Search

Summary Background Screening for Barrett's oesophagus relies on endoscopy, which is invasive and few who undergo the procedure are found to have the condition. We aimed to use machine learning techniques to develop and externally validate a simple risk prediction panel to screen individuals for Barrett's oesophagus. Methods In this prospective study, machine learning risk prediction in Barrett's oesophagus (MARK-BE), we used data from two case-control studies, BEST2 and BOOST, to compile training and validation datasets. From the BEST2 study, we analysed questionnaires from 1299 patients, of whom 880 (67·7%) had Barrett's oesophagus, including 40 with invasive oesophageal adenocarcinoma, and 419 (32·3%) were controls. We randomly split (6:4) the cohort using a computer algorithm into a training dataset of 776 patients and a testing dataset of 523 patients. We compiled an external validation cohort from the BOOST study, which included 398 patients, comprising 198 patients with Barrett's oesophagus (23 with oesophageal adenocarcinoma) and 200 controls. We identified independently important diagnostic features of Barrett's oesophagus using the machine learning techniques information gain and correlation-based feature selection. We assessed multiple classification tools to create a multivariable risk prediction model. Internal validation of the model using the BEST2 testing dataset was followed by external validation using the BOOST external validation dataset. From these data we created a prediction panel to identify at-risk individuals. Findings The BEST2 study included 40 diagnostic features. Of these, 19 added information gain but after correlation-based feature selection only eight showed independent diagnostic value including age, sex, cigarette smoking, waist circumference, frequency of stomach pain, duration of heartburn and acidic taste, and taking antireflux medication, of which all were associated with increased risk of Barrett's oesophagus, except frequency of stomach pain, with was inversely associated in a case-control population. Logistic regression offered the highest prediction quality with an area under the receiver-operator curve (AUC) of 0·87 (95% CI 0·84–0·90; sensitivity set at 90%; specificity of 68%). In the testing dataset, AUC was 0·86 (0·83–0·89; sensitivity set at 90%; specificity of 65%). In the external validation dataset, the AUC was 0·81 (0·74–0·84; sensitivity set at 90%; specificity of 58%). Interpretation Our diagnostic model offers valid predictions of diagnosis of Barrett's oesophagus in patients with symptomatic gastro-oesophageal reflux disease, assisting in identifying who should go forward to invasive confirmatory testing. Our predictive panel suggests that overweight men who have been taking antireflux medication for a long time might merit particular consideration for further testing. Our risk prediction panel is quick and simple to administer but will need further calibration and validation in a prospective study in primary care. Funding Charles Wolfson Charitable Trust and Guts UK.

Introduction The EndoringsTM is a distal attachment consisting of two layers of circular flexible rings that evert mucosal folds. Aims; to investigate if Endorings Colonoscopy (ER) improves polyp and adenoma detection compared to standard colonoscopy (SC). Methods This multi-centre, parallel group, randomized controlled trial included screening, surveillance and symptomatic patients. Primary outcome; number of polyps per patient. Secondary outcomes; number of adenomas per patient, adenoma/polyp detection rates and withdrawal times. Results Total of 556 patients (214 females, 342 males) randomized to ER (275) or SC (281). Mean age 67. Colonoscopy completed 532/556 (96%) cases. EndoRings removed in 74/275 (27%) patients. In 66/74 (89%) cases removal was performed due to difficulties with sigmoid intubation. Remainder removed to facilitate retroflexion or polyp removal/retrieval. Total number of polyps in ER limb 571 vs 444 in SC limb. Total number of adenomas in ER limb 361 vs 343 for SC limb. Our study showed a statistically significant difference in the mean number of polyps per patient in both the Intention to Treat (ITT) (1.8 SC vs 2.1 ER, p-value 0.02) and Per Protocol (PP) (1.8 SC vs 2.25 ER, p-value 0.009). There was a trend towards a greater polyp detection rate in the ER colonoscopy (67.5% SC vs 75.2% ER, p-value 0.05). Conclusions Despite the high removal rate of Endorings, there was a statistically significant increase in the mean number of polyps in the ER limb compared to the SC limb. Our study shows promise for the EndoRingsTM device to improve polyp detection.

