Update on Endoscopic Research Presented at 2006 Digestive Disease Week
Update on Endoscopic Research Presented at 2006 Digestive Disease Week
Prepared by Gastroenterology fellows from New York City and Boston area training programs.
Edited by Jonathan Cohen, MD, New York University School of Medicine
| Topic | Authors |
|---|---|
| Introduction and Summary | Jonathan Cohen, New York University School of Medicine |
| Plenary ASGE session | Pascale Anglade, MD, Montefiore Medical Center Suresh Pitchumoni, MD, New York University School of Medicine |
| New Technology - Treatment | Jordan Karlitz, MD, Montefiore Medical Center, Albert Einstein College of Medicine |
| GERD and Other Esophageal Disorders | Ponni Perumalswami, MD, New York University School of Medicine |
| Colorectal Cancer Screening: Topics in Standard and Virtual Colonoscopy | Jordan Karlitz, MD, Montefiore Medical Center, Albert Einstein College of Medicine |
| Pancreatic ERCP | Ronald Palmon, MD, Mt. Sinai School of Medicine |
| Esophagus - Barrett's/Cancer | Ilseung Cho, MD, New York University School of Medicine |
| New Technology - Diagnosis | Sang Lee, MD, Montefiore Medical Center, Albert Einstein College of Medicine |
| Imaging Advances in the Biliary Tree | David Hass, MD, Montefiore Medical Center, Albert Einstein College of Medicine |
| Contemporary Approaches to Small Bowel Imaging | David Hass, MD, Montefiore Medical Center, Albert Einstein College of Medicine |
| Pediatric Endoscopy | Tamara Feldman, MD, Montefiore Medical Center, Albert Einstein College of Medicine |
| Pancreatic EUS in Clinical and Translational Research | Christopher Di Maio, MD, New York Presbyterian Medical Center |
| Clinical Innovations in EUS | Daniel Gassert, MD, Montefiore Medical Center, Albert Einstein College of Medicine |
| Endoscopic Safety and Complications | Sang Lee, MD, Montefiore Medical Center |
| New Technology - NOTES | Debbie Hillman, MD, Montefiore Medical Center |
| Colonoscopy Potpourri | Saad Jazrawi, MD, St. Vincents Hospital, NY |
| AGA Session: Imaging and Advanced Technologies | Reina Pai, MD and Derek Fong, MD, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA |
| Summary | Jonathan Cohen, New York University School of Medicine |
Introduction
This May, at Digestive Disease Week in Los Angeles, California, endoscopic researchers presented a wide array of interesting and at times provocative papers. This report will attempt to summarize the highlights of the latest developments in endoscopic research as presented at the meetings. It was compiled through the efforts of gastroenterology fellows from the New York and Boston area whose names and institutions are listed in the table at the end of this document. The goal of this report is to provide an overview for OMED members around the world of the key areas of current endoscopy related investigation. The following paragraphs include detailed summaries of all of the original papers from each of the 13 ASGE research topic fora, three additional original endoscopic research papers presented in the AGA Session: Imaging and Advanced Technologies, and the original research included in the ASGE plenary session.
Summary
The specific research activities and progress summarized above collectively creates an impressive glimpse of major developments within endoscopic research that appear underway. The papers included a wide array of new technological development in colonoscopy, moving towards the reality of sedation free, loop free colonoscopy and optical and targeted biopsies. The new area of NOTES, merely a provocative concept 5 years ago, is exploding with collaborative research involving gastroenterologists and surgeons as well as new devices to carry out the new techniques that are contemplated. The number of abstracts involving wireless capsule endoscopy and double balloon endoscopy abounded, and the presentation of a prototype remote controlled capsule made the concept of therapeutic capsule endoscopy seem a not too distant possibility. The endoscopic ultrasonographers have expanded the horizon of potential therapeutic interventions with reports of FNA injection of gene therapy vectors and chemotherapeutic agents. Long term follow-up data was presented for now established therapies such as PDT and EMR for Barrett’s esophagus with dysplasia, and further more aggressive techniques of endoscopic submucosal dissection were described. Finally, an increasing focus on research and discussion on important quality and outcome measures of endoscopic procedures, such as adenoma detection rates was evident.
This report hopefully conveys the general sense imparted during this meeting that the field of endoscopy is ensconced in a time of change, both incremental and disruptive. The research presented reflects the ongoing efforts of investigators to chronicle, to embrace, and to shape these ongoing developments.
ASGE Plenary Session: From Colonoscopy to Surgery-Endoscopy Today and Tomorrow
Prospective analysis of withdrawal time and polyp detection rates during screening colonoscopy and a comparison to US multi-society task force recommendations: A community-based continuous quality improvement study
Withdrawal time during colonoscopy may be a measure of quality. In a community practice of nine gastroenterologists, this study aimed to determine if withdrawal time was associated with number and type of polyp detected. This study did not show a statistically significant association between detection rate and withdrawal time, though most of the withdrawal times in the study were under 5 minutes. The only statistically significant variable was bowel preparation: for unclear reasons, a worse preparation predicted a higher number of adenomas detected. The authors speculated that this observation might reflect increased time spent cleaning the mucosa. These results seemed to question several other studies that have been reported showing a linear relationship between withdrawal time and the detection of total polyps as well as adenomas.
Magnified observation if microvascular architecture using NBI for differential diagnosis between non-neoplastic and neoplastic colorectal lesion: A prospective study
NBI allows imaging of mucosal capillaries without the use of dye. Once a polyp is detected NBI can be used to determine the likelihood of it being neoplastic or non-neoplastic by examination of capillary patterns. This study aimed to determine usefulness of NBI in screening colonoscopies. NBI evaluation of polyps was compared with histologic examination and results showed a high sensitivity (96%) specificity (92%) for the NBI diagnosis of neoplasia. NBI may become an efficient way to determine the presence of neoplasia within polyps without histologic examination. The NBI features of serrated polyps have not yet been well characterized.
A novel retrograde-viewing auxillary imaging device (3rd eye retroscope) improved the detection of simulated polyps in anatomical models of the colon
Colonoscopy has a known miss rate. Most missed lesions are located on the proximal aspects of folds, outside the viewing range of a conventional colonoscope. The third eye retroscope is a disposable unit used through the instrument channel of the colonoscope, which allows visualization at a180 degree angle. The performance of six gastroenterologists using anatomical colon models with simulated polyps (obvious or “hidden” on proximal aspects of folds) with and without the 3rd eye retroscope was studied. The number of polyps found on the proximal aspects of folds was significantly higher than the number found using conventional colonoscopes (81% vs 12%). This new technology may decrease the miss rate associated with colonoscopy. Of course, with the present prototype instrument, any lesions detected would then need to be located and removed by front-viewing and retroflexion after the retroscope is removed from the biopsy channel of the colonoscope. The ability of this device to decrease the polyp miss rate in real subjects has not yet been tested.
An endoscopic implantable device stimulates the LES on-demand by remote control in canine model
GI pacing devices have traditionally been placed surgically. This study looked at the feasibility and efficacy of an endoscopically placed LES pacing device providing on-demand stimulation. After submucosal injection of saline above the LES and needle-knife incision, the device was implanted in three dogs. Manometry studies revealed increased LES pressure in all dogs when using a stimulation amplitude of 10 mA. It remains unclear if increased LES pressure on demand will result in clinically improved GERD symptoms.
Endoscopic temporary gastric stimulation- Results of 140 consecutive patients
Surgically placed gastric pacing device are used for the treatment of gastroparesis. This study looked at the feasibility and efficacy of an endoscopically placed temporary pacing device in 140 pts with gastroparesis. Results showed a statistically significant improvement of nausea and vomiting in all patients studied after the device was placed. This technology may help to determine which patients will respond to a surgically placed permanent gastric pacing device for gastroparesis.
Blinded Comparison of Esophageal Capsule Endoscopy Versus Conventional Endoscopy for Diagnosis of Barretts Esophagus in Patients with Chronic Gastroesophageal Reflux
In recent years, there has been a lot of interest in an esophageal capsule endoscope for the detection and surveillance of Barrett’s esophagus. Such reasons include its safety, lack of a need for sedation, and overall patient preference. Recent studies in 2004 and 2005 showed that the esophageal capsule endoscope could be a viable option for studying Barrett’s esophagus. But the studies were not prospective, not properly blinded or not confirmed with biopsies. So the authors of this study set forth to fix these problems and perform a blinded, prospective study on the utility of esophageal capsule endoscopy (versus conventional endoscopy) for the diagnosis of Barrett’s esophagus. They performed the study on 90 patients with GERD and performed both capsule endoscopy and EGD with biopsy on the same day. Results showed that capsule endoscopy had a significantly lower sensitivity and specificity for the detection of Barrett’s esophagus than did EGD with biopsy. Although the results were slightly disappointing for the esophageal capsule, the results still show that there is great potential for this technology.
Design, Development, and Testing of a Remote-Controlled, Stereoscopic (three-dimensional) Imaging, Self-Propelled, Wireless Capsule Endoscope
In the year 2000, the journal Nature published a paper on wireless capsule endoscopy. This paper caught the interest of several gastroenterologists at Stanford University. They saw the obvious benefits of such a system, but were bothered by the inherent limitations of such a capsule, which could easily miss areas of interest, such as an ulcer or varix, just being propelled along the GI tract through peristalsis. The gastroenterologists approached their colleagues in the Aerospace Engineering department, who were intrigued by this concept. Together, they decided to embark on a project to overcome these inherent flaws in the capsule by producing a remote-controlled capsule endoscope.
