Good morning everyone, good morning, I'm Dr. Karim Ben Ali from Saint-Étienne University Hospital in France and I'm pleased with my co-chairman Dr. Richard Schilling from the Bath Heart Center to welcome you for this early morning session on updates on ablation strategies in percentage of ablation. And our first speaker will be Dr. Jonathan Coleman from Royal Melbourne Hospital to discuss about PVA. Richard Karim, thanks very much and good morning everybody. It says it's loading the presentation, there we go. So it's worthwhile really starting with a look at what the expert consensus statement on catheter ablation from 2024 late last year says and really about the only advice to do is electrical isolation of the pulmonary veins for persistent atrial fibrillation and that you need to achieve entrance block and beyond there we head into areas of uncertainty. If we look to the ablation strategies beyond the pulmonary veins, it looks pretty much advice to do electrical isolation of the veins and entrance block and if we move then to ablation strategies beyond the pulmonary veins, it looks as limited as it did in 2017. So if linear lesions are to be performed, then of course achieve block. That's advice to do which is the traditional 1A. If there's a trigger that's active in the time of the procedure, then of course you map and ablate that. If there's a perimitral flutter and you're doing a lateral line, then it may be appropriate to inject the vein of Marshall but areas of uncertainty, voltage and substrate mapping, vein of Marshall infusion for persistent atrial fibrillation rather than flutter, attempting to induce non-PV triggers, isolation of the posterior wall, all areas of uncertainty. So this is a slide that I put together some years ago now and you can see I wrote to date no ablation strategy for persistent AF proven to be better than PVI alone and we're going to hear some interesting work that's coming out, interesting and exciting work I should say but I would just remind you that we've met the false messiah on this long and winding road more than once and so this meta-analysis, looking at all the different approaches, single procedure, multiple procedure, nothing adds much to PV antral isolation alone. If you recall the STAR-AF2 study, no benefit of CAFE or Lyons, possibly worse, CHASE-AF, that version of the stepwise, no benefit, of course there are improvements and updates. This is another meta-analysis and it's a really interesting one because it's a sort of time sequential analysis and it looks at all of the studies over the past 25 years, each dot on this graph is an arm of a study and there's PVI only in blue and PVI plus in red and over time we're doing better, we're definitely doing better, 1.8% improved outcome per year, per publication year but additional lesion sets, it's a non-significant trend to worse outcomes and so we look at posterior wall linear drivers, stepwise, fibrosis guided, no benefit, CAFE, worse outcomes and then we've got a couple of small trials that maybe do better. Over time, yep, procedure times, fluoroscopy times coming down as technology improves and this is even prior to the advent of pulse field ablation but again if we look at each individual trial from this meta-analysis then on the red side of the line we have CAFE and linear and stepwise and we'll hear more about that and I think it's getting a little bit more optimistic and then on the green side of the line but not reaching significance of the different approaches we've taken to drivers and posterior wall isolation and fibrosis guided and then we have two very small trials with large confidence intervals that I think we can't really say too much about at this stage. Now I'd like to talk to you about outcomes and how we measure outcomes because we always say you know PVI alone it's terrible for persistent atrial fibrillation ablation. This is a patient who had one of those old Medtronic devices in the AT500 and I did this ablation simple pulmonary vein isolation 2005 and what you see is the patient was pretty much in persistent atrial fibrillation, occasional short periods of sinus rhythm, pulmonary vein isolation. It almost all goes away but at around six months you see that two minute episode of AF. This is a failure on every 30 second outcome in every trial this patient's in the bottom arm. This is in the failed arm this patient. But I'll show you the follow up now 2010, 2016, 2023 keeps coming back for his pacemaker revisions and reviews and he's been completely free of atrial fibrillation other than that two minute episode and I can tell you I saw the electrograms that was definitely atrial fibrillation. So that 30 second outcome we've really hampered ourselves in terms of we've undersold the impact of PVI alone and if we look to the Jason's data, Jason Andrade's data in paroxysmal atrial fibrillation circuit dose, 52, 53 per cent efficacy, disappointing but burden, median burden, zero in all arms of this trial. Now let's go to the Kapler data and this is Pete's trial and really just focus on pulmonary vein isolation because that's what I'm talking about. Pete will talk about the posterior wall. 52 per cent efficacy at one year, disappointing, right? Median burden, zero. And we'll talk about that median burden a little bit more and it's a pretty typical population, age 65, three quarters male, BMI 29, longest continuous AF five months. These aren't the advanced patients, that's for sure. We're not talking about patients with advanced structural disease, Chad's VASC of two. Let's look at the nature of the recurrence and again let's focus on PVI only. It's around two thirds of these persistent AF patients have PVI only come back with paroxysmal recurrence, not with persistent recurrence and that matters because if you look at the healthcare utilisation in the green bars, that's the persistent recurrence, high levels of healthcare utilisation, the paroxysmal recurrence much lower. And if we then, not down to the levels of course of those patients who are AF free, and if we then look to the impact on quality of life, the AF EQT, just look to the bottom right. For those patients free of atrial fibrillation, 32 point improvement in AF EQT, remembering that five is the minimal clinically relevant improvement in