Our next speaker is Dr. Andy Thao. Dr. Thao completed his MGI fellowship and chief residency, a chief fellowship at Baylor College of Medicine. He is currently a partner at Austin Gastroenterology, a GI hospitalist, and the Ascension Section Network Chief of Gastroenterology in Austin, Texas. He'll be talking to us today about esophageal leaks and fistula, choosing your intervention, stent versus not. Thank you. It's a great honor for me to talk to you today about this topic, which is near and dear to my heart. Here are my disclosures. This is my agenda. I want to discuss to you the principles of defect closure. These are the four D's. Then I want to take you through how to choose the modality. These are the three D's. And then I want to take you through some cases where we apply the D's, okay? Principles of defect closure. There are four D's. You have to secure all four of them if you want to guarantee success. Drainage, diversion, diet, and then you must address distal obstruction. Drainage can come in a variety of ways. Percutaneous is most common. That's a chest tube or mediastinal drain. Next come internal drainage. This can be double pigtail or endoscopic vacuum therapy, which is in some ways internal-external if you think about it. Diversion comes in the form of a stent. It also can come in the form of EVT. They make a U-turn. They swallow the secretions. They get sucked back out, as well as placing the patient NPO. Diet also has to be secured. So important to eye often on the index endoscopy, place a PEG-J tube, a direct PEG, or acutely if they're very ill, TPN. You also must address distal obstruction because oftentimes hydrodynamically that is the cause of the development of the fistula. There's a distal obstruction. So you can address that with a stent or a dilation. Notice that stent is listed there twice. EVT is also listed there twice. So some of the most common modalities address multiple Ds. If all four of these are addressed, success is maximized. You know what you notice there? There's no C for closure. There's Ds, but no C, no closure. Why is that? Ladies and gentlemen, if you secure the four Ds, you don't need to close. They will heal on their own. Now obviously if you close, you secure diversion and you can reintroduce diet and so on. But you must secure these if you have any shot at closing a chronic fistula in particular. So notice the fully covered stent, it gets diversion and it can resolve distal obstruction. Two birds with one stone, diet kind of, I don't feed them when they're on the stent, but you could put a Dobhoff tube down for example. Like vacuum therapy, you both get drainage and you get diversion. Now you say, well, what do you mean diversion? Well, you just, the secretions make a U-turn, they just get suctioned out instead of getting bypassed like the stent. So I normally tell you to say this out loud, but there's too many people here. Say this, I see pus, I drain pus. Do not say this, I see hole, I close hole. If you see hole and you close hole and it's soiled, I will come to your house and punch a hole in the drywall, take a dump behind the wall and then patch it up. And we'll see how that goes. It's not going to go very well. So don't do that. And I know a lot of people in this room are advanced endoscopists. You want to manage everything on your own, right? But resist, this must be a multidisciplinary problem. And here's the reason, mediastinitis is the pancreatitis of the chest. If you think pancreatitis is bad at creating a raging systemic inflammatory response, you haven't seen nothing yet. These people will die within 48 hours if you don't get the mediastinum source under control. So you have to drain the mediastinum, you must get a chest tube if the pleura has been compromised. The drainage of the pleural cavity, and I learned this the hard way, is not always equivalent to the drainage of the mediastinal cavity. Some patients only have mediastinal abscess, they haven't broken into the pleural space. Some people have broken through both. You drain the pleura with the chest tube, there's still a mediastinal abscess. So you may need both. Utilize CT scans very generously in the management of these patients to ensure no undrained connections or collections are left behind. I sometimes just use non-contrast low-resolution studies. Now let's choose a modality. These are also three Ds. You must think duration, defect size, and drainability. When you choose a modality, you want to think is it acute or chronic, that's duration. Then you want to think the defect size. The defect size is the mouth of the opening, not the actual size of the cavity, it's the blue one. And then you want to think drainability. And I really mean percutaneous drainability. If percutaneous drainage is not possible, like there's no window, now you may say there's always a window. No, there's these things called ribs in the way, they're difficult to drain. So if there's no drainage possible, then internal drainage with double pigtail or EVT are your only options. You can exit the exercise, those are your only two options. If drainage is not possible and there's a soiled cavity, do not stent it. If you stent it, I will poop in your wall. Do not stent it. And do not try to primarily close it. It will fail, pus will find a way out, guys. This is a picture, this is just to show you what I mean by cavity. This is a limestone cave system, it actually surprisingly looks a lot like the mediastinum. That's what a cavity is, that's what a defect is. The karst spring is drainage. So for an acute leak or perforation, these are detected within days, two or three days, postoperatively or iatrogenically, you accidentally made them, and you can close them and they respond favorably with primary closure methods. In discopoly, that means clips of all varieties and sutures of all varieties. In general, they're easier to close technically, but they may reopen if you don't address the four