The team of aerospace engineers first created a mathematical model of a propulsion system. They were ultimately able to convert this into an actual propulsion system, using an ultrasonic transducer, which is able to push and pull the capsule through the aqueous environment of the stomach, creating a net force. They tested the system, and it worked. The next task was to add depth to the pictures generated. So the team added stereoscopic vision, producing a three-dimensional view of the GI tract. And finally, the team was able to shrink all this down to fit into a relatively small capsule, controlled by a joystick. Movement is still limited to a liquid medium, and it remains a bit too large for easy swallowing.
While this provocative creation is not yet ready for human trials, this paper illustrated that a remote-controlled, 3-D capsule endoscope may well be available for clinical use in the near future.
Deconstructing the Endoscope: Intragastric, Transgastric and Laparoscopic Wireless Endosurgery Using Manipulable Attached and Free Capsule Imagers
In recent years, gastroenterologists have grown accustomed to the concept of using a capsule endoscope to view the GI tract. Many are familiar with the progression of images of the esophagus, stomach, small intestines, and even the colon, as the capsule advances through the GI tract. But a group of gastroenterologists stretched this concept, thinking outside the proverbial box, by imagining the potential uses of a fixed capsule with real-time viewing capability—thereby “deconstructing” the endoscope. The group conceptualized placing and fixing a capsule in various locations—inside the gastric wall, outside the gastric wall, and in the peritoneum. They explained the rationale for such a novel concept—prolonged observation of a defined region. For example, if a capsule is placed in the stomach wall, the capsule can observe a gastric ulcer for potential bleeding or rebleeding. If the capsule is implanted in the peritoneum, it can provide assistance during laparoscopic procedures. Such observation can be extended, as it may be possible to control the capsule’s position with a remote-control, thereby allowing adjustments in your view. The next challenge will be creating a battery with a longer life to allow long-term observation or even one with the ability to turn off and on to prolong potential viewing time. In conclusion, such “deconstruction” allows spatial separation of the imaging and therapeutic functions of the flexible endoscopy and may thereby provide functions never imagined before.
New Technology: Treatment
The therapeutics session for the ASGE consisted of 6 abstracts examining novel endoscopic techniques. 2 studies focused on the endoscopic creation of anastomoses between the stomach and the small bowel. A study entitled “Double Balloon Anastomosis: A New Method for Transgastric Gastro-Jejunostomy” demonstrated a technique in which 2 balloons, one in the small bowel (placed directly by way of the peritoneal cavity after needle knife excision of the gastric wall) and one in the stomach, were positioned side by side to compress intervening tissue, forming a gastrojejunostomy over time. A second study from the University of Pennsylvania described a technique in which a loop snare is advanced into the small bowel through a gastrostomy. A guidewire is advanced into the snare via transgastric EUS and ultimately stent placement over the guidewire is performed allowing “real time” gastroenteric anastomosis creation.
Two studies focusing on the submucosal space were also presented. An abstract from Germany entitled “Submucosal Endoscopy- A Novel Approach to En Bloc Endoscopic Mucosal Resection” examined the use of a small endoscope introduced into the submucosal space by way of a needle knife incision in pig esophagus. Once in the submucosal space, tissue dissection was performed, allowing visualization of the muscularis and underside of the mucosa. This technique may ultimately be used to visualize the depth of mucosal lesions and facilitate their removal in humans. The second study, also from Germany, entitled “Submucosal Injection of Autologous Blood Before Endoscopic Resection- An Option to NACL and Hyaluronate? Results of an Ex-Vivo Pilot Study” demonstrated the use of autologous blood injection to create a submucosal cushion for snare resection in the porcine stomach. Results showed longer mucosal elevation compared with 0.9% NACL but shorter than with Hyaluronate.
Two other abstracts presented at the session included a study from Toronto entitled “Endoscopic Cryotherapy for the Management of Gastric Antral Vascular Ectasia (GAVE): A Pilot Study” and an abstract from the University of Cincinnati entitled “Novel Transmural Endoscopic Suturing Device.” In the GAVE study, endoscopic treatment with cryotherapy was utilized in 9 patients with previously diagnosed GAVE, 8 of whom had already been treated with APC. Results demonstrated a complete response in 6/9 patients and a partial response in 3/9 patients in 3 endoscopic sessions. The follow up was short and limitations due to overtube requirement and difficulty treating in certain locations were addressed. In the suture study, an endoscopic device was used to place full thickness sutures in a pig stomach. Out of 27 successful needle passes, 14/27 sutures were visible in the gastric serosa and 5/27 in the muscularis thus demonstrating successful transmural placement.
GERD and other Esophageal Disorders
Endoscopic Treatment of Stomal Stenosis Following Roux-en-Y Gastric Bypass
Abessi et al presented a retrospective review of 11 patients treated with stomal stenosis after roux-en-y gastric bypass with serial endoscopic balloon dilation. The results showed that serial balloon dilation until symptoms improvement was tolerated well by patients and these patients continued to lose weight. Patients who had recurrence of symptoms underwent a repeat dilation with improvement of symptoms. Patients were followed for average of 29 months and underwent 2-4 dilations total. The mean final balloon size for the dilations was 16 mm. It will be interesting to see whether any of these patients regain weight following dilation of the stoma to this diameter, or what the mean diameter would be if these patients undergo repeat endoscopy other than for symptoms of restenosis during their follow-up.
Prevalence of Esophageal Findings in Non-Cardiac Chest Pain Patients
Up to 25% of the US population has non-cardiac chest pain (NCCP). Esophageal findings are reportedly uncommon (<10%). The aim of this study of Dickman et al. was to determine the prevalence of endoscopic findings in patient undergoing upper endoscopy for NCCP using the Clinical Outcomes Research Database and compare the findings to those with GERD. Significant endoscopic findings in non-cardiac chest pain patients included Barrett’s (4.4%), erosive esophagitis (19.4%, hiatal hernia (28.6%), and 44.1% had normal findings. The authors concluded that esophageal findings are common in patients with NCCP, though significantly less so that in patients with classical GERD symptoms alone.
Use of an Artificial Sphincter for the Treatment of GERD- Preliminary Study
Ganz and colleagues conducted a feasibility study looking at the use of a surgically implanted titanium magnetic lower esophageal sphincter necklace in 10 pigs. The necklace is composed of expansile string of miniature magnetic beads placed surgically at the distal esophagus at the GE junction. The varying magnetic strengths studied showed normal EGD and barium study findings when pigs were followed for several weeks and then sacrificed. LES and yield pressure were increased by the prosthesis with no effects on sphincter opening or evidence of dysphagia in the pigs. The histological findings were consistent with fibrotic changes. There was good separation and re-approximation when the necklace was studied suggesting that this is potential future alternative for fundoplication pending further human studies.
The Budget Impact of Screening for Esophageal Varices (EV) with Capsule Endoscopy
This was a study using the Clinical Outcomes Research Initiative (CORI) database looking at budget impact models comparing esophageal capsule, beta blocker, upper endoscopy and endoscopic band ligation for the treatment of esophageal varices. The authors found that medical therapy with beta blockade was the cheapest therapy when looking at one year cost. They also found that capsule examination might be cost effective compared to endoscopic strategies if medical therapy is to be used as primary prophylaxis based on the results instead of band ligation. EGD would be more efficient if EBL were to be considered if varices are found. Of note, the capsule strategy was not assigned a GI consult fee in this model whereas the screening EGD had consult fee included; screening EGD on demand without a consultation would have made the EGD strategy less expensive than the capsule approach.
Successful Treatment of Globus Sensation by Argon Plasma Coagulation of Gastric Inlet Patches in Cervical Esophagus- Pilot Trial
Meining et al. reported an intriguing but small study of their experience using thermal ablation of inlet patches in an attempt to treat refractory globus sensation. The authors looked at 10 patients who underwent treatment of gastric inlet patches with APC with respect to patient symptoms. Though the follow up was limited, they found that patients with histologically proven gastric inlet patch who were treated with APC had improvement in symptoms (using a visual analog score) of globus sensation and sore throat following treatment. Other symptoms associated with reflux such as respiratory symptoms, hoarseness, and heartburn were not affected. It would be useful to have an effective treatment for the globus sensation and the refractory sore throats that persist despite full anti-secretory therapy for presumed reflux; this study suggests that a controlled study with sham treatment and prolonged follow-up is warranted to test the hypothesis that APC ablation of inlet patches is effective in this setting.
Regenerative Medicine for Widespread Endoscopic Mucosal Resection of the Esophagus
Sakurai and Hori presented their work performing endoscopic mucosal resection (EMR) in a swine model using regenerative medicine techniques to prevent stricture of the esophagus after EMR of broad lesions. Regenerative medicine techniques used in this study involved injection or autologous implantation of cultured buccal mucosa keratinocytes at the site of resection. Each pig had a control lesions consisting of EMR without any implantation. The pigs were then sacrificed 2 weeks later. Grossly, scar and stricture formation was found in control lesions while the implanted lesions had intact epithelium and smooth luminal surfaces histologically. This feasibility study using a novel technique, if it can be translated to humans, might lead to important advances in the prevention of strictures following wide EMR.
Colorectal Cancer Screening: Topics in Standard and Virtual Colonoscopy.
