Hello, everybody. Welcome back from lunch. Please take your seats when you're ready so we can get started with our next session. We've reached beyond the halfway point in the course today. We're going to – we're into session five, which is stomach and small bowel. I'd like to invite my speakers to come up and join me on stage. We have Dr. Marvin Yu, Dr. Allison Shulman, and Dr. Praveen Chahal. Our first lecture is by Dr. Yu. Dr. Yu received his M.D. from Harvard Medical School in 2003. He's an associate professor of medicine at Harvard, where his focus is advanced endoscopic research, specifically in procedure and device development. Welcome, Marvin. Thank you, Jonathan and the course organizers, for the very kind invitation to speak on gastric varices. These are my disclosures. We will start with some background information on gastric varices, and then we'll spend the bulk of our time talking about treatment options and treatment strategies. As this audience is well aware, gastric varices represent dilated semicostal veins arising usually in the setting of portal hypertension. In cirrhosis, the prevalence of gastric varices is low, 17 to 25 percent reported, in contrast to the higher prevalence of esophageal varices approaching up to 85 percent. However, compared to esophageal varices, gastric variceal bleeds happen less frequently but much more severely. The incidence of gastric variceal bleeding is 16 to 45 percent reported at three years, and the mortality rate is much higher. Overall, there are fewer well-established guidelines for gastric varices. In terms of classification system, most of us are familiar with the traditional serine classification system, which is based on endoscopic appearance. This is divided into GOVs and IGVs. The GOVs represent esophageal varices that extend into the stomach, whereas isolated gastric varices are only in the stomach. Now, this nomenclature is being replaced by simpler nomenclature. GOV1s, which are essentially esophageal varices extending into the lesser curve, are now called lesser curve gastric varices. This represents the bulk of gastric varices, about 80 percent. These can be treated like esophageal varices. If they're small, they can also be banded. GOV2s and IGV1s are now called cardiofundal gastric varices. These represent about 20 percent of all gastric varices. These can be large. They can look like clusters of grapes at the top of the stomach, are usually too large for banding, and are treated with direct endoscopic injection of glue, usually, or endoscopic ultrasound-guided needle injection. Now, distal gastric varices are the rarest and usually occur in the setting of portal vein or splenic vein thrombosis. The other classification system is radiographically based. The Sod-Caldwell classification system is the most common radiographic classification system. Overall, we would recommend a CT or an MRI using portal venous phase of contrast, which will usually reveal a highly variable, highly complex vascular anatomy. The Sod-Caldwell classification system has several types. Type 1 loosely correlates with the lesser curve gastric varices. Type 2 usually loosely correlates with IGV1 gastric varices, and then Type 3 loosely with GOV2 gastric varices. You'll also notice that there is a Type A and a Type B. The B represents the presence of something called a gastrorenal shunt, as denoted here with the red asterisks. This is a shunt that connects the portal system to the cable system, and is actually an anatomic prerequisite for an IR procedure called BRTO, which we'll get into shortly. Overall, these get very complex very quickly, and so multidisciplinary discussion is highly recommended. So, let's talk about treatment options. We'll start with endovascular therapy, TIPS specifically, transjugular intraepatic portosystemic shunt. This is an IR procedure where a stent is placed connecting the hepatic vein to the portal vein for decompression. This is not as effective for a cardiofundal gastric variceal bleeding, as these specific subtype tend to bleed at lower portal pressures, and further pressure reduction with TIPS usually leads to a sizable re-bleed rate. They may be, however, more effective for lesser curve gastric variceal bleeds that are refractory to banding, similar to esophageal varices. Adverse events include encephalopathy and hepatic ischemia, which may actually be higher in the presence of a gastrorenal shunt. So, if you have a gastrorenal shunt and you are committed to an IR procedure, the IR procedure of choice would be something called a BRTO, balloon-assisted retrograde transvenous obliteration. This is where the interventional radiologist utilizes a balloon occlusion catheter, inflates it at the base of the shunt, and sclerosis the entire shunt with the top of the varices included. This procedure leads to a very high rate of cessation of active bleeding, north of 90%, with a re-bleed rate less than 5% to 7% at one year. However, there are adverse events to be aware of. First and foremost, esophageal varices are exacerbated in 35% post-BRTO, and this actually becomes a common cause of upper GI bleed afterwards. Additionally, clinically evident ascites or hydrothorax presents in 35% to 40% of these patients within one year post-BRTO. The historical gold standard for endoscopic therapy has been direct endoscopic injection of glue or cyanoacrylate. Histoacryl or N-butyl-2 cyanoacrylate has been the most commonly used cyanoacrylate. This is a rapid polymerizer. In this country, we have used germabond mostly. This is two-octyl cyanoacrylate, and this is a slower polymerizer of note. Neither types of glue are FDA-approved for this indication, but it's still widely used. The endoscopic technique requires a clear field of view, and intravascular needle placement may be inaccurate, especially if you are working in retroflexion for these cardiofundal varices. Adverse events include embolic complications from the glue, as well as irreversible scope damage and a decent re-bleeding rate. Now, lipidol, which is a radiopaque oil, is sometimes mixed with histoacryl to delay polymerization, but not typically used with two-octyl cyanoacrylate. If you're going to embark on direct endoscopic glue injection, it's recommended to have multiple scleroneedles