Hi, I'm Pugal Vijayaraman from Geisinger Health System. Welcome to Core Concepts in Electrophysiology. The topic for today is cardiac conduction system pacing theory and practice. These are my disclosures. Performing conduction system pacing, it's quite important to understand the anatomy of the His-Pukinje conduction system. If you look at the His bundle, starting from the AV nodal region, and it transits across the atrial side of the membranous septum, and then crosses into the ventricular aspect of the membranous septum. So both portions are available for access for His bundle pacing. And histologically, you can find that a good portion of the His bundle is on the atrial side while the tricuspid valve, and that's in the main membranous part of the interventricular atrial septal region. It's His bundle pacing in a patient with AV nodal block is fairly easy to understand. As you can see here, a patient with complete heart block with his escape rhythm preceding every QRS activation and pacing from the His bundle can restore normal conduction below the His-Pukinje conduction system and results in perfectly normal looking QRS. On the other hand, if you look at His bundle pacing in 2 to 1 AV block in the setting of HV block, as shown in this intracardiac electrogram from the His bundle pacing, which shows a His electrogram followed by no ventricular actuation, and then a His followed by ventricular actuation, suggesting 2 to 1 HV block. What is interesting is pacing at this side can restore conduction through the distal previously blocked His bundle. And that's because majority of the conduction disease in HV block occurs in the main His bundle. And often these lesions are fairly discreet, and you can bypass or capture the distal conduction system by pacing at the side of conduction block or even slightly proximal to it. Well, ideally, pacing a distant to the side of block would be much superior in terms of threshold and long-term stability. While performing His bundle pacing, it's important to understand the various anatomic variations of the His bundle. There are three different variations that's being described based on autopsy studies. And in the first one, we can see type 1, the His bundle courses along the inferior aspect of the membrane, a septum, and it's covered by a very thin layer of myocardial fibers. And clinically equivalent of this, you may be seeing about 47% of the patients. In these patients, often there is His bundle and associated RV myocardial capture, but usually with minimal fusion as the amount of myocardial tissue around these His bundles are very small. And often you find non-selective to selective His bundle pacing during threshold testing. Here's an example of a patient. You can see that pacing at about 1 volt, you can see small degree of ventricular fusion with the delta waves in lead B1, B2. And as you come down on the output, you can see the loss of ventricular capture, and this is selective capture. And the degree of fusion is fairly small, does not increase the cure's duration significantly, as you can see in this example. On the other hand, in type 2, the His bundles entirely encased in a thick sheet of ventricular myocardium. In these patients, this is seen about 30% of the patients. And often you will find in addition to His bundle injury current, you would see a significant myocardial injury current. So it's almost impossible to get selective myocardial capture or selective His bundle capture in these patients. And the amount of ventricular fusion may also be significant. The delta waves may be more prominent in these patients. And often the RV capture threshold is lower than the His bundle capture threshold. And here's an example of a patient with non-selective His bundle pacing, still see the delta fairly prominent. And as you come down on the output, you see loss of His bundle capture and fairly widened cure's duration. And you can see the degree of fusion in these patients. On the type 3 is the least common type, which is the most desired type where the His bundle is often called naked His bundle, where the entire His bundle is seen on the membrane septum, fairly superficial, not much surrounding myocardial fibers. So in these patients, this is fairly sub-endocardial. And often you find large His electrogram, good His bundle injury current, and you often get selective His bundle capture in these patients. Here's an example of a patient, simultaneous display of the His electrograms. And even with a little bit of lead fixation, you can see a huge current of injury on the His bundle. And pacing here can result in selective capture of the His bundle alone without any myocardial captures. You can see discrete myocardial content in the His bundle electrogram. In many of the features of His bundle pacing has been somewhat standardized based on this recent recommendation from a multi-center His bundle pacing collaborating working group for standardization of definitions. And this is somewhat a busy slide, but that describes various scenarios and various types of His bundle capture with or without correction of underlying bundle branch blocks or His book injection disease and when to call it selective His bundle pacing and when to call non-selective His bundle pacing. We'll go over some of these examples. And here's His bundle pacing in patients with normal His book injury conduction. What criteria do you look for? So when you have normal book injury conduction, and if you have selective His bundle pacing, in this situation, often the stimulus to QRS duration will be same as His electrogram