The colonoscopy session consisted of 6 abstracts examining a variety of topics including colonoscopic withdrawal time and its impact on polyp detection and the role of virtual colonoscopy in various clinical settings. A study entitled “Impact of Endoscopic Withdrawal Speed on Polyp Yield: Implications for Optimal Colonoscopy Withdrawal Time” was conducted at the Mayo Clinic. In the study, 10,955 colonoscopies were performed by 43 endoscopists. The median polyp detection rate for this group of endoscopists was 42.7%, corresponding to a median withdrawal time of 6.7 minutes. Longer withdrawal time increased polyp yield, though this applied more for detecting smaller sized polyps. The authors concluded that a median withdrawal time of 7 minutes will lead to polyp detection rates above the median performance among endoscopic peers. A second study entitled “Effect of a Pre-Specified Minimum Colonoscopic Withdrawal Time on Adenoma Detection Rates During Screening Colonoscopy” was conducted at the University of Illinois. In the study, 12 experienced endoscopists performed a total of 818 colonoscopies on average risk patients with withdrawal times of at least 8 minutes. When compared to routine colonoscopies performed on an average risk control population, study endoscopists had longer withdrawal times (12.5 vs. 9.3 minutes, P< 0.0001) and had higher adenoma detection rates (0.73 vs. 0.58 adenomas per subject screened, P < 0.001).
Abstracts focusing on virtual colonoscopy included a study from the University of Wisconsin which compared optical colonoscopy (OC) and virtual colonoscopy (VC) in average risk patients. In the study, 1110 patients were screened with VC and 1132 with OC. 933 total polyps (255 of which were adenomas) were seen in the OC group whereas 153 total polyps (61 of which were adenomas) were seen in the VC group. There were significantly more small polyps (less than or equal to 9 mm) detected in the OC group compared with VC, but the detection of larger lesions (greater than 10 mm) was the same with both modalities. A second study, also from the University of Wisconsin, examined the impact of VC screening programs on OC in clinical practice. These investigators found that the number of colonoscopies performed for screening or for polypectomy did not change 14 months after the introduction of screening VC in the community. However, the number of referrals for OC did significantly decrease which may be an early indicator of decreased demand in the future. Interestingly, a significant number of patients with polyps found on VC were not referred for OC. A third abstract concerning VC entitled “Risk Stratification for Advanced Colonic Neoplasia: A Screening Strategy Using Colonoscopy and CT Colonography” described 3005 asymptomatic patients who had already undergone screening colonoscopy. These patients were retrospectively randomized to derivation and validation groups and 3 screening strategies were examined: universal colonoscopy, universal CT colonography and a stratified strategy of colonoscopy for high risk and CT colonography for low risk patients. Investigators concluded that compared with universal colonoscopy, the stratified screening strategy resulted in almost the same sensitivity for advanced neoplasia while reducing the number of needed colonoscopies by 1/3.
An abstract from Norway entitled “Colonoscopic Screening for Colorectal Cancer- Impact on Later Need for Usual Care Lower Gastrointestinal Endoscopies” focused on the increased demand for endoscopic resources. The investigators compared a group of 451 individuals who were invited for a colonoscopic screening evaluation with a control group who received “ordinary care” through the local health service. After following the patients for 9 years, it was noted that patients who had initially undergone screening colonoscopy received 50% fewer non-surveillance related colonoscopies than the control group (63 vs. 110 examinations, P < 0.001). In other words, the act of undergoing a screening colonoscopy prevented patients from having additional procedures in the future for indications other than surveillance.
Pancreatic ERCP
Several interesting studies were presented at the “Pancreatic ERCP” session at DDW 2006. Two studies focused on the use of techniques to avoid pancreatitis after ERCP. By using the Indiana University prospective database of almost 15,000 ERCP’s, it was shown that the extent of opacification of the pancreatic duct correlated strongly with the risk of post-ERCP pancreatitis. Other factors that contributed to the risk of pancreatitis in the multivariate analysis included suspected sphincter of Oddi dysfunction (SOD) and therapeutic vs. diagnostic ERCP. Female gender and young age which have been important risk factors in other studies did not independently contribute to pancreatitis risk in this cohort. The other study from the University of Alabama was a prospective randomized trial comparing needle-knife over stent vs. pull-papillotome followed by stent for pancreatic sphincterotomy in patients with manometrically documented pancreatic sphincter hypertension. The trial was halted early after 24 patients had been randomized to each group when the pull-papillotome group had 7 patients with pancreatitis while none were observed in the needle-knife group. One criticism of this study is that blended current was used in both groups; in some studies of biliary sphincterotomy, blended current has been shown to cause more pancreatitis than pure cut current.
A prospective cohort of 128 patients from the University of Virginia undergoing ERCP for pancreatic duct disruption was analyzed to identify risk factors for successful endoscopic treatment. Success was achieved in 90% of patients and was defined by resolution of symptoms: pancreatic ascites, fistula, or pseudocyst. Factors associated with success by logistic regression analysis included age < 55 years and placement of pancreatic stent. According to the presenter, the effect of age is lessened in their most recent unpublished updated analysis.
A small Japanese study examined the histologic findings in the major duodenal papilla with regard to diagnosis of autoimmune pancreatitis (AIP). In 3 surgical specimens and 3 endoscopically-obtained biopsies of patients with known AIP, the level of IgG4-positive plasma cells by immunohistochemistry was much greater than in controls. Thus, biopsy of the papilla may be a useful adjunct to serum IgG4 and/or pancreatic biopsy in the diagnosis of this rare condition.
A study from Minneapolis described improved outcome after gastric pacemaker implantation in 7 patients with refractory confirmed gastroparesis who had previously undergone endoscopic therapy for SOD or pancreatic divisum. While pain was often relieved by the initial pancreaticobiliary endotherapy (4/7), the refractory nausea and vomiting in these patients had a good or excellent response in 4 of 7, with all 3 TPN dependent patients getting off TPN. The authors stressed that gastroparesis may be part of a spectrum of motility diseases including those of the pancreas and should be investigated in these patients. Additionally, they viewed this preliminary small series as promising for the potential role of gastric pacemakers in this setting.
Lastly, an animal study from Finland showed the feasibility of placing a biodegradable pancreatic stent. These stents were placed surgically in 5 swine; by 3 months the stents were no longer visible radiographically. Laparotomy at 6 months showed total absence of the stent in 4 of 5 swine, and histopathological analysis showed only mild ductal dilation at the pancreatic head. Similar stents have been placed endoscopically into the bile duct in swine. Clinical trials are beginning in humans; these stents may prevent the need for repeat endoscopy after ERCP for stent removal, and may have fewer complications than traditional plastic stents.
Barrett’s Esophagus and Esophageal Cancer
Development of Esophageal Adenocarcinoma in Patients with Barrett’s Esophagus and High Grade Dysplasia Undergoing Surveillance: A Meta-Analysis and Systematic Review
This presentation reported on a meta-analysis covering multiple databases (Medline, Pubmed, Ovid, Cochrane, pharmaceutical abstracts) that included patients with histologically proven Barrett’s and high grade dysplasia who did not undergo endoscopic or surgical therapy and who did not have esophageal cancer at the time of enrollment or within 6 months. Follow-up was reported in patient years. Four articles were included in the analysis covering 236 patients and 1240.5 patient years. A total of 69 esophageal cancers occurred with a crude incidence rate of 5.57 per 100 patient-years and a direct standardized incidence rate of 6.58 per 100 patient-years.
A 5-Year Randomized Phase III Trial of Efficacy and Safety of Photodynamic Therapy Using Porfirmer Sodium in High Grade Dysplasia in Barrett’s Esophagus
This paper reported the 5 year follow up data from a previously published 2 year data of this group’s trial of PDT therapy in patients with HGD in Barrett’s esophagus. In that initial report, 77% of patients receiving PDT had ablation of their HGD compared to 39% of patients receiving omeprazole alone (p < 0.0001). The protocol enrolled 138 patients receiving PDT vs 70 patients receiving omeprazole alone. All patients had 4 quadrant jumbo biopsies every 3 months. At 5 years, the PDT group had similar significant improvement in efficacy in eliminating HGD compared to the omeprazole group alone to the differences observed after 2 years (p < 0.0001) as well as significant long term efficacy in preventing progression to cancer, [15% vs 29% (p = 0.027)] though the study was not powered to evaluate progression to cancer.
Curative Endoscopic Therapy for Barrett’s Early Cancer and High Grade Dysplasia: Long Term Results in 304 Patients
This paper reflected a long term follow up of results of an extensive series of EMR ro early cancer and high grade dysplasia in Barrett’s from a center with some of the largest experience with this technique. 304 patients with either HGD (n=45) or early adenocarcinoma (n=259) were included in this study from 1996-2001. EMR was performed in 215 patients with 605 resections and PDT was used in 72 patients. 12 patients had combination PDT and EMR. 5 patients had APC ablation. Mean follow-up was 69.5 months. Complete remission was achieved in 262 patients (86%). Metachronous lesions found in 64 patients (21%) upon follow-up with all but 5 receiving repeat endoscopic therapy. The 5 remaining underwent surgical resection for deep infiltration. It is unclear if the metachronous lesions were truly metachronous or local recurrences. 34 patients died but only 2 from esophageal malignancy. 5-year survival rate was 89%. Overall complication rate was 15% with significant bleeding occurring in only 4 patients and no perforations reported. It will be interesting to see how generalizable these results will be to other centers with less volume and expertise in EMR.
