Update on Endoscopic Research Presented at 2008 Digestive Disease Week
Edited by Jonathan Cohen, MD, New York University School of Medicine
Introduction
The following report is the third annual overview of the endoscopic research presented at Digestive Disease Week as reported by GI fellows from the New York City area training programs. This year, the abstracts selected for oral presentation at 16 ASGE topic fora as well as the scientific papers presented at the ASGE Plenary session are included in this review. Readers will note that this is just a sample of the breadth of innovative scientific work in the area of endoscopic research presented at DDW. However, the abstracts covered in this review should provide a good snapshot of the current areas of investigation—from initial reports of new techniques, to careful controlled trials subjecting new methods to rigorous validation, to presentations of data addressing many of the key clinical practice issues we face today.
Special thanks are due to all of the fellows who contributed the reports that comprise this summary. In particular, the effort of Shashin Shah, MD of Columbia University Medical Center for his help in organizing the effort this year.
Section 1: Barrett’s Esophagus: Detection, Management, and Follow-Up
This forum involved the presentation of 6 abstracts, which covered a diverse spectrum from diagnosis to endoscopic treatment of dysplastic Barrett’s esophagus. There was a particular emphasis on the importance of incorporating newer technologies in the diagnostic and treatment algorithms, including endoscopic resection modalities and cryotherapy.
Prospective Multicenter Study on the Incidence of Neoplastic Progression in Barrett Esophagus Patients
Marjolein Sikkema presented this study, which followed a cohort of Dutch patients to determine the incidence of progression from non-dysplastic Barrett’s esophagus to at least low grade dysplasia (LGD), as well as the incidence of low grade dysplasia progressing to high grade dysplasia (HGD) or early esophageal cancer. Only patients with at least 2 cm of Barrett’s were included in the study. A total of 664 non-dysplastic patients and 199 LGD patients were followed, with 60 non-dysplastic and 20 LGD patients lost to follow-up after 2 years. 37 non-dysplastic patients progressed to LGD (a rate of 3.1% per year) and another 10 progressed to HGD or early cancer (a rate of 0.9% per year). 12 of 99 LGD patients progressed, leading to a rate of 6.7% per year. The risk of progressing from either non-dysplastic Barrett’s or LGD to early cancer was calculated to be 0.8%. An important critique was raised in questioning that it is unclear whether the cases described were incident or prevalent, which can greatly affect progression rates.
A Randomized Prospective Trial Comparing the Cap-Technique and Multi-Band Mucosectomy Technique for Piecemeal Endoscopic Resection in Barrett’s Esophagus
Roos Pouw discussed the findings of this study, whose aim was to compare the safety and efficacy of the two endoscopic resection techniques. Patients with either HGD or intramucosal carcinoma, grade IIa-c were included, but there could be no suspicion or suggestion of submucosal invasion, lymphadenopathy or metastasis. The results showed that while the same average number of resections was required, the multi-band device took less time to complete the resection. The cap and multi-band device achieved the same depth of resection, but the average specimen diameter was slightly larger with the cap method. However, though not statistically different, there were 2 perforations with the cap and none with the multi-band device. The authors concluded that the multi-band device may be preferred for flat lesions without risk of submucosal invasion, while the cap method appears better for nodular lesions or those with a higher risk of submucosal invasion.
Study of Preoperative and Postoperative Pathology in High Grade Barrett’s Dysplasia
Robert Schoen presented this paper from the University of Pittsburgh, which examined whether our ability to detect esophageal adenocarcinoma (EAC) improved with technological advancements over the past 14 or so years. The charts of all patients who underwent esophagectomy for HGD from 1993 to 2007 were reviewed by the research team in order to determine the prevalence of EAC. 12 patients, or 18%, were found to have concurrent EAC, while 2 patients were “down-graded” to LGD. The other 80% had confirmed HGD on their esophagectomy specimens. Of note, 10 of the 12 patients with EAC were determined to be “early stage” (though intramucosal carcinoma and T-Ia lesions were not included in this category). 15% of 1993-2000 specimens and 19% of 2001-2007 specimens contained EAC, a non-significant difference. Similarly, 20% of 1993-2003 and only 14% of 2004-2007 specimens had EAC, also non-significant. The authors concluded that detection rates for early EAC had not improved with time.
Are Patients with “Low Risk” Submucosal Invasion of Early Barrett’s Carcinoma Eligible for Curative Endoscopic Therapy? Outcomes of Endoscopic Therapy and Surgery in 80 Patients with Suspected or Definite Diagnosis of Submucosal Barrett’s Cancer
Hendrik Manner presented the results of this study, which explored whether endoscopic therapy could achieve results equivalent to surgical outcomes for “low risk” esophageal cancers that extended into the submucosa. A total of 21 cases met the criteria for a “low risk” lesion on initial diagnostic endoscopic resection, namely invasion only into the most superficial third of the submucosa (SM1) without evidence of lymphatic or vascular involvement. Of the 21 patients who met criteria, 19 participated in the study and underwent endoscopic resection with the goal of obtaining complete local remission (CLR). 18 of 19 patients achieved CLR, and the other patient was referred for surgery after 2 sessions of endoscopic resection failed to achieve CLR. Intensive surveillance for a mean of 69 months followed, and secondary lesions were found in 28% of the 18 patients, who were then treated endoscopically if possible. The 5 year survival rate for these patients was 66%.
Endoscopic Cryotherapy Ablation is Safe and Well-Tolerated in Barrett’s Esophagus, Esophageal Dysplasia, and Esophageal Cancer
Bruce Greenwald presented this abstract to discuss early observations with a low pressure liquid nitrogen spray being used to ablate Barrett’s esophagus. The aim of the study was to evaluate the safety and tolerability of a large cohort of patients receiving this treatment. Patients with non-dysplastic Barrett’s, LGD and HGD were included. After receiving treatments with either 10 or 20 second pulsed cycles of treatment, patients underwent self-assessment in post-op days 1 to 7. A total of 77 patients were treated, and 29% reported no side effects (a total of 48% of all procedures). Chest pain was the most common reported complaint (present in 18% of patients), followed by dysphagia, odynophagia and sore throat. Most were characterized by patients as being of “mild” severity. Dysphagia lasted about 5 days, with other complaints present for fewer days. Side effects were more likely if Barrett’s was greater than 6 cm in length. More significant side effects were seen in only 4 patients: 1 patient with Marfan’s syndrome developed a lower gastric perforation, thought to be due to overinflation of the stomach and not the actual cryoablative therapy. 3 others developed a stricture, and all resolved with endoscopic dilation. The authors concluded that cryotherapy was generally as safe and well tolerated (if not better) than other ablative therapies, and deserves further study.
Low Rate of Invasive Cancer in Patients Undergoing Mucosectomy for Barrett’s Esophagus (BE) and High Grade Dysplasia (HGD) or Intramucosal Cancer (IMC)
Vani J. Konda presented the results of this study, a retrospective review of a prospectively collected database to evaluate the utility of endoscopic resection for eradication of advanced neoplasia. Lymphatic channel invasion has emerged as a risk factor for metastatic spread, and previous studies have shown an incidence of occult invasive cancer of 12%, with a higher risk of occult cancer if visible lesions are present. 50 patients were included in the study, 37 with HGD and 13 with IMC (though 2 more IMC were picked up on resection specimens). Of these, 6.4% of visible lesions had submucosal invasion, and another 6.4% had lymphatic spread. There was no evidence of invasive adenocarcinoma on any of the surveillance endoscopic resections or biopsy specimens obtained. The overall rate of metastatic spread was 4%, which was much lower than previously published data showing rates of around 40%. The authors concluded that, given this low rate of occult invasive disease, endoscopic resection is an appropriate therapy for removal of HGD or IMC.
Evaluation of the Diagnostic Accuracy Rate of Minute Colonic Adenomas: High-Resolution Magnifying Chromoendoscopy vs. Histopathology of a Biopsy Forceps Specimen
This prospective study presented by Tamura from Kochi, Japan identified 255 adenomas 3mm or smaller. High resolution magnifying chromoendoscopy with 0.2% indigo carmine was used to identify type III Kudo pit patterns, which in previous studies has been shown to correspond to adenoma on histology. They were then resected with biopsy forceps and separated into two groups- half were placed directly in fixative (20% buffered formalin), while the other half were flattened using the forceps before being placed in fixative. Adenoma was diagnosed histologically in 100% of the “flattened” group vs. Only 85% of the un-flattened group. From this latter group, the 19 specimens initially diagnosed as non-neoplastic were fully sectioned (i.e. The complete paraffin wax blocks) and reexamined, and 17/19 (89%) were reclassified as adenomas. The authors concluded that high magnification chromoendoscopy is more accurate than standard histopathology in the evaluation of very small colon polyps.
Section 2: Improving the Quality of Your Colonoscopy Practice: How Can We Do It Better?
This session focused primarily on improving the efficacy and safety of colonoscopy by various changes to our current practices; in addition, the one upper gastrointestinal study evaluated the use of fibrin adhesion for acute ulcer bleeding. Three of the studies emphasized changes to and quality in bowel prep as a means of improving the quality of colonoscopy. One study evaluated the use of a transparent cap attached to the tip of the scope in decreasing sigmoid looping and patient discomfort during colonoscopy. Another study evaluated a UK accreditation program for ensuring competent and safe colonoscopy. The last study confirmed that changing old habits is hard as shown in sustaining improved withdrawal times of less than 6 minutes to conform with current guidelines.
Duration of the Interval Between Completion of the Bowel Preparation and Start of the Colonoscopy as a Better Predictor of Bowel Preparation Quality than Colonoscopy Start Time Alone
This study presented by Kenneth Yang aimed to study whether the quality of bowel preparation is affected by the interval between the end of bowel prep and the start of the colonoscopy. The investigators prospectively gathered data on 378 patients undergoing outpatient colonoscopies. Compliance with the bowel preparation and dietary restrictions, and the time since the last dose of bowel prep agent was recorded immediately prior to colonoscopy. The quality of the prep was measured in the right colon on a five-point scale by the endoscopist and endoscopy nurse. The investigators found no significant difference in bowel prep quality between morning and afternoon colonoscopies or the duration of time from the last solid meal to the start of the colonoscopy. They found that in patients with excellent/good bowel prep quality, the interval between the time of the last dose of bowel prep agent and the start time of the colonoscopy was significantly shorter than in those whose preparations were scored fair/poor/inadequate (p = 0.013). The authors concluded that bowel prep quality worsens as the interval from the end of bowel prep and start of colonoscopy (shorter duration from bowel prep to colonoscopy leads to better prep quality).
Improving the Quality of Colonoscopy in term of Visibility by adding Simethicone to Sodium Phosphate for Bowel Preparation
Sasinee Tongprasert presented this study evaluating the effects of simethicone (an antifoaming agent used to reduce gas bubbles) on visibility and efficacy during colonoscopy. One hundred and twenty-four patients were enrolled and randomized in a prospective, double-blind, placebo-controlled study. All patients received a 1 day liquid diet and were randomized to receive either sodium phosphate (45ml) plus simethicone tablets (240mg) or sodium phosphate (45ml) plus placebo, dosed twice, once the evening before and once the morning of the procedure. Visibility and haziness were graded by a single, blinded investigator. Endoscopist satisfaction was evaluated by a questionnaire and patient satisfaction was measured by a visual analog scale . There were no differences in success rates, total duration of colonoscopy and side-effects of medications. The study showed improved visibility by diminished presence of air bubbles in the sodium phosphate plus simethicone group compared to the sodium phosphate plus placebo group (1.6% vs. 57.6%, p<0.0001). Haziness was not shown to have been diminished in the simethicone group (9.8% vs. 18.6%, p=0.15). The study also showed improved endoscopist (95.1% vs. 42.4%, p<0.0001) and patient satisfaction (8.7+1.8 vs. 7.5+1.9 p=0.0004). The authors concluded that the addition of simethicone to a sodium phosphate prep is beneficial to bowel preparation by diminishing air bubbles, therefore enhancing visibility and improving the quality of colonoscopy. In addition, simethicone also heightened endoscopist and patient satisfaction.
Accrediting competence in colonoscopy: validity and reliability of the UK Joint Advisory Group/NHS Bowel Cancer Screening Programme accreditation assessment
Roger Barton presented this study evaluating the validity and reliability of the UK Joint Advisory Group (JAG)/ NHS Bowel Cancer Screening Programme accreditation assessment in colonoscopy. The assessment tool was created to select the screening program’s future colonoscopists and consists of performance data, a knowledge test (MCQ) and a Direct Observation of Procedural Skills (DOPS) assessment. DOPS forms from two national training centers were reviewed and important colonoscopy skills were identified into domains (i.e., awareness of patient’s consciousness and pain, adequate mucosal visualization, loops). Detailed descriptors for all skill levels were written for each domain. Performance data, demographic data, MCQ and DOPS over two cases by two assessors were collected prospectively from candidates. Questionnaires on the validity of the assessment were also collected from candidates and assessors. 3000 paired judgments on over 100 candidates during more than 200 cases were reviewed and there was 96% congruence of agreement across the pass/fail divide (98% for major domains). Absolute congruence across individual domains between assessors was between 60-83%. Of the candidates, 93% felt the DOPS assessment was overall fair or very fair, while 86% felt the MCQ was fair or very fair. Of the assessors, 75% felt the DOPS was valid or very valid, while others felt it was not testing enough nor reliably assessing therapy. 100% of assessors overall felt the process was fair or very fair. The author concluded DOPS accreditation has good validity and high reliability across the pass/fail divide. This study stands out as an important step towards the development of reliable tools to objectively assess competency among trainees.
