So, good afternoon to everybody. I think we can start the session. My name is Francesco Tannone from Italy, and my co-chair is Dr. Batul from Geisinger Group at US. So this interesting session is understanding PACE-QRS complex and conduction system pacing. Why? Because there is strong evidence that we need to know what we are doing. So we need to classify, especially if we are doing A, B or C, and not put all together. So we have five experts on this topic, and I want to invite the first, which is Dr. Ponsunami from India, which will talk on HEG conduction system pacing, an essential tool for success. Good afternoon, everyone. For the next 10 minutes, I'm going to discuss about understanding the pace QRS morphologies in patients undergoing conduction system pacing. So these are my disclosures. I'll just start with a question. So where is the pacing lead here? It's an easy guess. It's in the left bundle branch area. So it's an easy guess. We can know that the lead is right in the left bundle area. Where is the pacing lead here? Any guess? I'll give you the options. Non-selective his bundle capture, parazine myocardial capture, non-selective left bundle capture, LV septal capture, or just like me, I'm confused. Let me sort out the confusion in the next 10 minutes. So conduction system pacing essentially involves capturing the cardiac conduction system fibers starting from the his bundle until the vesicular fibers. So basically, you can see different patterns of ECG while capturing the same site. So this is because the pacing lead is essentially going to capture two different substrates. One is a thick myocardial fibers. Another one is a thin conduction system fibers. So we know that the conduction system fiber and the myocardial fiber has different conduction velocity, has different capture threshold, and has different refractory period. Essentially, all these three things will put together to generate a different pattern of ECG depending upon which substrate is getting captured. So why is it important? So we know that studies have essentially shown both his bundle as well as left bundle branch pacing is much, much superior as compared to the RV pacing in terms of reducing pacing-induced cardiomyopathy. So his bundle pacing, so depending upon the anatomical orientation, we can have three different patterns of ECG. So one is a selective capture where the impulses will travel down only along the his bundle fibers to reach the ventricles after a delay of 35 to 55 milliseconds. So this is essentially a selective capture. In non-selective capture, there will be capture of surrounding myocardial tissues as well. So the pacing artifact will be immediately followed by a delta wave, which is just like a WPW syndrome like delta wave, which will coincide along with the QRS onset. So here, the impulse will travel down both via the conduction system fiber as well as the myocardial fibers. The third variety, where there will be only a myocardial capture, there will not be any activation of the conduction system fibers. So the ECG will look like a classical left bundle branch block like ECG along with multiple notches and slurry. So we can have at least four different patterns of ECG transitions when you're pacing the his bundle. So if the lead is right on the top of the his bundle, we can have either a transition from selective capture at high output to loss of capture at a very low output. But more commonly, you will have a non-selective capture at high output. Once you reduce the output, you'll have a selective capture. Further reduction will result in loss of capture. But if the lead is in the his bundle zone, just one or two millimeters beyond the his bundle zone, at high output, you'll have a non-selective capture. But once you reduce the output, you'll lose the conduction system capture. It will be only a myocardial capture. And at very low output, you lose the capture. Loss of capture will be seen. In patients with an associated wide QRS, can have a non-selective his bundle capture with complete correction at high output. Once you reduce the output, you'll have a selective capture with complete correction. Once you reduce the output still further, you'll have a selective capture without correction, ultimately resulting in loss of capture. But these three things are very important. You need to know part of the things which may not be seen, which cannot be seen. You can go only from up to down. You cannot go from below to upwards. You cannot have a selective to non-selective capture transition. You cannot have a selective to myocardial capture transition. And you cannot have without bundle band correction to with bundle band correction transition. So the most burning question is to differentiate this non-selective his bundle pacing from myocardial pacing. So we know that both QRS will be a little bit wider. But how to differentiate? There are two important things to appreciate. So one is by either reducing the output or reducing the extra stimuli cycle length. So you'll be able to appreciate a change in QRS morphology. So you'll have a more sharp R waves, more sharp S waves when you have a conduction system captured in non-selective his bundle pacing. But once you reduce the output, once there is a loss of conduction system capture, the QRS duration will be prolonged and you may be able to see, appreciate the roundening of R wave and slurring of upstroke along with notches on the downstroke. So all these indicates a myocardial capture rather than a non-selective capture. This is one way of differentiating a non-selective his bundle capture from myocardial capture. But the easiest thing is what Marek has shown, measure the R wave peak time in B6. And though both overlaps, a magic number of 100 milliseconds will help us to differentiate a conduction system capture from myocardial capture. So anything less than 100 milliseconds, it implies a non-selective capture of his bundle and anything more than 100 milliseconds, it implies a pure myocardial capture. So the most important question which arises is to differentiating a non-selective his bundle pacing versus the myocardial pacing. So you can use these two important criteria. The QRS morphology, whether it is very short or very broadened and the R wave peak time, whether it is less than 100 milliseconds or more than 100 milliseconds. Let us move on to my second half of the topic, this lumbar branch area pacing. We have recently a good definition for a lumbar branch area pacing. It's defined as capture of LV subendocardial fiber on the inter-vertical septum without capturing the conduction system. Remember, this includes LV septal pacing. So lumbar branch area pacing includes both lumbar branch pacing, vesicular pacing, as well as the LV septal pacing. Though a lot of criterias are there, I'm not going to discuss detail about the criterias. The role of ECG in lumbar branch pacing comes in both