the AF EQT. So this is dramatic. The paroxysmal recurrence is 23, the persistent recurrence is 12, much less improvement. And if we look to the three year outcomes, it's the same impact. You start to see efficacy for the 32nd outcomes in around 40% range, but the median burden remains zero. PVI alone at three years median burden, zero. And if you look at the impact on quality of life, and this is a paper in review and Peter's called this the sustained improvement or the increasing improvement, that's no sham because of the people telling us that this is AF ablation, the year of telling us it's placebo effect. Well, look at this impact out to three years. It keeps improving as patients get off their drugs and return to full activity and get their confidence back and all of that impact out to three years still improving. If we look at the burden for a cut point for improvement in quality of life, it's at around 5%. Below 5, greater than 5% burden, no improvement in quality of life. Less than 5% burden, you start to see improvement. But note, top right, less than 0.1% burden is where you see the most dramatic improvements in quality of life. Now we're going to hear about some of the later trials and I'm going to steal a bit of Pierre's thunder here and talk about PROMPT-AF. But just briefly, this is a randomised trial, an excellent large randomised trial, PVI alone versus PVI plus vanomarshal injection and a series of linear ablations. Single procedure, atrial arrhythmia recurrence free, clearly better in the plus arm, the ethanol infusion plus linear ablation arm, 0.045. But if you look at the multiple procedure freedom, that benefit goes away and that's just 8% redo in the PVI only group. So if you just do redo touch-up PVI in 8% of your population, you don't have to do any of that additional work. And if we look to the burden outcome, and that includes redos and includes antiarrhythmics, it's 0 in both arms, 0% in both arms. And if you look at the AFEQT down the bottom right, p-value 0.94, 26 and 27 point improvement in both arms, absolutely no difference in those kind of key endpoints. So we really do need to be looking to the more important endpoints, I think. So let me just come back to this slide. And of course, what you're all thinking about is, well, what about these patients? And these are the patients where we do need to think about what else we can do and what else we have to do. Those who recur persistent, 27% burden. Those who recur paroxysmal, 0.9% burden. But that's still potentially going to be a significant issue. How do we predict the patients? Well, in this trial, the ejection fraction, lower ejection fraction, larger LA volumes, of course. But I want to talk a little bit, one slide on substrate. And on the left here, I really just want to highlight this study from Pierre, because it's 20 years old almost now. And it's the first description of small circuit reentry. It's the first description of anterior substrate. And it's the first description of recording an entire circuit on a single catheter, and really an important study for us. And since then, many have reported the same thing. I put up this study from Chris Lu and Bruce Lerman, 35 patients with that extensive anterior wall substrate. Localized reentry, it's also the substrate for perimitral reentry. If you target those continuous fractionated signals, you can do it with a single application. But when it looks like this, I think you need to take the approach that Hans Kottkamp suggested, and that's in addition to the anterior line box isolation of that substrate. And I put this up because no one thinks that this patient is going to do well with PVI alone. I would never suggest that. So in conclusion, we have to think about that group of patients, those who are above that zero line. It might be 30%. It might be 40% of our persistent AF patients. And how do we pick them? Advanced remodeling, long-lasting AF, cardioversion dependent, the nature of the structural heart disease. And how do we design the next trials to target those patients? How do we find them without losing the impact by including all those who are going to do well with PVI only? And perhaps it's redo procedures, ongoing AF with isolated veins, and perhaps it's just focusing on those high-risk patients. And I'm going to stop there and thank you. Thanks, John. We're going to move straight on to the next talk and have the questions and discussion at the end. So it's a pleasure to introduce Pierre Gé, who's going to talk about the Marshall Plan. Thank you very much, Richard, Karim. Pleasure to be here. So maybe the first thing I have to do is to acknowledge my younger colleagues, Nicolas Derval and Thomas Pombrin. You see their picture here because they are conducting most of the hard work when it comes to vein of Marshall alkalization. And they choose this name, Marshall Plan, which I like particularly nowadays. It reminds us that just after the Second World War, the U.S. were with us and have been extremely generous with Europe. And as a result, this strategy, this plan was extremely fruitful for everyone. Things to remember probably nowadays. So there has been quite some background showing the importance of the vein of Marshall. It may act as a trigger and it may bridge the endocardial block that you can create when delivering a mitral isthmus line and render it incomplete. And most efforts, at least with RF, would remain unsuccessful without taking care of the vein of Marshall. And the concept that we are advocating is that there are localized reentries, should I say, that can anchor around the veins and micro-reentries that can loop around the veins or around the mitral annulus and also some other circuits that take advantage of the CS musculature and vein of Marshall. And I'm absolutely convinced that all of this exists during AF. We just don't see it because it's impossible to map, at least at present time, we can't really map atrial fibrillation. But I think these circuits are in fact existing and contributing to AF, and this is why we have an impact on the fibrillation itself by targeting those areas. Now it is also important not to destroy too much. And this is, in my opinion, something we have to keep in mind, particularly now that we have mass destruction PFA