Evaluation of Patient Pain During Colonoscopy for Types of Looping Between Hood Attached Group and Non-Hood Attached Group: An Analysis Using Magnetic Endoscope Imaging (Mei)
Koichiro Sato presented this study evaluating patient pain and looping during cecal intubation with an oblique transparent cap (OTC) using magnetic endoscope imaging. 483 consecutive patients were randomized to undergo colonoscopy with or without the OTC by two experienced endoscopists who were blinded to the magnetic imager view. Patients with colonic resections, poor bowel preparation and severe strictures in the colon were excluded. No sedation was utilized during the colonoscopies and the patients completed a visual analog scale assessing the degree of pain associated with intubation of the cecum. With the magnetic endoscope imaging, the degree of pain in each type of loop and frequency of loop formation in the sigmoid colon was retrospectively analyzed. Rate of completion was the same and cecal intubation times were similar between the two groups (5.6 minutes with OTC and 5.9 minutes without OTC, p=0.241). Examinations were significantly less painful with the OTC than without (22.8mm vs. 27.1mm on VAS, p=0.014) and there was less looping in the sigmoid colon in the OTC group compared to the non-OTC group (132/242 vs. 183/241, p=0.0246). The degree of pain was significantly higher in reverse alpha loop (53.9mm, p<0.001), N loop (33.1mm, p<0.001), and alpha loop (38.3mm, p<0.001) than in non-loop (9.8mm). The authors conclude that by attaching an OTC to tip of the colonoscope (which maintains the distance between the lens and lumen) there is less looping (N, alpha, reverse alpha) resulting in less painful examinations compared to conventional colonoscopy.
Colonoscopy in the Elderly: Worse Preparation, Better Tolerability
This study presented by Kinesh Patel aimed to study the effectiveness of colonoscopy for complete colonic examination, defined by cecal intubation, in patients over 75 years of age. All colonoscopies performed in a teaching hospital (17,926 colonoscopies) were analyzed for rates of completion and any reasons for failure were categorized and recorded. In patients under the age of 75, 14.1 % had incomplete colonoscopies compared to 24.6% in patients over 75 (p<0.001). Reasons for failed colonoscopies included poor prep (8.6% in patients over 75 vs. 4.5% under 75, p<0.0001), technical problems, organic disease and patient discomfort (1.9% in patients over 75 vs. 3% in those under 75, p=0.001). The authors concluded that poor bowel preparation is the main factor hindering completion of colonoscopy in the elderly and further studies are needed to optimize bowel preparation to improve the efficacy of colonoscopy in this population where organic pathology is common. Also of note is that, contrary to prior beliefs, colonoscopy is less frequently aborted secondary to poor tolerance and better tolerated in the elderly compared to younger patients.
Treatment of Acute Ulcer Bleeding in the Upper Gastrointestinal Tract with Fibrin Adhesion
The only upper gastrointestinal study in this session was presented by Deniz Uyak. The study aimed to examine the efficacy of fibrin adhesion in the management of acute ulcer bleeds of the upper gastrointestinal tract. A total of 266 patients underwent endoscopic fibrin adhesion therapy between 1992 and 2005. Bleeding was classified using the Forrest classification system; 9.7% had ulcer with active bleeding, spurting vessel (Ia), 48.5% had an oozing ulcer (Ib) and 42.8% had an ulcer with visible vessel, clot or flat spot (II). Definitive initial hemostasis was achieved in 86% of patients and 93.3% after a second endoscopic intervention. Rebleeding after fibrin adhesion therapy overall was noted in 29 patients out of 266 (10.9%); 42.3% in Ia, 12.4% in Ib and 1.8% in II. Surgical intervention was required in 11 patients (4.1%); 30.7% with Ia ulcers, 2.3% with Ib and none with type II ulcers. Total case fatality was reported as 3.7%, 11.5% with Ia, 4.7% with Ib and 0.9% with II. Autopsies on 2 of the 4 patients who died following surgical intervention revealed pulmonary embolism and multi-organ failure as the cause of death. Of the 6 out 255 who did not require surgical intervention, autopsy on 4 revealed the cause of death as pulmonary embolism (1), coronary occlusion (1), overlooked bleeding ulcer(1) and small intestinal bleed (1). The author concluded that fibrin adhesion therapy yields good results in the treatment of acute gastroduodenal ulcers, particularly type Ib and II bleeds. Prospective trials comparing its use to standard modalities in terms of efficacy and cost will be needed.
Changing Colonoscopy Performance to Meet the Quality Indicator of Withdrawal Time: Results of a Clinic-Based Improvement Intervention
This study presented by Kirsten Weiser aimed to assess the sustainability of improved withdrawal times (≥ 6 minutes as per current guidelines) after instituting interventions including education, measurement, and feedback, by applying principles of continuous quality improvement. Baseline withdrawal times (WT) during screening colonoscopies were recorded by nurses over a two month period, without endoscopist knowledge. During an initial academic detailing session, the evidence for WT ≥ 6 min was reviewed. Baseline WTs were discussed with the endoscopist, and subsequent WT were measured and reviewed by the endoscopist after the procedure (rapid feedback loop). Summary data and an update on performance was provided to the endoscopists on a weekly basis. Further education interventions were made at the end of each week, as necessary, to improve WTs. No further changes were made after the 4th week. WTs were measured during weeks 4 through 8 and were available to the endoscopist, if requested. At baseline, WT in 72.1% of colonoscopies was ≥ 6 minutes with a mean WT of 7 minutes 19 seconds (± 3:08 min). At weeks 3 and 4, with 91% and 93.5% of colonoscopy WT ≥ 6 minutes. At the end of the 8 week period, there was no significant difference in WT compared to baseline. The authors concluded that although interventions including education, measurement and immediate feedback influenced endoscopist performance, there was a rapid regression to mean and non-durable results. Sustained improvement likely depends on interventions that provide continuous intra-procedure feedback, such as a timer. This study provides compelling evidence, however, that the very act of measuring one’s performance can impact that performance. As such, it illustrates one way in which objective self-assessment can be beneficial, if sustained.
Section 3: Pediatric Endoscopy
This session included six papers on different pediatric GI issues. The topics included per-oral suturing for adolescent obesity, ketamine sedation, natural history of esophageal polyps, capsule endoscopy retention, impact of capsule endoscopy on clinical management, and correlation of optical coherence tomography with histology in celiac disease diagnosis.
Per-Oral Suturing As a Potential Treatment for Adolescent Obesity - A Pilot Investigation of 12 Patients with 6-Months Follow-Up
It was a pilot study of 12 adolescents between 14-17 years of age who underwent Endoluminal Vertical Gastroplasty (EVG) for obesity. During EVG, Endocinch was used under general anesthesia to appose anterior and posterior gastric walls in order to reduce the stomach volume. The average baseline BMI was 38.1 ± 7.7 with weight range 80-158 kg. There were no procedure-related complications reported. The data was analyzed at 1, 3, and 6 months. All patients lost weight. The average weight losses was 18%, 38%, and 60%; and mean BMI was 35, 32 and 27.5 at 1, 3 and 6 months follow-up. Hence, EVG may be an alternate to invasive surgery. However, longer follow-up is needed.
Ketamine Versus Midazolam/Fentanyl Sedation for Pediatric Endoscopy: Comparison of Patient Movement, Need for Restraint and Vocalization of Distress
This study compared ketamine vs. midazolam/ fentanyl sedation in children undergoing both upper and lower endoscopy. A cohort of 17 children undergoing ketamine sedation (1 mg/kg/bolus dose, max bolus dose 70mg, max # boluses 2) from 3/06-9/07 was compared with 20 children who received midazolam (.05-.3 mg/kg IV, max dose 15 mg) and fentanyl (1-5 μ/kg, max dose 250 μ) from 12/03-11/04. The Ohio State University Behavioral Rating Scale was used as a validated measure of patient sedation. The researchers found out that children undergoing ketamine sedation exhibited body movements and required restraint for the same percentage of time as those with midazolam/ fentanyl. However, children were less likely to vocalize distress with ketamine sedation. Large trials are needed to further compare the use of ketamine sedaion to standard regimens in the pediatric population.
Esophageal Polyps in Pediatric Patients Undergoing Upper Gastrointestinal Endoscopy
This was a retrospective chart-review performed to analyze the incidence and natural history of esophageal polyps in children <21 years of age undergoing routine EGD. Out of 9093 EGD’s performed at three institutions, only 12 patients had esophageal polyps (0.13%). The mean age was 8.7 years, and the majority were male (9). The presenting complaints varied: vomiting (41%), upper GI bleed (25%), abdominal pain (175), and failure to thrive (17%). The majority (7 of 12) were located at GEJ. The major histologic subtype was inflammatory (58%); followed by squamous papilloma (17%), hemartomatous (8%), and indeterminate (7%). The majority (75%) of them had esophagitis, though this may reflect that asymptomatic children were not included in this series. On repeat EGD, they tend to persist overtime, and might not respond to anti-reflux medications. This study provides limited insight onto the clinical significance of these lesions.
Wireless Capsule Retention in Pediatric Patients - What We Have Learned
It is the largest wireless capsule endoscopy (WCE) study reported to date that analyzed the incidence and risk factors for capsule retention in pediatric patients. The data was gathered retrospectively on 209 WCE procedures performed from 01/200 to 11/2007 in pediatric population with a mean age of 14.7 years (range 8-23). All patients swallowed the capsule. The mean follow-up was 2 months. Capsule retention occurred in only 3 of 209 procedures (1.4%). All had known Crohn’s disease (CD). Only 1 patient required surgery. The risk of retention with known CD was 5.1% (3 of 59). The risk increased to 37.5% with known CD and abnormal small bowel series. Despite the retrospective study design, these data support the practice of obtaining small bowel radiologic imaging prior to capsule endoscopy examination in children with suspected CD.
Impact On Management of Small Intestinal Disorders By Wireless Capsule Endoscopy in Young Children Less Than Age 8: A European Multicenter Study
This European multicenter study analyzed the impact on clinical management of small bowel disorders by wireless capsule endoscopy (WCE) in young children <8 years of age. 70 patients with mean age 5.5 years (range 1.5-7.9) were included. The capsule was delivered by swallowing (if ≥4 years old), via Advance[US endoscopy], or using a Roth Net for delivery. WCE changed the management in overall 57% of the cases. While positive findings altered management in approximately 50% of those with suspected Crohn’s disease or GI bleeding, both positive and negative findings influenced the management in 90% of children with recurrent abdominal pain. This is a notable difference from the clinically relevant findings when WCE has been utilized in adult populations for the indication of unexplained abdominal pain.
Evaluation Study On the Correlation Between Optical Coherence Tomography and Histology in the Diagnosis of Celiac Disease in Pediatric Patients
This prospective study analyzed the correlation of optical coherence tomography (OCT) with histology in diagnosing Celiac disease (CD) in pediatric population. OCT uses infrared waves to visualize villous morphology. Out of total 134 patients, 67 had positive serology for CD, while 67 were controls. Multiple images in 4 quadrants from descending duodenum were taken; followed by biopsies from the same area. The villous morphology was categorized as either no atrophy (pattern 1), mild atrophy (pattern 2), or marked atrophy (pattern 3). The overall concordance rates were 91.6 and 100% in suspected CD and controls respectively. In suspected CD children, the concordance rates for pattern 1, 2, and 3 were 100%, 94% and 92% respectively. Hence, OCT can be helpful in diagnosing CD. It is unclear how much better this modality is in assessing small bowel villi when compared to careful endoscopic examination of folds and villi, particularly with improved high resolution endoscopy.
Section 4: Colonoscopic Screening and Polypectomy
This section focused on colonoscopy quality measures. Studies focused on how well polyps both large and small are completely resected and how colonoscopy compares to screening with FOBT. Several studies investigated how differences in withdrawal time affect polyp detection rate and interval development of neoplasia on subsequent surveillance examinations and if monitoring withdrawal time can influence an endoscopist’s withdrawal time and/or polyp detection rate.
Quality Polyp Resection in Colonoscopy: Are we Achieving Polyp Clearance?
The authors evaluated the resection quality of 31 polyps in 22 patients. Polyp sizes ranged from 2-13 mm. Polyps were removed with polyp biopsy (standard or jumbo forceps) or snare polypectomy. The polyp perimeter and polypectomy base were biopsied to assess whether the polyp was completely resected. 6 polyps (19%) were incompletely removed, 5 out of 6 of these polyps were removed with standard biopsy forceps; there was no clear association between size of polyp and incomplete resection. The authors concluded that standard biopsy forceps may not allow for complete polyp resection and that alternative techniques may improve polyp resection. One limitation of this study is that the authors did not comment on the use of hot biopsy forceps or the use of hot or cold snare polypectomy technique. While the clinical implications of incomplete resection of these small polyps is unclear, this small study raises questions about the contribution of residual adenoma in the development of interval cancers during screening.
Clinical and Economic Impact of a Tertiary Referral Colonoscopic Polypectomy Service (TRCPS)
The authors studied the impact of creating a referral service for endoscopic resection of large sessile colonic polyps that were not able to be completely removed by experienced referring gastroenterologists (average 15 years in practice) as an alternative to surgical resection. 110 sessile lesions found in 98 patients were removed endoscopically (piecemeal resection for lesions >25 mm). 94% of patients had adenomatous polyps (30% with high grade dysplasia). Seven patients [ 6%] of patients had invasive cancer which was completely removed by endoscopic mucosal resection (2 out of these patients had no cancer detected at subsequent surgery, and 5 elected not to have surgery and are being followed). No perforations occurred, 5 patients required overnight admission, 9 patients had bleeding controlled during the procedure, and 5 had delayed bleeding; 12 patients went straight to surgery without colonoscopy mostly because of concern for cancer. Overall 87.6% of patients avoided surgery. According to Australian cost estimates, this strategy of endoscopic resection of large sessile polyps instead of primary surgical resection provided cost savings of approximately $11,000 per patient. This data established the feasibility and safety of such a tertiary referral practice, and follow up colonoscopy data from this cohort will be needed to support the apparent excellent complete removal achieved.
Mandated Screening of All Veterans for Colorectal Cancer: The FOBT Versus the C-Scope
This study compared FOBT to colonoscopy in asymptomatic age >50 year old veterans who were referred for colonoscopy. 7% of patients were referred secondary to positive FOBT done as a screening test and all patients were given FOBT cards prior to colonoscopy to compare the yield and “miss rates” of FOBT compared to colonoscopy. FOBT detected CRC in 8% of positive patients; 1.5% of colonoscopy patients had CRC diagnosed. FOBT missed 53% of CRC’s. 93% of CRC patients diagnosed by colonoscopy were early stage (Stage I/II) compared to only 72% of positive FOBT patients. 70% of the CRC patients with advanced stage detected by positive FOBT were dead at an average of 18 months, this survival was significantly worse than patients with advanced stage diagnosed by screening colonoscopy, independent of patient age. The authors concluded that FOBT is no longer a viable screening option.