to QRS duration with an isoelectric interval from the stimulus to QRS onset. And during His bundle pacing, the local ventricular electrogram, since it's not being captured, should be discrete with the stimulus to ventricular interval, similar to the His to local ventricular interval. And often the pace QRS duration will be identical to the native QRS duration. And morphology, since this is primarily selective His bundle pacing, you only have one capture threshold. It will be only His bundle capture. So here's an example for patient. It's a faster sweep speed at 100 millimeters per second. You have an HV interval of 40 milliseconds, QRS duration of 90 milliseconds. And during pacing, the stimulus to QRS onset is about 40 milliseconds. And the local electrogram is discrete, as you can see by the arrow markers. And the pace QRS is identical to the baseline QRS morphology and duration. When you have non-selective His bundle pacing, the stimulus to QRS is often short, usually close to zero, because the myocardial capture starts inscribing the ventricular activation immediately after the pacing stimulus. And most often you do not have an isoelectric interval, but sometimes you may have a pseudo-isoelectric interval, because the delta waves are so low amplitude, as in the case of type 1 His bundle anatomical variation. And because the non-selective His bundle pacing, in addition to His bundle capture, there is capture of the local ventricular electrogram, you will not see a discrete ventricular signal following the stimulus artifact. And the pace QRS duration is usually wider than the native QRS because of the duration of ventricular fusion occurring during the His to ventricular conduction interval. And you will find two distinct capture threshold, one for His bundle capture, another for RV capture. So here's an example of a patient with baseline normal QRS, 80 milliseconds, and at 1.2 volts, there is non-selective His bundle pacing, and the QRS duration is 120 milliseconds, inclusive of both 80 millisecond QRS and a 40 millisecond isoelectric HV interval time, during which there is myocardial fusion. And at a lower output, you get only RV capture, and there's loss of His bundle capture. So His capture threshold is 1.2 volts, and ventricular capture is less than 1 volt. And you can see that the stimulus to atrial activation time is somewhat prolonged with the loss of His bundle capture. And here is another form of non-selective His bundle pacing, where the transition happens from non-selective His capture to selective His capture. In this case, the myocardial threshold is 1.5 volts, and His bundle capture threshold is less than 1 volt. And because His is captured during both situations, the stimulus to atrial timing interval is unchanged at 150 milliseconds. How about His bundle pacing in patients with underlying conduction disease, and with correction of underlying conduction disease? So in a selective His bundle pacing, with correction of His bookend conduction disease, the stimulus to QRS interval will be less than the His to QRS interval, with an isoelectric interval. It can be identical to His to QRS interval, depending on how much correction of the HV timing happens. The ventricular electrogram is usually discrete. The pace QRS duration is generally narrower than the native QRS, because of the correction of underlying bundle branch block. And you often have two distinct capture thresholds, selective His capture, with bundle branch block correction, and then His bundle pacing without bundle branch block correction. So you can see in this example, patient with a wide right bundle branch block of 180 milliseconds, pacing at 1.2 volts, you see selective capture with complete correction of the underlying right bundle branch block, even though there is no shortening of the stimulus to ventricular activation times. And at the lower output, there is capture of the His bundle without correction of the right bundle branch block. So you would want to know the threshold required to correct the right bundle branch block. And how about here in a patient with a wide left bundle branch block? You can see at 2 volts, there is selective capture, no fusion, no delta waves, and normalization of the QRS. And at lower than 2 volts, there is still capture of the His bundle with underlying left bundle branch block without correction of the left bundle branch block. So your ideal threshold to pace this patient would be above the threshold required to correct the left bundle branch block. Here's a patient with 2 to 1 HV block, and here there is selective capture with complete correction of the underlying 2 to 1 HV block. And how about non-selective His bundle pacing with correction of His Purkinje conduction disease? The clinical criteria are similar to what we saw with normal His Purkinje conduction, but the most important thing is that there will be three distinct capture thresholds, one for His bundle pacing with bundle branch block correction, and then without correction and with RV capture threshold. So these, the last two can vary depending on how close the myocardial stimulation is. Here's an example of a patient with the right bundle branch block. At 1.2 volts, you see complete correction and a non-selective capture with the fusion and a delta wave, as you can see in V1 and V2. At a lower output, there is loss of right bundle capture, but still there is myocardial capture. So you have both left bundle and RV capture, so loss of right bundle branch block correction. And at a lower output, you lose the remaining His capture with left