Welcome to the Pericardial Access for Epicardial Mapping and Ablation section of the Core Concepts in EP. I'm Roderick Tung, Professor of Medicine at the University of Arizona College of Medicine in Phoenix. And what we'll be talking about is really the why, the when, and the hows of epicardial mapping and a lot of the anatomic considerations. I have no disclosures related to this topic. And the real question is why, in 1996, was that first puncture heard around the world to Bannishagas disease? Why was that necessary? And if you look at ischemic cardiomyopathy, which was the other subgroup that was required in inferior infarctions done by Sosa and colleagues, our success rates in VT simply just are not great. And in 2008, the Thermocol registry showed that the best of the best at that time with an irrigated catheter had a success rate of around 50% at six months. So that's really pretty much of a coin toss. And if you think about why we're not more successful, if we understand the mechanism, it could have been that we just don't have the best surrogates during sinus rhythm. We can't always map VT. There's multiple VTs, but it also could be that ablation fails to be able to destroy the substrate transmurally, and the substrate could be anywhere within the myocardial wall. So if you look at pattern A, which is transmural, this is very different than a subendocardial in B, a D, which is subepicardial, and C, which is midmyocardial. The worst scenario is trying to ablate D from the endocardium. Pattern C has a better chance, although pattern C can be very challenging as well, these midmyocardial substrates. So the idea that even non ischemics are now the predominant form that are referred, there can be scar at any layer, any depth, because it can be compartmentalized within the mid, the epi, or the endocardium. And in a very seminal paper by the UPenn group, Kano and Marchlinsky showed that in non ischemic cardiomyopathy patients, which typically have a basal lateral paravalvular substrate, the epicardial low voltage abnormalities typically outnumber the amount of scar area on the endocardium. Again, a seminal paper to describe the need to be able to get into the pericardium to be able to address, map, characterize, and destroy this substrate in these basal lateral predominant subtypes of non ischemic cardiomyopathy. So when we think about epicardial mapping and ablation, it's a two-dimensional approach to a very complex three-dimensional structure. And if you give the patient an analogy that if a wall has a crack in it, would you rather patch one side or other side, you could clearly make a case that looking at both surfaces and trying to address from both sides is going to improve the ability to create a durable and more transmural impact on the substrate. So when do we consider epicardial mapping and ablation? Well, here in this very nice summary by Dr. Boyle and Shiv Kumar from UCLA, you can look at different characteristics. And many people look at the ECG to be able to see it wide and slurring. And we'll talk about the limitations of that. Because obviously the farther you are from the conduction system, you might have an initial intrinsical deflection that's more slurred and broad because it's muscle to muscle. Prior unsuccessful endocardial ablation often signifies an indication for epicardial ablation, but perhaps imaging, being able to see that there's actually epicardial scar might be the most important, as well as the underlying substrate in which ARVC and non ischemics tend to have more epicardial substrate than ischemic cardiomyopathy. So in looking at ischemic and non ischemic, the critical differences are threefold. Number one, the scars are more subendocardially based because the wavefront of necrosis described by Reimer and Jennings goes from endo to epi when you occlude a coronary artery. Number two, the voltage abnormalities are more extensive and more dense in ischemic cardiomyopathy. And number three, as a result, the lateness of late potentials is often relatively modest in non ischemic, which tends to have more mixed scars, smaller scars. So some of these late potentials are not as abnormal in the non ischemic population. In this analysis, we actually found that those that undergo an epicardial approach with endo compared to endo alone had improved outcomes in a non-randomized experience at UCLA now actually 10 years ago. And here you can see that we can achieve great success over 80%, much better than the 50% seen in Thermacool compared to something just below 60, which is around 56, 57% in a small group of patients. So this is relevant in the sense that even ischemic cardiomyopathy may have significant improvement in their freedom from VT with an epicardial approach, again, some referral bias if there's been previously failed ablation. Here would be an example in which you're showing that there's complete epicardial re-entry that here is indicated by the mid-diastolic and pre-systolic epicardial potentials. And on endocardium, you don't see much of these signals across the way sub-endocardially. So this really bucks the notion that ischemic cardiomyopathy does not have epicardial substrate because in fact, many cases actually have complete epicardial re-entry, even in the setting of ischemic cardiomyopathy. Now we started taking this one step further and looking at both surfaces, simultaneous epi and endo mapping