weapons. We can wipe out the anterior left atrium, and that's not a good idea. So the best ablation strategy, in my opinion, is the one that restores sinus rhythm at the least tissue cost because those cells are useful and it's important to preserve the atrial physiology as much as we can, particularly in older patients where the LV gets stiffer and where the atrial kick is becoming critical to their life. So this is the randomized trial that Nicola and Toma end up with, 120 patients randomized to PVI only or to the Marshall plan that would include, in addition to the PVI, the roof line and the mitral isthmus ablation with the alkalization of the vein of Marshall. So this is showing you the set of lesions. Interestingly enough, we start the ablation by the alkalization of the vein of Marshall. It allows you to understand whether the vein is there or not, whether it's visible or not. And it gives you the opportunity to map after that, which is always very interesting. It's a viable amount of impact on the voltage map that you will witness, but it's interesting to understand, in fact, how much bigger the network of this vein is as compared to the trunk, which appears to be a very small structure. And then, of course, it's the isolation of the veins, the roof line, the mitral isthmus line, the caveotracheal isthmus line, and some epicardial ablation from within the coronary sinus that we display systematically in order to have more durable block in this region. These are the characteristics. So typically, you see here the maximum length of uninterrupted AF duration and longstanding. It's pretty homogeneous distribution, 18% in both arms, PVI or PVI+. And these are the numbers when it comes to the LA volume. You see that it's significantly increased over 180 milliliters. This is the feasibility of the vein of Marshall alkalization in the Marshall plan arm, 97%. That's a lot. They are really good. They have developed some expertise, and I must confess that it's not always easy. And the vein of Marshall is not always something you can find, and you may have issues with some dissections, et cetera. So there are limitations, no question. But the ablation set ended up to be complete with the Marshall plan in 88%. And with no surprise, it was 100% when the ablation was limited to PVI. You see that it takes slightly more, so 36 minutes, almost 37 minutes of RF delivery with the Marshall plan as compared to slightly less than 30 minutes with the PVI only. What I found interesting is that the mitralismus, which was the most difficult one before the era of vein of Marshall alkalization, became in fact much easier than the roof line. And the duration is pretty much the same for mitralismus. The RF duration is pretty much the same for mitralismus as what it is for the CTI. And the trial is very significantly positive. You can see here in intention to treat that it's a 20% difference in favor of the Marshall plan strategy. This is single procedure with or without anti-arachnic drugs. Now the drugs are something incredibly difficult to deal with because some patients are so anxious of interrupting them that even if they have no AF at all, they can't just stop it. So it's always a difficult thing to judge. And so one of the very interesting aspects of this approach is that you do not distort too much the atrial physiology. You see here the activation and sinus rhythm after the ablation, positioning this mitralismus line here, pre-posterior and lateral, is in fact where the activation would collide anyway. So you do not create too much destruction or too much impairment of the activation as compared to an anterior line. Now, I must admit that sometimes the anterior wall is completely remodeled with a lot of scarring, and then it could make sense to deliver here, but even more importantly, the atrial kick is very visible after ablation, and you see that it improves pretty significantly from just after the ablation, three months in one year. In some patients, it takes six months to one year to really recover, and there are some parameters of improvement of the LV function as well. Even in those without impaired ejection fraction. So what to do in absence of the vein of Marshall? Well, this is not so rare, and interestingly enough, the group of Sebastian Knecht demonstrated that in fact you can ablate the mitralismus and be pretty successful. You see here that they reached 97% acute block in those without vein of Marshall visible, and that is superior to an approach where you don't even look whether the vein of Marshall is present or not. So it's really telling you that this vein is a problem to complete the mitralismus, but when it's not there, then mitralismus becomes reasonably easy. And John alluded to this PROMPT-AEF trial where they had much more patients included. That's almost 500 patients. That were randomized to PVI and PVI+, and they've got positive results, as said before. Now, I think that all of this has to be reassessed with the new tools that we now have, and I can share here my experience, or the broader experience, that is the PHAPLS device is not really good at doing linear lesions. It has not been designed for that, and it is not very effective when it comes to mitralismus ablation. We had plenty of reconnections. The Sphere 9 seems better, and I have no doubt that everything that I'm discussing here are first generation of PFA devices. This will only improve with the other generations, but it looks like this is already a pretty good tool for this purpose. Finally, I want to mention this trial by Ilad Anter that was published last year where he did persistent AEF ablation randomly using either RF or PFA, and PFA was using the Sphere 9 technology, and you can see here that there was a bias. Way more mitralismus line in the PFA group as compared to RF. Well, no surprise, it's so much easier. So this trial is not perfect, far from that, but they missed the demonstration for superiority of PFA by two patients only. So we're close to that, and in this effort by Vivek Reddy, you can see that the durability of the mitralismus line was very good when using this technology at three months systematic remapping with 68% durability. This is probably going to be a game changer. So I'd like to conclude that this Marshall