bundle stimulation, and there's only RV myocardial capture with prolongation of retrograde conduction times. How about a patient, another patient with right bundle branch block, higher output, you have complete correction with fusion, and lower output, there is selective capture with correction, and it is still lower output selective capture of the right bundle without capture of the left bundle fascicle. So there can be different combination, different morphological changes. It's important to understand what is most essential for that particular patient and program above that output required to correct the underlying conduction disease. There's another patient with left bundle branch block, a wide QRS with multiple fractionated QRS suggesting mixed conduction disease. At high output, there is non-selective hispital pacing with correction of the left bundle branch block because overall QRS duration is narrow, and the PKLV activation time is somewhat shortened here. And at a lower output, there is selective capture with correction of the left bundle branch block, but only partial correction. There is still certain degree of IVCD in this patient to 125 milliseconds. And then at a lower output, there is even further loss of capture of the left bundle correction, and there is underlying left bundle branch block displayed. So there's still three different outputs in this patient. We want to program above the threshold required to correct the left bundle branch block in this patient. There's another patient with complete HV block. You can see AH and ventricular activation is dissociated with the underlying right bundle branch block. The escape rhythm seems to arise from the left bundle fascicles. And here pacing at 1.5 volts, you have complete correction of both the right bundle branch block and HV block. And subsequently, you have selective capture with the underlying right bundle branch block pattern. So here you have loss of RV capture and right bundle capture. So it's important to program above 1.5 volts in this patient to achieve complete capture of the distal Hispokiniae conduction system. And of course, you can also have Hispokiniae pacing in patient with more distal conduction disease. So you don't correct any of the underlying either bundle branch block or IVCT. In these patients, you can either have non-selective Hispokiniae pacing without correction, or you can even have selective Hispokiniae pacing without correction. And you will see different capture threshold as outlined before. Here's in this patient, the example of right bundle branch block. At higher output, you have partial correction of right bundle branch block more from RV fusion. At lower output, there is selective capture without correction of the right bundle branch block. In this patient, we combined Hispokiniae pacing with RV pacing in a sequential fashion to achieve the perfect narrowing of the QRS. So it's important to understand the fundamentals of assessing Hispokiniae pacing using ECG. And during pacing threshold, it's quite helpful to have a 12-lead ECG to assess the morphology changes. So in patient with selective Hispokiniae pacing, there will be only one transition. In non-selective Hispokiniae pacing, there is usually two transitions. And then in patients with one branch block, you can have up to three transitions as we talked about. So changing pacing or using a 12 lead to assess various capture is important. In some patients, both his bundle and RV capture threshold can be identical in these patients. Program stimulation can be helpful. And then sometime looking at the various ECG features to confirm his bundle capture is also helpful. Just using QRS duration alone is not very helpful. As you can see, when you have RV capture only compared to non-selective capture, you can have a significant overlap in this distribution codes of QRS duration. Similarly, assessing peak R wave times in V6 can also be significantly overlapping. But certain parameters are very helpful in confirming conduction system capture in these patients. So for example, compared to non-selective his bundle pacing versus RV pacing, you can see the lead one R waves are fairly pointed, rapid peaking compared to plateaued appearance or notching of lead V1 or V4 with the slur. And similarly, V5, you can see these slurs when there is correction of underlying bundle branch block versus when there is no correction. And similarly, peak R wave times are generally longer than a hundred milliseconds in patients with RV only capture. In difficult cases, you can do program stimulation. You can see non-selective his capture turning into RV only capture as you become more premature. This is exploiting the differences in refractive period of tissues, both myocardium and his bundle. His bundle generally tends to have a longer refractive period compared to myocardial tissue. Sometimes using rapid pacing, in this case, you can see pacing at less than 330 millisecond result in RV only capture and loss of his bundle capture is a quick way to determine if the patient has non-selective capture or RV only capture. While his bundle pacing is quite interesting and also quite effective in reproducing normal conduction, there are quite a few challenges. Because of its anatomical location of the his bundle in a fibrous tissue, high pacing thresholds are not uncommon in about 10% of patients. More importantly, in another 10% of patients, there could be unpredictable threshold increase during follow-up. And many of these patients may end up needing lead revisions in about 5% to 7%. And because of