Multi-Center Clinical Trail Using Endoscopy and Endoscopic Ultrasound Guided Fine Needle Injection of Anti-Tumor Agent (TNFerade) in Patients with Locally Advanced Esophageal Cancer
This study reported on a Phase I trial using TNFerade injections against locally advanced esophageal cancer. TNFerade is a 2nd generation replication-deficient adenovector that carries the transgene encoding for human TNF. It is inducible by radiation via Egr-1. This was a dose escalation study that looked at the maximum tolerated dose, safety, and pathologic response rate. All participants had resection after the end of radiation. 24 patients were enrolled in the study. Majority of tumors were adenocarcinomas (20/24), T3 (23/24) and N1 (18/24). In the three higher doses, partial complete response was seen in 6/15 (40%) with disease free states in 10/17 (59%). Mean follow-up was 21 months. Adverse events included fatigue (54%), fever (38%), thromboembolic events (33%). This is the second paper presented at this meeting in which this gene therapy chemoradiation sensitizer is being introduced via EUS-guided fine needle injection, reflecting a new avenue in the expanding realm of EUS guided therapy.
Endoscopic Submucosal Dissection for Early Esophageal Cancer
This paper presented a large series of endoscopic submucosal dissection (ESD) with a hook knife for consideration of this technique as a potentially preferable alternative to the use of EMR for early esophageal cancer. While the authors noted that EMR is currently recommended by the Japanese Esophageal Association as the treatment of choice for early esophageal cancers, they pointed out that the lateral growth of these tumors often precludes effective en bloc resection by EMR. This study looked at 130 patients who underwent endoscopic submucosal dissection for superficial SCC of the esophagus. The median size of the resected specimen was 32mm (range 8-76 mm) while the cancer was 28mm (4-64mm). Complete resection, defined as en block resection with negative vertical and lateral margins, was seen in 93.8%. The local recurrence rate with a mean follow-up of 28 months was 0%. In terms of complications, 10 cases of mediastinal emphysema were conservatively managed. Despite these impressive results, further study of the widespread safety and difficulty of this technique and controlled comparison with EMR appears warranted.
Endotherapy vs Esophagectomy for Barrett’s Esophagus with High Grade Dysplasia or Intra-mucosal Carcinoma: A Review of Outcomes
This study compared the efficacy, risks, and costs of endoscopic therapy versus surgical intervention in the treatment of Barrett’s esophagus with high-grade dysplasia or intramucosal carcinoma. The study was a single center design and prospectively followed 64 patients who underwent endoscopic therapy and 32 patients who underwent esophagectomy. Average follow-up time was 20 months for the endoscopic group and 12 months for the surgical group. There were no deaths attributable to esophageal cancer in either group. Invasive cancer developed in 6% of the endoscopic group while no distant cancers developed in the surgical group. There were more complications, both major and minor in the surgical group (31% and 65% respectively) when compared against endoscopy (9% and 29%). Median costs were also higher in the surgical group ($66,060) when compared to the endoscopic group ($40,079). While this comparison was not randomized, it well characterized some of the relative benefits and drawbacks of the two approaches.
New Technology: Diagnosis
This session included several papers on emerging technologies to perform optical biopsies using ultra high magnification probes, zoom narrow band imaging, optical coherence tomography (OCT), and Raman spectroscopy. Can endo-cytoscope system (ECS) predict histology in neoplastic lesions? In this paper, Eberl et al. describe a probe based system which produces images at 450x and 1100x magnification. This method required methylene blue pre-staining using a spray catheter. This series examined the findings in over 70 patients, and found a 91% sensitivity for diagnosing pathology in the esophagus and colon, with somewhat less ability in the stomach, perhaps due to secretion.
Bittinger et al presented Zoom-narrow band imaging (NBI) as a tool for the detection of Barrett’s metaplasia (BM)—helpful or unnecessary? In this paper, the authors showed that NBI significantly increased the sensitivity and negative predictive value in detecting the presence of Barrett’s esophagus to values of 93% and 92%. Of interest, Barrett’s was found with high frequency in this group—43% of the 35 individuals studied with upper gastrointestinal complaints, perhaps due to the protocol of taking biopsies just the GE junction as the gold standard and comparing blinded interpretations of videotaped examinations under white light and NBI filtered light to predict the presence or absence of Barrett’s. The authors also noted that they relied more upon the NBI mode than the zoom function to diagnose Barrett’s. The methodology was interesting in that NBI-targeted biopsies to increase the yield of positive Barrett’s diagnosis were not performed.
Intraductal optical coherence tomography (OCT) for investigating main pancreatic duct strictures.
In this presentation, Pieralberto et al. placed an OCT probe though a standard ERCP catheter and showed images of the pancreatic duct wall in 12 patients with either normal MPD, chronic pancreatitis, or neoplasm, with the probe inside and outside of the catheter. Loss of the distinct pattern of 3 layers was accurate in all cases of neoplasm. The authors claimed that the technique was relatively easy to learn and perform, though it does require successful negotiation of a wire and catheter across the MPD stricture.
In their paper, Diagnostic accuracy of Raman spectroscopy for the classification of dysplastic lesions in Barrett’s esophagus, Song et al. have performed extensive targeted biopsies in a large group of patients with Barrett’s using a reusable flexible probe using near infrared light to create spectral algorithms for predicting the presence of dysplasia using this contact probe with high sensitivity, specificity, and accuracy. It is unclear whether this will be a practical technique for screening a large area of Barrett’s or more of a confirmatory test for areas suspicious on other optical techniques geared to highlight suspected areas of dysplasia during an initial scan.
In addition, one new technique was presented to allow access to the gastric remnant following gastric bypass surgery for diagnosis and therapy. In their paper, Endoscopic evaluation of the defunctionalized stomach using Shapelock technology, Pai et al. showed a means of using the Shapelock device previously shown to prevent looping during colonoscopy to allow relatively rapid passage of an enterosocpe into the distal pancreaticobiliary limb of 3 of 4 patients who had undergone Roux Y gastric bypass surgery. Present models are 16 mm wide which may pose some concern regarding passage via the gastrojejunal anastamosis, and can only accommodate an enterosocpe inside the overtube. While previously accessible with great time and difficulty, a quick and safe method of reaching well up pancreaticiobiliary limbs will have an important impact on both diagnostic imaging and therapeutic capabilities in this group of patients.
Finally, a new method of performing mucosal oximetry using visible light spectroscopy was validated in a comparison of normals with individuals with mesenteric ischemia pre and post remedial stenting. Diagnosis of chronic mesenteric ischemia by visible light spectroscopy during endoscopy. In this paper, oximetry of mucosal capillaries was assessed using a new visible light spectroscopy catheter through the biopsy channel. The observed improvement in mucosal oxygen saturation after successful angioplasty in 3 patients was an impressive preliminary demonstration of the potential utility of this new area for endoscopic diagnosis.
In a time when a number of disruptive technologies are threatening a decline in standard diagnostic endoscopy, these papers offered a glimpse of how technological advances might allow endoscopists to make diagnoses more efficiently, in previously difficult to assess locations, and for conditions not previously possible via the endoscope.
Imaging Advances in the Biliary Tree:
Pilot Evaluation of a Direct Visualization Probe for Intraductal Examination of the Biliary System
The purpose of this study was to evaluate the feasibility of using a miniprobe for direct intra-ductal visualization of the biliary tree in patients referred for ERCP. The fiber optic miniprobe system (PolyDiagnost, Inc.) consisted of a 0.9mm, 6000pixel, 70? viewing field cholangioscope. 29 patients were enrolled in the study and standard ERCP was performed. Successful biliary access was achieved in 23/29 (79%) via the miniprobe system. The most common indications for ERCP included obstructive jaundice (65%) and pancreatitis (13%). Complete visualization of the area of interest was rated as extremely clear in 78% of exams. Technical limitations were experienced in 8.7% of cases. Direct miniprobe cholangioscopy increased total procedure time by an average of 8.4 minutes. Direct visualization of the biliary tree altered the final diagnosis in 43.5% of cases. No adverse events were noted. The authors concluded that direct visualization of the biliary system using the miniprobe system is safe, effective, and may alter diagnosis and/or refine the therapeutic plan in a significant portion of patients.
A Novel Peroral Cholangiopancreatoscopy System (Spyglass™) with Four-Way Deflection for Direct Access, Visualization, and Biopsy
Results from the First Human Use Clinical Series Utilizing a New Peroral Cholangiopancreatoscopy System (Spyglass™ Direct Visualization System)
Peroral cholangiopancreatoscopy (PC) has been used effectively for direct visualization of the pancreaticobiliary system , direct tissue sampling, and therapeutic interventions such as lithotripsy of large bile duct stones. This paper presented a new system for PC called Spyglass™ which confers several theoretical advantages over existing means of PC via 9-10 mm diameter baby scopes. Spyglass™, manufactured by Boston Scientific allows for a single operator, and 4-way deflected steering with separate working and accessory channels. Prior to the development of Spyglass™, cholangioscopes were restricted to two-way deflection, and dependence on multiple operators. This paper described a bench simulation designed to test the four-way steerability of Spyglass™. The system allowed access to all endoscopic visual quadrants in 100% of attempts, whereas a conventional cholangioscope (used as a control) could access target lesions in all 4 quadrants in only 48% of attempts when a biopsy forceps was loaded. Spyglass™ directed biopsies obtained an adequate gross specimen in 91% of cases and an excellent to adequate histology sample in 90% of cases. Thus, Spyglass™ was effective in this model for direct access, visualization, and biopsy of pre-specified targets.