Background and study aims Up to 25 % colorectal adenomas are missed during colonoscopy. The aim of this study was to investigate whether the endocuff could improve polyp detection in an organized bowel cancer screening program (BCSP). Patients and methods This parallel group, single-blinded, randomized controlled trial included patients with positive fecal occult blood test (FOBT) who were attending for BCSP colonoscopy. The primary outcome was the number of polyps per patient. Secondary outcomes included the number of adenomas per patient, adenoma and polyp detection rates, and withdrawal times. Results A total of 534 BCSP patients were randomized to endocuff-assisted or standard colonoscopy. The mean age was 67 years and the male to female ratio was 1.8:1. We detected no significant difference in the number of polyps per patient (standard 1.8, endocuff 1.6; P = 0.44), adenomas per patient (standard 1.4, endocuff 1.3; P = 0.54), polyp detection rate (standard 69.8 %, endocuff 70.3 %; P = 0.93), adenoma detection rate (standard 63.0 %, endocuff 60.9 %; P = 0.85), advanced adenoma detection rate (standard 18.5 %, endocuff 16.9 %; P = 0.81), and cancer detection rate (standard 5.7 %, endocuff 5.3 %; P = 0.85). The mean withdrawal time was significantly shorter among patients in the endocuff group compared with the standard colonoscopy group (16.9 vs. 19.5 minutes; P Conclusions This study did not find improved polyp or adenoma detection with endocuff-assisted colonoscopy in the FOBT-positive BCSP population. A shorter withdrawal time with endocuff may reflect improved views and stability provided by the endocuff.Trial registered at ClinicalTrials.gov (NCT02529007).

Endoscopic mucosal resection (EMR) is widely practiced in western countries. Endoscopic submucosal dissection (ESD) is very effective for treating complex polyps but colonic ESD in the western setting remains challenging. We have developed a novel technique of knife-assisted snare resection (KAR) for the resection of these complex lesions. Here we aim to describe the technique, evaluate its outcomes, identify outcome predictors and define its learning curve.We conducted a prospective cohort study of patients who had large and refractory polyps resected by KAR at our institution from 2007 to 2013. Polyp characteristics and procedure details were recorded. Endoscopic follow-up was performed to identify recurrence.A total of 170 patients with polyps 20-170 mm in size were treated by KAR and followed up for a mean of 31.5 months (range 12-84 months). 29% of the polyps were50 mm, 22% had fibrosis from previous unsuccessful interventions and 25% were in the right colon. The perforation rate (1.2%) and bleeding rate (4.7%) were acceptable and managed conservatively in most patients. Recurrence rate after the first attempt was 13.1%. Recurrence was significantly increased by polyp size50 mm (p = 0.008; OR 5.03, 95% CI 1.54-16.48), presence of fibrosis (p = 0.002; OR 6.59, 95% CI 1.97-22.07) and piecemeal resection (p 0.001; OR 0.31, CI 0.078-1.12). Cure rates were 87% after the first attempt, improving to 95.6% with further attempts. En bloc resection rate showed a linear increase and reached almost 80% as the endoscopist's cumulative experience approached 100 cases.This is the largest reported Western series on KAR in the colon. We have demonstrated the feasibility, efficacy and safety of this technique in the treatment of complex polyps, with or without fibrosis and at all sites. KAR has shown better outcomes than either EMR or ESD. We have also managed to identify significant outcome predictors and define the learning curve.