its unique location at the AV annulus, often the R waves are fairly small, which can lead to ventricular undersensing or atrial oversensing. So this has led to development of more distal conduction system pacing called the left bundle branch area pacing. As you can see, the LV septal surface, the left bundle branch is a much wider well distributed conduction tissue compared to the very narrow his bundle, as you can see on the RV septal surface. And sometimes these anatomical distribution can be quite variable, but you can still get into the conduction system fairly early in the LV septal myocardium compared to his bundle. And this has led to a recent surge in the utilization of left bundle branch area pacing compared to his bundle pacing, because his bundle pacing has a narrow target, while left bundle pacing has a much wider landing zone. And it is surrounded by extensive myocardium. So the threshold increase or issues associated with may be less common. While his bundle pacing may be physiologically superior, left bundle branch pacing still provides a fair amount of left ventricular synchrony. And it can be used in any form of conduction disease, especially in patient intraneural blocks or bundle branch block. Left bundle branch pacing may be much better in achieving better thresholds. And the lead division rates overall being very, very small in the left bundle branch area pacing groups. So this option and possibility of achieving stable load threshold and ability to pace beyond the site block, especially in intraneural blocks and in patient left bundle branch block, with no concerns about over-sensing, this has become a go-to therapy in many patients. And because of the wider target zone, it's a lot easier to get into the conduction system compared to the more challenging his bundle pacing location. And before we go into the details of left bundle branch pacing, I want to clarify some of the definitions as this area is still in the development, and there's a lot of misconceptions. And we'll try to clarify that as best as possible. So when we have clear evidence for capture of the left bundle or its branches, it's considered as left bundle branch pacing. And invariably, all of these patients will have LV septal capture. So the question of selective and non-selective left bundle branch capture does not come into clinical practice because almost all of these patients have so-called non-selective left bundle branch pacing, capturing the left bundle and the surrounding LV septal myocardium. In many patients, it may be difficult or impossible to prove there is clear left bundle branch conduction system capture. In these patients, we call them LV septal pacing, and there is no evidence for left bundle branch capture. But if you want to combine the whole as a group, both left bundle branch pacing and LV septal pacing, then you call it left bundle branch area pacing. These leads place in the deep septal LV septal myocardial region and have access to the left bundle branch either directly or indirectly. So let's talk about some of the criteria for left bundle branch pacing versus LV septal pacing. So again, here, the challenges are to identify these findings are easier in patients without the left bundle branch block compared to patients with left bundle branch block. So we will tackle them separately. So for left bundle branch pacing, the lead needs to be located at the proximal in the deep septal region, usually about one to three centimeters from the distal his bundle. This can be confirmed on fluoroscopy with contrast injection or echocardiography or CT imaging. About practical purposes, fluoroscopy contrast injection can confirm this at the time of implant. And most commonly, there should be, when you have left bundle branch pacing, unipolar paced QRS morphology should have a right bundle branch block or a right bundle branch conduction delay pattern, one which includes a QR pattern or RSR pattern. Very rarely, you may see QS pattern with narrow complex with the lead located in the deep septal location. Should not exclude left bundle branch pacing, but for practical purposes, most often you have this right bundle branch block type morphology. And in patients who don't have left bundle branch block, you should be able to demonstrate left bundle branch potential. In addition to those features, you got to show at least one direct evidence for left bundle branch capture. So if you have capture of left bundle branch and the LV septal myocardium, you would usually see transition from non-selective to selective left bundle branch capture doing threshold testing, usually very close to near threshold outputs. And this is associated with the changes in QRS morphology, possibly changes in the appearance of discrete local electrogram without changes in R-wave peak times. If there is a transition from non-selective to selective, the R-wave peak time should remain same. There is a transition from non-selective to LV septal capture, usually the R-wave peak time will prolong because of delayed entrance into the conduction system. And during lead implantation, from a mid-septal to deep septal implantation, there should be an abrupt shortening of peak LV activation time once you have a right bundle branch block pattern. So at a higher output, you will have a shorter LV activation time on a lower output, again simulating the non-selective to septal transitions. And when you have a HISS electrogram to compare, you will see that when pacing from the left bundle branch, you should have a retrograde HISS potential within 35 milliseconds