Plan is a comprehensive ablation strategy. It's anatomical, but PVA is anatomical as well, and so far no much difference. It's just the extent of the ablation that is different. It aims to provide a complete lesion set that is systematically performed in all persistent AF patients, and I'm not entirely comfortable with that, I can tell you. I think that in the future, we will have to identify those patients in whom PVI only will work, and that's a very important piece of information, and there are signs that are predicting this, and I'm sure that artificial intelligence will do much better than what I do. So again, superiority was demonstrated in this study, and the remaining question is probably the role of the vein of Marshall alkalization. Once we have more powerful tools to complete the mitralismus line without using the alkalization, what is left of the benefit of the alkalization outside the simple mitralismus block? Thank you so much. Thank you. Thank you, Pierre. Thank you for this very nice presentation, and our first speaker will be Dr. Kessler, Peter Kessler from the Alfred Hospital in Australia, to discuss about posterior wall isolation and substress-based ablation. Thank you very much. Good morning, everybody, and it's a great pleasure to talk at this session on posterior wall isolation and substrate ablation. Well, the posterior wall's certainly a very attractive target and one that electrophysiologists seem particularly enamoured by, and there are a bunch of reasons for that. It has a common embryologic origin to the pomary veins. It's a recognised source of non-pomary vein AF triggers. It houses the parasympathetic ganglia, may be a site of proximal rotational activity, gives us a belt and braces approach to the posterior pomary vein isolation, antral lines, and provides some debulking of the atrium. So there are a lot of reasons why isolating this structure should work. It also houses the septopomary bundle, which John has shown previously is an important structure to lead to wavefront detours and fibrillation. So with that in mind, there's been a bunch of strategies developed over more than 10 years, and the one that we used in Kaplow was this linear approach you can see there in the top-left corner, but there's also the single-ring approach. There's the use of the cryo-balloon. There's a kind of annihilation, sort of polka-dot approach you can see there down the bottom. And then, of course, what we've alluded to is the use of pulse-field ablation, but there's also the surgical hybrid approach. Sorry. So there've been a couple of randomised trials. Again, the Kaplow trial on the right there, which we've all become familiar with, but prior to that, Hui-Nam Park's group in Korea also did a randomised trial, a smaller trial, but similarly showed absolutely no sign of benefit with posterior wall isolation. In that particular study, they included an anterior mitral line in the majority, which might be considered pro-arrhythmic, but nonetheless, there was no signal to suggest any benefit with posterior wall isolation. John's shown the long-term outcomes of Kaplow, and you can see there, if anything, the group that were randomised to posterior wall isolation had more recurrences than those with PVI alone. He's also highlighted the 0% burden at a mean follow-up of 3.6 years. And also, if you look down in that bottom right corner, not only do you see the very low AF burden at that long-term follow-up, but you see that about nearly 90% of patients are in sinus rhythm. If we have a look at the redos over that long-term follow-up, the redo rate was nearly one-third, and what I'd like to highlight is that the use of contact force RF catheters still performed pretty well. Again, we're convinced that pulse field ablation gives us better durable pulmonary vein isolation, and as Pierre's just said, we're really at the beginning of our journey with pulse field, but really, the reconnection rates are equivalent between RF and PF at this stage of our journey. You'll also see there that the rates of posterior wall re-isolation, which has been highlighted, and I agree is a limitation of Kaplow, was very high at 75%. So based on Kaplow and Pobey, I would suggest that there's really no role for empiric posterior wall isolation in all comers with persistent AF, but there may be some exceptions that we should consider, and certainly the long-standing persistent group, that with low posterior wall voltage, severely enlarged left atrium, and then we're going to hear more about certain electrogram characteristics which might highlight the posterior wall as a reasonable target for ablation, and then there's also the group who come back with recurrent AF in the presence of enduring PVI, and there's certainly this perception, without a lot of evidence, that if we don't lock out the posterior wall at the time of PF, that all the patients are going to come back with roof-dependent flutter. So the only signal that we saw on Kaplow for perhaps a benefit with posterior wall isolation was that long-standing persistent group, and you can see the confidence intervals are fairly wide there because that was a relatively small population of about 18% in the Kaplow group who had long-standing persistent, but nonetheless, you can see the hazard ratio there was 0.54. The confidence intervals crossed unity, but I think it is a group that we can think about maybe really trying to drill down on as perhaps that which may benefit from that strategy. This is work that David Cheng did as part of the Kaplow study, and what David looked at here was specifically the population with low posterior wall voltages, and again, as has been shown in a number of studies, those with low posterior wall voltage do worse, but what we found in Kaplow was that laying on posterior wall isolation in that group made absolutely no difference, and what Dave showed very nicely was that if you have low posterior wall voltage, then nearly 92% of patients will have low voltage elsewhere, and I think that's a really key explanation for that finding. Louise Segan, also in our group, looked at the complexity of electrograms, which is something that John really pushed us to do and look at, and what Louise found was that when we recorded AF on the posterior wall over a