Background Currently, various advanced endoscopic techniques are available with varying success rates. These technologies are manufacturer dependent, which has financial implications in the current era of austerity. Acetic acid is a commonly available dye that has been used in the detection of neoplasia within Barrett's esophagus. It has been shown to be effective in detecting neoplasia in high-risk subgroups, but its efficacy in a low-prevalence surveillance population remains unproven. Objective This study aimed to investigate the effectiveness of acetic acid chromoendoscopy in a Barrett's esophagus surveillance population. We aimed to compare the neoplasia yield of acetic acid chromoendoscopy (AAC) with the neoplasia yield from standardized random biopsy (SBP) protocol–guided biopsies in the routine surveillance of patients with Barrett's esophagus. Design Retrospective cohort study. Setting Tertiary referral hospital in the United Kingdom. Patients Patients 18 years of age and older with a diagnosis of Barrett's esophagus undergoing surveillance gastroscopy. Interventions AAC versus standardized random biopsy protocol (SBP) for Barrett's esophagus surveillance. Main Outcome Measurements Neoplasia detection in 2 groups. Results The overall neoplasia detection rates for all grades of neoplasia were 13 of 655 (2%) in the SBP-guided biopsy cohort and 41 of 327 (12.5%) in the AAC cohort ( P = .0001). On per-patient analysis, a 6.5-fold gain in neoplasia detection was seen in the AAC cohort compared with the SBP cohort (0.13 vs 0.02, P = .000). In the SBP cohort, a total of 13 of 655 (2%) neoplasias were detected, of which 3 of 655 patients (0.5%) had low-grade dysplasia, 7 of 655 (1%) had high-grade dysplasia, and 3 of 655 (0.5%) were found to have superficial cancer (T1a/T1b). In the AAC cohort, a total of 41 of 327 neoplasias (12.5%) were found, of which 9 of 327 patients (2.7%) had low-grade dysplasia, 18 of 327 (5.5%) had high-grade dysplasia, and 14 of 327 (4.2%) were found to have superficial cancer. The number of biopsies required to detect 1 neoplasia was 15 times lower in the AAC cohort (40 biopsies) than in the SBP cohort (604 biopsies). On per-biopsy analysis, a 14.7-fold increase in neoplasia detection was seen in the AAC cohort per biopsy compared with the SBP cohort (0.025 vs 0.0017; P = .000). Limitations Not a randomized, controlled study. Conclusions Our study demonstrates that acetic acid detects more neoplasias than conventional protocol-guided mapping biopsies and requires 15 times fewer biopsies per neoplasia detected.

Introduction Up to 25% polyps are missed during colonoscopy. The Endocuff Vision® is a cap with soft flexible arms that attaches to the colonoscope tip and improves views during withdrawal. We have performed the first randomised controlled trial to identify the role of Endocuff Vision® in improving polyp detection. We aim to investigate the impact of Endocuff Vision® assisted colonoscopy on polyp detection, as compared to standard colonoscopy, in the UK Bowel Cancer Screening Programme (BCSP). Methods Single centre, parallel group, randomised controlled trial. Ethics ref: 14/SC/0207. Adopted on UKCRN portfolio (ID: 16985). Patients attending for BCSP colonoscopy were stratified based on attendance for index screening colonoscopy or for polyp surveillance. Within each stratum participants were randomised to either Standard or Endocuff assisted colonoscopy. All procedures were performed by accredited BSCP endoscopists, who have carried out > 5000 colonoscopies and have caecal intubation rates of >90%. Results 534 patients recruited from Sep 2014 to Sep 2015. 3 excluded due to new diagnosis of polyposis syndrome. 531 were included and randomised to the 2 study arms. No significant difference was seen between the 2 groups for the primary endpoint of number of polyps per patient. Secondary endpoints: No significant difference was observed between the 2 groups for adenoma detection rate (ADR) or number of adenomas per patient (Table 1). No significant adverse events were encountered during the study in either arm. The cecal intubation time was not prolonged and patients did not experience any additional discomfort due to the Endocuff Vision. Conclusion In the UK, bowel cancer screening is performed by highly experienced endoscopists. Our results suggest that in expert hands, ADR exceeds 60% even without Endocuff. In such settings, Endocuff Vision did not improve polyp detection rates (PDR) or ADR. However, it did not cause any adverse events, prolong procedure duration or cause additional discomfort. These data demonstrate the safety and feasibility of Endocuff. However, no additional gain was demonstrated in expert hands. Disclosure of Interest None Declared