from pacing. And if you have any kind of mapping or other catheters in the LV septum, you should be able to demonstrate left conduction system potential within 25 milliseconds of your pacing artifact, suggesting that you're capturing the conduction system. In difficult cases, you can use program stimulation to confirm selective left bundle or LV septal capture, similar to our HISS bundle pacing program simulation techniques. And often we use arbitrary peak R-wave times, there is some evidence for this, we can talk about it in subsequent slides. So let's go through some of these criteria. This is the location of the left bundle branch pacing lead, about 1-3 centimeters from the HISS bundle region. And you can see during pacing from this site, you have a right bundle branch block morphology, especially unipolar pacing. And you see this left bundle branch potential is recordable. And you can see retrograde HISS potential that's within 35 milliseconds of the pacing artifact, suggesting early entrance into the conduction system. And simultaneously coinciding with proof of deep septal lead placement, as in this case, you can see on the left anterior oblique view, once the lead is placed, you can inject contrast, see how deep the lead is engaged in the system. In this case, there's at least 13 millimeters of the lead into the interventricular septum. So some of the concepts of non-selective left bundle pacing, there is no latency. As soon as you stimulate, there is ventricle activation starts. But when there is left bundle branch capture, the stimulus to PKL reactivation time is short. In this case, it's about 84 milliseconds. And at a lower output, you have only capture of the conduction system. And there is a stimulus to QRS latency and a discrete local electrogram, as you can see in this situation. And the stimulus to PKL reactivation time is still about 84 milliseconds, identical to your non-selective left bundle branch pacing. Here's another example of the left bundle potential in a patient with a normal QRS. Here, pacing during a mid-septal location, and you have a glimpse of a right bundle branch block morphology at a low output. But a high output that's more prominent, but you can see the stimulus to PKL reactivation time is significantly shortened. And the lead is advanced a little bit deeper. You have a more prominent left bundle potential with the current of injury. And now, at a very low output, you see a non-selective to selective transition with a short stimulus to PKL reactivation time. And a right bundle branch block pattern changes to a selective right bundle capture with an RSR factor in this situation. You can also see retrograde his capture with a short stimulus to his activation time. Conforming all of the findings that we talked about for left bundle branch capture. How do you conform left bundle branch pacing in patient left bundle branch block? This becomes a little more challenging with some empirical evidence have to be used. Apart from using the same criteria we talked about, the lead location being deep and proximal septal, and paced QRS morphology in lead B1, you may be able to demonstrate left bundle branch potential if you correct the underlying left bundle branch block with his bundle pacing. Now, restoring conduction in the left bundle, you should be able to see a potential. Or if the patient has native conduction, normalizing during a PVC or other times. While PVC is originating from the left bundle branch region would give a left bundle branch potential on the electrogram. In addition to that, you need to have the other traditional criteria for left bundle branch capture. So transition during threshold testing becomes important, proving that there is non-selective to selective or non-selective to LV septal capture. And then during implantation, the same abrupt shortening of the left ventricular activation time or peak R wave times and conforming distal potentials or with program stimulation. But in patients with left bundle branch block, especially in patients with cardiomyopathy, the peak LV activation time can be somewhat longer. We want it to be less than 90 milliseconds. Here's an example of a patient with a wide left bundle branch block. There may be sometimes the retrograde left bundle branch potentials in the left bundle electrograms. And with his bundle pacing, you can see demonstration of left bundle branch potentials just in conduction through the left bundle when there is correction of underlying left bundle branch block. And at low output, you can see transition from non-selective to selective capture with a short LV activation times, in this case measured with an LV coronary sinus lead. Here's another patient with left bundle branch block. When you pace his bundle and correct the left bundle branch block, you can see prominent left bundle branch potentials as you can see demonstrated in this patient. Or at other times, program stimulation can help show the difference between non-selective capture with myocardial only capture in this situation. The other way, this is more of a research interest. You can see that do a decouple of catheter placed along the LV septal surface. And during initial lead placement in this patient with normal QRS at baseline, you can see septal left conduction system activation with Purkinje potentials and the lead in the left bundle branch location demonstrating left bundle branch potential. And during initial lead implantation, during mid-septal lead placement, there is a semblance of right bundle branch block pattern, but at the only LV septal