permeable in animal intraperitoneally injected with TNF-α. Other studies have shown that the gaps density in the termial ileum of IBD patients was significantly higher than controls by CLE imaging, and the increased epithelial gaps could predict the relapse of disease. Probiotics could maintain intestinal epithelial barrier by increasing tight junction protein expression, decrease epithelial cell apoptosis induced by inflammatory factors and necrotizing enterocolitis, and decrease the intestinal permeability. However, to date, no study has investigate the influence of probiotics on intestinal epithelial gaps by in vivo observation and the possible mechanisms.The aims of the current study were: (1) to evaluate the use of CLE for in vivo molecular imaging of EGFR in patients with colorectal neoplasia and to perform a comparative analysis of in vitro detection of EGFR expression with IHC and in vivo EGFR molecular imaging using CLE;(2) to investigate the gaps density in patients with IBD by CLE, and to evaluate the function of probiotics application to the epithelial gaps in IBD rat model using CLE, and discuss the possible mechanisms.Methods:Part One: In vivo molecular imaging of colorectal neoplasia by confocal laser endomicroscopy.From January2011to June2011,40consecutive patients known to have colorectal neoplasia from previous examinations as outpatients or inpatients at our hospital were enrolled in our study. In all patients, surgical or endoscopic resection was planned in accordance with the evaluative results of the previous colonoscopic examination. Study was divided into2stages.Stage Ⅰ: pilot studyIn order to establish the proper antibody concentration and incubation time, the first3participants with previously confirmed CRC were recruited for a pilot study. After topical application of5ml of Alexa Fluor488labeled anti-EGFR antibody to the CRC at concentrations of1:500,1:100, or1:50, CLE (Pentax EC-3870K) imaging was performed after a10and15min incubation. Targeted biopsy was taken from the observed site. Molecular imaging with high quality was achieved at a dilution of1:50and incubation time of10min.Stage Ⅱ:prospective studyThirty-seven patients with known colorectal neoplasia were enrolled in the prospective study. After fully rinsing, EGFR antibody at the concentration of1:50was topically sprayed to the neoplastic tissue. Molecular imaging was conducted by CLE after10min incubation. Confocal imaging of normal mucosa adjacent to the neoplastic lesions from the same patients was also performed in10cases. Targeted biopsy was taken at all observed sites for immunohistochemistry (IHC) and histology. Fluorescence intensity of confocal images was quantified offline using Image J. Three ROIs of60*60μm with the strongest fluorescence in the representative image were selected for calculation. The grading of H&E results was performed by an experienced gastrointestinal pathologist in a blinded manner according to the modified Vienna classification. Serum samples from4patients were taken4-6weeks after molecular imaging and were tested for human anti-mouse antibodies (HAMAs).Part Two: Functional imaging of epithelial gaps in small intestine of inflammatory bowel disease using confocal laser endomicroscopy. Study was divided into2stages, clinical study and animal study.Stage Ⅰ: clinical studyThe study group included outpatient with known or suspected to have a diagnosis of IBD. Asymptomatic individuals undergoing colonoscopy for health surveillance or follow up after polypectomy were included as controls. After thorough bowel preparation and conscious sedation, confocal laser endoscopy (Pentax EC3870CIK, Tokyo, Japan) was carried out with the patients. After successful intubation into the terminal ileum,6ml of fluoresceince sodium was injected intravenously. A minimum of5different, normal-appearing sites in terminal ileum were imaged using CLE. CLE images were collected and stored for further evaluation of gaps density.Stage Ⅱ: animal studyRats were randomly assigned to3groups: IBD group, normal control group, and IBD treated with VSL#3group. Confocal images was carried out by rigid confocal mini-microscope FIVE1in terminal ileum of rats for gaps density (number of gaps per1000cells). Biopsy samples