at the time of implantation as well as for confirming the capture. So during implantation, either you can give rapid rotation to generate PVCs or much more easily, you can continuously pace while deploying the lead. So as the lead traverses from the right side to the left side of the septum, you'll be able to appreciate the change in QRS morphology from QS when the lead is on the right side of the septum to QR once the lead reaches the left side of the septum. So this QR morphology beats, which we label it as either a template beat or fixation beat will imply that the lead has already reached the lumbar branch area. So all you need to do is look for capture confirmation criteria and give further rotation only if it is indicated. The QRS morphology changes depending upon the site of pacing. So whether you are going to capture the proximal main bundle or one of its pre-cycle. If you're pacing and capturing the proximal main bundle, the QRS will almost look like a normal appearing QRS, but for the RBBB delay pattern. We know that the RBBB delay pattern is the hallmark of capturing the left bundle. So apart from that, the QRS will appear normal, but in patients with a fascicular capture, we'll have a shift in axis, either towards the leftward or rightward, depending upon the fibers getting captured. So though the QRS duration differs in a main trunk pacing versus the fascicular pacing, we don't know the long-term clinical outcomes because we don't have any hemodynamic studies to demonstrate one is superior to the other. The capture transition, just like we have discussed in the HisModel pacing can be demonstrated in left bundle bench pacing, and though it is difficult to appreciate in a 12-lead ECG at a 25-millimeter speed, once you increase the sweep speed to either 100 or 150, you'll be able to see both the changes at the local EGM. You can see a distinct artifact followed by an isoelectric interval, followed by the appearance of the ventricular electrogram, as well as on the surface ECG. The surface ECG, we tend to appreciate three important changes. One, in lead V1, you'll be seeing a QR pattern changing to RSR or M pattern. And the second important thing, which is easily appreciable is increase in amplitude of S wave in lead V6. So this is much more appreciable as compared to the change in V1. You'll be able to see appearance of a new S wave or increase in amplitude of the previously existing S wave, both in terms of amplitude as well as duration in lead V6. Similarly, the terminal R component of lead AVR will be much more rounded if there is a selective capture as compared to the non-selective capture. Non-selective to septal transition is easily appreciable. You just have to monitor the lead V1, just trace lead V1 once you reduce the output, the R wave amplitude will tend to come down so ultimately resulting in a loss of R wave and you'll see only a QS complex in lead V1 along with prolongation of QRS duration and disappearance of S wave in the lateral leads. Physiology based ECG criteria, just to put it in a simple term, if the native R wave peak time is equal to the based R wave peak time, that will tell you that you have captured the cardiac conduction system. This V6, V1 inter-peak interval, this is just to play around between the activation delay of LV and RV. In non-selective capture, the LV will be activated first, followed by RV. In selective capture, the LV activation will remain the same, but the RV activation will be a little more pushed. But in LV septal activation, both will be pushed. So if you see the difference between this activation delay, in selective capture, the activation delay will be much more appreciable as compared to the LV septal pacing. So again, a magic number given by Marek is 33 milliseconds. If it is more than 33 milliseconds, it is of conduction system capture. And if it is less than 33 milliseconds, it is of a septal pacing. Programmed deep septal stimulation, again, the same principle. Give extra stimuli, reduce the cycle length. At one particular cycle length, you will have loss of conduction system capture. So that will be reflected on the surface ECG as a change in QRS morphology, duration, as well as access. Again, the burning question, which is very difficult to answer. Differentiating this lumbar branch pacing from LV septal pacing. I just put all the criterias together. So PACE QRS morphology, it's QR in almost all patients with a lumbar branch pacing as compared to 44% of the patient in LV septal pacing. And the important thing to appreciate is a terminal SVM in lead B1, which is never seen in patients with a lumbar branch pacing, but it is almost seen in 60% of the patient with LV septal pacing. R-wave peak time will be short and constant. It'll be different at different output. And an absolute value is given by Dr. Wong, less than 75 milliseconds in non-LV patient and less than 85 milliseconds in LV patient, but will be prolonged in patients with LV septal pacing. Potentials are not good for differentiating lumbar branch pacing versus LV septal pacing. It may or may not be seen in LV septal pacing, but capture transitions are. So you'll have either non-selective to selective or non-selective to septal transition in lumbar branch capture as compared to no transition in patients with the septal capture. And V6V1 interplay, it'll be less than 33 milliseconds and more than 33 milliseconds. And a physiology-based ECG criteria, native R-wave peak time and paced R-wave peak time will be almost equal in patients with lumbar branch capture as compared to a delay of more than 10 milliseconds in septal pacing. Programmed deep septal stimulation, if nothing works, if nothing works, no other criteria works, you can do the programmed deep septal stimulation to show change in QRS morphology and axis. Again, the principle is the same. We're capturing two different substrates. Let us go back to our patient, which I've shown in the first slide. So this is a little cumbersome case, though I'll try to explain it in the next one minute. So this was at a high output, three volt, and this was at a low output and one volt. So with this alone, it'll be very difficult to diagnose, but with this, I know at least 50% of the crowd will be able to diagnose what exactly it is. So at a low output, at a high output, you are seeing a QRS in lead V1, and at low output, you are seeing a QR in lead V1. So this will happen in at least 5% of the patient with a labrador branch pacing. This is what is called a masked RBB delay pattern. We have shown incidents of close to 5%, maracas stone and incidents of close to 8%. So this is an important thing to appreciate. So whenever you're doing a labrador branch pacing, try to pace at both high output and low output to see this masked RBB delay pattern, which is very, very common. So the theory which we proposed is either because of a retrograde conduction or a distal bifurcation