Colonoscopy Withdrawal Time and Risk of Neoplasia on Follow-Up Colonoscopy: Results from VA Cooperative StudyProgram 380
The authors examined the effect of colonoscopy withdrawal time (WT), namely to test the hypothesis that WT is inversely associated with interval neoplasia on subsequent colonoscopy. 1193 out of 3121 patients who had undergone screening colonoscopy returned for surveillance colonoscopy within 5.5 years. Of patients with no neoplasia on initial exam, 16% developed neoplasia on repeat colonoscopy (one with invasive cancer). There was no difference between baseline WT in these subjects with and without neoplasia on repeat exam. On a center level analysis, WT’s correlated with adenoma detection rate on baseline colonoscopy, but not advanced adenomas. WT at baseline was not correlated with finding interval adenomas or advanced adenomas. One explanation given for these findings in discussion at the symposia was the fact that most endoscopists had an acceptable WT and therefore increases in WT above what is recommended may not give further improvement in adenoma/advanced adenoma detection rates.
Simply Recording Colonoscopy Withdrawal Time Increases Polyp Detection Rate
Given recommendations that WT should be 6-10 minutes on average for colonoscopies without intervention in order to maximize polyp detection rate, the authors examined if simply knowing WT would be measured would increase the a physician’s polyp detection rate. Colonoscopies performed 5 months before (Group A) and 5 months after (Group B) WT began to be recorded were examined to determine differences in polyp size, number, location and histology. Polyps were found in 34% of patients in Group A and 40% in Group B, relative increase of 14% (p=0.002). Average WT was 11 minutes in Group A and 14 minutes in Group B, full statistics not given. Data on histology, location and polyp size is pending. There appears to be a small increase in polyp detection rate with recording WT, however it is not known whether this increase will yield more adenomatous polyps. This study also confirms the finding that knowing withdrawal time is being measured leasds to longer withdrawal time, supporting calls for more widespread measurement of procedure related performance data.
The Effect of Periodic Monitoring on Screening Colonoscopy Withdrawal Times, Polyp Detection Rates and Patient Satisfaction Scores
This study sought to assess the effect of monitoring and feedback on WT, polyp detection rate and patient satisfaction and to determine the relationship between WT, polyp detection rate and patient satisfaction. The authors hypothesized that shorter WT would be a marker of a hurried doctor-patient relationship. Colonoscopies performed by 10 experienced gastroenterologists before and after WT was monitored with feedback given were examined. There was a statistically significant increase in negative colonoscopy WT from 6.8 to 8.4 minutes (p=0.03) before and after monitoring, respectively and a small non-significant increase in polyp detection rate (20 to 23%). There was no effect on WT and patient satisfaction surveys; there was little difference in patient satisfaction between gastroenterologists. There was a wide range in WT between gastroenterologists (5.3 to 9.9 minutes) and advanced neoplasia detection rates (2.4% to 7.4%). WT >6 minutes was associated with a higher advanced neoplasia detections rate (6% vs. 2.8%, p=0.04). The authors conclude that monitoring and feedback does increase WT and that WT is correlated with advanced neoplasia detection rate, but does not impact patient satisfaction. Of course if patients knew that their doctors spent more time looking for polyps, or even measured their own performance to optimize quality, a large potential benefit on patient satisfaction could be realized.
Section 5: Upper GI Bleeding: Current Therapeutic Trends
This session explored the use of currently available as well as novel technologies in the treatment of acute upper gastrointestinal bleeding.
For Severe UGI Hemorrhage Doppler Ultrasound Probe Is More Accurate for Risk Stratification and Helpful for Complete Endoscopic Hemostasis Than Lesion Stigmata Alone
Dennis Jensen presented this abstract which evaluated the use of a Doppler ultrasound probe (DUP) for evaluation of active or recently bleeding lesions. He mentioned that for 30 years, we have risk stratified ulcers based on visual cues, but actual rebleeding rates are based on below-surface pathology. 79 patients were enrolled in this study, whose aim was to determine the prevalence of underlying blood flow in high risk UGI bleeding and follow the patients for 30 days. The DUP used in this study was passed through the endoscope channel and placed adjacent to the lesion for evaluation. 67% of the patients had an ulcer, while 14% had a portal hypertension-associated lesion. Of the visually “benign” ulcers, 49% had a positive DUP exam, which often led to the discovery of previously missed visual findings. 30 patients were found to have major stigmata, and 20% of these patients were DUP positive even after treatment. Of the 23 patients with a negative DUP, none rebled. The overall rebleeding rate for DUP-guided hemostasis was 5.2%, compared to 29.7% for the patients who did not have DUP as part of their procedure.
Initial Evaluation of a Novel Prototype Forward-Viewing Echoendoscope in a Porcine Arterial Bleeding Model
Michael Pollack discussed this abstract, which assessed an Olympus prototype forward-viewing endoscope which contained a 90 degree range linear array ultrasound in line with a therapeutic channel. This set-up allowed for simultaneous visualization of both the bleeding site as well as the vessel deep to the surface. The aim of the study was to determine if the equipment could visualize a vessel in a porcine model, allowing for real time therapy. The bleeding model was prepared by severing an artery feeding the stomach and tunneling it into the submucosa. Hemostasis was obtained by injecting the vessel using a sclerotherapy needle, followed by bipolar electrocautery, leading to a cessation of flow on ultrasound. 7 total cases were performed as part of the study, with therapy successful in 2 of 4 cases. The greatest challenge to using this technology, according to the authors, was maintaining direct acoustical contact between the probe and the gastric wall.
Prokinetic Agent Metoclopromide Increases the Yield of Upper Endoscopy and Decreases the Need for Second Look Endoscopy in Patients with Upper Gastrointestinal Bleed: a Randomized Study.
This abstract was not presented during the session. It was a randomized controlled study of IV metoclopramized along with pantoprazole and octreotide vs. a solution without metoclopramide. 100 patients were randomized. The protocol used 10 mg of IV metoclompramide every 4 hours for 1-2 total doses. The use of IV metoclopramide was associated with several significant advantages: increased finding of each of the major stigmata of bleeding ulcers, increased hemostatis success, decreased need to performe a second lok endoscopy, decreased need for further transfusions, and decreased length of hospital stay form 5-6 days as compared to 1-2 weeks.
Does Endoscopic Therapy of Bleeding Peptic Ulcers Influence Need for Repeat Endoscopy?
Brintha Enestvedt presented this paper, which aimed to evaluate the frequency of repeat EGD for a bleeding ulcer. 4 years of data from the CORI database were reviewed, with a focus on the type of intervention utilized in the initial procedure. A repeat bleed was defined as occurring in less than 72 hours, and clean-based ulcers were excluded given their very low rate of repeat bleeding. Of 3,792 patient records where ulcers were seen on an EGD, 40% did not have a clean-based lesion and remained in the analysis. Of note, monotherapy was used in 23% of actively bleeding ulcers, and 12% of visible vessel cases. 120 total patients had a repeat bleed, where 31% of ulcers were clean-based and 28% were actively bleeding. Patients who were treated initially only with epinephrine injection had a rebleeding rate of 12.2%, twice that of thermal monotherapy or combination therapy. Similarly, 39 actively bleeding lesions required a repeat EGD, and monotherapy with injection again had a higher rate of rebleeding. The authors discovered that more than 95% of lesions with high risk stigmata received only therapy with 1 treatment modality. They believe this study strongly suggests avoiding monotherapy with injection and strongly considering bimodal therapy to decrease rebleeding rates. The use of the CORI database to ascertain physician practice patterns and, as in this case to indicate a potential area in which such a practice—injection monotherapy—might be leading to worse outcomes is a helpful adjunt to existing controlled efficacy trials to influence future endoscopic practice.
A Method for Endoscopic Obturation of Gastric Varices That Maximizes Effectiveness Without Risks of Systemic Embolization
Bimaljit Sandhu discussed the results of this prospective pilot study, which aimed to minimize the risk of systemic passage of cyanoacrylate after injection into a gastric varix and before coagulation. 30 patients were enrolled after informed consent was obtained, including discussion of the alternative TIPS procedure. The initial goal after the endoscopy began was to attempt to identify the culprit vessel, then inject at the origin of blood flow to minimize the risk of systemic spread. The authors discussed finding certain patterns which led them to the origin points. Following injection, the varix was probed to see if it had hardened or still was soft, in which case it was injected a second time. After 4 weeks, a second EGD was performed, at which time the gastric varix again was injected if it had not hardened. The trial was deemed a success, as after 11 months of follow-up, only 1 patient had any repeat bleeding.
Section 6: Advances in Endoscopic Practice: Maximizing Patient Safety and Comfort.
The papers presented in the forum covered a variety of ways that the endoscopist can monitor, assess and prevent adverse events during our procedures. A total of five abstracts were presented representing international research on a variety of topics including but not limited to capnography, sedation, feeding after PEG placement, and perforations during EUS.
Capnography Prevents Hypoxemia During ERCP and EUS: A Randomized Controlled Trial.
Hypoxemia is one of the most common cardiovascular complications of EUS/ERCP which can occur 40-70% of patients. Usually hypoventilation occurs prior to hypoxemia and as such capnography can be used as an early warning system to pick up early hypoxemia. This was a randomized controlled trial to see if capnographic monitoring reduced hypoxemia in patients undergoing ERCP/EUS using Microstream Capnographic monitoring. 63 patients were randomized to the capnography blinded arm and 59 were randomized to the titration arm. Within the titration arm, the endoscopic team was aware of capnographic changes as they occurred while in the capnography-blinded arm the endoscopic team was unaware of any abnormalities for up to 30 seconds of apnea. The primary endpoint was the proportion of patients with hypoxemia (<90% oxygen saturation for greater than 15 seconds) in both arms. They found that hypoxemia and apnea occurred more frequently in the blinded arm than in the titration arm. They concluded that microstream capnographic monitoring of respiratory activity can significantly reduce hypoxemia, apnea and oxygen requirements in patients undergoing ERCP/EUS’s.
Gastroenterologist-Directed, Balanced Propafol Sedation for EGD and Colonoscopy: An Analysis of Patient Safety in 15,286 Patients.
Balanced propafol sedation combines small doses of opioid and benzodiazepine with propafol for moderate sedation while monitoring the patient for respiratory abnormalities with capnography. The endoscopist is involved in all decision making in regards to the dosing of propafol. The goal of this study was to evaluate the safety of gastroenterologist-directed balanced propafol sedation during a routine colonoscopy and EGD in an office based setting. This was a retrospective review of EGD/Colonsocopies from January 1st 2003 to June 30th 2007 by three gastroenterologists. The patients were moderately [not deeply] sedated with midazolam, fentanyl and propafol with dose reductions based on age. Adverse events such as hypoxemia, myocardial infarction, seizures, aspiration or use of a reversal agent were collected via chart review or endoscopist recollection. 15,286 endoscopic procedures were reviewed that used gastroenterologist-directed balanced propafol sedation. There were no serious adverse events and only three patients required the use of a reversal agent. The authors concluded that gastroenterologist-directed balanced propafol sedation has an acceptable patient safety profile when undergoing elective EGD/Colonoscopies, cautioning that doses should be reduced based on age.
Early Versus Delayed Feeding After Placement of a Percutaneous Endoscopic Gastrostomy: A Meta-Analysis
Historically after placement of a percutaneous endoscopic gastrostomy, feedings would be delayed for up to twenty-four hours. Earlier feedings may be an option based on certain trails. This group conducted a meta-analysis to look at the effects of early feedings (less than or equal to 4 hours) after PEG placement in adult patients. MEDLINE, Cochrane Central Register of Controlled Trials & Database of Systemic Reviews, DARE, OVID Healthstar & Journals, CINAHL, PubMed and recent abstracts from major conference proceedings were searched. They surveyed a total of 231 articles, 225 of which were excluded which left 6 articles for analysis. Meta-analysis for the effect of early versus late feedings were analyzed by calculating pooled estimates of complications, death less than seventy-two hours and increased post-procedural gastric residual volume during the first 72 hours. There was no statistically significant differences for all complications between the two groups (p=0.63) or death less than 72 hours (p=0.31). A statistical significance was only noted in the gastric residual group (p =0.04). In conclusion, early feeding within four hours is mostly safe and appropriate.
Cervical Esophagus Perforations At the Time of Endoscopic Ultrasound: A Prospective Evaluation of Frequency, Outcomes and Patient Management.
There is minimal data on the risk of cervical esophageal perforation at the time of endoscopic ultrasound (CEP-EUS). The objective of this study was to determine the frequency of cervical esophageal perforations during endoscopic ultrasounds. This was a prospective study done over the course of seven years. The procedures were performed by a single experienced endosonographer and the perforation repairs were performed by a single surgeon. The indications and complications of the procedures were recorded prospectively. The patients were divided into two groups divided by time to analyze the effect of a learning curve. A total of 4,894 upper EUS procedures were performed. The average age of the patients was 59.7 years with an equal sex distribution with a Caucasian predominance. Three patients suffered cervical esophageal perforations and all were treated successfully with a neck incision and recovered. This series showed that CEP-EUS are rare in expert hands, and potentially devastating. The incidence may be two to three times more common than has been reported in the survey literature and early identification and treatment is crucial for recovery.
Drainage Percutaneous Endoscopic Gastrostomy Tube Placement in the Setting of Ascites.
Three percent of patients with advanced malignancies develop malignant bowl obstruction with ascites. Usually in these situations NG tubes are poorly tolerated and surgery is often ineffective with the multifocal points of obstruction. Drainage percutaneous endoscopic gastrostomy tubes have been shown to be effective in the palliation of symptoms associated with malignant bowel obstruction. Ascites generally have been viewed as a relative contraindication to placement of a PEG tube. The goal of this study was to analyze the safety of placing a PEG tube in patients with malignant bowel obstruction and ascites. The study was done retrospectively identifying all patients who underwent successful drainage PEG placements for malignant bowel obstruction at Memorial Sloan-Kettering Cancer Center. Of these patients 35 were identified with moderate ascites, 16 with small ascites, 44 with large ascites and 4 with loculated ascites for a total of 100 patients. Forty patients underwent a pre-PEG paracentesis and the PEG’s were placed using the “pull” method. Five patients suffered complications, 4 minor and 1 major (sepsis/death) complication giving a 5% complication rate. In comparison other studies done in 1992 and 2005 with PEG tube placement without ascites had a complication rate of 1-3%. In conclusion, drainage PEG tubes can be safely placed in patients with ascites with complication rate comparable to regular PEG tubes. This data cannot be extrapolated to patients with ascites secondary to portal hypertension.