minute, we found that patients who had very rapid activity on the back wall had a worse outcome, and in fact, in that group, if they underwent posterior wall isolation, they had improved success compared to the group with slower posterior wall activation. Now, this is just hypothesis generating. It was a retrospective look back, but nonetheless, I think interesting data. So Pierre and John have already alluded to the transformation in our field of the arrival of pulse field ablation and certainly the durability in lesion sets that pulse field looks like it's going to bring. We heard just a couple of days ago Vivek shared this data with us, and this is now in publication in CERC, showing that a strategy of PVI and posterior wall isolation, importantly with implantable loop recorders, in Kapler we used twice-daily ECGs but not continuous monitoring, gave us a success of around 73%, but then if you look on the right panel there, you can see a bunch of cut points depending on our definitions of freedom from AF. And I'd like to suggest that perhaps the difference is not all that far away from what we saw with Kapler using RF. We also reported a success of around 52% based on a cut point of 30 seconds with twice-daily ECGs. We had an AF duration anywhere from seven days to three years, not less than 12 months as with the ADVANTAGE-AF study. We included a whole lot of patients with heart failure, which again were largely excluded from the ADVANTAGE-AF study. So I think the results stand up pretty well. So what data do we have with respect to PF with posterior wall isolation? The manifest registry certainly didn't show a hint of improvement with including the posterior wall. It does look like it's more durable on that right side there from the Swiss group. And perhaps in the interest of time, I was going to cover a fair bit on substrate, but maybe we'll confine it to posterior wall. Well, there's definitely this trend, if many of you in the audience, given PFA is so fast now, why not just take out the posterior wall at the same time? Well, I'd like to suggest there's no improvement in outcome. Pierre's already given us a really important concept of achieving the best results whilst maintaining the integrity of the atrial myocytes. It no doubt increases procedure time. And we're still at this phase where we're used to long procedures and PF has given us such short procedures that we remember those dark days of long procedures. But why increase your procedure time if you don't need to? There's some data there from the Advantage AF Phase 2 study. So the number of PF applications for PVI only 46 with a duration of 22 minutes. Adding posterior wall isolation pretty much doubled it up. So another 36 lesions, another 16 minutes. More PF, more hemolysis, can be pro-arrhythmic and also probably requires ice and 3D mapping. And my last slide is just to share some data that we looked at, again, with John from Michael Lim, which showed that if you isolate the posterior wall, you will double the incidence of left atrial flutter. Thank you very much for your attention. APPLAUSE Thanks, Peter, and sorry for harassing you. It's a real pleasure to introduce Elaine Wan from Columbia, who's going to be talking about the AI-guided or tailored approach. Great. Good morning. Thank you so much for the invitation. My name is Elaine Wan. I'm an associate professor at Columbia University. I'm a physician scientist as well as a practising electrophysiologist. And I am so privileged to be able to give sort of the summary of this session about the new pathways forward, including AI-guided or the tailored approach, although I had nothing to do with the tailored AF trial. These are my disclosures. So I actually typed into chat CBT, a form of artificial intelligence, what are the ablation strategies that they would recommend for persistent atrial fibrillation? And luckily, it really just summarised all of the wonderful trials that my previous speakers and colleagues talked about. And definitely number one, tried and true, is pulmonary vein isolation and all these other trials that were published in reputable journals and have really good evidence behind that. And ever since I was a fellow, there has been so many different ablation strategies that have been proposed in new technologies. And of course, there's posterior wall isolation, linear lesions, although some trials which were mentioned, for example, like STAR-AF, there's been questions as whether more and more lines is actually helpful. Obviously, there's also in the past CAFE ablations, vein of marshal ablations. And there was a time period where we really wanted to investigate whether or not ablating areas of rotational re-entry, like rotor and driver ablation, was the key to success. And then there's ongoing trials as whether or not, how do we identify non-pulmonary vein trigger ablations? Now, I just want to take one step to just look at these trials is that we're focusing here on just electrograms during imaging, or electrograms while we are doing mapping. But we're in this new age of AI where we're not just looking at electrograms itself, but also this incoming era of new imaging modalities. So first, there was a recent trial that was published just this month on a tailored AF trial, which focused on a multi-centered randomized control trial double blind and using the Volta AI program. And they looked at targeting special temporal electrogram dispersion was detected by artificial intelligence. And the question was whether or not if we were able to look at areas of dispersion and ablate these areas, whether or not there would be increased freedom from atrial fibrillation. And it was a superiority trial which showed that after one year, using this strategy, AI and PVI of looking at these areas of spatial temporal dispersion, that freedom from atrial fibrillation after one year was superior using AI. But however, when you looked at overall any freedom from any atrial arrhythmia, there really was no difference. And there was no difference in safety endpoint differences. And as my previous speaker talked about, the procedure time ablation time was twice as long. So it brings the question is, what actually are we ablating? So this is an example of the same group's 2017 publication on what are we