The evaluation of endoscopic technologies in the investigation and management of colonic neoplasia and obscure gastrointestinal bleeding

© 2010 Dr. Marios EfthymiouThe aim of this work is to evaluate the use of endoscopic technology in colorectal neoplasia and obscure gastrointestinal bleeding. The focus is on the specific evaluation of cold forceps polypectomy, balloon enteroscopy, narrow band imaging and capsule endoscopy. The accuracy of cold forceps polypectomy for diminutive polyps (≤5mm) is assessed in a prospective study using a novel protocol. Patients referred for colonoscopy with one or more diminutive polyps, underwent cold forceps polypectomy followed by endoscopic mucosal resection of the polypectomy site with a 2-3mm resection margin. The two specimens were collected in separate pathology jars and assessed by a pathologist blinded to the study protocol. Comparison between the two specimens allowed for unequivocal assessment of resection efficacy. Fifty-two patients were enrolled and fifty-four polyps were resected. The complete resection rate with cold forceps polypectomy was 39%. Histology was predictive of resection efficacy, with the resection rate for adenomas at 62% compared to 24% for hyperplastic polyps, P=0.008. Polyp size and the number of bites of the forceps were not predictive of resection. A pilot study is then presented, assessing the feasibility of a protocol comparing narrow band imaging and chromoendoscopy, for dysplasia surveillance in patients with chronic colitis. The secondary aim of the study was to assess the accuracy of the Kudo classification (using narrow band imaging) in predicting the neoplastic potential of detected lesions. Twenty-two patients were enrolled in the study. The colon was divided into segments, using the flexures. Each segment was independently examined using narrow band imaging followed by examination with chromoendoscopy. Seven dysplastic lesions were detected in four patients. Both narrow band imaging and chromoendoscopy detected six of the seven dysplastic lesions. Narrow band imaging failed to diagnose one dysplastic lesion detected by chromoendoscopy. Both narrow band imaging and chromoendoscopy detected far more benign lesions than neoplastic. The sensitivity and specificity of the Kudo classification for neoplasia, (using narrow band imaging) was 60% and 79% respectively with a negative predictive value of 95%. Attention is then turned to capsule endoscopy, in a prospective assessment of the accuracy of faecal occult blood tests (FOBT) as a screening test pre-capsule endoscopy in obscure gastrointestinal bleeding. Patients referred for capsule endoscopy underwent two FOBTs within a fortnight preceding capsule endoscopy. The results of the FOBTs were correlated with clinically significant findings on capsule endoscopy. The primary aim of the study was to calculate the predictive value of FOBT in this setting. Fifty-six patients were enrolled. Fifty four percent had occult obscure gastrointestinal bleeding and the remainder overt obscure gastrointestinal bleeding. The overall yield of capsule endoscopy for clinically significant findings was 41%. The combination of immunochemical and guaiac FOBT had only modest sensitivity (63%), specificity (59%), negative predictive value (68%) and positive predictive value (54%). The last project presented in this thesis, is a randomised controlled study comparing single balloon enteroscopy (SBE) and double balloon enteroscopy (DBE) for the investigation and treatment of small bowel conditions. Patients referred for balloon enteroscopy were randomized in 1:1 allocation to single balloon enteroscopy or double balloon enteroscopy. Ninety-nine patients were enrolled and in total, 110 procedures were undertaken, 46 with SBE and 64 with DBE. The primary endpoint was diagnostic yield. Secondary endpoints were therapeutic yield, depth of insertion and procedural characteristics. The diagnostic yield of single and double balloon enteroscopy for small bowel findings was similar at 46% and 41% respectively, P=0.60. Therapeutic interventions were undertaken in 28% of those with SBE and 23% of those with DBE, P=0.57. The depth of insertion was longer for antegrade double balloon enteroscopy, however this was not statistically significant (200cm for SBE and 250cm for DBE, P=0.13). Insertion depth for retrograde procedures was 70cm median for SBE and 85cm for DBE, P=0.86. Despite shorter endoscope setup times with single balloon enteroscopy (3min vs. 11min, P