TSE “Mask” Improves Oxygenation and Decreases Requirements for Airway Assistance in Deeply Sedated Patients During Lengthy Upper Endoscopic Procedures
Hypoxemia is a common occurrence in patients receiving moderate to deep sedation for upper endoscopic procedures. Supplemental oxygen via nasal cannula is often insufficient to maintain oxygenation with the bite block and endoscope in place. The TSE mask is a simple plastic face tent that can increase FiO2 by 40-60%. This can be used to pre-oxygenate the patient 1-2 minutes prior to inserting the endoscope. The TSE mask is a 12 x 10 inch plastic tent that is used to cover the patient’s nose and mouth. The hypothesis was that the TSE mask can improve oxygenation and decrease the need for airway assistance during lengthy upper endoscopic procedures. This was a prospective, randomized controlled clinical trial. Patients were assigned to the control arm (oxygen via a nasal cannula) and the TSE Mask arm (oxygen via nasal cannula with the presence of a mask). 73 patients were enrolled in the control arm and 74 patients in the “mask” arm with an average age between 57-61 years. Moderate to deep sedation was achieved with the use of propofol. The primary outcomes were oxygenation and oxygen requirements. The mean lowest oxygen saturation was 83% in the control group and 93% in the Mask group with a p value of less than 0.05. The need for airway assistance was higher in the control group (69%) versus the Mask group (48%). In conclusion the TSE mask improved oxygenation and decreased the need for airway assistance in patients undergoing propofol sedation for lengthy upper endoscopic procedures.
Section 7: GERD: Detection, Management and Follow-up
This session briefly discussed treatment and risk factors for GERD. One focus of the final three presentations was on the detection and classification of early pathological changes in patients who are either asymptomatic or mildly symptomatic with GERD.
Endoscopic Full-Thickness Plication for the Treatment of GERD: Five Year Multi-Center Results
This abstract was listed on the schedule but was not formally presented. The following is a summary of the abstract:
This was an open label long term follow-up study from Lenox Hill Hospital looking at treatment of patients with chronic heartburn who underwent full thickness plication 1 cm distal to the gastro-esophageal junction. The results were very similar to those published with three year follow-up. Of the thirty three subjects, there were no adverse events at 5 years and 67% remained off PPI with a significant improvement in their GERD related health quality of life (GERD-HRQL) scores from baseline. In addition, more than half of the patients GERD HRQL had sustained improvement by 50% or more with long term follow-up. The authors conclude that the plicator is a safe and effective long-term modality to reduce GERD symptoms and medication use.
The Association Between Cholecystectomy and Gastroesophageal Reflux Symptoms
Dr. Lin presented a prospective multi-center trial of 302 patients comparing both reflux symptom score (RSS) and gastrointestinal symptom rating scale (GSRS) both one to two weeks prior to and one to three months following hernia surgery (control group) and cholecystectomy (study group). Although baseline GSRS and RSS scores were significantly higher in the cholecystectomy group, the changes in these scores were not significantly different between the two groups with the exception of the GSRS pain sub-score which decreased significantly more in the cholecystecomy group. Therefore the authors conclude that, despite previous reports, cholecystecomy probably does not lead to worsened reflux symptoms. Post-presentation discussion pointed out that weight gain is one probable etiology of GERD post cholecystecomy and the timing of post-surgical endoscopy in this study was too soon to allow sufficient time for patient’s to gain weight and therefore this factor was not controlled for in the study.
Underdiagnosed Esophageal Ulcers in Patients Treated with Cardiac Catheter Radiofrequency Ablation for Atrial Fibrillation: Results of An Ongoing Prospective Trial
Dr. Seres presented a prospective trial of consecutive patients presenting to the University of Oklahoma Health Sciences Center for minimal dose radiofrequency ablation as treatment for either paroxysmal or persistent atrial fibrillation. Previous canine models and case series show frequent post-procedure esophageal ulceration with the potential risk of left atrial-esophageal fistula development (24 case reports in the literature). This group assessed 64 patients for esophageal mucosal injury via endoscopy one day following radiofrequency ablation. 32.8% of the patients had ulcers between 3 and 12 mm in length with none reporting any long term sequelae or adverse events. Since this prospective trial showed a similar incidence of ulceration as previously published case series, this group concludes that the observed ulceration is a validated post-procedure complication and an early risk to the rare but frequently fatal fistulae formation. The authors feel endoscopist’s should be aware of this potential complication as patients will present with either frank UGIB or melena for several weeks leading up to possible hemorrhage. They recommend CT scan prior to endoscopy when considering this diagnosis.
Efficacy of Intelligent Chromo Endoscopy for Detection of Minimal Mucosal Breaks in Patients with Typical Symptoms of Gastroesophageal Reflux Disease
Dr. Chaiteerakij presented this study from Chulalongkorn University comparing white light with “Fujinon intelligent chromo endoscopy” (FICE) for the detection of mild gastro-esophageal reflux disease (MERD). FICE is a tool that utilizes both a high definition endoscope and adjustable wavelengths of light for enhanced detection of abnormal lesions. By imaging the GE junction of each patient with white light and two other settings of the FICE system designed to detect esophagitis, 9 patients without gastro-esophageal reflux symptoms (GERD) and 21 patients with MERD were compared. The images were shown to five experienced endoscopists unaware of the patient’s symptom history, to assess for evidence of mild GERD by looking for indentation into the squamous mucosa from the villiform columnar mucosa at the Z-line (“triangular lesions”). Although the FICE images showed higher sensitivity, negative predictive value and accuracy in predicting those with MERD, intra-observer variability was high and therefore more training and experience is required prior to implementing this detection system into daily practice.
Dilated Intrapapillary Capillary Loops By Magnifying Endoscopy: Usefulness for Diagnosis of GERD
Dr. Kosaka presented a study looking at magnifying endoscopy for the detection of intrapapillary capillary loops (IPCL) as an indicator of early mucosal injury related to GERD. 25 consecutive patients with GERD symptoms were assessed via endoscopy with magnification at 5, 10, 15 and 20 cm proximal to the Z-line. Patients with IPCL (confirmed by two experienced endoscopists) had biopsies drawn from the sites of greatest dilation while those without IPCL had biopsies drawn 5 cm proximal to the Z-line. Each biopsy was then analyzed via light microscopy and transmission electron microscopy for dilated intercapillary space (IS) as an indicator of inflammation. The 15 patients who showed endoscopic evidence of IPCL had significantly larger IS via both microscopic modalities compared with the 10 patients who did not have IPCL. The authors conclude that IPCL observed via magnifying endoscopy is a useful indicator of early mucosal change in patients with GERD.
The Reliability of Ultra-Thin Endoscope for Diagnosis of Milder Form of Esophagitis Compared to Standard Endoscope
Dr. Hongo presented this prospective study comparing the standard (9.6 mm diameter) and ultra-thin (5.9 mm diameter) endoscopes ability to detect early mucosal changes of reflux disease in patients with mild symptoms. Of the 614 consecutive patients who were given the choice of which endoscope would be used, 259 chose the ultra-thin compared with 355 who opted for the standard. There was no anesthesia administered and mucosal observations were made at deep inspiration. In order to differentiate subtle findings, the Los Angeles grade M lesions (minimal) were sub-classified into minimal with redness (MR) and minimal with white blurring (MW), as per previous presentation by this group at DDW in 2007. There were no significant differences in the frequency of detection of any class from the Los Angeles grading system, including the MW and MR sub-groups. The authors conclude that ultra-thin endoscopes can be used to screen for subtle early esophageal mucosal changes in patients with mild symptoms. Furthermore, the group argues that the consistency of findings between endoscopes further validates the sub-classification schema of the Los Angeles criteria presented previously.
Section 8: Gastrointestinal Bleeding
This session focused on GI bleeding and how the epidemiology has changed in the era of H. pylori treatment and PPIs. It also looked on the impact on clinical management of pre-bleeding PPI administration, the impact of endoscopy on mortality, and the role that endoscopy might play in the evaluation of young patients. The presentations also suggested that the optimization of therapy in patients with acute UGIB should aim at reducing the risk of multi-organ failure and cardiopulmonary death instead of focusing merely on successful hemostasis.
Prospective Evaluation of the Upper and Lower Gastrointestinal Tract in Young Patients with Iron Deficiency Anemia (IDA)
Although endoscopic evaluation in older patients with IDA is standard of care in the US, the approach to younger patients with IDA is less clear. Looking at a New York VA population, Chander and Bini examined the utility of endoscopic evaluation in 1432 patients with iron deficiency anemia from 1998-2005, specifically focusing on the yield of upper and lower endoscopy in different age groups. 272 patients were under 50 years old (19%) and a vast majority were male (92%). Individuals with a previously normal colonoscopy and a history of heavy menses were excluded. 49.6% of these young patients had clinically significant lesions, mostly adenomas (20%). In this young population, colon cancer was found in 6%, colitis in 5%, gastric ulcers in 4.4%, and no celiac disease was found. They saw that the types of lesions varied by age, and the number of lesions increased with age in the lower GI tract but not in the upper GI tract. The authors concluded that because a significant number of clinically important lesions may be found, all patients with iron deficiency anemia should undergo GI evaluation, regardless of age.
Localization, Efficacy of Therapy, and Outcomes of Dieulafoy Lesions of the GI Tract- The UT Southwestern GI Bleed Team Experience.
The aim of this study was to further characterize the features of Dieulafoy lesions (DL) in a retrospective and prospective cohort study from (1/03- 10/07). Lara et al found that of the 54 patients with DLs in that time period, DLs made up 3% of all GIB during that time period, and 4.5% of all upper GI bleeds. Of their 54 patients with DLs, 52% were on aspirin, NSAIDs, or warfarin and required a mean number of 4 transfusions. The mean age was 57 years old. 57% of DLs were in the stomach, 24% were in the duodenum, and 9% were in the colon (mostly rectum). The mean Rockall score was 5.3 (1 to 9). Primary hemostasis was achieved in 98% of patients (EGD was performed in <12hours for 73% of patients with DLs) with only 1 patient having rebleeding. They observed a 9% mortality rate with bleeding DLs, but only 1 patient death was directly attributable to bleeding from the DL. These mortality and rebleeding rates seen with DLs in this study were appreciably better that those associated with bleeding from peptic ulcer disease. The authors concluded that GI bleeding from DLs is associated with a good clinical outcome in their center.
Retrospective observational study of patients admitted with acute upper GI bleed
This study examined the advantage, if any, of patients admitted with an acute non-variceal UGIB already established on oral PPIs prior to admission. Data was collected in patients from 2004-2007 who fulfilled criteria including medications on admission (specifically PPIs, NSAIDs, aspirin, clopidigrel, and warfarin) and Rockall scores. Outcomes were compared between the difference in severity of bleeding between patients on and patients not on pre-admission PPI. Patients were excluded if they had bled during a prolonged hospital stay, did not undergo EGD, if admitted for rebleeding, or if they had a variceal bleed. Comparing 106 patients not on a pre-admission PPI to 46 patients on a pre-admission PPI, there was no significant difference between the Rockall scores of the two groups (3.69 v 3.09, NS). They did find that length of stay (LOS) correlated with the pre-clinical Rockall score, and that patients on aspirin and clopidigrel had overall worse Rockall scores. However, as opposed to their preliminary data, they found no difference in Rockall scores depending on pre-admission PPI. This study is ongoing, and the authors hope that as their power increases, a clear beneficial impact of outpatient pre-admission PPIs in acute upper GI bleeding might be seen with lower Rockall scores and shorter lengths of stay.
Decline of incidence and mortality of PUD bleeding between 1983-85 and 2002-04 in Northeast Italy.
The investigators saw a decrease of the incidence of GIB events secondary to PUD during 2 different management eras (PPIs and H pylori treatment), comparing 347 consecutive patients admitted during 1983-85 with 286 patients in 2002-04. Over the 20 year period, the overall annual incidence of ulcer bleeding decreased from 66.5/100,000 persons to 47.6/100,000, corresponding to a 41.6% decrease. The mean age of patients with PUD-related UGIB increased by 9 years, and the use of aspirin and NSAIDS and number of co-morbidities in these patients increased as well. Not surprisingly, the use of endoscopy increased from 51% to 72% as did the use of therapeutic maneuvers during endoscopy. In 1983-85, antisecretory pharmacotherapy consisted of only H2-antagonists drugs, while in 2002-04 49.6% of the patients were treated with proton pump inhibitors. Rebleeding rates and the need for surgery related to UGIB decreased and the annual mortality decreased from 6.3% to 3.5%. This mortality decrease was seen exclusively in patients presenting through the emergency department rather than from the inpatient population. Similarly, the decreased incidence in PUD bleeding was observed only in patients under 70 years. Over a twenty year time period, this hospital in NE Italy saw a decrease in incidence in GIB bleeding and GI bleeding-related mortality, despite an overall increase in in-hospital GI bleeding, NSAID use, comorbidities, and average age of patients. H pylori status over this time period could not be examined due to the retrospective nature of the study.
Analysis of Cause of Death in Peptic Ulcer Bleeding Patients: Study of a Cohort of 10,451 Cases
This group of investigators from this Hong Kong tertiary center prospectively enrolled patients in a GI bleeding registry from 1993-2005 and looked specifically at the cause of death (bleeding versus non-bleeding related), in-hospital mortality, and 30-day mortality. They included over 10,000 patients admitted for non-variceal upper GI bleeding and found the in-hospital mortality to be 4.5% (77% non-bleeding related; 22% bleeding related) and 30-day mortality to be 5.5% (80% non-bleeding related; 18% bleeding related). Patients with a history of PUD and older patients were more likely to die of bleeding-related causes. The use of NSAIDs and aspirin was associated with more bleeding related deaths and higher in-hospital mortality, but this was not seen with events associated with H pylori infection. Rebleeding and the need for endoscopic therapeutic maneuvers did not predict bleeding related death. Among those died of non-bleeding mortality, terminal malignancy (23.5%) and multi-organ failure (29.7%), pulmonary diseases (27.2%) and cardiac failure (14.6%) were the most common causes. The authors concluded that the optimization of treatment for serious UGI bleeding should aim at the prevention of multi-organ failure and cardiovascular complications since most patients who died did not die of bleeding-related causes.