The same investigators went on to report the first human use of Spyglass™ to determine the clinical utility of the system for direct visualization, diagnosis, and therapy as compared with conventional ERCP methods. 22 consecutive patients were enrolled undergoing ERCP for various indications including indeterminate stricture, suspected malignancy, and EHL stone therapy. CP directed biopsy was accomplished in 91% of patients. In 91% of patients, Spyglass™ was felt to have provided some benefit over standard ERCP techniques, such as the ability to confirm or exclude malignancy, diagnose the etiology of indeterminate filling defects, and perform EHL therapy in those who had failed conventional stone extraction in ERCP. They concluded that Spyglass™ is a safe and technically feasible tool for pancreaticobiliary investigation. With further evaluation and refinement, the implication is that direct visualization may gain an increasing role relative to fluoroscopy in evaluating the pancreaticobiliary tree.
Characterization of Dominant Bile Duct Stenoses in Patients with PSC Using Cholangioscopy
Conventional ERCP assisted sampling of the biliary tree often cannot distinguish between benign and malignant strictures in patients with PSC. These authors evaluated the role of cholangioscopy as a tool for distinguishing between benign and malignant dominant bile duct strictures in 53 patients with PSC. Twelve of these patients were found to have cholangiocarcinoma. The sensitivity and specificity of cholangioscopy in making the diagnosis was 92% and 93%, respectively. The positive predictive value (PPV) and negative predictive value (NPV) of cholangioscopy for cancer was 79% and 97%, respectively. This was significantly better than conventional ERCP, and suggests that transpapillary cholangioscopy be considered as an underutilized tool for distinguishing benign from malignant strictures in PSC.
Prospective Evaluation of Intraductal Ultrasound (IDUS), Digital Image Analysis (DIA), Fluorescnecve in situ Hybridization (FISH), CA 19-9, and ERCP with Routine Cytology and Intraductal Biopsy in the Evaluation of Indeterminate Bile Duct Strictures.
Standard techniques for evaluating indeterminate bile duct strictures are notoriously low in sensitivity. The objective of this study was to determine the sensitivity and specificity of newer diagnostic modalities to assist in distinguishing between benign and malignant biliary strictures. Eighty-six patients were evaluated. The modality with the highest relative sensitivity was IDUS at 87%. Specificity was rated as 92%. IDUS, and tests that identify aneuploidy such as DIA, and FISH, were found to be useful adjuncts to standard techniques in assessing indeterminate bile duct strictures, diagnosing malignancy in 67% of patients with false negative cytology and histology.
Defining the Difficult ERCP: Performance of the ASGE (Modified Schultz) Scoring System in a Prospectively Studied Cohort of 5264 Procedures.
This paper stemmed from ASGE recommendations that endoscopists should record their technical success and unplanned events for patients undergoing ERCP, and the importance of correctly factoring in for procedure difficulty when assessing success and outcome data. The authors aimed to evaluate the performance of the modified Schultz score as a means by which to document ERCP difficulty. They reported data from 5264 unselected ERCPS, performed by 182 endoscopists. 3,707/5,264 [70%] ERCP’s were performed successfully. Scoring divided procedures into three separate groups: 1) Standard diagnostic/therapeutic; 2) Billroth II (diagnostic), stones >1cm, IHD stent; 3)Previous Whipple, therapeutic Billroth II, pancreatic duct therapy, IHD stones, lithotripsy, manometry. 74% of patients in Group 1, 68% of patients in Group 2, and 51% of patients in Group 3 were deemed to have undergone successful ERCP’s. Complication rates were 4.8,4.9%, and 7.2% in Groups 1,2, and 3, respectively. The outcomes correlated with modified Schultz score, but the differences observed were small, particularly between type 1 and type 2 cases.
Acquisition of ERCP Skills During Fellowship Training
This study looked at the important question of acquisition skills in ERCP during gastroenterology fellowship training. The ASGE has defined competency as the endoscopist’s ability to successfully complete the technical task in 80% of cases. This study prospectively evaluated 3,529 ERCP’s performed by fellows during training. The fellows each performed a median of 179 ERCP’s. Competency, as defined by the above criteria, was achieved by all fellows evaluated in this study. The study supports the ASGE position that while numbers alone do not confer or guarantee competency, 180-200 cases are generally required before competency can be assessed. This paper also reinforced the concept that there is marked variability in the rate at which each individual is able to achieve competency in these skills, again supporting to emphasis in competency assessment today on demonstrating actual procedure success rather than on total number of cases performed.
Contemporary Approaches to Small Bowel Imaging
This session reviewed the most recent data evaluating the cost-effectiveness and utility of double-balloon enteroscopy (DBE), capsule endoscopy (CE), and the most efficacious approach to imaging the small bowel. Mehdizadeh and colleagues presented data comparing DBE with CE in a multi-center trial, entitled “Double Balloon Enteroscopy (DBE) Compared to Capsule Endoscopy (CE) Among Patients with Obscure Gastrointestinal Bleeding (OGIB): A Multicenter U.S. Experience.” These authors evaluated all cases from six tertiary care centers within a 1-year period. Of 188 eligible patients, DBE was performed in 69.1% (130/188). 54% of patients undergoing DBE had a potential bleeding source seen on previous CE. 45% of patients who underwent DBE had negative CE prior to DBE. Of the patients with a possible bleeding source seen on CE, 65% had a lesion seen on DBE as well. Of the patients with an initial negative CE, 30% had lesions found on DBE. To determine agreement between the individual imaging modalities, kappa statistics were performed and revealed a k = 0.74 for ulcers (0.52-0.96) suggesting high agreement between the two modalities, and a k = 0.59, and 0.42 for large masses and AVMs, respectively, suggesting moderate agreement. Thus, DBE and CE appear complementary in that CE can be used as an initial screening test, while DBE can serve as a confirmatory and/or therapeutic modality.
Pennazio et al. in their study “Outcome of Double-Balloon Enteroscopy after Capsule Endoscopy in Patients with Suspected Small-Bowel Disease,” assessed technical and clinical outcomes of DBE in patients with previous CE and suspected small bowel disease. 44 patients with various conditions such as OGIB, polyposis syndromes, refractory celiac disease, or suspected Crohn’s Disease were evaluated. DBE was able to confirm findings seen in CE in 77% of patients. A new diagnosis was made in 14% of patients. A suspected diagnosis was excluded in 9% of cases. DBE findings played an important role in 32/44 (77%) of patients. The conclusion reached from this data was that DBE served as a complementary modality to CE helpful to further characterize and manage small bowel lesions seen on CE. Prior CE can also help determine whether peroral or rectal DBE should be performed first depending on the location of the lesion seen on CE.
The yield of back-to back capsule endoscopy was also evaluated during this session with Kimble et al. discussing their data entitled “Variation in Diagnostic Yield of Back-to-Back Capsule Endoscopy in Obscure GI Bleeding: Preliminary Results.” Eighteen patients were evaluated, all of whom suffered from obscure GI bleeding. Two capsule endoscopies were performed within 4 days of one another. 6/15 of the second capsule studies detected new lesions which were not seen on the first capsule study. Of those lesions seen on this second capsule study, five were deemed to be clinically significant including angioectasia, erosions, and ulcers. This study suggests that repeating capsule studies may be helpful in visualizing lesions not seen on an initial CE, and thus may increase the sensitivity of capsule modality as a modality of small bowel imaging.
Kamal et al. detailed in a cost-benefit analysis that in a standard patient with obscure occult GI bleeding, DBE via oral insertion was the most cost effective approach for an initial evaluation with a cost-effectiveness ratio of $6000 per patient successfully treated compared to no therapy.
Saperas et al. further compared modalities for diagnosing the source of obscure GI bleeding to determine whether CE was superior to conventional and CT angiography in the work-up of obscure occult GI bleeding. In their study entitled “A Prospective Comparison of Capsule Endoscopy, Helical CT Angiography, and Mesenteric Angiography for the Diagnosis of Obscure Gastrointestinal Bleeding (OGIB),” CE was able to determine a source of bleeding in a greater proportion of patients than CT-angiography, conventional angiography, and CT-angio + conventional angiography (68% vs. 21% vs. 50% vs. 61%,) respectively. The most frequent CE findings were angioectasias, fresh blood, and jejunal ulcers. CE is superior to CT-angiography in the diagnosis of OGIB. This would imply that CE may be a better first line diagnostic modality in patients with OGIB who are hemodynamically stable.
Finally, Lai et al. described the long-term follow up of patients with an initial negative CE. Forty-nine patients who underwent CE were followed for at least 1 year for clinical overt and occult bleeding. Median follow-up was 19 months. Cumulative rebleeding rate in those patients with an initial negative CE was significantly lower than in those with a positive CE [5.6% vs. 48.4% p=0.03], respectively. The negative predictive value of an initial negative CE was calculated to be 94.4%. As patients with obscure GI bleeding with an initial negative CE had a very low rebleeding rate, this suggests that such patients can be deferred from further invasive investigations.