considering spatial temporal dispersion. When we're looking at re-entry, they're looking actually at the electrograms. And we look at these electrical waveforms, there's many ways to look at how long they are, the voltage, the frequency. And using their propriety algorithm, they decided that these were areas of spatial temporal dispersion after various areas of mapping. And this was actually a response by Pichini et al at Nature Medicine about what are the strengths and challenges of having this AI-guided therapy. Well, number one, the team is definitely to be congratulated to show a new form of targeting areas of AI and showing that there was freedom from atrial fibrillation using AI plus PVI over conventional methods. But one of the issues was also in that in one single procedure, there was really no difference in arrhythmia recurrence. And it's really also unclear, as the previous speakers talked about, in this era of pulse field ablation where we're ablating more areas, how does this really translate? And also, as I mentioned earlier, when we have longer procedural times and we still have to do re-dos, how will this help with recurring arrhythmia such as atrial tachycardia? So this is their data from the AI-guided ablation strategy, which was also talked about yesterday in two of the sessions, which looked at the tailored AF approach where they used artificial intelligence, showed improvement versus just conventional PVI using an anatomical ablation, and that in different patient populations, there is superiority. So it brings into our next area, like what do we wanna see from HRS 2026? And I think one of the great things that we've seen in this session so far is the number one, the advancement of imaging. Not only do we have 3D, but 4D ice, as I've been able to see in the science and translational basic science sessions, we're incorporating more of the investigation into the substrate abnormalities that some of the previous speakers talked about. Not only are we looking at fibrosis, but also fat deposition, and also the advent of digital twins, some of the pioneering work by my computational science colleagues, looking at using MRI, CT scan, or inflammatory markers to see where areas that might be incorporated into computationally guided or digital twins, prediction of where to actually ablate. So I really look forward to the next HRS and the next upfroing scientific sessions, how we go beyond just electrical dispersion, electrical electrograms, and using computational modeling. But as some of the speakers brought up, a lot of these abnormalities we're looking, substrate abnormalities such as scarring, low voltage, and changes in conduction velocity. What other areas can we target? Fibrosis and adipose tissue, and at the end, are we just using more powerful weapons to ablate more areas, and how can we be more precise? And how do we target just looking at imaging in young patients which really have little to no substrate abnormalities, and how can we combine pre-procedural imaging to real-time information that we get in active mapping? So what other factors for AI should be upcoming? So as I mentioned, I thought that this was a good review image, is definitely scar, areas of decreased conductional velocity, decreased voltage. And a lot of this mapping is done during persistent atrial fibrillation is when the patient is in atrial fibrillation. How about when we restore normal sinus rhythm, are there still areas that persist that is suggestive of trigger activity? And we may wanna look, as some of the previous speakers talked about, areas of flow turbulence, for example, when there's mitral valve regurgitation or high turbulence to different obstruction areas, for example, in hypertrophic cardiomyopathy. So these may be upcoming. So I wanted to progress a little bit further. So at Columbia University, one of the topics that we're working on on AI is not just for atrial fibrillation, but for atrial tachycardia. So I just bring this as to light because the tailored AF trial showed that after AI predictions for AF, there was still recurrent arrhythmias for atrial arrhythmias. So are there ways to predict areas for atrial tachycardia? And here in around 20-something patients, we looked at areas, minimal conduction speed, distance to scar, distance to block. And this was seminal work with my colleague, Dr. Deepak Sluja. And we looked at voltage in all these different areas. And we were able to look at what we taught, the machine learning convolution neural network, what were areas of possible re-entry to where they predicted. And the blue areas where they predicted, and whereas red areas were where we taught it to learn. And we can see that there was significant overlap and it was very teachable as for an algorithm to learn where was the areas of success for ablation. And the areas that are circled are around seven to eight millimeters, which shows very good precision and refinement. So in conclusion, just thought for the future is, it seems like we're getting more and more personalization of AF ablation strategies may improve outcomes. This age of AI may definitely help us to consider areas of electrical dispersion, as well as substrate abnormalities. And that further refinement of AI-guided ablation strategies may be helpful for atrial fibrillation, as well as other atrial arrhythmias. Thank you very much. Thanks, Elaine. That was a really lovely wrap-up. We've now got a nice period of time for discussion. Before we go on to some of the audience questions, I'd just like to ask in general, I mean, if we look around this audience here, we can see very clearly human beings are different. And why would we expect the mechanism of AF to be the same? And it strikes me that the struggle we've had in applying an empiric ablation strategy to these patients is that it's not going to suit everyone. And so that gets lost in the randomized trials. And maybe machines will help us to differentiate the complex data that allows us to customize the procedure a little bit more for that individual patient. Pierre, you're looking skeptical, so I want you to sort of argue against me. Yeah, absolutely. I think this is off. Okay, thanks. I only partly agree with what you just said, reason being that the only thing we all agree on