of the His bundle, or maybe because the right bundle itself is coming from the left side of the septum and moving towards the right side. So we don't know the exact etiology, but still it is appreciable in almost 5% of the patient. So here, the pacing lead is in the labrador branch area. Yeah, so last time. So just to conclude, both the Hispon-KG system and the myocardium differ in their capture thresholds and refractoriness. So this can be utilized to see the changes in the ECG morphology and to confirm the conduction system capture. Thank you. So, in the interest of time, we'll just skip the questions towards the end. Our next speaker is Dr. Marek Czesrewski, and he will be highlighting a topic which has always confused me a little bit, and that's program stimulation to diagnose conduction system capture. Thank you for inviting me. My talk will be about program stimulation of the conduction system. This is a technique that was developed some five years ago in my lab to diagnose if you capture or not capture the conduction system. I think that this technique is very useful. On many occasions all other criteria will keep you in the dark but this technique can shed a light on your case and really provide a correct answer. It all have started with a simple random ECG like this one with asynchronous pacing, an ECG that required explanation why this third QRS is suddenly different. And when the answer came, and the answer is quite simple, that this final QRS is different because the pacing stimulus this time encroaches on the refractoriness of one of the two tissues that work together to create this non-selective QRS complex. When this answer occurred to us, I immediately knew that this observation has a diagnostic potential and it will have a future in conduction system pacing. So at first we used it for this random His-Bundle pacing cases where the conduction system capture can't be diagnosed with this simple threshold test because the threshold test exploits the differences in excitability and like 5-7% of the cases you have the same threshold, so there is no change. And in these cases when you provide a closely coupled extra stimulus then around the coupling interval of 330, 360 even, you see a dramatic change of the QRS complex because you capture only one tissue. This might be myocardial only capture or conduction system only capture. And that was the diagnostic clear-cut response and basically in 100% of the cases you can see this response and have a diagnosis. It was simple but not very useful because most of the time the threshold test does the job. But the real usefulness of that technique came when we started in 2018 to do left bundle branch pacing. The criteria were not really well developed at that time and we were looking for the program stimulation to provide the answer to the capture. As you see in this case the program stimulation provides the same QRS transition here as you can obtain with the program stimulation provides the same QRS transition as you can obtain with the threshold test. The same QRS is here as here in V1 and in V6. The same transition only you use different method instead of differences in the threshold test you use differences in the duration of the absolute refractory period. So it seemed simple. However, it was a bit more difficult to use than for his bundle pacing because in this picture you see the S2 delivered at more and more shorter coupling interval and you see how the QRS is changing in terms of V6 or with peak time prolongation and QRS prolongation. And some people complain that they use this technique and what they see is just a progressive QRS prolongation without clear cut transition. And the way out of this situation is first you can measure very closely because actually there is a diagnostic transition in this row of QRSs. And the second is to augment the difference in refractoriness. So we try to do that. And I recollected this paper by Denker from the 80s. I knew it very well because it's described in the Mark Josephson's clinical electrophysiology textbook. And this paper basically tells you that the refractoriness of the conduction system and refractoriness of the working myocardium behaves differently in response to changes in the preceding cycle length. To put it simply, the drive cycle length influences refractoriness of the working myocardium while the refractoriness of the conduction system is mainly influenced by the one preceding cycle. By using these differences in response to preceding cycles, you can increase the refractoriness of the tissue that you want. And we did that. So basically when you use a very fast drive, and then have a pause, so you create like an Ashman phenomenon, and then the S3 will find the left bundle refractory, and only the septal myocardium will be activated, as you see this preceding potential is now after the QRS. That's a myocardial response, selective myocardial response. But if you have a long, slow drive, and then one closely coupled extra stimulus to shorten the refractoriness of the left bundle, then the S3 will capture only the conduction system, only the left bundle, because the refractoriness of the working myocardium will be prolonged. So using this, you can really make the difference, you can augment the difference between the refractoriness of these two tissues, and the agnostic yield of this maneuver will be higher. However, there is one caveat. You can't really provide an extra stimulus, see a change in QRS, and see that this is the agnostic. I see this quite often during talks over the internet. This is not true. There is one caveat. You have to measure carefully, and not all changes are the agnostic. Why not? This slide is probably most important to explain this. So again, you see the QRS complexes as they change when you shorten the coupling interval. And at coupling interval of around 300, you see that the R-wave peak time from the baseline 87 prolonged over 10 milliseconds. So this is the agnostic. At that moment, you can say that you can have a diagnostic response. This is already a myocardial response. This QRS that you obtain here is the same as you can obtain during the threshold days in this patient. So you have a diagnosis. From that moment on, all that you have is myocardial capture, because the refractoriness of the left bundle makes it a pure myocardial capture. So LV septal pacing. Yet, despite that this is LV septal pacing, the QRS is still changing when you shorten the coupling interval. So you can easily see that if you have LV septal pacing, you will also see a QRS prolongation. So how to solve this situation? Why do you see it? Because there is a thing like relative refractory period of the working myocardium as well. So you may lose the left bundle, but you will still have a QRS prolongation because the refractoriness of the myocardium also plays a role. So the way out of that is to reject the last three