Pediatric Endoscopy
There were six clinical topics discussed at the ASGE sponsored forum entitled Pediatric Endoscopy. The first paper entitled Safety of Endoscopist Versus Anesthesiologist Administered Sedation for Pediatric Endoscopy: In and Outside the Operating Room was presented by Jennifer Lightdale, MD from Boston Children’s Hospital. This was a prospective analysis which looked at the adverse events of endoscopist versus anesthesiologist administered sedation in and outside of the operating room. Adverse events were reported in forty one percent of the cases (2366/5793), predominantly agitation (44%) followed by respiratory distress (22%). Anesthesiologist-associated adverse events occurred more frequently in the endoscopy suite than in the operating room. Agitation was more commonly associated with endoscopist administered sedation, whereas, respiratory complications were more often associated with anesthesiologist administered sedation. Unique Endoscopic Features of Corpus Gastritis with H. Pylori Infection in Children and Adolescents; Close Observation By Conventional Endoscopy and Magnified Endoscopy was presented by Yoshiko Nakayama from Japan. With a magnifying endoscope, the endoscopic appearance of punctate erythema in the corpus correlated with a regular arrangement of collecting venules (RAC), which is found in normal gastric tissue. This study reviewed the endoscopic findings seen in H pylori gastritis and suggested that the lack of corpus punctate erythema may be the only endoscopic aberrant finding seen in some cases of H. pylori gastritis. Another topic entitled Immediate Complications of Upper Endoscopy in Children was a multi-center retrospective database survey conducted at Baylor College of Medicine from 1999-2003 that enrolled 13 institutions. The complication rate was estimated to be 2.3% based on 10,236 upper endoscopies. The most common complications included hypoxia (66%), bleeding (12%), and respiratory distress (4%). Although there was no difference in the complication rate amongst gender and race, those children who received intravenous sedation, who had younger age, or had a higher ASA class were more likely to develop complications. There were two topics presented from Loma Linda University Children’s Hospital: Wireless Capsule Endoscopy (WCE) in Pediatric Gastroenterology Practice and Endoscopic and Surgical Findings of Failed Fundoplication in Children: Experience from Re-Do Fundoplication. WCE was a retrospective review of 44 pediatric patients who underwent WCE; Thirty four children were able to swallow the capsule, whereas, 10 capsules were placed endoscopically. Debilitating abdominal pain and chronic diarrhea were the most common symptoms. Twenty-five patients had findings consistent with Crohn’s disease. It appears that WCE may be a valuable diagnostic tool among the pediatric population and helpful in the evaluation of children with suspected Crohn’s disease. The other topic was a retrospective chart review from 1997-2005 that involved 40 patients with a fundoplication. The most common presenting symptoms suggestive of a failed fundoplication were feeding intolerance, recurrent pneumonia, and upper gastrointestinal bleeding. Early endoscopic findings of a failed fundoplication is often splitting or fanning of the fundoplication fold. The last topic entitled Prospective Evaluation of Impact of Endoscopic Management on Sphincter of Oddi Dysfunction in Children was a prospective study conducted in Alabama that included six patients with suspected sphincter of Oddi dysfunction (SOD). Although the reference standard for SOD in children is unknown, the mean basal pressure in this subset of children was 110mmHg (112-182). The authors claimed that pancreatic and/or ductal sphincterotomy may be effective and safe in expert hands in a small subgroup of children with sphincter of Oddi dysfunction.
Pancreatic EUS in Clinical and Translational Research
Endoscopic ultrasound is increasingly being utilized to assist in the diagnosis and management of pancreatic diseases. In this session, a number of exciting developments regarding the application of EUS in chronic pancreatitis, pancreatic cysts, and pancreatic cancer were discussed.
Varadarajulu et al presented their findings regarding the correlation of EUS findings with histopathological specimens in the diagnoses of non-calcific chronic pancreatitis. The group prospectively analyzed all patients undergoing EUS for pancreatico-biliary indications and who underwent subsequent pancreatic surgery. Patients were excluded if they had calcific chronic pancreatitis or received external beam radiation pre-operatively. The final results demonstrated that of the 42 patients meeting inclusion criteria, 50% had a histopathologic diagnosis of non-calcific chronic pancreatitis. Four or more EUS criteria provided the best sensitivity (90.5%), specificity (85.7%), and accuracy (88.1%). EUS features significantly associated with histological findings included the following: parenchymal foci, stranding, and lobularity; dilated and/or irregular main and/or side branches. Furthermore, the number of EUS features correlated with the presence of and severity of non-calcific chronic pancreatitis on histological analysis. Overall, the results of this study strongly support the use of EUS in diagnosing and assessing the severity of chronic pancreatitis, and may soon be the initial test of choice in those patients with suggestive symptoms and lack of classical radiologic findings.
An interim report by the multicenter pancreatic cyst DNA analysis (PANDA) study group was presented. The investigators described the utilization of DNA analysis on pancreatic cyst fluid obtained via EUS-FNA to detect the presence of malignancy in patients with confirmed histology of mucinous cystic neoplasms (MCN) and intraductal papillary mucinous neoplasms (IPMN). The modalities specifically evaluated included DNA quantification with optical density, DNA quality analysis by cycle threshold on qPCR, allelic loss analysis of 15 microsatellites, and k-ras point mutation analysis by direct sequencing. The results revealed that of the 187 patients enrolled in the study, 41 had confirmed histological diagnosis of MCN or IPMN. The investigators demonstrated that premalignant lesions were differentiated from malignant lesions on the basis of DNA quantity, DNA quality, and k-ras allelic loss, providing further support that detailed DNA analysis may be essential adjuncts to the radiographic and endoscopic evaluation in characterizing cystic lesions of the pancreas.
The results of two studies demonstrating the use of EUS in the management of pancreatic cancers were presented. Farrell et al described their experience with intra-tumoral delivery of a gene therapy agent in patients with locally advanced pancreatic cancer. In this phase II study, they compared the effects of TNFerade, an adenoviral vector containing the human TNF-alpha gene and its chemoradiation-inducible promoter Egr-1, by comparing its delivery by either percutaneous injection or by EUS injection, followed by chemoradiation. The results demonstrated that the maximally-tolerated dose of this therapy was associated with greater loco-regional control of the tumor, longer progression-free survival, higher proportion of patients with stable or decreasing CA 19-9, greater percentage of patients going to resection, and overall modestly improved median survival. Furthermore, modes of delivery were similarly safe with no difference in disease progression between the two groups. Ashida et al presented the results of their study of DNA microarray analysis on pancreatic cancer tissue obtained at EUS-FNA. They demonstrated the ability to measure gemcitabine-resistance-related genes in obtained tissue and how this modality can help predict chemotherapeutic response in patients with advanced pancreatic carcinoma. Overall, the results of these two studies provide a glimpse into the future regarding the use of EUS in helping to tailor chemotherapeutic regimens for pancreatic cancer patients based on their genotype as well as to actually deliver specific therapeutic agents directly into the tumor.
The final two abstracts presented in the symposium evaluated the diagnostic capabilities of EUS. Canto et al looked at a group of high-risk individuals for pancreatic cancer (strong family history or history of Peutz-Jeghers syndrome) and correlated their EUS findings with that of ERCP, pancreatic function, and pathology. They found that 72.3% of high-risk individuals had EUS findings consistent with chronic pancreatitis; 85.6% of patients who underwent both EUS and ERCP had mild ERCP abnormalities of duct irregularity, main duct dilatation, branch duct ectasia, or saccules. Pancreatic function was measured by serum trypsin values, which were similar in high-risk individuals when compared to patients with known chronic pancreatitis or pancreatic cancer, but overall lower when compared to normal controls. However, mean peak bicarbonate levels in pancreatic juice were similar in between the high-risk individual group and normal controls. EUS features of chronic pancreatitis also correlated with PanIN density and PanIn associated lobulocentric atrophy on pathologic analysis. Finally, Puli et al performed a meta-analysis of the diagnostic accuracy of EUS for vascular invasion in periampullary and pancreatic cancers. Pooled analysis of results from 29 studies revealed an overall sensitivity of 73% and specificity of 90.2% for detection of vascular invasion by tumor. The authors concluded that these numbers are not as high as has been suggested and that further refinements in technique and technology may be warranted to improve detection of vascular invasion.
Clinical Innovations in EUS
Five of the six abstracts presented during the “Clinical Innovations in EUS” session examined the role of endoscopic ultrasound along with fine needle aspiration (FNA) in the staging of cancer.
Dr. Michael Wallace from the Mayo Clinic, Jacksonville presented an abstract entitled “Complete ‘Medical Mediastinoscopy’ Under Conscious Sedation.” The hypothesis was that endoscopic and endobronchial ultrasound (EBUS) with FNA was superior to blind bronchoscopic (“Wang”) FNA, and that the combination of EUS and EBUS allows for complete mediastinal lymph node staging in lung cancer. Thirty five patients with 51 malignant lymph nodes were examined, of whom 24 underwent surgical resection. On a per lymph node basis, the sensitivity of EUS/FNA was 69% compared with 75% and 35% for EBUS/FNA and bronchoscopy/FNA, respectively. Combining EUS and EBUS yielded a sensitivity of 98% on a per lymph node basis. The authors concluded that the combination of EUS and EBUS allowed for near-complete, non-surgical staging of the mediastinum in patients with lung cancer.