is that PVI is the thing to do when you want to get rid of atrial fibrillation. And this is an anatomical kind of universal approach, and it's the only one that we know is working. I don't disagree, but it's really interesting that we don't really understand why PVI works. How much is autonomic modulation a factor? And if we're using PFA, will it have the same effects? I would agree with that. And how common is this in the entire field of medicine? There are plenty of things that we discover, and we don't yet know why they work, but they do work. The point I'd like to bring is that... I mean, your point is well taken. Everything that fails with PVI only remains unclear. So this is where tailoring to the need of the patient is meaningful and desirable. No question. Now, the thing I'd like to emphasize, and I think that both Peter and John very nicely pointed out, is that we are wrong in the first place. We were wrong in 2007 when we decided that 30 seconds would be the definition for failure. This is ridiculous. It has absolutely no medical significance. And because of that, everything that we assess is polluted. And I think that we have now enough background to say that AEF burden is the thing we should look at, because AEF burden is what is going to determine the risk for complications and side effects of atrial fibrillation to our patients. So this should be revisited, and the other aspect I think where we are completely wrong is that everything we did in the guidelines so far is pointing out to symptoms and favors ablation of paroxysmal AEF in young patients. And this is not the population who will have the greatest benefits of returning in durable sinus rhythm. So I think this is kind of impacting also the aspect you just mentioned. Can I just pick you up on what should be the end point for clinical trials like the Marshall Plan that you just suggested? I suppose we're left with this conundrum between... If you look at the CAFLA data, for example, the AEF burden is 0% with PBI only at nearly four years. How can you get better than zero? Yeah, well, you can, because zero is for at least 50% of the population, meaning that the other 50% is outside this number and is still having arrhythmia and is not happy with that and is still exposed to the complications. How do you think we should reflect that, though? Should we just look at AEF burden in recurrence? Yeah, AEF burden is going to be the first step towards a more desirable approach in thinking. It is not the only one, in my opinion, because AEF burden is exposing... It is correlated to the risk for a given patient, but it's most certainly not the only indicator. And we should come up with some, you know, multi-parametric score that will better capture this extra risk. I want to pull us away from outcomes, because everyone's come here to learn from the experts what we should do. And before I pass over to Karim, I want to explain why, you know, dispersion has been shown to have a significant impact on outcome. Why are we not just doing that in everyone? I mean, why are we not just looking at... Why isn't everyone just using that technology and ablating, using that to guide ablation? You mean the tailored AEF? Yeah. Well, I mean, I think it's definitely a great start for using artificial intelligence to help us out. I think the data shows, or the trial at least, gives us first food for thought, that it definitely reduces freedom from AEF, but then we have all these other atrial arrhythmias to target. And I think it was a first step, and that shows that perhaps even after when we do PBI, if there's an atrial flutter, we still target that. So maybe there needs to be a second arm to that after AEF if there's a prediction for all the other atrial tachycardies or flutters that we should target that also. But then I just want to bring to the great point that Piero talks about, is are we just ablating more and more and using these technologies? But if we can find exactly the areas of re-entry and focus our target on that, even with these new ablation catheters, can we improve outcomes? Well, I guess that was the point that you were making, that being targeted and preserving atrial function is really important. And, yeah, so I take that point very much. Sorry, I've been hogging the microphone. I completely agree. And now maybe you can take some questions from the online platform. And the first one is more technical, and it's for you, Piero, regarding the Marshall Plan strategy. Do you think that now, with the advent of PFA and transmural lesion optimization, that we still need Marshall bundle infusion with alcohol in your experience? That's a very good question. We're working on that, and I can't answer very scientifically, in fact, yet. I hope we will in the next few months. My feeling is that we will not get 100% durable lesions there. Even with Gen 2 of these powerful devices, I suspect that 100% doesn't exist in medicine in general. So we won't be there. It will improve. Now, what I think is that the vein of Marshall is way more than just a bridge over the endocardial ablation that we are delivering. So even if you cut the trunk, you still have the impact on the autonomic nervous system and the trigger that may arise from the network, et cetera. So my bet is that there will possibly still be some benefit due to the alkalization of the vein of Marshall. How big is this going to be? That's the entire question. Maybe it's going to be minor and to be delivered only in some patients and not in the majority. But again, what I think is that we are largely ignorant on whom we should consider doing more than just PVI. That's really the key question to me. And then comes the techniques, but secondary. Yeah. Thank you. We have another question that is also related to trigger and initiation of AF on the online platform. And maybe, John, you can answer about the role of SVC isolation in patients with persistent AF. Is there still a place for this strategy? Yeah, that's an interesting question because all these years down the track, there's still pretty limited randomized evidence that this is going to be beneficial as a first time, even as a redo strategy. I think we've all got patients where it's been dramatically effective. Isolated veins come back, you isolate the SVC, or