or four QRS complexes, because that's non-specific response. Although the QRS very nicely changes from this R wave that you see here is absolutely absent here. You reject this last four, and then you go out of the relative refractory period and just look at these preceding cycles if you have a diagnostic response. And you have if you have left bundle branch pacing. And you will have not if this is just LV septal pacing. So this is the way how you should interpret the myocardial response. Take into consideration relative refractoriness of the working myocardium, and it will be diagnostic for you. So let me conclude that with a case. So you have implanted a left bundle branch area. Pacing lead, you have this QRS complex. It doesn't look that bad, but if you measure it carefully, you will see that the V6 R wave peak time is 96 milliseconds. But the use criteria, this suggests or actually diagnoses almost LV septal pacing. So it's just LV septal pacing. You look at the V6 V1 interpeak, and it's short. So again, it indicates LV septal pacing. There was no LVB potential and no QRS transition during the threshold test. So all that is consistent with LV septal pacing. So should we reposition the lead? No. What we should do is program stimulation. A simple protocol, my favorite, double extra stimuli on intrinsic rhythm. Intrinsic rhythm is in a way the slowest drive that you can have, so it will facilitate the selective response. Double extra stimuli, and what do we see? We see a dramatic change in V1, prolongation of the time to the peak, broad R wave. This is typical for selective response. You see completely changed axis, typical for selective response. And in the cardinal channel, you see the diffused potential, local potential is now separated. You see discrete local potential confirming that this is a selective response. And selective response is 100% pathognomonic diagnostic of conduction system capture. So what do we have? All these criteria have kept you in the dark. We're misleading. And program stimulation was a light in the dark. It really shed a light on your case and told you that your case was, after all, successful. So use this method because it will shed a light on your case as well. Thank you very much. Thank you, Marek. Thank you very much. The question and answer at the end. So it's my pleasure to introduce Ugal, which is talking on electrocardiographic marker of conduction system pacing importance of the V6 wave peak time. Thank you Francisco. Thank you Atika. It's a pleasure to be here and I know this is a quite a bit of a nuanced talk in terms of trying to figure out whether you have conduction system capture or not and some of them going to be repetitive. Hopefully the repetition will make it easier to understand. So my topic is very specific just our way peak time because when we started doing left bundle branch pacing, r-wave peak times or peak LV activation time became a common theme. So if you go back to the original description of r-wave peak times the definition is basically the instant at which the area of cardiac muscle immediately below a unipolar electrode is completely depolarized. It's a very old description but it still holds true and in the early literature v6 r-wave peak time or greater than 50 millisecond was indicative of eccentric left ventricular hypertrophy not concentric but more of an eccentric LV hypertrophy. So in a native QRS the longer r-wave peak times usually the result of prolongation of a transmural conduction from endocardium to epicardium and it can be dependent on multiple factors and which is critical because for you to understand how it behaves when we do left bundle branch pacing or even his bundle pacing. So how thick the muscle is, is there a degree of fibrosis in the myocardium, is the LV dilated and what happens to the conduction velocity especially when you use anti-arrhythmic drugs. All of this makes a difference. So here in systolic or concentric LVH the r-wave peak time is generally not as prolonged. However in patients with left bundle branch block or in patients who have eccentric left ventricular hypertrophy the peak r-wave times are often prolonged. So how do we use that in understanding conduction system capture? Some of this is going to be repetitive again. His bundle pacing and left bundle branch pacing. Often as Merak and Shanmugha pointed out, diagnosing his bundle capture is very straightforward. Often you see a transition from non-selective his bundle pacing to either myocardial or selective capture which is seen more than 90 percent of the time. In about 10 or 8 percent of the patient where there's no transition during threshold testing despite either adjusting the output or pulse width, then you resort to other features and Merak has done a lot of this work on r-wave peak times and analyzing the ECG spending countless hours. I don't know if I have the patience to do all of that. What he came out with is that if you have no notches or slurs in lead 1 v1 v4 to v6 along with r-wave peak times less than 100 milliseconds it's very pathogenic and diagnostic of his bundle capture. And so you can clearly say the difference is to know whether you have his bundle capture or is it just myocardial r-wave pacing. So there are slurs, notches, plateaus in the lead 1 v1 v4 6 or r-wave peak times more than 110 milliseconds. Except in the situation where there's prolonged conduction time long hv interval this may not hold true but most often this is very clear that this myocardial pacing. So I would suggest you look at this particular paper as you can see changes in the r-waves in lead 1 the plateau appearing same thing in v6 v4 those notches are very helpful. And moving on so one of the things that you can look at also these graphs distribution codes for the r-wave times you can see there is a zone that this is helpful to know that there are going to be patients where these numbers don't hold true. So you have to individualize these numbers for that particular patient. So you have to pay attention to what is the native his bundle to r-wave peak time and then when what happens when you do pacing. So that is what his next proposed criteria is that look at the r-wave peak times from his electrogram to r-waves in v6 and then what happens during pacing so if you have selective or non-selective capture that number is within 10 milliseconds and this number magic number was 12 milliseconds but you have only rv capture then that number will be a lot more than 10 12 milliseconds. So individualizing for that pacing for that particular patient is easier when you have his bundle pacing and because you almost always record his bundle potential. This becomes more challenging when you start looking at patients with left bundle branch pacing. So again you can see what you can differentiate this particular area of commonality where it's harder to differentiate using individual delta r-wave