MICHAEL LEVY: Hello, I'm Mike Levy, a gastroenterologist at Mayo Clinic specializing in pancreaticobiliary disease. My areas of clinical expertise include pancreatic disorders, endoscopic ultrasound, endoscopic retrograde cholangiopancreatography, and tumor ablation. Today, I'll be sharing the latest information on a EUS-guided angiotherapy. When we manage gastrointestinal bleeding, there are several options. Most undergo standard endoscopic intervention. When there's rebleeding or failed detection of the lesion, we sometimes have to pursue interventional radiological or surgical alternatives.

There is now a new alternative, which is EUS-guided angiotherapy. This list includes the type of patients that we've treated all at Mayo Clinic, and include varices from a number of different sites, esophageal and other luminal cancers, dieulafoy lesions, stromal tumors, luminal anastomotic bleeding, ulcers of various sites, of Brunner's gland hematoma, duodenal metastases, pseudoaneurysms of various sites, and a number of vascular abnormalities, as well as a prostate cancer that eroded into the rectal wall.

I should note that EUS was almost never the initial modality in these patients. We pursued it after standard techniques failed or there was significant rebleeding. What I'd like to do is take you through five cases, five patients and their care, and I think it highlights the potential role, the utility, as well as limitations of EUS in this setting.

The first was a 70-year-old gentleman with diabetes, hypertension, obesity, hyperlipidemia, hypothyroidism, GERD, and Barrett's esophagus. He had two prior bleeds. One was clinically significant, and had received seven units of packed red blood cells. His prior evaluation included a negative exam, two EGDs, one colonoscopy, and a CT of the abdomen. He was undergoing his third EGD, which was negative except for the finding of Barrett's as expected, and was undergoing fourth-quadrant biopsy.

There was an unexpected clinically significant bleed. The endoscopist was not able to perform epinephrine injection or banding because though vascular abnormality was identified, he attempted Gold probe and sent the patient to the intensive care unit. He performed-- or asked me to perform an EUS, which demonstrates blood flow in the left ventricle, azygous, and aorta as expected, but unexpected were very large columns of varices and multiple varices. Grayscale, color, power, and pulse Doppler all show the vascular flow, and then coils were placed into the varix.

This is an anechoic, or black-appearing vessel. That's how they typically appear. This is the needle, and then through the needle a coil is advanced. And you look at this and might think, is this coil too small, the correct size, or too large? I'm not going to show an ideal example. I'm actually showing a less than ideal, which is the insertion of too small of a coil. We know it's too small because once it's injected, it appears and has the same shape as it does outside of the patient. Ideally, that should not be the case. If it's too small, there's a risk of embolization.

Happy to say this did not occur in this patient. There would be a risk in placing too small of a coil. And in fact, what you want to do is place a coil that's about 1.2 to 1.6 times the diameter of the vessel. Therefore, it's a little difficult to insert and assumes a geographic or three-dimensional shape, and doing so, it's more likely to stay in place and not embolize.

We're really concerned about embolization. We can place a little bit of the coil distal to the wall, as well as most within the vessel, and some in the proximity and intervening tissues. This shows a nice treatment response pre- and post-therapy, very active flow, and absence of flow following therapy. This gentleman did well. He was discharged two days later following EUS, and has had no recurrence more than six months following intervention.

Now, I'm going to show the care of a second patient, a 26-year-old gentleman with a central thrombocytosis. He has thrombosis of his portal vein, superior mesenteric vein, and spenic vein, and varices, both esophageal and biliary, resulting biliary stricture. He had numerous prior bleeds and numerous transfusions of packed red blood cells. ERCP was performed on several occasions with covered metal stent placement hoping to tamponade the blood vessels. Unfortunately, that did not work well. You can see here, this is not only air, but esophageal blood clots within.

Shunt surgery was considered, but considered technically not feasible in this gentleman. He was a candidate for transplant, but not at that time. They were looking for a temporizing measure, and therefore asked me to perform EUS. The EUS, you can see the hyperechoic or bright white struts of the metal stent. And it's surrounded by a feeding and then numerous pericholedochal varices. There were so many varices I did not want to treat all of them. I thought that might even make the stricture even worse.