Ds. For example, if there's a distal obstruction and you close it, it's going to open up again. For chronic leaks or what I call fistulas, primary endoscopic closure is less successful for chronic leaks because they have this epithelialized fibrotic tissue, it's like cheese. And you guys have tried this before, suturing it and then you tighten it and it just tears. It's really not favorable. So the paradigm for chronic leaks and fistulas focuses on optimizing pressure gradients to allow for internal drainage with closure of the cavity by what's called secondary intention through granulation tissue. You're allowing to heal from the base up. And the three main endoscopic offerings there are EVT, endoscopic vacuum therapy, the internal drainage of the double pigtail stents, and then fully covered self-expanding metal stents, which mostly do diversion and form a scaffolding for the granulation tissue to attach to. Just to recall, primary intention, you close the hole directly. Those are for acute small defects. Secondary intention, you close indirectly by allowing granulation tissue to come up from the base. Okay. So ladies and gentlemen, this is the most important slide. Okay. These are the four quadrants of options. Acute, you have many options. This is primary closure. So if this injury just happened within a few days, you can close this so long as there's no soiled cavity. If there's a soiled cavity, you've got to get a drain in there. So these are the options. If it's large, then you need to use suturing systems. Okay. If it's small, in the bottom left quadrant, well, you can use some of these tissue appositions through the scope clips or the over the scope clips. Okay. If it's chronic, endoscopic vacuum therapy, if it's a very large defect. Okay. And if it's a very small defect and it's hard to get a sponge in there, then internal drainage with double pigtail stent. Now, a fully covered stent is also in the option set, but remember, don't stent a dirty soiled cavity. Okay. So if you can, so you can see there, primary closure on the left, secondary closure on the right. Okay. So let's apply this. Let's do the exercise. So I'm going to give you four most common scenarios I'm going to go through. The first one is acute perforation, usually micro perforation, mediastinal air, but no soilage. This often happens, you know, midnight misadventure, food bolus. You know, if you see the food bolus, you're like, I just push it in there and push it in there. And oh my God, look at the rent I created. And the patient has pneumomediastinum, but there's no drainage on the CTS soft grip. Okay. With that one, you can do a primary closure, right? You can use clips, right? Here's a traditional clip. Sometimes these defects are hard to close. You know, you look at scratch and it didn't open. Okay. So you can use like the, I hate to say it, the mantis clip. I'm not supposed to say that, but you can use these tissue apposition clips. Okay. With the mantis clip, I like to grab it, pull it back into the scope a little bit, do a little bit of a turn to set the hook. And then what, if you have any fishermen out here, you know what I'm talking about. And then you can reopen it and it won't let the original hook go. You can go to the other side and then close it as I do here. And as you guys know, many of you guys do palm. Obviously the first clip is so critical, isn't it? You want to get that puckered up space so that you can grab further tissue. Okay. Another option of course is just to use over the scope, I'm sorry, the Apollo overstitch system or the Boston system, right? This can get full thickness bites. This I remember, this was a misadventure from a TEE probe. You ever get that consult? They pass the TEE probe and they have difficulty going and they keep pushing and then they make a huge rent. This was that one. And we closed it without any issues. So that's primary closure. Here are through the scope tissue apposition devices. I made this for you because these are new. Okay. These are relatively new in the last few years. And the highlights are that the mantis clip is rotatable, maneuverable, good in tight spaces, but it's tissue. Those little hooks, the mantis hooks, they sometimes tear and they sometimes only grab mucosa, but it can fit through a diagnostic gastroscope. The DAT clip from Microtech, it's not rotatable. It's pretty stiff. It tends to be on fonts. It's not good in tight spaces, but holy cow, it can really, the two opposing little arms, they can really grab a lot of tissue and pull over. And so that's, that's its benefit. And then the X-TAC system gets modest thickness bites, but it's really good in tight spaces. Okay. And it has unmatched defect size closure because you can get really wide spaces between the tacks. Okay. Second scenario, acute perforation, small defect size, but the cavity is soiled. CT esophagram shows a leak. Okay. That is classic for like a gastric sleeve leak. Okay. I know this is not exactly an esophageal preparation, but as everyone in this room maybe has encountered, those leaks are very close to the esophagus. It's usually right on that left side, just below the esophagus. For those, of course, you need to get percutaneous drainage first, and then you could try to close it with an over the scope clip. Here I sometimes use methylene blue so that I can clearly identify where the fistula is. And then I prepare the fistula tract using APC, and I cauterize it with pretty high energy to try to change the histology to be a little bit more acute. And then we close it with the over the scope clip. If I could give you one piece of advice, before you install the over the scope clip, think about what the lips of the defect are. For example, if you wanted to shut me up from this lecture right now, you're going to install the clips like this, right, close me up like that. If you installed them like this, I'm just going to keep on talking, right? So you need to think