Early Endoscopy in Nonvariceal Upper GI Hemorrhage: Prevalence and Outcome in a Population-Based Sample
This study examined the benefits of early endoscopy (defined as within 24 hours) in non-variceal UGI bleeding, demonstrating a reduction in length of hospital stay but not decreased mortality. This was a national observational study based on 2004 Medicare claims data and thus the population was age 65 or older. Patients underwent at least 1 endoscopy with evidence of either a Mallory-Weiss tear or peptic ulcer disease. Of 2752 UGIB episodes, the mean age was 78 years old, 55% women, 94% with PUD, 6% with Mallory Weiss tears. 71.6% of their population underwent early EGD, increased among hospitalized patients and those with an increased number of co-morbidities. Gender, age, and source of hemorrhage were not predictors of early endoscopy. Early endoscopy was associated with significantly shorter lengths of stay (4 v 6 days), significantly fewer surgeries for UGIB, but not a significantly decreased mortality rate (6% v 7.5%). This study examined inpatients and outpatients in a national setting, but only included older patients as this was a Medicare-based study.
Section 9: New Techniques to Enhance Polyp Detection
This session presented observations and innovations to help improve the sensitivity and technical success rate of optical colonoscopy.
A Randomized Controlled Study Comparing Cap-Assisted Colonoscopy Versus Conventional Colonoscopy: the Final Result
Dr. Lai from The Chinese University of Hong Kong presented results from a randomized controlled study comparing conventional colonoscopy to colonoscopy performed with a mucosectomy cap attached to the colonoscope tip, otherwise known as “cap-assisted colonoscopy”. Patients were randomized to receive colonoscopy (CC) or cap-assisted colonoscopy (CAC). The primary endpoints were the rate of complete examination, cecal and terminal ileal intubation time, and total colonoscopy time. A total of 1000 patients were studied. The cecal intubation time was significantly shorter in the CAC group compared to the CC group (6.0 +/- 3.9s vs.7.2 +/- 4.8s, p<0.001). Total colonoscopy time was also decreased in the CAC group. There was no difference in the rate of complete examination or the number of total polyps, hyperplastic polyps, adenomas, or carcinomas between the two groups.
A Comparison of Overall Adenoma Detection and Recovery Rate of Optical Colonoscopy and CT Colonography
This study presented by Dr. Pfau compared outcomes from patients receiving CRC screening from optical colonoscopy versus CT colonography at the University of Wisconsin Hospital. Adenoma detection rates from 942 patients who received colonoscopy were compared to results from 848 CTC exams. Both groups were average risk for CRC and were age and gender matched. In the OC group, all polyps were removed. In the CTC group, patients with polyps > 10 mm were referred to OC for polypectomy. However, patients with polyps measuring 6-9 mm were offered OC with polypectomy or CTC surveillance, and polyps 5 mm or less were not reported. The polyp detection rate for OC was 0.4 (374 polyps detected in 942 patients) vs. 0.16 n the CTC group (136 polyps in 848 patients). Of the 136 polyps detected by CTC, 61 were still in surveillance or awaiting resection. Not surprisingly, there were more ≤5 mm adenomas detected by OC than by CTC (204/942 vs. 54/848, p< 0.01). A small number of advanced lesions < 10 mm were detected in both groups. There was a trend toward increased detection of advanced < 10 mm lesions in the OC group (10/942 vs. 3/848, p=0.07). Within the OC group, 6.8 % of the 6-9 mm adenomas had advanced pathology, with 3 advanced adenomas ≤ 5mm. The presenter concluded that optical colonoscopy detects and recovers a significant 2.5 fold greater overall number of adenomas compared to a CTC.
Fellow Participation Increases Adenoma Detection Rate During Colonoscopy
The aim of this study presented by Dr. Rogart from Yale was to assess whether GI fellow participation during colonoscopy affects adenoma detection rate. A retrospective review of data from 309 patients who had been prospectively enrolled in a different colonoscopy study not involving polyp detection was performed. A total of 126 colonoscopies performed by a GI attending alone were compared to 183 performed by a GI fellow supervised by one of the same four GI attendings. There were more patients with a poor prep in the fellow group (17% vs. 4%, p<0.001). The adenoma and polyp detection rates were higher when a fellow was involved (37% vs. 23%, p<0.01; and 49% vs. 37%, p<0.01, respectively). However, there as no difference in the number of advanced adenomas detected in each group (7.1% for fellow and attending vs. 5.6% for attending alone, p=0.16). The adenomas detected when fellows participated were also smaller (mean size 4.4mm vs. 5.8mm, p<0.05) and more likely to be sessile (80.6% vs. 64.9%, p<0.05). The presenters concluded that fellow involvement in colonoscopy increases the adenoma detection rate, and the detection of smaller and flatter adenomas. The presenter also discussed whether this finding could be attributed directly to fellow involvement or whether simply “two sets of [trained] eyes” would be sufficient to achieve the same result.
Usefulness of a Transparent Retractable Extension Device On Colorectal Adenoma Detection
Dr. Horiuchi from Showa Inan General Hospital in Japan presented a study evaluating the affect of using a retractable cap fitted to the tip of a colonoscope on cecal intubation time, withdrawal time, and the number, size, and location of adenomas detected. Colonoscopy with or without the transparent retractable extension device was performed by one endoscopist (Dr. Horiuchi) on 835 patients. Sixty of these patients with colonic adenomas were randomized to repeat colonoscopy within three months after the initial study with or without a transparent retractable extension device. Cecal intubation time, withdrawal time and the proportion of patients with adenomas were similar between the two techniques. However, more adenomas were found when the transparent retractable extension device was used compared to without the extension (205 vs. 150, P=0.04). In the 60 patients randomized to repeat colonoscopy with or without the device within 3 months, 20% more adenomas were detected using the retractable cap compared to a 4% increase using standard colonoscopy (P=0.029). The presenter concluded the transparent retractable extension device improved the adenoma detection rate without affecting intubation and withdrawal times.
Colonoscopy in Obese Patients: A Growing Problem
This study presented by Mr. Desormeaux, a Nurse at the Livermore VA Hospital in Palo Alto, California. The aim of the study was to assess whether starting colonoscopy on obese patients in the prone position improves cecal intubation time. The idea for this study came to the presenter after he noticed that many obese patients at his institution finished colonoscopy in the prone position because the cecum could not be intubated in the left lateral or supine position. For this reason, some physicians, at the nurse’s request, began starting colonoscopy in the prone position on obese patients. A retrospective review of this practice was performed. Charts from 75 obese patients (BMI>30) who started colonoscopy in the prone position and 75 obese male patients starting colonscopy in the left lateral position were reviewed. BMI and reported cecal intubation time, sedation doses and pain level were reviewed and compared. Average BMI was slightly higher in the patients started in the prone compared to the left lateral position (38 vs. 36 p < 0.05). Median time to cecum was lower in the patients started in the prone position (5 min vs. 11 min., p < 0.001). Maximal pain level was also decreased in prone position (1.3 vs. 3.0 on the 0-10 scale, p < 0.001). Average midazolam and Fentanyl doses were similar in each group. This intriguing study suggests that starting colonoscopy in the prone position in obese male patients decreases cecal intubation time and decreases pain level at similar doses of sedation. However, the study is limited by it’s retrospective design.
Section 10: New technology
This section consisted of 6 abstracts presenting new endoscopic techniques developed for benign esophageal strictures, gastroplasty for GERD and submucosal dissection. Two studies showing the efficacy of endoscopic incisional therapy for benign esophageal stricture were presented. A new transoral gastroplasty device for GERD was presented using ex vivo and in vivo animal and human tissues. A German and two Japanese groups introduced novel devices and techniques that make endoscopic mucosal dissection more feasible, easier and safer.
A randomized comparative study between dilation by electrocautery incision and savary Bougie for benign anastomotic gastro-esophageal strictures
This aim of this study by Hordijk et al. was to compare the efficacy of dilations with electrocautery incision (EI) or Savary Bougie (SB) in patients with an anastomotic stricture after esophageal resection. 62 patients with grade II dysphagia or worse were randomized to EI or SB and followed up after 1,3 and 6 months. Overall weight change was significant in favor of EI and there was no significant difference in success rate (96.2% vs. 80.8%, p=0.08) and time interval between dilations (median 6 weeks, 95% CI 3.1-8.9 vs. median 6 weeks, CI 4.4-7.6, p=0.73) between EI and SB. However, for the subgroup of patients with extreme stenosis, EI showed a significant higher success rate (88.9% vs. 37.5%, p=0.03) and longer time interval between dilations (mean 6 weeks, CI 4.1-7.9 vs. median 4 weeks, CI 3.3-4.7, p=0.04). Overall patients with anastomotic esophageal strictures preferred EI to SB and patients with extreme stenosis required less dilations with EI.
Endoscopic incisional therapy with iso-tome and transparent hood in patients wit benign anastomotic esophageal stricture
This study by Lee et al. from South Korea evaluated the efficacy of endoscopic incisional therapy using an Iso-Tome and an end-viewing endoscope fitted with a transparent hood to ensure safety margin in 24 patients with benign anastomotic esophageal stricture after total gastrectomy. Radial incisions parallel to the longitude of esophagus were performed by pulling up the Iso-Tome. All 21 patients with stricture < 1 centimeter in length were successfully dilated after a single incisional treatment without repeated treatment during the follow up 18 months. 3 patients with > 1 centimeter experienced restenosis in a mean of 1.6 months and 2 sessions of treatment were repeated. This technique was described by the authors as easy, safe and appears to maintain longer patency than bougie or balloon dilation.
Randomized controlled study of endoscopic mucosal resection (EMR) versus endoscopic submucosal dissection with a waterjet hybridknife (ESDH) of esophageal lesions in a porcine model
The objective of this study as to compare a new simplified ESD technique with conventional EMR. A total of 25 esophageal areas with a diameter of 20 mm were randomized to either EMR (n=13) by use of the cap technique or ESD (n=12) with a hybrid knife which allows cutting/coagulation as well as injection/flushing through axial water jet channel with a preselected pressure of a high pressure waterjet system. There was no perforation in both groups. This randomized controlled trial shows that ESDH significantly achieves complete resection of esophageal lesions more frequently (10/12 vs. 6/13, p=0.05) with less number of specimen than EMR (1 vs. 2.5+/-0.9, p<0.05). The easy use of water-jet assisted injection of large amount of fluids may explain why ESDH was as safe as EMR in spite of thin esophageal wall in the porcine model.
A new transoral gastroplasty device for GERD and obesity
The authors of this study developed a device that allows a physician to create an endoscopic, full thickness gastroplasty under conscious sedation. The gastroplasty was formed by a new dilator shaped device (SafesStitch Medical Inc.). The instrument consists of a flexible tube with an integrated excision and suture capsule on the distal end. They performed preclinical procedures with ex vivo porcine (n=5) and ex vivo human tissue (n=10) as well as in vivo porcine (n=5) and canine procedures (n=5). In each model, the authors were able to create full thickness gastroplasties at the GE junction and within the proximal stomach.
A novel device, counteraction-spring, is easier and safer for endoscopic submucosal dissection of mucosal stomach cancer
The authors in this study from Japan invented a novel counteracting device that consists of a stainless steel spring that is 2 mm in diameter, 2 centimeter long and has a wire diameter of 0.1 mm and a nylon loop attached to each ends. This spring passes through 2.8 mm forceps channel and counteraction is applied by attaching one of the loops to one end of the stomach cancer with a clip and attaching the other loop to the healthy mucosa on the other side. ESD with this novel device was performed for 10 patients. 8 of 10 cases were type IIa early gastric cancer. Dissection was possible with sufficient counteraction and a good field of view in every case. Pathologic findings showed 8 well differentiated mucosal cancers with negative resection margin and no vascular invasion, 1 adenoma and 1 submucosal invasion with positive vertical margin. This novel device provided favorable counteracting for submucosal dissection of mucosal stomach cancer and made the dissection technique easier and safer.
Magnetic force assisted endoscopic submucosal dissection with Permalloy tissue anchors
This is another study from a Japanese group that used a novel device to provide counteraction during submucosal dissection. The authors developed a “multipurpose magnetic device for fixing multiple surgical tools and organs” (m-DMSTO). This system allowed intuitive manipulation of multiple surgical tools within body cavities. Surgical tools in this system are attached to super-paramagnetic material (permalloy) tissue anchor and remotely maneuvered with a needle shaped magnetic manipulator. The technical feasibility of ESD using m-DMSTO was studied in an anesthetized in vivo porcine model. 40 mm mucosal lesion was safely removed less than 10 minutes. The overlying mucosa was effectively deflected away from the dissection plane during submucosal dissection with optimal multidirectional tissue traction by magnetic force. There were no complications and magnetic force assisted ESD is technically feasible, safe and easy.
Section 11: Contemporary Small Bowel Imaging
This session contained a number of studies involvoing various modalities for investingating small bowel pathology , the issue of volume associated variability in the accuracy of capsule endoscopy interpretation, and localization of angioectasias within the small bowel. There was also a presentation of a series using the novel spiral overtube technique of enteroscopy.
