Now, I look forward to kind of introducing the second speaker, who is Dr. Uzma Siddiqui, who's Professor of Medicine at the University of Chicago, Associate Director of the Center for Endoscopic Research and Therapeutics, and Director of Advanced Endoscopic Training. If we can get Uzma's talk there, she will present the Best Upper and Small Bowel Abstracts. Good afternoon, everyone. I want to thank the ASGE and Dr. Vanessa Chammy for the invitation to speak today. I have the pleasure of presenting some interesting DDW abstracts related to the upper GI tract and small bowel. Here are my disclosures. Now, I divided the abstracts that I'm going to present into four major categories, as listed here. And we'll start with lesion detection. Artificial intelligence, or AI, was a hot topic at this year's DDW. And while we know its use in the colon, its applications are making its way into the upper GI tract. Detection of early gastric cancers remains a clinical challenge. And this Japanese study aimed to utilize AI to help with not only identifying gastric lesions, but also classifying them into those that are cancer versus those that are not cancer. They took thousands of still images from endoscopies to create an AI system that was used to then make a training set to confirm accuracy in lesion detection and classification. And this AI system was used real time and compared against endoscopists, both trainees and experts, and their findings. In the validation set, the AI system was pretty good at detecting cancerous lesions with a sensitivity of 89% and an accuracy of 84%. But it did not do as well in detection and classification of non-cancerous lesions. When used real time and compared to endoscopists, the AI system had a higher accuracy and specificity compared to the endoscopist findings. And it was also quicker at detecting lesions. So AI in detecting gastric cancer is still a work in progress. But as the authors suggest in their title of this abstract, this would simply be a technology used to support the endoscopist's own detailed exam. Our next two abstracts will look at ways to make small bowel enteroscopy less cumbersome. As we know, single and double balloon enteroscopy can be technically challenging and require special equipment, endoscopist expertise, and a lot of time. So this study looked at using a novel rigid overtube that's flexible initially, but then when needed, can be stiffened to allow further scope passage, and most importantly, can be used with a standard pediatric colonoscope, therefore making it accessible to the average endoscopist in almost any unit when needed. This is initial data on 13 patients over 11 months that showed this technique was effective in both antegrade and retrograde directions and was reported as easy to use. This next abstract also looks at making small bowel enteroscopy easier and also quicker using a motorized spiral enteroscopy device. There have been prior studies on a manual spiral device, which had long procedure times and reports of mucosal injury and trauma. So this newer motorized spiral enteroscopy device has been around for about five years with limited published data. So this abstract looked to validate the device's efficacy in not only detecting small bowel lesions, but also in being able to carry out interventions. The device has a short spiral overtube, and movement back and forth is controlled by tapping a foot pedal. So in theory, making for a shorter procedure with less ergonomic stress on the endoscopist. They used the device in 21 patients with 100% technical success rate and relatively short procedure times with a median time of 62 minutes for antegrade enteroscopy and only 40 minutes for retrograde use. They detected lesions in 67% and made an intervention in 86%, which compares favorably to other small bowel techniques. But most importantly, there were only two serious adverse events. One aspiration despite the patient being intubated, and one jejunal perforation in a patient with lupus. Moving on to lesion resection, this study aimed to look at long-term survival in patients with a variety of early gastric cancer pathologies. They presented the results of a multi-center web-based registry including expert Japanese endoscopist cases at multiple centers. Previously, this group had published good short-term results. They were able to identify over 10,000 resected early gastric cancer lesions, which has to be one of the largest sample sizes on this topic. And then they looked at five-year survival rates and also looked at survival and outcomes and various early cancer pathologies as shown in the bottom table. This included expanded indications for endoscopic resection, such as those tumors that extend slightly into the submucosa known as SM1 tumors. Using this huge data set, they found that there were good long-term outcomes with five-year survival rates of 89%. And there was no significant difference in outcomes between any of the curability pathology categories mentioned