were taken in each animal group for ex vivo IHC of occludin and ZO-1and Enzyme-linked Immunosorbent Assay (ELISA) of IL-1β. Effect of IL-1β on epithelial gaps formation was evaluated by CLE imaging after intraperitoneal injection of IL-1(3in doses of5and10μg/kg in rats.Results:Part One: In vivo molecular imaging of colorectal neoplasia by confocal laser endomicroscopy.A total of37patients (21men,16women) with colorectal neoplastic lesions were enrolled in prospective study. After topical application of the labeled anti-EGFR antibody to all lesions, a specific EGFR fluorescence signal could be detected in18(94.7%) of the19CRC and in12(66.7%) of the18colorectal adenomas. Eight CLE-positive carcinomas included8well-differentiated carcinomas (8/8),10moderately differentiated (10/11), and0poorly differentiated (0/0). Six out of10high-grade intraepithelial neoplasia (6/10) and six out of eight low-grade intraepithelial neoplasia (6/8) were CLE positive. The mean fluorescence intensity was52.84±5.73for CRC and44.31±4.86for colorectal adenomas (p=0.329). Normal mucosa showed no specific fluorescence signal in7cases and weak fluorescence signal in3cases. The mean fluorescence signal intensity of normal mucosa was36.23±7.90compared with53.78±7.06in neoplastic tissues of the same10patients (p<0.001). The kappa value demonstrated a substantial agreement between in vivo EGFR imaging by CLE and ex vivo EGFR staining by IHC (κ=0.788). None of the4serum samples were found to be increased for HAMAs titers.Part Two:Functional imaging of epithelial gaps in small intestine of inflammatory bowel disease using confocal laser endomicroscopy.A total of43participants (19UC patients,4CD patients, and20controls) were enrolled. Compared with controls (4.45gaps/1000cells), the mean gap density in the terminal ileum of IBD patients was significantly increased (52.74gaps/1000cells, p <0.001). Compared with normal rats (14.17±4.17gaps/1000cells)., the mean gap density for the IBD ratswas significantly higher (47.83±7.57gaps/1000cells, p<0.001). For IBD treated with VSL#3rats, the mean gap density was significantly lower compared with IBD rats (19.33±4.84gaps/1000cells, p<0.001). Compared with controls, the mean gap densities in rats intraperitoneally injected with IL-1β in the dosages of10and5μg/kg were both significantly increased (205.00±23.48gaps/1000cells and157.67±28.36gaps/1000cells). Compared with normal control group, the expressions level of occludin and ZO-1were increased, and the expression of IL-1β was decreased in IBD treated with VSL#3group (p=0.028, p<0.001, p<0.001).Conclusions:1. CLE is a novel technique that can be used in molecular imaging in vivo with specific EGFR molecular probe in patients with colorectal neoplasia. This new technique shows a promising imaging approach for targeted therapies of CRC.2. Current study demonstrates that CLE can be used in functional imaging of epithelial gaps in vivo. Patients with IBD show increased epithelial gap than control group observed using CLE. Our study shows firstly that probiotics VSL#3could reduce the epithelial gap density in terminal ileum of IBD rat model by decreasing proinflammatory cytokine, and increasing tight junction protein expression.Significance:This study demonstrates, for the first time, that molecular imaging of EGFR is feasible in vivo using a fluorescent-labeled EGFR antibody in combination with CLE imaging in patients with colorectal neoplasia. Specific cellular signal could be observed by CLE molecular imaging. A substantial agreement between in vivo EGFR imaging by CLE and ex vivo EGFR staining by IHC was achieved. Calculation of mean grey scale of the CLE images could differentiate colorectal neoplasia and normal colorecatal mucosa. This new imaging technique could contribute to early detection, differentiatial diagnosis and targeted individual therapy of CRC. Moreover, CLE can be used in functional imaging of epithelial gaps in vivo. Our findings also demonstrate firstly that probiotics VSL#3could decrease the epithelial gap density observed by CLE in terminal ileum of IBD rat model by reducing proinflammatory cytokine, and increasing tight junction protein expression. This study shows new evidences for probiotics treatment in IBD patients.