Introduction Zenker’s diverticulum is a sac-like outpouching of the mucosa and submucosa through an area of muscular weakness between the thyropharyngeus and cricopharyngeus muscles. Traditional treatments have been surgical with open or intraluminal approaches to cricopharyngeal myotomy where the overall complication rate is 9.6%. An alternative approach is to use a flexible endoscope to perform a diverticulotomy. We present a video series to illustrate the principles of flexible endoscopic diverticulotomy. Method We reviewed prospectively collected data of patients with Zenker’s diverticulum referred to our department for endotherapy between January 2014 and January 2015. All patients had Zenker’s diverticulum confirmed on barium swallow. Every procedure was carried out under general anaesthesia by a single endoscopist (PB). A guide wire is inserted into the stomach under direct vision, over which a double-lipped overtube (ZD overtube, ZDO-22–30; Cook Endoscopy, Winston-Salem, North Carolina) is threaded. Under direct vision, the overtube is advanced until the long flap is positioned in the oesophageal lumen and the short flap is in the diverticulum. The septum is clearly visualised and stabilised between the two flaps. The septal mucosa is cut in the middle using a needle knife or diathermy scissors. The mucosal incision is extended until the cricopharyngeal muscle fibres are completely incised. Following this, the cut is extended to approximately 1 cm from the base. Prophylactic endoclips are placed to prevent perforation or bleeding. Patients resume a liquid diet 12 h post procedure. Results A total of 5 patients underwent flexible endoscopic diverticulotomy [female 3, median age 76 years (range 69–84 years)]. 2 patients had previous failed surgical interventions. The mean size of diverticulae was 48 mm (range 30–70 mm) and the mean duration of each procedure was 33 min (range 30–40 mins). 1 patient required an overnight stay as he lived outside our catchment area. There were no procedure related complications or mortality. All patients were able to drink within 12 h and start on a soft diet within 48 h post procedure. Prior to the diverticulotomy, 80% of patients experienced dysphagia with every meal and 60% suffered with regurgitation several times a week. All patients were asymptomatic at follow up at 3. Where 12-month data is available, all patients (3/5) remain asymptomatic. Conclusion In expert hands, flexible endoscopic diverticulotomy is a novel, and safe treatment for Zenker’s diverticulum. It obviates the need for invasive surgery and a majority of patients can be treated on a day-case basis. Initial results are encouraging with no reported complications and excellent short-term success rates. Disclosure of interest None Declared.

Introduction Endoscopic papillectomy is an alternative to radical surgery (10% mortality risk) in the management of ampullary adenomas. Very few centres in the UK have the expertise or experience of performing this procedure and there is certainly no published literature from the UK. We aim to report outcomes after endoscopic papillectomy in a single tertiary endoscopy unit in the UK. Method The prospectively collected data of all patients who underwent endoscopic papillectomy between 2005 and 2015 in Queen Alexandra Hospital, Portsmouth were reviewed. All procedures were carried out by PB (papillectomy) and PG (ERCP) using a standard duodenoscope. The pancreatic and biliary ducts were cannulated. A dilute methylene blue dye was injected into the pancreatic duct prior to papillectomy. Submucosal injection was performed in all cases prior to snare resection of the ampullary neoplasia. 5 Fr pigtail pancreatic stent insertion was attempted in all cases after resection of the neoplasia. Results A total of thirty-five patients were referred for papillectomy but only twenty-two patients (13 female, median age 72 years) underwent a total of 24 papillectomies. En-bloc resection was achieved in 17 patients (77%) with lesion sizes ranging from 8–25 mm (Median 16 mm). Pancreatic stent placement was successful in 82% of all papillectomies. Three patients experienced complications (12.5%); 2 bleeding and 1 acute pancreatitis. There was no procedure related deaths and no one required emergency surgery. There were no local recurrences in 77% of patients. Two patients required 2 attempts to achieve complete clearance of their adenomas. Histology of the resected lesions revealed low grade dysplasia (54.6%), high grade dysplasia (27.2%), cancer (9%), gangliocytic paraganglioma (4.6%), and neuroendocrine tumour (4.6%). Of the two patients who were found to have invasive cancer, the first patient went on to have a pancreaticoduodenectomy and subsequently died of post-operative complications and the second patient was palliated. The total curative resection rate was 86%. Conclusion This is the largest reported UK series of endoscopic papillectomy. Our data demonstrates that this method is a feasible, safe and efficacious means of treating ampullary neoplasia in expert hands. It obviates the need for pancreaticoduodenectomy with its inherent morbidity and mortality. Complications can be serious and expertise is required to deal with them. Disclosure of interest None Declared. References Suzuki K, Kantou U, Murakami Y. Two cases with ampullary cancer who underwent endoscopic excision. Prog Dig Endosc. 1983;23:236–239 De Palma GD. Endoscopic papillectomy: indications, techniques, and results. World J Gastroenterol. 2014;20(6):1537–1543