Capsule Endoscopy (CE) Re-Assessment in a High Volume Center (HVC) Identifies Significant Misinterpretations of the Community Gastroenterologists (CG) That May Impact Clinical Care
Patricio Ibanez et al from Cedars-Sinai Medical Center addressed the issue of the agreement between High Volume Center (HVC) vs. community gastroenterologists (CG) interpretation of capsule endoscopy (CE) images. This was a prospective study evaluating consecutive CE readings referred to their institution from CG for second-opinions and/or DBE between July and November 2007. The HVC readers (100-3,000 CE reading experience), initially blinded to the CG readings, gave interpretations on CE through initial review and subsequent group consensus at a weekly CE conference. The CG interpretation was then unblinded. A disagreement between the HVC and CG reads was classified as major if it led to a difference in recommended clinical management and minor if it did not. Double balloon endoscopy (DBE) when performed was considered the gold standard for assessing CE findings. The results show that of the 43 consecutive CE studies, 38.1% had a major disagreement, 31.7% had a minor one, and 8% had both. In the 8 subsequent DBEs obtained in the major disagreement group, the HVC accuracy was 100% as compared to 13% by CG (p=0.0014). CG interpretations in this group had a false positive and negative rate of 38% and 80% respectively including the missed interpretation of 1 adenocarcinoma, 1 sarcoma, 1 NSAID-stricture, and 1 probable lymphoma. They concluded CE readings referred to HVC from CG will often have reinterpretation of findings, frequently with a change in management indicated.
Correlation of CT Enterography (CTE) and Capsule Endoscopy (CE) Findings in Patients Being Evaluated for Suspected Small Bowel (SB) Disease
In this retrospective study Mishkin et al from Boston University looked at 100 patients who received both CT enterography (CTE) and capsule endoscopy (CE) in the work-up of known or suspected small bowel disease. In the “CTE then CE” group (88 pts) the most common indications included gastrointestinal bleeding (GIB) and iron deficiency anemia (IDA) (55 pts) with the most common finding being vascular ectasias. For those with initially negative CTE, a positive CE finding was found in 36/49 (73.5%) patients. For patients sent with known or suspected Crohn’s Disease (16 pts) there was a concordance of findings in 12/16 (75%) cases. In the “CE then CTE” group (12 pts) the most common indications were also GIB and IDA. CTE was often performed in these instances to clarify diagnosis (6 pts) and to evaluate an intralumenal “bulge” seen on CE (4 pts). There was concordance in 7/11 (63.6%) of the cases. Their conclusion was that in patients with known or suspected CD who clinically required both imaging modalities, the concordance between CTE and CE appeared to be high.
Do Clinical Characteristics Predict Small Bowel Angiodysplasias On Capsule Endoscopy?
DeBenedet et al from University of Michigan set out to investigate whether clinical characteristics can predict the presence of small bowel AVMs by capsule endoscopy (CE) in patients with previously established AVMs by EGD or colonoscopy. Their retrospective study included only patients with complete exams (EGD, colonoscopy and CE) and excluded patients with a history of radiation therapy to the GI tract and hereditary angiodysplasia syndrome. 120 patients (93% with GI blood loss) were found to have 1 or more AVMs on EGD and/or colonoscopy and met inclusion criteria. The mean age was 69 years, and 64% of the patients were women. 36% of patients were found to have ≥1 jejunal AVMs, and 15% were found to have ≥1 AVMs. In a multivariate model including age, gender, creatinine, history of aortic stenosis, and presence of AVMs on traditional endoscopy; age ≥ 65 (OR 2.63, p=0.05) and presence of AVMs on EGD (OR 5.56, p=0.01) were significant predictors of jejunal AVMs, but not ileal AVMs. AVMs on colonoscopy alone were not predictive of jejunal or ileal AVMs. Their conclusion was that patients with AVMs on EGD are at increased risk for jejunal AVMs on CE, particularly if they are elderly. In these patients, it may be reasonable to bypass CE and proceed directly to therapeutic SB enteroscopy if blood loss persists after treatment of AVMs on endoscopy.
Location and Distribution of Arteriovenous Malformations in the Small Bowel Based On Capsule Endoscopy As An Aid to Determine Anterograde Vs Retrograde Approach for Balloon Enteroscopy
Isenberg et al from Case Western Reserve University set out to examine the distribution of AVMs in the small bowel based on capsule endoscopy (CE) as a possible aid in the clinical decision of whether anterograde vs. retrograde balloon enteroscopy should be performed first for therapeutic intervention of obscure bleeding. The study design was that of a retrospective chart review of 784 patients who underwent CE between May 2002 and October 2007. 107 patients (58 men; mean age of 68, range 39-89) were ultimately diagnosed with small bowel AVMs by CE. The distribution was 37 proximal, 20 proximal-middle, 18 middle, 6 middle-distal and 13 distal. 17 of the diagnosed small bowel AVMs were actively bleeding. The distribution by location of bleeding small bowel AVMs was 9 proximal, 4 proximal-middle, 1 middle, 2 middle-distal, and 2 distal. Their conclusion was non-bleeding and bleeding AVMs occurred most frequently in the proximal small bowel and least frequently in the distal portion of the small bowel thus adding credence to the practice of performing anterograde balloon enteroscopy prior to the retrograde approach.
Post-Procedure Elevated Amylase and Lipase Levels After Double Balloon Enteroscopy: Relations with the Double Balloon Technique
In this prospective study from Yeditepe University in Turkey, Pata et al attempted to study the relationship between post-double balloon enteroscopy (DBE) amylase and lipase elevations with DBE technique. Peroral (48 pts) and anal (8 pts) DBEs were performed in 56 patients with blood samples taken before the procedure and at 4h and 12 h afterwards. Patients were evaluated for abdominal pain at 4h using a visual analogue scale (VAS). The route and duration of DBE; the total insertion length of the scope; the insertion length at first balloon inflation; and the duration between the first and second balloon inflations were noted. In the anal DBE group 0/8 patients had hyperamylasemia, defined as >upper limit of normal (>65 IU/ml). In the peroral DBE group hyperamylasemia was noted in 12 patients (25%). “Pancreatitis-like abdominal pain” occurred in six patients (12.5%). Using a definition of pancreatitis as an amylase level three times above normal limits (>200 IU/ml), a VAS score >5 at 4h was 100% sensitive and 96% specific for diagnosing pancreatitis. Following multivariate regression analysis, the one independent predictor of post-oral DBE pancreatitis was duration between the first and second balloon inflation (p=0.000). They concluded hyperamylasemia was a common peroral DBE complication which might be prevented by reducing the time between the first and second balloon inflations.
The Spiral Enteroscopy Experience in 101 Consecutive Patients: Safety and Efficacy Using the Discovery SB
Dr. Paul Akerman presented his group’s experience since November 2005 with Spiral Enteroscopy (SE), a technology involving the use of a spiral on a rotating overtube (Discovery SB) to pleat small bowel on the enteroscope and advance through the bowel. 101 consecutive outpatients patients were all examined for an indication of obscure GI blood loss. They were 52 males (51.5%), with a mean age of 45 (range 19 – 78), mean height of 164cm, mean weight of 71.4kg. 86 patients (85.1%) received MAC with propofol, fentanyl and versed while 15 patients required general anesthesia. The 57F Discovery SB was used with pedi colonoscope (26pts), and the new 48F Discovery SB was used with a Olympus 200cm 9.2mm enteroscope(25pts) and a Fujinon 200cm 9.4mm enteroscope (50pts). 3 patients had their procedure aborted, 1 due to an esophageal stricture, and 2 due to inability to intubate. Average depth of insertion past the Ligament of Treitz (LOT), estimated by endoscopic criteria, was 224cm(50-400cm). Average time to maximal insertion was 21.4min. and average total procedure time 32.1min. The findings included AVMs (24pts), strongyloides (4 pts), tumors (2 pts), duodenal ulcers (2 pts). The two patients with tumors had the distance past the LOT measured at surgery (280cm and 100cm) and compared to the endoscopically estimated distances (240cm and 75cm). Complications included intussusceptions (3pts) all resolved during the procedure; mild mucosal damage - limited to mucosa (17 pts); moderate mucosal damage - limited to submucosa (5 pts), and sore throats (22pts). One patient was admitted after 24 hours with abdominal pain. Their conclusion was SE is a safe and effective method of visualizing the small bowel.
Section 12: ERCP: New Innovations
This session covered papers on new techniques and concepts in ERCP. It included six studies including timing of precut sphincterotomy, biliary stent removal techniques, importance of biliary stent diameter, double guidewire technique, endoscopic papillary large balloon dilation for removal of large biliary stones, and the development of an ERCP quality network.
Timing of Precut Procedure Does Not Influence Success Rate and Complications of ERCP Procedure. A Prospective Randomized Comparative Study
It is a prospective randomized study that compared the success and complications of early vs. delayed precutting during ERCP. Out of total 1078 procedures during the two year study period, 146 patients (that failed biliary cannulation after 5 minutes) were randomized in 1:3 ratio into group A (early precut) or group B (B1 if cannulation was achieved by standard approach in further 20 minutes; otherwise delayed pre-cut was carried out- B2). There were 36 patients in group A and 110 in group B. Deep biliary cannulation was achieved: 92% in group A, while 95% in group B (B1 100%, B2 81%). The overall complication rate was: 8% in group A, and 6% in group B. The authors concluded that the timing of precut has no influence on success and complication rates of ERCP. However, this was an under-powered study.
Endoscopic Removal of Uncovered, Partially & Fully Covered Biliary SEMS: Problems & Salvage Techniques. A GEPED Multicenter Descriptive Study
This study analyzed the feasibility and techniques of biliary SEMS removal during ERCP. It was a retrospective review of 68 SEMS removals in 62 patients from 01/2002 to 11/2007 at three sites. The removal techniques were classified as either basic (rat tooth/snare), advanced (cautery/intraductal forceps), or salvage (any unreported). The time interval was classified as early (< 1 month), intermediate (1-4), or late (> 4). Initial success was obtained with basic techniques in 48 (72%) and in a further 5 with advanced techniques (80.3%). Initial failure was more likely in late vs intermediate vs early attempts at removal, but salvage by a second coaxial SEMS was equally effective. Failure was caused by embedding due to ingrowth, mesh break-up, or angulation w/o ingrowth. Hence, SEMS removal is feasible, safe, and affected by time interval. Furthermore, covered stents do not prevent embedding. A second coaxial SEMS is an effective salvage technique.
Comparative Performance of Uncoated Self Expanding Metal Biliary Stents (SEMS) of Different Designs in Two Diameters: Final Patency Results of An International Multi-Center Randomized Controlled Trial (RCT)
This multicenter RCT included 183 patients with unresectable malignant biliary strictures below the bifurcation. The patients were treated with one of the three uncoated SEMS: 6 mm Zilver, 10 mm Zilver, or 10 mm Wallstent (control). 6 mm group was closed as interim analysis revealed excessive occlusions. Stent occlusion occurred more frequently (39%) and earlier in the 6 mm arm, but proved to be equal between the 10 mm arms (21-24%). The main mechanism for occlusion was ingrowth, followed-up by overgrowth, debris and hemorrhage. Difficulty of placement, malposition, complications and mortality were not statistically significant amongst all groups. The authors concluded that it’s the diameter and not the stent design or material that determines stent patency.
The Usefulness and Safety of Pancreatic Guidewire Placement for Achieving Deep Cannulation of the Bile Duct
This was a retrospective study that analyzed the usefulness and safety of pancreatic guidewire placement (PGP) in difficult biliary cannulation that was defined as unsuccessful cannulation in 15 minutes. Out of total 3955 ERCP’s from 12/01 to 10/07, PEP was attempted in 113 cases. The success rate was 82%. The overall post-ERCP pancreatitis rate was 12%. Prophylactic pancreatic stent (5 Fr x 4 cm) was attempted after PGP in approximately 2/3rd of cases. No or failed pancreatic stenting was a significant risk factor for pancreatitis. Spontaneous stent dislodgement rate at 2 weeks was 93%. Hence, PGP is an effective and safe strategy; and successful prophylactic pancreatic stenting can significantly reduce the risk of pancreatitis.
SES+EPLBD Method Makes the Treatment of Large Common Bile Duct Stones Easy and Long Term Result Was Acceptable 5-Th Report
This study reported the short & long-term results of small endoscopic sphincterotomy combined with endoscopic papillary large balloon dilation (SES+EPLBD) to remove large or multiple biliary stones without lithotripsy. After SES, papillary dilation was slowly performed with a large 20 mm balloon to match the size of bile duct; followed by stone retrieval with balloon and basket. The success rate was 96%. Mean: stone size 13.6 mm, number of stones 2.4, and procedure time 32 min. Complications: bleeding 2%, post-ERCP pancreatitis 1%, no perforation. After mean follow-up of 30 months in 55 cases, stone recurrence 2%, sludge 2%, no bile duct stenosis. Hence, SES+EPLBD is a safe and effective strategy. This data reflects the use of somewhat smaller diameter balloons and smaller size stones that some of the previously reported series with few but serious complications reported from EPLBD.
The ERCP Quality Network. A Pilot Project for Comparing ERCP Practices and Performance
This presention reported on an ongoing project supported by Olympus Corp. that is designed to compare the practicality and quality of ERCP performed by individual US endoscopists. Information including the experience and practice environment of the endoscopists, details on each ERCP procedure, and complications is gathered and loaded onto a secure website. Participants can view their “report card”, and compare their performance with others (i.e., benchmarking). So far, data has been gathered from 44 US endoscopists. 70% of the ERCPs were biliary. The success rates: deep biliary cannulation 96% (range 78-100); biliary stone removal 97% (for stones <10 mm) and 80% (for stones >10 mm). This project will allow comparisons across the spectrum of ERCP practice, and facilitate development of a national system for quality reporting.
Section 13: What's New in Lower GI Bleeding and Polyps
This session included papers covering a diverse range of topics. These included endoscopic findings over time in patients undergoing radiation for prostate cancer, risk stratification for patients presenting with lower GI bleeding, the yield and clinical importance of incidental colon findings on CT/PET scans, the use of ESD techniques applied to laterally spreading colon polyps, and an novel technique of using suction to create pseudopolyps to facilitate the complete snare excision of flat colon polyps.
The Effect of Radiotherapy for Prostate Cancer on Later Rectal Injury
In this study of the natural history of radiation induced rectal injury, 146 patients had flexible sigmoidoscopies at both 16 months and 3 years following ratdiation. Nearly half of all patients included in their cohort (48 % of patients) developed sigmoidoscopic evidence of radiation injury at 16 months following prostate radiotherapy but moderate or severe changes were uncommon. Radiation induced injury worsened over time with sigmoidoscopic findings at 3 years significantly worse that at 16 months. Clinical symptoms did not correlate with sigmoidoscopic findings. Endoscopic evaluation of bleeding is mandatory to verify radiation damage and to exclude other pathology. In view of the frequency of polyps (20.2%) and cancers (2.5%), pre-radiation lower GI endoscopy appears to be strongly advised.