The second abstract from the University of Alabama, Birmingham was titled “Impact of Staging Endosonography on the Treatment and Survival in Patients with Non-Small Cell Lung Cancer.” Dr. Mohamad Eloubeidi presented prospective data on 125 patient with non-small cell lung cancer (NSCLC) who were referred for EUS/FNA . EUS confirmed malignant N2 or N3 disease in 52% of patients. Those with positive lymph nodes were more likely to undergo chemotherapy and/or radiation therapy and less likely to undergo surgery. The median survival in patients with node positive disease was 1.18 years compared with 1.67 years for those with node negative disease. The conclusion was that EUS/FNA was the most important predictor of survival in patient with NSCLC.
The third abstract, “EUS Can Accurately Detect Liver Metastases,” was presented by Dr. Pankaj Singh from Central Texas Veterans Health Care System. CT scan and EUS/FNA were performed on 130 consecutive patients with primary cancers of the lung, pancreas, esophagus, stomach and colon; 26 of 132 patients had liver metastases. EUS/FNA was able to identify 24 of the 26 patients, yielding a sensitivity of 92% (compared with 96% for CT scan). However, specificity of EUS was 100% (vs. 92%), positive predictive value of EUS was 100% (vs. 74%) and diagnostic accuracy of EUS was 98% (vs. 93%). The authors concluded that EUS/FNA is a safe and accurate tool in the diagnosis of liver metastases and may be especially useful for lesions that are too small to characterize by CT scan.
The fourth presentation was the only non-cancer related EUS abstract. Dr. Douglas Faigel from Oregon Health and Science University discussed “Economic Realities of EUS in an Academic GI Practice.” Given that EUS is both time-intensive and technically demanding, the aim of the study was to determine the relative reimbursement of EUS as compared with EGD, colonoscopy and colonoscopy with polypectomy. Medicare had higher professional fee reimbursements for EUS compared to EGD and colonoscopy without polypectomy. However when actual reimbursements were calculated for an average ½ day of EUS versus a ½ day of standard endoscopy, it was determined that compensation for standard endoscopy was 2.1 times higher. This was due to the time allotted for procedures: EUS required 90 minutes per procedure (allowing for only 3 procedures per ½ day) while standard endoscopy required only 30 minutes per procedure (allowing for 7-10 procedures per ½ day).
“EUS for Prostate Cancer” was presented by Dr. Everson Artifon from the University of Sao Paulo. Fifteen patients scheduled to undergo radical prostatectomy had EUS/FNA performed. The tumor staging of EUS yielded a sensitivity of 75%, a specificity of 91%, and a diagnostic accuracy of 84%. For nodal staging, EUS had a sensitivity of 61%, specificity of 76% and a diagnostic accuracy of 64%. The authors concluded that comparative studies between EUS and conventional rigid rectal probe are warranted.
Lastly, Dr. Michael Levy from the Mayo Clinic, Rochester, presented “EUS FNA Detection of Malignant Iliac Lymph Nodes in Rectal Cancer.” While the role of EUS for T and N staging in rectal cancer is well established, its use in evaluating iliac lymph nodes (which designate metastatic disease – M1) has not been investigated. In this prospective study, 457 patients with rectal cancer underwent T, N, M staging by EUS. Suspicious iliac lymph nodes (ILN) were identified in 32 of 457 patients (7%), of which 15 of 32 (47%) were found to be malignant by FNA. CT scan detected ILN in only 7 of the 15 with confirmed malignant ILN. The finding of malignant iliac lymph nodes altered care in 85% of patients, usually leading to expansion of the radiation field or extended lymphadenectomy. These data support the routine examination of ILN by EUS and highlight the inadequacy of rigid instruments in the staging of rectal cancer.
Endoscopic Safety and Complications
Randomized Controlled Trial of Sedation for Colonoscopy: Entonox Versus Intravenous Sedation
Inhaled Entonox (50% nitrous oxide: 50% oxygen) was compared to IV Midazolam/Fentanyl for sedation for colonoscopy. Elective colonoscopies for patients without a history of colonic resection were included. VAS pain score, endoscopist/nurse assessment, patient satisfaction score, recovery assessment (letter cancellation test), and recovery times were better for Entonox as compared to IV sedation. Completion rates were similar. No complications are reported for either groups (131 patients so far)
Prospective Determination of Bacteremia and Associated Complications in Patients Undergoing EUS FNA of Rectal and Perirectal Lesions
The need for routine use of prophylactic antibiotics for EUS FNA of rectal and perirectal lesions has not been studied previously. 76 patients undergoing above procedure received 3 sets of blood cultures: prior to study, after diagnostic flex-sig, 15min after EUS-FNA. 4 patients developed positive blood cultures: 2 were considered to be contaminants, 2 were considered to be positive (1 before FNA, 1 after). The results suggest that routine use of prophylactic antibiotics is not necessary for EUS FNA of rectal and perirectal lesions.
Endoscopic Resection in Esophagus and Stomach Is Safe: A Prospective Analysis of 303 Procedures
303 endoscopic mucosal resection procedures were prospectively evaluated for complications. Most (261) were for Barrett’s esophagus. 231 ER’s were performed with ER-cap technique, 66 with multi-band mucosectomy, and 6 with the lift-and-snare technique. Acute complications occurred in 19% of procedures, mostly bleeding episodes which were stopped endoscopically. 3 perforations were noted – 2 in the esophagus treated conservatively and 1 in the stomach closed surgically. 1% of patients developed “delayed” bleeding (>48hrs).
Comparison of Proton Pump Inhibitor with H2 Receptor Antagonist for Prevention of Bleeding from Ulcer After Endoscopic Submucosal Dissection for Early Gastric Cancers: A Non-Blind Randomized Controlled Trial Noriya Uedo et al
ESD provides more complete resection for Early Gastric Cancer than Endoscopic Mucosal Resection, but is associated with higher complication rates, especially for delayed bleeding (>5%). 105 patients received either PPI (labeprazole 40mg po bid) or H2RA (cimetidine 800mg po bid) from 1 day prior to ESD and continued for 8 weeks. Bleeding rate was significantly lower in PPI group (1.8%) than H2RA group (12%).
Prospective Randomized Trial Evaluating Ketamine for Advanced Endoscopic Procedures in Difficult to Sedate Patients
Ketamine, a dissociative anesthetic, was studied in patients receiving EUS/ERCP who were considered to be difficult to sedate (meperidine 50mg, midazolam 5mg, diazepam 5mg). 175 patients were included in the study – randomized to Ketamine 20mg q3-5min vs continuing standard IV sedation. 35.5% of standard sedation had to be crossed-over to Ketamine therapy for failed sedation as compared to 3.7% randomized to Ketamine. The results favored Ketamine for physician rating of sedation, patient discomfort, and technical difficulty of the procedure. Procedure completion rate, procedure time, and adverse events did not differ between the groups.
Use of Hemostatic Clips in Patients Undergoing Colonoscopy in the Setting of Coumadin Anticoagulation Therapy
Patients on coumadin anticoagulation (CAC) are thought to have 10-fold increase in risk of post-polypectomy bleeding. 100 patients on CAC undergoing colonoscopy were included in the study. ASGE criteria for level of risk of interruption of CAC was used: low risk patients held coumadin for 4 days; high risk patients held coumadin for 48 hrs. 68 of 100 patients had polyps; 70% had at least 1 adenoma. 153 polyps were removed – size ranged from 5 to 20mm. 3 different clips were used at the discretion of the endoscopist (all three were available free of charge). Clips were used on all polyps requiring intervention beyond forceps biopsy and small cold snares. Coumadin was restarted on same day in 90% of patients. Delayed bleeding occurred in 1 patient. No thromboembolic events were noted.
Natural Orifice Translumenal Endoscopic Surgery
Measurements of Intraperitoneal Pressure During Flexible Transgastric Surgery and the Development of a Feed-Back Control Valve for Regulating Pressure.
Unlike conventional laparoscopic surgery, transgastric surgery does not monitor intra-peritoneal pressure. Over-inflation can compromise respiratory and cardiovascular function. In this study a prototype valve, attached to a double channel endoscope that monitors and allows feedback control of intra-peritoneal pressure was compared with a pressure recording trans-abdominal needle connected to conventional laparoscopic insufflator. Unacceptably high pressures (>15mmHg) were recorded in various transgastric surgeries, often as high as 30mmHg and caused hemodynamic compromise in some cases. Pressures between the intra-peritoneal needle and prototype valve correlated well. Use of the prototype valve on the double channel endoscope halved peak pressures recorded. Conclusion: Over-inflation is common during transgastric surgery and use of a valve allows ongoing measurement and regulation of intra-peritoneal pressures.
Closure of Gastric Perforation with a Novel Tissue Anchoring Device.
This study assessed the efficacy of a full thickness tissue anchoring device to close large porcine gastric perforations. The device consisted of two needle catheters and a split nylon suture with two distal t-tags and a proximal t-tag. Full thickness puncture with each distal t-tag is made on either side of the defect and the proximal t-tag is pushed down toward the two t-tags to cinch the defect close. This novel device was tested on 6 pigs each with two 2cm gastric perforations. Each perforation required 3-5 tissue anchors, for a total of 48 anchors. 3 of 24 anchors used on the anterior wall penetrated neighboring liver or abdominal wall. Conclusion: The device successfully repaired large defects, though adjacent structures can be penetrated.
Initial Experience with a Novel Endoscopic Device Allowing Intragastric Manipulation and Plication.