Introduction Buried Barrett’s’ or Subsquamous Intestinal Metaplasia (SSIM) refers to glands which are ‘buried’ underneath the squamous epithelium. Buried Barrett’s can pose significant diagnostic and surveillance challenges. Buried Barrett’s has mainly been reported in the post ablation context (APC, RFA, PDT). We aim to evaluate its prevalence in patients who are ablation naive, and understand the reasons behind it. Method This is a prospective cohort study. We investigated our Barrett’s database for patients who were referred for endoscopic treatment (EMR) of Barrett’s neoplasia between June 2006 to June 2014. We assessed histology reports before and after endoscopic therapy (EMR), specifically looking for evidence of buried Barrett’s. Biopsy: Biopsies were first obtained from any suspicious looking area. Following this, biopsies were then obtained from the neosquamous area. Finally, random biopsies were obtained. These were sent in separate cassettes. Histopathology was reported by two independent GI pathologists and was prospectively recorded on a central pathology database. Buried Barrett’s was defined as any metaplastic or glandular tissue beneath the squamous epithelium. Pathology specimens were reported by 2 independent, accredited GI pathologists. Results Our study shows that in the pre-EMR cohort, there was an overall prevalence of 12.2% of buried Barrett’s and a 9.1% prevalence of buried Barrett’s with high grade neoplasia (HGD or IMC). Our results in the post EMR cohort shows an overall prevalence of 16.8% of buried Barrett’s with 6.1% prevalence of buried high grade neoplasia (HGD or IMC). This has significant implications for post EMR endoscopic assessment and surveillance. Conclusion Buried Barrett’s and Barrett’s cancer are seen in endotherapy naive patients. This is likely to be related to intensive biopsies. EMR, despite being a non ablation technique, still results in buried Barrett’s and Barrett’s cancer. The overall prevalence of buried Barrett’s is higher than previously reported. We need to be aware of this while assessing Barrett’s patients. Buried Barrett’s glands after ablation (APC/RFA/PDT) are well reported. This is the first study to report on the prevalence of Barrett’s in endotherapy naive patients and in the post EMR cohort. Disclosure of interest None Declared.

Introduction Despite advances in endoscope technology there is still a significant miss rate of neoplastic lesions during gastroscopy. Mucosal views are frequently impaired by residual bubbles and mucus. Method We conducted an RCT in 126 patients attending for routine gastroscopy. Subjects were randomised in a 1:1:1 ratio to receive a pre-procedure drink of water, simeticone and n-acetylcysteine (Group A), water alone (Group B) or no preparation (Group C). Study endoscopists were blinded to group allocation. 4 digital images were taken at pre-defined locations during the procedure – lower oesophagus, upper body, antrum and fundus. Images were rated for mucosal visibility (MV) using a 4 point scale (1 = best, 4 = worst) by 4 blinded experienced endoscopists. Primary outcome measure was mean MV score. Secondary outcome measures were procedure duration and volume of flush required to achieve adequate mucosal views. Results Groups were well matched for age, gender or indication for endoscopy. Mean MV score for group A was significantly better than for group B and group C (p Subgroup analysis of MV scores at each location demonstrated significantly better mucosal visibility in group A compared to group B and group C at all locations (p Conclusion A pre-procedure drink containing simeticone and n-acetlycysteine significantly improves mucosal visibility during routine gastroscopy and reduces the need for flushes during the procedure. This may improve detection of early neoplasia and other pathology during gastroscopy. Subanalysis of separate locations demonstrates significant benefit in the lower oesophagus, demonstrating potential benefit in Barrett’s surveillance procedures. Disclosure of interest None Declared.