for different VTs. And you can see in this very simplified schematic that a true two-dimensional VT actually has re-entry on one surface and a focal breakthrough on the other. Then you might have situations in which you have partial re-entry on one surface, like the middle, and then you have complete on the other. And this is really more of a three-dimensional VT because it involves more than one surface where there are components of the isthmus on both epi and endo. And then lastly, the most confusing one is that if you take the isthmus and you turn it 90 degrees and rotate it, it can be completely confined within the mid-myocardium and you only see exit focal activation on the epi and the endo. This new construct of looking at human VT from a three-dimensional perspective also challenges the notion that VTs should be categorized in a binary fashion as endo or epi. What if there are isthmus components on both? What if there's exit activation on both? Is that an epicardial VT or an endocardial VT? And how do we define what an epicardial VT is? Does it require entrainment? Does it require termination? Does it require all the circuitry or does it require some of the circuitry? The bottom line is that the vast majority of VTs in over 85% of cases have critical circuitry on both surfaces of the heart. And here you can see an ischemic cardiomyopathy, three-dimensional VTs comprise 73% and non-ischemic 49%. When you look at the scenario of ARVC, this very important multicenter retrospective study showed that the combination of an endo and epi approach outperforms an endocardial based strategy alone in terms of freedom from VT in a very significant fashion here, over 20% difference. So this again highlights the fact that ARVC is sometimes the opposite of ischemic cardiomyopathy where the abnormality starts from the epicardial surface and goes to the endocardial surface. And the Hopkins experience shows in 2007 before irrigate catheters, high resolution mapping, success rate from VT was below 20% at one and two years. And you can see that with the increasing implementation of epicardial strategies, the success rate started increasing to 84% at one year. And then more recently in 2015, they actually were mapping the circuits and show that 69% of the circuits were epicardial and the success rates were approximately 73%. So with increasing implementation of epicardial, we get improved outcomes in an ARVC sort of substrate. Now, what about QRS with EKG signatures? Here's an example of a very wide VT with a QRS complex of 300 milliseconds. You could quickly say, well, this looks to be delayed engagement of the hysperkinesis system. It's wide. There's some slurring, but if you pacemap from the endo and the epi, in both situations, you get wide QRS pacemaps here, 290 and the other 285, not much of a difference. So it might be an oversimplification to say that an epicardial pacemap is just a lot wider than the endo, where in some of these patients that are very sick hearts with antiarrhythmics, you might get a wide pace morphology regardless of where you pace from. And in this situation with a very wide QRS, we actually had an endocardial termination of the VT. So again, this all depends when you think about what the gold standard for an epicardial VT really is, because here it looks to be epicardial, but it's actually successful endocardially. One of the reasons why QRS characteristics probably fail is this concept of TR fusion. When the T wave starts merging into the next R wave, it's very hard to distinguish or discern where the QRS onset is. Here you can see it looks like it's a complete pseudo delta wave in simulated VT that's just pacing rapidly. But if you look at the first beat prior to the pacing more rapid in which the T fuses into the R, you can see their separation of the T wave and then the R wave, and there's a bona fide isoelectric segment. This here is where the QRS activation starts. But in the pericordium, when they're slurring in, you don't know where that level is. So that's why we like the intrinsicoid notches in like V2. So these we call pseudo pseudo delta waves. And then you can understand that the inner observer variability for determining TRS onset might be quite high, which really confounds the ability for these QRS characteristics to be successful. And as a matter of fact, Martinick and Stevenson showed that none of the criteria that have been useful in idiopathic PVCs were useful in the setting of scar related VT with regard to pseudo delta intrinsicoid deflections, MDIs, to be able to predict an epicardial VT. Again, many limitations of these analyses because what is the definition of an epicardial VT? Is it a case mapping? Is it required entrainment mapping? Does it require the entire isthmus or part of the isthmus? And in the 3D VT paper, we offer six different options for what an epicardial VT could be. Most pragmatically, maybe a termination. That's what most people care about. But that's also not mechanistically the most important. And again, the fact that you could have non-binary activation on both epi and endo at different parts of the circuit also introduces that a binary categorization of this also may be challenging and inaccurate. Here you can see that with the combined epi and endo approach in structural heart disease, there's actually improved evidence, evidence of improved mortality and VT recurrence in