Clinical Signficance of Incidental 18F-FDG uptake in the colon on PET/CT
This paper reported on 58 patients who underwent colonoscopy following PET/CT studies which had incidental findings and compared colonoscopic findings to those of control patients having colonoscopies following normal PET/CT scans. Colonosocpy revealed neoplasms in 48% vs 3% of controls. These authors concluded that colonoscopy should be performed when incidental colonic activity on PET/CT is found despite a false positive rate of 1 in 3. The SUV readings were not helpful in distinguishing benign from malignant lesions. Also, a negative PET/CT was not adequate to rule out significant colonic neoplasm, lymphoma or inflammatory conditions.
Development and Validation of a Simple Risk Classification Rule for Patients with Acute Lower Gastrointestinal Bleeding
The authors reviewed data from over 15,000 patients hospitalized with acute lowe gi bleeding to identify clinical parameters predictive of mortality. They developed a risk classification tool which incorporated altered mental status, BUN, creatinine and hemodynamics which was able to stratified mortality risk for these patients on admission.
Family History of Single First-Degree Relative with Colorectal CancerDoes Not Predict Increased Risk of Advanced Adneomas on Screening colonoscopy in a Southwestern U.S. Population
This prospective study of individuals referred for screening colonoscopy looked at the impact of a family history of advanced adenomas. 4377 patients were included in this study. The authors found that having a single first-degree relative with an advanced adenoma did not predict a significantly increased finding of an advanced adenoma in the screening population.
Treatment Strategy for Laterally Spreading Tumors: Historical Overview Since the Introduction of Colorectal Endoscopic Submucosal Dissection
This study compared the surgical resection rate and the en bloc resection rate of laterally spreading colon tumors > 20 mm before and after the advent of endoscopic submucosal dissection (ESD) techniques developed initially to resect early gastric cancers en bloc. 166 patients were included in this historical, non-controlled comparison. In this experience, ESD techniques led to fewer surgical resections and increased en bloc resection rate.
Prospective Study and Description of the Suction Pseudo-polyp Technique for the Removal of Small Flat Polyps of the colon and Rectum.
This series detailed a new technique for resecting flat polyps. 59 polyps in 47 patients were small and flat and all were removed successfully using a suction technique to create a psueodopolyp facilitating subsequent snare polypectomy. In this study, the authors found that small flat polyps of the colon and rectum can be safely removed by suctioning the polyp into the suction channel with gentle retraction of the colonoscope followed by snare resection (cold if < 7mm and hot snare of > 7mm) as to remove the polyp and allow for adequate margins.
Of note, over half of these small flat polyps were found to be adenomas and no significant complications were noted from this new technique.
Section 14: Current Trends in Upper Gastrointestinal Neoplasia
This session included papers on various topics relavent to the diagnosis and management of upper GI neoplasia. Topics ranged from exploring the yield of diagnositic work up for alarm symptoms in younger patients, current distribution of pathological types of gastric polyps that are encountered, a comparison of EMR to ESD techniques for early esophageal cancer, the outcome of gastric ESD in a large cohort, and the implications of incomplete EMR upon subsequent gastric surgery.
Therapeutic Outcome in 1000 Cases of Endoscopic Submucosal Dissection (ESD) for Early Gastric Neoplasms; Korean ESD Study Group (Kesg) Multi-Center Study
The authors conducted a multi-center study of 1000 cases of early gastric neoplasms which were resected via ESD and monitored for therapeutic outcome. In the immediate term, ESD was successful in curative resection in 87% of cases, with over 95% of cases resected en-bloc. Lesions with a statistically significant higher en-bloc resection success rate were 1) located more distally, 2) did not have a scar present, or 3) were adenomatous instead of carcinomas. Lesions located in the proximal stomach were more frequently associated with delayed bleeding. Proximally located lesions also had longer procedure times. The perforation rate was 1.2%. The authors conclude that ESD is safe an effective in the management of early gastric neoplasms, and that complication rates were acceptable, but an experienced ESD-trained endoscopist should perform the procedure.
Which Endoscopic Resection Method Is Best to Treat Small Esophageal Cancers?
This single center study by Ishihara et. al. was conducted o compare the treatment of small esophageal cancers using EMR vs. ESD. The authors treated 341 esophageal cancers in 241 patients by EMR or ESD. Inclusion criteria were: lesions with HGD or EC that were 20mm or less in diameter, sm1 invasion without lymph node involvement, in patients with no prior history of treatment with chemotherapy. The study found that ESD was successful in resecting the lesion en-bloc 100% of the time, and resulted in curative resection 97% of the time. This was significantly higher than EMR in successful en-bloc resection (87%) and curative resection (71%) rates. The authors found the procedure time to be significantly longer using ESD. They conclude that for en-bloc or curative resection, lesions between 15-20mm should be treated by ESD for higher success rates.
A Cohort-Based Study of Missed and New Cancers After Gastroscopy
This study by Raftopoulos et. al. evaluated the hypothesis that normal upper endoscopy rarely misses upper gastrointestinal cancer (UGC). The authors used an electronic database of all patients that underwent EGD between 1990 and 2004. They collected data regarding site and histology of cancer, findings at EGD, indication for EGD, interval to malignancy, operator experience, and EGD equipment used. They defined missed cancers as those diagnosed within 1 year of EGD, possible missed lesions as those diagnosed 1-3 years following EGD and new cancers as those diagnosed greater than 3 years after EGD. Of 27500 EGDs performed, UGC was diagnosted in 115 cases. There were 28 missed cancers, 25 possible missed cancers and 74 new cancers. 11 of the missed cancers were esophageal, 14 were gastric and 3 were duodenal. In 68% of missed cancers, an abnormality was described at the site of malignancy. In 57% the indication for EGD was dysphagia or blood loss. The authors found that operator experience or new equipment were not significantly related to missed cancers. They conclude that UGC is rare after EGD and that EGD is highly accurate in detecting cancerous lesions.
Are “High Risk” Features Associated with Increased Gastrointestinal Pathology in Patients Aged Less Than 50 Years with Dyspepsia?
The authors conducted a study to determine if EGD should be performed in all patients, irrespective of age, in patients with “high risk” features such as weight loss, anemia, dysphagia and vomiting. Current ASGE guidelines currently recommend EGD for high risk symptoms in patients over 50, but the merit of EGD in younger patients has not yet been defined. The authors evaluated endoscopic findings in all patients undergoing EGD for dyspepsia over a 23 year period. They identified 10,641 patients under 50 with dyspepsia, 1119 had high risk features. Overall, cancer prevalence was very low in patients aged <50 years with dyspepsia; but was significantly higher in patients with high risk features (1.4%) compared to those without high risk features (0.4%). They found that The number of endoscopies required to diagnose 1 cancer was 14.5 in patients aged ≥50 years with high risk features, and 69.9 in patients aged <50 years with high risk features and 264.5 <50 years without high risk features. They concluded that while the rate of cancer found via EGD performed in patients under 50 years of age with dyspepsia was low, it was still a cost-effective strategy for detecting cancer when performed on patients with high risk features.
Prognosis of Patients with Early Gastric Cancer Who Underwent Additional Gastrectomy After Incomplete Endoscopic Mucosal Resection
This study was performed to determine the outcome of patients undergoing gastrectomy after incomplete EMR. 53 patients were followed up for an average of 47 months after failed EMR. Curative resection with LN dissection was performed in 51 patients and 2 patients underwent wedge resection. 28 patients had evidence of residual cancer in the stomach. Lymph node metastases were found in 3 patients. There was one case of cancer recurrence after gastrectomy. This study suggests that the outcome of additional gastrectomy following incomplete EMR may be similar to that of primary gastrectomy for EGC.
The Changing Balance of Gastric Polyps: A Study of 5,972 Patients
This is the largest study conducted to date on the frequency and pathology of gastric polyps. The authors analyzed 5972 gastric polyps found via a multi-center electronic database of patients undergoing endoscopy during a 12 month period from 2006 to 2007. They found that 71% of polyps were fundic gland polyps, 16.9% were hyperplastic, 9.1% were foveolar hyperplasia/polypoid gastritis, 1.4% were xanthomas and 0.96% were adenomas. Patients with polyps were older compared to the base population. They conclude by noting that older studies on smaller cohorts stated that 75% of gastric polyps were hyperplastic, polypoid gastritis was second most frequent followed by fundic gland polyps and adenomas. Their study counters these numbers as fundic gland polyps reperesented 2/3 of all resected polyps. The authors suggest that the increased and widespread use of PPIs probably accounts for this observed shift in gastric polyp pathology.
Section 15: Endoscopic Ultrasound – Pancreas
This topic forum featured innovation ideas on applications of EUS in the pancreas, such as the effort to utilize FNA DNA information to guide chemotherapy decisions, to better differentiate carcinoma from chronic pancreatitis, and to gain prognositic information in pancreatic endocrine tumors. The role of elastography coupled with neural networks was reported in increasing the detection of malignancy and consensus data on the EUS criteria for chronic pancreatits was presented.
Gene Analysis of Advanced Pancreatic Carcinoma Using EUS-Guided FNA with Focused DNA Array: Possible Predictor of the Chemotherapeutic Efficacy
Ashida et al used a microarray platform containing 133 genes pre-selected for their suspected role in drug resistance (a.k.a. a focused DNA array) to predict response to gemcitabine therapy. They enrolled 95 patients whom underwent EUS-FNA of a pancreatic mass and were able to isolate RNA and amplify cDNA from 81 of them. 35 of these patients received gemcitabine for at least 2 courses and these patients composed the final study population. 12 of these patients were defined as responders (partial response and stable disease) and 23 were non-responders. Based on the relative expression of a 6 gene panel (DCK, ENT1, ENT2, DCD, NT5, RRM1) as compared to average expression of these genes in the 81 samples they were able to establish a GEM-sensitivity score with sensitivity, specificity and accuracy of 67%, 74% and 71%. This important study showed the potential utility of EUS-FNA samples as not only diagnostic but also valuable in prognostication of chemotherapeutic response.
DNA Microsatellite Loss in Endoscopic Ultrasound-Guided Fine Needle Aspirate from Pancreatic Endocrine Tumors Is Associated with Increased Mortality
Fasanella et al evaluated the role of microsatellite loss analysis in this retrospective study of patients with confirmed pancreatic endocrine tumors. Their data was limited to 23 patients whom had either died or had at least one year post-operative follow up and for whom LOH analysis was successful. As previously published by this group, they expressed the number of LOH mutations per valid LOH in each sample (fractional allelic loss or FAL), and used a cut-off of >0.2 (or 20 % of evaluated LOH showing anueploidy). This value correlated with both disease progression and mortality. Using a cut-off of FAL>0.33 correlated with a 20 fold increased chance of death. The authors conclude that this analysis may have a role on pre-operative risk stratification in certain PET patients.
Measuring Mesothelin Protein Level in Endoscopic Ultrasound-Guided Fine Needle Aspiration Specimens-A Novel Method in Pancreatic Adenocarcinoma Detection
Mesothelin is expressed relatively late in the molecular cascade of pancreatic cancer progression. It is also known that cytologic evaluation of EUS-FNA specimens of pancreatic masses may have a false negative rate of 10-20%. This work by Wang et al evaluated the value of testing tissue mesothelin expression by immunohistochemistry and ELISA and serum levels by ELISA. They found the levels in tissue to be significantly higher in pancreatic cancer then in chronic pancreatitis. The levels in both groups were much lower in the serum and there was no difference between the two groups. They conclude that mesothelin is a tissue specific marker that may have a role in further distinguishing CP from pancreatic cancer in situations of non-diagnostic cytology.
Standardized Weighted Criteria for EUS Features of Chronic Pancreatitis: The Rosemont Classification
This work by Hernandez et al. presented the expert consensus opinion of an EUS-based diagnostic criteria for chronic pancreatitis (CP). 5 experts presented a series of statements and questions to 45 recognized endosonographers and established the following grading system based on at least >2/3 agreement.
EUS features were grouped and defined as follows:
| Major A (A) | Major B (B) | Minor (Min) |
Hyperechoic foci + shadowing |
Lobularity + honeycombing |
Cysts; dilated MPD>3.5 mm; irregular PD; dilated side branch>1mm; hyperechoic duct wall; strands; non-shadowing hyperechoic foci; lobularity with non-contagious lobules |
Classification:
| Most consistent with CP | Suggestive of CP |
Indeterminate of CP |
Normal |
1(A)+>3(Min);1(A)+1(B);2(A) |
1(A)+<3(Min); (B)+>3(Min); >5(Min) |
>2 but <5 (Min); (B) |
<2(Min) |
Cytologic Evaluation of Cystic Pancreatic Lesions (CPL): Comparison of EchoBrush Vs. Standard EUS-FNA Techniques in a Blinded Prospective Study
This prospective study by Gill KR et al compared the diagnostic yield of EchoBrush with standard FNA in cystic pancreatic lesions. The cysts were aspirated to 50% volume by FNA, followed by the EchoBrush and then the remaining fluid was aspirated to cyst collapse. The pathologists were blinded to the samples and evaluated intracellular mucin (ICM) and grade of dysplasia. Out of 22 patients the EchoBrush identified ICM in 6 whom had negative FNAs. Of those with ICM 2 were found to have high grade dysplasia by EchoBrush alone. The study suggested that EchoBrush may have higher yield for the diagnosis of ICM and dysplasia then standard FNA.
Neural Network Analysis of Dynamic Sequences of EUS Elastography Used for the Differential Diagnosis of Chronic Pancreatitis and Pancreatic Cancer
Saftoiu et al used EUS elastography to differentiate benign (normal or chronic pancreatitis) and malignant (adenocarcinoma and NETs) tissue characteristics. They utilized the publicly available ImageJ software to calculate hue histograms and established the sensitivity, specificity and accuracy between confirmed benign and malignant tissue as 91%, 87% and 89% respectively. Using these results they trained a neural network to classify benign vs malignant disease and achieved a 95-97% average performance. As shown for the differentiation of benign and malignant lymph nodes this study demonstrated the promise of EUS elastography in distinguishing malignant from benign lesions and points to its potential role in better targeting of biopsies.