Introduction Buried Barrett’s’ or Subsquamous intestinal metaplasia (SSIM) refers to glands which are ‘buried’ underneath the squamous epithelium. High dose acid suppressive therapy and lack of acid exposure can result in squamous re-epithelialisation over the Barrett’s mucosa. Buried Barrett’s can pose significant diagnostic and surveillence challenges. Data on the prevalance of Buried Barrett’s in endoscopic therapy naive patients is limited. Like wise there is limted data on the prevalance of Buried Barrett’s in patients following EMR. We aim to study and compare the prevalence of Buried Barrett’s in these two groups of patients. Methods Inclusion Criteria: Patients with Barrett’s referred for endoscopic treatment between June 06 and June 12 Patients with Barrett’s dysplasia following EMR procedure. Biopsy: Biopsies were first obtained from any suspicious looking area. Following this, biopsies were then obtained from the neosquamous area. Finally, random biopsies were obtained. These were sent in separate cassettes. Histopathology was reported by two independent GI pathologists and was prospectively recorded on a central pathology database. Buried Barrett’s was defined as any metaplastic or glandular tissue beneath the squamous epithelium. Pathology specimens were reported by 2 independent, accredited GI pathologists. Results Conclusion Our study shows that in the pre-EMR cohort, there was an overall prevalence of 15.7% of buried Barrett’s and a 14.5% prevalence of buried Barrett’s with high grade neoplasia (HGD or IMC). Our results in the post EMR cohort shows an overall prevalence of 33.7% of buried Barrett’s with 9.6% prevalence of buried high grade neoplasia (HGD or IMC) suggesting that a third of patients undergoing EMR for Barrett’s dysplasia harbour buried Barrett’s and a third of these patients harbour high grade neoplasia. This has significant implications for post EMR endoscopic assessment and surveillance. The results from our study shows that there is a need to develop and maintain proficiency in sampling techniques in patients with Barrett’s oesophagus. It also shows that the biopsies particularly from those with dysplasia should be carefully reviewed by gastrointestinal pathologists who devote specific attention to identifying buried Barrett’s and buried dysplasia. Disclosure of Interest None Declared

Introduction Cost effectiveness of Barrett’s surveillance has recently being questioned due to the low neoplasia detection rate. Acetic acid chromoscopy (AAC) has been shown to improve neoplasia detection in Barrett’s oesophagus but not in surveillance population. The aim of this study is to compare the effectiveness of AAC with Cleveland clinic protocol (2 cm quadrantic) guided biopsies at detecting high risk neoplasia during Barrett’s surveillance. Methods Prospective Cohort study comparing two different Barrett’s surveillance strategies. All patients who underwent Barrett’s surveillance between 2008–12 were recorded on a Barrett’s database. All neoplasias were independently reviewed by two GI Pathologists. Barrett’s surveillance patients were randomly allocated to acetic acid chromoscopy lists (cohort B) or protocol guided biopsy (Cohort A) lists. AAC involved targeted biopsy of area of concern & 3 additional biopsies from lower, middle & top end of Barrett’s. Protocol guided were taken as quadrantic biopsies every 2 cm & any visible abnormality. Fisher’s exact test was used for statistical analysis. Results N = 982 Barrett’s surveillance gastroscopy between 2008–12. Median age was 66 years & Median Barrett’s length was 4.5 cm (range: 1–20). Male: Female = 3.3:1 . Protocol guided Cohort A N = 655/982(66.7%). 7/655 (1%) patients were found to have high grade dysplasia (HGD) & 3/655(0.4%) had T-1 cancers with an overall high risk neoplasia detection rate of 10/655(1.5%). Acetic acid Cohort B N = 327/982(33.2%). 18/327(5.5%) patients were found to have HGD & 14/327(4.2%) had T-1 cancers with an overall high risk neoplasia detection rate of 32/327(9.7%). This shows a statistically significant 6.5 fold (p = 0.0001) increased detection of high risk neoplasia with acetic acid guided biopsies as compared to protocol guided biopsies in Barrett’s surveillance. Conclusion This is the first report from a large exclusively Barrett’s surveillance population. Our data demonstrates that acetic acid chromoscopy significantly (6.5 fold) improves the detection of high risk neoplasia in Barrett’s surveillance as compared to the current standard of 2 cm quadrantic biopsies. AAC also results in significantly less number of biopsies taken so overall it will be very cost-effective. This questions the validity of the current standard of non targeted protocol guided biopsies during Barrett’s surveillance. Disclosure of Interest None Declared