capture, there is no conduction system captures. You don't see any potential, but at a higher output, we can demonstrate some of these potentials. But as the lead is advanced deeper, both at high and low output, you see these potentials are present. This is what we are striving to get. Conduction system capture at any pacing output demonstrates non-selective and selective conduction step capture in this patient. And similarly, in a patient with left bundle branch block, you don't see a left bundle branch potential. But pacing at high output, you can see that as the QRS is narrowed, there is conduction system capture and demonstration of conduction system potentials in this patient. And same way at once in the final place, any output results in demonstration of conduction system potential in the patient on the left bundle branch block. And during his bundle pacing also, you can demonstrate the same conduction potential confirming that you have captured the conduction system or your lead is present at the level of the conduction system. So when you don't have any of the evidence for direct left bundle branch captures, many of the criteria that we talked about, in addition to deep septal lead placement, then you call it LV septal pacing only. And because of the challenges associated with confirming conduction system capture, we have several indirect criteria. So some of them are so-called physiology-based criteria using similar criteria as in his bundle pacing using left bundle potential to peak R wave times. Native enduring pacing, we can see the difference between left bundle capture versus the LV septal pacing. So here's a patient with a narrow complex native, the QRS onset to peak R wave time is 56 milliseconds. And during left bundle branch pacing, if you take the real QRS onset, excluding the initial phase, then time to peak activation time should be similar. When you don't have conduction stem capture, that timing gets prolonged here. And here's another case where potential to onset is about peak R wave time is 80 millisecond. And during stimulation, that stimulus to peak LV activation time is shorter compared to LV septal pacing, which gets longer than the potential to QRS duration. And same phenomena demonstrated in this situation, differentiating LV septal pacing versus non-selective LV branch pacing. So there's a good correlation between using the R wave peak times between native QRS and LV branch-based QRS. Similarly, with the left bundle potential to R wave peak times versus stimulus to R wave peak times during non-selective LV branch pacing. However, there is a certain degree of overlap. So if you want to use only peak wave R wave cut times, anything less than 74 milliseconds gives you a very high degree of specificity. In patient with left bundle branch block, anything less than 80 milliseconds will provide you some evidence that there is left bundle branch capture. This would be indirect of empirical criteria that you can use. However, in patient with left bundle branch block, we can become more specific. We can measure peak activation time during his pacing with correction of left bundle branch block. In this case, it's about 103 milliseconds. So left bundle branch being more distal in the conduction time, the stimulus to peak LV activation time should be shorter. In this case, it's about 78 milliseconds. While with LV septal pacing, it is same as his bundle pacing. So left bundle branch pacing being further in the conduction system gets quicker to the LV activation time. So results in shorter peak activation time. And here is another example of a patient with left bundle branch pacing to LV septal pacing. You can see his pacing gets a peak R-wave times of 96 milliseconds. However, during non-selective left bundle branch pacing, that's about 67 milliseconds. And with septal pacing, it's about 90 milliseconds. So we plotted this delta R-wave times between his bundle pacing and left bundle pacing. So in this case, you can see the overlap between LV septal pacing and his bundle pacing, but the difference between left bundle branch capture was LV septal pacing. There's a significant degree of overlap. However, when we use only the difference between his bundle pacing and left bundle branch area pacing, in order to have left bundle branch capture, that difference should be less than eight milliseconds to have high sensitivity and specificity. But if this difference is more than 10 milliseconds, that has about 100% specificity. So our goal is to identify the peak R-wave times during corrective his bundle pacing in a patient with left bundle branch block, and then shoot for a number that will be at least 10 milliseconds less than what you achieve with his bundle pacing. Now that we understood the various forms of left bundle branch pacing and his bundle pacing, where do we use conduction system pacing? The most common indication is in lieu of right ventricular pacing. And why should we do? Looking at a series of 333 patients, we're able to perform conduction system pacing in about 97% of these patients. And we were successful with his bundle pacing in 42% and left bundle branch area pacing in 55%. You're also able to identify where's the site of conduction disease, was intrahysion in 89% of the patients. Well, intrahysion was only seen in about 4% of patients. That's more distilled diffuse conduction disease. So majority of patients have conduction disease in the AV node or in the proximal his bundle region, in the main his bundle region. So all of these patients are amenable to conduction system pacing with either