about how you install them, right? Because the over the scope OTSC system, it's not like the padlock system where it's circular. It has a directionality that you need to think about. Okay. Number three, chronic small defect, soiled cavity, and then the IR says, sorry, Andy, can't get a drain in that little mediastinal abscess. What do you do there? Well, you use double pigtail stent, okay? This is just like walled off pancreatic necrosis. So this was a puncture from an EUS. This can also happen from with a dilation or ESD. Here I have a little clear cap. This is the distal esophagus, just above the G junction here. The patient had a stricture. I think that's why it happened. And you can see, I will isolate the defect for you right there. And you see there was pus in there. You can see that little pus and I suction a little bit. There's pus in there. Okay. Don't clip this. I will poop in your wall. Don't clip this. Okay. I know you want to clip it. Look at how those lips, they're just begging for a clip, aren't they? You do not clip this. Okay. I injected contrast. It showed this kind of amorphous defect, abscess, if you will. The patient had fever and chills, of course. And so we put a double pigtail stent in there. That's what you should do. Okay. Let that drain for three or four weeks. The patient comes back, you pull the drain out. It's closed like a little raisin, just sucked up. Then, yes, you can put a clip on it and finish it off. Okay. Here's another scenario. This one's the bad one, guys. Boerhoff syndrome, acute perforation, large defect, large cavity, heavily soiled. We get a lot of these in Texas, in the barbecue belt. We get EOE and then people eat too much, they don't swallow, and they blow the whole sausage into their mediastinal cavity. Okay. For this one, there's no choice. Ladies and gentlemen, there's no choice. You must do EVT or the patient has to go for a radical diversion. Okay. You have to use endoscopic vacuum therapy. This is a real case, food bolus, a young man with EOE, that's two-day-old sausage and brisket. It was terrible. We sucked it all out. We debrided it. We put the vacuum sponge in there. We came back the second time. It had formed a granulation tunnel. This is progress. Okay. Much cleaner. Unfortunately, we found a second tunnel. That's the nature of the mediastinum. It's not an obvious circular-shaped structure. There's a second tunnel. I think there's some pimento cheese in there. The sausages we eat in Texas are crazy. There's cheese and jalapeno. Anyways, we put another sponge in there. And then lo and behold, the patient got sicker and there was a new right mediastinal abscess that wasn't drained by the chest tube. So I had them put a mediastinal drain in there. And on the third endoscopy, I went in there, that's the right mediastinum. You can see the chest tube and the pigtail drain in there. And then I kind of had to do a little transoral vats here to get that drain unclogged. We flushed that out. Okay. It was replaced. And then when I went back the fourth time, this is a beefy red tunnel. That's granulation tissue. That is the human glue. That is the glue of the human body and the mediastinum. Okay. And the other little separate tunnel, it closed. Okay. So I put another sponge in there. I put an intraluminal sponge because I knew that that was the glue and there was a little defect left. And I used combo therapy to close it with an OTSC. And the patient did very well. And this was like, this is the highlight of my career right there. I saved this young man and that's why I'm here today to tell you about it. 21 days. Okay. So briefly, EVT, I'm not going to teach you how to do it. You can do intraluminal or extraluminal. Extraluminal is in the cavity. That's better. This is logistically how you do it. I don't have time to teach you how to do it. You cut a little sponge. Anytime you replace it, three to five days later, you cut it down smaller and smaller. This is same as wound care concepts. And you make sure you want to tie a little suture loop so you can drag it down there. This is a post-Iver-Lewis esophagectomy leak. This is the second. This is not true Borhab's. This is much more chronic. The patient come in one to two weeks later, chest pain and fever. What you just saw there was actually the defect. That wasn't the lumen. That was actually the cavity. The lumen's on the right over here. So in this situation, again, we put the sponge in there. You can see it at the bottom. The cavity's very soiled. You can see that. And after the first one, you can see some granulation tissue. That stippling is starting to form. But it's not beefy red yet. This demonstrates an important concept. There's an ischemic ulcer there right at the turn here. And you have to debride that. The sponge can't take care of that. You need to kind of debride that away. And after I debrided that away, I put the sponge back in there. And we sucked on it for another probably another two weeks. Until finally we got this. And that was closure. But it had this little lip there. So I ended up suturing that and then placing a stent over the top. Okay. I'm not going to show you the suturing. But here's the stent. I suture and I stent on top. Because I cannot afford a recurrence. Okay. So I throw the whole kitchen sink. So this is my modality for big fistula closures. I do EVT for two or three sessions to clean it out and change the histology until I get that beefy red granulation tissue. And then when I get that granulation tissue and I know it's soft and acute histology, then I close it with a suturing system. And then I place a stent across that and I suture that in. And I make sure that they have diet. These are my take-home points. The 4Ds, the 3Ds. And consider using combination therapy. It's really such a pleasure and a luxury to be able to talk to you today. Thank you.

Dr. Andy Thao

esophageal leaks

fistulas

endoscopic techniques

multidisciplinary approaches