peak times is helpful. This is going to be helpful even at his bundle pacing I mean with left bundle branch pacing also. So as again pointed out by earlier speakers diagnosing left bundle branch capture is challenging. There's a whole different question. Is it important to know that you have left bundle branch capture? Many of us who are conduction system pacing enthusiasts will still believe in confirming and proving there's left bundle branch capture although maybe LV septal pacing may be adequate in many of those patients with bradycardia indication but still uh holding true to the physiology we want to prove that. And so threshold testing transition that we see more than 90 percent of the time in his bundle pacing doesn't happen with left bundle branch pacing. So that's why all of those multiple criteria that we use to try to confirm left bundle branch capture. And again another important caveat is just because you see left bundle branch potential is not enough to diagnose left bundle branch capture. It's clearly confirmed that it's a left bundle branch area pacing. But whether you have capture or not then you need to use other criteria. Early on if you look at the literature for left bundle branch pacing they came up with QRS duration if it's less than 130 milliseconds or V6 R wave peak time less than 75 milliseconds. So this led to a lot of challenges because QRS more than 50 percent of the time more than 130 milliseconds with left bundle branch pacing. And about 40 50 percent of patients don't have R wave peak time less than 75 here in the proximal left bundle branch region. So again this nice distribution curve shows that QRS duration for non-selective left bundle pacing or LV septal pacing almost identical. So you can use QRS duration in conduction system capture for left bundle branch pacing. So how about R wave peak times and then that again if you look at it there is a significant overlap for the R wave peak times. More than the his bundle pacing. So how do we use these numbers so you can see even with LV septal pacing you can have a lot of 60 70 milliseconds. So that is the challenge. And when you looked at series of 100 plus patients the cutoff that he came with was 75 milliseconds at a very good specificity for left bundle branch capture. While you want good sensitivity and specificity so anything less than 83 milliseconds was all fairly suggestive of left bundle branch capture. So if you want to use those of us who want to use absolute numbers and want to move on quickly in a long day with multiple procedures then this may be reasonably good enough. And then when you see that when you lose left bundle branch capture in an individual patient usually there's a 20 milliseconds or longer prolongation of the R wave peak times. And with left bundle branch block patients then that number was 85 milliseconds. Often when you have left bundle branch block there is dilatation there's eccentric hypertrophy and then conduction delay so that's why the higher number. And this has a good specificity but the sensitivity may not be good enough. And you can see that with non-selective capture to selective capture here in this particular patient 84 milliseconds in a narrow QRS complex and with selective capture that number remains the same. And you see the local isoelectric interval. And these are helpful if you see it but you may not always see it. So one of the things we use is we can use the difference between his bundle pacing or left bundle potential to QRS timing versus compared to when you what you see with left bundle branch pacing. In this particular patient with narrow QRS you can see at three walls it's a stem to peak activation time is of 105 millisecond while left bundle to activation time is 80 milliseconds. At high out even though there is a left bundle potential there is no left bundle branch capture at three walls. Only at eight volts you have. So you advance the lead a little bit further. This way it's helpful to know this when to stop. And so now you have a left bundle potential with injury current which is almost always associated with left bundle branch capture. And then now you can see at a low output non-selective selective transition and the stimulus to peak R wave time is equal. And wide complex rhythm is where you want to use a different number. And here in this patient you can see when you're in mid septum left septal capture only gives 96 milliseconds. When you have higher output there's left bundle branch capture. This again suggests that you need to advance the lead a little further. And once you advance the lead further now at low output or high output you have the same stimulus to peak LV activation time. So you need to individualize for that patient and that's often quite helpful. And lastly I want to finish off here. So sometimes here is a patient with left bundle branch block. Left bundle, well this deep septal left bundle branch area pacing has a stimulus to peak of 75 millisecond. Is it good enough? I see that the R wave times related to the local LV activation time through biotronic vision wire shows about 100 millisecond. That's a significant reduction from baseline. But when I do his bundle pacing you can see that that number is 97 millisecond. And the stimulus to peak activation time is 70 milliseconds. So what we got earlier was 75 milliseconds. That's not left bundle branch capture. That was left septal pacing. So we could have advanced this patient a little bit further. So we use a difference between his bundle capture to when you do left bundle branch pacing. We need to have difference of more than 10 milliseconds to confirm that you have left bundle branch capture. So I'll stop there in the interest of time. Thank you very much. Our next speaker is Margarita Lopez from University of Barcelona. And she's going to talk about body surface mapping and ultra high frequency ECG in conduction system pacing. Good afternoon. Good afternoon. Thank you so much for the invitation. It's my pleasure to present a topic that I love, which is the electrocardiographic imaging and ultra high frequency ECG in conduction system pacing. It's my pleasure to present this topic, electrocardiographic imaging and ultra high frequency ECG in conduction system pacing. My disclosures. I would like to start my presentation with talking about the two electrocardiographic techniques, body surface mapping and ultra high frequency ECG. And after that we are going to talk about the applications of both techniques in conduction system pacing. Electrocardiographic imaging uses a body surface potentials in order to obtain a picardial maps. And to go from here to the maps, we use mathematical algorithms to solve the inverse solution and we obtain the electrocardiographic imaging maps. It's a electrocardiographic imaging, it's non-invasive, it's multi-lead and some systems need CT and MRI, but some systems no longer need imaging