Well, thank you and to the course organizers for having me speak. I think perhaps the topic might be better coined as West meets East, because we've really learned a lot from the East. Those are my disclosures. Now for gastric cancer, we know that the incidence of gastric cancer has been slowly decreasing, but it's still a significant problem. The SEER data projects that for this year in the United States, there'll be almost 30,000 new cases of gastric cancer, 11,000 deaths, and the five-year survival rate has really stalled at around 30 percent. Worldwide it's still a significant problem, fifth in incidence, but still the third commonest cause of cancer death worldwide. There are some very significant racial differences. We know that in Asians, Japanese, Korean, and Chinese natives that there's a high risk of gastric cancer. In native North Americans or non-Hispanic whites, it's very low, and in Japanese and Korean Americans who have emigrated, it's moderate. Hispanics also have a high risk of cancer, and they were not included in this study. Now cancer develops through a paradigm of dysplasia that starts out with chronic gastritis. This can be induced by H. pylori or perhaps autoimmune. It progresses to atrophy, intestinal metaplasia, and then that's the precursor lesion for intestinal-type adenocarcinoma. A Swedish study did suggest that as many as one in 40 patients with intestinal metaplasia or one in 20 who have dysplastic IM will develop cancer within the 20 years of the index endoscopy. It's important that if we can find these early lesions and treat them endoscopically, we hopefully can prevent cancer. And that requires a very close and detailed examination of the mucosa, and I think you're going to see that as a paradigm for a lot of what we're doing in terms of detecting early lesions. We have to learn to look carefully, and our Asian colleagues have taught us this. You can see on your left is a normal cardiac and body of the stomach, and on the right, clearly quite marked atrophy with loss of rugal folds and very prominent vascularity, all indicative of atrophic gastritis. To find the intestinal metaplasia, we're going to look for subtle abnormalities within that field, irregularities, nodularity. If we get close enough, we may see some villiform pattern. And this is an area where digital chromatoscopy, I think, has become critical, and using a blue light chromoendoscopy, whether that's NBI, BLI, or one of the eye scan modes, depending on the equipment you have, we're again looking for slightly raised pale areas compared to the background and changes in the pit pattern. This is just an example. This is a patient I saw last week who came for surveillance, and you can see that in the antrum, this is pretty obvious, but the mucosa's not normal. It's irregular and nodular. The narrow band imaging really highlights the areas of intestinal metaplasia, and we get close with near focus, you can see the villiform pit pattern. And with NBI, this blue light crest sign, all of which are indications of intestinal metaplasia. Now, unfortunately, not all patients are that obvious, and in high-risk patients, there are biopsy protocols that have been suggested to help us find IM in high-risk individuals, and this is the Sydney system, where it suggests we should be taking biopsies from the greater and lesser curve in the antrum, in the corpus, and separately at the incisura, and that is a protocol in high-risk patients will help us find patients with intestinal metaplasia. Now, which one of these is best? And this is a paper published by Lorne Lane a few years ago, and they looked at a high-quality white light examination, a second examination under narrow band imaging by an endoscopist blinded to the findings of the first, and then mapping biopsies were done. And you can see, in terms of the yield, that for single modalities, the mapping biopsies were very good and detected 76 percent of patients, NBI was slightly less, and white light alone was not as good. When you combine the modalities, though, a combination of a detailed NBI exam with mapping biopsies, we were able to detect all the patients with intestinal metaplasia, and I think, moving forward, we have to learn that I think this is the paradigm we should be using in examining high-risk patients. We then want to look at those areas to find dysplasia, because these are the patients that we're going to want to be treating, and there's not a lot of time to go into that today, but you're, again, it's training your eye now not just to see the normal mucosa but to see irregularities in either the vascular or the surface pattern that will sort of clue us into that these may be areas where there's dysplasia or early cancer. And this can be very subtle. This is a case from our unit a little while ago. You can see there was obviously a nodule. This was biopsied. It had high-grade dysplasia, possible intramucosal carcinoma, and it was sent to us for removal. The question is, where is the full lesion? It's not just that nodule. If you look very closely with NBI, you'll see that there's a very subtle demarcation from normal to abnormal right here, and, in fact, that's the entire lesion. So we have to learn from our colleagues in the East how to look at the mucosa and find these lesions. Now, should we be screening? Clearly, there are established programs in Asia where the incidence is high, and these are useful. In North America, there are no recommendations for screening patients, but recent cost-effective studies suggest that in high-risk subpopulations, it's probably cost-effective, while it is not in groups that have a low incidence such as non-Hispanic whites. Next, we've identified patients with IM. Should we be surveilling them? Our current ASG guidelines do not recommend surveillance of patients with nondysplastic IM unless other risk factors are present, but the Europeans have just published new guidelines in February of this year, and they do recommend that in patients with extensive IM, that a repeat endoscopy should be done at about a three-year interval or sooner if there's a family history. Patients with low-grade dysplasia and no visible lesion should have surveillance in a year, and if there's a visible dysplastic lesion, whether that's low or high-grade, it's recommended for resection. So