this systematic metal analysis performed by Jorge Romero. So here we now talk about the how. And one of the problems with the LV summit, which is often thought to be the epicardium, is that there's always fat and overlying coronary arteries in this region, which really impair the ability to have adequate contact and radiofrequency delivery to the epicardial LV summit. And these coronaries and fat is a common theme as to why epicardial ablation can be challenging. That is the reason why mapping and ablation from within the coronary venous system, which is immediately adjacent to the actual sub-epicardial surface, might be better because it's underneath and within the fat to be able to record in the regions of the LV summit. And this is why an epicardial approach is seldomly useful in situations in which you need to manage the LV summit. How do we do this? Well, you can clearly use the TUI needle or you can use a micropuncture and do a needle-in-needle approach. One of them, you give up the tactile feedback when it's a small needle, but it's obviously a smaller hole, which is why if you have a complication, micropuncture makes sense. However, you're trained, you're using an LAO view, trying to use a 45-degree angle, contrast is injected. And once you're able to puncture the pericardium, then you're in the pericardial space. And then you see that the map can then direct you with the camera angulation towards LAO towards the left heart border. And it gives you an LAO projection, which must be seen in order to confirm that you have not punctured the right ventricle. This was the technique popularized by Sosa and colleagues at all in 1996, and is implemented throughout the world in this fashion. It's important to remember that an RAO, you can only tell the depth of the angulation. So shallow gives you something more apical and something deeper will give you something more basal if you have the steepness of this changing. So if you want to access the basal portion, it's deeper, apical is more shallow. If you want to get anterior, you need to go almost parallel, immediately below the xiphoid. In the LAO view, you don't want to go directly 90 degrees in because this increases the risk of RV puncture. You'd rather be a little bit more tangential, but you don't want to be too septal because we have had inadvertent punctures of the PDA or the MCV when you're too septal. and this can be useful because the RV lead also signifies where the septum is at the very least. So you don't want to be two septal and you don't want to be 90 degrees in. Here would just be an example of tenting, introducing the wire through and being able to see the staining with the wire then coming through in close-up magnification. Here is an example of the human anatomy from the bodies exhibit which actually shows an anterior approach can be performed safely. It is immediately parallel, almost like a subclavian below the subxiphoid process. The problem is that you have a lima and a superior epigastric continuation of the lima which is just a little bit leftward of the center midline point which could cause a bleed and there's the dreaded double RV puncture in which you can puncture in and then puncture back out and not be aware that you've actually skewered the right ventricle. Posterior approach is preferable to be able to access the more basal portions and the inferior portions of the heart immediately but this also risks liver and the setting of hepatomegaly that could be a problem in heart failure and right heart failure and diaphragmatic puncture and there can be small diaphragmatic vessels that can create an abdominal bleed. Here would be an example from an image by Jacob Carruth and Andre Davila of a liver hematoma. This could be a problem and sometimes the sheath can go through the liver. That would be a complication and then with an anterior approach you minimize the risk of liver puncture. You do a cross table lateral and here you can see access to the anterior apical portion of the right ventricle sneaking underneath almost parallel to the sub xiphoid region. The dreaded double RV puncture is one in which you don't know that you have a problem going in. You believe you're in the epicardial space and then there's a bleed out once you pull the sheath because now you have two holes in. The key here is that this typically happens more with anterior punctures and that's why you want to just sneak up with the puffs of contrast rather than going too far over the cardiac silhouette. Pericardial adhesions can be a problem and this here looks like loculated pockets of contrast when there's adhesions, whereas a pericardiogram can be useful to see if there's layering or if there's just normal circulation that's equally distributed. Here would be example again of these loculated contrast pockets and sometimes you can take a wire and perform adhesional lysis to be able to get it slowly in, but this does increase the risk of bleeding. Sub xiphoid windows were then innovated by Sojima and Stevenson in the EP lab in order to mitigate that risk of bleeding by being able to do manual lysis of adhesions to get into the space and here's just an example of how you have very nice access to the diaphragmatic inferior surface of the heart through a sub xiphoid window. If you have a history of cardiac surgery, the general practice is not to attempt percutaneous access because this is usually