Hello everyone, thank you for joining us for the very first episode of LEAD Podcast. This is the latest production of the Digital Education Committee of the Heart Rhythm Society. This is a bite-sized journal club where we will be discussing the latest and greatest articles in the field of electrophysiology. Today we'll be discussing left bundle branch area pacing versus biventricular pacing as an initial implant strategy for cardiac resynchronization therapy results from the Synchrony Collaborative Group. This article has been presented today at the Heart Rhythm Society meeting 2023 and will be published in JAC-EP. My name is Dinesh Kaila. I'm a clinical cardiac electrophysiologist at the Mayo Clinic. I'll be hosting this first episode along with my co-host, Dr. Jason Jacobson. He's the Director of Complex Arrhythmia Program at Westchester Medical Center in New York. And today, my guests are Dr. Roderick Tung. Dr. Roderick Tung serves as the Division Chief of Cardiology and Director of Cardiovascular Clinical Research at the University of Arizona College of Medicine and Banner University Medical Center Phoenix. Along with, we are joined by Professor Rajiv Pathak. He's the Lead of Cardiac Electrophysiology at the Canberra Hospital in Australia. Welcome. Thank you. Good to be here. So, I'll just give a brief synopsis of this article. So, this was a prospective non-randomized registry and what was done is patients needing cardiac resynchronization therapy, left bundle branch block, EF less than 35 percent or anticipated RV pacing of greater than 40 percent. Patients were, the initial implant strategy in these patients for resynchronization was either left bundle area pacing versus biventricular pacing with traditional corny sinus LV lead. This was at the discretion of the implanter. And this was a multi-center prospective non-randomized registry. The primary endpoint was composite of heart failure hospitalization and all cause mortality. A total of 371 patients were enrolled, 128 in the left bundle branch area pacing group and 243 in the biventricular pacing area group. Initial success was statistically significant, higher in the biventricular pacing group, 94.7 percent versus 84.4 percent in the left bundle branch area group. And the primary endpoint favored in the left bundle branch area pacing group, 24.2 percent versus 42.4 percent in the biventricular pacing group. And this was predominantly driven by heart failure hospitalizations, 22.6 percent in the left bundle branch area pacing group versus 39.5 percent in the biventricular pacing group. The fluoroscopy time, the procedural time were shorter in the left bundle branch area pacing group versus biventricular pacing group. And the QRS duration was 123.7 millisecond in the left bundle branch area pacing group versus biventricular pacing group at 149.3 milliseconds. Long-term complications similar in both groups, 9.4 percent in left bundle branch area pacing group versus 15.2 percent in the biventricular pacing group. And the authors conclude that left bundle branch area pacing may be used as an initial implant strategy for cardiac resynchronization therapy. So I'll hand the mic over to our guest. Well, I think that clearly there's so much interest in conduction system pacing right now and you're seeing very small single center go to larger multi-center. But still this is observational. It's non-randomized. It's comparative. But you can even see there's an imbalance and it's coming from multiple centers. But really good to see some data come not from Pugil, Vijay Raman's groups, et cetera. Because sometimes you have to think about generalizability. And a lot of this is coming, led by Jorge Romero from South America as well, which is great to see the global uptick of this. I think what's really remarkable about this, number one, is that these are hard endpoints. We were always trying to look at ejection fraction of improvement of greater than five percent. And all that echo stuff is great because we like to see that. But in the end, it's got to be heart failure and mortality, which this was really driven by heart failure. Number two, the procedures were 30 minutes shorter. So if you can save a half an hour, why wouldn't you do that even for something equivalent? But this is actually, looks like it's superior. But I think this is as far as you can really go with something that is non-randomized. I think the field is clearly there. And then today, you know, we see Pugil, Vijay Raman get up and show almost, you know, over 2,000 cases of which they found the exact same trend, which is really nice on the same day to see something similar. I agree completely. I think the more observational data which we continue to get, the time has come when we should look into having a proper RCT, which I think is coming. And these all results are very exciting. I agree that improvement in LV function, diastolic systolic volumes are all something which we all like. But hard out points, hospitalization, reduction in health care utilization, improvement in quality of life are the meaningful results which will help. And these studies which give us indication, gives us confidence that we should try to start looking at that in a randomized fashion. I agree fluoroscopy time reduction, procedural time reduction, and especially more 30 minutes, it's quite a significant amount of time. And for operators who are doing fluoroscopy time being less with these, of course, helps us and these are good results from the left model perspective. The major limitations of anything that's not randomized, you have no crossovers in the study. Why? Because they weren't randomized. And that's really what has been the crutch for a lot of this. And the real question also, which would be great to see more of, is what type of left bundles these are. We're understanding that there's different left bundle phenotypes. And not everything that says block is actually blocked physiologically. There's definitely IVCDs in here too. So I think that being able to understand even some subgroups when we get to big randomized trials will