tool. On the bottom of the slide, we have three systems. On the right, we can see a system that is FDA approved. On the left, it's a research device, it's a belt. And in the middle is the device that we use in our center. We have moved from the first device and now this doesn't need an imaging technique. It reconstructs the image of the torso with artificial intelligence. This is the first patient that we did with his bundle pacing and that we perform electrocardiographic imaging. Moving to ultra-high-frequency ECG, ultra-high-frequency ECG uses the spectrum of the high frequency. And we can see here in a patient with left bundle range block that we have shift on time the B1 and B6 activation. This is in contraposition that we obtain with a normal ECG that we see B1 and B6 overlapped in time. In a patient with a narrow QRS, what we obtain is overlapping of the both signals, B1 and B6 are in the same time. To make information easy to read, we transcript this information into this polarization maps. And what we obtain is this kind of maps. We are using 12-lit ECG. And we are using the 12-lit ECG is this kind of maps. We are using 12-lit ECG plus two more lits, B7 and B8. And then we have information from the right ventricle, apex, and free wall of the left ventricle. And what we see here is the first point of activation and the last point of activation. And we can calculate the desynchrony index. This desynchrony index is different if we have a left bundle range block or right bundle range block. If we have left bundle range block, we have first activation of the right ventricle septum and delayed activation of the left ventricle. It's a positive desynchrony index. And if you have right bundle range block, then it's on the other side. It's a desynchrony index that it's negative. Now we are going to move to talk about the applications of body surface mapping and ultra-high frequency ECG in conduction system pacing. We have five topics to talk. The first is assessment of resynchronization itself as an endpoint with electrocardiographic imaging. Second, pre-implant, for implant optimization, and follow-up. First, in relation to showing resynchronization itself as an endpoint, we have this beautiful study performed by a group in the Imperial College in London. They performed ventricular pacing and an accurate study with his pacing. The lead was not screwed. And what they obtained is that with his bundle pacing, the decrease in left ventricular activation time and the decrease in left ventricular desynchrony index was greater with his bundle pacing compared to the ventricular pacing. The other study that we have used as an endpoint, the left ventricular activation time with electrocardiographic imaging is the LEVELAT trial. This is a randomized trial published last year, 35 patients to conduction system pacing and 35 to ventricular pacing. And we obtained a similar decrease of left ventricular activation time with conduction system pacing and with ventricular pacing. Here we have a beautiful map with homogeneous activation of the left ventricle with conduction system pacing. This is what we obtained with intention to treat. It was non-inferiority data for conduction system pacing compared to ventricular pacing. And if we do a treatment-received analysis, conduction system pacing was superior. We are moving now to pre-implant for patient selection. Strozzi et al. did this study with computational modeling. They simulated conduction system pacing, CRT, hot CRT, and low CRT. And they showed that if you have conduction system pacing, it's perfect. But if you have a delay in the His-Polk engine, then conduction system pacing is negatively affected. And in this case, hot CRT and low CRT could be better in this context. On the other hand, if we have septal scar simulated with this computational model, CRT will be better compared to His-Polk pacing and left underground pacing. In this context, we are working with our team in order to propose with electrocardiographic imaging if we can show that maybe some patients could benefit more from conduction system pacing than ventricular pacing. We have to look for the niche for the patients that will need conduction system pacing. And we think we do think that electrocardiographic imaging could be important here. There are some index that can be calculated with electrocardiographic imaging. There is the difference between the left ventricular activation and the right ventricular activation. If it's close to zero, there is less dis-synchrony. And if you have greater VEW, more than 50, then we have more dis-synchrony. And these patients showed to have higher response. We have analyzed this in the level at trial. And we have seen that patients, the same that in the previous study, we have seen that patients with baseline VEW that it's higher, we can obtain better response. Moving now to the in-plan application of the body surface mapping and conduction system pacing. We are working hard in this line in our center hospital clinic. We are using electrocardiographic imaging, the system that doesn't need imaging CT. And we screw into the septum. We obtain a map with baseline, with left underbranch block. And then we screw and we obtain different maps. We can see the change in the pattern of activation and the decrease in the left ventricular activation time. We can do a similar study with ultra-high frequency ECG. This is a video that Carol Kurila sent me. Ultra-high frequency ECG is so fast you can obtain a map and see the dis-synchrony index if it corrects or not with left underbranch space. The last points are optimization and follow-up. For optimization, in our center we are using fusion optimized intervals to optimize the ECG. We choose the better A-B interval to obtain the narrower QRS. And if we do that with conduction system pacing, what we obtain with left underbranch pacing, if we prolong the A-B interval, we can resynchronize better both ventricles. We can see here that right ventricle and left ventricle activate at the same time if we prolong the A-B interval. Carol Kurila has worked in optimization with ultra-high frequency ECG. If they program bipolar, anodal, left underbranch pacing, they have seen that it preserves better the physiological activation. We can see here the map. We have here the septum. With unipolar, we have like right underbranch block morphology and this is delayed activation. But with anodal capture, we can obtain more physiological activation. And last point is for follow-up. We can change in the clinic. We are not currently doing that, but maybe in the future it could be an option. If the patient comes to the clinic and we see that the ECG is not perfect or changes between the ECG now and the ECG obtained during implant, we can use electrocardiographic imaging or ultra-high frequency ECG to see the changes and maybe modify the programming. To conclude, body surface mapping