limited by dense pericardial adhesions. It is possible to get in, but then you typically have a lot of work to do in terms of being able to free up these adhesions and it increases the risk of a complication. If the access point is inferior, you can only get to the subdiaphragmatic surface of the heart. If the area of interest is actually lateral or anterior, we recommend doing a limited anterior thoracotomy which is just below the level of the nipple here and here you can see then you gain access to an anterior region in a patient here that had a congenital disease, an LED infarction from a ligation issue and then here you can see that there's basic concealed entrainment at an exit region here anteriorly afforded by a limited anterior thoracotomy. So there's 10 tips for safe epicardial approach. Number one, know the complications from the regional anatomy and we talked about the liver and the track of the needle and the course of that puncture. Hold on to coagulation as possible. Some antiplatelets are the ones that bleed more than actually anticoagulants and always have surgical backup. In the rare event that you have a complication, this can be really a time-sensitive complication that could be bailed out if you're able to do a sternotomy or a stitch in the lab and avoid surgery, avoid this after surgery if you know there's cabbage or a valve because there's going to be a lot of adhesions and it does increase the risk for artery punctures and other bleeds. Always check the wire and LAO to make sure you go far left lateral and a pericardiogram can be useful to be able to see adhesions and prior to putting a large sheath in just a simple fire French can confirm that you're in the pericardium and see how the contrast distributes. We always implement a double wire technique because anytime you're able to gain access into a very privileged space, it's better to be able to keep a second wire in, clip it to the side, put a tegaderm over it. So in the case of an inadvertent sheath that backs out of the pericardial space, you don't double your risk of a complication by having a stick again. And in many cases, if you're endocardial, you're already on heparinization or heparin and you have to reverse and then do it again. Never leave a sheath exposed and I'll show an image provided in which you can get a laceration with a naked sheath that doesn't have something that a space filler like a catheter. And prior to ablation in the epicardium, you want to ask yourself two questions. What can I damage in terms of a collateral incident and collateral problem? And that's coronaries and the phrenic nerve. And the last thing is to mitigate the risk of adhesions and significant pericarditis and increase the chances of being able to get back into the pericardial space. You want to irrigate any bleeding clear with saline and then also administer interpericardial steroids, which was shown by Andre Davila to decrease the chances in a preclinical canine model for pericarditis. In general, it can be challenging when you're in the pericardial space to be able to discern low voltage from fat or scar. But if you have late potentials, and if you have significant fractionation, you can use this as a potential tool to differentiate scar from just insulated fat. Here's an example in which there's a critical amount of fat here over four to five millimeters of fat actually is required to be able to insulate the myocardium enough to show up as low voltage. Here in this example in which anterior interventricular fat here is only three millimeters does not show up as actual low voltage. When there is fat, Andre Davila has taught us that this actually greatly impairs the ability to create a lesion. Here you can see the zone of the ablation and over the areas of thicker fat, the depth of penetration is much more shallow. This here is an example of the type of dimensions we see, six millimeters deep in normal tissue, four millimeters over fat. So it trims this off by 2.6 millimeters if you have overlying fat with cool tip radiofrequency ablation. The phrenic nerve is the other thing to think about when you're ablating and understand that in this beautiful paper by Sanchez-Quintana and Cabrera, that there are three courses either anterior to the left atrial appendage, the expected course which is at the base of the left atrial appendage, and then you can actually get it even posterior to left atrial appendage. So the phrenic nerve can be anywhere and that's why in these non-ischemic cardiomyopathy patients, anytime you have a lateral wall, anterolateral, or infralateral substrate, you want to pace map and be able to see if you can see capture and pace map the phrenic and then tag those areas of where the phrenic nerve is so you can mitigate the risk for left phrenic injury. One option if the phrenic nerve is there over a critical area is to introduce an epicardial balloon. These are techniques in order to create increased separation. This could also be done with contrast as well as air. With regard to coronary arteries, a safe distance as endorsed by experts and the heart rhythm consensus statement is greater than five millimeters throughout all heart cycles, throughout the cardiac cycle, and here you can see that ablation directly over coronary arteries in preclinical models has been shown to accelerate intimal hyperplasia. With very detailed analysis of coronary arteries