which lesions can we resect endoscopically? I think the rationale is that endoscopic treatment is appropriate. When the mucosal lesion can be fully resected, we want to do an R0 resection with minimal risk of lymph node involvement. So we have to sort that out, and what data do we have? This is the SEER data, and it's data on over 1,500 patients with early gastric cancer who went for gastrectomy and had lymph node dissection, and you can see with small lesions, two sonometers or less, the risk of lymph node involvement was less than 2%. As the lesions got larger or there was submucosal involvement, that lymph node risk increased. Now unfortunately what this data doesn't give us are other risk factors such as the presence of ulceration, lymphovascular invasion, and that's where our Japanese colleagues have provided us with a lot more data. This is Gotoda's study from over a decade ago. This is over 5,000 patients with early gastric cancer undergoing gastrectomy and lymph node dissection, and you can see that for well-differentiated intestinal types, in the absence of mucosal ulceration, lymphovascular invasion, or submucosal invasion, that those patients, even with large lesions, the rate of lymph node involvement is less than 1%. For ulcerated lesions that are small, less than three centimeters, and no other high risk features, it remains in that range. Once we get into larger ulcerated lesions, the rate of lymph node involvement increases and gets to between 3% and 4%, and even for lesions that have very early submucosal involvement or less than 500 microns, if they're small lesions with no other risk factors, the lymph node rate is in the range of 2% to 3%, as well as for poorly differentiated or diffused types that are small and intramucosal, the lymph node rate is in that 2% to 3% range, and if we accept that the operative mortality in many of these patients is also in that same range, it certainly raises the likelihood that endoscopic treatment may be appropriate, and based upon these data are the indications for endoscopic resection through the Japanese society and the European guidelines mirror these, so basically all lesions with high-grade dysplasia regardless of size, what has been referred to as the absolute indication was a lesion less than two centimeters, well differentiated with no other risk factors, and the two centimeter size at the time was really chosen because that was the limit of what we could remove on block with EMR. We know from the Japanese data that I just presented to you that the lymph node rate is less with even larger lesions, so the expanded indications now include any size that's differentiated with no ulceration or submucosal invasion, less than three centimeter ulcerated lesions that are differentiated, if there's early submucosal invasion in the absence of other risk factors, that can be considered for endoscopic removal, and the small and differentiated intramucosal lesion. What should we do prior to resection? Well, the detailed mucosal examination is really the most important, and we should try to limit biopsies for suspicious lesions because we don't want to induce too much submucosal fibrosis. Endoscopic ultrasound is not routinely used or recommended as it tends to be inaccurate for these superficial lesions, can overstage them, but we do consider using it for atypical lesions when we are concerned there may be more advanced disease underneath or to confirm that, and CT or other cross-sectional imaging is not routinely required but can be considered for patients with lesions where the indications are borderline. How to resect them? We have EMR and ESD. The EMR can be done either by cap or band technique, and lesions two centimeters and less can usually be removed on block, but larger lesions require piecemeal resection. I guess the advantage is that there are many physicians who are very familiar with esophageal EMR, and this can be used in the stomach. The downside is the pathologic assessment, especially for piecemeal resections, becomes limiting. Endoscopic submucosal dissection allows for on-block resections and can be used for lesions almost of any size. It produces a specimen which gives a very detailed pathologic assessment, but it's technically more difficult and does require advanced training, and although things are improving, it still has somewhat limited access in Western countries. Comparing the two, endoscopic submucosal dissection clearly allows for a better on-block and R0 resection with a reduced local recurrence rate at the exposure of a higher procedural time compared to EMR. There's a slightly increased risk of perforation, but the bleeding risk is about the same. Overall, we think ESD is superior if it's available. The outcomes of ESD have been published by Dr. Yahagi. This was last year. This is over 5,000 patients who've had ESD for gastric lesions. Under the absolute indications, remember this is a Tucson lesion or less, then the results are excellent with almost no recurrence or metastatic recurrence. As we expand the indications, not surprisingly, there's a slight reduction in the curative resection rate, increase in local recurrence, and a small number of metastatic recurrences, but overall the results are very good. As we are not resecting the stomach, we're leaving a lot of mucosa at risk in place, so regardless of how we remove these endoscopically, these patients are at risk for metachronous lesions and they require ongoing surveillance. Gastroscopy at three to six months after resection and at least annually until it's appropriate to stop. We can consider a follow-up CAT scan in patients who have expanded indications where their lymph node metastasis rate may be borderline or high. So in closing then, I think intestinal metaplegia is the precursor lesion for intestinal type gastric cancer. High-risk populations should be considered for screening. That detecting intestinal metaplegia requires very careful endoscopic examination. We have to train our eyes, use digital chromoendoscopy with blue light and mapping biopsies. That patients with high-grade dysplasia and early cancer should be treated endoscopically and likely with ESD as the preferred method. Thank you. Thank you.

gastric cancer

Asian populations

endoscopic examination

precancerous lesions

surveillance guidelines