and ultra-high frequency ECG will provide bit-by-bit activation map assessing left ventricular resynchronization and synchrony improvement. Electrocardiographic imaging showed left ventricular resynchronization and decrease in the synchrony with conduction system pacing. And both techniques will be useful for pre-implant in order to personalize therapy during implant and starting resynchronization for optimization, selecting the best configuration and for follow-up and checking if we have changes. And maybe I would like to conclude that maybe ECG, electrocardiographic imaging or ultra-high frequency ECG could be the ECGs of the 21st century. Thank you so much to all my colleagues in the lab in Barcelona and thank you so much to my mentors. Thank you very much and I would like to introduce the last presenter Taya Glosser from I don't know from... and it's talking on electrograms from troubleshooting and conduction systems. So, thank you for this invitation. I'm speaking here with a lot of esteemed colleagues, and I hope some of what I say will be understandable to the common man in the audience, common man and common woman. So first, just to give you a little history, 1956, Earl Bakken designed the first endocardial pacemaker. In 1987, Maurer got the first patent for biventricular pacing, and since that time, we've had tens of thousands of patients that have showed that biventricular pacing improves survival, reduces heart failure hospitalization, and does all of these amazing things. His bundle pacing was first described in 2000 by Deshmukh, and it was FDA-approved just in 2018, and left bundle area pacing was first described in 2017, and it just got FDA approval last year, just to give you some sense of how quickly this field is moving. So this is the ideal scenario for his bundle pacing. You know, we see a HV, and we see a pace to V that has the same duration as the HV, and the QRS complex looks identical. And as we've learned, during threshold testing, we can determine selective or not selective his bundle pacing. You see non-selective his bundle pacing in the far left with a slurred upstroke, selective his bundle pacing his only with loss of the slur, and a pace to QRS interval that's flat, loss of capture, and then if you look at the intrinsic QRS, it's completely different. And this is a cute slide showing his bundle capture patterns in patients with an underlying left bundle. You can see in yellow, there's non-selective his bundle capture with correction of the QRS interval, but a slurred upstroke. Then you can see selective his bundle capture with reproduction of the patient's native left bundle and an isoelectric interval from the pace to the QRS, and you can actually see selective his bundle capture with a right bundle branch block configuration. And this is a slide Dr. Vijay Raman shared with me, looking at his bundle capture in a pacemaker. So here on the left side of the slide, when you're doing V pacing, you can see retrograde atrial activation, there's a VA interval that's short, and as you turn down the output and lose his bundle capture and only capturing myocardium, suddenly the VA interval prolongs, as does the super QRS, showing you that you now just have myocardial capture. And sometimes when you're doing his bundle pacing, 20% of the time, you may actually get right bundle branch pacing. And if you look at the bottom of the slide, you have non-selective his bundle capture on the left, and then as you put the lead in further, you may have non-selective right bundle branch capture, and you know that you've lost his bundle capture because you've prolonged the R-wave peak time from 83 to 94, and then as you screw the lead in a little further, you may have selective right bundle capture by, if you look in leads two and three, the dramatic change in the QRS interval, and now an isoelectric interval between the pacing artifact and the QRS complex. I would love to have a mouse, a pointer. I guess I don't get a pointer. Okay. So now we're going to talk about left bundle area pacing, which as everybody said, is a little more complicated to understand. This is an example of... Can I have a pointer? Oh, use this. Thank you. Thank you very much. Thank you. Okay. So when you start pacing on the right side of the septum in V1, you see this QS interval. As you continue to pace and you begin to get into the deep septum, the QRS complex changes. When you get over to the left side, you start to see an R-wave in V1, and when you're capturing the left bundle, you have an even sharper, larger R-wave in V1. And during this transition, pacing from the right side of the septum to the left side of the septum, you will see shortening in the R-wave peak time. This is a beautiful example of someone who did it perfectly. This image was shown to you earlier, but I think it is a beautiful image that is in the recent era guidelines on conduction system pacing, where you can see the main His bundle here and the time between a potential and a QRS interval, if you're lucky enough to see it. As you get to the left bundle, that interval gets a little bit shorter, and as you get more distal in the conduction system, the potential to QRS interval shortens even more. And if we look at the paced QRS morphology in leads 2 and 3, it will be different depending on which part of the conduction system you're capturing. If you're capturing the anterior fascicle, 2 and 3 will be positive. The left septal fascicle, 2 may be positive, 3 may be negative. And the left posterior fascicle, 2 and 3 can both be negative. So now if you look at that example, if you're pacing the left bundle branch block, it's the left bundle itself, you'll see a potential to QRS interval of about 30 milliseconds. And leads 2 and 3 can be positive in 2 and negative in 3. And as you go deeper down, more distal in the conduction system, if you're pacing the left anterior fascicle, 2 and 3 will be positive. The left septal fascicle, 2 will be positive, 3 will be negative. And it's actually okay to get a negative QRS in 2 and 3. It just means you're pacing the left posterior fascicle. And again, the potential to QRS interval will be even shorter as you're more distal in the conduction system. The other thing that I found interesting is this will hold pacing on the right side. If you start on the right side and your paced QRS is positive in 2 and negative in 3, when you go over to the left side to get your R wave, you'll have the same result. And the same is true if you start and your 2 and 3 are negative when you start pacing with your W and V1. By the time you get over to the left side, 2 and 3 will likely still be negative, meaning you're pacing the left posterior fascicle. We heard a lot about threshold testing. I'm not going to do too much detail here, but if you go from non-selective left bundle pacing to LV septal pacing, it's obvious that the R wave peak time will be longer