and epicardial fat, the two major barriers to widespread epicardial ablation, Katja Zeppenfeld and Van Hul's Van Texas showed that the two most common reasons for failure is when the fat is thick over seven millimeters in the presence of coronary arteries. This is the major issue with epicardial ablation is the acknowledgement and the awareness of these two critical structures. This is just the reference by Andradavila and Vivek Reddy showing that in 25 healthy swine, the introduction of triamcinolone, of hyaluronic acid and triamcinolone significantly reduced the incidence of pericardial adhesions 14 days after ablation and here you can see in A, B, and C, and D, there are nasty hemorrhagic pericarditis responses and with the intermediate acting corticosteroids, they were able to prevent and get a clean looking pericardium that was not irritated in panels E and F. Lastly, the known complications of epicardial ablation are really related to the access site, the things that you can injure along the track, and then the dreaded RV puncture. If you go in too deep, you simply can end up in the right ventricle. It's important to understand that this has been estimated to be in one out of five punctures, but this was also a very early experience now, hard to believe 13 years ago. Fortunately, the vast majority of these punctures do not translate into significant adverse events because if you're in a contractile portion of the RV, it can actually be sealing itself up with every contraction. Therefore, the incidence of intrapericardial bleeding is defined by greater than 80 milliliters was only 4.5 in this multi-center experience. Delayed pericardial reactions and effusions can happen, but again the RV puncture without consequence here was noted in almost 20 percent in 17 percent of cases. This is an image that I believe is very worthwhile taking a mental snapshot of because here you can see that the wire in LAO does not reach the left heart border where the coronary sinus lead is. If you cannot get to the left heart border, you have not proven that you are not in the right ventricle and having the septum be the barrier between you and the left heart border. This is an example of an RV puncture in which the wire traverses what appears to be multiple chambers but doesn't get to the left side of the heart and that's because it's going through the RV retrograde up into the RVOT and then out the right pulmonary artery. This easily can fool you because you believe it's traversing multiple cardiac chambers throughout the silhouette, but again it does violate the major rule that you have to see the wire go to LAO and putting a sheath in this can cause a catastrophic complication, whereas early recognition of this you can just pull the wire back and try to gain access to the pericardium as if it's a through and through stick. This is a beautiful image by Young-Hoon Kim in which they were able to save the patient from a laceration on a vein because of an unprotected sheath. You could imagine that with every systole and diastole an unprotected sheath actually could be abrasive and actually if you're unlucky create a small laceration or a perforation of a vein and that's why we always advocate that the sheaths anytime you have an exchange you have some other catheter a mapping and ablation catheter to go into that space. The unusual complications of this are not only interpericardial bleeding but also abdominal bleeding and if you have pericardial clots it can really destroy your ability or minimize your ability to actually drain all the pericardial fluid and this also decreases the chances that you're going to have resolution. We do advocate a double suction technique to be able to try to drain the pericardium dry so that anything that's bleeding can tack up against and oppose the pericardium whereas allowing for the formation of a pericardial thrombus with kind of retraction of the overall heart not allowing it to tack down the pericardium is a bad way to resolve tamponade. So in conclusion ladies and gentlemen the knowledge of the needle path from skin to the subxiphoid maybe the liver maybe the superior epigastric maybe the diaphragm is so critical to understand where you might have complications and to recognize them. Number two follow those top 10 tips double wire LAO maybe sneak up with a very small fire french before putting a large dilator to be 100 sure you're in the pericardial space and assess for adhesions. In the setting of surgery we do not advocate trying to get access because this is usually restricted by pericardial effusions and increases the potential for complications that cannot be resolved without expanding this into a sternotomy. ARVC and anastemics typically have more epicardial substrate they're enriched for epicardial success but I want to remind you that our definition of epicardial VT is continually a moving definition that does not have a gold standard. There are different ways to define this and the surface criteria are not good based on our new understandings of 3D VT and how sometimes you can have exits on both surfaces of the heart. I hope this has been helpful for you we appreciate you tuning into core concepts and thank you very much for your attention.

Epicardial mapping

Epicardial ablation

Ventricular tachycardia

Ischemic cardiomyopathy

Two-dimensional approach

Epicardial scar tissue

Epicardial access techniques