you can demonstrate a trigger in the SVC and you isolate the SVC and it eliminates the problem. And I think it's become much easier for us in the era of pulse field ablation with the basket lower risk to the phrenic nerve, still a risk to the sinus node. So I think we still have a lot to learn there. And I think it still fits broadly into that category of non-pulmonary vein triggers. Do we do this blindly? How do we initiate triggers? How do we find triggers? And again, all these years down the track and all these different approaches and studies and methods to initiate non-PV triggers, and we're still a bit struggling in the dark there, I think. The one thing that I would say about it's easy to do the SVC now is just remember it's not clear that you can't damage the phrenic. And it is pretty clear that if you're in the wrong spot, in a patient who might have somewhat compromised sinus pacemaker complex function, that you might wind up with profound bradycardia. So I think it's a great question, and I think there are a lot of questions that arise from it. Thanks, John. And Peter, one of the questions was talking about the success rate being 70% to 80%. What are we missing? But actually I think we've addressed that to some degree, that we're measuring the outcomes. But I think one of the fascinating things in the data you presented was that the quality of life continues to get better in both groups. And is there other things that patients do when they know they've got AF that improves their quality of life? Is it substrate modification as well? Are they losing weight or getting fitter? What's the explanation for that? Yeah, I suppose, and John showed this data, we certainly didn't see that the additional substrate ablation had any impact on their quality of life. I think John mentioned some of the points already about getting off antiarrhythmic drugs, the confidence of not having recurrent symptoms. Again, some of John's work in the past, there's still a lot of focus on physical symptoms with AF and the recognition of the major contribution of psychological symptoms, the incidence of anxiety, even suicidal ideation, that the remedial studies showed beautifully that are alleviated by PVI versus medical therapy. So I'm not convinced that it's the additional substrate ablation but rather the freedom from AF and the dramatic reductions in burden that contribute to quality of life. So they're not... I mean, certainly my experience is that patients tend to take care of themselves a little bit more once they realise they've got a heart problem, and I wonder whether that's a factor. They try and lose weight, try and get fitter, and it's actually... Sometimes the AF can save their life by making them take care of themselves. Maybe we can take some questions from the audience, if you have some hot questions. So, online there is another question about the martial plan strategy. Pierre, for you, what is the role of the roofline in the martial legend set, given the fact that the posterior wall isolation in the CAPLA study didn't show any benefit? Is there a link for you between these? Yeah, that's a very good question. Now, every randomised comparison we have, that is PVI versus PVI plus linear ablation, is using RF catheters. And in those catheters... I mean, RF, should I say, is in fact producing... There are two issues. The first one is that it's difficult. It's difficult, it takes time, and not everyone is very committed to get the block. And we all have, in our mind, the risk for fistulas to happen in this specific region. So, it prevents also us from delivering enough energy to make durable block there. PF is completely different. So, again, we have to reassess the benefit of these linear lesions, and particularly the one in the posterior wall, now that we have PF, where we can, like, overtreat somehow to ensure durability without the risk for fistula formation. This is, I think, absolutely key. Now, there are several aspects here. Number one is that, because of this risk of fistula, RF is used in a very superior position, real roof, where the distance with the oesophagus is higher, so it's safer there. But it's also way more difficult, because you have the bulk of the septopermary bundle that bridges epicardially, and it's extremely difficult to get complete block, or durable block, should I say, in this location. This is why we also sometimes have to apply a much lower linear lesion, where we have no septopermary bundle issue, but the distance to the oesophagus is much less, so the risk is much higher. So, again, I don't think we can say that PF for linear lesion in the low posterior wall has been assessed. We have some data from the registry, from manifest, where there is no benefit of the posterior wall debulking, because this is debulking. It's not a linear lesion. But this is a registry. It's not a randomized comparison. So, really, I don't think we are in a position to say much about the efficacy of this linear lesion using PFA. It will have to be reassessed. Great. Thank you. So, we've run out of time, but I'd like to summarize by suggesting that the speakers have told us that, you know, there's a brave new world coming, where Elaine and her colleagues will come up with an AI way of trawling the patient's demographics and notes and planning who's going to succeed and who's going to fail and what you should do. And then Pierre and his colleagues are going to give us a simple way of giving them an improvement. But I think the danger we have is, as the point that Peter made, as the procedures get faster and faster, we're going to spend less and less time thinking about that individual patient's electrophysiology. And probably in some patients, maybe identified by Elaine and AI, those that are likely to get recurrence, we should probably spend a little bit more time customizing the procedure to suit them and get them a better outcome. I'd really like to thank Karim and the speakers for a really excellent session and the audience for their great questions. Thank you very much.

ablation strategies

atrial fibrillation

pulmonary vein isolation

Vein of Marshall

posterior wall isolation

radiofrequency ablation

pulse field ablation

AI-guided strategies

electrophysiological profiles

personalized treatment