because you're no longer capturing the left bundle branch block. You're just capturing the LV septum, and the R wave peak time will prolong. Going from non-selective left bundle pacing to selective left bundle pacing is more difficult to see, but you will have the same R wave peak time because you're capturing the left bundle in both instances, but you may have in V1 an RSR instead of just a QR, and a little bit of a deeper S wave in V6, which I think is subtle to see. I'm talking about EKGs. I think these fixation beats when you're screwing the lead across the septum and you see these right bundle beats, it's time to stop and test. It means you've gotten over to the left side, and I think they happen very frequently. These are just two examples of patients who had native right bundle branch blocks, left bundle pacing in patients with native right bundles. In this case, the patient had a left anterior hemiblock, and the lead was put into the left posterior fascicle, and the left anterior hemiblock persisted. And in this case, the patient did not have a left anterior hemiblock, and the lead was in the main left bundle not having the left anterior hemiblock. And this is just an example of mine. I just thought it was interesting to see native right bundle branch block with a left anterior hemiblock, and when you get over to the left side, you have positive R waves here. You lose the left anterior hemiblock, so the lead is in the main left bundle. So I also want to say that sometimes conduction system pacing can fail because patients may have diffuse disease in the conduction system, and the conduction time is long, and you may not be able to narrow the QRS like you want to. This is another example of pitfalls of LV-only pacing in a patient who has a biventricular ICD. So this is kind of cute. If you do unipolar LV-only pacing, and the RV lead is off, and it's only sensing, there is cross-channel blanking on the RV channel up to 120 milliseconds. But if you make a very short AV delay with your left bundle pacing, you'll result in more delayed retrograde right ventricular activation, and the RV sense on this lead that is off will fall outside of the blanking, and now you'll see it as a VF sense. And the problem with this is it will then delay the subsequent atrial pacing and slow your atrial pacing rate. So there's a conundrum of programming the AV delay short enough to allow left bundle capture, but long enough to allow some fusion with the native right ventricle so you don't have this happen. And one way to fix this would be to just turn on RV pacing at 80 to 100 milliseconds later than your left bundle, and then it will be pacing, but it won't actually be capturing. And another example of this programming LV-only pacing in a bivy ICD that Dr. Vijay Raman shared with me, in this case, the patient suddenly lost capture management. Suddenly there was no longer any capture management on this LV-only lead, and when the patient came into clinic, you can see an A sense, V pace, failure to capture, and a sense of the native QRS. And if you look at the intracardiac electrograms, the device electrograms, there's A sense by V pace, but it's failing to capture, and then, again, fibrillation sense, which again will delay your timing, and this was evidence that the LV pacing lead had dislodged. Another tip, simple tip, that I think is clear to everybody, if you think about, you know, I'm doing left bundle pacing, my nurses hand me the CKG and say, here's your paced EKG, isn't it beautiful? And I say, no, that, you know, it's paced, but it's not actually left bundle capture, and sometimes you have to program the patient VVI to see not a fusion complex, but an actually paced complex to confirm left bundle capture. So, in conclusion, I'm just going to make a plug for this randomized clinical trial that is going to start shortly. Cardiac resynchronization therapy using His left bundle pacing versus by V pacing in patients who have heart failure with EF less than 50, wide QRS, and anticipated pacing more than 40%. It's going to be funded by PCORI and sponsored by the Baylor College of Medicine, and it will actually evaluate His or left bundle pacing compared to by V pacing on quality of life, exercise capacity, hospitalization for heart failure, and mortality. And I think it will be very, very exciting for us all to learn if left bundle branch or area pacing is equivalent or maybe even better than by V pacing. And I thank you for your attention. So we have time only for one short question from the audience, if there is one. Gregorio, a short question, please. Dr. Zastrzewski, sometimes the notch, the QRS notch or the slur, is it diagnosed that we don't stimulate the His bundle? Sometimes I believe we do stimulate, but it depends. If we have an ischemic cardiomyopathy, we will have some slurs. As we have said in our paper about the diagnostic value of notches in left lateral leads during His bundle pacing, it indicates either loss of capture or capture, but problem on the left side. So you can capture the His bundle, but if there is, as you mentioned, scar or, for example, not corrected left bundle branch block, you will still have a notch. So the notch tells you that you may lose the His or have no His bundle pacing, or you pace the His, but it doesn't do a good job on the left side because there is disease conduction system on the left side. So in both ways, it will inform you that your pacing is not exactly physiological. Who's going to answer? Usually in left-bottom-batch pacing, we usually leave the pacing modality in bipolar. So depending upon the requirement of anodal pacing, you can program the output. So you don't have to leave it in unipolar pacing. Suppose if you are doing a pacing for an underlying right-bottom-batch block morphology, you may have to utilize the anode, so you have to program it probably at 2.5 or 3 volt plus to capture the anode also. But if you don't want to capture the anode, you can always program it at an output of less than 2 volt to avoid the anodal capture. And regarding the second part you asked, second question? Yeah, so at a clinically-programmed output, all left-bottom-batch pacing is almost always non-selective because selective capture occurs at an output of less than 0.5 volt, which is not the clinically-programmed output. So we tend to pace at 1.5 plus output, so which is always going to be non-selective. So in left bundle, it is almost always at all clinically-programmed output, it's going to be a non-selective capture. Okay, I think we can conclude, and just a few words to summarize. I think that the main message is that we all have to go back to electrograms and look better for all the modification. So the best tool is electrograms.

electrograms

diagnosing

troubleshooting

conduction system pacing

his bundle pacing

ECG patterns

selective capture

non-selective capture

QRS morphology

R-wave peak time