be really important. What's also noted is that the LVAT is not that short. It exceeds 85, which is kind of what Weijun Wang's been teaching us to say. That's left conduction system capture. And the interpeak delta is typically over 44 by America Shremsky and they're not all over 44, which means that not all of these are actually correcting the bundle with conduction system bypass. It might just be LV septal pacing. But then that's promising. Because if you don't get it perfect, there's still an overall potential benefit for LV septal. Exactly right. I mean, they have looked into that data. They have said that where they managed to get a perfect left bundle capture and where they have not. And they found that the effect size and the results were much better in the patient's subset group where they managed to find it, which is again encouraging. But I agree with you. Even in the patient's subset group where they did not get a perfect results, the results were still better than by V. I guess that is a very encouraging result from what we can see here. I'd like to maybe add a note of caution here. I think these data are very promising for a technique that really is taking off or so it seems. And certainly this population is good population. It's a heart failure population, a population that standard LV lead placement has been less than ideal for so long. However, I think we know this is being applied to many different types of patients without any significant conduction disease, maybe even just sinus node dysfunction. My question long term is what is the durability of leads in this position that aren't designed to have the stresses of that hinge point? What will it be like to extract these leads in the future? And there's also emerging data regarding tricuspid regurgitation, the closer you are to the base of the septum. I think all these potential complicated issues need to be considered and maybe give us a little bit of pause without a randomized study with hard end points and a very extensive follow up. And we have been doing this now for what, seven, six, seven years. There should be some data that we can start to see regarding durability of leads and experience with extraction. I completely agree and the long term data will obviously tell the story. You have the labeling now for Medtronic at least with 3830 with a lot of flex and hinge point testing, et cetera, that shows that this is probably okay and some of the initial data with extraction seems like you pull and it comes out. So that's the good news, but we obviously won't know until we really know. The only other thing to talk about is obviously recall and observation bias, et cetera, where these truly consecutive patients, you leave one or two of these out and everything changes. So that's always the challenge, which is why when you really go to randomization, everyone's accounted for. Yeah. Whereas one here, if there's five that are not accounted for, two, it can really change the calculus of all this. I completely agree. These are inherent bias, what we always get that in observational study. Rod, while you're here, I'll also ask, there are some European studies showing loss of R wave in V1 and loss of pure left bundle capture. These are heart failure patients. They do get RV dilatation, LV dilatation. What is your personal experience? Have you seen those loss of R wave in V1? With over time, you're saying, you're saying as if we're losing correction. Well, again, I'm not even sure a lot of these are corrected in the first place. It all might be LV septal. Sometimes even when we go from our recording system to another EKG or even EKG to EKG, there's some slight changes. It also, we need a better understanding of how things are programmed. The one thing that I would put a plea to is, what are the AV delays? Are these anodal capture? Are these unipolar? Are they bipolar? People have built a whole career on AV optimization and we don't talk about AV optimization here at all. How much of this is still fusing with the intrinsic right bundle versus not? How much of this is anodal? Those would be reasons why you could lose R wave. In the end, it's too complex to look at every single one of these factors. In the end, we got to put a lead that it could consistently capture and see if overall it's better. Think about the complexities of a BIV, which vein, what the output is there, the relationship to the scar, to it, all of it, but in the end, you have to show that in a big series of patients that get CRT, there was mortality benefit, which we did. I think those are great things for us to think about, but the field is moving exactly where we were thinking it was. It's studies like this that are critical to continue adding to all of the evidence here, and then we can talk about what level of evidence with the new conduction system pacing guidelines. I was going to ask you about this. Recently, they endorsed it as 2B. Do you think this changes for you or for the listeners? Would you have ranked it higher with this evidence coming out and the other late breaking in? Until we have RCTs, it will become very difficult. Large observational data are useful. Once you start having smaller randomized studies and larger randomized studies, that's when I think from 2B to 2N and class 1 indication will most certainly need a good randomized study for us to be able to use that. I mean, I guess technically 2B historically is risk and benefit are comparable. You just can't cause harm, so I think that's a very safe play. If you look at all the non-randomized observational, it looks like it may be superior, so I think that it might be too conservative. I think a 2A is reasonable, but clearly the level of evidence is not high. We had the left bundle re-sync that came out last year, but that was 40 to 50 patients, so no question the level of evidence is probably a C, but we're going to get there with left versus left and other studies, but I think 2A is maybe physiologically warranted, but again, are we really capturing the conduction system? I'm not certain, but I'm comfortable with the 2B, but I