his bundle pacing or left bundle branch area pacing, especially in patient HV disease. In a clinical outcome study, we looked at two groups of patients, one set of patients in one hospital getting all his bundle pacing while the other group getting RV pacing. So we were successful in using his bundle pacing or early series for 92% of the patients. Well, RV pacing was performed predominantly in another hospital. During a mean follow-up of two years, we looked at the primary outcome of death, heart failure, hospitalization, or need for upgrade to biventricular pacing. These are all the bad outcomes associated with right ventricular pacing. We were able to demonstrate about 7% absolute reduction in the primary outcome. And this was about 29% relative risk reduction. And this difference was primarily in patients who needed more than 20% ventricular pacing with a 35% risk reduction for death, heart failure, hospitalization, or need for upgrade to bi-V pacing. And this difference also stems from primarily reduction in heart failure, hospitalization. There was no mortality benefit when you look at individual outpoints, although there was a trend towards improved mortality in these patients. In a similar series on a smaller subgroup of patients, with a five-year follow-up, with 80% success, we were able to demonstrate that the incidence of pacing-induced cardiomyopathy was very low in patients with his spinal pacing compared to 22% in patients with RV pacing. How about left spinal branch pacing? We also looked at the outcome differences between RV pacing versus left spinal branch pacing in a multi-center study. There were about 783 patients. These were not randomized. These were retrospective observational series. Looking at the same clinical outcome of death, heart failure, hospitalization, or upgrade to bi-V pacing was significantly reduced in those patients who underwent left spinal branch area pacing compared to RV pacing. And this was also associated with improvement in all-cause mortality and heart failure, hospitalization between the two groups. And all of these outcome differences came in patients with more than 20% ventricular pacing. So I think both his spinal and left spinal branch pacing is very helpful in preventing some of the RV pacing-induced adverse events. How about treating patients who have previously undergone right ventricular pacing and then develop pacing-induced cardiomyopathy? This was a multi-center series looking at 60 patients who had pacing-induced cardiomyopathy who were successful at this spinal pacing, 95% of them. The duration of RV pacing was about six to seven years in a mean. And these patients, a third of patients had infra-neural AV block. Despite having a long history of HV block, were able to still perform a spinal pacing and correct underlying conduction disease, narrow the QRS duration from 178 millisecond, 114 millisecond. More importantly, we were able to reverse the RV pacing-induced LV dysfunction from 34% to 48% during follow-up. It's an individual patient with EF improvement following conduction system pacing using his spinal pacing in this series. So based on some of these studies and other outcome of the 2018 Heart Rhythm Society guidelines on the evaluation and management of patient with cardiac conduction disease and bradycardia have recommended his spinal pacing or biventricular pacing in patient with greater than 40% need for ventricular pacing and an LV ejection fraction less than 50%. And ESC guidelines even recommend his spinal pacing as a first-line therapy when you undergo AV nodal ablation. And how about conduction system pacing indication in lieu of biventricular pacing? That's a whole group of patients where conduction system pacing holds a lot of promise. Some of the basic work on his spinal pacing comes from this elegant study that looked at patients with left spinal branch block and heart failure and gave his spinal pacing and biventricular pacing in the same patient for 18 patients looked at ECG imaging guided electrical activation and hemodynamic changes. They were able to demonstrate that in this group of patients, his spinal pacing achieved greater reduction in QRS duration on top of what could be achieved with biventricular pacing. With greater reduction in LV activation times and improvement in LV dyssynchrony index associated with greater hemodynamic improvement on top of what could be achieved with a well-placed biventricular pacing. And this was tested in a randomized clinical study. This is his alternate study. It's a small pilot study of 50 patients with left spinal branch block and with defined by STROSS criteria, half and half randomized to his CRT versus biV CRT. They were successful in correcting left spinal branch block in 96% of these patients, 24 or 25 patients, but because of high threshold, they were able to place their histate only in 72% and others crossed over to biV pacing. And one patient in the biV crossed over to his spinal pacing. So on intention to treat analysis, his spinal pacing resulted in similar improvement in QRS duration and LV ejection fraction in this group of patients. More importantly, on a per protocol analysis, those with his CRT result in greater LV ejection fraction improvement than biventricular pacing, suggesting that the conduction stem pacing has greater potential on top of what could be achieved with biventricular pacing. And because of the limitations of his spinal pacing, the higher thresholds and challenges in implanting the lead, the left spinal branch pacing offers a greater opportunity in this patient. Here's