on hand, as well as multiple small syringes, 2-milliliter syringes. You want to prime your needle with lipidol or distilled water. You don't want to use saline, as saline can cause premature polymerization of the glue inside your catheter. You want to inject in small aliquots of 0.5 to 1.0 mils, and then immediately flush with distilled water. Back bleeding is common, as seen here, and you want to really keep your cool and re-inject as necessary, and then probe with your catheter for local induration to check your progress. Now, EUS guided coil injection therapy is considered the new kid on the block, although we have, at this point, at least 10 years of data supporting it. EUS is attractive for this particular problem because it enhances intravaricyl targeting, even during active hemorrhage. Additionally, it expands the options for your injectate, namely the use of coils, which we will inject in multiple, followed by an adjunct agent. Historically, that's been glue, but other agents like gel foam and thrombin have been utilized, and Doppler with the EUS scope is useful for gauging real-time hemostasis. Coils, if you've not seen coils, they look like this. They're constructed of soft platinum wires with wool-like attachments. They pack within the vessel, and the wool-like attachments activate the clotting cascade. These are borrowed from interventional radiology. They come in varying sizes, shapes, and lengths, and importantly, they are FDA approved for intravascular use. This is a representative video of a patient hemorrhaging from gastroenteritis, where endoscopic therapy is not permissible because of the amount of blood, but under EUS guidance, we can visualize a target very clearly. A needle is used to access the variceal nest, and intravariceal needle placement is confirmed, at which point the coils are extruded, and then Doppler is again used to gauge the progress of the hemostasis. Fluoroscopy can be helpful as a secondary agent or adjunct modality, and here is the reduction of Doppler flow and an injection of a secondary agent. Again that can be glue, but in this case it was gel foam. The evidence at this point supports EUS-based gastrovariceal therapies over direct endoscopic glue injection. It specifically supports the notion of improved control of active bleeding, durability of hemostasis, and lower complication rates. At this point, there are at least two meta-analyses that support any EUS therapy being superior to direct endoscopic injection, and there is one meta-analysis that supports EUS combination therapy with coil and glue specifically as the preferred strategy. The literature at this point also includes a recent U.S. multicenter study for EUS coil therapy that we were privileged to be a part of. This was the first U.S. multicenter study of its kind. 106 patients across 10 U.S. tertiary care centers. Many of these patients were very sick, actively bleeding, requiring ICU hospitalization and blood transfusions. The majority had cirrhosis and IgV1 varices. You can see that the technical and clinical success were 100% and 89% respectively. The adverse event rate was 6.5%, which did include a case of a fatal pulmonary embolism, speaking to the tenuous nature of these patients and also the persistent risk of embolic complications despite the use of coils. Recurrent bleeding occurred in 14% of patients at a mean of 32 days. However, the majority of these patients were successfully retreated with repeat EUS coil therapy. Importantly, there were no significant differences between high-volume and low-volume centers, which I think speak to the potential for scalability at centers with the requisite interventional EUS expertise. So we'll pivot and talk about treatment strategies. We'll start with primary prophylaxis or prevention of initial bleed. There's limited data for primary prophylaxis. Beta blockers is a weak recommendation. There is one RCT that shows endotherapy may be beneficial as primary prophylaxis. So at this point, the most recent ESGE guideline supports endotherapy as a consideration for patients intolerant of beta blockers. And of note, neither TIPS or BRTO are recommended for primary prophylaxis. For acute gastric variceal bleed, the recommendations are very similar to what you will see for acute esophageal variceal bleeding. So specifically, a restricted hemoglobin goal of 7 to 9 grams per deciliter, ceftrioxone, octreotide, and endoscopy within 12 hours to diagnose and potentially treat. If you see something on the lesser curve and it's small enough, banding is the recommended treatment. If you see something on the cardiofundal side, this will require direct endoscopic glue injection or EUS-guided therapy where available. And if the patient is unstable or the endoscopy is unsuccessful, these patients will then go on to interventional radiology. For secondary prophylaxis, data is also limited. There is one randomized controlled trial showing the superiority of glue over beta blockers and another randomized controlled trial that shows the superiority of glue over TIPS. I'm sorry, TIPS over glue, but nothing for EUS-guided therapy. So at this point, the ESGE 2022 guidelines recommends an individualized approach based on patient factors and local expertise. So a suggested algorithm at this point for bleeding gastric varices would be the following. For anything on the lesser curve, again, treat these as esophageal varices if they're small enough and you can do banding or direct endoscopic glue injection if they're large. If they're cardiofundal, then these would be treated with direct endoscopic glue injection or EUS-guided therapy where available. Assuming you can get control of the bleeding, then secondary prophylaxis would be more of the same successful strategy. And if you can't get control of the bleeding, these patients would go on for TIPS or BRTO. So to conclude, evolving guidelines for gastric variceal management. Importantly, treatment decisions should continue to be made in a multidisciplinary setting. Endoscopic treatment entails banding and direct endoscopic glue injection, but there is growing evidence for inclusion of EUS-guided therapy at centers of expertise. Thank you.

gastric varices

portal hypertension

TIPS

endoscopic therapy

multidisciplinary approach