think that's a little bit on the conservative end for me. No, I agree with you, and there are so many other factors, as Rod was talking about. If patient has AV conduction, if patient has no AV conduction, how you're programming those, and there are so many different factors when you look into and to make sure that you are truly capturing left bundle. I was going to ask you, STLT2 inhibitors, all the CRT trials were before that. Now with the new conduction system basing trials that we are going to do, do you think that should be accounted for, and is it going to be harder for us to achieve a mortality benefit or looking at it in these trials because of STLT2, or the numbers are going to be much larger that will be needed? I mean, the point is well taken, of course, but we are talking here after optimal heart failure medication, when these patients are offered device, most of the physicians will include those drugs as part of it, so going forward, the new trials, which will be randomized trials, I think they will, most of these patients, if they tolerate these medications, they will be on the STLT2 inhibitor drugs as well, and I guess that's when we will see whether we still have the benefit. My personal feeling is yes, we will have that, because it doesn't matter once we have decided they have left bundle branch block and have ejection fraction, which is still poor, and resynchronization, if better with left bundle lead, is going to still show a good effect. And I think you're always going to see this, where we really fail with four pillars of GDMT. Our heart failure colleagues are always going to say it's still not adequate, we're looking at 26% and 33% of ARNI, so it's not even an STLT2 inhibitor, but we know in the real world the uptake is also quite low as well, so one could always make the argument there could be better medical optimization in these patients, and then we'll never really know what happens with really optimal background therapy until we really know. Yeah, I agree. One of the questions that my colleague raised here was, like, left bundle branch area pacing 2017 is roughly the time when it came around. Are we moving too fast with this, like that we are talking about that as an initial implant strategy, we don't have the randomized data, and then how to talk to the patient, what is the message from this paper and the experts in the field to an implanter out there who is working in a community hospital in the country or out back, for example, in Australia, what should he do and go from there? Out back in Australia, there are a few centers where we are doing that as a first line, we have just had our experience which we have put into this conference where we have 50 patients with a left bundle CRT which we did it, and we have had similar results, so I'm very excited with what we are finding. I completely agree, the center has to be experienced, to be prepared, to have discussed with the patient, be comfortable, and the operator himself has to be comfortable, then only he should offer it rather than just a new next exciting thing. I think the majority of centers are still using Bi-V as first line because it's truly evidence-based, but feeling more comfortable that as a bailout, and maybe the threshold for bailout is lower, rather than spending three hours and then bailing out, also we used to dump these leads into the AIV, or the MCV, and I think that is a good message to say, instead of accepting what we know is suboptimal, then maybe have a lower threshold to bail out and go to this, and I think that's what readers should take from this, is to not do this as first line for every patient, because again, you don't know how much the LVAD needs to be, or when is it good enough, and measuring the QRS widths are notoriously challenging here, I mean they had an average QRS width of over 120, but that's typically what it is if you measure from the stem, but everyone measures that differently as well, so it's hard to know what the electrical endpoint really should be for this. I agree with you, those are the things, I think bailout strategy, we have enough evidence that we should do first line, we should be cautious and slowly adopting these things. I guess just one last point is that, you know, morphology of the QRS is certainly kind of a moving target as you alluded to, is anodal capture a bad thing? You're maybe resynchronizing better, and have you seen additional right bundle capture from anodal capture, and is that something we should be aiming for? Well, there's been two small studies about anodal capture, and they go in different directions. Zach, when it's looked at this with blood pressure, and shown that it might be a little bit better, and then there's been other ones that show it might be a little bit worse in terms of the synchrony that was Karol Carillo, so it's not clear if anodal is better or worse. It seems to me like it's also non-physiologic, because it's probably, you know, the Chinese called it bilateral bundle branch capture, which is probably too ambitious to say that I'm now on the right and left at exactly the same time. It was probably anodal capture the whole time. So I don't think we know, we've actually been doing some strain imaging acutely right after the implant, we bring the echo in, and we'll do anodal versus not. And in an unpublished experience that's still very small, anodal actually seems like it's got worse parameters for strain. So the jury's still out, and we'll see. But programming is something we need to focus on as well. I agree, absolutely. The programming is something which we have so much experience with by we, and little experience with this. As the time goes on, that'll be something we need to focus on. Thank you. I would like to congratulate Dr. Diaz and Dr. Romero for their excellent work, and this is very helpful in this area. And I would like to thank Dr. Patek and Dr. Tang for joining us for this very first episode of The LEAD Podcast. Thank you. Thank you.

LEAD Podcast

study

left bundle branch area pacing

biventricular pacing

cardiac resynchronization therapy