This was a report on the use of two novel devices, TransPort & gProx, for intraluminal tissue approximation. They are a multichannel, flexible, steerable device with a tissue approximation device that is passed through one of the channels. Multiple areas in the stomach of a pig were treated with the novel devices which enabled tissue plication via retroflexed and antegrade approach. This technique was easy to learn for two endoscopists new to the procedure. The plication was full thickness or down to the muscularis propria. Conclusion: TransPort & gProx are effective at accessing and plicating tissue in any area of the stomach.
Transgastric Endoscopic Pyloroplasty with Full-Thickness Gastric and Duodenal Myotomy and Suture Closure.
Transgastric endoscopic pyloroplasty was performed in 8 pigs. Using a double channel gastroscope and needle knife, a 1.5-2cm full thickness incision across the pyloric muscle was made. One stitch was placed at the distal duodenal end of the incision and another at the proximal gastric end of the incision, then tied together. With this, the pylorus was opened and the remaining defect was closed with stitches on either side of the initial stitch. 5 pigs survived, 1 suffered pyloric stenosis, and others had an easily traversed pylorus. Average time of the procedure was 30 minutes. Post-mortem exam showed complete healing and no signs of leak or peritonitis.
Transcolonic Hepatic Wedge Resection in a Porcine Model.
This study evaluated the efficacy of a transcolonic approach to hepatic wedge resection in pigs. Four pigs were prepared for the procedure with water, cefazolin and betadine enemas then colonoscoped with Olympus endoscope. Using abdominal wall palpation, the anterior wall of the colon at 20cm was located and then incised with an insulated tip needle knife. After the peritoneal cavity was insufflated, the liver was located. A targeted resection of specimens up to 2.5 cm in size from the left lateral and left medial lobes was taken with a “snare over forceps” technique. Minor resection site bleeding was treated in two animals using cautery and endoscopic clips. The colon was then closed with a prototype device or clips. The investigators obtained specimens in four out of four cases. All four animals survived fourteen days without apparent sequelae.. Necropsy showed no perforations, however abscesses at the biopsy and incision sites were found in one animal. In conclusion, transcolonic hepatic wedge resection is a technically feasible, but not yet sterile technique.
Development and Testing of a New Platform for Retroflexed Flexible Transgastric Surgery: Cholecystectomy, Fundoplication, Gastric Restriction, and Diaphragmatic Repair.
This was a test of a novel device for retroflexed transgastric surgery, a technique employed for difficult to reach areas. Usually the retroflexed scope bows away from the target tissue as any force is applied. In addition, retroflexion is limited by insertion of instruments. This novel device, the TransPort, is a flexible, multi-channel endoscope with Shapelock technology, which allows it to lock the body of the scope while steering the most distal portion. This device was used to perform cholecystectomy, diaphragmatic repair, gastric restriction, and fundoplication in pigs. In conclusion, the TransPort device enhanced retroflexed transgastric surgery in difficult to access areas.
Colonoscopy Potpourri
Is Variable Stiffness Colonoscope (VSC) Better Than Regular Adult Colonoscope (AC) for Colonoscopy? Meta-Analysis of Randomized Controlled Trials (RCT).
In a study compared variable stiffness colonoscope (VSC) with regular adult colonoscope, Othman et al, reviewed the published literature from Pubmed (1966-2005). six RCT met inclusion criteria. The study authors compared cecal intubation rate and time, abdominal pain scores, and use of ancillary maneuvers. VSC did no differ from adult colonoscope in all parameters compared. There was subgroup analysis which showed that small advantage in time to reach the cecum when using pediatric VSC. Further studies are needed to characterize those subgroups that may benefit from VSC.
The Cost Effectiveness of Colonic Stenting (CS) As a Bridge to Curative Surgery in Patients with Acute Left-Sided Malignant Colonic Obstruction: A Canadian Perspective
In the 2nd study, Singh et al, presented his cost effective analysis comparing colonic stenting as a bridge to curative surgery in patients with acute left sided malignant colonic obstruction. He compared this with the Canadian experience of resective surgery or emergent colostomy with later re-anastamosis. Colonic stenting is comparable in cost to surgical options but less likelihood of stomas and may also be associated with lower procedure mortality. These data while seems promising, their emergency colostomy data contained patients who presented with acute diverticulitis which may affect the outcome to some degree.
Response of Crohn’s Strictures of Different Locations to Endoscopic Balloon Dilation
In the 3rd study presented, Mueller and colleagues looked on their cohort of Crohn’s disease that underwent balloon dilations 1999-2005. They evaluated the need of surgery and the long term success rate of dilation. Balloon dilation was successful in 94%. Long term success was observed during a median follow up of 14months (1-65), in 50% of ileal strictures and 80% of ileocecal ones and all colonic and duodenal strictures. One patient had perforation of terminal ileal stenosis. 26% need later surgery due to recurrent obstructive symptoms. This study offers us a safe and effective option for stricture treatment in Crohn’s patients, and can avoid or postpone surgery.
Outcomes of Colonoscopy in The Elderly
An important retrospective study by Kahi and colleague on the endoscopic findings and survival of elderly after colonoscopy. In this study, the medical record of elderly patients underwent colonoscopy at a large VA and county hospital in Indianapolis 1/1999-1/2001 were reviewed. Known IBD and CRC patients were excluded. A total 404 patients over the age of 75 [mean age 79] were included. 15% had advanced neoplasm (9 with CRC and 4 need surgical treatment). There were 167 (41%) deaths within the mean follow up of 4.1±1yrs, while cardiac was the cause of 52 (35%), colon cancer was in 3 patients only (2%). Of note, the high mortality rate overall was perhaps greater than expected from this cohort of patients for whom the calculated life expectancy was @ 9 years. Despite the colonic findings in 15%, the limited survival of this age group raises questions about the utility of repeat screening and surveillance colonoscopy in these patients.
Computer-Assisted Colonoscopy (The NeoGuide System): Results of the First Human Clinical Trial
In the final study by Van Dam and colleagues, a feasibility study of first human trial of using computer assisted colonoscopy (The NeoGuide system). In this prospective, non-randomized, un-blinded study, fives physicians with variable experience participated. 10 consecutive patients in need of diagnostic or screening colonoscopy included. The cecum was reached in all ten patients and terminal ileum in nine. Findings included diverticular disease in two and multiple polyps in another two (removed during the same session). Post procedure assessment at discharge, 48hrs, and 30days revealed no complications. This new technology will provide us with an added armamentarium of methods to aid us to screen for colorectal cancer, and has the potential to make sedation free lower cost colonoscopy more of a reality.
AGA Session: Imaging and Advanced Technologies
Endoscopic Anastomotic Reduction after Roux-en-Y Gastric Bypass Surgery: A Potential Treatment for Weight Regain
Dilation of the gastrojejunal anastomosis is a structural complication of the surgical pouch that may contribute to weight regain after Roux-en-Y Gastric Bypass (RYGB) surgery. Endoscopic anastomotic reduction using a Bard EndoCinch suturing device is a novel procedure that may be a potential treatment for weight regain in patients with outlet dilation. In this analysis of this group’s early experience with this technique in fifteen patients, eleven patients lost weight with an overall mean excess weight loss of 14.2% at 6 month follow-up. There were no serious complications from this procedure. This paper demonstrated that endoscopic anastomotic reduction may be a useful treatment option for weight regain in selected patients after RYGB surgery, however, the durability of weight loss and the need for subsequent anastomotic reductions will need to be examined with prospective randomized studies.
Transcolonic Endoscopic Abdominal Exploration
Natural Orifice Transluminal Endoscopic Surgery (NOTES) using a transcolonic approach may obviate technical limitations associated with a transgastric approach including the need for awkward retroflexion to navigate the upper abdomen. In this study, transcolonic endoscopic abdominal exploration was performed in six pigs through a colonic incision. All of the animals survived fourteen days without apparent sequelae. At necropsy, salpingocolonic and colovesicular adhesions were identified in 5 of 6 animals and histologic examination revealed the presence of microsabscesses at the colonic closure site. In contrast to the transgastric route, a transcolonic approach provides en face orientation to organs in the upper abdomen and allows for better visualization, scope stability, and reach. The findings of adhesions and microabscesses at the colonic incision site may be suggestive of inadequate procedural sterility and refinements in colonic preparation and sterile technique will need to be addressed in future studies.
Natural Orifice Transluminal Endoscopic Surgery (NOTES) Cholecystectomy: A Transcolonic Survival Study in a Porcine Model
Transgastric cholecystectomy has been reported in two non-survival studies which detail substantial technical limitations. The aim of this study was to demonstrate feasibility and evaluate technical limitations of a transcolonic approach to cholecystectomy. Under general anesthesia, 5 adult Yorkshire pigs were prepped with multiple tap water enemas, per-anal instillation of an antibiotic and betadine rinse, and external betadine scrub. A sterile dual-channel endoscope (Olympus™) was introduced through the anus and advanced through a 2 cm, anterior, trans-colonic incision created by a needle knife approximately 15 - 20 cm from the anal verge. In all 5 animals, cystic duct and artery ligation followed by dissection and removal of the gallbladder was successful. The one complication in the series was related to inability to successfully close the colonic incision. This study demonstrates the technical feasibility of transcolonic organ resection via a single incision. For this approach to be translated to humans, a sterile conduit, secure closure device and better instruments for triangulation are necessary.