earlier. This indicated that endoscopic resection can provide a successful long-term outcome for early gastric cancer patients. And it can also include those tumors that invade slightly into the submucosa. This next abstract examined long-term outcomes of endoscopic resection of duodenal neuroendocrine tumors. Clinically, we're taught to remove all duodenal neuroendocrine tumors because of the potential risk for lymph node metastases. And then those of us that endoscopically resect these types of lesions often worry when a resection margin is positive. So this study looked at 34 non-ampullary neuroendocrine tumors that were endoscopically resected, and then examined long-term outcomes and compared them to nine surgically resected lesions to see if there were any risk factors that could be identified for predicting lymph node metastases. All but one lesion was endoscopically resected with EMR. And 14 out of 34 had positive resection margins. Overall, five-year survival was excellent amongst the endoscopically resected group. And there was no recurrence or metastases found regardless of whether the endoscopic resection margin was positive or negative, which is quite reassuring in this small sample size. Now, a couple of risk factors for presence of lymph node metastases included lesions greater than 10 millimeters and those that invaded the muscularis propria. Moving on to upper GI bleeding, this abstract from Hong Kong was a nicely designed study where they randomized 100 patients with upper GI bleeding from peptic ulcers to either standard hemostasis therapy first or over-the-scope CLIP or OTSC therapy first. They found that most of the bleeding peptic ulcers in this study were larger sized at more than 1.5 centimeters and that there was no difference in overall clinical bleeding rates between the two groups. However, they did find that in the group that had OTSC therapy first, the patients that failed this initial therapy all had re-bleeding when crossed over to standard therapy, versus the group that had standard therapy first had no re-bleeding once an OTSC was applied in a crossover arm. So the authors concluded that for these larger bleeding ulcers, it may be better to stick with standard hemostasis therapy and then use OTSC only if the initial therapy fails. And they presume that this may be due to the fact that the OTSC devices may be a little bit more challenging to apply. Now, hemospray has also been garnering more attention lately, especially since it got FDA approval in the US. But this multicenter international study looked at adding hemospray to standard therapy in GI bleeds. They prospectively enrolled 230 patients from 18 international centers and looked at immediate bleeding cessation and re-bleeding rates. Most bleeds in this study were due to peptic ulcer disease in a high-risk cohort. And in this particular group, hemospray was most commonly used with adrenaline and had a 93% hemostasis rate. It still had a 19% re-bleeding rate, though. Now, if you look specifically at malignancy-related bleeds, you had similarly high rates of hemostasis at 92%, but only one case of re-bleeding. So the argument the authors made was that in this subcategory of malignancy-related bleeding, perhaps hemospray could be used as monotherapy. And my final category involves improving gastric outflow. There were a number of abstracts at DDW looking at the tunneling technique of G-POM to perform myotomy at the pylorus. This abstract from Brigham and Women's included a meta-analysis with 20 studies that had various etiologies for delayed gastric emptying. They demonstrated that G-POM does have benefit by improving gastric emptying on scintigraphy and also improving patient symptom scores. One additional G-POM abstract looked at a subgroup of post-lung transplant patients who, interestingly, have a high rate of developing gastroparesis as a side effect after surgery. But even in this subgroup, G-POM did have significant benefit. And finally, this is another randomized controlled trial out of Hong Kong that compared traditional, uncovered duodenal stents that are prone to tumor ingrowth over time to new partially covered duodenal stents. They enrolled 117 patients from 10 centers and demonstrated that there was a trend toward lower rate of tumor ingrowth at one year in the partially covered duodenal stent group. However, this did not confer any clinical benefit in outcomes. Obviously, we're gonna need larger scale studies to know if this new type of duodenal stent will have any additional clinical benefit. So in conclusion, there were a number of great upper GI and small bowel abstracts presented at this year's DDW. They discussed promising new technologies, devices, and techniques. And I'm sure we're gonna be hearing a lot more about all of these in the future, especially once we get some more data on each of these topics. Thank you.

artificial intelligence

lesion detection

small bowel enteroscopy

gastric cancers

upper GI bleeding