Section 16: New Imaging in the Upper GI Tract
This session focused primarily on endoscopic techniques which could better predict histology in diseases such as Barrett’s esophagus, GERD, and submucosal lesions in the upper GI tract. The overall theme was the concept of virtual biopsies thereby transitioning the endoscopist to the “endopathologist” utilizing these new techniques.
Confocal Laser Endomicroscopy for In Vivo Diagnosis of Barrett’s Esophagus and Associated Neoplasia: An Ongoing Prospective Study
An Italian study which took 39 patients (35 with known BE) and compared fluroscein aided confocal laser microendoscopy to histologic findings. The confocal images were classified using the Kiesslich Confocal Barrett Classification. The accuracy was 80.2% and the agreement was substantial (X= 0.74). They concluded that this technique increases the yield of biopsies as compared to the standard of care random 4 quadrant technique with an impressive ~ 96% sensitivity.
Confocal Laser Endomicroscopy in Barrett’s Esophagus and Associated Neoplasia: A Prospective Randomized Controlled Crossover Trial
The purpose of this crossover study done at Johns Hopkins compared targeted mucosal biopsies via confocal laser endomicroscopy (CLE) with standard 4 quadrant random biopsies in 30 patients with either known or possible Barrett’s esophagus. One group consisted of those who were referred for high grade dysplasia and the other routine surveillance. They found that the mean number of biopsies taken in the was significantly lower in the CLE group as compared to the standard group (10.75 vs. 28.4, p=0.004). The overall diagnostic yield was significantly higher in the CLE group as compared to the standard group (32.4% vs. 19.7%, p=0.03).
In Vivo Imaging of the Living Cancer Cell and Evaluation for Tissue Atypia Using Endocystoscopy: ECA (Endocytoscopic Atypia) Classification in the Esophagus and Stomach
This Japanese study by Dr. Inoue and colleagues used endocytoscopy to distinguish cellular features (in particular nuclear morphology) with the aide of tissue staining in 130 patients containing superficial esophageal and gastric lesions. The changes were characterized using the endocytoscopic atypical classification, a five tier classification assessing nuclear morphology where ECA-1 is normal and ECA-5 is cancer. ECA 1-3 were considered benign and ECA 4-5 were malignant. The overall accuracy in distinguishing benign vs. malignant lesions was 87.1 %. The PPV and NPV was 86% and 89.7% respectively. The researchers concluded that this technique may have the potential to make histologic diagnosis without actual biopsy and that this technique may be more important to those patients who are anticoagulated.
High Definition Endoscopy (HD+) with I-Scan and Lugol’s Solution for the Detection of Inflammation and Neoplasia in Patients with GERD
This German study assessed 41 patients with heartburn and compared standard high definition (HD) endoscopy vs. enhanced imaging with either chromoendoscopy (via Lugol’s solution) or I-Scan. I-Scan utilizes a new EPKi processor (Pentax, Japan) which allows physicians to better assess the mucosal topography and vasculature of the epithelium with the touch of a button. Using the LA classification, they found a non-significant trend in detecting mucosal breaks with I-scan as compared to standard HD endoscopy (12 vs 8, p=ns). More importantly was a statistically significant increase in the number of mucosal breaks detected with chromoendoscopy (21 vs 8, p = 0.029). A secondary endpoint was the detection of circumferential lesions which were detected more with chromoendoscopy followed by I-scan and standard HD endoscopy (70 vs. 41 vs. 19, p < .001). The authors point out that although the enhanced imaging detected more lesions, there was no increase in the amount of neoplastic changes.
Digital Image Analysis (DIA) Is a Useful Adjunct to Endosonographic Diagnosis of Submucosal Lesions (SML) of the Gastro-Intestinal Tract
Dr. Das and colleagues from the Mayo Clinic performed this retrospective study evaluating submucosal lesions in the upper GI tract by reviewing EUS images. The aim was to see whether subtle changes in the EUS morphology could be better detected via changes in pixel patterns (utilizing a software program) and hence predict “virtual” histology without need for biopsy. A total of 46 patients with histologically documented SML’s consisting of GIST (28), lipoma (8), and carcinoids (10). Using three types of pixel distribution they found that DIA had a sensitivity of between 79-90% and a specificity of 77-81%, concluding that this technique is a useful adjunct to EUS in diagnosing SML’s.
In Vivo 3D Comprehensive Microscopy of the Human Esophagus for the Management of Barrett’s Patients
This study from Harvard evaluated using optical frequency domain imaging (OFDI) in the management of Barrett’s esophagus. Using a special fiberoptic probe, three-dimensional microscopic views of the inner surfaces of vessels and organs were obtained. Forty patients were investigated with this technique examining ~ 6 cm of the distal esophagus. The researchers found that in 92% (37) patients achieved satisfactory imaging when correlated to histology, concluding that this technique is a potential means of obtaining a “virtual biopsy”.
Section 17: ASGE Plenary Session, Part I
The first group of endoscopic papers presented in the ASGE Plenary Session addressed some of the most pressing topics in endoscopy, in particular related to cost-effective strategies of colorectal cancer prevention, as relates to the management of small polyps found on colonosocpy and on CTC. A novel low-tech way to facilitate colonoscope passage was presented along with a meticulous survival animal study comparing various methods of perforation closure.
Cost Savings of Removing Diminutive Polyps Without Histologic Assessment
This interesting study out of Indiana University examined the cost savings of using “real-time” histology, as opposed to pathologic evaluation, to determine the classification of small polyps. Since most colorectal polyps are dimunitive (< 6mm in size) and rarely contain cancer, the histologic reports are mainly used to guide post-polypectomy colonoscopy intervals. The study evaluated 4474 colonoscopies in which at least one diminutive polyp was removed, each of which was evaluated by both “real-time” histology and pathology. They found that by using “real-time” histology, $181 could be saved per colonoscopy, with $116 million in annual savings in the US. Only 3% of patients with polyps would be mislabeled using on-site assessment, and only 12% would be scheduled for follow-up at a non-recommended interval (of which, most would undergoing surveillance at a too early interval). Less than 1 in 1100 patients with a diminutive polyp removed would have an undetected cancer in any removed polyp. Based on this data, it may be more effective to not send small polyps for pathologic examination.
American College of Radiology (ACR) Recommendations for CT Colonography (CTC) Interpretation: Implications for Resection of High Risk Adenoma Findings
This study, also out of Indiana University, evaluated the implications of the ACR recommendations for CT colonography. The ACR recommends that polyps ≤ 5mm in size not be reported on CT colonography studies, while patients with 1 or 2 polyps 6-9 mm in size can be offered “CTC surveillance” in 3 years in lieu of polypectomy. In this study, 10,780 polyps were removed endoscopically from 10,034 patients over a 5 year interval. There were 774 patients fit the category of 1 or 2 polyps 6-9 mm in size and no polyps ≥ 10 mm in size. Of these patients, 184 actually had either 3 or more adenomas ≤ 9 mm in size or an advanced adenoma ≤ 9mm in size or both findings. There were 269 patients who had no polyps ≥ 6mm, but had either 3 adenomas ≤ 5mm or an advanced adenoma ≤ 5mm or both. They concluded that if CTC were used in this population according to ACR interpretation recommendations, then a significant number of patients with high risk adenoma findings would be incorrectly interpreted as either being normal or would have polypectomy delayed at least 3 years.
Water Immersion Vs Conventional Colonoscopy Insertion: a Randomized Controlled Trial Showing Promise for Minimal-Sedation Colonoscopy
This study assessed the feasibility of minimal-sedation colonoscopy using a water immersion colonoscopy insertion technique. Using the water immersion technique, 300cc of tap water, instead of air, is instilled in the rectum. Ninety-eight patients at the Palo Alto VAMC were randomized for either water immersion or standard technique of colonoscopy, and premedicated with 2mg midazolam. Procedure “success” was defined as completion of the colonoscopy to the cecum without additional medication or attending physician intervention. The researchers found that cecal intubation time was significantly shorter in the water group, although intubation rate was not. Less sedation was needed in the water group. There was also a trend towards increased procedure “success” in the water group, especially for procedures performed by fellows. Patient satisfaction survey responses did not significantly differ between the groups. They concluded that the water immersion significantly improved cecal intubation time while using less sedation and without sacrificing patient satisfaction. It will be interesting to see if this technique is more widely adopted.
Clip Closure Versus Endoscopic Suturing Versus Thoracoscopic Repair of An Esophageal Perforation: Randomized Comparative Long Term Survival Study in a Porcine Model
This study out of Germany evaluated different methods of closing a full thickness gastrointestinal wall incision, as would be necessary for NOTES. Clip closure (CC) has been used in some patients, but surgical repair (TC) remains the gold standard. A new endoscopic suturing (EC) technique has been reported in animal models. This study compared the 3 techniques on a standardized 2 cm esophageal incision performed in 18 pigs (6 in each group). The pigs survived for 6 weeks after closure. They were then re-evaluated endoscopically before pathologists performed post-mortem examinations with histology and morphometric analysis. The researchers found that esophageal repair was successful in all groups. At autopsy CC showed the best results with hardly any periesophageal adhesions. Two animals died of mediastinitis (1 EC animal due to food having refluxed into the mediastinal space during repair and 1 TC pig due to suture inadequacy). On histology, EC actually performed significantly better than either TC or CC, with shorter and less wide defects with fibrous replacement. They concluded that CC and EC had no more complications than TC, and EC actually was significantly better histologically. The results are encouraging and suggest that endoscopic esophageal wall repair may be feasible.
Section 18: ASGE Plenary Session Part 2
The second set of endoscopic research abstracts included a carefully conducted animal trial in the emerging field of NOTES, a randomized trial to determine the optimal hemostasis technique in the era of hemoclips, a compilation of a tremendous volume of experience in non-anesthesiologist administration of propofol, and a randomized trial of CO2 insufflation for ERCP. A common thread to these papers is the application of good scientific methodology to address questions that are timely and clinically important areas for the present and future of the field.
A Prospective Randomized Controlled Trial of NOTES versus Laparoscopic Distal Pancreatectomy, Preliminary Results.
This study compared outcomes of endoscopic natural orifice vs. laparoscopic distal pancreatectomy in pigs. Standard laparoscopic distal pancreatectomy has a reported morbidity up to 52% with leaks occuring in up to 30% of cases. The authors aimed to compare a NOTES procedure to a laparoscopic procedure with relatively high risk in an attempt to detect a potential difference in complications. After a run-in phase, 28 pigs were randomized to the two groups. This was an 8 day survival study. One animal died in each of the 2 groups. No leaks occurred in the endoscopic group and one leak was observed in the laparoscopic arm. Endoscopic distal pancreatectomy was more time consuming. This study demonstrated that endoscopic transgastric distal pancreatectomy is feasible and could be done safely in a pig model. At least in this survival animal model, non-inferiority of this method was demonstrated.
Hemostasis of Very High Risk Patients with Severe Non-Variceal UGI Hemorrhage Comparing Injection Hemoclipping with Injection MPEC.
Dr Jensen presented the results of this trial which was notable for the focus on a population of patients with very high co-morbidity, a group often excluded from controlled hemostasis trials. The purpose was to compare efficiency and safety of the two hemostasis methods in this high risk population. This was an intention to treat analysis. Excluded patients were those with malignancy, varices, and angiomatous disease. All patients received PPI and H. Pylori treatment if indicated. Thirty-two patients were in the Epi: MPEC arm and forty-seven in the Epi: Boston Scientific Hemoclipping arm. The number of patients with clinical decompensation and hypotensive shock were the same. One perforation was noted in the Epi: MPEC arm. The re-bleed rate was 43% in the EPI:MPEC arm and was 20% in the Epi:Clip arm. Failure of hemostasis was 23% in the Epi: MPEC arm and 3.5% in the Epi:clip group. Surgery for re-bleed was needed in 20% of the epi: MPEC and only 3.5 % of the epi:clip group. The main observation was that the combination of Epinephrine and hemoclipping appears to be superior to the combination of Epinephrine and MPEC and should be considered in the treatment of severe, non-variceal hemorrhagic upper gastrointestinal bleeds.
Non-Anesthesiologist Administered Propofol Sedation for Endoscopic Procedures: A Worldwide Safety Review.
This was a robust literature review performed with the goal of providing safety data for the use of propofol by non-anesthesiologists. Of 519, 627 cases of non-anesthesiologist administered propofol for GI procedures, there were three deaths, one of which was in a patient with severe mitral regurgitation, one with metastatic pancreatic cancer and one with a history of IVDA. Additionally there were 5 endotracheal intubations which were required, 1 seizure reported. 349 Bag-Mask ventilations out of 462, 055 cases were reported of which no intubations were required in this group. All complications occurred in upper endoscopic procedures. No complications occurred in lower GI procedures. Limitations of review include the underreporting of adverse events and the dose of propofol which was given was not always documented in the reviews. The main point of this literature review is that non-anesthesiologist administered propofol is safe and effective. Outcomes from this analysis (limited by retrospective nature) were better than the currently used non-anesthesiologist administered combination of opioid and benzodiazepines.
Carbon Dioxide (C02) Insufflation During ERCP for the reduction of Post-Procedure Pain: Preliminary Results of a Randomized, Double-Blind Controlled Trial.
CO2 diffuses 35 times more quickly than nitrogen (major component in air). These authors conducted a study to determine whether insufflation with CO2 leads to any difference in pain experienced 1 hour following ERCP than when air insufflation is used. The Ez- EM CO2 insufflator was used at 3 liters/min. Patients were monitored for pain and nausea. They were treated for pain when pain scores exceeded 4. Waist circumference to monitor for abdominal distention and follow-up phone interviews were performed. 105 patients were enrolled in this study to date. One hour post procedure pain was significantly less in the CO2 arm but no change was noted on post procedure day 1. Waist circumference was also significantly greater in the air insufflation group. More pain meds were needed in the air group but this difference was not significant. The authors concluded that CO2 insufflation for ERCP is safe and the observed benefits in post-procedure pain could theoretically influence patient satisfaction and need for hospital admission.