So I performed an intraductal ultrasound, and here's the small probe in the bile duct, to try and identify the vessels that were most likely to be resulting in the bleeding. And on magnified view, you can see the entire lumen is obliterated by a very large varix. Using both EUS and fluoroscopy to guide treatment, targeted that vessel, and it was successful. Fortunately, transplant could be delayed 15 months, at which time he was an appropriate candidate.

Showing the care of a third patient, a 70-year-old female with PSE and cirrhosis. She had had several clinically significant bleeds, had been managed with somatostatin, non-selective beta blockers, and topical silver nitrate. EUS was performed by placing the probe actually on the skin of the patient. And here you can see peristomal varices. Coils were then inserted with the probe, not only on the skin, but slightly in as well, and placement of the coils. And fortunately, in the follow-up period of eight months, she developed no rebleeds.

There insufficient data, I would say, for EUS-guided angiotherapy. Not enough centers have done a large enough volume to put together meaningful data for most indications and applications, but there are data for gastric varices. The technical success of using EUS-guided angiotherapy is approximately 80% to 100%, based on the literature. The treatment success after one treatment session is approximately 50% to 80%, and it's approximately 90% to 95% for all treatment sessions. The rebleeding rate is a little difficult to discern, because studies have somewhat limited follow-up period, and it's often been used for primary prophylaxis.

There are a number of adverse events that have been reported. The one we worry most about is glue embolization. This was a small study, but they evaluated 19 patients who underwent glue injection for gastric varices, obtained a chest CT in all 19. And although all patients were asymptomatic, nine developed occult embolization, and occasionally there is clinical sequelae. So this is something we always keep in mind when considering this therapy.

There remain some uncertainties. While it is at least as good as standard endoscopy, how much better still remains to be determined. Indications are somewhat uncertain in some centers, in terms of patient selection, lesion selection, and timing. The technique has not been standardized in terms of the use of coil and/or glue, and the number of coils and volume of glue. The costs must be considered for both EUS and fluoroscopy. And again, while the efficacy and safety appear to be at least as good as standard endoscopy, more data are needed to determine how much better EUS-guided approaches are.

I'm going to take you through the care of two final patients to show some unique uses. 37-year-old gentleman who is an alcoholic, had recurrent acute and chronic pancreatitis diagnosed in 2012. Had multiple episodes of hemosuccus pancreaticus. There were attempts to control with interventional radiology, at least four attempts at a referring hospital, one at our institution. CT shows a densely calcified pancreas here, and interventional radiology revealed the bleeding lesion. They placed coils, thinking this would result in a cessation of bleeding. Unfortunately, this was not the case.

In follow-up imaging, both CT and MR revealed the enhancing of blood flow within the pseudoaneurysm, and these were the corals that had been placed. EUS was then performed. Again, here is the pseudoaneurysm with vigorous blood flow, shown on Power and pulse Doppler. And under EUS-guidance a needle was advanced into the mass and a coil placed. You can see that soon thereafter the pseudoaneurysm was full of clot, and this was successful therapy as demonstrated on follow-up interventional radiology, and the absence of bleeding or rebleeding over a 15-month follow-up.

And finally, there was a 60-year-old gentleman with alcohol-induced cirrhosis. He had more than five episodes of hemosuccus pancreaticus, had received more than 18 units of packed red blood cells. And again, he had had four or more somewhere at referring hospitals, attempt at interventional radiology control. CT shows the pseudoaneurysm here, as well as seen on EUS. And under EUS-guidance, multiple coils were placed in the pseudoaneurysm. There's now a clot formation and cessation of bleeding.

So the presented cases demonstrate the utility of EUS-guided therapy in the appropriate clinical setting. I would say, for patients who are evaluated at the Mayo Clinic, we approach them in a multidisciplinary manner, and consider whether standard endoscopy, EUS, or surgery is the most appropriate, and I think this optimizes patient care and outcome. Thank you.