an example of a patient with a fairly wide left spinal branch block, deep location of the his spinal pacing lead, oh, sorry, left spinal branch pacing lead resulted in subsequent QRS narrowing significantly to 104 milliseconds. And we looked at this in a multi-center approach and retrospective study about 325 patients with indication for cardiac resynchronization therapy were successful in 85% of these patients till the early phase of the learning curve, challenges in treating patients with non-left spinal branch block. Nonetheless, the study showed that these patients achieved greater QRS duration reduction and more so in patients with left spinal branch block and greater functional class improvement, LV ejection fraction improvement and changes in LV and diastolic dimensions. This was recently tested in another study where they tried his spinal or left spinal branch pacing for CRT from a Spanish group, looking at 70 patients. There was a 76% success rate in performing conduction system pacing. The primary endpoint was looking at LV activation time. They demonstrated that using ECG imaging, they showed that conduction system pacing resulted in similar reduction in LV activation times and intention to treat versus protocol analysis showed even greater reduction in LV activation time with conduction system pacing. The similar improvement in LV ejection fraction and functional class was noted in this study. This was a non-inferiority design that met its endpoint. Another trial was left spinal branch pacing resync trial. Another pilot study looking at 40 patients with non-ischemic cardiomyopathy and complete left spinal branch block. One-to-one randomization. This study is more interesting because of higher success rate. They were successful in 90% of the patients and about 10% crossed over to bi-V pacing compared to 80% success in bi-V pacing with four crossover to left spinal branch pacing. They showed significant improvement in LV ejection fraction, but much greater improvement in LV ejection fraction with left spinal branch pacing. And they used untreatment analysis in this group of patients. And QRS duration reduction was also greater in this group of patients and greater degree of a hyper-response and LV remodeling measures were achieved with left spinal branch pacing compared to bi-ventricular pacing. We looked at clinical outcomes of conduction system pacing versus bi-ventricular pacing in a larger cohort of patient in a retrospective fashion in a multi-center study. Looking at 477 patients in whom 238 patient underwent bi-V pacing first versus CSB in 239 patients. And there was a great degree of crossover in this real world experience. And some of the bi-V failures were referred from multiple centers. So overall initial success somewhat lower in the bi-V pacing group. This could be falsely misinterpreted, but nonetheless, the final group of patient included 219 patients that underwent bi-ventricular pacing and 258 patients that underwent conduction system pacing. So the primary endpoint of death to heart failure hospitalization were analyzed during a two-year follow-up in this patient demonstrated that overall the primary endpoint of death to heart failure hospitalization was significantly reduced in those patients who had conduction system pacing. With a hazard ratio of 1.52 for bi-ventricular pacing with P value of 0.013. This outcome difference was even more magnified in patients with left frontal branch block with a hazard ratio of 2.1 and a P value of 0.006. The secondary outcome of heart failure hospitalization was also significantly reduced in patients, all patients with conduction system pacing compared to bi-ventricular pacing. Well, there was a trend towards a reduction in mortality to start reaching statistical significance. And more importantly, in patients with left frontal branch block, these outcome differences were even more prominent as seen here for the heart failure hospitalization. Looking at QRS duration reduction, this was much greater with conduction system pacing, more so in patients with left frontal branch block compared to bi-ventricular pacing. And this resulted in significantly greater improvement in LV ejection fraction, both in all comers as well as in those with left frontal branch block, suggesting that conduction system can maximize resynchronization therapy in patients with frontal branch block or other needs for cardiac resynchronization therapy. Nonetheless, the current evidence for efficacy of cardiac resynchronization therapy in heart failure patients with left frontal branch block suggests that bi-V pacing has overwhelming clinical data with randomized controlled clinical trials. But CSP, his or left frontal branch pacing has great potential, but that needs to be proved with randomized controlled clinical trials. In summary, I want to say that HSPOC-NG conduction system pacing should be considered in all patients who require ventricular pacing. And conduction system pacing can be considered as an alternative to traditional bi-ventricular pacing. And in those patients who fail bi-ventricular pacing who are non-responders to bi-ventricular pacing. Randomized clinical trials are necessary and clinical trials are necessary in both groups of patients with cardiac cardio and CRT indication. Thank you very much for your attention.

Cardiac conduction system pacing

His-Bundle Pacing

Left-Bundle Branch area pacing

QRS morphology

heart failure hospitalizations

left ventricular ejection fraction

conduction system disease

bundle branch block

pacing-induced cardiomyopathy